Professional Documents
Culture Documents
HANDBOOK
of
PUBLIC POLICY
Peter-3383-Prelims.qxd 6/8/2006 7:36 PM Page ii
Peter-3383-Prelims.qxd 6/8/2006 7:36 PM Page iii
HANDBOOK
of
PUBLIC POLICY
Edited by
B. GUY PETERS
and JON PIERRE
SAGE Publications
London ● Thousand Oaks ● New Delhi
Peter-3383-Prelims.qxd 6/8/2006 7:36 PM Page iv
Introduction © B. Guy Peters and Jon Pierre 2006 Chapter 15 © Kevin V. Mulcahy 2006
Chapter 1 © Peter L. Hupe and Michael J. Hill 2006 Chapter 16 © B. Guy Peters 2006
Chapter 2 © Peter DeLeon and Christine R. Chapter 17 © Richard D. Bingham 2006
Martell 2006 Chapter 18 © Wyn Grant 2006
Chapter 3 © Bryan D. Jones, Graeme Boushey and Chapter 19 © Kenneth Button 2006
Samuel Workman 2006 Chapter 20 © Walter Carlsnaes 2006
Chapter 4 © Davis B. Bobrow 2006 Chapter 21 © Tim Newburn 2006
Chapter 5 © Peter Bogason 2006 Chapter 22 © Ian Thynne 2006
Chapter 6 © B. Guy Peters 2006 Chapter 23 © Evert Vedung 2006
Chapter 7 © Irene S. Rubin 2006 Chapter 24 © Aiden R. Vining and David
Chapter 8 © Søren C. Winter 2006 L. Weimer 2006
Chapter 9 © John Uhr 2006 Chapter 25 © Gary Bryner 2006
Chapter 10 © Helen Fawcett 2006 Chapter 26 © Geert Bouckaert and John
Chapter 11 © Harold L. Wilensky 2006 Halligan 2006
Chapter 12 © Michael Moran 2006 Chapter 27 © Herbert Gottweis 2006
Chapter 13 © Susan Marton 2006 Chapter 28 © Jon Pierre 2006
Chapter 14 © Christoph Knill 2006
Contents
List of Contributors ix
Introduction 1
B. Guy Peters and Jon Pierre
7 Budgeting 139
Irene S. Rubin
8 Implementation 151
Søren C. Winter
CONTENTS vii
Index 493
Peter-3383-Prelims.qxd 6/8/2006 7:36 PM Page viii
Peter-3383-Prelims.qxd 6/8/2006 7:36 PM Page ix
List of Contributors
Graeme Boushey is a Graduate Research Fellow in the Center for American Politics
and Public Policy at the University of Washington, USA.
Kenneth Button is Professor of Public Policy at the School of Public Policy, George
Mason University, USA
Ian Thynne is in the Governance Program, Faculty of Law, Business and Arts,
Charles Darwin University, Australia.
John Uhr is Senior Fellow in the Political Science Program, Research School of
Social Sciences, at the Australian National University, Canberra.
LIST OF CONTRIBUTORS xi
Evert Verdung is Professor of Political Science at the Institute for Housing and
Urban Research (IBF) and Department of Government, Uppsala University,
Sweden.
Introduction
The Handbook of Public Policy is an attempt the academic aspects of this area of inquiry.
to cover the area of public policy studies. That Practitioners generally want answers to their
is an ambitious goal, and perhaps a too ambi- day to day problems, are little concerned with
tious one. Both the real world of public policy theory, and think that academics are hopelessly
and the academic study of policy are sizeable, enisled in their ivory towers. On the other
complex and differentiated bodies of practice hand, however, academics often argue that the
and of knowledge. If anything the body of experts in particular policy areas are obsessed
information that any treatment of public with those policies and cannot see the bigger
policy must attempt to cover has been increas- political and economic picture about the
ing. On the one hand, governments have impact and meaning of those policies. We have
altered the manner in which they consider attempted to balance these various perspectives
policies substantively, changing some estab- within the Handbook as a whole, but relatively
lished fields such as economic policy to few of the individual chapters have been able to
become “competitiveness policy”, and/or “sus- strike the balance between these two poles.
tainable development policy”, and also are
attempting to integrate better the range of
policies that have existed in the past to gener-
THE COMPLEXITY OF PUBLIC POLICY
ate more strategic approaches to governing.
The academic study of public policy has also
been expanding to include a wider range of The study of public policy is a very complex
academic disciplines and approaches. For topic, and any attempt to force policy into any
example, anthropology has become a more narrow theoretical frame should be considered
central player in understanding the processes with some skepticism. On the one hand there
by which policies are selected, as well as under- are some real virtues for policy as an area of
standing some substantive aspects of policy inquiry for the social sciences, given that it is
such as those directed toward immigrants. amenable to so many different perspectives.
Further, the range of theoretical approaches to On the other hand, however, this complexity
public policy has expanded as the limits of the requires bringing together a wide range of
more conventional ways of thinking about theoretical and analytical perspectives to gain
policy have become more apparent. For exam- any sort of understanding of what is happen-
ple, the chapter on constructivist approaches ing in any policy area. Both academic disci-
to policy (Gottweis, this volume) reflects the plines and substantive policy concerns tend to
emergence of a strong strand of theorizing that narrow the vision and to limit the ability of
has been developing to supplement, and to analysts to understand the underlying com-
contradict, to more rationalist perspectives on plexity of most policies.
policy.1 The typical manner of approaching public
The final major question that faces policy policy is to consider the various areas of gov-
studies is how to put together the practical and ernment activity one by one – health, defense,
Peter-3383-Introduction.qxd 6/8/2006 7:44 PM Page 2
taxation. Studies of that sort are certainly especially in Northern European countries in
valuable, but the first important aspect of the which the policy networks (Sorenson and
complexity of public policy is that focussing Torfing, 2003) and corporatism (Wiarda,
on the “single lonely policy” may vastly over- 1997) institutionalize the involvement of a
simplify the interactions of multiple policies in variety of social actors in making and imple-
producing outcomes for citizens. For example, menting public (sic?) policies. Even in the less-
if government wants to improve the quality of developed countries there is growing evidence
health for its population, the obvious area for of the use of social actors as a means of assist-
investment is in hospitals and other aspects of ing government to make and deliver public
the “health care industry”. On the other hand, policies, and that this involvement of non-
however, improving nutrition, or enhancing government actors has enhanced the legiti-
opportunities for exercise, may actually pro- macy of the State.
duce greater health benefits. Therefore, there is A final complexity concerns what do we
a need to think carefully about the interactions study when we attempt to study public policy?
of policies and means of coordinating policies Do we focus only on the decisions that are
to create more effective, if more complex, made and the processes that produce them, the
responses to policy problems. programs that governments develop to deliver
A second dimension of complexity in public services and to influence the society, or the
policy studies is the need to examine policy tools that governments use in their implemen-
questions from a range of theoretical perspec- tation? Ultimately, as Harold Lasswell (1935)
tives. For example, in his classic study of inter- pointed out some 70 years ago, that politics is
national politics and foreign policy in the about “who gets what”, and if that is true of
Cuban missile crisis, Graham Allison (1971) politics then it is certainly true for public poli-
discussed the virtues of “triangulation” and cies. Therefore, in the end we should be con-
the use of multiple theoretical perspectives to cerned with the impacts of policy choices on
understand decisions. His more comprehen- people, and the distribution of benefits in the
sive2 approach has rarely been used for other society. As numerous students of government
policy areas, but would make a major contri- performance have pointed out, however,
bution to understanding exactly what is hap- assessing those impacts is difficult. Some
pening in the policy area. The difficulty may be impacts, for example the benefits of education,
that few individuals are able to bring a wide- may not be fully apparent until many years
ranging theoretical grounding for the study of later, and some may be so diffuse (culture and
policy in this manner, so we all focus on what arts policy?) that measurement for other than
we know best and attempt to do it well. academic purposes may not be particularly
Scholars and practitioners may be able to build valuable.4
triangulated understandings of policies by
cumulating the work of individuals using their
own approaches, but this assumes that those
THE ELUSIVENESS OF PUBLIC
individuals will have examined exactly the
POLICY AS OBJECT AND THEORY
same policies and have made their findings
sufficiently transparent to do that sort of
comparison.3 To gain some understanding of public policy,
A third complexity involved in the study of which – as this volume will show – is both a
public policy arises from the word “public”. very heterogeneous concept and also a field of
There has been a tendency in the analysis of the social sciences that displays a number of
policy to consider primarily, or solely, the role different approaches, we must first remind
of the public sector and official actors in the ourselves of the significance of the substantive
process, and to ignore the role of private sector and analytical dimensions of public policy.
actors. In fairness, that tendency has been less Substantively, public policy has undergone
apparent in recent years than in the past, some rather fundamental changes over the
Peter-3383-Introduction.qxd 6/8/2006 7:44 PM Page 3
INTRODUCTION 3
past couple of decades in terms of policy about what causes changes in public policy.
design, selection of policy instruments and the The growing interest in policy diffusion and
role of the state in society more broadly. As we policy learning depends in part on its claim to
will discuss later in this Introduction, public be a rewarding avenue into these issues. Most
policy in general has become less intervention- observers would, however, probably argue that
ist, controlling and obtrusive. Instead, there is such patterns of policy transfer among states
today a stronger emphasis of the regulatory offer only a limited explanation to the broader
role of the state. The Zeitgeist of much of the question of policy change. If anything, policy
1980s and 1990s has been that of rolling back transfer probably occurs once a changed
the state, unleashing markets, cutting taxes course of policy has been decided and is not so
and public spending and reducing control. much part of the drive for change per se.
Globalization – real or imagined – has contin- Instead, it appears as if the main sources of
ued to fuel this model of public policy. policy change are either economic or related to
Analytically, a number of approaches have voters’ demands. The economic drivers of
emerged, highlighting different elements of policy change – either in terms of changes in
public policy and offering different types of the tax base and the public budget or in terms
explanations to policy design and efficiency. of global pressures on tax levels and public
This volume presents the leading approaches in spending – represent a powerful pressure on
policy research but it does so without detailing policy makers. Voters’ preferences matter too,
the chronological development of this research but probably less in terms of policy instrument
field. Policy researchers have, rightly, been con- selection and policy design than in the rela-
cerned with process throughout most of the tionship between taxes and benefits for the
post-war period. Indeed, even when institu- main constituencies in society. People are
tional analysis has been the predominant theo- probably not very concerned with how services
retical perspective on public policy there has and programs are delivered as long as they are
been an undercurrent of studies calling for provided.
“process-tracing” to supplement the institu- Most importantly, there is plenty of proof
tional perspective. Process remains extremely that politics matters in shaping policy choice.
important; actors’ assessment of policy chal- That means that governments of different ide-
lenges and various strategies to ameliorate ological orientations tend to make different
those problems today are largely a reflection of choices with respect to how the state should
decisions and assessments made yesterday. allocate its resources and how those resources
Policy evolves through process, hence it is diffi- should be mobilized. These observations hold
cult for studies that ignore process to arrive at a true even in an era where it is often said that
deeper understanding of policy. We will return policy challenges are becoming more similar
to these issues later in this chapter. among different countries, whether because of
The tool kit of policy analysts includes an demographic patterns, economic moderniza-
impressing number of theories, models and tion or global economic pressures (see, for
frameworks. Unlike many other subfields of instance, Castles, 1998). National political and
the social sciences, however, these approaches institutional contexts still offer much in terms
seem to be complementary much more than of explanation of policy choice. Those choices
they are contending, mainly because of many are made in the context of national systems of
different types of questions which scholars structures and values that have proven quite
raise in policy analysis. As a result, much of the resilient in the midst of globalization and
controversy that has evolved in political economic deregulation.
science over rational choice versus other The second question we need to address has
theories never made much of an imprint in the to do with the development of policy analysis
field of public policy. as a social science research field. Public policy
These trends and developments raise two research has been concerned with changes in
sets of questions. First, we need to know more public policy across time and space. This
Peter-3383-Introduction.qxd 6/8/2006 7:44 PM Page 4
applies to the entire range of variables in policy somewhat malfunctioning rifle. The target is
research, from theories of societal problems to changing course frequently – or does not
selection of policy instruments. Students of change course when we expect it to – and our
public policy, like all social scientists, have fads weapon is far from perfect in composition and
and fashions with respect to what theories and robustness. However, it is safe to say that the
models are “in” and which are not. diversity of approaches to public policy that we
It could well be argued that institutional have seen develop has helped us understand
approaches, ideational approaches or, indeed, public policy to a greater extent than would
most or all of the approaches covered in this have been the case had the discipline been
Handbook are to a large extent more a reflec- more homogenous.
tion of changes in the preferences among the
observers of public policy than a reflection of
changes in policy itself. That having been said,
THE DEVELOPMENT OF PUBLIC
however, these linkages do exist between the
POLICY PROPER
dimensions, i.e. between what is studied and
how it is studied. Thus, for instance, the grow-
ing interest in implementation followed fairly The debate among students of public policy
logically on the expansion of federal programs about how to best approach this topic has
in the mid- and late 1960s under the heading evolved alongside significant developments in
of the Strong Society. Similarly, there has been public policy itself. A quick look back in time
an increasing interest among scholars in regu- suggests that over the past several decades we
lation, or even in the emergence of a “regula- have seen profound changes in policy objec-
tory state”, partly as a result of states’ increasing tives and instruments. The 1960s and 1970s
use of regulation and less emphasis on using saw most Western governments expanding the
more costly policy instruments. public sector. The preferred policy instruments
Another example of the linkage between tended to be fairly coercive. New public pro-
substance and theory can be found in the study grams were launched to address societal prob-
of policy change. Adaptation to changing eco- lems. The growth in the economy generated
nomic preconditions is a surprisingly slow increasing incomes to the state which provided
process, and although politics matters in the the financial base for expanding programs in
aggregate, changes in individual programs can sectors like education, social welfare, housing
be a frustrating experience for incoming gov- policy and research and development. With
ernments because of the inertia in the public only slight exaggeration we could say that the
sector. This has drawn scholars’ attention to Western societies of the 1960s and 1970s were
the significance of different types of obstacles more state-centric compared to those of the
to policy changes. Such obstacles are typically 1980s and 1990s. This applies – with signifi-
believed to be institutions or processes which, cant variation – to jurisdictions on both sides
de facto if not de jure, lock in or protect areas of the Atlantic; in the case of the United States
of public spending from reassessment. Again, we need only compare the Strong Society pro-
it appears as if the increasing interest in insti- grams under Lyndon Johnson with the more
tutions has generated a broader and more market-embracing policy style of Reagan to
inclusive understanding of what shapes and note how much has changed in a rather short
sustains public policy. In some ways, it could period of time.
be that social scientists are just beginning to Although many policy objectives have not
discover what practitioners in the policy been abandoned, contemporary public policy
process have known for a long time; politics tends to draw on other instruments and insti-
matters but institutions mitigate the impact of tutional models compared to twenty years ago.
changes in the political leadership. Indeed, one could argue that in some cases the
Thus, the study of public policy could be social theory underlying public policy has been
described as shooting at a mobile target with a reassessed. In the early post-war period, the
Peter-3383-Introduction.qxd 6/8/2006 7:44 PM Page 5
INTRODUCTION 5
predominant social theory of public policy beyond that role should be carefully assessed
emphasized the positive contribution of policy and, if possible, transferred to the market or to
and state action to resolve societal problems. In civil society. In the United States, the Reagan
the contemporary policy vernacular, there is a administration sought to rid the federal level
much greater awareness of the contingencies of all responsibilities and functions which
and complexities surrounding policy design could not be defined as part of the core role of
and implementation. government. Similarly, in Britain, Mrs. Thatcher
Looking more broadly at these issues, there embarked on a project aiming at drastically
seem to be three types of changes that summa- reducing the scope of government in society
rize the developments over the past couple of and to “unleash” the market. While these
decades. The first of these changes is a shift types of reform helped the state to curb budget
away from a policy style characterized by com- deficits and to stimulate the economy, they also
mand and control by the state towards a regu- caused considerable political friction as con-
latory state (Moran, 2003). The budgetary stituencies deprived of financial support took
cutbacks in welfare programs that represent a to political arms.
rather uniform pattern across the western The third and final change in policy style
hemisphere have been accompanied by a grow- which we will discuss here is in many ways
ing emphasis on the regulatory role of the derived from the former two; the so-called
state. This development has also been driven shift from government to governance. Some-
by the tendency to open up for market-based times labelled “the new governance”, this devel-
systems of resource allocations and the rela- opment refers to a tendency among most west
tionship between the providers and consumers European governments during the late 1980s
of public services. Thus, if the previously pre- and 1990s to produce and deliver public
dominant policy emphasized political solu- service in concert with the market and civil
tions to societal problems, the current strategy society. The role of the state in the governance
is more based on a notion of rolling back the model is not to produce all those services itself
state, allowing the market to play a greater role but rather to coordinate public and private
in society. action so as to ensure that those services and
Another overarching development in policy programs are delivered. Collective goals and
style could be described as “back to basics”. objectives remain defined by political institu-
During the first couple of decades following tions but the pursuit of those goals needs
World War II, most western states – as well as not be a matter solely for the state. Thus, the
subnational institutions – assumed roles and governance perspective highlights concerted
responsibilities that could not be said to action, shared resources and negotiation as an
belong to the core roles of the state in society. alternative to the state-centric model which
Thus, many states owned enterprises providing features a clear separation of state and society
collective good such as electricity, railway and a state strategy primarily based on com-
transportation, telecommunication services, mand and control.
and the like. Also, states provided financial The increasing interest in “new governance”
support to an array of functions and activities is part of a cluster of issues which we and many
in civil society and the culture sector. The others have discussed elsewhere (Pierre and
1980s and 1990s saw somewhat of a reversal of Peters, 2000, 2004; Pierre, 2000; see also Kjaer,
this expansion of the scope of the state. The 2004; Rhodes, 1997). The significance of the
previously mentioned stronger emphasis on governance model in the current analysis is
the regulatory role of the state is one element that it represents a very different foundation of
of this “purification” of the state (see Premfors, public policy compared to the traditional gov-
1999). The state should no longer own enter- ernment model. In a governance perspective,
prises, not even if the services produced were public policy is to a large extent the outcome of
collective goods. Instead, the state should bargaining among political institutions and
return to its core role in society; any function societal actors. Students of neo-corporatism
Peter-3383-Introduction.qxd 6/8/2006 7:44 PM Page 6
find little news in this observation and, to be designed and implemented. By studying the
sure, what has been labelled “new governance” different stages of the policy process – agenda
is not very new in many European national setting, decision making, legitimation, imple-
contexts. What does appear to be novel about mentation, and so on – observers develop an
the model of governance which evolved during understanding of why policy looks the way it
the 1990s is that it involves the market to a does (see, for instance, Dye, 1987; Kingdon,
larger extent than the neo-corporatist model, 1995). Each stage has its own set of actors; for
which mainly saw peak organizations engage instance, decision makers are rarely involved in
in deliberation with the state. If anything, the the implementation of policy. Also, each indi-
“new governance” is more contextual and vidual stage displays its own types of potential
messier than most previous models of gover- conflict and political opportunity structures.
nance. That development poses a significant The overall rationale of the stages approach is
challenge to observers of policy making and that what actors do at one stage of the policy
public policy. While the state remains the process is to a large extent framed by what
undisputed center of representation, account- other actors have done earlier in the process.
ability and coordination, it faces a more com- Historical institutionalists rightly point out
plex environment with which it interacts. that the first stages in any process have a major
These three developments in the role of the influence on what happens next. We should,
state have redefined the range and scope of however, recognize that, although there are
public policy and, more broadly, have refor- powerful lock-in factors in the policy process,
mulated the social theory which public policy there is some room for political and strategic
rests on. That development, in turn, redefines consideration at all the stages of the process.
policy objectives and influences the selection For instance, as the implementation literature
of policy instruments. Most importantly, per- tells us, identical policies can be implemented
haps, is that all models of state-society in a number of different ways – with different
exchanges become institutionalized; citizens outcomes – in different institutional settings.
and organized interests learn what to expect There is a great deal of logic in the process
from the state in terms of problem-solving and approach. What happens at a particular stage is
resource mobilization. In societies with “weak very much shaped by decisions and actions at
states” (Migdal, 1998), expectations on the the previous stages. Policy makers probably
state are low and societal actors develop non- assess policy alternatives and political strategies
political solutions to various problems. in a process stage model and, indeed, the formal
Similarly, in countries with “strong states”, procedure of policy making defines in great
people become socialized into turning to the detail the process. These formal rules of policy
state for help with almost any problem. This making shape actors’ behavior and it therefore
institutionalization makes a transition from a also makes sense for policy analysts to structure
government-centered model of governance to their observations according to that perspective.
a market-based or network-based governance On the other hand, however, individuals also
model difficult and politically quite complex. play a role in the process, and act as policy
entrepreneurs to attempt to have their own
policy preferences enacted into law. Thus, policy
making represents a complex interaction of
THE STUDY OF PUBLIC POLICY
individuals, institutions, ideas and interests.
In addition to looking at policy, the process
A key purpose of this volume is to present the by which policies are formed and implemented,
multitude of approaches to public policy. In we can also ask a number of other questions
much of the public policy analysis of the past about policy. One of the more important com-
several decades, the focus has been on the ponents of the study of public policy is the use
process through which public policy is of analytic techniques that can be used to assist
Peter-3383-Introduction.qxd 6/8/2006 7:44 PM Page 7
INTRODUCTION 7
INTRODUCTION 9
Section One
Making Policy
Peter-3383-Chapter-01.qxd 6/8/2006 7:31 PM Page 12
Peter-3383-Chapter-01.qxd 6/8/2006 7:31 PM Page 13
1
The Three Action Levels of
Governance: Re-framing the Policy
Process Beyond the Stages Model
(p. xvi), ‘frameworks and methods or Meanwhile, on the other hand, we see
approaches’ (p. xvii), ‘analytical frameworks’ theoretical work which suggests that the policy
(p. 32, heading), ‘theoretical frameworks’ process departs so far from rationality that the
(p. 32, text). Under the heading ‘Models, maps most appropriate model is a ‘garbage can’
and metaphors’ Parsons explicitly addresses (Cohen, March and Olsen, 1972) or that at
the question of how to label the different ways least parts of the policy process, such as agenda
of organizing one’s ideas in the study of the setting, are ‘dynamic, fluid and loosely jointed’
policy process. As an overarching and most (Kingdon, 1995: 230). Post-modernist theory
general term he chooses ‘frameworks’, ‘within goes even further, not merely to unmask the
which and through which we can think and rationalist discourse, but indeed to suggest the
explain’ (p. 57). Parsons distinguishes here impossibility of rational processes at all (Fox
between explanatory, ‘ideal-type’ and norma- and Miller, 1995; see also Fischer, 2003). The
tive frameworks (pp. 57–8). argument we will develop is obviously closely
We begin with theories, models, mental maps, meta-
linked with the former of the alternative per-
phors; and to think analytically about public policy spectives just mentioned. At the same time we
we have to be sensitive to the existence of ‘reality’ as a consider there can be a reconciliation between
construction within a multiplicity of frameworks. The these apparent extremes.
activity of theorizing about public policy is, therefore, Our point of departure is the view that in
like drawing a map… (p. 58).
Western countries in many fields public poli-
Given the varied landscape as pictured there is cies are working. Such policies are only given
more than reason to have adequate maps. For, public attention when there is a crisis, or even
since the founding fathers of the study of a disaster. For academics there has been a ten-
public administration, the discourse has devel- dency to take their leads from publicity about
oped in diverging ways that in many respects the negative aspects of government perfor-
can be described as a ‘dialogue of the deaf ’. On mance. Since Selznick published TVA and the
the one hand, the rational assumptions about Grass Roots in 1949, many social science stud-
public decision making, as once postulated by ies, certainly ones concentrating on implemen-
authors like Wilson (1887) and Simon (1945), tation, have found disappointing results of
seem to remain persistent. While implying public policies in practise. Often this dis-
‘stagist’ relationships between sets of activities appointment about what happened to good
within the realm of politics and administra- intentions has been expressed in more or less
tion, these assumptions above all refer to a straightforwardly gloomy titles. Setting the
clear division of labour. The way in which tone were New Towns in Town: Why a Federal
practising politicians and administrators in Program Failed (Derthick, 1972) and Imple-
many countries, for example, have adopted mentation: How Great Expectations Are Dashed
contracting out and similar meta-policies, in Oakland (…) (Pressman and Wildavsky,
aiming at the separation of ‘implementation’ 1973; see Hill and Hupe, 2003, for a further
from ‘policy’, resembles those assumptions. At discussion of the issues about ‘implementation
the same time, reporting on ‘policy fiascos’, failure’). In this context Linder and Peters speak
journalists tend to confront what is happening of the ‘horrors of war’ approach to the study of
at the implementation level of a policy public policy (1987: 460). This approach, with
straightforwardly with what has been agreed its emphasis on ‘failures’, has turned many of
upon in the legislative stage. Moreover, many these studies into what Rothstein calls ‘misery
academics are quite happy to do work which research’ (1998: 62–5).
would seem to share assumptions of this kind. Before we move to look at the stages model
We have in mind here the substantial evalua- more precisely, it is important to highlight a
tion literature, particularly the current concern problem about the study of the policy process,
to explore ways to make policy more evidence which has made the debate about that model
based. In the UK this trend is visible under the both complex and intense. As indicated, the
slogan ‘What works?’ (Davies et al., 2000). systematic study of the policy process has its
Peter-3383-Chapter-01.qxd 6/8/2006 7:31 PM Page 15
roots in the classical literature on public plants, aircraft carriers and submarines’ (2003:
administration. The founding elements in that 80). Quoting Perrow (1999) on ‘normal acci-
literature had two contentious characteristics. dents’ they say: ‘There will be failures and there
One was the notion that there could be a clear will be accidents, simple probability demon-
separation between politics and administra- strates that it is so … But every day we enjoy
tion, making the latter the careful formulation the modern miracle of high-reliability systems’
of specific activities based upon the policy (Ibid.: 81). We could similarly refer to social
goals set by democratically accountable politi- policy systems. All over the world complex sys-
cians (Wilson, 1887). The other was that the tems have been designed to deliver social ben-
policy process could, or should, be a rational efits, in accordance with explicit rules, which
one, in the sense in which that term was used function uncontroversially much of the time.
by Simon (1945). This would mean that a care- Debate focuses upon either their occasional
ful and thorough examination of the relation- failures or upon those (often marginal) parts
ship between ends and means is carried out by of the systems where rules are hard to operate.
those required to translate policies into action. If we try to take a balanced view between the
Both of these notions, of course, came under positive and the negative image of policy
attack. It was argued that Wilson’s ideal of the processes we need a way of theorising about
separation of administration from politics was both its more systematic aspects and about the
unattainable, that administration is inevitably phenomena that may undermine that. In our
a political process (Waldo, 1946) and that the stance we join the many students of the policy
translation of policies into action is a very process who want to try to sort out approaches
much more haphazard process than Simon to specifying the issues at stake and the scope
suggested (Lindblom, 1959). for their systematic study. One key element in
debates about how to do this has involved
sharp differences of view about the extent to
Systematic study requires which it is possible to separate the policy
a general map process into ‘stages’. The so called ‘stages
model’ has been offered as a map, around
We put aside here the arguments, particularly which a consensus about theory building may
used on the ‘rationalist side’, that their models be developed. A question then is whether this
represent ways policy process should occur. map building can be a shared process, as is
Rather our stance is based upon a view that implicit in the positivist model of social
both sides in the argument tend to distort real- science, or whether we have a series of compet-
ity. The case against the rational actor can be ing maps (or at worst each our own map)
easily made: indeed much of the work of its which cannot be reconciled with each other.
opponents deploys good evidence of haphaz- Our objective here is first to explore the
ard and irrational decision making processes nature, functions and limitations of what has
and of the manipulative deployment of dis- been called the stages model, and then to see
courses. The case that needs to be made more whether an alternative general framework for
clearly, however, is on the other side, that there the analysis of policy processes may be built
is a tendency for contemporary analyses of upon it. The central questions are:
policy processes to focus upon mistakes and
disasters and to disregard the large number of • What functions does the so called ‘stages
examples of stable, successful policy processes, model’ fulfil in the study of the policy
involving deliberate design work worthy of process?
Simon’s model. Frederickson and Smith speak • Which limitations to its use can be
of ‘the high-reliability systems; the best exam- identified?
ples include commercial air travel, the provi- • And how may these limitations be over-
sion of electricity, gas, and cable television come in an alternative framework for
services; and the operation of nuclear power analysis?
Peter-3383-Chapter-01.qxd 6/8/2006 7:31 PM Page 16
to policy initiation, an early part in the stages decisions and actions that occur around a
heuristic (though that hardly seems fair to the policy’ (Schlager, 1999: 239). But if it is a typol-
former). DeLeon gives more examples of per- ogy, what then are the parameters? In our view
ceived alternative approaches to the stages the key to understanding the nature of the
heuristic (for instance Fischer, 1995, on evalu- stages heuristic must be sought in its origins:
ation), which, in fact, relate well to it instead of Lasswell’s ‘policy orientation’ (1951a). Focusing
replacing it. If this is so, may it be then perhaps on what Dewey (1927) once described as ‘the
that the stages heuristic is something of a dif- public and its problems’, public policy is an
ferent order? object of scholarly inquiry with a scope going
It seems useful to note here how carelessly beyond scientific theory. This implies that the
notions of theory are used in policy studies. stages heuristic can be seen as part of a wide-
Elinor Ostrom (1999) argues that it is impor- spread, in fact pre-scientific, general conception
tant to distinguish within theoretical work of public policy. The term ‘public policy’ seems
three levels of specificity, which are often con- to imply a set of underlying assumptions of
fused: frameworks, theories and models. which, in an ideal-typical form, the following
Hence she identifies these as follows: construction can be made.
Within the trias politica the executive power
• A framework helps to identify the elements
executes what the legislative power has formu-
needed for more systematic analysis, provid-
lated and decided upon. Actors other than the
ing a list of variables and ‘metaphorical lan-
designer/decision maker execute public poli-
guage that can be used to compare theories’
cies according to the intentions legitimately
(Ibid.: 40).
laid down in laws or other official documents.
• Theories ‘enable the analyst to specify
Preconditions for effective policy are adequate
which elements of the framework are
knowledge and information, a solid basis of
particularly relevant to certain kinds of
power and cooperation, and an obedient
questions and to make general working
field of application (Van Gunsteren, 1980).
assumptions about these elements. Thus,
Through the implementation process the
theories focus on a framework and make
intentions of a public policy are, literally,
specific assumptions that are necessary for
realised. Overall, the realisation of these inten-
an analyst to diagnose a phenomenon,
tions takes the form of what is called a policy
explain its processes, and predict outcomes.
process: a series of orderly staged successive
Several theories are usually compatible
sequences, going from articulation via forma-
with any framework’ (Ibid.).
tion and implementation to evaluation (and
• Models ‘make precise assumptions about a
back). A policy is seen as ‘a hypothesis contain-
limited set of parameters and variables’
ing initial conditions and predicted conse-
(Ibid.).
quences. If X is done at time t1, then Y will
Therefore, perhaps we may argue that the result at time t2’ (Pressman and Wildavsky,
stages heuristic is a framework rather than a 1973: xiii). Despite the use of the term ‘policy
theory. This seems the obvious position to take cycle’, actually the more appropriate metaphor
about the value of the heuristic as a device to is that of a chain: both in time and often in
facilitate research, and, as John (1998) adds, space, vertically linking various ‘clearances’.
teaching. Public servants, especially those with an imple-
Schlager, however, offers a slightly different mentation task, fulfil their jobs within a hier-
approach to this issue. She sees the stages notion archical setting, with fixed competences, led by
as a ‘useful categorization of behaviour and documents and guided by rules. If what is
action within entire policy processes’ but not a achieved is not what was expected, shortcom-
framework because ‘general classes of variables, ings in implementation are to blame, particu-
or “universal elements”, and general relationships larly insufficient rule compliance.
among them’ are lacking. She thus describes it as For policy researchers a few consequences
‘a typology that completely describes policy follow (for an elaboration see Hill and Hupe,
Peter-3383-Chapter-01.qxd 6/8/2006 7:31 PM Page 19
2003): Policy determines action, while policy model which denies that policy making goes
intentions predict policy performance in a 1:1 on at later stages, nevertheless this is the impli-
relationship. cation of its casual or uncritical use. It may,
With the general conception of public policy, indeed, be ‘implementation’; but it could also
as constructed here, we obviously are back to the be ongoing ‘policy formation’. While it may
idea of the administrative process described in have been originally inherent to the stages
our introduction as the traditional model linked heuristic that there was a fixed, a priori locali-
with the work of Wilson, Simon and many sation of specific stages to certain administra-
others. We accept then the way in which tive layers, this empirically needs not be the
researchers have in various respects argued that case. Second, related to this problem is what
this ideal-typical ‘image’ does not represent the has been designated as ‘the fallacy of the wrong
real world of public policy. Yet people continue layer’ (Hill and Hupe, 2003). If at a ‘lower’ layer
to hold this conception of public policy, founded the stage of ‘policy formation’ de facto can be
as it is in institutionalised normative views observed, this must not be seen as implying a
on democracy and the rule of law, as highly judgement about the legitimacy of that partic-
attractive. The conception seems so widespread, ipation in the policy process. The latter is a
not only because it is rooted in the political- normative question that should be distin-
administrative culture and institutions of the guished from the empirical observation. The
Western Rechtsstaat, but also because it appeals to stages heuristic seems to imply for each stage a
both a general quest for control (Van Gunsteren, specific, presupposed set of actors, a third
1976) and a psychological desire for a rational methodological limitation. Where empirical
order of things2. In this context there is no need and normative matters thus are confused, one
to be surprised about the persistency of the stages may fall into a pitfall which could be called the
heuristic. If that is the perspective, is the heuristic ‘control trap’ (Hill and Hupe, 2003).
then suited for the study of the policy process as Our position is that the stages heuristic is
it is, after all; or should it be adapted? rooted in the logically defensible assumption
DeLeon, though an advocate, acknowledges that decisions are followed by actions, which
that the negative aspect of the invitation from require a cumulative process if anything is to
the stages heuristic to researchers to look at ‘just occur at all. But such a view does not preclude
one stage at a time’ is the neglect of the policy the idea that there will be subsequent decisions
process as a whole (1999: 23). Furthermore, the that may undermine the original one. Then
heuristic may lead to a view of the policy does this imply a need to abandon the stages
process as ‘disjointed’ and ‘episodic’, taking place notion or to use it with care?
in the relatively short term (a key aspect of Despite his criticism of the stages heuristic
Sabatier’s critique from the point of view of his Sabatier observes ‘Given the staggering com-
‘advocacy coalition framework’). Besides, the plexity of the policy process, the analyst must
picture of stages inappropriately implies a cer- find some way of simplifying the situation in
tain linearity to many, instead of ‘feedback order to have any chance of understanding it’
actions and recursive loops’ (Ibid.). Also, when (1999: 4). Parsons sees the value of the stages
one takes the stages heuristic as a general map heuristic to cope with ‘multi-framed activity’
for the analysis of policy processes, as indicated (1995: 80), but at the same time underlines the
above, it potentially gives rise to some misun- necessity to look beyond it towards
derstandings that may have consequences for
the mapping of the wider contexts of problems, social
research findings. process, values and institutions within which policy-
First, the picture of stages seems to reify the making and policy analysis (take) place (p. 81).
scope of each separate stage. This conflicts with
a view taken for a long while, and particularly Hence he argues
since Lipsky’s seminal work (1980), that policy Given the existence of a complex policy reality framed
making goes on even at the street-level. While, by a range of theories, models, explanations, values and
in fact, there is nothing implicit to the stages ideologies, the problem is not with the policy cycle
Peter-3383-Chapter-01.qxd 6/8/2006 7:31 PM Page 20
[stages may be substituted in this context – the authors] doing so, it is important to recognise that
per se, as with the need to incorporate or include in some respects these still involve a ‘stages’
models and approaches which are or may be deployed in
policy analysis (Parsons, 1995: 81).
perspective, albeit a modified one.
Lynn (1981: 146–9; 1987) uses the concept
In other words, given the need for a relatively of ‘nested games’ to assign the various parts of
‘empty’ framework general enough to com- the policy process. There is the ‘high game’, in
prise multiple theoretical approaches, Parsons which it is decided whether or not a policy will
sees a remaining function for the, perhaps be made. Then, in the ‘middle game’, the direc-
amended, stages heuristic. tion of the policy is determined. The ‘low
On the basis of this overview we suggest the game’ is about the practical side of the policy
following criteria for a ‘framework’ as a ‘general making; implementation is central here.
map’. First, the latter must have the capacity to Parsons (1995: xvii) distinguishes ‘three
encompass conceptually the ‘multiple multiplic- broad and overlapping levels or dimensions of
ity’ character of public policy processes. This analysis’. Each of those ‘may be seen through a
involves both a variety of actors and very often a variety of different frameworks and approaches’
variety of administrative layers, and horizontally (p. 82). The way in which issues and problems
linked organisations. The scholarly analysis of are defined and policy agendas are set Parsons
these uses a variety of disciplinary perspectives, calls ‘meso-analysis’. As a level of analysis cut-
‘lenses’, and focuses specifically on various parts ting through the various phases of the policy
of the policy process. Second, a ‘general map’ or process, it is ‘meso’ because it:
framework must enable specified (‘localised’)
explores approaches which link the input side of the
theory formation around selected sets of vari- policy-making process with the policy/decision-making
ables, rather than compel the building of one and output process focusing on the relationship
grand theory. Third, while the first criterion between the ‘pre-decisional’ dimensions of policy-
requires the framework to be as comprehensive making and its decisional and post-decisional contexts
(Parsons, 1995: 82).
as possible, it at the same time needs to be open
in the empirical sense. Rather than implying a The analysis of ‘how decisions are taken and
‘managerialist’ or ‘top-down’ view of the policy policies are made and how analysis is used
process (Parsons, 1995: 81), a general framework within the decision-making process’ Parsons
for analysis must facilitate systematic and nor- calls ‘decision analysis’ (p. 82). Then ‘delivery
matively open empirical research. analysis’ refers to ‘how policies are adminis-
Assessing the stages heuristic according to tered, managed, implemented, evaluated and
these criteria we may conclude that it has diffi- terminated’ (p. 82).
culty in meeting the first and third criteria. Finally, there is the ‘institutional analysis and
DeLeon’s reference to the way in which it has development’ (IAD) framework developed by
facilitated some classic pieces of research indi- Elinor Ostrom and her colleagues, presented
cates that it meets the second one. Overall, we as such for the first time in 1982 by her and
may characterise the stages heuristic as a multi- Larry Kiser. Stemming from micro-institutional
dimensional framework for the analysis of policy analysis Kiser and Ostrom (1982: 184) specify
processes, of which not all relevant dimensions three related but distinct levels of analysis.
have been articulated. Can we build on it to pro-
The operational level, which explains the world of
duce a more helpful overarching framework? action. The collective choice level, which explains the
world of ‘authoritative decision-making’. The third is
the constitutional level, explaining ‘the design of collec-
Alternative analytical frameworks tive choice mechanisms’ (p. 184). It may be noted that
as ‘general maps’ these are listed in the original formulation in an order
that reverses the conventional order of the stages model.
Individuals at the operational level ‘either take direct
Though the stages heuristic seems a widely
action or adopt a strategy for future actions, depending
used general framework for the study of the on expected contingencies.’ They are often ‘authorized
policy process, some alternatives have been to take a wide variety of actions at this level without
developed. We will look at these here. But in prior agreement with other individuals’ (pp. 207–8).
Peter-3383-Chapter-01.qxd 6/8/2006 7:31 PM Page 21
Collective decisions are made ‘by officials (including Governance ‘is a descriptive label that is used to high-
citizens acting as officials) to determine, enforce, continue, light the changing nature of the policy process in recent
or alter actions authorized within institutional arrange- decades. In particular, it sensitizes us to the ever-increasing
ments’. These collective decisions are plans for future variety of terrains and actors involved in the making of
action. ‘Unlike individual strategies, collective decisions public policy. Thus, it demands that we consider all the
are enforceable against nonconforming individuals … actors and locations beyond the “core executive” involved
The authority to impose sanctions is a key attribute of the in the policy making process’ (2002: 3).
collective choice level of decision-making’ (p. 208).
On the third level ‘Constitutional decisions are collec- While for Milward and Provan
tive choices about rules governing future collective deci- ‘Governance (…) is concerned with creating the condi-
sions to authorize actions. Constitutional choices, in tions for ordered rules and collective action, often
other words, are decisions about decision rules’ (p. 208). including agents in the private and nonprofits sectors, as
In the framework institutional arrangements are ‘link- well as within the public sector. The essence of gover-
ing each level of decision making to the next level. nance is its focus on governing mechanisms – grants,
Constitutional decisions establish institutional arrange- contracts, agreements – that do not rest solely on the
ments and their enforcement for collective choice. authority and sanctions of government’ (1999: 3; for
Collective decisions, in turn, establish institutional arrange- other definitions see also Kooiman, ed., 1993; 1999;
ments and their enforcement for individual action’ (p. 209). 2003; Pierre and Peters, 2000; Heinrich and Lynn, eds,
2000; Lynn et al., 2001).
In this framework it is only at the operational
level where an action in the physical world A contribution towards developing a gover-
flows directly from a decision. nance perspective to the study of policy
In an updated discussion of the framework, processes, founding and grounding their con-
Ostrom (1999: 36–9) elaborates some ‘key dif- ceptualisation in relation to the state of the art
ficulties’ in the study of institutions. They of implementation theory and research, was
involve the multiple meanings of the term made by Hill and Hupe (2002). What follows in
institution, the invisibility of institutions, the this section is partly based on that but offers a
multiplicity of inputs coming from various further elaboration of the insights presented
disciplines and the corresponding need to there. The concept of governance is designed ‘to
develop a specific ‘language’, and the configu- incorporate a more complete understanding of
ration character of relationships. An additional the multiple levels of action and kinds of vari-
matter Ostrom mentions is the multiplicity of ables that can be expected to influence perfor-
levels of analysis. Particularly important here is mance’ (O’Toole, 2000: 276). That makes the
the analytical treatment of the nested structure concept well suited to incorporation into a gen-
of the framework (1999: 38–9). eral framework for the multi-dimensional
Further to these ways of specifying a frame- analysis of policy processes. Thus, such analysis
work for policy analysis we think there is room can be seen as ‘governance research’ (see
for an additional framework as another alterna- Heinrich and Lynn, eds, 2000; Lynn et al., 2001).
tive for the stages heuristic, driven by the aim of Hill and Hupe (2002: 15) sketch the conse-
all three but particularly inspired conceptually quences of that. First, a clear distinction is made
by Ostrom’s framework. In our view for that between the how and the what of scholarly
purpose some amending of that framework is attention. Focusing upon governing as action,
necessary. Specifically, there is a need to make an rather than as government as institution, leaves
explicit link with the concept of ‘governance’. empirically open who is the acting actor (may
be a public, may be a private one). Second, dif-
ferentiating between administrative layers is
4. THE POLICY PROCESS AS important. Third, the act of management is
MULTIPLE GOVERNANCE taken seriously; in principle it can be observed
in all the loci of political-societal relations.
Governance Therefore levels of analysis have to be specified.
Summarising, the stucture of a policy process
Various authors have given definitions of the can be seen as consisting of various elements:
concept of ‘governance’. Richards and Smith, actors, sets of activities, action situations, and
for instance, say: layers. Each concept will be elaborated here.
Peter-3383-Chapter-01.qxd 6/8/2006 7:31 PM Page 22
Table 1.2 Alternative general analytical frameworks for the study of the policy process
Institutional rational choice
Nested games Kiser and Ostrom Multiple stages Multiple governance
Lynn (1981) (1982; Ostrom, 1999) (Parsons, 1995) (Hill and Hupe, 2002)
High game Constitutional level Meso analysis Constitutive governance
The framework provides flexibility to choose argue that public policy processes need to be
the appropriate unit of analysis. On the side able to hold in tension multiple ‘accountabi-
of the independent variables, too, the Multiple lities’. In this respect, as in the cases of public
Governance Framework offers a way of struc- policy for health and education mentioned
turing these rather than prescribing what they above, it may be important to try to develop
should be. O’Toole’s (1986) scan of the imple- policies so that they follow complex pathways
mentation literature suggested a massive list of across the items in Table 1.1. The issues about
variables, attracting the criticism that here is a how, and by whom, the overall structure and
subject in need of parsimony (see Matland, funding should be determined may differ from
1995; also Meier, 1999). The identification of those about local arrangements for policy
the framework within which action may occur delivery, whilst each again may differ from the
brings this list down to a limited range; Hill concerns about the discretionary behaviour of
and Hupe (2002: 123) have suggested about practitioners. All are connected, but there are
seven categories of independent variables5. many options about how those connections
This narrower range makes it possible to select may be made: this goes for policy research as
a small number of variables thought relevant well as in policy practise.
to the research question at hand.
6. CONCLUSIONS
Aiding the identification of
action choices
Hardly any other insight from public adminis-
A great deal of the implementation literature tration or political science has been so gener-
has been preoccupied with a normative argu- ally adopted by practitioners as the so-called
ment between ‘top-down’ and ‘bottom-up’ stages model of the policy process. This seems
perspectives. Both stem from deeply held views to exemplify Lasswell’s stress on the extended
about democratic accountability. Faced with scope of what he called the ‘policy orientation’,
what they regard as defects in the implementa- going beyond the borders of science. The idea
tion process both perspectives concern them- of policy making as proceeding in successive
selves with efforts to increase the capacity to stages influenced Pressman and Wildavsky in
steer the policy processes, either from the formulating their ‘chain’ hypothesis, implying
top or the bottom. By contrast, an analytical the likelihood of ‘implementation gaps’
perspective that recognises that influencing (Bowen, 1982). By consequence it invited the
the policy process can involve adjustments to a kind of ‘misery research’ we mentioned above.
complex nested system of levels, loci and The stages notion is an important element in a
layers, may help actors to identify alternatives broader conception of public policy that is
for action. Those with a strong top-down per- both persistent and, in empirical research, to a
spective may be assisted by recognising that certain extent misleading. That it is persistent
they have choices between fundamental has to do, as we have shown, with three factors.
restructuring, the adjustment of specific First, the conception is normatively attractive
arrangements, or new ways of curbing street- because it has been founded on the principles
level discretion. Conversely, from a bottom-up of democracy and the trias politica. Second,
perspective there will be questions about it appeals to what Van Gunsteren (1976)
whether what is crucial is to devise new ways of describes as a general quest for control in public
making street-level decisions, or whether there affairs. Third, the orderly neatness of particu-
are institutional and/or structural modifica- larly the stages notion seems psychologically
tions that would need to be made before these attractive.
would be feasible. At the same time, especially inasmuch as this
Between those strong ideological positions conception of public policy points automati-
many will, following Day and Klein (1987), cally to the ‘black box’ of the implementation
Peter-3383-Chapter-01.qxd 6/8/2006 7:31 PM Page 27
of a specific policy as ‘explaining’ why the studies, empirical questions about its validity
results of that policy are judged as disappoint- are to be answered. Apart from enabling con-
ing or worse, it is misleading. In this respect textual theory formation we have indicated
this conception may keep away from open and that it may aid contextual reflection by practi-
systematic research into what has happened, tioners. Furthermore, the Multiple Governance
how and why. This ‘control trap’ seems to lay at Framework may also fulfil institutional func-
the basis of much of the failure discourse tions; one could think of a research program-
around public policy processes. It is particu- matic use or a function as a map for public
larly this restricted explanatory power that agendas. In fact, the framework is designed to
forms the heart of the criticisms frequently combine two essential tasks: one within social
expressed with regard to the stages ‘model’. science, and one in a broader societal context.
We accept the case for the stages heuristic The first task is providing what Ostrom (1999:
functioning as a general map for the analysis of 41) calls a ‘multi-tier conceptual map’. The
policy processes. As such it is a very general other task goes beyond that: to function as a
map – not a ‘model’ and certainly not a causal multiple heuristic in an institutional sense.
‘theory’ – that has been used to good effect in Then, after all, we are back again with the
some classic policy studies. At the same time, essence of Lasswell’s ‘policy orientation’; double
according to the standards of a moderate bound as it is.
positivism in social science, it is conceptually
neither multi-dimensional nor empirically
open enough to enable and enhance systematic NOTES
theoretical-empirical research in the study of
the policy process. Given the varied landscape 1. Following up Implementing Public Policy that Michael
of the study of the policy process, our stance Hill and Peter Hupe published in 2002, this chapter elabo-
is that such research is both necessary and rates on some of the insights developed in that book. The
possible. major work on it was done during a stay of the latter as a
visiting fellow, March and April 2003, at the Public
Therefore, building further upon Elinor
Management Institute, Katholieke Universiteit Leuven. He
Ostrom’s Institutional Analysis and Devel- thanks Geert Bouckaert, the director of the Institute, and
opment framework, we have linked that Rudolf Maes for their generosity. Wim Hafkamp, the Dean
framework into the mainstream of the politics- of the School of Social Sciences, Erasmus University
and-administration focused policy sciences con- Rotterdam, and Victor Bekkers and Jan Hakvoort, of the
Department of Public Administration, are thanked for
necting it to an explicit governance orientation.
their support of Peter Hupe’s two months stay in Leuven.
As such the Multiple Governance Framework On the 10th of June 2003 Peter Hupe presented a draft of
has a multidimensional as well as a nested char- this chapter in the Centre for Public Governance, research
acter. It provides organising concepts that can group of the Department of Public Administration,
assist low- or middle-range theory formation Erasmus University Rotterdam. He thanks the participants
in that session, particularly Menno Fenger, for their com-
and systematic empirical research with a
ments. On a later version comments were given by Elinor
‘localised’ character. Among other things this Ostrom and George Frederickson, which were highly
means that one level of action, such as the ‘stage’ appreciated by the authors.
of implementation, is not seen as a priori con- 2. It is against this background that we could formulate
fined to one administrative layer. the implementation follows formulation and decision theorem.
This says nothing about who and where, but is a matter of
The framework presented here does not
logic (Hill and Hupe, 2002: 4).
break altogether with the concerns of the many 3. Originally (Hill and Hupe, 2002) we used the term
scholars who have recognised the logical ‘nest- ‘constitutional governance’ here. Now we think the possi-
ing’ or ‘institutional pathways’ affecting many ble reference to ‘the constitution’ – though not intended –
decision processes. In that respect it can still be may make the connotation of this term too fixed and
formal. Essential is the general notion of the Latin verb
said to embody an idea of stages, though a
‘constituere’: forming, building, designing. Therefore we
loose one. When the framework is put forward propose here, instead, to speak of constitutive governance.
as a conceptual device to assist with the fram- 4. Acknowledging two things seems appropriate here.
ing of subsequent theoretical-empirical First, the labels used in this chapter purposely are made
Peter-3383-Chapter-01.qxd 6/8/2006 7:31 PM Page 28
more ‘empty’ or abstract than in the book (2002) version. Dewey, J. (1927) The Public and its Problems.
Aiming at maximal analytical applicability, in the designa- New York: Holt.
tion of the dimensions of the analytical framework any Dror, Y. (1989) Public Policymaking Reexamined,
connotation with real world epitheta (cf. ‘street-level’) has second edition (first edition: 1968). New
been avoided. Eliminating connotations with real layers
Brunswick, N.J.: Transaction Publishers.
within the framework makes the latter more applicable in
research, in principle, at any layer. Second, at the same time
Fischer, F. (1995) Evaluating Public Policy. Chicago:
the nested character is reconfirmed: the distinguished loci Nelson Hall.
and actors can be observed at each empirical layer in the Fischer, F. (2003) Reframing Public Policy, Oxford:
real world of public administration, formal or not. Oxford University Press.
5. On the basis of an updated extensive literature scan Fox, C.J. and H.T. Miller (1995) Postmodern Public
Hill and Hupe (2002: 123) identified seven categories of Administration: Towards Discourse. Thousand
independent variables seen as having an impact on the Oaks Calif.: Sage.
results of policy processes. These categories are the follow- Frederickson, H.G. and K.B. Smith (2003) The
ing: a) variables dealing with policy characteristics; b) vari- Public Administration Theory Primer. Boulder,
ables on the policy formation sub process; c) variables
Colorado: Westview Press.
about the kind of ‘vertical public administration’ con-
cerned (number of layers; their legitimate competence,
Graaf, H., van de and R. Hoppe (1989) Beleid en
actual relationships, etc.); d) variables regarding the politiek. Bussum: Coutinho.
response from implementers to the policy involved; Gunsteren, H.R. van (1976) The Quest for Control:
e) variables dealing with the horizontal inter-organizational A Critique of the Rational Central Rule-Approach
relationships; f) variables on the response from those in Public Affairs. London: John Wiley and Sons.
affected by the policy, and g) variables regarding the con- Gunsteren, H.R. van (1980) ‘Planning in de verzorg-
figuration of wider macro-environmental factors. ingsstaat: Van chaotisch naar systematisch falen’
in J.K.M. Gevers and R.J. in ’t Veld (eds.), Planning
als maatschappelijke vormgeving Deventer: Van
REFERENCES Loghum Slaterus: 15–32.
Heinrich, C.J. and L.E. Lynn, Jr (eds.) (2000)
Governance and Performance; New Perspectives.
Anderson, J.E. (1975) Public Policy-making. Washington, D.C.: Georgetown University Press.
New York: Holt, Praeger. Hill, M.J. and P.L. Hupe (2002) Implementing Public
Baumgartner, F.R. and B.D. Jones (1993) Agendas Policy: Governance in Theory and Practice. London:
and Instability in American Politics. Chicago: Sage.
University of Chicago Press. Hill, M.J. and P.L. Hupe (2003) ‘The multi-layer
Bowen, E.R. (1982) ‘The Pressman-Wildavsky problem in implementation research’. Public
paradox’. Journal of Public Policy, 2 (1): 1–21. Management Review, 5 (4): 469–488.
Brewer, G.D. (1974) ‘The policy sciences emerge: To Hogwood, B.W. and L. Gunn (1984) Policy Analysis for
nurture and structure a discipline’. Policy Sciences, the Real World. Oxford: Oxford University Press.
5 (3) (September): 239–44. Hoogerwerf, A. (ed.) (1978) Overheidsbeleid: Een
Brewer, G.D. and P. DeLeon (1983) The Foundation inleiding in de beleidswetenschap. Alphen aan den
of Policy Analysis. Monterey, California: Brooks/ Rijn: Samsom.
Cole. Howlett, M. and M. Ramesh (2003) Studying Public
Cohen, M.D., March, J.G. and Olsen, J.P. (1972) Policy: Policy Cycles and Policy Subsystems. Second
‘A garbage can model of organizational choice’, edition, Ontario: Oxford University Press.
Administrative Science Quarterly, 17 (1): 1–25. Jenkins, W.I. (1978) Policy Analysis: A Political and
Day, P. and R. Klein (1987) Accountabilities. London: Organisational Perspective. London: Martin
Tavistock. Robertson.
Davies, H.T.O., S.M. Nutley and P.C. Smith (eds.) Jenkins-Smith, H. and P.A. Sabatier (1993) ‘The
(2000) What Works? Bristol: Policy Press. study of the public policy process’ in P.A. Sabatier
DeLeon, P. (1999) The Stages Approach to the Policy and H.C. Jenkins-Smith (eds.) (1993) Policy
Process: What Has it Done? Where Is It Going? Change and Learning: An Advocacy Coalition
in: P.A. Sabatier (ed.) (1999) Theories of the Approach. Boulder, Colorado: Westview Press.
Policy Process. Boulder, Colorado: Westview Press, John, P. (1998) Analysing Public Policy. London:
19–32. Pinter.
Derthick, M. (1972) New Towns in Town: Why a Jones, Ch. (1970) An Introduction to the Study of
Federal Program Failed. Washington, D.C.: Urban Public Policy (first edition). Belmont, California:
Institute. Wadsworth.
Peter-3383-Chapter-01.qxd 6/8/2006 7:31 PM Page 29
Kingdon, J.D. (1995) Agendas, Alternatives and Mack, R. (1971) Planning and Uncertainty.
Public Policies (second edition). Boston: Little, New York: John Wiley.
Brown and Company. Matland, R.E. (1995) ‘Synthesizing the implemen-
Kiser, L.L. and E. Ostrom (1982) ‘The Three Worlds of tation literature: The ambiguity-conflict model
Action: A Metatheoretical Synthesis of Institutional of policy implementation’, Journal of Public
Approaches’ in E. Ostrom (ed.) (1982) Strategies of Administration Research and Theory, 5 (2): 145–74.
Political Inquiry. Beverly Hills, California: Sage. May, J. and A.B. Wildavsky (eds.) (1978) The Policy
Kooiman, J. (ed.) (1993) Modern Governance: New Cycle. Beverly Hills, California: Sage Publications.
Government-Society Interactions. London: Sage. Meier, K.J. (1999) ‘Are we sure Lasswell did it this
Kooiman, J. (1999) ‘Social-Political Governance: way? Lester, Goggin and implementation
Overview, Reflections and Design’, Public research’, Policy Currents, 9 (1): 5–8.
Management (Review), 1 (1): 67–92. Milward, H.B. and K.G. Provan (1999) ‘How net-
Kooiman, J. (2003) Governing as Governance. London: works are governed’, unpublished paper.
Sage. Nakamura, R. (1987) ‘The textbook policy process
Kuypers, G. (1980) Beginselen van beleidsontwikkel- and implementation research’. Policy Studies
ing (two volumes). Muiderberg: Coutinho. Review, 7 (2) (Autumn): 142–54.
Lasswell, H.D. (1951a) ‘The policy orientation’ in Ostrom, E. (1999) ‘Institutional Rational Choice: An
Lerner, D. and H.D. Laswell (eds.) (1951) The Assessment of the Institutional Analysis and
Policy Sciences. Stanford, California: Stanford Development Framework’ in P.A. Sabatier (ed.)
University Press. (1999) Theories of the Policy Process. Boulder,
Lasswell, H.D. (1951b) ‘Democratic character’ in Colorado: Westview Press, 35–71.
The Political Writings of Harold D. Lasswell. O’Toole, L.J., Jr (1986) ‘Policy recommendations for
Glencoe, Illinois: Free Press. multi-actor implementation: An assessment of
Lasswell, H.D. (1956) The Decision Process: Seven the field’. Journal of Public Policy, 6 (2): 181–210.
Categories of Functional Analysis. College Park, O’Toole, L.J., Jr (2000) ‘Research on Policy Imple-
Maryland: University of Maryland. mentation: Assessment and Prospects’. Journal
Lasswell, H.D. (1968) ‘The policy sciences’ in The of Public Administration Research and Theory,
Encyclopedia of the Social Sciences, vol. 12. 10 (2): 263–88.
New York: Macmillan/Free Press. Palumbo, D. (1988) Public Policy in America. New
Lasswell, H.D. (1970) ‘The emerging conception of York: Harcourt, Brace, Jovanovich.
the policy sciences’. Policy Sciences, 1: 3–14. Parsons, W. (1995) Public Policy: An Introduction to
Lasswell, H.D. (1971) A Pre-view of Policy Sciences. the Theory and Practice of Policy Analysis.
New York: Elsevier. Cheltenham, UK: Edward Elgar.
Linder, S.H. and B.G. Peters (1987) ‘A design per- Perrow, C. (1999) Normal Accidents: Living with
spective on policy implementation: The fallacies High Risk Technologies. Princeton: Princeton
of misplaced prescription’. Policy Studies Review, University Press.
6 (3): 459–75. Pierre, J. and B.G. Peters (2000) Governance, Politics
Lindblom, C.E. (1959) The science of muddling and the State. New York: St Martin’s Press.
through, Public Administration Review, 19: 78–88. Pressman J. and A. Wildavsky (1973, first edition)
Lindblom, C.E. (1968) The Policy-making Process. Implementation: How Great Expectations Are
Englewood Cliffs, N.J.: Prentice-Hall. Dashed in Oakland; Or, Why It’s Amazing that
Lindblom, C.E. and E.J. Woodhouse (1993) The Federal Programs Work at All, This Being a Saga of
Policy-making Process, third edition. Englewood the Economic Development Administration as Told
Cliffs, N.J.: Prentice-Hall. by Two Sympathetic Observers Who seek to Build
Lipsky, M. (1980) Street-level Bureaucracy: Dilemmas Morals on a Foundation of Ruined Hopes.
of the individual in public services. New York: Berkeley: University of California Press.
Russell Sage Foundation. Richards, D. and M.J. Smith (2002) Governance and
Lynn, L.E., Jr (1981) Managing the Public’s Business. Public Policy in the UK. Oxford: Oxford University
New York: Free Press. Press.
Lynn, L.E., Jr (1987) Managing Public Policy. Boston, Rose, R. (1973) ‘Comparing public policy: An
Mass.: Little, Brown and Company. overview’. European Journal of Political Research,
Lynn, L.E., Jr, C. Heinrich and C.J. Hill (2001) 1: 67–94.
Improving Governance: A New Logic for Empirical Rothstein, B. (1998) Just Institutions Matter: The
Research. Washington DC: Georgetown University Moral and Political Logic of the Universal Welfare
Press. State. Cambridge: Cambridge University Press.
Peter-3383-Chapter-01.qxd 6/8/2006 7:31 PM Page 30
Sabatier, P.A. (1991) ‘Towards better theories of the Schlager, E. (1999) ‘A Comparison of Frameworks,
policy process’. Political Science and Politics, 24 (2) Theories, and Models of Policy Processes’ in
(June): 147–56. P.A. Sabatier (ed.) (1999), Theories of the Policy
Sabatier, P.A. (1999) ‘The need for better theories’ in Process. Boulder, Colorado: Westview Press, 233–60.
P.A. Sabatier (ed.) (1999) Theories of the Policy Selznick, P. (1949) TVA and the Grass Roots.
Process. Boulder, Colorado: Westview Press, 3–17. Berkeley: University of California Press.
Sabatier, P.A. (ed.) (1999) Theories of the Policy Simon, H.A. (1945) Administrative Behavior. First
Process. Boulder, Colorado: Westview Press. edition. New York: Free Press.
Sabatier, P.A. and H. Jenkins-Smith (1999) ‘The Stone D.A. (1989) ‘Causal stories and the formation
Advocacy Coalition Framework: An Assessment’ of policy agendas’, Political Science Quarterly,
in P.A. Sabatier (ed.) (1999) Theories of the Policy 104: 281–300.
Process. Boulder, Colorado: Westview Press, Waldo, D. (1946) The Administrative State: A Study of
117–66. the Political Theory of American Public Adminis-
Schlager, E. (1997) ‘A response to Kim Quaile Hill’s tration. New York: The Ronald Press Company.
In Search of Policy Theory’, Policy Currents, Wilson, W. (1887) ‘The study of administration’,
7 (June): 14–5. Political Science Quarterly, 2: 197–222.
Peter-3383-Chapter-02.qxd 6/8/2006 7:31 PM Page 31
2
The Policy Sciences: Past,
Present, and Future
In contrast, the American university at the later articulated by Brewer (1974) and, subse-
turn of the 20th century housed a number of quently, Brewer and DeLeon (1983) (also see
disciplinary approaches, such as political Anderson 1975/1979 and Jones 1970/1984)
science, anthropology, geography, law, psy- includes Policy Initiation, Policy Estimation,
chology, sociology, and public health, that were Selection, Program Implementation, Program
the natural precursors to the study of public Evaluation, and Policy Termination. Providing
affairs in general and the activities of govern- a conceptual breakdown of policy formulation
ment in particular. Heineman et al. 2002 (also and execution, with each stage possessing
Fischer 2003) have singled out public adminis- unique characteristics, the stages approach
tration and political science as progenitors in (referred to as a ‘heuristic’ by Sabatier 1999
this particular focus. However, the policy and Fischer 2003) offers a mechanism to
sciences approach and their authors have care- achieve a multidisciplinary and value-oriented
fully distinguished themselves from these early approach to policy. In practice, however,
disciplinary contributions by offering three researchers have broken the stages into dis-
defining characteristics of the approach: jointed units, as we shall see below, betraying
the holistic intent of Lasswell’s process and
1. The policy sciences are explicitly problem- resulting in an implied linear rationality
oriented, rejecting the study of a specific devoid of idea and value (DeLeon 1999).
phenomenon per se; the societal or politi- Paul Sabatier (1993 and 1999), along with
cal question of ‘so what’ has always been Robert Nakamura (1987) and others, have been
integral to the policy sciences. Likewise, very critical of the stages process, noting,
problems occur in particular contexts that among other things, that it neglects ‘the role of
must be considered in terms of both the ideas – particularly ideas involving the rela-
analysis and later recommendations. tively technical aspects of the policy debates –
2. The policy sciences are distinctively multi- in policy evolution’ (Sabatier 1993: 15). He
disciplinary in their intellectual and practi- (and co-author Hank Jenkins-Smith) have
cal approaches; virtually every social or severely criticized the policy process frame-
political problem has components tied to work for its theoretic shortcomings, specifi-
varying academic disciplines without cally (Jenkins-Smith and Sabatier 1993: 3–4;
falling clearly into any one discipline’s emphases in original):
exclusive domain.
3. The policy sciences’ approach is explicitly • ‘The stages model is not really a causal
value oriented; in many cases, the central model at all.’ That is, it did not lend itself to
theme deals with the democratic ethos and prediction or even how one stage transi-
human dignity, thus denying the strictures tioned to another.
of logical positivism that were so prevalent • ‘The stages model does not provide a clear
in the American social sciences in the 20th basis for empirical hypothesis testing.’ That
century.3 This value orientation recognizes is, it is not amenable to amendment, con-
that no social problem is without a value firmation, or verification.
component. As such, in order to under- • ‘The stages heuristic suffers from descriptive
stand a problem, one must acknowledge inaccuracy in posing a series of stages …’
its value components. Similarly, no policy • ‘The stages metaphor suffers from a built-
analyst is without her/his own values, in legalistic, top-down focus.’
which also must be addressed (Amy 1984; • ‘The stages metaphor inappropriately
Stone 1998).4 emphasizes the policy cycle as the temporal
unit of analysis.’
The policy sciences have been operational- • ‘The stages metaphor fails to provide a
ized as a process delineated in terms of discrete good vehicle for integrating the roles of
stages in the policy process. The decision policy analysis and policy-oriented learning
process originally proposed by Lasswell (1956),5 throughout the public policy process.’
Peter-3383-Chapter-02.qxd 6/8/2006 7:31 PM Page 33
Many of Sabatier’s observations are correct Lindblom and Cohen 1979), or, if it is, to what
when one views the policy process approach effect or purpose, as David Kirp (1991) talks
(or what Sabatier labels a ‘metaphor’ or semi-facetiously (one hopes) about the ‘end of
‘heuristic’) but there is little evidence that such policy analysis’? Heineman and his colleagues
a set of criteria was ever intended (DeLeon (2002, 1, 9) speak of this concern:
1999: 24 and Brunner 1991) or even appropri- … despite the development of sophisticated methods of
ate. Rather, the stages approach is designed to inquiry, policy analysis has not had a major substantive
feature different stages of the policy process, impact on policymakers. Policy analysts have remained
highlighting their distinct functions and fea- distant from power centers where policy decisions are
made… . In this environment, the values of analytical
tures, ranging from Policy Initiation to Policy rigor and logic have given way to political necessities.
Termination, and provide the necessary guide-
lines. For example, different mechanisms Radin (2000) provides a counter to these
attend policy estimation compared to policy charges of despair, arguing that policy analysis
implementation. A review of the policy litera- has not only produced excellent and effective
ture over the past forty years indicates that the policy research but has had a distinct effect on
stages approach has done precisely that (see policymaking, although not as much as a pro-
DeLeon 1999: 22). In that sense, Lasswell’s ponent would have preferred. And the policy
model continues as a beacon, although, as we ‘market’ place would be supportive, in terms of
will see below, the stages’ particular roles have the number of policy analysts employed in a
been amended by lessons drawn from various lengthy list of policy agencies.
political events. We need not necessarily agree with all of the
claims of the demise of the policy sciences and
certainly not the utility of policy research in
THE APPLICATION OF THE general. Still, one can assert that the Lasswellian
POLICY SCIENCES charge for the policy sciences in either applica-
tion or concept has not been universally real-
ized. Let us take, then, a moment to chronicle
Moving the policy sciences from the halls of the political events that have had a noticeable
academe to the offices of government largely effect on the policy sciences to better appreci-
occurred on the federal level during the 1960s ate their evolution.
(see Radin 2000), such that, by the 1980s,
virtually every federal office had an analytic
office. Since then, many states (including
THE GROWTH OF THE POLICY SCIENCES
memberships in inter-state consortia, such as
the National Conference of State Legislatures)
have built up policy analysis shops to the In general, two paths have been proposed to
extent budgets permit. In addition, for-hire outline the development of the policy sciences.
‘think tanks’ of most every political orientation Beryl Radin (2000) has characterized the insti-
have proliferated. Every public sector official tutional growth of the policy approach, largely
would agree that more information on which relying on the (fictional) histories of an ‘old
to base decisions and policies is better than school’ economist cum policy analyst juxta-
less. To serve that demand, virtually every uni- posed with a ‘younger,’ university-trained policy
versity has a graduate program in public affairs analyst. Through them, she casts an institutional
(or has re-tooled its public administration framework on the policy sciences, indicating
program) to fill the apparent need for sophis- their march from a limited analytic approach,
ticated policy analysis. Yet the turn of the 21st practiced by relatively few practitioners, to a
century has hardly ushered in a Golden Age of growing number of government institutions.
Policy Advice. One needs to ask why policy Specifically, Radin notes the emergence of ana-
scientists increasingly voice the perception that lytic studies from the RAND Corporation to the
their work is not being utilized (Weiss 1980; US Department of Defense (DoD) in the early
Peter-3383-Chapter-02.qxd 6/8/2006 7:31 PM Page 34
1960s (under the guise of ‘systems analysis’ and campaign. However, after the war, while the
a Programmed Planning and Budget System, ‘supply’ side of the policy equation was ready,
PPBS). From its apparent success in the Defense there was little on the ‘demand’ side. Policy-
Department, President Lyndon Johnson man- makers, perhaps tired of the wartime exigen-
dated the government-wide diffusion of PPBS, cies or perhaps enveloped with a return to
most visibly in the Department of Health, peacetime ‘normalcy,’ did not take these newly
Education, and Welfare, in the mid-1960s. honed skills into consideration.
Although success in the DoD of PPBS’s DoD Still, these wartime activities established an
was never duplicated elsewhere (see Wildavsky important illustration of the ability of the
1984; Schick 1973), the analysis orientation social sciences to direct problem-oriented
soon was adopted by a number of federal analysis to urgent public issues, in this case
offices, state agencies, and a number of analytic assuring victory over the Axis powers. As a
consultant groups (see Fischer 1993 and Ricci point of interest, Lasswell and Kaplan spent the
1984). Thus, Radin views the growth of the war with the Library of Congress studying how
policy sciences as a ‘growth industry,’ in which a to best utilize (and protect against) propa-
few select government agencies first adopted an ganda. These realizations led directly to the
explicitly innovative analytic approach, others formation of the National Science Foundation
adopted similar approaches, and a correspond- and the Council of Economic Advisors (see
ing industry developed. Polsby 1984), as well as research facilities such
DeLeon (1988) has cast the growth of the as The RAND Corporation (Smith 1966). Yet,
policy analysis in a consonant but more com- as a result of the imbalance of supply and
plicated manner, in which he ties the growth of demand, the policy analytic approach was
specific analytic ‘lessons learned’ to given more or less quiescent until the 1960s, when
political events. In his view, political condi- the assassination of President John Kennedy
tions effectively supplied analysts with particu- and the succession of President Lyndon Johnson
lar scenarios and data to which they could turn conspired to declare a War on Poverty.
their skills, thus impressing both their imme- The policy sciences faced another opportu-
diate policymaking clients and the larger pop- nity to practice their skills during the War on
ulation, as well with the perspicacity of the Poverty, namely the need to confront social
approach. In particular, he suggested that the complexity and identify the central problem,
resulting policy initiatives (which he termed namely, the pervasive poverty – largely fueled
‘supply’) and policymakers’ requirements by the emerging civil rights demonstrations –
(‘demand’) collaborated to define the specific afflicting what Harrington (1960) called ‘the
development of the policy sciences. The two, other America.’ Even though poverty had
he posits, must be synchronous for a synergis- always been a part of the American fabric, US
tic relationship to develop. policymakers found that they were remark-
DeLeon (1988) initially set forth five politi- ably uninformed about the conditions and
cal conditions that articulated the policy extent of poverty in America. Social scientists
sciences: the Second World War; Lyndon moved aggressively into this knowledge gap
Johnson’s ‘War on Poverty;’ America’s involve- with unbridled enthusiasm, if not always rele-
ment in the Vietnam War; the ‘Watergate vant insights, producing what Moynihan
Affair’ and the ensuing impeachment of (1969) called ‘maximum feasible misunder-
President Richard Nixon; and the 1970’s standing.’ Policymakers proved to be inher-
Energy Crisis. ently limited in their views by their unique set
During the Second World War, the United of experiences.
States marshaled an unprecedented array of To engage the campaign against poverty, a
social scientists – economists, political scien- vast (if not necessarily coordinated) number
tists, psychologists, etc. – to support the war of social programs (Model Cities, VISTA,
effort, ranging from managing the domestic Headstart, and a host of programs out of the
economy to coordinating the strategic bombing Office of Economic Opportunity, OEO) was
Peter-3383-Chapter-02.qxd 6/8/2006 7:31 PM Page 35
initiated, with important milestones being Vietnam from the early 1960s until the early
achieved, especially in the improved statistical 1970s. The decade-long Vietnam war brought
measures of what constituted poverty and the detached policy analysis instruments,
evaluation measures to assess the various anti- including applied systems analytic techniques,
poverty programs (Rivlin 1970), but poverty to the intimate horrors of combat, with politi-
was insistently endemic. Walter Williams cal conditions exacerbated by the growing
(1998), looking back on his days in the OEO, domestic civil unrest as to its conduct and, of
has suggested that these were the ‘glory days’ of course, the loss of lives. The war was closely
policy analysis. Other OEO veterans, such as monitored and managed by the Secretary of
Robert Levine (1970), were more reserved, Defense’s office, with close oversight from a
while some, such as Murray (1984), indicated succession of presidents. It became increas-
that, with the advent of the anti-poverty pro- ingly obvious that analytic rigor – specified in
grams, the American poor was actually ‘losing terms such as ‘body counts,’ sorties flown, and
ground.’ At best, policy analysts were forced to hamlets ‘pacified’ – and ‘rational’ decisionmak-
confront the immense complexity of the social ing were not indicative of the growing rancor
condition. Later, DeLeon was to ask ‘if ten of the war. There was repeated evidence that
years and billions of dollars had produced any ‘hard and fast’ numbers were being manipu-
discernible, let alone effective, relief ’ (DeLeon lated to serve political purposes. Moreover, sys-
1988: 61). tems analysis was not intellectually able to
The ‘policy lessons’ ascribed to the War capture analytically the almost daily changes in
on Poverty were three-fold. In the first place, the war’s activities, occurring on both the
the policy sciences were thwarted by policy- international and domestic arenas (see Gelb
makers’ inability to understand and respond to and Betts 1979).
the complexity of poverty as presented to To return to DeLeon’s metaphor, during the
them. There was an inability to formulate per- Vietnam War and its domestic ramifications, the
suasive arguments in the policy initiation policy ‘supply’ could not square with the politi-
stage. Second, policymakers and analysts ‘dis- cal ‘demand.’ In terms of policy estimation, sys-
covered’ the vagaries of implementation (see tems analysis, one of the apparent US advantages
Pressman and Wildavsky 1984 for a particularly of defense policymaking, was surprisingly
cogent example); in retrospect, this ‘imple- myopic and was a partial contributor to the
mentation blinder’ could have been foretold by ultimate US failures in Vietnam (Gray 1971).
the public administration scholars had they Department of Defense analysts could not
been engaged, but the difficulties encountered reflect the required (and respective) political
eviscerated many programs (see Derthick wills necessary to triumph, as The Pentagon
1972). Finally, and arguably the most success- Papers (Sheehan 1971) subsequently showed. On
ful learning experience, policymakers learned the other hand, Frances FitzGerald’s Fire in the
to demand thorough evaluations of a variety of Lake (1972) foretold the inevitable American
policy programs, even though, at that time, military disaster, as the North Vietnamese were
evaluators were still methodologically naïve; willing to incur whatever losses were needed in
more to the point, however, both parties failed what they saw as the defense of their nation.
to realize the political nature inherent in these Even if the war effort itself had been well con-
exercises. In effect, policy soldiers in the War ducted (surely an arguable point), the manner in
on Poverty failed to frame the right questions, which the war was visualized and projected by
account for the factors that affected implemen- the analytic community left much to be desired,
tation, and were unable to evaluate the pro- a shortcoming widely noted in the domestic
grams with discernment. In short, the War on anti-war community and, ultimately, in policy-
Poverty served as an annealing agent for policy making circles.
research. The policy sciences learned that, in spite of
The disappointment of the policy sciences the best analysis, good policy analysis is insepa-
recurred with the unfortunate experiences in rable from values. The events surrounding the
Peter-3383-Chapter-02.qxd 6/8/2006 7:31 PM Page 36
These constellations of events have mani- context always matters. And policymakers
fested themselves in a worrisome position for must be willing to see and accept the products
the policy sciences, that is, a general disillusion of the policy sciences and their proponents.
in the way in which the American people view Confronting complex problems not only
their government and its processes and, as a requires a multidisciplinary approach. Many
result, the role of the policy sciences. From the have held that the key to the policy process
immense national pride that characterized the basically must accommodate a broad under-
victory over totalitarian forces in the Second standing with accurate problem framing
World War, the American voter has suffered a (Schön and Rein 1994). In spite of a set of
series of on-going disappointments, ranging strong analytic skills, the policy sciences are
from what many consider to be a failed War on inseparable from values, normative concepts,
Poverty to a failed war in Southeast Asia, to the and political ideology. Yet, the policy sciences
unprecedented (in living memory) attacks on have not regularly integrated complexity and
Washington DC and New York City, to the fail- values with policymaking. One needs to ask:
ure of US troops to be treated as ‘liberators’ in Why should the nominal recipients of the
Iraq. Watergate cast a darkening pall on the policy sciences subscribe to them if they do not
American body politic; the Clinton Adminis- manifest the values and intuitions of the client
tration did little to dispel those clouds; and policymaker? To this question, one needs to
President Bush – who was elected in 2000 on a add the question of democratic procedures, a
platform of lowering political dissonance in tenet virtually everybody would agree upon
Washington – has not been able to reduce the until the important issues of detail emerge (see
partisan tensions. Thus, scholars like E.J. DeLeon 1997; Barber 1984; Dahl 1970/1990),
Dionne writes Why Americans Hate Politics e.g., does direct democracy have a realistic
(1991) or Joseph Nye (and colleagues) edit a place in a representative democracy?
book Why Americans Don’t Trust Their Govern-
ment (1997) disparaging the American body
politic and, with it, the policy advice industry.
THERE ARE NO EASY ANSWERS
Most damaging to the policy sciences’ tradi-
tion is Christopher Lasch’s pointed and hardly
irrelevant question: ‘does democracy have a Let us submit the following proposition: that,
future? … It isn’t a question of whether democ- on face value, the policy sciences approach has
racy can survive … [it] is whether democracy inherent strengths, both in ‘knowledge of ’ gov-
deserves to survive’ (Lasch 1995: 1 and 85; ernment (i.e., understanding the processes),
emphases added). and ‘knowledge in government’ (what they
To be sure, political activities are not syn- offer policymakers in terms of substance). But
onymous with the practice of the policy what we have seen above is that the juncture of
sciences. But the two indisputably reside in the the Lasswellian vision of the policy sciences
same policy space. For the policy sciences to with workaday policymaking has not been
meet the goals of improving the processes and realized, often because the analytic ‘supply’ has
results of government through a rigorous not coincided with the policymaking ‘demand.’
application of its central themes, the failures of So how can one best prescribe the policy ana-
the body politics naturally must be at least par- lytic skills that policymakers request from their
tially ascribed to the policy advice industry, advisors and how can the policy sciences best
which includes the policy sciences. Historical respond with integrity? Inherent in this ques-
examples have shown that the supply and tion is a principal assumption: policy scien-
demand conditions for the policy sciences are tists, in the words of Aaron Wildavsky (1979),
necessary but not sufficient for good policy. must ‘speak truth to power.’ Without access to
Supply and demand for policy analysis needs and trust from policymakers, the policy
to be coordinated around the right issue at the sciences lose their sine quo non. They are, from
right time to the right person/agency. As such, their earliest iteration, an applied discipline; if
Peter-3383-Chapter-02.qxd 6/8/2006 7:31 PM Page 39
the policy sciences become irrelevant through identity, economic instability, environmental
lack of application or, to borrow another degradation, militaristic ideals, and a history
metaphor, if (policy) advice does not match and fears of marginalization, or what Samuel
(political) consent, then the policy sciences Huntington (1996) has referred to a as ‘clash of
will have failed to meet the challenges spelled civilizations and the remaking of world order.’
out by its earliest advocates. The ideological debates between public and pri-
Of course, we should not necessarily abide vate sectors continue, with the relatively new
by a counsel of despair. As policy scientists, infusion of the nonprofit sector (both nation-
we need to recognize that a variety of condi- ally and internationally) assuming an increas-
tions have changed (witness, for one, the ingly important role in service provision. The
revolution in technologies that directly affect resulting process makes it clear that no specific
the productivity levels of the American sector has a monopoly on the processes or prod-
worker) and, moreover, that no one has ever ucts of governance. Moreover, the bifurcation
suggested that the policy sciences must of the American body politic along largely par-
remain constant to their original vision; tisan lines often makes agreement on specific
mutatis mutandis in terms of context and policy issues problematic.
processes must be part of the policy sciences.
In this section, then, let us outline a postpos-
itivist approach to enhance the policy Postpositivism and the
sciences and a few relatively new methodolo- Policy Sciences
gies and approaches (e.g., social network
analysis, participatory policy analysis, and Q- The policy sciences community has never been
methodology) that the policy sciences might blind to the presence of and competition
wish to apply to a changing world. But, as we between competing values, but perhaps values
will show, none of these (or the combination) have been under-represented in policy research
will act as an analytic Rosetta Stone; the policy taken as a whole. This amendment, then, is to
community, its issues, and its membership are find conceptual approaches and tools that
too diverse and, in some cases, oppositional. accommodate the diversity. To that end,
There are, in short, no easy answers. numerous authors (Fischer 1998; Schneider
Let us first offer a few milestones from which and Ingram 1997; Bobrow and Dryzek 1987;
to view the landscape as a means to assess in Schön and Rein 1994; Dryzek 1990; Hajer and
what ways conditions have changed since Wagenaar 2003; Forester 1999) have advanced
Lasswell. The early twenty-first century brings a a postpostivist perspective. Although not uni-
world that is increasingly interdependent, where versally accepted (see Sabatier 1999 and Lynn
regional issues have global reach. Economies 1999), its advantages warrant discussion.
and social systems are inextricably connected In response to the shortcomings of framing
and interdependent, as transnational economic and issue understanding, as well as presenting
activities spur local responses; for instance, basi- a more encompassing epistemological per-
cally invisible (and all but invulnerable) life spective, the postpositivist perspective – which
forms called ‘prions’ can result in ‘bovine includes a variety of differing methodologies,
spongiform encephalopathy’ (a degenerative such as ‘deliberative’ (Forester 1999), policy
and fatal brain disease, or Creutzfeldt-Jacob discourse (Hajer 1993), argumentation (Fischer
syndrome in cattle, or what is popularly referred and Forester 1993), interpretative, and narra-
to as ‘mad cow’ disease), a local event that affects tive (Roe 1994) – provides a more thorough
the food supply of nations oceans away, upset- prescription for dealing with diverse, intercon-
ting international commerce and threatening nected, value-laden policy issues.7 We will deal
public health regimes. While in general the end with two arguments in order, the first being a
of the Cold War has brought a shift from com- rejection of the positivist orientation, the
munistic to democratic thought, the realization second being more constructive (dare we say
of these benefits is challenged by issues of ethnic ‘positive’?) in nature.
Peter-3383-Chapter-02.qxd 6/8/2006 7:31 PM Page 40
Regarding the first assessment, Fischer The policy researcher must recognize that
(1998: 143; also see Fischer 2003) provides a not all empirical research is to be discarded in
model postpositivist rationale when he writes the postpositivist dustbin of policy analysis. An
that ‘Postpositivism … not only offer[s] a immense corpus of analytic research over the
theory of the social sciences that is identifiable past half-century, as well as highly skilled
readily in our existing practices, it also consti- analysts, have added greatly to our knowledge;
tutes an incorporation of new methods and in many cases, we should be loathe to relin-
approaches rather than a simple rejection of quish these contributions and to deny them,
old ones.’ While empirical or behavioral in Lynn’s (1999) words, a ‘place at the table.’
approaches which have been lumped under a Edward Lawlor (1996) poses the hard ques-
positivist label identify causal relationships as a tions that policy scientists need to ask them-
means of deriving an aggregated, predictive selves before professionally signing on the any
relationship, postpositivist approaches identify new research orientation: ‘What separates the
causal mechanisms as a means understanding a analyst from the journalist or consumer advo-
relationship (DeLeon 1998). Ann Chin Lin cate under this new argumentative turn? What
suggests that ‘[I]nterpretive work reconstructs separates the policy analyst from literary theo-
categories that are organic to the context it rist and critic in the case of narrative policy
studies, and thus is much less likely to be led analysis?’ He continues:
astray by preconceived notions that stem from To disconnect policy analysts from their disciplinary
inappropriate generalizations’ (Lin 1998: 164). roots and charge them with the general communicative
Other authors – Frank Fischer (2003), John functions espoused by the new argumentative school
Dryzek (1990 and 2000), Ronald Brunner would not only remove ‘tools’ as a defining feature of
the field, it would further undermine the already shaky
(1991), Maarten Hajer (with Wagenaar 2003) –
intellectual identity of the field. Postpostivism and
are more strident in their criticisms and have so-called postmodernism in policy analysis is a swamp
identified what they describe as serious episte- of ambiguity, relativism, and self-doubt … creating
mological shortcomings of the positivist more problems for the policy analysis business than it
approach, assumptions, and results, offering solves (Lawlor 1996: 120).
historical examples that attest to its deficien- Thus, without abandoning the positivist tools
cies. Dryzek (1990: 4–6) has been particularly that are appropriate for specific situations –
outspoken in his assessments of positivism, remember, context always counts (see, for
especially of what he (and others) call ‘instru- example, DeLeon 1998) – let us explore some
mental rationality,’ which he claims, research tools that move the policy sciences
destroys the more congenial, spontaneous, egalitarian, forward in the postpositivist direction.
and intrinsically meaningful aspects of human associa-
tion … represses individuals … is ineffective when con-
fronted with complex social problems … makes Social Network Theory, Participatory
effective and appropriate policy analysis impossible … Policy and Q-Methodology
[and, most critically] is antidemocratic.
sectors, working (hopefully but not always) attention to the structural context and content
cooperatively towards a consensus resolution. of exchange, in favor of research on transac-
As such, networks reflect interconnectedness tions. Previous research shows a separation of
among actors, issues, or groups and provide the connection of networks and the process
a mechanism to conceptualize the complex and outcome variables (Scharpf 1990: 161).
relationships among these societal elements Atkinson and Coleman (1992: 160) pose the
(Coleman and Skogstad 1990; Heclo 1978; question: ‘Are there relationships of power and
Rhodes 1990). Castells (1996: 468) summa- dependency that transcend and color individ-
rizes: ‘Networks constitute the new social mor- ual transactions?’ Answers to this and similar
phology of our societies and the diffusion of questions would go far to build a predictive
network logic substantially modifies the oper- theory of policy outcomes.
ation and outcomes in processes of produc- As all theories are refined over time, network
tion, experience, power, and culture.’ theory will continue to be developed. A future
As with other approaches, social network research agenda would address some outstand-
analysis imposes a semblance of order on a ing issues. The first is how policy networks
chaotic reality. It provides form and identity to change and how that change affects the policy
relationships and analysis, yet recognizes the outcomes, whereby research would address the
dynamic nature of boundaries. Networks variables of boundary shifts and inclusion and
strive to address who will participate in which exclusion (Atkinson and Coleman 1992). In this
event. That is, whereas other policy analysis vein, a similar agendum considers the connec-
approaches tend to focus on the hierarchical tions between policy communities and policy
processes that have characterized the process networks, where the former represent a variety
in the past, a network approach examines the of actors and potential actors who are interested
policy process in terms of the horizontal rela- in the policy issue and share interest and beliefs,
tionships that increasingly describe policy though not necessarily concordant, about the
issues. As Heclo (1978: 104) notes, ‘it is policy solution. Policy networks are a subset of
through networks of people who regard each policy communities.8 They are formed on the
other as knowledgeable, or at least needing to basis of exchange, of primarily information and
be answered, that public policy issues tend to resources, and influence and represent the body
be refined, evidence debated, and alternative of actors that interacts regularly (Fischer 2003).
options worked out – though rarely in any Howlett and Rayner (1995) hypothesize that
controlled, well-organized way.’ policy change occurs most readily when the
To address Sabatier’s (1999) criticisms that policy community and networks are unified.
the stages associated with the policy sciences A second area of study is to understand the
are disjointed, networks provide a more fluid ways in which network and community actors
view of the policy process and the contingent develop, ascribe, and share meaning of back-
actors, as well as addressing the complexity ground assumptions, ideas towards scientific
created by conditions of reciprocal interdepen- knowledge, and their role of involvement
dence (Atkinson and Coleman 1992; Scharpf (Fischer 2003). This area is premised on an
1990). Hajer and Wagenaar (2003: 13) place interpretive community, where knowledge is
network analysis at a key juncture of future created not just by the relationships among
policy research when they ask ‘what kind of selected data and variables, but by the inter-
policy analysis might be relevant to under- pretation and situational context in which
standing governance in the emerging network those variables are applied (Innes 1998).
society.’ The third deals with the impact of political
While the identification of network actors ideas and theories about policies. Particularly,
and relationships is, in and of itself, important, attention must shift to ‘the dominant values
its real value to the policy sciences is the iden- guiding public policy, the knowledge base
tification of the content (if not the intensity) of available to policymakers, and the norms that
relationships. Policy research has paid less legitimize various approaches to policy’
Peter-3383-Chapter-02.qxd 6/8/2006 7:31 PM Page 42
(Atkinson and Coleman 1992: 174). Norms and identified and meaning ascribed to the artifacts
values may change over time or differ per actor. is discerned for different members of the
As such, ‘analysts must seek to ascertain the community so that points of conflict, reflect-
more general principles and norms underlying ing different interpretations, can be identified
interpretation of the policy field’ (Atkinson (Fischer 2003; also see Stone 1998). In this way,
and Coleman 1992: 175). stakeholders help analysts understand the con-
As a conceptual and methodological tool, text of analysis (Durning 1999), in theory, on a
networks offer promise. Interestingly enough, more equitable, discursive basis, or what
the injection needed to remedy and develop Habermas (1983) refers to as an ‘ideal speech
this approach is consistent with a postpositivist situation.’ In a complementary basis, Emery
prescription. The value of postpositivism over Roe (1994) has indicated that ‘narrative’ policy
positivistic approaches is the ability to discern analyses can examine competing narratives
context, power, and the answer to ‘why’. Thus, and help frame problems differently, especially
structural interpretations, change processes, in the context of a highly uncertain, charged,
and norms and values of networks can possi- or power-laden issue. The methodology places
bly be handled with postpositivist approaches. an emphasis on learning from ambiguity,
Within the policy tool kit should be mecha- but one needs to carefully extrapolate to other
nisms to enhance participatory policy analysis policy situations (Schram 1995).
(DeLeon 1997), discourse analysis (Torgerson Still, one needs to caution that deliberative
2003; Hajer 1993), and policy learning. Fischer democracy must be treated with the same
(1998: 143) posed the issue directly: ‘Holding skepticism that underlies all policy methodol-
out the possibility of redeeming or realizing a ogy (see Lynn 1999), especially from its propo-
policy science of democracy, [postpositivism] nents. Certainly greater public participation
calls for participatory institutions and prac- will threaten the extant bureaucracies and their
tices that open spaces for citizen deliberation proponents, but the democratic promises of
on contextual assumptions, empirical out- the initiative and referenda have been sorely
comes, and the social meaning of conclusions’ undercut by highly organized political groups
(also see Fischer 2003). But just how one oper- (Broder 2000; Ellis 2002). Torgerson (2003:
ationally reaches the saddle point between par- 119) cautions: ‘The institutionalization of a
ticipation and governing has been a question discursive design … has the potential to influ-
for eternities. Still, participatory policy at least ence the power context from which it emerges
explicitly addresses the issue. Indeed, Torgerson and may … be opposed because of this. Both
(2003) makes the argument that increased the feasibility of discursive designs and their
efforts in participatory policy analysis would co-optive tendencies thus need to be consid-
narrow the perceived gap between politics and ered in terms of the power relationships found
policy, since both parties will be able to address in a particular setting.’ Moreover, few would
their preferences more openly and delibera- advocate for a policy discourse in which sheer
tively. A participative approach has the poten- volume and empty rhetoric outweighed delib-
tial to provide better information that may eration (DeLeon 1997), but the prevalence of
ultimately open the door to broader and more that behavior is seen daily in the public media
appropriate policy solutions. Thus, a parti- and its effects on the body politic.
cipatory discourse or inquiry allows for Of the postpositivist tools, Q-methodology
decreased conflict, increased trust, multiple is consistent with narrative analysis that gets
viewpoints, and normative interpretations, beyond the polarized framing of issues by
and portends towards a greater democratic reframing policy issues in ways not initially
underpinning, along the lines of a ‘one person, perceived, largely by subverting the assump-
one vote’ direct democracy (Barber 1984). tions of objectivism. The method provides
A participatory approach opens the door to information on public values and positions,
interpretive policy analysis, whereby the artifacts providing policy makers with a broader range
that carry meaning to the policy community are of policy options (Durning 1999; van Eeten
Peter-3383-Chapter-02.qxd 6/8/2006 7:31 PM Page 43
2001). ‘Q-methodology is useful in the The history of the policy sciences naturally
prescription context because it can allow a affects the present day analyst/client relation-
public dialogue to take place regarding values ships. We can further say that a continuation of
and can then contribute to the stabilization of the status quo will not be to the benefit of
expectations needed to achieve prescription either party. The ‘promise’ then of the policy
outcomes’ (Steelman and Maguire 1999: 365). sciences is that, to gain in stature and accep-
Q-methodology identifies patterns of subjec- tance, some important new directions must be
tive perspectives across individuals. It groups grafted to the policy sciences’ approach.
individuals with like views to discern how that Moreover, most observers can agree that the
subgroup perceives an issue (Durning 1999). policy sciences cannot surrender their dedica-
Important to its application is to conceptu- tion to a democratic ethos. As Hajer and
alize the relationship between the policy Wagenaar (2003: 15) state, ‘Whatever we have
analyst and the decisionmaker – as consulta- to say about the nature and foundation of
tive or participative – and the stage of the the policy sciences, its litmus test will be that
policy process for which information is gath- it must ‘work’ for the everyday reality of mod-
ered (Steelman and Maguire 1999; Durning ern democracy’ (also see Torgerson 2003).
1999; Roberts 1995; Lasswell 1971). The tech- Given the three defining characteristics of
nique offers policymakers a tool to increase the policy sciences – problem oriented, multi-
the knowledge base and understanding of a disciplinary, and normative – we propose that
situation’s context and meaning. The chal- these directions will more clearly articulate the
lenge to Q-methodology will be to aggregate value-orientation vector. It is important that
preferences to survive a collective process of the value component of policy decisions be
deliberation (Lynn 1999). understood and made more transparent for all
concerned.
To these ends, we have suggested ways in
which this avenue can be accommodated most
IN CLOSING …
readily, through the use of a more ecumenical
postpositivist approach, a more participatory
The policy sciences are unquestionably an set of guidelines, the application of social net-
approach in transition, but it is not clear what work analysis, and perhaps a greater use of
the end point will resemble. The past has been tools such as Q-methodology for specific pur-
promising in many ways, but the policy poses. The common denominator among these
sciences have not achieved the prominence in approaches is that all encourage a less reduc-
policymaking circles that its early proponents tionist, more democratic voice in the policy
might have wished; in addition, at times, the process. These will not come particularly easy,
means to these ends have been problematic. as they require skill sets somewhat different
Part of this was due to the policy-defining con- from those currently practiced. For instance,
ditions that have proven to be remarkably participatory policy analysis requires a certain
complex and analytically unwieldy, a situation adroitness at group processes and mediation.
only worsened by the wholesale movement of We would not want to suggest that these are
the policy analysis community to economics the only new avenues for the policy sciences to
and behavioralism, or what we have called here consider; surely there is a world of alternatives
‘positivism.’ But part of this dyspepsia is that to ponder. However, we do wish to make the
the policy sciences’ offerings were consistently point that, if the policy sciences are to legiti-
not timely nor resonant with the requirement mately aspire to their original claims, featuring
of the policymaking community. Thus, neither ‘knowledge of and knowledge in’ government,
the policy advising nor the policymaking then the ‘same old same old’ will only lead to the
communities have been satisfied with the marginalization of the policy sciences, a fate
applications of the policy sciences, hardly an Lasswell and succeeding generations of policy
encouraging condition. scientists and policymakers can only regret.
Peter-3383-Chapter-02.qxd 6/8/2006 7:31 PM Page 44
Durning, Dan (1999). ‘The transition from tradi- Hajer, Maarten A., and Hendrik Wagenaar (2003).
tional to postpositivist policy analysis: A role for ‘Introduction.’ In Maarten A. Hajer and Hendrik
Q-methodology.’ Journal of Policy Analysis and Wagenaar (eds.), Deliberative Policy Analysis:
Management. Vol. 18, No. 33. Pp. 389–410. Understanding Governance in the Network Society.
Drew, Elizabeth (1994). On the Edge: The Clinton Cambridge, NY: Cambridge University Press.
Presidency. New York: Simon & Schuster/ Pp. 1–33.
Touchstone. Hajer, Maarten A., and Hendrik Wagenaar (eds.)
Dror, Yezekhel (1971). Design for the Policy Sciences. (2003) Deliberative Policy Analysis: Understanding
New York: American Elsevier. Governance in the Network Society. Cambridge,
Dryzek, John S. (1990). Discursive Democracy: NY: Cambridge University Press.
Politics, Policy, and Political Science. Cambridge, Harrington, Michael (1960). The Other America:
UK: Cambridge University Press. Poverty in the United States. New York: Penguin.
—— (2000). Deliberative Democracy and Beyond. Heclo, Hugh (1978). ‘Issue Networks and the
Oxford, UK: Oxford University Press. Executive Establishment.’ In The American
Ellis, Richard J. (2002). Democratic Delusions: The Political System, Anthony King (ed.). Washington:
Initiative Process in America. Lawrence: University American Enterprise Institute.
of Kansas Press. Heineman, Robert A., William T. Bluhm, Steven A.
Fischer, Frank (1993). ‘Policy Discourse and the Peterson, and Edward N. Kearny (2002). The World
Politics of Washington Think Tanks.’ In Frank of the Policy Analyst. Chatham, NJ: Chatham
Fischer and John Forester (eds.), The Argumentative House Publishers. Third Edition.
Turn in Policy Analysis and Planning. Durham, NC: Howlett, Michael and Jeremy Rayner (1995). ‘Do
Duke University Press. Pp. 21–42. ideas matter? Policy network configurations and
—— (1998). ‘Beyond Empiricism: Policy Inquiry in resistance to policy change in the Canadian forest
Postpositivist Perspective.’ Policy Sciences Journal. sector.’ Canadian Public Administration. Vol. 38,
Vol. 26, No. 1 (Spring). Pp. 129–146. No. 3. Pp. 382–410.
—— (2003). Reframing Public Policy. Oxford: Huntington, Samuel P. (1996). The Clash of Civiliza-
Oxford University Press. tions and the Remaking of World Order. New York:
—— and John Forester (eds.) (1993). The Argumen- Simon & Schuster.
tative Turn in Policy Analysis and Planning. Innes, Judith J. (1998). ‘Information on Communica-
Durham, NC: Duke University Press. tive Planning.’ Journal of the American Planning
FitzGerald, Frances (1972). Fire in the Lake. Boston: Association, Vol. 64, No. 1. Pp. 52–63.
Little, Brown. Janis, Irving (1983). GroupThink. Boston: Houghton
Forester, John (1999). The Deliberative Practitioner: Mifflin. 2nd Edition.
Encouraging Participative Planning Processes. Jones, Charles (1970/1984). An Introduction to the
Cambridge, MA: The MIT Press. Study of Public Policy. Belmont, CA: Wadsworth.
Gelb, Leslie, with Richard K. Betts (1979). The Irony 3rd Edition.
of Vietnam: The System Worked. Washington DC: Jenkins-Smith, Hank C., and Paul A. Sabatier
The Brookings Institution. (1993). ‘The Study of the Policy Process.’ In Paul
Goldhamer, Herbert (1978). The Adviser. New York: A. Sabatier and Hank C. Jenkins-Smith (eds.),
American Elsevier. Policy Change and Learning. Boulder, CO:
Gray, Colin (1971). ‘What Has Rand Wrought?’ Westview Press. Chap. 1.
Foreign Policy. No. 4 (Fall). Pp. 111–129. Johnson, Hayes (2001). The Best of Times: America
Greenberger, Martin, Garry D. Brewer, and Thomas in the Clinton Years. New York: Harcourt.
Schelling (1984). Caught Unawares: The Energy Kirp, David L. (1991). ‘The End of Policy Analysis:
Decade in Retrospect. Cambridge, MA: Ballinger. With Apologies to Daniel (The End of Ideology)
Habermas, Jürgen (1983). The Theory of Communica- and Frank (The End of History) Fukuyama.’
tive Action: Reasons and the Rationalization of Journal of Policy Analysis and Management. Vol. 11,
Society. Thomas McCarthy (Trans.). Boston: Beacon No. 4 (Fall). Pp. 693–696.
Press. Landsberg, Hans (1979). Energy: The Next Twenty
Hajer, Maarten A. (1993). ‘Discourse Coalitions and Years. Cambridge: Ballinger for Resources for the
the Institutionalization of Practice: The Case of Future.
Acid Rain in Great Britain.’ In Frank Fischer and Lasch, Christopher (1995). The Revolt of the Elites and
John Forester (eds.), The Argumentative Turn in the Betrayal of Democracy. New York: W.W. Norton.
Policy Analysis and Planning. Durham, NC: Duke Lasswell, Harold D. (1951a). ‘The Policy Orienta-
University Press. Pp. 43–76. tion.’ In Daniel Lerner and Harold D. Lasswell
Peter-3383-Chapter-02.qxd 6/8/2006 7:31 PM Page 46
(eds.), The Policy Sciences. Palo Alto, CA: Stanford Radin, Beryl A. (2000). Beyond Machiavelli: Policy
University Press. Chap. 1. Analysis Comes of Age. Washington, DC: Georgetown
—— (1951b). The World Revolution of Our Time: A University Press.
Framework for Basic Policy Research. Palo Alto, Rhodes, R.A.W. (1990).’Policy Networks: A British
CA: Stanford University Press. Reprinted in Perspective.’ Journal of Theoretical Politics. Vol. 2.
Harold D. Lasswell and Daniel Lerner (eds.), Pp. 293–317.
World Revolutionary Elites. Cambridge, MA: The Ricci, David M. (1984). The Transformation of
MIT Press 1965. Chap. 2. American Politics: The New Washington and
—— (1956). The Decision Process. College Park: the Rise of Think Tanks. New Haven, CT: Yale
University of Maryland Press. University Press.
—— (1971). A Pre-View of Policy Sciences. New Rivlin, Alice M. (1970). Systematic Thinking for
York: American Elsevier. Social Action. Washington DC: The Brookings
—— and Abraham Kaplan (1950). Power and Institution.
Society. New Haven, CT: Yale University Press. Roberts, R. (1995). ‘Public Involvement: From
Lawlor, Edward F. (1996). ‘Book Review.’ Journal of Consultation to Participation.’ In F. Vanclay and
Policy Analysis and Management. Vol. 15, No. 1 D.A. Bronstein (eds.). Environmental and Social
(Winter). Pp. 110–121. Impact Assessment. New York: John Wiley & Sons.
Levine, Robert A. (1970). The Poor Ye Need Not Roe, Emery (1994). Narrative Policy Analysis.
Have With You: Lessons from the War on Poverty. Durham, NC: Duke University Press.
Cambridge, MA: The MIT Press. Sabatier, Paul A. (1993). ‘Policy Change over a
Lin, Ann Chih (1998). ‘Bridging Positivist and Decade or More.’ In Paul A. Sabatier and Hank C.
Interpretivist Approaches to Qualitative Methods.’ Jenkins-Smith (eds.), Policy Change and Learning.
Policy Studies Journal. Vol. 26, No. 1 (Spring). Boulder, CO: Westview Press. Chap. 2.
Pp. 162–180. —— (ed.) (1999). Theories of the Policy Process.
Lindblom, Charles E. and David K. Cohen, (1979). Boulder, CO: Westview Press.
Usable Knowledge: Social Science and Social Problem ——, and Hank C. Jenkins-Smith (eds.) (1993).
Solving. New Haven: Yale University Press. Policy Change and Learning. Boulder, CO:
Lynn, Laurence E., Jr. (1999). ‘A Place at the Table: Westview Press.
Policy Analysis, Its Postpositive Critics, and the Scharpf, Fritz (1990). ‘Games Real Actors Could
Future of Practice.’ Journal of Policy Analysis Play: The Problem of Connectedness.’ Köln:
and Management. Vol. 16, No. 3 (Summer). Max-Planck-Institute für Gesellschaftsforschung,
Pp. 411–425. Paper 90/8.
Milward, H. Brinton and Gary L. Walmsley (1984). Schick, Allen (1973).‘A Death in the Bureaucracy: The
‘Policy Subsystems, Networks and the Tools of Demise of Federal PPB.’ Public Administration
Public Management.’ In R. Eyestone (ed.), Public Review. Vol. 33, No. 2 (March/April). Pp. 146–156.
Policy Formation. Greenwich, CT: JAI Press. Schneider, Anne Larason and Helen Ingram (1997).
Moynihan, Daniel P. (1969). Maximum Feasible Policy Design for Democracy. Lawrence, KS:
Misunderstanding: Community Action in the War University of Kansas.
on Poverty. New York: The Free Press. Schön, Donald and Martin Rein (1994). Frame
Murray, Charles (1984). Losing Ground. New York: Reflection. New York, NY: Basic Books.
BasicBooks. Schram, Stanford F. (1995). ‘Against Policy Analysis:
Nakaruma, Robert (1987).‘The Textbook Polic Process Critical Reason and Poststructural Resistance.’ Policy
and Implementation Research.’ Policy Studies Sciences. Vol. 28, No. 4 (November). Pp. 375–384.
Research. Vol 7, No. 2 (Autumn). Pp. 142–154. Sckopal, Theda (1997). Boomerang: Health Care
Nye, Joseph S., Jr., Philip D. Zelikow, and David C. Reform and the Turn Against Government. New
King (1997) (eds.). Why People Don’t Trust Govern- York: W.W. Norton.
ment. Cambridge, MA: Harvard University Press. Sheehan, Neil (1971). The Pentagon Papers. New
Olson, Keith W. (2003). Watergate: The Presidential York: Bantam and the New York Times.
Scandal that Shook America. Lawrence, KS: The Smith, Bruce L.R. (1966). The RAND Corporation.
University Press of Kansas. Cambridge, MA: Harvard University Press.
Polsby, Nelson W. (1984). Political Innovation in Sobel, Dava (1995). Longitude. New York: Walker
America. New Haven: Yale University Press. and Co.
Pressman, Jeffrey, and Aaron Wildavsky (1984). Steelman, Toddi A. and Lynn A. Maguire (1999).
Implementation, Berkeley: University of California ‘Understanding Participant Perspectives: Q-
Press. 3rd Edition. Methodology in National Forest Management.’
Peter-3383-Chapter-02.qxd 6/8/2006 7:31 PM Page 47
Journal of Policy Analysis and Management. Weiss, Carol H. (1980). ‘Knowledge creep and
Vol. 16, No. 3 (Summer). Pp. 361–388. decision accretion.’ Knowledge. Vol. 1, No. 3.
Stone, Deborah (1998). Policy Paradox. New York: Pp. 341–404.
Norton. Wildavsky, Aaron (1979). Speaking Truth to Power.
Torgerson, Douglas (2003). ‘Democracy through Boston: Little, Brown.
Policy Discourse.’ In Maarten A. Hajer and Hendrik —— (1984). The Politics of the Budgetary Process.
Wagenaar (eds.), Deliberative Policy Analysis: Boston: Little Brown.
Understanding Governance in the Network Society. ——, and Ellen Tanenbaum (1981). The Politics of
Cambridge: Cambridge University Press. Chap 4. Mistrust. Beverly Hills: Sage.
van Eeten, Michel J. G. (2001). ‘Recasting Intractable Williams, Walter (1998). Honest Numbers and Demo-
Policy Issues: The Wider Implications of The cracy. Washington DC: Georgetown University
Netherlands Civil Aviation Controversy.’ Journal Press.
of Policy Analysis and Management, Vol. 20, No. 3. Woodward, Bob (1994). The Agenda. New York:
Pp. 391–414. Simon & Schuster.
Weyant, John P. (1980). ‘Quantitative Models in —— (2004). Plan of Attack. New York: Simon &
Energy Policy.’ Policy Analysis. Vol. 6, No. 2 (Spring). Schuster.
Pp. 221–234.
Peter-3383-Chapter-02.qxd 6/8/2006 7:31 PM Page 48
Peter-3383-Chapter-03.qxd 6/8/2006 7:32 PM Page 49
3
Behavioral Rationality and the Policy
Processes: Toward A New Model of
Organizational Information Processing
B R YA N D . J O N E S , G R A E M E B O U S H E Y , A N D
SAMUEL WORKMAN
the problems of delegation and control in evolved from the efforts of such diverse
bureaucracy. The obsession with control is theorists as Herbert Simon, Mancur Olson,
symptomatic of the larger pathologies of ratio- and Anthony Downs to link individual human
nal choice theory – the core assumptions of indi- decision-making with broader macropolitical
vidual utility maximization simply do not outcomes (Simon 1947; Olson 1982; Downs
capture the complexities of human decision- 1957). Although the competing models of
making. By focusing narrowly on questions gen- rationality disagreed about the fundamental
erated from the control trap, rational choice motivations behind individual choice, they
theory ignores a wide range of intriguing ques- were unified in their belief that the processes
tions about information processing and policy and policy outcomes be most powerfully
dynamics in bureaucratic decision-making. understood through exploring the role of indi-
In the final part of this essay, we point toward vidual behavior in collective decision-making.
an information-processing model of the policy A theory of rationality which anticipated indi-
process that avoids the obsession with control that vidual choices in the context of the larger
permeates rational theories of policymaking. We political process would present a significant
argue that it is both more scientifically sound and advancement in the study of government –
fundamentally more parsimonious than models such a set of assumptions would act as a theo-
based on rational choice. retical tool-box that could be used, not only to
cut through the complexity of public decision-
making, but also to locate areas of conflict
within organizations, and potentially predict
PART I: BOUNDED RATIONALITY
future political outcomes. Both rational choice
AND RATIONAL CHOICE THEORY:
theory and bounded rationality pushed the
AN OVERVIEW
study of public policy away from case studies
and atheoretical descriptions of public admin-
In the behavioral sciences, such as psychology istration toward a generalized theory of public
and behavioral biology, the major topic of policy.
inquiry is individual behavior. In the social Despite these common aims, the theories of
sciences, such as political science and econom- rationality are deeply divided over the most
ics, the aim is to understand social systems of basic assumptions of individual choice.
interacting individuals. Social scientists need Rational choice theory borrows heavily from
both a model of individual behavior and an economic assumptions of individual prefer-
understanding of how individuals interact ences, and believes that a sufficient behavioral
to produce social outcomes, such as those model could be drawn from deductions of an
produced by free markets or policymaking individual’s self interested utility maximiza-
processes in government. As a consequence, tion. Bounded rationality began as a critique of
social scientists cannot dally with the nuances comprehensive rationality, and grew from an
of human behavior that do not impact on the effort to reconcile the reductionist economic
interactions of people in organizations. Any assumptions of rational choice with observed
model of individual choice in policymaking psychological constraints on human decision-
processes must be parsimonious, and they making. The following sections outline the dis-
must link individual level actions to policy- tinguishing characteristics of both rational
making outcomes. choice theory and bounded rationality as
The behavioral models of comprehensive applied to the study of public policy. This cri-
rationality and bounded rationality, which today tique intends not only to cast light on the
divide the theories of public policy, were driven major differences between the two theories,
by a common desire to improve the rigor of but also calls attention to exciting empirical
political analysis. Applications of both rational findings which we believe are fueling increas-
choice theory and bounded rationality in ing theoretical convergence around a single
the study of public policy and administration positive theory of human choice.
Peter-3383-Chapter-03.qxd 6/8/2006 7:32 PM Page 51
behavior. By assuming that actors are that is not the point. It was clearly and cleanly
singularly motivated to maximize gains avail- stated in a generalized manner that facilitated
able in political decisions, a wide range of con- testing against empirical observation.
siderations that have long complicated the study This leads to the third advantage. The ratio-
of political science – social class, partisanship, or nal choice method is a comparably rigorous
cultural values – become peripheral to a deci- approach to the study of political processes.
sion-maker’s preferences. The challenge of Rational choice researchers derive formal mathe-
rational choice theory is to characterize accu- matical models from a set of assumptions
rately the payoffs available to political actors in about individual preferences. These models
a given arena. A classic example of this approach derive a set of hypotheses of anticipating a spe-
can be drawn from David Mayhew’s study of cific outcome in politics, and then test them
congressional voting behavior in Congress: the against the self interested behavior of actors
Electoral Connection. In his analysis, democratic put forward in the theoretical assumptions of
representation is motivated singularly by a con- rational choice theory—in this regard, rational
gressman’s desire to retain political power choice hypotheses are verified or falsified. This
(Mayhew 1974). Ideology and policy-making approach forces researchers to advance well-
matter only insomuch as they instruct a politi- specified models of political choice. Rational
cian as to where his source of political power is choice models focus only on those most crucial
located, and how to make appropriate votes to elements of the political process that are neces-
retain electoral support. sary to explain outcomes.
The second advantage of the rational choice Both the simplicity and the promise of theo-
method is its broad theoretical generalizability. retical generalizablity have made rational
Unlike descriptive studies of political behavior, choice theory a popular tool for the study of
the baseline assumptions of rational choice government. The rational choice approach has
methods are characterized by a ‘universalism been applied to such subfields as organiza-
that reveals generalizable implications beyond tional behavior (Bendor and Moe 1985; Moe
those under immediate investigation’ (Levi 1984) congress (Mayhew 1974; Arnold 1990),
1997, 20). Deductions from rational choice parties and elections (Downs 1967) and collec-
theory are not specific to a particular time and tive action problems (Olson 1982). The
case, but should offer insights wherever similar breadth of this research is suggestive of the
conditions can be observed. Mancur Olson theoretical power of the model. By exploring
argues that ‘the persuasiveness of a theory the strategic behavior of self-interested individ-
depends not only on how many facts are uals, rational choice theorists have produced a
explained, but also how diverse are the kinds of rich and theoretically unified body of research
facts explained’. In Olson’s view, the rational in a discipline once marked by methodological
choice approach has the double benefit of pre- eclecticism.
serving the parsimony of a theory by removing
‘any inessential premises or complexities that
ought to be removed from an argument,’ while
BOUNDED RATIONALITY AND THE
retaining a high degree of explanatory power
THEORY OF POLICY PROCESSES
(Olson 1982: 12–13). Drawing upon a relatively
small set of assumptions about human behav-
ior, Olson is able to advance a series of predic- Perhaps because rational choice theory takes a
tions that serve as the basis of a theory of how decidedly reductionist approach to the study
the formation of interest groups in democratic of government—one which is ‘willing to sac-
societies detract economic growth. That demo- rifice nuance for generalizability, detail for
cratic nations have sustained high levels of logic’ (Levi 1997: 21)—dissenting researchers
growth, even as interest groups have prolifer- have long charged that these sparse assump-
ated, means that Olson’s theory was wrong, but tions of individual utility maximization distort
Peter-3383-Chapter-03.qxd 6/8/2006 7:32 PM Page 53
the complexity of both individual behavior moving between individual decisions and
and organizational decision-making. Field organizational outcomes. Bounded rationality
researchers complain that observations of retains the hallmarks of a theoretical model—
individual behavior rarely match the calculat- it captures only those aspects of human
ing self-interested actor posited in rational behavior needed to understand collective
choice theory (Brehm and Gates 1997; Lipsky decision-making.
1980). The behavioral norm of individual util- As with so many researchers who followed
ity maximization simply does not seem to him, Simon noticed that the assumptions of
reflect accurately the actions of politicians, expected utility analysis failed to match his own
bureaucrats and voters, whose choices so often observations of real world economic decision-
seem to be motivated by risk aversion, sense making (Simon 1999). Looking at budgeting in
of mission, identity, fairness, or altruism. For Milwaukee, Simon observed that relatively little
these critics, rational choice theory is at best individual behavior matched ‘substantive or
idiosyncratic—applicable only to discrete objective rationality, that is, behavior that can be
institutions such as the US congress where adjudged to be optimally adapted to the situa-
political behavior can be safely matched to tion’ (Simon 1985, 294). He instead found that
self-interest (Rockman 2000). At worst rational the processes of both individual and organiza-
choice theory is misleading, as it reduces tional decision-making were a good deal
potentially interesting social behavior such messier than rational expectations would have
as altruism to an individual’s self-interest us believe. Bureaucratic budgets were often
(Monroe 1996). adjusted incrementally, using the prior year’s
Researchers who actually study individual budgeting as a benchmark for future spending
decision making find the rational choice needs (Thompson and Green 2001; Simon
approach curious. Cognitive psychologists find 1991). Organizational decisions were made
the economic assumptions of stable ordered more through horse trading and bargaining
preferences, transitivity, and utility maximiza- than a process even remotely resembling fully
tion to be a strange abstraction of human rational decision-making (Simon 1999). Orga-
decision-making. Findings in psychology indi- nizations and individuals proved poor at
cate that people are poor at forming prefer- generating complex alternatives or making trade-
ences, generating alternatives, and making offs. Environmental factors, such as issue
decisions (Sniderman et al. 1991; Tetlock 2000; salience, individual attention, and time con-
Jones 2001). Studies show that individuals often straints, shaped the depth of solution searches
lack even the most basic tools with which to in organizations.
make informed rational decisions. Preferences Simon (1985) believed the most glaring
and choices are bounded more by emotion and problem with comprehensive rationality was
environmental context than by rational analysis its focus on outcomes rather than the process
(Jones 2001). Research in cognitive psychology of individual decision-making. By ignoring the
makes one wonder how people are able to form procedures of choice, rational choice theory
preferences or make decisions at all. blithely accepted that ‘rational’ outcome
Herbert Simon (1947) developed bounded emerged from self-interested behavior. Such an
rationality as an effort to reconcile the strict approach is especially dangerous when
economic assumptions of comprehensive ratio- attempting to understand the broader mecha-
nality with actual decision-making revealed by nisms of collective policy-making, where the
the empirical study of organizations. Preferences motivations for preference formation and
and choices seemed bounded by cognitive and political behavior are at least as important as
emotional constraints that interfered with the the outcome itself. Simon (1985: 294) explains:
process of purely rational decision-making.
Much like comprehensive rationality, bounded There is a fundamental difference between substantive
rationality offers an efficient method for and procedural rationality. To deduce substantively, or
Peter-3383-Chapter-03.qxd 6/8/2006 7:32 PM Page 54
objectively, rational choice in a given situation, we need research have improved our understanding of
to know only the choosing organism’s goals and the the behavioral model of choice, these four
objective characteristics of the situation. We need know
absolutely nothing else about the organism, nor would
central tenets remain largely valid.
such information be of any use to us, for it could not
affect the objectively rational behavior in any way.
To deduce the procedurally or boundedly rational The Principle of Intended Rationality
choice in a situation, we must know the choosing organ-
ism’s goals, the information and conceptualization it has The principle of intentionality suggests that we
of the situation, and its abilities to draw inferences from
look at the goal directed behavior of people,
the information it possesses. We need know nothing
about the objective situation in which the organism and investigate the manner in which their cog-
finds itself, except insofar as that situation influences the nitive and emotional constitutions either help
subjective representation. or hinder their goal-seeking. This distinguishes
bounded rationality from psychological
Simon argued that a theoretical model ori- theories, which generally focus only on the
ented toward understanding procedurally or limitations of individual choice-makers. While
boundedly rational decision-making would pro- comprehensive rationality assumes single-
vide a more realistic bridge between individual minded maximization, the principle of
and collective choice. A ‘behavioral’ model of intended rationality allows researchers to
rationality would take a decidedly scientific and distinguish between careful cost benefit analy-
inductive approach to understanding decision- sis that closely approximates utility maximiz-
making. In order to make efficient generaliza- ing decision-making, quick decisions based on
tions connecting individual psychological heuristic cues, unthinking reliance on past
processes to collective political and economic strategies, or even spontaneous decisions that
choice, research needed to follow strict scientific seem to make no reference to potential gains or
guidelines. Underlying theoretical assumptions losses. Given the time, costs, and demands of a
of human behavior needed to be tested and specific decision, humans may rely on hard-
retested. Those assumptions that were verified wired biological responses, generalized decision-
through scientific research would be preserved in making strategies, or full information searches.
the scientific model, while others would be mod- ‘Cognitive architecture is most obvious when
ified or abandoned. What bounded rationality action occurs at short time scales. As one
lost in parsimony it would gain in accuracy. moves toward actions that take longer times,
Unlike the artificial behavioral assumptions of cognitive architecture is less and less evident,
comprehensive rationality, bounded rationality and the nature of the task takes on more and
would capture the biological, emotional, and more importance in explaining action’ (Jones
environmental constraints that constrained the 2001, 56). Humans may intend to be rational,
procedures of decision-making. but their decision-making capabilities break
As with rational choice theory, bounded down under time constraints or very high
rationality has been widely applied in the study information costs. Moreover, there is no good
of public policy. While bounded rationalists evidence that people are more rational when
might emphasize different elements of behav- the stakes are high, as some rational choice
ioral model in their research, virtually all theorists have maintained. In direct defiance of
research in bounded rationality draws from that claim, state lotteries sell more tickets when
four core principles: the Principal of Intended the pot is large, lowering the probabilities and
Rationality, the Principal of Adaptation, the the expected return.
Principle of Uncertainty and the Principle of
Trade-offs (Jones 2003). From these principles,
modern researchers have advanced a vigorous Principle of Adaptation
research program that explores how both
people and institutions behave. As we will The principle of adaptation is closely related to
see, even though experimental and empirical the notion of intentionality. Much of human
Peter-3383-Chapter-03.qxd 6/8/2006 7:32 PM Page 55
behavior is explained by the nature of the ‘task on procedures, and may explain instances of
environment’ surrounding a decision. With extremely risk adverse behavior.
time, human decision-making adapts to the
specific nature of the problems they face in
The Principle of Trade-Offs
a specific circumstance. The more time and
learning an individual invests in a specific
The final notion central to models of bounded
problem, the less constrained they become by
rationality is the principle of trade-offs. Unlike
environmental or biological constraints. This
comprehensive rationality, which suggests
notion of adaptation accounts for changes in
that individuals are able to move seamlessly
decision-making efficiency in a single problem
between ranked goals, bounded rationalists
space over time. When a problem is iterated
argue that people find it difficult to trade off
over time, people learn or develop coping
one goal against another when forming prefer-
strategies. Even more intuitively, the principle
ences and making choices (Slovak 1990;
of adaptation may explain why organizations
Tetlock 2000). This critique is a major shift
encourage specialization in areas of complex-
away from the transitivity assumption in ratio-
ity, and routinization in decisions under severe
nal choice theory; however it is important to
time constraints.
capture the volatile shifts in preferences
A rich tradition of research in political
observed by both behavioral economists
science and psychology explores the use of
(Kahneman and Tversky 1986) and public
heuristics in political decision-making—
opinion researchers (Zaller 1992). Preferences
attempting to identify cost-cutting cues indi-
are determined by emotional and cognitive
viduals rely upon in order to limit investment
cues, and are rarely as stable as rational choice
costs from making complex decisions in
theorists have us believe. Because of these
low-information environments. The key issue
trade-off difficulties, Simon argued that indi-
is whether these heuristics are maximally
viduals, and by extension organizations, ‘satis-
adaptive (Lupia, McCubbins and Popkin 2000;
fice,’ quickly choosing an option that is ‘good
Gigerenzer et al. 1999). While some have
enough’ rather than searching for one that
argued that heuristics invariably follow what a
weighs the payoff of every possible choice.
fully rational individual would choose, that is
Because bounded rationality is concerned
quite clearly not the case. ‘Buy a lottery ticket
with the procedures of individual choice, the
when the stakes are high’ is a classic heuristic
school of thought tends to approach the study
that leads to lower expected returns than not
of politics by looking to how individuals
buying at all, or buying when the stakes are
and organizations respond to changes in their
low. It is likely that some heuristics are adap-
problem environments. Because individual
tive calculational crutches and some are mis-
decision-makers have limited attention for
leading and even mal-adaptive.
problem solving, they must address problems
serially, one-at-a-time, which means they are
The Principle of Uncertainty forever juggling inputs, prioritizing them via
the allocation of attention and the sense of
Individuals operate in an environment of urgency that inputs generate. The salience of a
almost constant risk and uncertainty. Because particular problem is almost always generated
of human cognitive architecture, uncertainty is by non-rational elements in politics—by scan-
far more fundamental to choice than expected dal, by crisis, by the mobilization of critics—
utility theory admits. Not only are individuals rather than calm decision to allocate the scarce
unaware of the outcomes that will result from resource of attentiveness.
strategic choices, but they are uncertain of In governments, as well as in all organiza-
the procedures of choice themselves and are tions, attention is allocated in a process politi-
even uncertain about their own preferences. cal scientists call agenda setting. Organizations
Uncertain outcomes may produce dependence also suffer from limited attention spans, and
Peter-3383-Chapter-03.qxd 6/8/2006 7:32 PM Page 56
must process at least major problems serially. to actual human strategies in real world
Routine problems may be delegated and han- games?
dled according to rules, but fresh problems can-
not be handled this way. It is common for many
problems to press forward on the governmental FRAMING EFFECTS AND PUBLIC POLICY
agenda, and whether the most severe problems
receive the most attention is an empirical ques-
Studies linking framing effects present an espe-
tion. In individuals, as well as in organizations,
cially intriguing challenge to our understand-
the allocation of decision-making attention is
ing of rational political behavior. Framing
crucial for understanding the immediate and
studies have demonstrated that an individual’s
future behavior of an organization.
understanding of a policy idea depends heavily
Because it is explicitly oriented toward under-
on the context and heuristic cues surrounding
standing the processes of decision-making,
a policy problem (Garrett 2003; Jones 1994;
bounded rationality has been widely applied to
Kahneman and Tversky 2000). Rather than
the study of public policy and public adminis-
holding stable opinions about public policies,
tration. Bounded rationalists have explored a
individual opinions are volatile, and shift
range of policy problems, ranging from agenda
according to the framing of the political issue
setting (Kingdon 1995; Baumgartner and Jones
(Zaller 1992). The relationship between a voter
1993), congressional decision-making (Kingdon
and a policy idea is thus at least partially
1973), federal budgets (Padgett 1980), incre-
contingent on the context in which the voter
mentalism (Wildavsky 1964; Davis, Murray,
is presented the policy choice. By ‘directing
Dempster and Widalsky 1966) and risk aversion
attention to one attribute in a complex
(Kahneman and Tversky 2000). As a model of
problem space,’ policy entrepreneurs can pro-
choice, bounded rationality is at least as broadly
duce desired changes to public responses to
applicable as comprehensive rationality.
policy problems (Jones 1994, 104). Public
receptivity to public policies seems highly
responsive to shifts in the framing of a policy
PART II: BEHAVIORAL DECISION
problem.
THEORY AND THE SHORTCOMINGS
OF RATIONALITY
environmental context. People would hold the manner in which people receive and then
different preferences for virtually identical psychologically allocate the use of money has
decisions based on how the problem was pre- an important bearing on how the money is
sented. Kahneman explains: perceived and spent. ‘The effect on current
Preferences appeared to be determined by attitudes to consumption … of winning the $300 football
gains and losses, defined relative to a reference point, pool should be the same as having a stock in
but Bernoulli’s theory and its successors did not incor- which I own 100 shares increase by 3 dollars
porate a reference point. We therefore proposed an a share, or having the value of my pension
alternative theory of risk, in which the carriers of utility
increase by $300. The marginal propensity to
are gains and losses—changes of wealth rather than states
of wealth. Prospect theory (Kahneman and Tversky, 1979) consume all types of wealth is supposed to be
embraces the idea that preferences are reference depen- equal’ (Thaler 1992, 108–109).
dent. (461–462). Yet people do not approach the gain of $300
equally. The $300 won in lottery or betting
Approaching the study of risky choice from
is seen as a sudden windfall, and might be
the perspective of changes in wealth rather
spent with clear conscience on celebration. A
than overall states of wealth, Kahneman and
sudden $300 increase in the pension plan will
Tversky arrive at the intuitively appealing
be treated with miserly caution, an essential
notion that humans cope with gains and losses
part of a savings account linked to concerns for
differently—that people ‘are risk adverse in the
future financial stability. Although the amount
domain of gains, and risk seeking in the domain
is the same, people perceive it differently
of losses’ (Quattrone and Tversky 1988, 723).
because it was accounted for differently. Money
Prospect theory further demonstrates that
can thus be better thought of as placed in
people make different choices depending on
various mental accounts. Money appropriated
how decisions are framed—the weights they
for one use is spent or saved differently than
give in decision-making shift in response to
money appropriated for another (Thaler
how their attention is directed. ‘An individual
1992).
would prefer a sure gain of $80 over an 85%
Findings from prospect theory and mental
chance to win $100. The same individual
accounts are compelling because they provide
would prefer a gamble offering an 85%
empirical evidence that individual decision-
chance of losing $100 to a certain loss of $80’
making is strongly constrained by framing
(Berejikian 2002, 762.) Although the chance to
cues which illicit emotional responses. Unlike
win $100 is the same as the chance of losing the
the calculating maximizer posited by rational
$100, individuals generally make different
choice theory, ‘in real life, generating maxi-
choices. In the domain of gains people value
mum utility is neither simple nor smooth, and
certain gains over possible gains. In the
is affected by the cognitive and emotional con-
domain of losses, people will make riskier deci-
stitution of the decision maker’ (Jones 2001,
sions in the hopes of avoiding any loss. These
26). The strategic choices pursued by individu-
findings clearly contradict the tenets of ratio-
als may not reflect calculations of optimality,
nal choice theory, which would assume that a
but rather their disposition towards loss aver-
rational chooser would prefer certain gains or
sion or risk acceptance, or even their emo-
certain losses equally.
tional response to a sudden windfall.
In a similar vein, Richard Thaler’s research
on mental accounting further suggests that
individual preferences and expectations are
EXPERIMENTAL GAME THEORY
shaped by emotional cues. Basic economic
theory suggests that money is fungible, that
money ‘has no labels’ and is treated the same Unlike studies of framing effects, which examine
way regardless of how it is earned (Thaler how individual preferences shift in response
1992, 108). However, Richard Thaler draws a to external emotional cues, experimental
series of hypothetical scenarios to show that game theory began by asking if people behave
Peter-3383-Chapter-03.qxd 6/8/2006 7:32 PM Page 58
in a manner consistent with game theoretic maximizing behavior in individuals in the lab,
expectations of rational self-interest. To test but they also provide strong evidence that
accurately assumptions of economic choice, human decision-making cannot be fit neatly
individuals must be given clear guidelines and into economic assumptions of comprehensive
all relevant information with which to form rationality. Individual trading behavior in ulti-
strategic preferences. In spite of this conven- matum games has been shown to be highly
tion, the most popular experimental games context-specific. Repeated play games suggest
have provided mixed evidence at best that that learning, social interaction, and norms of
individuals behave in a manner consistent with reciprocity strongly shape ultimatum trading
game theoretic assumptions of strategic pref- (Nowak, Page and Sigmund 2000; Goeree and
erences. Players with full information about Holt 2000). A sense of entitlement seems to
the rules and objectives of the game frequently shape how a proposer or responder plays the
prove themselves to be inefficient in forming game. Players who earn the responsibility of
game theoretic strategies (Fudenberg and proposer engage in distinctly different trading
Levine 1997). behavior than those who are simply designated
By far the most intriguing findings have the role of proposer (Hoffman and Spitzer
come from research in the now famous ‘divide- 1985). Researchers have demonstrated the
the-dollar’ games, which captured the imagi- influences of culture (Roth et al. 1991; Henrich
nation of decision theorists exactly because 2000), and proposer and respondent charac-
they have failed to substantiate the most basic teristics such as sex (Solnick 2001) and age
hypothesis drawn from expected utility theory.1 (Murnigham and Saxon 1994) in ultimatum
In the simplest form of the game, the ultima- trading strategies. Post-game surveys of players
tum game, a proposer and a responder are suggest that norms of fairness, reciprocity,
asked to share a sum of money. The proposer cooperation, and generosity express them-
must divide the sum however he sees fit with selves strongly, even in single shot anonymous
the responder. Because neither player receives trading.
any money if the responder rejects the offer, The brunt of experimental research in
rational choice theory would lead us to predict divide-the-dollar games has remained nar-
that the smallest permissible offers would be rowly focused, first on revealing the patholo-
made and accepted. The proposer would ratio- gies of comprehensive rationality, and then in
nally seek to maximize the monetary gains revealing altruistic, fair or other ‘anomalous’
available in game play, while the responder behavior in research subjects. The discussion
would prefer even an incremental gain over no and conclusion sections of most ultimatum
gain at all. game research almost invariably speculate that
Yet mounting evidence suggests that indi- evolutionary norms of fairness, cooperation,
viduals rarely if ever play as pure income max- or rational risk aversion constrain responders
imizers. Proposers frequently offer up to half from making decidedly one-sided offers. This
of the sum being divided, while responders points to a common complaint of experimen-
routinely reject offers of under one-third of tal research—it is a powerful critique of one
the total—even when neither player knows the model of rationality, but so far the tradition
other, and often simply play against a com- has done little to generate and test new and sci-
puter (Nowak, Page and Sigmund 2000). Even entifically realistic alternatives of choice.
in the dictator form of the divide-the-dollar Experiments are fantastic devices for testing
game, where the responder is required to particular behavioral assumptions, but they
accept any offer made by the proposer, an are less useful in generating new ones.
impressive 76% of responders chose to divide Despite these limitations, efforts to replicate
the money equally (Kahneman, Knetsch and assumptions of comprehensive rationality have
Thaler 1986). succeeded in fueling a movement to integrate
Ultimatum game experiments not only rou- findings of actual behavior into a more realis-
tinely fail to reproduce purely rational utility tic model of human choice. Findings from
Peter-3383-Chapter-03.qxd 6/8/2006 7:32 PM Page 59
cognitive psychology, cultural anthropology, reality albeit at the sacrifice of a neater and more
political science, and behavioral economics, parsimonious model. Nonetheless, as the maximand
more closely approximates reality, the more reality it
demonstrate how considerations outside of should capture’ (Levi 1997, 24–25).
strategic rationality influence the procedures of
individual decision-making. Experts across dis- Individual preferences may be modified to
ciplines increasingly agree that a successful capture not only an individual’s desire to max-
model of human choice must integrate empiri- imize gains, but an equally strong ‘ethical’
cal findings about the effects of culture, attention belief that the individual’s behavior harms no
streams and issue salience, emotional cues, fram- other individual’s. These rational choice models
ing effects, and identity and social perspective integrate assumptions of fairness, altruism,
into a parsimonious model of decision-making. reciprocity, justice, etc., into the ordered pref-
This research agenda has led to renewed interest erences of their rational actor. The individual
in established theories of decision-making, such continues to hold ordered preferences, and
as bounded rationality and prospect theory, as continue to make decisions based on whatever
well as a wealth of new and modified theories will yield their most preferred outcome. The
from the rational choice tradition. problem, of course, is that trade-offs among
these various goals must be explicitly built into
the model, and one of the primary findings of
RATIONAL RESPONSE: THIN behavioral decision theory centers on the diffi-
RATIONALITY AND NEW culties people have with such trade-offs.
INSTITUTIONALISM Another hallmark of recent research in
rational choice theory has been to give greater
attention to the institutions in which political
Although a few dogmatic economists and
actors make choices. Institutional Rational
political scientists have continued to conduct
Choice Theory examines how institutions pro-
research through assumptions of comprehen-
vide specific information about the gains and
sive rationality, most recent rational choice
losses available to actors in a given sphere.
studies in political science have attempted to
Institutional rational choice theory is espe-
integrate empirical findings of individual pref-
cially concerned with understanding how
erences into renewed models of individual
organizational cultures permit cooperation in
rationality. It is clear that the traditional ratio-
places where selfish maximization might prove
nal choice assumption of self-interest has
problematic (Miller 1992; Ostrom 1998). In
failed—that the self-interest axiom cannot
many situations, rational maximization at the
explain instances of altruism, ethical restraint,
individual level leads to collective disaster—as
or fairness—has lead an increasing number of
in Garret Hardin’s (1968) tragedy of the com-
theorists to allow maximization for any goal.
mons, where unlimited common grazing
The analyst refuses to make assumptions about
rights lead to destruction of the common pas-
motives for maximization. These steps may
ture illustrates. Through learning or experi-
complicate the model or limit broad theoreti-
mentation, organizational rules may develop
cal generalizability; however, the growing evi-
that encourage collective outcomes that overcome
dence documenting inconsistencies between
the “tragedy” of individual self-maximization
rational choice assumptions and actual human
(Ostrom 1999a). Once in place, these rules are
behavior warrants a revision of the theory to
enforced through a system of formal and
include the potential rewards outside of mate-
informal incentives that control selfish behav-
rial or status gain. Margaret Levi points out:
ior. This framework provides ‘a general lan-
The addition of non-egoistic considerations or motiva- guage about how rules, physical and material
tional norms [as community standards or fairness prin-
conditions, and attributes of the community
ciples] does increase the complexity and difficulty of
analysis… . The advantage of an assumption in which affect the structure of action arenas, the incen-
actors consider net payoffs that include both material tives individuals face, and the resulting outcomes’
and ethical factors is that it may better approximate (Ostrom 1999b, 59).
Peter-3383-Chapter-03.qxd 6/8/2006 7:32 PM Page 60
These institutional and cultural extensions management; however we fear the privileged
are the most important additions to rational place such research now holds in the study of
choice theory to be put forward in recent years. organizations provides students with a simplis-
Yet, by incorporating so many “soft” variables tic and misleading view of the dynamics of
they move quite briskly away from the tradi- bureaucratic policy-making.
tional approach. How can one predict whether A more comprehensive understanding of
an organizational culture will emerge to over- bureaucratic decision-making can be gained
come rational self-interest? Clearly not always, through the lens of bounded rationality, which
and hence the institutional structure is depen- offers a scientifically sound alternative model
dent on some other context that is left unspec- of behavior while sacrificing little parsimony.
ified in the model. Rather than focusing exclusively on control,
bounded rationality focuses on how changes in
the external environment shape information
processing both within and across organiza-
PART III: PRINCIPAL AGENCY AND
tions. This approach to the study of organiza-
ORGANIZATIONAL BEHAVIOR
tions will invariably yield a more sophisticated
understanding of bureaucracy, as it is better
In spite of improvements, rational choice suited to link the procedures of human choice
models simply are insufficient to capture criti- with broader policy processes.
cal aspects of organizational behavior. Because
this research remains focused on outcomes
rather than processes, rational choice theories
POLICY IMPLEMENTATION AND
are overwhelmingly focused on problems of
AGENCY THEORY
collective action and individual utility maxi-
mization, which provide only limited insights
into organizational behavior. Bounded ratio- Rational choice decision-making as applied to
nality can examine the roots of collective the implementation stage of the policy process
action problems, but it also can shift research generally takes the form of principal-agent
attention toward the more interesting ques- models of the bureaucracy’s interactions with
tions of attention allocation and organiza- the legislative, executive, and judicial branches
tional information processing. of government. In this section we argue that
Recent studies in public administration have the underlying assumptions of the principal-
focused a good deal of attention on problems of agent model of politico-bureaucratic relations
delegation and control in organizational decision- render it inaccurate, not only in terms of ade-
making. Indeed, virtually all theoretical discus- quately explaining the relationship, but also
sions of policy implementation have centered as a description of the relationship and policy
on these questions to the exclusion of such processes more generally at the implementa-
important topics as adaptive problem-solving tion stage.
and information-processing. While principal- Principal-agent models of behavior were
agent dilemmas illuminate some important originally developed in the economic analysis
aspects of bureaucratic behavior, its promi- of firms (Alchian and Demsetz 1972);
nence in studies of public policy is partially an although an emphasis on efficient production
artifact of the rational choice model of behavior. and incentive structures in organizations dates
The fundamental problems of principal-agent much earlier (Taylor 1911). The principal- agent
dilemmas—those of information asymmetries, relationship is essentially a contract. In order
moral hazards, and incentive structures—are to increase production, the principal enters
those that map neatly onto the most basic into an agreement with the agent for the pro-
assumptions of individual utility maximization. duction of the good. The principal’s goal is to
Pioneering work in principal-agent models ensure the efficient production of the good;
have spurred a rich discussion in public meanwhile, the agent’s goal is to avoid time,
Peter-3383-Chapter-03.qxd 6/8/2006 7:32 PM Page 61
effort, and resources necessary for producing about the relationship of the actors to their
the good. goals. The principal-agent model assumes that
For example, if we were examining a busi- a given actor’s goals are known. Formal principal-
ness firm, we might consider the employer the agent modeling has rightly recognized that
principal and the employees as agents. In the there exists uncertainty about the goals of the
model, employees have incentive to “shirk”, that other actors in the relationship (Bendor and
is, loaf. This means that managers must closely Moe 1985; Bendor, Taylor and Van Gaalen
monitor them to insure efficient production of 1985; Bendor, Taylor and Van Gaalen 1987a;
a given good. No system of monitoring can Bendor, Taylor and Van Gaalen 1987b; Moe
completely eliminate all free-riding, because 1984), but this does not take account of the
the resources for such monitoring face the law uncertainty that shrouds an actor’s knowledge
of diminishing marginal returns. This is the of her own goals. Actors may not be able to
essence of the principal-agent problem. define their own goals, especially in contexts
Why do principals face the problem of free in which trade-offs must be made between
riding? The answer lies in information asymme- competing goals.
tries found in the relationship between the prin- The model also assumes that strategies for
cipal and the agent. In the principal-agent attaining given goals are easily attached to the
model, the agent holds two key informational goals. The principal’s goal of efficient produc-
advantages over the principal. The first infor- tion is attained by developing the appropriate
mational advantage lies in the selection process. incentive structure such that it is to the agent’s
Principals cannot know a priori the exact quali- material disadvantage to shirk rather than work.
fications of a given agent, while the agent has a Likewise, the agent’s goal is to maximize the
better idea of his or her competence. The agent benefit to be gained from the contractual rela-
may exaggerate her ability in order to attain the tionship with the principal through mani-
position, and even if most exaggerations are pulation or information. The final assumption
caught by the employer’s personnel policies, concerning goal oriented behavior is complete
some incompetents will slip through. This first information. The model assumes that if the
informational asymmetry is usually termed principal is able to observe agent outputs, the
adverse selection. The second informational principal is able to judge whether or not these
advantage held by the agent involves the moni- outputs conform to the goal of efficient produc-
toring of the agent after the contract has been tion (Worsham, Eisner and Ringquist 1997).
established. Moral hazard results when the prin- Finally, the principal-agent approach makes
cipal lacks the resources or ability to engage in certain assumptions about the nature of
constant monitoring of the agent. The agent is the principal-agent relationship over time. The
then able to shirk, reducing the efficiency of the principal-agent approach is based in compara-
firm. The task of the principal is to induce com- tive statics (Bendor and Moe 1985). Rela-
pliance on the part of the agent through a tionships change only when exogenous ‘shocks’
system of incentives. to a system disturb a system otherwise in equi-
In order to invoke the principal-agent librium. It should be noted that this is not
framework, the analyst must assume goal entirely a problem of principal-agent theory
incongruence. The goal of the principal is the per se. Comparative statics and equilibrium
production of some good, while the goal of the analysis also has roots in systems theories in
agent is to reap the rewards of the contract both political science and sociology (Eisner,
with the principal without expending the Worsham and Ringquist 1996; Worsham, Eisner
effort to produce the good. To put it simply, the and Ringquist 1997; Worsham, Rinquist and
goal of the agent is to shirk. If goal incongru- Eisner 1998).
ence does not exist, then principal-agent Given these assumptions, it is imperative to
models are inappropriate for examining the the success of the contractual relationship that
relationship. Underlying the assumption of the principal control the behavior of the agent.
goal incongruence is a further assumption There is no room for shared goals or learning
Peter-3383-Chapter-03.qxd 6/8/2006 7:32 PM Page 62
and adaptation of behavior over time, or the Moe (1984) argues that for principal-agent
potential counterproductive results on pro- theory to be useful in the study of implemen-
ductivity of an environment of continual dis- tation, we must take account of the fact that
trust and suspicion fostered by monitoring. politicians are not necessarily motivated by the
The theme of control has been central in sub- efficient production of public service; they may
sequent scholarship on the implementation be more concerned with political efficiency
stage of the policy process. rather than production efficiency. Moreover,
Rational choice approaches were used in the the major problem of control might not be
study of public policy long before the advent of shirking, but could involve several different
the now-dominant principal-agent approach, possibilities, including material benefit of
most notably in Niskanen’s (1968; 1971; 1975) some sort, ranging from budgetary slack to
model of the budget-maximizing bureaucrat and promotion, but might also be policy related. At
Anthony Downs’ (1967) study of information a minimum, scholars utilizing the principal-
limitations and bureaucracies. The principal- agent frameworks would need to grapple with
agent model really burst onto the scene in politi- the issues of political efficiency and diversity of
cal science with works by Mitnick (1980; see also goals—not trivial alterations.
Mitnick 1973) and Miller and Moe (1983a;
1983b). This culminated in an article by Moe
(1984), The New Economics of Organization,
EX ANTE CONTROL
which formally introduced principal-agent
OF IMPLEMENTATION
approaches to the study of policy implementa-
tion to political scientists. Moe insightfully uni-
fied principal-agency from the economic analyses With the formal introduction of principal-agent
of business firms to the issue of how democrati- theory, work on policy implementation col-
cally elected officials can control unelected public lapsed to analyses of how a political principal is
bureaucrats; hence principal-agency could be to control a shirking bureaucracy. One group of
used to illuminate a critical problem in democra- scholars argues that political control of the
tic theory. Unfortunately, it altered the perspec- bureaucracy is achieved ex ante (McCubbins
tive on bureaucracy from one of balancing 1985; McCubbins, Noll and Weingast 1989).
problem-solving, information-acquisition, and Democratically elected officials gain control of
the advantages and disadvantages of delegation, the bureaucracy through setting the ‘structure
to one stressing a single- minded emphasis on and process’ of organizations before their actual
formal procedures of monitoring and control. creation through detailed legislation specifying
It also reversed the old normative emphasis administrative procedure, personnel, and orga-
in public administration on keeping politics nizational structure (McCubbins, Noll and
out of administration. At least since Woodrow Weingast 1989; Banks and Weingast 1992). In this
Wilson (1887), political scientists have been view, Congress ‘stacks the deck’ against the agency
concerned with the distinction between poli- during its creation in order to ensure subservience
tics and administration, mostly from the per- to Congress or organized interests at later periods.
spective of trying to keep politicians from The deck-stacking thesis focuses on political
interfering with professional implementation efficiency—the political coalition in control
by demanding special favors from supposedly wishes to continue its advantage through later
neutral bureaucrats. Principal-agency suggests implementation. Other scholars argued that the
that all independent behavior of bureaucrats is political branches of government could also
not motivated by professional problem- achieve ex post control of the bureaucracy
solving, but is based in the desire to shirk. through budgeting, personnel, staffing, political
Principal-agency assumes that all attempts by appointments, ongoing interactions with the
politicians at control are motivated by “efficient bureau, and congressional oversight hearings
production” rather than crass political gains (Bendor and Moe 1985; Bendor, Taylor and Van
that add inefficiencies to public production. Gaalen 1987a; Miller and Moe 1983b).
Peter-3383-Chapter-03.qxd 6/8/2006 7:32 PM Page 63
But is congress the principal in the American is no room for politicians to prefer neutral
system of divided powers? Both congress and competence in administration—the major aim
the president attempt to influence the bureau- of the Progressive Movement so important
cracy (not to mention the judiciary). Some in American political life during the first quar-
scholars argue that congress is the institution ter of the 20th Century. To the extent that
wielding the most influence or control on the Progressives were successful (and in many
bureaucracy (McCubbins 1985; McCubbins, respects they were very successful), they left
Noll, and Weingast 1989; Weingast and Moran a legacy of removing political control and the
1983). Other scholars see the president, with his politicized administration that the deck-
ability to influence the bureaucracy through stacking thesis suggests.
budgets, OMB, and political appointments, as
the abler of the two branches to achieve control
and focus their theoretical and empirical atten-
DESCRIPTIVE CRITIQUES
tion on him (Golden 2000; Wood 1988; Wood
and Waterman 1991).
McCubbins, Noll, and Weingast (1989) put Beginning in the 1990s, scholars began to argue
forth three hypotheses concerning agency that descriptive flaws led agency theorists to
structure and process and its effects on the explanations of bureaucratic behavior that
subsequent regulatory environment. First, the were less than adequate (Eisner, Worsham, and
agency’s design should, ‘create a political envi- Ringquist 1996; Hindera and Young 1998;
ronment that mirrors the politics at the time of Krause 1996; Meier, Pennington, and Eller
enactment’ (McCubbins, Noll, and Weingast 2001; Potoski and Nemacheck 2001; Worsham,
1988, 444). That is to say, winners and losers at Eisner, and Ringquist 1997). Principal-agent
the formulation stage of the policy process analysis generally has difficulty accounting for
should benefit or lose in the same relative pro- the fragmented nature of policy processes in
portions at the implementation stage due to the American system of government. That is,
institutional design. Second, agency structure there are multiple, competing principals in the
and process will favor those constituencies of American system of separated powers (Choi
the winning coalition at the formulation stage. 2001). Moreover bureaucracies themselves
Finally, agencies will exhibit an ‘autopilot’ have constituencies and engage in advocacy
characteristic (McCubbins, Noll, and Weingast policy-wise for these constituencies (Hindera
1988, 444). As the preferences of the favored and Young 1998; Meier, Pennington, and Eller
constituency change, so will the preferences 2001; Potoski and Nemacheck 2001). These
and thus behavior of the agency. bureaucratic constituencies could also be con-
More recent scholarship is divided on the sidered the principals.
efficacy of structure and process as a means for All of these potential principals compete to
controlling the bureaucracy. Balla (1998) finds influence or control the bureaucracy. These
that the open comment procedures for the new competing principals often have multiple, con-
pay schedule and rules regarding Medicare did flicting goals, especially under divided govern-
not favor the groups presumably favored by ment. Once the political context of multiple,
the dominant coalition in congress. Balla and competing principals is taken into account,
Wright (2001) did find that advisory commis- bureaucratic intransigence becomes much
sions correctly represented the political fac- harder to pin down conceptually. To whom
tions involved in the legislation of drinking should the agency respond? Why should the
water, but these authors did not examine the agency respond to one principal to the exclusion
actions of these commissions. of others? We argue that the more appropriate
Note that in the development of the principal- question to ask is, how do bureaucracies make
agent literature, the descriptive fit of the trade-offs among the competing and contradic-
principal-agent model is somewhat strained. tory demands of their political environment?
For example, in the deck-stacking thesis there Answering this question requires that we take
Peter-3383-Chapter-03.qxd 6/8/2006 7:32 PM Page 64
an information processing approach to the How do principals know which strategies will
study of bureaucratic behavior. lead to the desired outcomes?
A second descriptive criticism of agency Furthermore, it was argued long before
theory involves the goals of the agent. The goals principal-agent theory came to dominate imple-
of the agency are assumed to be adverse to those mentation studies that success in terms of goals is
of the principal. Principal-agent modeling has difficult to measure (Lipsky 1980; Worsham,
difficulty addressing agents whose goals are not Eisner, and Ringquist 1997). How do principals
consistently adverse to those of the principal. know desirable outputs when they see them?
Recent research at the street level calls into ques- Monitoring outputs can be counterproductive as
tion the assumption that bureaucrats are pri- agents shift away from productive problem-
marily motivated by the goal of shirking solving (as well as shirking) toward the goal that
(Brehm and Gates 1997). Krause (1996) argues is monitored. This is a classic critique of stan-
that the development of goals over time is a dardized testing in schools as teachers “teach to
two-way process. Agencies influence and even the test”. The linkage between strategies and goals
help to refine the goals of political principals. cannot be understood outside a dynamic rela-
Second, using principal-agent models requires tionship. This suggests that bureaucratic out-
the researcher to acknowledge that both the comes, rather than outputs, are the more
principals’ and agents’ goals are known and important indicators of goal success, which, in
easily prioritized. Bender and Moe (1985) note turn, suggests the efficacy of a problem solving
that the relationship between the bureaucracy approach to implementation.
and its political principals is likely dynamic, and If bureaucrats’ interests are not necessarily
the goals of the actors are likely to evolve and are adverse to those of the principal, an emphasis
clarified over time, both by the interactions of on control as a description of the process of
the bureaucracy with the political principals implementation may be misleading. Golden
as well as through organizational factors. (2000) finds that bureaucrats in various regula-
Worsham, Eisner, and Ringquist (1997) have tory agencies in the Reagan administration for
argued that we cannot assume that bureaucratic the most part were responsive to the policy
agents are motivated to maximize material directions of the administration, even in the
return to themselves; they may have other absence of formal control procedures. Brehm
goals—some of which may correspond to those and Gates (1997) find that bureaucrats are not
of the policymaking branches. This raises fundamentally interested in shirking. They
several questions concerning the type of behav- argue that functional motivations and peers are
ior we should expect from the bureaucracy. the prime influences on intra-organizational
When confronted with a choice between behavior to the exclusion of influence by supe-
attaining one of many possible goals, how riors within the organization. Bureaucrats, for
does an actor choose? We are back to a serious the most part, work. Their research adds a
problem in all rational theories of choice: to strong empirical foundation to arguments advo-
make any progress, some very strong assump- cating taking organizational context seriously
tions must be made about what goals are used (Lipsky 1980; March and Simon 1958; Simon
in the process of maximization. A bounded 1947; Wilson 1989).
rational view of individual decision-making
takes account of the difficulties of making trade-
offs among multiple and competing goals.
PART IV: INFORMATION-PROCESSING
Principal-agent theories of policy processes
AND PUBLIC POLICY
also assume that the strategies, or means for
attaining goals, are easily attached to the goals
themselves. Although the studies outlined How are we to come to grips with the fact
above all note that goals are not easily that bureaucrats are often predisposed to be
achieved, the uncertainty that surrounds the responsive, but that policy outcomes are not
means to achieve the goals is rarely modeled. always exactly as political principals would
Peter-3383-Chapter-03.qxd 6/8/2006 7:32 PM Page 65
dynamic change in the policy outputs of relationships are not a caldron of conflict. The
organizations. In this regard, we find it curious environment is dynamic and conflict varies
that researchers in public administration are according to both the level of information that
calling to expand principal-agent models of principals and agents posses and the level of
bureaucratic behavior (Waterman and Meier goal conflict’ (197).
1998; Bendor, Glazer and Hammond 2001). Perhaps the most troubling set of problems
A scientifically sound and parsimonious identified by critics of principal-agent models
alternative approach connecting individual have been that they are founded in artificial
and organizational behavior already exists. To and even incorrect assumptions of human
demonstrate the value of bounded rationality behavior. Under empirical scrutiny, the assump-
in the study of organizations, we describe first tion of goal incongruence generally, and agent
how the model might be applied to principal- shirking specifically, is less pronounced than
agent dilemmas and to broader and more principal-agent theory would have us believe.
interesting questions about bureaucratic infor- In many organizations, agent behavior is deter-
mation processing. We contend that bounded mined more by the organizational culture and
rationality actually provides a more robust feelings of agent solidarity than through incen-
understanding of principal agency. More tive structures and principal monitoring
importantly, we argue that those researchers (Brehm and Gates 1993). The importance of
who study organizational behavior through institutional norms and cultures inspired Gary
the lens of bounded rationality look to a Miller’s Managerial Dilemmas, one of the most
greater and more appealing range of research exciting innovations in modern organization
questions than the problems of delegation and theory (Miller 1992). Miller’s work challenged
control. the Skinnerian notion that workers in agency
were most effectively induced to work through
incentives and punishment, arguing instead
that an organizational culture is far more effec-
BOUNDED RATIONALITY AND
tive in establishing worker compliance or non-
PRINCIPAL-AGENT MODELS
compliance with principal goals (Miller 1992).
Nothing in bounded rationality rejects that
Recent empirical challenges to the central incentives and monitoring within organiza-
assumptions of comprehensive rationality have tions can shape individual behavior—and
spurred calls for reform in either the method or Miller’s interest in organizational culture is
the behavioral model employed in studies of entirely consistent with procedural rational-
interpersonal relationships. In their elegant ity’s interest in the influence of external factors
review Theories of Delegation, Bendor, Glazer, in individual choice.
and Hamond (2001) conclude with a challenge Yet bounded rationality is not merely as good
to improve the study of principal-agent dilem- as new institutionalism in explaining principal-
mas, complaining in part that too little attention agent relationships—it performs better. This
is given to a) the ‘real world’ institutional con- superiority stems from the considerable atten-
text of principal-agent dilemmas (266–267); tion bounded rationality places in the cogni-
b) pre-existing information and policy pro- tive constraints on decision-making. New
grams that might constrain an actor’s behavior; institutionalism is an enormous improve-
and c) the complexity of the task environment ment over traditional principal-agent studies
facing actors in a given environment (267). because it built such exogenous factors as
Waterman and Meier (1998) attack the limita- culture into the decision-making model.
tions of standard principal-agent dilemmas Unfortunately it adds, in an ad hoc way, what-
from another direction, deriving an interesting ever the analyst thinks actors are “maximizing”,
set of hypotheses by relaxing assumptions and generally ignores the trade-offs implied
of information asymmetries and goal incon- between the standard assumption of self-interest
gruence. They argue ‘all political-bureaucratic and the other imputed goals that might get
Peter-3383-Chapter-03.qxd 6/8/2006 7:32 PM Page 67
maximized. Find a deficit in the theory, and a preferences; however, bounded rationality
new goal is added. distinguishes itself by capturing psychological
Bounded rationality takes but one addi- as well as environmental constraints on choice.
tional step by showing how biological hard- Although inductive research is more time
wiring and psychological traits shape how consuming, such an approach will ultimately
individuals act within organizations, including yield a more compelling portrayal of principal-
the complex relationship between incentives agent dilemmas. In bounded rationality, indi-
and response. Having moved so rapidly away vidual responses to monitoring, incentive
from simple models of self-interest maximiza- structures, routines, and organizational mis-
tion, and having recognized that “informal sion are no longer abstracted from artificial
norms” such as a sense of fairness can be models, they are grounded in the science of
maximized, why do rational choice theorists observation.
not simply take the next small step and intro-
duce the individual and organizational cogni-
tive architectures that bounded rationality
BOUNDED RATIONALITY AND
emphasizes?
POLICY OUTPUTS
Bruno Frey’s research on ‘the crowding out
effect’ (Frey 1993; Osterloh and Frey 2000) is
an excellent example of how bounded ratio- While bounded rationality may well improve
nality might improve models of principal- research in principal-agent dilemmas, the thrust
agent relationships, and it shows the practical of this argument is not to rescue principal-agent
danger of believing that agents simply respond theory. Instead, we hold that starting with the
to incentives provided by the principal. assumption of bounded rationality will stimu-
Drawing from psychological studies that show late a wide range of interesting research ques-
individuals respond negatively to control, Frey tions in organizational information processing,
argues that ‘under readily identifiable condi- which will ultimately provide us with a far more
tions, increased monitoring reduces an agents robust understanding of bureaucratic behavior.
overall work effort’ (1993, 663). Increased This final section looks to the unique emphasis
monitoring breaks the ‘psychological contract’ of bounded rationality on how individuals and
with the agent, lowering his or her incentive to organizations receive, process, and act upon a
respond to incentives or punishment. Not only variety of information cues.
is increased monitoring not terribly effective, The promise of bounded rationality is that it
but it might prove harmful to organizational allows researchers to connect individual
productivity—a blunder based on theory behavior to organizational output without the
rather than analysis. Frey then compares rigid and misleading constraints of rational
behavior across organizations to identify choice. For the study of public policy processes,
which types of organizations will produce such a model is enormously appealing. It
a positive (increased productivity) or a nega- allows researchers to explore political behavior
tive (decreased productivity) by increasing in the aggregate by making assumptions about
monitoring—a much more scientific and the behavior of the individual.
inductive approach than the standard deduc- The earliest advances in organizational pro-
tive stance taken by principal-agent theorists. cessing models emerged from the study of
By looking to the psychological and envi- public budgeting (Simon 1947; Wildavsky
ronmental constraints on individual choice, 1964; Davis, Murray, Dempster and Wilasky
bounded rationality allows researchers to cap- 1966); see Thompson and Green 2001 for a
ture realistic and counter-intuitive behavior critical review). Bounded rationality was used
that escapes rational choice analysis. New insti- to explain incremental changes in public bud-
tutionalism and bounded rationality have both gets. The high costs of generating alternatives
discovered the importance of organizational and making fully optimal decisions, coupled
rules and norms in delineating individual with the uncertainty of outcomes, led to what
Peter-3383-Chapter-03.qxd 6/8/2006 7:32 PM Page 68
might be best characterized as risk averse Low salience issues will generally follow an
policy incrementalism. At the core of many incremental pattern—little new information
explanations of budget incrementalism was or alternatives will be integrated into the solu-
Simon’s notion of satisficing. To limit the costs tion searches. When attention is focused on a
of making decisions, both individuals and high salience policy problem, we anticipate a
organizations looked to alternatives that were broader solution search, greater generation of
good enough. Budget incrementalism emerged alternatives, and increased legislative activity.
from the risk averse convention of referring to However, even when attention is squarely
the prior-year’s budget when creating a new focused on a problem, decision-making will
one. ‘Because the reference point for decision- be bounded by cognitive and emotional
making is always some point decision made in constraints.
the past, the outcomes of local search, com-
bined with a propensity to limit bargaining
and enforcement costs, are small or incremen-
TOWARD AN INFORMATION-PROCESSING
tal changes in policy’ (Thompson and Green
MODEL OF POLICY MAKING
2001: 3).
While early models of organizational infor-
mation processing focused on the politics of In The Politics of Attention, Bryan Jones and
incrementalism, more recent studies of Frank Baumgartner (2005) outline the compo-
bounded rationality have focused on the nents of a comprehensive model connecting
dynamics of policy change. Behavioral models individual and organizational behavior. This
here have been driven by an intriguing empir- model is drawn from bounded rationality, and
ical reality—neither budgets specifically (Padgett attempts to capture both the dynamics of indi-
1980; Carpenter 1996) nor policy agendas more vidual and collective choice at each stage of the
generally (Baumgartner and Jones 1993) policy process.
adhere to the strict pattern of incrementalism. Figure 3.1 depicts their information-
Instead, budgets and agendas follow the trajec- processing model. Notice at each stage the
tory of punctuated equilibriums—periods authors have characterized the decision-making
of incrementalism followed by a sudden flurry process at the individual and the systems level,
of policy activity (Baumgartner and Jones because organizations fall prey to the same
1993). These recent studies have improved our kinds of cognitive limits as do individuals
understanding of the policy processes by inte- (Jones 2001).
grating the role of shifting attention into At the core of the model is the problem of
models of organizational behavior (Jones issue attention. Organizational agendas reflect
2003). Students of how governments set policy what actors believe to be the most salient or
agendas emphasize the role of the allocation of pressing concerns. Because both organiza-
attention in determining the behavior of tional and individual attention is limited,
policy makers (Cohen, March, and Olsen 1972; agenda setting necessitates that organizations
Kingdon 1995; Baumgartner and Jones 1993). prioritize political problems, focusing first and
Shifts in the external environment produce foremost on the most urgent concerns. These
shifts in the preferences and goals of policy- concerns may be dictated less by informed rea-
makers. When an institution or organization’s soning and more by emotional responses to
attention is squarely focused on a specific political problems, ‘as emotions are critical in
policy domain, we might anticipate increased determining priorities’ (Jones 2001: 73–74). In
activity, legislation, and spending. However, organizations, the high priority issues will
because the allocation of attention is limited to receive the brunt of attention, while less salient
a few select issues, many policy programs con- concerns will fall to the wayside. If these less
tinue incrementally—and follow the conserva- salient issues demand political action, they
tive, risk averse path outlined by prior research may follow the path of incrementalism and
in public budgeting. emphasis on ‘pre-packaged’ solutions.
Peter-3383-Chapter-03.qxd 6/8/2006 7:32 PM Page 69
Which attributes
Problem characterization Problem definition
to incorporate?
Which solutions
Alternative generation Proposals and debate
to examine?
Choice
The next step in the information-processing The final two stages of the model describe
model takes the form of problem characteriza- how individuals and institutions reduce a
tion and organizational problem definition. problem to a manageable set of alternatives, and
Received information is rarely neutral— then select a policy solution. For example—
individuals and organizations must decide how policy proscriptions addressing failing schools
to characterize and approach a particular might take any number of forms—smaller class
problem. Understanding how individuals char- sizes, increased funding, increased oversight,
acterize and organizations define the nature of a school and student accountability, curriculum
political task is essential in the policy process. overhaul, etc. Because decision-makers struggle
For example, failing students in public educa- at evaluating and making trade-offs (Sniderman
tion might be seen primarily as a problem of et al., 1991; Tetlock 2000; Jones 2001), choices
waning parental involvement, or it might be must be reduced to a few actionable alternatives.
seen as a systemic failure of the school system. In organizations, this is often done through
The received information may be characterized debate. Failing schools can be solved through
in any number of ways, depending on the pre- increasing funding, or through increasing
conceived attitudes of the decision-maker, or accountability and oversight. The choice is then
the larger organizational culture. made according to institutional procedures.
Peter-3383-Chapter-03.qxd 6/8/2006 7:32 PM Page 70
Balla, Steven J., and John R. Wright. 2001. Interest of SEC. Paper read at The Annual Meeting of the
Groups, Advisory Committees, and Congressional American Political Science Association, August
Control of the Bureaucracy. American Journal of 30–September 2, 2001, at San Francisco, CA.
Political Science 45 (4): 799–812. Cohen, Michael, James G. March, and Johan P.
Banks, Jeffrey S., and Barry R. Weingast. 1992. Olsen. 1972. A Garbage Can Theory of Organi-
The Political Control of Bureaucracies Under zational Choice. Administrative Science Quarterly
Assymetric Information. American Journal of 17: 1–25.
Political Science 36 (2): 509–524. Davis, Otto A., Murray A. H. Dempster, and Aaron
Baumgartner, Frank R. and Bryan D. Jones. 1993. Wildavsky. 1966. A Theory of Budget Process.
Agendas and Instability in American Politics. American Political Science Review 60: 529–547.
Chicago: University of Chicago Press. Downs, Anthony. 1957. An Economic Theory of
Bendor, Jonathan, and Terry M. Moe. 1985. An Democracy. New York: Harper.
Adaptive Model of Bureaucratic Politics. The Downs, Anthony. 1967. Inside Bureaucracy. Edited
American Political Science Review 79 (3): 755–774. by A. R. C. R. Study. New York: Harper-Collins.
Bendor, Jonathan, Serge Taylor and Roland Van Eisner, Marc, Jeff Worsham, and Evan Ringquist. 1996.
Gaalen. 1985. Bureaucratic Expertise Versus Crossing the Organizational Void: The Limits of
Legislative Authority: A Model of Deception and Agency Theory in the Analysis of Regulatory
Monitoring in Budgeting. The American Political Control. Governance (October).
Science Review 79 (4): 1041–1060. Frey, Bruno. 1993. ‘Does Monitoring Increase Work
Bendor, Jonathan, Serge Taylor, and Roland Van Effort? The Rivalry with Trust and Loyalty.’
Gaalen. 1987a. Politicians, Bureaucrats, and Economic Inquiry 31: 663–670.
Asymmetric Information. American Journal of Friedman, Milton. 1953. ‘The Methodology of
Political Science 31 (4): 796–828. Positive Economics.’ In Milton Friedman, Essays
Bendor, Jonathan, Serge Taylor, and Roland Van in Positive Economics. Chicago: The University of
Gaalen. 1987b. Stacking the Deck: Bureaucratic Chicago Press.
Missions and Policy Design. The American Fudenberg, Drew and David Levine. 1997. ‘Measur-
Political Science Review 81 (3): 873–896. ing players’ losses in experimental games.’
Bendor, Jonathan, Glazer, A, and T. Hammond. Quarterly Journal of Economics 112 (2): 507–536.
2001. ‘Theories of Delegation. In Nelson Polsby, Garrett, Elizabeth. ‘The Law and Economics of
ed., Annual Review of Political Science. Palo Alto, Elections: The Law and Economics of ‘Informed
CA: Annual Reviews. Voter’ Ballot Notations’ (Symposium). Virginia
Berejekian, Jeffrey D. 2002. ‘Model Building with Law Review 85 (November 2003): 1553.
Prospect Theory: A Cognitive Approach to Inter- Gigerenzer, Gerd, Peter Todd, and the ABC Research
national Relations.’ Political Psychology 23: Group. 1999. Simple HeuristicsThat Make Us
759–786. Smart. Oxford University Press.
Brehm, John, and Scott Gates. 1993. ‘Donut Shops and Goeree, Jacob K. and Charles Holt. 2000. ‘Asymetric
Speed Traps: Evaluating Models of Supervision on inequality aversion and noisy behavior in
Police Behavior.’ American Journal of Political Science alternating-offer bargaining games,’ European
37: 555–81. Economic Review 44: 1079–1089.
Brehm, John, and Scott Gates. 1997. Working, Golden, Marissa Martino. 2000. What Motivates
Shirking, and Sabotage. Ann Arbor: University of Bureaucrats? Politics and Administration During
Michigan Press. the Reagan Years. New York: Columbia University
Brehm, John, and Scott Gates. 1997. Working, Press.
Shirking, and Sabotage: Bureaucratic Response to a Hardin, Garrett. 1968. ‘The Tragedy of the
Democratic Public. Ann Arbor: University of Commons.’ Science 162: 1243–1248.
Michigan Press. Henrich, Joseph. 2000. ‘Does Culture Matter in
Camerer, Colin and Richard Thaler. 1995. ‘Anomalies: Economic Behavior? Ultimatum Game Bargaining
Ultimatums, Dictators and Manners.’ Journal of among the Machiguenga of the Peruvian Amazon.’
Economic Perspectives 9: 209–219. The American Economic Review 90: 973–980.
Carpenter, Daniel P. 1996. Adaptive Signal Hindera, John J., and Cheryl D. Young. 1998.
Processing, Hierarchy, and Budgetary Control in Representative Bureaucracy: The Theoretical
Federal Regulation. The American Political Science Implications of Statistical Interaction. Political
Review 90 (2): 283–302. Science Quarterly 51 (3): 655–671.
Choi, Jin-Wook. 2001. Rethinking Political Control Hoffman, Elizabeth, and Mathew L. Spitzer. 1985.
Over Independent Regulatory Agencies: The Case ‘Entitlements, Rights and Fairness: An Experimental
Peter-3383-Chapter-03.qxd 6/8/2006 7:32 PM Page 72
Mitnick, Barry M. 1980. The Political Economy of Awards. Paper read at The Annual Meeting of the
Regulation. New York: Columbia University Press. American Political Science Association, August
Moe, Terry M. 1984. The New Economics of 30–September 2, 2001, at San Francisco, CA.
Organization. American Journal of Political Quattrone, George and Amos Tversky. 1988.
Science 28 (4): 739–777. ‘Contrasting Rational and Psychological Analyses
Monroe, Kristen R. 1996. The Heart of Altruism: of Political Choice,’ American Political Science
Perceptions of a Common Humanity. Princeton Review 82: 719–736.
University Press. Rockman, Bert. 2000. ‘Theory and Influence in the
Muringhan, Keith J. and Michael S. Saxon. 1994. Study of Bureaucracy: Micro- and Neoinstitutional
‘Ultimatum Bargaining by Children and Adults,’ Foundations of Choice.’ Journal of Public
unpublished working paper, University of British Administration Research and Theory 11 (1): 3–27.
Columbia Faculty of Commerce. (Cited in Thaler Roth, Alvn, Vesna Prasknikar, Masahiro Okuno-
1992). Fujiwara and Shmuel Zamir. 1991. ‘Bargaining
Muringhan, Keith J. and Madan Pillutla. 1995. and Market Behavior in Jerusalem Ljubljana,
‘Being Fair or Appearing Fair: Strategic Behavior Pittsburgh, and Tokyo: An experimental study.’
in Ultimatum Bargaining.’ Academy of Manage- American Economic Review 81: 1068–95.
ment Journal 38: 1408–1427. Simon, Herbert A. 1947. Administrative Behavior: A
Newell, Allen. 1990. Unified Theories of Cognition. Study of Decision-Making Processes in Adminis-
Cambridge, MA: Harvard University Press. trative Organizations. 4th ed. New York: Simon
Niskanen, William A. 1968. The Peculiar Economics and Schuster, Inc.
of Bureaucracy. Papers and Proceedings of the Simon, Herbert A. 1985. ‘Human Nature in Politics:
Eightieth Annuel Meeting of the American Economic The Dialogue of Psychology with Political
Association 58 (2): 293–305. Science.’ The American Political Science Review
Niskanen, William A. 1971. Bureaucracy and Repre- 79: 293–304.
sentative Government. Chicago: Aldine-Atherton, Simon, Herbert A. 1991. Models of My Life.
Inc. New York: Basic Books.
Niskanen, William A. 1975. Bureaucrats and Simon H.A. 1999. The potlatch between political
Politicians. Journal of Law and Economics 18: science and economics. In Competition and
617–644. Cooperation: Conversations with Nobelistsabout
Nowak, Martin A., Karen Page and Karl Sigmund. Economics and Political Science, J Alt, M Levi,
2000. ‘Fairness versus Reason in the Ultimatum E Ostrom, eds. Cambridge, UK: Cambridge
Game’ Science 289: 1773. University Press.
Olson, Mancur. 1982. The Rise and Decline of Slovak, Paul. 1990. Choice. In Thinking: An invita-
Nations. Yale University Press. tion to cognitive science, vol. 3. Daniel Osherson
Osterloh, Margeret and Bruno Frey. 2000. ‘Motivation, and Edward E. Smith, eds. Cambridge, Mass: MIT
Knowledge Transfer, and Organizational Forms.’ Press.
Organization Science 11: 538–550. Sniderman, Paul, et al. 1991. Reasoning and Choice.
Ostrom, Elinor. 1998. ‘A Behavioral Approach to the Cambridge: Cambridge University Press.
Rational Choice Theory of Collective Action’. Solnick, Sara. 2001. ‘Gender Differences in the
American Political Science Review 92: 1–22. Ultimatum Game.’ Economic Inquiry 39: 189.
Ostrom, Elinor. 1999a. ‘Coping with Tragedies of the Taylor, Frederick W. 1911. The Principles of Scientific
Commons’. In Nelson Polsby, ed., Annual Review of Management. New York: W. W. Norton.
Political Science 2. Palo Alto, CA: Annual Reviews. Tetlock, Philip. 2000. ‘Coping with Trade-offs:
Ostrom, Elinor. 1999b. ‘Institutional Rational Choice: Psychological Constraints and Political Impli-
An Assessment of the Institutional Analysis and cations’. In Lupia, Arthur, Mathew D. McCubbins,
Development Framework.’ In Theories of the and Samuel L. Popkin, eds., Elements of Reason.
Policy Process, Paul A. Sabatier, ed. Boulder, CO: Cambridge: Cambridge University Press.
Westview. Thaler, Richard H. 1992. The Winner’s Curse;
Padgett, John F. 1980. Bounded Rationality in Paradoxes and Anomalies of Economic Life.
Budgetary Research. American Political Science Princeton University Press.
Review 74: 354–372. Thompson, Fred and Mark Green. 2001.
Potoski, Matthew, and Christine L. Nemacheck. Organizational Process Models of Budgeting,
2001. Bureaucratic Influence in the Legislative Research in Public Administration, 55–81.
Arena: 1984 Drug Price Competition-Patent True James L., 2000. ‘Avalanches and Incre-
Restoration Act and FDA Distributive Program mentalism: Making Policy and Budgets in the
Peter-3383-Chapter-03.qxd 6/8/2006 7:32 PM Page 74
United States.’ American Review of Public Wood, B. Dan. 1988. Principals, Bureaucrats, and
Administration 30: 3–18. Responsiveness in Clean Air Enforcements.
Waterman, Richard and Kenneth Meier. 1998. American Political Science Review 82: 213–234.
Principal-agent Models: An Expansion? Journal Wood, B. Dan, and Richard W. Waterman. 1991. The
of Public Administration Research and Theory Dynamics of Political Control of the Bureaucracy.
2: 173–202. The American Political Science Review 85 (3):
Weingast, Barry R., and Mark J. Moran. 1983. 801–828.
Bureaucratic Discretion or Congressional Control?: Worsham, Jeff, Marc Allen Eisner, and Evan J.
Regulatory Policymaking by the Federal Trade Com- Ringquist. 1997. Assessing the Assumptions: A
mission. Journal of Political Economy 91: 765–800. Critical Analysis of Agency Theory. Administration
Wildavsky, Aaron. 1964. The Politics of the Budgetary and Society 28 (4): 419–440.
Process. Boston: Little, Brown. Worsham, Jeff, Evan Rinquist, and Marc Eisner, eds.
Wilson, James Q. 1989. Bureaucracy: What Govern- 1998. A Theory of Political Influence of the
ment Agencies Do Why They Do It. New York: Bureaucracy. Edited by J. D. White, Research in
Basic Books, Inc. Public Administration. Stamford, CT: JAI Press.
Wilson, Woodrow. 1887. The Study of Adminis- Zaller, John R. 1992. The Nature and Origins of Mass
tration. Political Science Quarterly 2: 197–222. Opinion. Cambridge University Press.
Peter-3383-Chapter-04.qxd 6/8/2006 7:32 PM Page 75
4
Policy Design: Ubiquitous,
Necessary and Difficult
D AV I S B . B O B R O W
Further, the instances of work explicitly both reasons is the notion that engineering,
labeled policy design are far outnumbered by unlike architecture, denies the possibilities for
those in which designs are prescribed or evalu- unwanted responses by those who must go along
ated without bearing that label in a prominent for the design to materialize and work as envi-
and explicit way. With or without a policy sioned by its engineers. Such an assumption of
design title, policy assessments by lay people a lack of voice, exit options, resistance strate-
and “policy wonks” applaud or bemoan the gies, and bargaining power is of course more
consequences attributed to such efforts. That is often than not extremely unrealistic (Stewart
no less the case for international concerns than and Ayres, 2001).
for the dominant focus of the literature, quint- Under either metaphor, designs are represen-
essential domestic policy. We are, of course, tations of sets of choices and instructions to select
well into a world in which designs which and apply materials sufficient to produce an
assume that consequences, causes, and policy intended result. ‘Designers put things together
options can be neatly segregated by national and bring new things into being, dealing in the
borders are doomed to disappointment. process with many variables and constraints’
Subsequent sections of this chapter address: (Schon, 1987: 42). Architecture of course draws
policy design as a messy imperative; root con- on available scientific knowledge, but does so in
ceptions of policy design; subsequent elabora- the pursuit of normative standards and makes
tions and differentiations of its conception; and use of a lot of intuitive craft and creativity. Policy
core strategies for its conduct – a suggested ten designers do so as well, and the quality of their
commandments. endeavors does not depend only on the technical
Before turning to those matters, it seems robustness of the “policy sciences” (Lerner and
helpful to recognize the core metaphors for Lasswell, 1951; Dror, 1971).
policy design, and the core activities of policy A grossly over-simplified skeleton which a
designers they imply. As an initial orientation stereotypic designer of policies qua engineer
for subsequent modification, central metaphors might flesh out in a representation appears in
are those of architectural drawings or engineer- Figure 4.1. Note it’s linear, mechanical, and
ing blueprints, i.e., the products of the design non-reflective appearance.
sciences (Perlmutter, 1965; Alexander, 1964; A policy design is stipulated in the expecta-
Alexander, 1982). Policy designs are representa- tion that adopting it will produce particular
tions of what might be turned into realities. The patterns of human or organizational activity
analogy with architecture and architects should and chains of consequences. Individually, or in
be taken seriously, if only because it appears combination, they will suffice to achieve cer-
again and again (e.g., Schon, 1983; Weimer, tain predicted desirable changes in targets
1992: 136; Meier and Smith, 1994: 440). (intermediate products) which as a set will
Architectural metaphors have in many quar- provide the necessary and sufficient conditions
ters gained favor over the engineering ones for for a valued outcome.
several reasons. One is the view of ‘social engi- This skeleton of course raises more questions
neering’ as inappropriately and inherently a than it answers. Why bother? Can we flesh it out
matter of arrogant command by those in adequately if at all? What ingredients go into a
control of positions of power, be they in policy? How do we specify a valued outcome?
Washington, Moscow, or Brussels, i.e., anti- Who are relevant persons, organizations, and
democratic. A second is the not necessarily collectivities? What are the pertinent human
warranted view of it as a failure in practice, and organizational actions and socially relevant
e.g., some constructions of what has happened chains of consequences? What determines
with efforts to engineer a welfare state in the them? What policies will have what effects on
U.S. and elsewhere. Social engineers are not as those determinants? What are the targets to be
wise or smart as they and those who employ achieved? What human actions and chains will
them might like to think about picking do that? What determines if those in a position
winners and rejecting losers. Related to to decide to pick a particular representation will
Peter-3383-Chapter-04.qxd 6/8/2006 7:32 PM Page 77
Policy design
choose a particular one? Even if they do, will the prospective occurrence of unwelcome eventu-
sketch or blueprint be followed as building or alities, and the failure to materialize of attrac-
implementation proceed? What body of theory, tive ones we expected. We look to ruling elites
strategies, techniques, procedures and processes to design or at least adopt designs for a desir-
will enable us to answer these and other ques- able future and invent or at least implement
tions in ways which are consistent, ethically ways of bringing it about. We may well reward
appropriate, and pragmatically feasible? There and punish them accordingly within our
are then prima-facie grounds for doubt about understanding of their responsibilities and
the normative and pragmatic status of the policy possibilities. On their part, ruling elites
approach in the strawman of Figure 4.1. Those usually try to anticipate what will trigger par-
are far from sufficient grounds for turning one’s ticular punishments and rewards. The market
back on policy design. for ways to pursue preferred or at least tolera-
ble futures never closes with the attendant
incentives for purveyors to enter and stay in
that market. Applied social scientists have long
POLICY DESIGN AS A MESSY IMPERATIVE
been among those purveyors, or at least staffed
them and tried to influence them and their
We often are far from indifferent about our per- customers.
sonal and collective futures as individuals, When those ways involve more or less
members of social groupings, and citizens of explicit ensembles of material and non-
political entities. We act, and delegate to others material ingredients, context conditions, and
the responsibility to act and induce us to act, processes, we can think of them as designs.
in ways which hold normative and pragmatic They can appropriately be thought of as public
promise in our eyes. That promise lies in policy designs when they involve government
forestalling, or limiting the damage potentially units or officials acting directly, or as a target
posed by, distressing outcomes and achieving or of or motivator for actions by others. In a
at least improving the odds of appealing ones. common-language sense, designs are recom-
We tend to hold those in positions of collec- mended plans, but ones not at all necessarily
tive responsibility accountable for the actual or involving dirigiste central planning.
Peter-3383-Chapter-04.qxd 6/8/2006 7:32 PM Page 78
Public policy design is ubiquitous since “way”, and never been labeled policy design. In
organized social and political action always the contemporary world, the difference is
involves ingredients, always has a context, always massive broadening in the range of others and
contains processes, and in modern society phenomena whose impacts affect us, accelera-
always involves some aspect of governance. That tion in the speed with which they do so, and near
is no less true for those who believe in a benign instant information about and media reporting
or unavoidable, “invisible hand” economic or of ostensibly outcome relevant developments.
theological. For them, policy designing is still Those secular trends increase the incentives
warranted if only to let the invisible hand come to engage in policy design while making the
closer to functioning at its ostensible best rather endeavor more difficult. The difficulties
than its imaginable worst. Policy design involve: 1) greater complexity in terms of the
attempts are not the monopoly of those who number and variety of actors and factors;
buy into any one particular notion of the role of 2) rapid changes raising the risks of designs
the state or embrace any particular view of being rendered obsolete or seriously flawed
desirable and just distributions of power, before they are finished; and 3) extreme vul-
wealth, status or rectitude. nerability to a loss of support due to declines
Policy design then claims our attention if two in issue priority and “from the get-go” expo-
conditions are met. The first is that we are not sure to organized opposition.
indifferent to ‘who gets what, when, and how’ The feasibility of relatively attractive paths
(Lasswell, 1961), a range of concern which into the future seems then to depend in increas-
includes processes, and aggregate and distribu- ingly important ways on interdependencies with
tional outcomes. The second is that we acknowl- and sensitivities to others and on contextual
edge the existence of more than one possible factors. From this perspective, policy design is
positively or negatively valued state of affairs in necessary for timely anticipation of and accom-
some slice of the future, rather than granting modation to what we cannot control, as well as
inevitability to a particular prospect (Dror, 1986). for making the most of what we can control. A
Defeating prophets of inevitability, e.g., about ship’s captain must steer in the face of expected
inequality, may be facilitated by showing that a or unexpected bad weather just as he or she must
present state of affairs was not itself inevitable but in a benign sea. The wise captain, naval architect
rather results from a set of previous actions, be and engineer tries to have built in resilience
they ones of commission and omission (e.g., to shocking circumstances imaginable and
Fischer et al., 1996). Those actions could follow unknowable (Demchak, 2002). Policy design
from an integrated, purposeful plan (deliberate tries to help “captains” and citizens by providing
intent) or merely follow from unintended second courses of action and material aids which can
or third order effects from foreseeable or unfore- be used and retain their efficacy in ‘turbulent
seeable relevant events. environments’ (Ackoff, 1979).
In sum, when both conditions seem to be met, Policy design also is necessary for anticipa-
the responsible course becomes to try to ‘tilt the tory policy analysis. Policy analysis broadly
odds’ toward more preferable outcomes. conceived is fundamentally the “scoring” of
Preferences can be for a previously unexperi- possible behaviors in terms of value criteria.
enced future, maintenance of the status quo, or Policy analysts are in search of at least one set
even resurrection of an ostensibly attractive past. of behaviors which seems likely to be at least
They, alternatively, can be to forestall some envi- satisfactory or ideally optimal. That quest is
sioned future, erode or shatter the status quo, or hostage to: the possible behaviors we envision;
forestall reversion to some past. Policy elites and the foresight with which we specify sequences
publics have then little choice but to frame and and quantities and qualities of behaviors; and
choose among alternative policy designs. That is the extent to which we encompass the range of
hardly new, even if it may in pre-industrial soci- considerations and priorities of outcomes
eties have emphasized an accommodation to and their associated valuations. Anticipatory
spirits and nature, as with the Navajo Indian policy analysis can be no better than the policy
Peter-3383-Chapter-04.qxd 6/8/2006 7:32 PM Page 79
designs it considers. Unfortunately, the necessity and intuition, determines what behaviors on
of policy design may often only be the mother the field are desirable, and tries to prepare and
of invention and not of success. Like even the persuade players (as agents) to make them.
best batters in baseball or strikers in soccer, the Like even the wisest coach, a policy designer
harsh statistical reality is that the vast majority has to live with the possibility of the actual
of the time our best efforts will not achieve a players putting advice into action well, poorly,
hit or a goal. Even when they do, that is no or not at all. The designer can, however, be
guarantee of future successful performance by more or less competent at fitting strategies
repeating the previously successful moves and techniques to the available players, the
(counter to Chung, 1993). nature of the opposition, field conditions, and
Many of the inherent difficulties are well the time remaining in a game, series, or season.
known – the lack of stable and fully revealed And, of course, like changing coaches, firing
goal preferences; limited understanding of and replacing designers may be a more com-
causality; at best partial knowledge of the mon response to disappointment than chang-
future; counter-measures deliberate and unin- ing those who adopt or implement policy
tentional of unanticipated vigor, skill and type; designs.
the presence of residual designs, institutional Second, policy design is often a competitive
interests, legacies, and memories rather than a sport. Experts compete for the chance to
clean slate; and hard to fathom complexity. design policies. If they get it, their designs are
Policy design is a “game of failure” and disap- pitted against alternatives. When designs go
pointment, and thus not for the easily discour- into the white water of the policy process they
aged or those who do not find the act of are placed in the hands of individuals who may
participation inherently gratifying compared well be competing for personal success, and of
to only being a spectator. teams (e.g., political parties and bureaus) who
That is not to say that technique and strat- compete to improve or at least do as well for
egy do not matter for achievement, that all fail- themselves as they have before. Ultimately,
ures must be repeated, or that an unsuccessful designs in practice and the political entities
specific effort cannot set the stage for subse- who have put them more or less into practice
quent relative success. At least in the long-run, are competing against benchmarks used by
the odds for both positive achievements and relevant audiences.
avoiding failures can be improved, if only at The policy designer has reason to take into
the margin. Since policy design is a game that account the benchmarking criteria at work and
will be played, it makes sense to seek ways to their interpretation, while recognizing his or
play more effectively. What those are and the her limited (but perhaps not non-existent)
capacity to put them into use will of course capacity to stipulate them. Flying in the face of
vary. Some strategies and techniques may have prevalent benchmarks and interpretations
merit almost generically and others only for increases the chances that designs will be
particular situational factors and available ignored, rejected, or distorted in implementa-
resources. An implication is that policy design- tion. Even if fulfilled, a design may be deprived
ers need to be adept at generically useful skills of the status of a model for emulation in other
and situational diagnosis, command a broad jurisdictions or time periods. In effect a policy
repertoire to draw on for particular situations, design provides a promise, albeit a hedged one.
and be realistic about available resources. The promise competes with held images of the
Some final prefatory implications of a team past, present, and future to elicit “gambles” or
sport metaphor are worth noting as they bear bets from elites and general populations.
on policy design. First, while the designer is a Further, there is no guarantee that: a) the
player at the stage of suggesting what to do, he design teams and their employers or sponsors
or she is most of the time not in a position to are all playing the same game, or b) if they
adopt or implement policy designs. The role is are, accept the same rules, or c) if they do,
more that of a coach who, through analysis accept the same measures and standards of
Peter-3383-Chapter-04.qxd 6/8/2006 7:32 PM Page 80
performance. This lack of homogeneity can conditions … an economy must possess certain
lead to “design wars” between experts rooted in basic characteristics … and some government
different disciplines, ideologies, and intellec- activity.’ A government must have at hand
tual traditions (see Bobrow and Dryzek, 1987; ‘a minimum of instruments of economic
Linder and Peters, 1992). It can also lead in the policy … and these must be properly used.’
world outside the design studio, think-tank, or Because the goals or values sought are never in
policy planning bureau to profound disap- reality singular but always multiple (a multi-
pointments, surprises, and conflicts. attribute utility function), the instruments
More encouragingly, a policy design and the should be multiple as well.
activities associated with designing need not Beyond those basics, design is informed by an
move fully from paper or hard drive into prac- ‘awareness of … [goal related] … potentialities
tice, or fully achieve their promise, to provide and advantages’ derived from descriptions of
benefits. Efforts at designing or applying a the past and present, and projections. It also
policy can raise performance norms for policy should be informed by awareness of ‘varying
content and process. Those experiences can circumstances. Depending on circumstances
enlighten us about what does not work as some elements of … policy will require more
desired, and focus us on vulnerabilities for emphasis and attention or will appear more or
future amelioration. More arguably, they can less promising than others.’ Further, the action
leave satisfaction with the benefits of the elements of designs (‘programs’ and their com-
endeavor (“a game well-played”) even if one’s ponent ‘projects’) ‘by their very nature have to
favorite loses and thus maintain or increase be guesses and must be revised periodically.’
enthusiasm for trying again. If policy design is (Tinbergen, 1958: 4–7). In sum, Tinbergen
ubiquitous and necessary, it makes sense to called attention to the starting points of uni-
think of it as being played not only in a long versal requirements or necessary conditions,
season, but for many seasons. government policy instruments as means,
recognition of alternative possible outcomes,
situational diagnosis, uncertainty, and sufficient
flexibility for revision in light of experience.
CONCEPTIONS OF POLICY DESIGN
While Tinbergen did emphasize economic
matters and governmental instruments, he
More than four decades ago, Jan Tinbergen made it quite clear that a far broader set of
(1958) addressed policy design. His specific actors and factors should be considered in
concern was with development, which can be policy design. These include the private sector,
thought of more broadly as progress toward a general population attitudes, the administra-
preferred but yet to be realized collective future. tive organs of government with their varied
His perspective continues to echo in subsequent proclivities and capabilities, current economic
conceptions of the motivation for and funda- structure, natural resources, geographic situa-
mental characteristics of policy design. He tion, particularly strong personalities, the size
advocated a turn to design as a replacement for of the entity involved, and its sensitivity to
‘decisions … taken on the basis of vague ideas of behavior external to it. Policy design then is the
general progress, and often somewhat haphaz- ‘solution of a jig-saw puzzle of considerable
ardly … a process of trial and error … [marked complexity.’ (Tinbergen, 1958: 35)
by] … setbacks and crises, and probably a good In light of these considerations, the challenges
deal of misplaced energy and effort …’ to the policy designer are to devise a set of steps
(Tinbergen, 1958: 3). Policy design effort fol- which together make up a ‘coherent and coordi-
lows from dissatisfaction with a record of nated whole.’ That, in turn, calls for avoiding the
behavior including its efficiency. pitfalls of sole reliance on quantitative data, wish-
Motivated by that dissatisfaction, the policy ful thinking about available resources, omission
designer builds on awareness that ultimate, of interactive effects, and inattentiveness to time-
broad goals require achieving ‘certain general lags and sequences of action. Policy designs are
Peter-3383-Chapter-04.qxd 6/8/2006 7:32 PM Page 81
working drafts to be scrutinized and modified, The Tinbergian perspective surely can
based on eclectic information, internal consis- encompass many more recently offered concep-
tency, completeness, and the direct and primary tions. Herbert Simon’s view of design as the
as well as indirect and secondary consequences process of ‘changing existing situations into pre-
of their constituent programs and projects ferred ones’ (Simon, 1972: 55) becomes com-
(Tinbergen, 1958: 9–33). Given the importance patible when one invokes his conception of
of what will not be known with certainty, the satisficing. If a current situation, and current
responsible designer will consider offering prospects for the future, are satisfactory there
‘alternative programs,’ each with a somewhat need be only limited interest in design. If
different base of assumptions. Failures will still they are not, designers need quest after only
result because of factors such as the ignorance of what would make them satisfactory, and not
politicians, pressure from vested interests, minis- after what would make them optimum.
terial rivalries, incentives offered by legislatures, Reconciliation for inclusion is not even required
and external interference. for conceptions such as those of Linder and
Policy designers should be modest about the Peters (1988) ‘the purposive or goal-directed
‘role to be played by scientific knowledge and rearrangement of a problem’s manipulable fea-
insight … The relevant facts of life are too many tures’; May’s (1981) ‘identification and mani-
and too varied to make it possible to reach pulation of key design variables to create viable
decisions without a strong intuitive feeling for alternatives’; or Wildavsky’s art of finding solu-
human relations’ (Tinbergen, 1958: 68–69). tions to policy problems that specify desirable
Policy designers engage more in a craft than a relationships between ‘manipulable means and
science. Policy design a la Tinbergen stands in attainable objectives’ (Wildavsky, 1979: 15–16 in
contrast to ‘an interventionist perspective, Weimer, 1993: 111). Tinbergen’s recognition of
requiring a precise forecast of events so an situational variety is a harbinger of the empha-
external hand can intervene to assure meeting sis on it in ‘case-wise’ prescriptions for policy
of the exogenously determined targets’ (Saeed, analysis and design (Bruner, 1986, 1983).
1994: 11). Nevertheless, Tinbergen’s perspective, wise as
it was and is, leaves open a wide and difficult
range of issues about the practice of policy
design. Some of these are about steps in the
ELABORATIONS AND DIFFERENTIATIONS
creation of a meritorious design. Others are
about what its contents should be. Still others are
Since Tinbergen’s volume appeared, much has about the intellectual sources of or approaches
been written in several intellectual communi- to providing those contents. Taken to exclusive
ties about what policy design should and can extremes, the different positions in these debates
be. In large measure, these contributions amount to different conceptions of policy
amount to elaborating and differentiating ele- design (for comparisons of some of the major
ments explicit or implicit in his perspective. contenders, see Bobrow and Dryzek, 1987;
Elaboration and differentiation do not Linder and Peters, 1992). Yet we need not, and
necessarily carry with them implications of should not, treat them as either/or choices.
universal superiority, or exclusive, most quin- Continued contention about what aspects of
tessential status. They can instead suggest the the policy design elephant to emphasize is then
opposite – a broadening of the scope of policy about claims to shares in a mix, rather than
design rather than a narrowing. After Schon’s something to be resolved by giving the mix a
‘reflective practitioner’ who engages in a con- more homogenous, less multi-faceted character.
versation with a design situation (Schon, 1983, Those claims or suggestions may contribute to
1987), the elaborations and differentiations an ‘armamentarium’ of ‘frames with which to
may as a whole suggest an enriched and more envisage’ coherent designs and ‘tools with
varied, rather than a more circumscribed and which to impose … [designer’s] … images on
narrowly focused conversation. situations’ (Schon, 1987: 218).
Peter-3383-Chapter-04.qxd 6/8/2006 7:32 PM Page 82
This position, applied to political entities and and negative outcomes (issues of representation)
not just economic ones and to designs for out- and the quantity and distribution of a varity of
comes not limited to economic ones, adopts benefits (goods) received and costs (bads)
and generalizes the counsel of Adelman and borne (e.g., for indigenous knowledge as intel-
Taylor (1986: 84) ‘Ultimately, appropriate lectual property in Norchi, 2000, or for wealth
forms … need to evolve which can be adapted creation and public health and welfare in
flexibly to economies and societies at vastly dif- Daniere and Takahashi, 1999).
ferent levels of development, having different From another angle, problem definition
objectives and constraints, operating with differ- involves a grasp of causality. What brings
ent information and power structures, and hav- about particular human or organizational
ing differing social norms and values. The actions and chains of consequences? Why do
evolution of such planning technologies requires they produce particular intermediate products
melding a historical and socio-political perspec- or performance on target phenomena? Why do
tive with the tools of operations research in a particular sets of products and performances
much more flexible and synergistic manner …’ produce one or another outcome? With weak
Major elaborations and differentiations and partial answers to these questions, we
modify, almost always by complicating, the simply cannot know with confidence answers
visualization introduced in Figure 4.1. I will to a host of very practical questions. What
first consider two major sorts of elaborations needs to be changed? By how much and how
which are by no means incompatible with each quickly? What will induce those changes with-
other. One proliferates elements of a policy out triggering responses which countervail
design, and thus tasks to be dealt with in the them? Consideration of causality in policy
design process. A second provides maxims, design also raises issues about factors external
wise guidance, for the undertaking of those to a particular policy system which can affect
tasks. Neither sort is the particular property of its behavior, the triggers of the bad weather or
a single technical discipline. Each introduces turbulence to which we referred earlier.
“cascades” of questions to be dealt with. Context altering events and developments
may be unanticipated in their entirety or mag-
nitude. We may be unable to specify their fre-
quency, time of onset, and duration. Consider
ADDITIONAL TASKS
the impacts of the collapse of the Soviet Union
on the foreign policy designs of most other
Some of the additional tasks involve problem nations, of a tide of refugees on welfare deliv-
definition (DeLeon, 1994; Schneider and Ingram, ery systems, of assassination or atrocity on
1989). What are we proposing the design to rem- inter-group relations, or of an agricultural
edy? A wrong or incomplete answer will lead to blight on food provision. The “exogeneous”
designs which are social failures, either because cannot be relied on to remain helpfully walled
they deal with a remedy for a different malady off. Yet some possibilities are surely imagin-
altogether or yield an outcome which the osten- able, and thus can be accommodated in a
sible beneficiaries view more as a threat or harm robust design. Others are not, but such possi-
than as a remedy or benefit. bilities can be accommodated at least in part
From this angle, problem definition involves by building in slack and redundancy, even at a
value clarification and identifying dissatisfac- sacrifice of efficiency. Still, other changes can
tions or as yet unmet or endangered desires as only be handled by a quick and thorough
goals (Lasswell, 1971). These are certainly reversion to a reconsideration of problem
matters of moral philosophy – the warranted redefinition (after Demchak, 2002).
claims of the poor, the rich, nationals, foreigners, From a third angle, problem definition
the living, future generations. They call for involves a reasonably accurate description of a
judgments about whose fate should have how current situation, including its probably irre-
much standing in defining what are positive versible trends. That may seem rather mundane,
Peter-3383-Chapter-04.qxd 6/8/2006 7:32 PM Page 83
but descriptive reality is often not well known or Yet neglect of it flies in the face of the reality of
a matter of consensus. For major policy the attention to presentation and public and
matters, domestic and international, it more media relations which take a large part of the
often is contested ground. And it is fought over time of policy-makers and takers, and of those
or obscured in large measure because our who would influence them.
understandings of it affect what even the best I suggest that adoption, implementation, and
processes of value clarification and causality interpretation usually should be integral parts
determination will find or have found. This of policy design. An architect adapts his design
aspect of problem definition can take a lot of to the proclivities of those who must agree to
hard work. Reflect for a moment on the diffi- buy it, use it, and build it – while trying to influ-
culties of: listing the relevant shapers and shak- ence their notions of what they want and
ers of foreign trade conduct in particular can have. This caveated position accepts that
industries and flows as more than general cate- on occasion it may be useful to formulate a
gories; determining the percentage of welfare design simply as a utopian, aspirations raising
recipients who genuinely would prefer to have endeavor. Far more frequently, adoption or
jobs under varying conditions; estimating the implementation or interpretation failure are
current national costs of mental illness let alone fates to be avoided. What is in, and interpreted
its incidence; or assessing the closeness in time to be in, a Figure 4.1 type design will affect the
(imminence) of any particular country having chances of adoption, and the fidelity of imple-
ready-to-use weapons of mass destruction (as mentation in spirit and letter. That does not
with Iraq, Iran and North Korea). If we don’t argue for taking the design course with the most
know where we are, how can we determine a popular “face” and established odds of adoption
course to move from there to someplace else or and implementation. Doing so could vitiate the
even to maintain the status quo? reason for undertaking the design effort in the
To compound matters further, the problem first place – to improve what would otherwise
may not be an absence of good designs in the be outcomes. It does call for steps in design
compass of Figure 4.1 but lie with some of the which will serve several purposes.
factors and actors which pose themselves The first is to give a design thought to be com-
between the design as representation and the mendable in terms of valued outcomes the best
design in practice. Such obstacles involve possible chance of being realized. The second is
adoption and implementation (Bardach, 1977; to help focus design efforts on critical obstacles
Pressman and Wildavsky, 1973) and interpre- to preferred outcomes. The ‘bottleneck’ obstacles
tation (Edelman, 1977, 1971). Will the design may lie with the processes of and incentives at
be adopted in full or in part by authorized work for adoption and implementation and pre-
authorities and processes and sustained in that dominant interpretations. In that case, adoption
status for its intended life? Even if it is, will and implementation institutions and symbolic
those persons, organizations, and social group- costumes are important parts of policy designs.
ings, who must act to make the adoption more If those in being are problematic, design efforts
than a piece of paper, do so in ways the design should re-invent them. The third is to filter alter-
assumes? How will the policy be portrayed native proto-designs thought to hold equivalent
along the way from design to adoption to promise for preferred outcomes. Adoption and
implementation to a verdict on its results? At implementation feasibility and interpretative
least by now few policy experts would deny the ease can provide a reasoned basis for selecting
importance of adoption and implementation. the design to be developed in detail, to be trans-
The pertinent issues for a conception of policy formed from a sketch into a blueprint. The
design are whether those matters are part of it, fourth is that what goes on in adoption and
and if so what role they should play in it. implementation will affect interpretation of the
Interpretation is less discussed with relation to differences made in beneficial results and in
policy design, and even abhorrent to some valued outcomes. We often care about processes
given its connotations of “spin” and deception. and not just outcomes.
Peter-3383-Chapter-04.qxd 6/8/2006 7:32 PM Page 84
Policy design
Interpretation
Policy adoption
Yes No
Interpretation
Interpretation
Interpretation
Interpretation
indirect effects
We then arrive at a visualization more like with reason to be rather sanguine about what
Figure 4.2. will result while we wait for the research to pan
The spaghetti in this picture contrasts with out. Another possibility is to “do” policy design
the neatness of Figure 4.1. Rather than linear in ways which take deficits in problem defini-
movement, we have chronic recursiveness to tion; and setting into account, a “yellow-light”
one or more previous tasks. Rather than unfil- position. That position does not gainsay the
tered impacts on subsequent stages we have need for better foundations for problem defini-
intervening interpretative processes, often tion, indeed it may amount to advocacy for
highly contested about their main content. efforts to “hold the fort” until those improve-
This template is a far more realistic skeleton of ments are available and provide openings to take
what goes on in and around public policy sys- advantage of them as they become available.
tems than that of Figure 4.1. Even so, making Tinbergen of course took the “proceed with
the visualization directly useful for specific caution” position. The wisdom to which we
policies, located specifically as to place and now turn elaborates his position and has
time, involves unpacking the elements and the implications for designing as an activity. It
flows and filters between them to an extent emphasizes exploration, making provision for
that defies representation in a single picture. “post-launch” adjustments or even “mission”
To return to the architectural analogy, consider cancellation, and hedging.
how a sketch of a building on a single page Exploratory activities include interactions
proliferates into increasingly detailed, volumi- and pre-negotiations with the parties who will
nous representations of parts of it. “live” in the design (e.g., users, clients, stakehold-
ers, adopters, implementors, target populations)
and best judgment impact statements about
policy consequences. Explorations include
GUIDING MAXIMS
attempts to understand the dynamics of the cur-
rent system and its participants. An example is
We can readily agree on the design situation most the ‘reference mode’ in systems dynamics mod-
conducive to sound problem definition and thus els capturing the ‘important feedback loops
to high confidence policy design. The fortunate existing between the … elements in the system
policy designer would have a well-filled physical that create the particular time variant patterns’
or electronic shelf. It would contain reason- already present (Saeed, 1994: 22). Others feature
ably clear specifications of values, based on probes to illuminate how the parties will behave
widely accepted patterns of actor representation. if placed into one or another designed future
It would provide strong causal theories and (“what if” games, exercises, simulations and
empirically established sensitivities to legislation, even “trial balloons”) under a range of possible
administrative guidance, judicial decisions, conditions, and with what results (e.g., labor
resource inputs, organizational reforms, and market measures in response to changes in tech-
public relations campaigns – together with their nology and external competition, as in Ho, 2000,
temporal leads and lags. It would make available or welfare reform financial incentives as in
detailed, timely and accurate profiles of current Robins, Michalopoulos, and Pan, 2001).
states of affairs. Dream on. Still others, as circumstances permit, include
Part of the policy design problematique deals “trying out” all or parts of the policy design
then with what to do in the substantial absence (prototypes in pilot studies) to determine its
of such attractive resources. One possibility performance (field experiments) or examining
defers policy design altogether, since it amounts situations where it in effect has been tried out
to laboring on “bound to fail” (or only to suc- (natural experiments). (See the classic work of
ceed by chance) illusions. This “further research Campbell, 1969; and Cook and Campbell, 1979,
is needed” position stands aside for an indeter- and recent examples with respect to rural
minate period of time from the pursuit of development programs in Fox and Gershman,
valued collective outcomes. It is tolerable only 2000, welfare to work programs, as in Loeb
Peter-3383-Chapter-04.qxd 6/8/2006 7:32 PM Page 86
and Corcoran, 2001 and Bloom et al., 2001, poses. Policy designers are usually invested at
occupational safety, as in Murphy-Greene and least psychologically in their products. So too are
Leip, 2002, and home financing to check urban those who adopted them and have gained rents
sprawl, as in Blackman and Krupnick, 2001.) by a prominent role in implementing them. So
Even if systematic try-out data are not available, too are the beneficiaries of the intermediate
a more informal scrutiny by experts knowledge- products. Getting any or all of them to respond
able about previous design efforts may be constructively to adjustment signals takes fore-
undertaken (e.g., for international economic thought about how to hold up to them persua-
sanctions, as in Miljkovic 2002, or the “murder sive prospects of reward and punishment.
boards” to which intelligence estimates and Adjustment is more likely when designs are
weapons designs are sometimes subjected). pursued on the basis of a widely shared
Provisions for adjustment bear on avoiding premise that adjustment is their normal and
locked-in commitment to all or part of a appropriate fate, given imperfect forecasting
design, and the supply of and responsiveness to capacities (as argued in Aaron, 2000). It is less
“back-talk from the situation” as the design is likely when doing so invokes an attribution of
followed. The former amounts to recommend- incompetence or evil. In the first situation,
ing flexibility in the design, keeping options adjustments amount to ‘upgrades’ or ‘product
open for choice as the design unfolds in prac- improvements’ which usefully take into
tice. That can be done by delegation in place of account experiences and changes in opportu-
specification, for example, granting adminis- nities. In the second, they are confessions of
trative discretion. It can be done by building defects which call into question the basic merit
conditionalities into the design to be triggered of the design itself. Such confessions simply
by specified states of affairs. It can be done by provide an enemy with weapons.
structuring the design in terms of diversified Of course, the ultimate in avoiding excessive
“stages of choices,” each one of which is selec- commitments is not to make commitments at
tive while maintaining more than one alterna- all, or only the most modest and limited ones.
tive (competitors) to be selected from at a later Extreme incrementalism is the ultimate hedge.
point when uncertainty reducing additional It argues for narrow and thin designs as they
information will be available. involve changes from what will otherwise be the
Back-talk (feedback) provision at its sim- case. The chances of disappointment are less if
plest involves designing in transparency one tries to change too little, than if one tries to
through such “sunshine” devices as public change too much. One is less likely to propound
access to information, whistle-blower protec- designs which will worsen matters. The chances
tions, and reporting and review requirements. of designs being realized in practice benefit
At another and more demanding level, it ben- from “going with the flow” rather than challeng-
efits from the design containing explicit ‘flags’ ing most prevailing norms and standard operat-
relevant to a need for adjustment. These may ing procedures – or the incentives which
be milestones (what should have happened by support them. The counsel of incrementalism
a certain point as the design was put into prac- also implies placing policy designs on a “short
tice) or indicators confirming or disconfirm- leash” (e.g., through such mechanisms as sunset
ing that the design is performing as intended. provisions, or one-year funding, or institutional
It often is desirable and necessary to design in “checks and balances”). Go through the yellow
organizational cells empowered to collect and light very slowly and, after that, consider what to
process such information.1 do at the next intersection.
Yet even if designs have ample provision for An incremental approach surely can be part
the supply of feedback, other considerations of a “learning strategy” but in that case it too
merit attention to increase the chances of requires some of the same information provid-
responsiveness to it. These are more matters of ing and responsiveness encouraging features as
the design management situation than of the the adjustment emphasis discussed previously
design itself and the incentives that situation and blurs into it.
Peter-3383-Chapter-04.qxd 6/8/2006 7:32 PM Page 87
Another form of hedging is ‘compromise’ difficult to reverse or require more time and
(Arrow, 1986). The design can still be broad effort to do so. Opponents can mobilize and
and deep, but does not try to optimize in terms engage in blocking actions. Stances of adjust-
of intermediate target products or valued ment and hedging limit the incentives to the
outcomes. Designers ‘respect the uncertainty parties to bet strongly on a particular design,
of their basic premises’ by trying to ‘strike a or to accept it more than superficially. While
balance between meeting all contingencies, such stances do avoid the discrediting risks of
with great waste of resources for the contin- unfulfilled grand promises, they also work
gencies that do not arise, and meeting none, against the sorts of appeals that can mobilize
with the penalties that follow. How the balance enthusiastic, intense support.
between safety and daring is struck depends on
the probabilities of the different possibilities,
the costs of providing against them, and the
PRIMACY FOR PROCESSES
penalties for failure to meet them’ (Arrow,
AND PROCEDURES
1986: 162, 165–66).
These maxims all have strong arguments in
their favor. Unfortunately, they also tend to have Other than further research, there are three non-
several drawbacks. First, they increase the tasks fatalistic responses to the challenges of policy
involved with policy design and thus the design a la Tinbergen introduced previously.
resources and time it takes. That is reasonably One is of course to proceed with due humility,
self-evident for the exploratory and adjustment convinced that there is no other socially respon-
providing features and for hedging through sible course and with modest hopes about
compromise. Less self-evidently, it is also improvement in the tools available to us. A
true for incrementalism. That road leads to end- second, which does not buy into inevitable
less, short, and tight deadline policy design policy failure, is to drop a “pretentious” rhetoric
cycles. Each one involves a full set of tasks, as in of design in favor of the “old-fashioned way” to
Figure 4.2. The parties to policies in practice deal with policy problems. Get a bunch of smart,
have ample reason to act in their own sweet motivated, and knowledgeable people together
time as their bargaining power increases as and hope that they will be creative, politically
deadlines approach. The proliferation of design agile, and lucky. The problem is that even the
tasks, and increases in the resources and time “best and the brightest” can turn out to be wrong
they consume, raise the chances of policy design or lack the mandate for effective agency. A third,
fatigue and lessen the energies and speed of design with a fundamental difference, is that to
actually moving out to realize a design. which we now turn.
Second, for reasons mentioned earlier, it is Design with a difference does not directly
not obvious that many of the envisioned tasks address policies as choices for classic content
are any more capable of successful completion areas (e.g., social welfare, public health, eco-
than those discussed with reference to pro- nomic growth, defense). What it designs
blem definition. Consider the determinations instead are processes and procedures for
involved in Arrow’s compromising. Nor is it any generic purposes of problem definition, value
clearer than it was for problem definition as to clarification, and policy adoption, implemen-
when we will have done these tasks well enough tation, and interpretation. Normative posi-
to give us warranted confidence in proceeding. tions and pragmatic judgments come together
Third, delay in making major policy design to make process and procedure design the road
commitments has its own risks (as with U.S. to better policy performance. While oversim-
health insurance reforms in Oliver, 1999, or the plified, it is helpful to distinguish two strands –
“Oslo process” intended to resolve the Israeli– institutionalists and, after Linder and Peters
Palestinian conflict). Windows of opportu- (1992), ‘deliberationists.’
nity can close. Situations can deteriorate and Pragmatically, both are impressed by the dif-
unwanted ones can become either more ficulties noted in meeting the demands of
Peter-3383-Chapter-04.qxd 6/8/2006 7:32 PM Page 88
Figure 4.2, the requisite “matching” of policy (Weimer, 1992). Those reforms ostensibly will
design elements to current and altered charac- shift distributions of power and information
teristics of the world in which the design is to so that the relevant parties behave better. That
be put into practice. They are pessimistic about will, however, only happen if the incentives
our ever having the control or foresight to do cross a hard to discern strength threshold and
so well enough. Normatively, institutionalists include clear signals about their future applica-
and deliberationists see the need for protec- tion, as with pro-environment technology
tions against exploitation of the roles of efforts by firms in Norberg-Bohm, 1999.
problem definer, policy designer, adopter, Deliberationists focus more on the quality
implementer, and interpreter by special inter- of the discourse from which policy emerges
ests. Where they differ is in how to deal with than on what in particular emerges. To “get the
those pragmatic and normative problems. discourse right”, some recommend ‘explicit
Institutionalists, after Douglas North standards of discourse’ (MacRae, 1988;
(1990: 3), attempt to prescribe the ‘rules of the Anderson, 1987). Others concentrate on par-
game’ designs include and the mechanisms to ticular ‘models of deliberation’ (Fischer, 1981;
make those rules operational, that is, institu- Paris and Reynolds, 1983; Linder and Peters,
tional designs. One example is the formation 1992: 226–7). The former focus on what ought
of partnerships to contain and resolve what to be discussed; the latter on who will partici-
would otherwise be policy, program, and pro- pate and their freedom to raise matters they
ject paralyzing conflicts, as with the World care about. The latter places in the hands of the
Commission on Dams (Brinkerhoff, 2002). parties who will be involved and affected the
A more general example is that of the turn in tasks of value and option formation, context
development thinking toward governance to appraisal and choice, rather than stipulating
emphasize actions to bring about ‘accountability, them in the design. That empowerment osten-
a conducive legal framework, and transparency’ sibly will bring to bear creativity, information,
(World Bank, 1991: ii) drawing on general prin- diagnostic skills, value clarifications, and a
ciples about conducive conditions and necessary sense of ownership which a temporally or
properties of each. Accountability, it is claimed, situationally distant professional elite of policy
has a better chance when procedures and designers cannot muster. Interactions between
processes make for continuing competition, e.g., the parties along process and procedure lines
through full or partial privatization. Participatory provided by a designer will itself make for
opportunities for affected publics to articulate ‘matching’ and is likely to do so better.
their interests are helped by designing in NGO These recipes seem very different from those
roles and decentralizing responsibility to more of Tinbergen and his descendants. Yet three
proximate bureaucracies and elected office- questions remain to be answered. Can institu-
holders. A legal framework is as it should be tionalist and deliberationist approaches be
when ‘rules are known in advance … [and are] married with his in ways which compliment
… actually in force, mechanisms … ensure appli- each others’ strengths and compensate for each
cation of the rules, conflicts … [are] … resolved others’ weaknesses? Would institutionalist and
through binding decisions of an independent deliberationist recommendations necessarily
judicial body or through arbitration, there are work as envisioned or are some additional
known procedures for amending the rules when properties required? Is making those proper-
they no longer serve their purpose.’ A design ties available and getting the procedures to be
which provides for these conditions will be more realized into practice similar to some of the
likely to ‘prevent predatory government actions challenges posed in Figure 4.2?
and agent rent-seeking’ (World Bank, 1991: iii). The answer to the first question is largely
How to get institutions to behave as desired positive. For institutional design, that is implied
fundamentally amounts to taking advantage of by our previous discussion of adoption, imple-
the incentives they already have and ones pro- mentation and interpretation. Tinbergen after
vided by “reforms” in procedures and processes all recognized the importance of the targets of
Peter-3383-Chapter-04.qxd 6/8/2006 7:32 PM Page 89
institutional design. Yet he also recognized that discourse, the investment they are willing to
policy satisfaction also includes policy results make in it, and how binding the understandings
in terms of particular tangible and psychic reached will be. One can imagine cases in which
benefits and costs and not just processes and the participants would be diverse and represen-
procedures. Most assessments of policy perfor- tative, have substantial if not identical commu-
mance find it to be the joint product of both nicative skills, give high priority to the dialogues,
policies and institutions involved with them and have substantial influence with those they
(e.g., Keiser and Meier, 1996). Satisfaction can represent (as illustrated in Walters, Aydelotte,
be denied by getting institutions wrong (Meier and Miller, 2000). One can also imagine cases of
and Smith, 1994), but achieving satisfaction the opposite (as illustrated in Pelletier et al.,
takes more than getting them right. Even if, for 1999). Relevant parties are advertently or inad-
example, one can make implementing bureau- vertently left out. Some participants, for all sorts
cracies quite compliant in a Weberian fashion, of reasons, are highly effective communicators
we may still find abhorrent or stupid the policy and others are “bumps on the log.” Attendance is
they are complying with. a pro-forma or watchdog activity rather than a
Deliberationists’ communicative suggestions constructive engagement. Those who go to the
could certainly contribute to an improved grasp meeting are the people their home group neither
of the values at work, defining policy problems, trusts or respects.
knowing current realities, and understanding The differentiations from Tinbergen’s posi-
the reasons for human and organizational tion of institutionalists and deliberationists call
actions (e.g., Berger, 1977). What would be con- for different rules of the policy design and policy
ducive interpretations and conditions for adop- process game than often prevail. How to get that
tion and effective implementation might be to happen and meet relevant conditionalities
made clearer. At the same time, communicative involves a system of constructions, actions and
discourse could be improved by the sorts of consequences, rather like those in Figure 4.2.
analyses and information at the heart of Institutionalists would be wise to consider
Tinbergian policy design, and be substantially the variety in what has happened in light of
disappointing in their absence. their rather similar prescriptions for post-
Whether the institutionalists’ or the delibera- Communist nations. They also might well con-
tionists’ recommendations would have the bene- sider the implications of historical experience,
fits they envision seems contingent. The former which shows that nations with ‘basic success
pose recipes which experience suggests are less stories … [with respect to meeting basic human
than guaranteed in their consequences across needs] … have had very different economies
situations. That is why we see alternating cycles from an institutional and political point of
of centralization and decentralization (Montias, view’ (Streeten, 1986: 29).
1986). Institutionalists’ recipes may fail to Deliberationists would do well to note the
acknowledge how learning from one pass at particular circumstances apparently needed to
them may lead to behavior which deprives them get a process of evolving policy through a more
of subsequent efficacy. For example, rationaliz- cooperative dialogue in place for environmen-
ing closing of U.S. military bases was helped in talists with ranchers and agriculturalists in the
one pass by establishing closing procedures American West under Department of Interior
“behind a veil of ignorance” where individual sponsorship. They may find sobering how
legislators did not know if and how their district some steps in line with a shared information,
would be affected (Weimer, 1992). While the transparency standard can, in particular, con-
procedures have remained in place for subse- texts countervail other parts of their prescrip-
quent rounds, legislators became fully aware of tions. Consider uses of courts and legislatures
the potential for adverse district consequences to obstruct or enhance pursuit of data sought
and moved to block further closings. by those already holding negative views about
The deliberationist approach depends on who the effects of smoking or particular industrial
gets to participate, the skills they bring to the products and processes. In those instances,
Peter-3383-Chapter-04.qxd 6/8/2006 7:32 PM Page 90
greater achievement of transparency could and Linder and Peters, 1988, 1984). From this
has had the consequence of deterring “emanci- perspective, the rules for inclusion in the policy
patory” policy dialogue (Hilts, 1999). designer’s basic repertoire emphasize relevance
Institutionalist and deliberationist approaches to the challenges posed by Figure 4.2 and to a
are potentially important and useful. They are wide range of policy domains in terms of both
not sufficient unto themselves, automatically issue areas and levels and forms of governance.
successful, or devoid of the challenges facing Any selection on those bases is of course open
other approaches to policy design. to argument, but I suggest ten core strategies.
They belong if only because going entirely with-
out any one of them exposes the designer and
his or her designs to substantial, hard to avoid
STRATEGIES – THE TEN
grief. The core ten are then in effect a necessary
COMMANDMENTS
part of the repertoire, not to be confused with a
complete one. For convenience they can be
When we consider the intellectual instruments labeled as: discipline breakout; minimum neces-
available to policy designers, it is possible to sary conditions; safeguards; placement; opposi-
find abundance or a rather thin portfolio. tional analysis; borrowing; tinkering; backward
If one includes the bodies of instruments mapping; forward mapping; and judgment
used in the relevant sciences and professions – shaping. They may seem banal, until one reflects
by no means limited to the social sciences – on how often they are slighted with, at least in
abundance seems an accurate description. retrospect, disappointing results.
Consider Dorfman’s (1986: 108) list of relevant Discipline breakout rejects as more than
specialties for environmental policy: engineer- stimulating grist for the designer’s mill the
ing; agronomy; meteorology; hydrology; chem- framing provided by any single technical and
istry; soil sciences; epidemiology; toxicology; professional disciplines. The designer must
oncology; zoology; biology; forestry; geology; break out from the constrained definitions of
economics; social sciences. Of course, the social problems, lists of actors, patterns of causation
sciences themselves contain many disciplines and correlation, and inventories of policy moves
and we can and should also recognize the perti- set by any single such discipline. Reality always is
nence of law, the information sciences, and more complicated than that and policy designs
political and social philosophy. If we consider are instruments to deal with reality rather than
the wealth of different policy issues for which scholastic gymnastics. The most efficacious
there is some demand for policy design the list designs will then be full of what are tangential
would get much longer still. Each has its instru- matters from any orthodox single frame per-
ments which are potentially applicable and spective, as argued in Ascher, 2002 and exempli-
examples can be found of their use in policy fied in Brewer and Kakalik, 1979.
design (e.g., for economics Cody, Kitchen and A minimum necessary conditions strategy
Weiss, 1990; Henderson, 1986; Cohen and has the policy designer identify and address
Wyplosz, 1995; Brainard and Martimort, 1996). what has to be the case, or be made the case, for
There are far too many strategies and tech- a preferred outcome or range of outcomes to
niques to review here, particularly if considered have much of a chance of realization. If those
with responsible cautions about the appropri- conditions are not met, the design should be a
ateness of their theoretical assumptions and non-starter. If those necessary conditions seem
feasibility of their information requirements. in reach via the design, it should receive seri-
The long-standing abundance just discussed ous consideration and warrants elaboration to
for practical purposes contrasts with a much enrich it with sufficient conditions. This strat-
smaller repertoire of options when we limit egy involves thinking through how to avoid
our consideration to strategies formulated, well-known, chronically possible failure by key
with a special focus on policy design and elements in any policy design, e.g., recurrent
generically applicable to it (Weimer, 1992: 135; types of government and market failure. It also
Peter-3383-Chapter-04.qxd 6/8/2006 7:32 PM Page 91
involves dealing with what are for the outcome including by legacy policy networks, e.g., for
the design pursues crucial parties, and crucial agricultural environmental regulations, as in
desiderata for them. Montpettit, 2002.
For example, in addressing the necessary Oppositional analysis assumes that attempts
conditions for Israeli–Palestinian peace, Yorke to pursue a policy design in practice will be the
(1990) focuses on the need to meet minimum target of deliberate foiling moves by adversaries
Israeli requirements (‘full peace and security’) at one or more points in Figure 4.2 (Wohlstetter,
and Palestinian ones (‘an independent state’). 1964, 1968). Thus, it becomes important to
She reasons that those minima can only be pro- envision the stimulus the policy as designed will
vided with normalization between Israel and itself pose to those who would object to it (mea-
some other states in the region. Given the estab- sure/counter-measure sequences). The active
lished suspicions and grievances, sustained nature of the environment also calls for antici-
movement toward peace requires simultaneous pation of secular trends and changes, for exam-
judgments by the parties that the advantages of ple, in technology, which may enhance the
peace outweigh the advantages of a continua- opportunities available to opponents. Still fur-
tion of the status quo … all have a stake in keep- ther, attention should go to the possibility of sit-
ing the peace and … that peace holds fewer risks uational changes (shocks), unplanned by any of
than the status quo’ (Yorke, 1990: 117). the affected parties, which may provide attrac-
That last phrase illustrates the third strategy, tive opportunities for opposition. Imagining
safeguards. The risks in following through such changes in the ‘terms of competition’ usu-
with a policy design do not fall primarily on its ally requires very detailed analysis of the end-to-
designers but on the affected parties. end chains of developments encompassed in a
Realization of the design in practice requires Figure 4.2 design and thus of the complete life-
their tolerance and often active steps on their cycle of a policy design in practice.
part. They are more likely to entrust their fate Design work usually rests on faith that there
to the design when it provides and is seen by are unachieved but, in principle, possible
them to provide means to limit damage should opportunities to do better than the status quo
it go awry or break down. For Yorke, those pro- and the looming future. It remains to go from
visions involve guarantees and commitments faith to credible demonstration of what those
by third parties external to the Middle East. are and the realistic availability of means to
Also, given the very mixed record of policy achieve a more preferred future. Borrowing is
design results, the responsible designer recog- a strategy for doing so (Rose, 1993; Schneider
nizes a similar need whatever the affected par- and Ingram, 1988; Weimer, 1993). It embraces
ties may think. The strategy of safeguards, it what seems to have worked well elsewhere in
should be noted, is not equivalent to incre- circumstances sufficiently analogous to those
mentalism, which carries its own risks. facing the policy designer. The premise is that
Placement begins from awareness that such “solutions” or “best practices” worked on
policy designs apply to systems with histories. similar problems in similar contexts, so they
That is true for the actors in them (individuals, should work for me.2 The wheel need not be
organizations, social groupings, politico-legal reinvented but only emulated. Warranted bor-
entities) and policy issues and options. The rowing clearly involves a host of judgments
designer starts not from a clean slate but from about performance elsewhere, and similarities
a clutter. Placement (Neustadt and May, 1986) to the situation facing the designer.
develops chronologies for pertinent actors and The greater the scope of what is borrowed,
issues of major experiences in their past. Those the harder it will be to get it into practice in the
provide no precise causal guidance as to prece- relevant situation. Even if what is borrowed is
dents which will be used in the present and quite limited and narrow, no two situations
future, but they are suggestive of how policy will be precisely alike in all pertinent respects.
designs alternatives will be treated differently These realities, plus the persistent appeal
from actor to actor and issue area to issue area, of incrementalism, argue for a strategy of
Peter-3383-Chapter-04.qxd 6/8/2006 7:32 PM Page 92
tinkering (Weimer, 1993). The idea is to tune uncertainties which would damp optimism;
or tweak what exists, or what is borrowed and attraction to sure gains and underweighting of
the context to receive it, without challenging only probable ones; and favoring small steps,
most of what is in place. Designers go for rela- and aggregates of small steps, away from status
tively ‘feasible manipulations’ of what already quo behavior over large steps singly and in
exists (May, 1981). The policy world, even its aggregate. A particularly misleading tendency is
reformist elements, usually is full of those who to extrapolate the ease or difficulty of the initial
prefer tinkering. It then will be hard to follow steps in a policy to that of the critical for out-
a less incremental design before considering comes later steps, e.g., as in much of the Bush
such options and discrediting them. Administration responses to 9/11, including the
Backward-mapping and forward-mapping Afghan and Iraq ventures (Bobrow, 2003).
each start from the same premise that “end-to- Moves to shape responses to designs include
end”, full life cycle treatment is needed for possibilities to take advantage of the ways in
robust design, a premise shared with opposi- which preferences are context-dependent.
tional analysis. The backward and forward Support for a particular policy design becomes
strategies differ in where they would have a contingent on the presence of alternative
designer start (Elmore, 1985). In backward designs. By presenting an increased number of
mapping the idea is to start just before the end alternatives, the designer calls into action the
in the production of outcomes and work dynamics of “tradeoff contrast” and “extreme-
toward the beginning, identifying who must do ness aversion”. For the former, imagine two
what and what must be the case all along the policy designs one of which offers both greater
way (Elmore, 1979–80). Forward-mapping benefits and costs than the other. Introducing a
starts at the initial point of realizing a policy third design with the greater benefit but at a
design and works toward its conclusion. It lays much higher cost will increase preferences for
out the transition from what exists, or is cur- the first design at the expense of the second. For
rently “programmed” by man and nature to extremeness aversion, imagine two policy
exist, to a more preferable outcome. Think of designs, each of which has a great positive puta-
backward or forward mapping as yielding a set tive consequence and a great negative one. If a
of stripmaps which together span the journey third design with more limited positive and
from origin to destination, but which can be negative consequences is introduced, it will pull
produced and read, starting with the last strip support away from one or both of the first two.
or the first. These strategies call for moving These tendencies operate for designers and
briskly from generalized principles to the not just for the parties affected by policy
detailed particulars of a policy context, taking design.3 In sum, understanding judgment
into account possible changes in it, such as shaping has substantial implications for how
those highlighted by oppositional analysis. we develop policy designs, for their quality,
The final strategy, judgment shaping, draws and for their chances of acceptance.
on research in behavioral decision theory At their extremes, any one of the ten com-
(Tversky and Kahneman, 1974; Kahneman mandment strategies is in some tension with
and Tversky, 1979; Kahneman and Lovallo, one or more of the others. Balanced and judi-
1993; Tversky and Simonson, 1993). That cious attention to all of them can, however,
research yields human tendencies in making improve capacity to deal with the challenges of
judgments. They imply correctives policy policy design.
designs should supply, and design moves to
shape the judgments of those involved in their
adoption, implementation and interpretation.
The tendencies differ from traditional models NOTES
of rational judgment.
Tendencies to correct for include: treating 1. Of course, pledges for continuing assessment may
problems as isolated unique instances; denial of not actually have been implemented, as with the waivers
Peter-3383-Chapter-04.qxd 6/8/2006 7:32 PM Page 93
granted to American states for departures from previous Bobrow, Davis B. and John S. Dryzek (1987). Policy
federal health care program guidelines. Analysis by Design. Pittsburgh: University of
2. Related to this strategy is a filtering one of “counter- Pittsburgh Press.
borrowing”. In it, the search is for failures in similar Brainard, S. Lael and David Martimort (1996).
contexts and what is found should then be rejected as a
‘Strategic Trade Policy Design with Asymmetric
design alternative.
3. The implications for the self-awareness and discipline
Information and Public Contracts.’ Review of
of policy designers are significant, with implications anal- Economic Studies, 63: 81–105.
ogous to those incorporated into the professional prepara- Brewer, Gary D. and James S. Kakalik (1979).
tion of psychiatrists. Handicapped Children: Strategies for Improving
Services. New York: McGraw-Hill.
Brinkerhoff, J.M. (2002). ‘Global Public Policy,
REFERENCES Partnership, and the Case of the World Commission
on Dams.’ Public Administration Review, 62: 324–36.
Bruner, Ronald D. (1986). ‘Case-wise Policy
Aaron, Henry J. (2000). ‘Seeing Through the Fog: Information Systems: Redefining Poverty.’ Policy
Policymaking with Uncertain Forecasts.’ Journal Sciences, 19: 201–23.
of Policy Analysis and Management, 19: 193–206. —— (1983). ‘Case-Wise Policy Analysis: Another
Ackoff, Russell (1979). ‘The Future of Operational Look at the Burden of High Energy Costs.’ Policy
Research is Past.’ Journal of the Operational Sciences, 16: 97–125.
Research Society, 30: 93–104. Campbell, Donald T. (1969). ‘Reforms as
Adelman, Irma and J. Edward Taylor, eds. (1986). Experiments.’ American Psychologist, 24: 409–29.
The Design of Alternative Development Strategies. Castro, Mary and International Labor Office.
Rohtak, India: Jan Tinbergen Institute of Devel- (1978). Migration in Brazil: Approaches to Analysis
opment Planning. and Policy Design. Liege: Ordina.
Alexander, Christopher (1964). Notes on the Synthesis Chung, Kyong Won (1993). ‘The Miracle of Han
of Form. Cambridge: Harvard University Press. River: Government Policy and Design Manage-
Alexander, Ernst A. (1982). ‘Design in the Decision- ment in the Motor Industry.’ Design Management
Making Process.’ Policy Sciences, 14: 279–92. Journal, 4: 41–47.
Anderson, Charles (1987). ‘Political Philosophy, Cody, John, Richard Kitchen and John Weiss (1990).
Practical Reason, and Policy Analysis,’ In Frank Policy Design and Price Reform in Developing
Fischer and John Forester, eds. Confronting Values Countries: Guidelines with Special Reference to
in Policy Analysis. Beverly Hills: Sage, 22–44. Industry. New York: St. Martin’s.
Arrow, Kenneth J. (1986). ‘Planning and Uncertainty,’ Cohen, Daniel and Charles Wyplosz (1995). ‘Price
In Adelman and Taylor, op.cit., 161–70. and Trade Effects of Exchange Rate Fluctuations
Ascher, William (2002). ‘The Brewer-Kakalik and the Design of Policy Coordination.’ Journal of
Research on Services for Handicapped Children: International Money and Finance, 14: 331–47.
A Paradigm of Applied Policy Sciences.’ Policy Cook, Thomas D. and Donald T. Campbell (1979).
Sciences, 35: 305–10. Quasi-Experimentation: Design and Analysis for
Bardach, Eugene (1977). The Implementation Game. Field Settings. Boston: Houghton Mifflin.
Cambridge: MIT Press. Daniere, Amrita G. and Lois M. Takahashi (1999).
Berger, Thomas R. (1977). Northern Frontier, ‘Public Policy and Human Dignity in Thailand:
Northern Homeland: Report of the Mackenzie Environmental Policies and Human Values in
River Inquiry. Toronto: James Lorimer. Bangkok.’ Policy Sciences, 32: 247–68.
Blackman, Allen and Alan Krupnick (2001). DeLeon, Peter (1994). ‘Reinventing the Policy
‘Location-Efficient Mortgages: Is the Rationale Sciences: Three Steps Back to the Future.’ Policy
Sound?’ Journal of Policy Analysis And Manage- Sciences, 27: 77–95.
ment, 20: 633–49. Demchak, Chris C. (2002). ‘Un-Muddling Homeland
Bloom, Howard S., et al. (2001). ‘Testing a Financial Security: Design Principles for National Security
Incentive to Promote Re-employment among in a Complex World.’ The Forum, 1. www.bepress.
Displaced Workers: The Canadian Earnings com/forum.
Supplement Project (ESP).’ Journal of Policy Dorfman, Robert (1986). ‘Benefit/Cost Analysis of
Analysis and Management, 20: 505–23. Environmental Problems,’ In Adelman and
Bobrow, Davis B. (2003). ‘Losing to Terrorism: Taylor, op.cit., 103–23.
An American Work in Progress.’ Metaphilosophy, Dror, Yehezkel (1986). Policy Making Under
34: 387–406. Adversity. New Brunswick, NJ: Transaction.
Peter-3383-Chapter-04.qxd 6/8/2006 7:32 PM Page 94
—— (1971). Design for Policy Sciences. New York: —— and Amos Tversky (1979). ‘Intuitive
American Elsevier. Prediction: Biases and Corrective Procedures.’
Edelman, Murray (1977). Political Language: Words Management Science, 12: 313–27.
that Succeed and Policies that Fail. New York: Keiser, Lael R. and Kenneth J. Meier (1996). ‘Policy
Academic. Design, Bureaucratic Incentives, and Public
—— (1971). Politics as Symbolic Action. New York: Management: The Case of Child Support
Academic. Enforcement.’ Journal of Public Administration
Elmore, Richard F. (1985). ‘Forward and Backward Research and Theory, 6: 337–64.
Mapping: Reversible Logic,’ In Kenneth Hanf and Killick, Tony (1995). IMF Programmes in Developing
Theo A.J. Toonen, eds. Policy Implementation in Countries: Design and Impact. New York: Routledge.
Federal and Unitary Systems. Boston: Martinus Lasswell, Harold D. (1971). A Pre-View of the Policy
Nijhoff, 33–70. Sciences. New York: Elsevier.
—— (1979–80). ‘Backward Mapping: Imple- —— (1961). Politics: Who Gets What, When, and
mentation Research and Policy Design.’ Political How. Cleveland: World.
Science Quarterly, 94: 601–16. Laudon, Kenneth C. (1986). Dossier Society: Value
Fischer, Claude S., Michael Hout, Martine Sanchez Choices in of the Design of National Information
Jankowski, Samuel R. Lucas, Ann Swidler, and Systems. New York: Columbia University Press.
Kim Voss (1996). Inequality by Design: Cracking Lerner, Daniel and Harold D. Lasswell, eds. (1951).
the Bell Curve Myth. Princeton: Princeton The Policy Sciences: Recent Developments in
University Press. Scope and Methods. Stanford: Stanford University
Fischer, Frank (1992). ‘Participatory Expertise: Press.
Toward the Democratization of Policy Science,’ In Linder, Stephen H. and B. Guy Peters (1992).
William N. Dunn and Rita Mae Kelly, eds. Advances ‘A Metatheoretic Analysis of Policy Design.’ In
in Policy Studies Since 1950, Policy Studies Review Dunn and Kelly, op.cit., 201–37.
Annual Volume 10. New Brunswick: Transaction, —— (1988). ‘The Analysis of Design or the Design
351–76. of Analysis.’ Policy Studies Review, 7: 738–50.
—— (1981). Politics, Values and Public Policy. —— (1984). ‘From Social Theory to Policy Design.’
Boulder: Westview. Journal of Public Policy, 9: 237–59.
Fox, Jonathan and John Gershman (2000). ‘The Loeb, Susanna and Mary Corcoran (2001). ‘Welfare,
World Bank and Social Capital: Lessons from Ten Work Experience, and Economic Self-Sufficiency.’
Rural Development Projects in the Philippines Journal of Policy Analysis and Management,
and Mexico.’ Policy Sciences, 33: 399–419. 20: 1–20.
Goggin, Malcolm L. (1987). Policy Design and the MacRae, Duncan (1988). ‘Professional Knowledge for
Politics of Implementation: The Case of Child Policy Discourse.’ Knowledge in Society, 1: 6–24.
Health Care in the American States. Knoxville: May, Peter J. (1981). ‘Hints for Crafting Alternative
University of Tennessee Press. Policies.’ Policy Analysis, 7: 227–44.
Henderson, David (1986). Innocence and Design: Meier, Kenneth J. and Kevin B. Smith (1994). ‘Say It
The Influence of Economic Ideas on Policy. Oxford: Ain’t So, Moe: Institutional Design, Policy
Blackwell. Effectiveness, and Drug Policy.’ Journal of Public
Hilts, Philip J. (1999). ‘Law on Access to Research Administration Research and Theory, 4: 429–42.
Data Pleases Business, Alarms Science.’ New York Miljkovic, Dragan (2002). ‘Economic Sanctions as
Times, July 31, A1ff. the Propositional Sustainability Problem.’ Policy
Ho, Lok Sang (2000). ‘Wage Subsidies as a Labour Sciences, 35: 1–15.
Market Policy Tool.’ Policy Sciences, 33: 89–100. Montias, J. Michael (1986). ‘On the Centralization
Hoag, Malcolm W. (1976). United States Foreign and Decentralization of Economic Activities,’ In
Policy: Why Not Project Interdependence by Design?; Adelman and Taylor, op.cit., 171–90.
Los Angeles: International Institute for Economic Montpetit, Eric (2002). ‘Policy Networks, Federal
Research. Arrangements, and the Development of Environ-
Hurley, Robert E., Deborah A. Freund, and John E. mental Regulations: A Comparison of the
Paul. (1993). Managed Care in Medicaid: Lessons Canadian and American Agricultural Sectors.’
for Policy and Program Design. Ann Arbor: Health Governance, 15: 1–20.
Administration Press. Murphy-Greene, C. and L.A. Leip (2002). ‘Assessing
Kahneman, Daniel and Dan Lovallo (1993). ‘Timid the Effectiveness of Executive Order 12898:
Choices and Bold Forecasts: A Cognitive Perspective Environmental Justice for All?’ Public
on Risk Taking.’ Management Science, 39: 17–31. Administration Review, 62: 679–87.
Peter-3383-Chapter-04.qxd 6/8/2006 7:32 PM Page 95
Neustadt, Richard E. and Ernest R. May (1986). —— (1988). ‘Systematically Pinching Ideas: A
Thinking in Time: The Uses of History for Decision Comparative Approach to Policy Design.’ Journal
Makers. New York: Free Press. of Public Policy, 8: 61–80.
Norberg-Bohm, Vicki (1999). ‘Stimulating “Green” Schon, Donald A. (1987). Educating the Reflective
Technological Innovation: An Analysis of Practitioner: Toward a New Design for Teaching
Alternative Policy Mechanisms.’ Policy Sciences, and Learning in the Professions. San Francisco:
32: 13–38. Jossey-Bass.
Norchi, Charles H. (2000). ‘Indigenous Knowledge —— (1983). The Reflective Practitioner: How
as Intellectual Property.’ Policy Sciences, Vol. 33 Professionals Think in Action. New York: Basic
no. 3–4, pp. 387–398. Books.
North, Douglas (1990). Institutions, Institutional Simon, Herbert A. (1972). The Sciences of the
Change, and Economic Performance. New York: Artificial. Cambridge: MIT Press.
Cambridge University Press. Stewart, Jenny and Russell Ayres (2001). ‘Systems
Ohls, James C. and Harold Beebout. (1993). The Theory and Policy Practice: An Exploration.’
Food Stamp Program: Design Tradeoffs, Policy, and Policy Sciences, 34: 79–94.
Impact: A Mathematica Policy Research Study. Streeten, Paul (1986). ‘Basic Needs: The Lessons,’ In
Washington: Urban Institute Press. Adelman and Taylor, op.cit., 27–37.
Oliver, Thomas R. (1999) ‘The Dilemmas of Taylor, John B. (1993). Macroeconomic Policy in a
Incrementalism: Logical and Political Constraints World Economy: From Econometric Design to
in the Design of Health Insurance Reforms.’ Journal Practical Operation. New York: W.W. Norton.
of Policy Analysis and Management, 18: 652–83. Tinbergen, Jan (1958). The Design of Development.
Paris, David C. and James F. Reynolds (1983). The Baltimore: Johns Hopkins University Press.
Logic of Policy Enquiry. New York: Longman. Tversky, Amos and Itamar Simonson (1993).
Pelletier, David et al. (1999).‘The Shaping of Collective ‘Context-Dependent Preferences.’ Management
Values through Deliberative Democracy: An Science, 39: 1179–89.
Empirical Study from New York’s North —— and Daniel Kahneman (1974). ‘Judgment
Country.’ Policy Sciences, 32: 103–31. Under Uncertainty: Heuristics and Biases,’ In
Pellikan, Huib and Robert J. van der Veen. (2002). Richard Zeckhauser, et al., eds., Benefit-Cost and
Environmental Dilemmas and Policy Design. New Policy Analysis. Chicago: Aldine, 295–307.
York: Cambridge University Press. Walters, Lawrence C., James Aydelotte, and Messica
Perlmutter, Howard V. (1965). Toward a Theory of Miller (2000). ‘Putting More Public in Policy
Social Architecture, Tavistock Pamphlet 12. Analysis.’ Public Administration Review, 60: 349–59.
London: Tavistock. Weimer, David L. (1993).‘The Current State of Design
Pressman, Jeffrey and Aaron Wildavsky (1973). Craft: Borrowing, Tinkering, and Problem Solving.’
Implementation. Berkeley: University of California Public Administration Review, 53: 110–20.
Press. —— (1992). ‘Claiming Races, Broiler Contracts,
Roessner, J. David, ed. (1988) Government Inno- Heresthetics, and Habits: Ten Concepts for Policy
vation Policy: Design, Implementation, Evaluation. Design.’ Policy Sciences, 25: 135–59.
New York: St. Martin’s. Wildavsky, Aaron (1979). Speaking Truth to Power.
Robins, Philip K., Charles Michalopoulos, Elsie Pan. Boston: Little, Brown.
‘Financial Incentives and Welfare Reform in the Wohlstetter, Albert (1968). ‘Theory and Opposed
United States.’ Journal of Policy Analysis and Systems Design,’ In Morton A. Kaplan, ed., New
Management, 20: 129–50. Approaches to International Relations. New York:
Rose, Richard (1993). Lesson-Drawing in Public St. Martin’s, 19–53.
Policy: A Guide to Learning Across Time and Space. —— (1964). ‘Analysis and Design of Conflict
Chatham, NJ: Chatham House. Systems,’ In Edward S. Quade, ed., Analysis for
Saeed, Khalid (1994). Development Planning and Military Decisions. Santa Monica: RAND, P-
Policy Design: A System Dynamics Approach. 387-PR, 103–48.
Aldershot: Avebury. World Bank (1991). Managing Development: the
Schneider, Anne and Helen Ingram. (1997). Policy Governance Dimension, Discussion Paper.
Design for Democracy. Lawrence: University Washington: The World Bank.
Press of Kansas. Yorke, Valerie (1990). ‘Imagining a Palestinian State:
—— (1989). ‘The Powers of Problem Definition: An International Security Plan.’ International
The Case of Government Paperwork.’ Policy Affairs, 66: 115–36.
Sciences, 22: 97–121.
Peter-3383-Chapter-04.qxd 6/8/2006 7:32 PM Page 96
Peter-3383-Chapter-05.qxd 6/8/2006 7:32 PM Page 97
5
Networks and Bargaining in
Policy Analysis
PETER BOGASON
compete for attention within their scholarly policy analysis. In particular, the fourth
community. Policy analysts also react to and generation of evaluation (Guba and Lincoln
analyze the same empirical phenomena in 1989) is closely linked to governance and
society. Thus, they create competition, innova- deliberative policy analysis. Fischer (1995)
tion and the diffusion of ideas, which often brings some of the pieces together nicely.
bear considerable resemblance to one another,
and which are discussed in groups of scholars
who share some fundamental views on social ORGANIZED ORDER VERSUS
theory. They then apply the theories with some MUDDLED PROCESSES
variation, according to the circumstances of
their empirical research.
The classical, rational and the mutual adjust-
In short, theory is contingent on time and
ment models form the backbone of this
space, and thus the present network under-
chapter. In this chapter, we shall focus on how
standing of policy has come about as a result
these models treat the decision-making
of scholars interacting and discussing the
processes and the interaction between actors.
possible interpretations of social phenomena –
in this case policy processes. This chapter
explores some trends that have been present Rational Policy-making
for the last 30 years or so without pretending
any full coverage, since the theme of policy “Speaking Truth to Power” (Wildawsky 1979)
networks is vast. And since the author was pre- indicates a troubled relationship between
sent most of the time, participating in several science and politics, between those finding the
networks, the critical reader may find some true state of the world and those wanting to
autobiographical biases. The reader will find rule it. Indeed, much of the policy literature is
other recent accounts of the development of concerned with authority, expertise and order
the theme in Hoppe (1999), Hajer and (Colebatch 1998). First, the policy literature
Wagenaar (2003), and Fischer (2003), each tai- deals with core activities of governments, set-
lored to a specific context (and all critical of ting up authority relations to back up the ideas
traditional policy analysis). In addition, Hill of the policy principles so that they may be
and Hupe (2002) provide a general discussion carried through authoritatively. Second, it dis-
from the angle of implementation, particularly cusses that policy principles do not come from
the tensions between top-downers and bottom- an empty space, they are based on in-depth
uppers (explained below). knowledge of the affairs the policy aims at reg-
The discussion will be selective, it is not ulating. This knowledge may come from gov-
possible to digest all types of network policy ernmental or external sources, but it is brought
analysis within one short chapter. We have together in the contents of the policy. And
omitted the trends towards a transnationaliza- third, the literature expects the policy to aim at
tion of domestic policies, which has been due solving a number of important problems
to international regimes, like the EU, the UN, within the target area, thus creating some sort
the WTO etc., discussing how policies are of order in that segment of society. In the end,
negotiated in complex settings involving many the policy may not be successful, but still,
actors, including various NGOs; see for exam- problem-solving is an important aspect of the
ple Linkage Politics (Rosenau 1969). Following general understanding of policy.
patterns towards institutionalization within Mostly authority, expertise and order has
the EU, there has been a merge of literature on been dealt with in the orderly fashion brought
intra- and inter-state relations, to some degree about by a top-down perspective, using a
captured by the concept of multi-level gover- sequential model of policy-making. Policy is
nance (Hooghe and Marks 2001). We also created, decided upon and implemented step
ignored the evaluation literature which, of by step by collecting information, weighing the
course, is relevant for methodology within pros and cons of various possible ways of
Peter-3383-Chapter-05.qxd 6/8/2006 7:32 PM Page 99
acting, and then deciding on the course of execution of the policy, executing it, and
action that – in the vein of Pareto-equilibrium – evaluating the results.
will provide most people with most happiness The feedback elements of the model give it a
for the lowest costs. Public (sub)agencies then dynamic feature, and Dror stresses the demands
execute the policy without much further ado. for iterative processes. He also leaves room for
This model, often named rational,2 consti- “extra-rational” behavior based on limited
tutes a core in the sequential model of policy- resources, uncertainty, and lack of knowledge as
making (see chapter by Charles O. Jones), well as creativity and intuition (Dror 1968,
a model with good heuristic qualities, and a 157–158), but the aim of the model is to limit
model that fits the picture which has domi- the importance of such elements in order to
nated constitutions separating politics and enhance optimal policy-making – understood
administration, as well as the minds of man- as “one that is not distorted by the noise that is
agers, and their supporting management con- in fact inherent in all, and especially complex,
sultants and also much of the literature on structures” (Dror 1968, 200). The task, then, is
management. It is a model of leaders being in to organize processes so that at least one unit
control at the apex of the organization, from contributes to each phase, and so that the con-
where they can design the processes desired to tributions of various units add up to an overall
obtain the goals of the organization. A good optimal operation at low costs and with little
example of how these lines of thought have distortion. However, there is no one single
been used in the literature is provided by model for organizing – one may use hierarchy
Yehezkel Dror, who in 1968 published his or polycentric structures in various forms,
Public Policymaking Reexamined (Dror 1968), depending on the demands of the situation. The
followed in 1971 by two companion books judgment of success or failure rests on the con-
(Dror 1971; Dror 1971) to substantiate some tribution of the participants to the process, not
of the contentions of the first book. His aim to a particular organizational form.
was twofold: to advance the study of policy- In other words, Dror does not subscribe to a
making and to contribute to the improvement monolithic hierarchy. Nevertheless, he empha-
of public policymaking – which lacks the sizes the need for overall systems management,
proper use of knowledge. metapolicy-making and comprehensive public
Dror’s optimal model has three major stages policy-making in order to promote adjustment
(Dror 1968, 163–196): Metapolicy-making, and take advantage of new knowledge, and to
policy-making and post-policy-making, and prevent sub-optimization by single units. The
within those there are eighteen sub-stages, one key to such a demand is better personnel: pro-
of which is continuous communication and fessional staffs, units to survey and retrieve
feedback channels interconnecting all phases. knowledge, and units for policy-oriented
Metapolicy-making involves seven stages of research. This form of manpower is to be sup-
processing values, processing reality, process- ported by computerized systems and it must
ing problems, developing resources, designing be managed in new ways (this is 1968), “in
the policy-making system, allocating problems, order to stimulate interprofessional teamwork
values and resources, and finally determining and creativity” (Dror 1968, 274). In addition,
the policy-making strategy. Policy-making there must be some systematic evaluation and
involves another seven stages of suballocating learning feedback from experience.
resources, making and prioritizing operational The model, then, relies on our capabilities to
goals, ditto for other significant values, prepar- produce knowledge based on science, and to
ing a set of major alternative policies (includ- feed it into the policy-making process in order
ing some “good” ones), predicting benefits and to enhance enlightened choices within a com-
costs of those policies, identifying the best prehensive system, and in order to avoid incre-
policies in that light, and then deciding mental policy-making (see next section) which
whether the best alternatives are “good” poli- in Dror’s opinion amounts to nothing but con-
cies. Post-policy-making involves motivating the servatism in disguise. In a later edition of the
Peter-3383-Chapter-05.qxd 6/8/2006 7:32 PM Page 100
book (Dror 1983), Dror has added an making a comprehensive analysis of all possible
introduction in which he laments the lack of means to obtain an end, the administrator
advance of the policy sciences in the direction resorts to comparing only a few which often do
he has recommended. He also acknowledges not deviate much from past uses, and the one
that active participation in governmental roles selected is the one that creates agreement
has, in the meantime, taught him some impor- among the participants in the policy-making
tant lessons about policymaking: “Social process, no matter what their ideological
science studies from the outside do not pene- standpoints might have told them to do.
trate into the realities of central high-level For the purposes of this article, Lindblom’s
decision making”, and “… ominous policy- discussion of how agreement comes about is
making weaknesses are built-in into core com- crucial. It is not a long or even deep analysis. It
ponents of governance, with present policy is a short, nearly an ideal type description of
predicaments overtaxing maximum policy- how almost every interest in the USA has its
making capacities.” (Dror 1983, x–xi). The watchdog, and that in the formation of a policy
problem he faces is that of research and advice a process of mutual adjustment takes place
versus politics of all sorts. The original book is among various interest groups and public agen-
based on the ideal of science as an integral part cies; and even though all these actors may not
of the desired model optimal policy-making, have an explicit focus on a particular policy
and the political dimensions were not treated goal, the result of the processes will be a viable
in-depth – a problem Dror did not solve, policy. Thus there is no comprehensive income
no matter how many times he paid heed to policy in the USA, but “a process of mutual
other sources of information, including extra- adjustment among … (various actors) …
rational forms3. accomplishes a distribution of income, in which
particular income problems neglected at one
point in the decision process becomes central at
Mutual adjustment in policy-making another point.” (Lindblom 1959). Furthermore,
policies are not made once and for all, but
This line of argumentation in the policy litera- changed and adapted in a never-ending and
ture is concerned with the empirical character- continuous process in which those who lost at
istics of the policy process in a political setting. one point may gain at another. Moreover, since
Analysis of policy cannot be understood in changes are incremental, losses (and gains) for
isolation from the ways politicians, administra- each policy process are endurable.
tors and representatives of interest in society at The underlying understanding of this
large interact about themes of common inter- process is one of a large number of actors, con-
est. One core argument, promulgated by tinuously interacting about a host of themes,
Charles E. Lindblom, is that the information rarely coordinated by any central agency, but
rendered in and by such processes has as much rather performing according to some analogy
value as information produced by researchers of the hidden hand of the economic market.
and other experts. So, where proponents of the Lindblom indicates this without really concep-
rational model recommend problem-solving tualizing it in footnote number 7 in the article:
based on the authority of expertise, followers of “The link between the practice of successive
mutual adjustment advise problem-solving limited comparisons and mutual adjustment
based on the authority of agreements reached of interests in a highly fragmented decision-
among interested parties. making process adds a new facet to pluralist
Lindblom’s most famous text is, undoubt- theories of government and administration.”
edly, “The Science of Muddling Through” Lindblom expanded this line of thinking in his
(Lindblom 1959), originally published in The intelligence of Democracy (Lindblom 1965)
Public Administration Review, but reprinted in with the subtitle Decision making through
numerous Readers. The message is relatively mutual adjustment. The book sets the tone on
simple, but also highly contested; instead of page 3: “… people can coordinate with each
Peter-3383-Chapter-05.qxd 6/8/2006 7:32 PM Page 101
other without anyone’s coordinating them, they differ sharply in their interpretation: the
without a dominant common purpose, and rational model subscribes to comprehensive
without rules that fully prescribe their rela- uses of scientific knowledge, whenever possible,
tions to each other.” the model of mutual adjustment puts science on
The principles of this argument are found in a par with any other type of knowledge. This
a much earlier paper from 1955, “Bargaining. does not mean that Dror’s model ignores other
The hidden hand in Government” (Lindblom means of acquiring knowledge, but any infor-
1988), and it is a largely un-referenced, but mation should be put into a context of priorities
insightful, discussion of how bargaining coordi- set beforehand. The model of mutual adjust-
nates policy, how it takes place in and among ment does not rely on pre-set goals, but on
public agencies, and how actors are motivated agreement acquired during the process.
for that particular behavior. The key is that no Both models are created for Western demo-
one trusts hierarchy to bring forward “every fact cratic and pluralistic societies. Therefore, they
and value favorable to him. We want a social both contain elements of communication and
mechanism in which every man can speak for interaction which are useful for our subse-
himself or find someone to speak for him.” So quent discussion of networks. However, their
bargaining involves actors and brings forward understanding of how to play a role in a demo-
more aspects to a matter. In other words, the cracy is quite different. One is based on tech-
policy-making process is a matter of politics in nocratic knowledge, depending on how
the broadest sense, and in politics there is not politicians allow it to be expressed. The other
only one truth available. Researchers mostly fol- one is based on knowledge in the demos,
low the political master designated by the hier- depending on how it may express itself.
archy, but other parties interested in the matter Rational models are often seen as command-
may contribute with other views. and-control systems, featuring the (democratic
In the quote above, we find one clue to and elected) top. Dror does not subscribe to
Lindblom’s subsequent career of advocating such a view, but recommends interaction
for pluralism in policy analysis. There is between stages and between actors in the
more, of course; the arguments are unfolded in process – within the frames of goal-setting. The
The Intelligence of Democracy and used in model presupposes that the politicians ulti-
Lindblom’s and Cohens’ Usable Knowledge mately are in control of the bureaucracy and
(Lindblom and Cohen 1979). One basic hence, in Dror’s terms, they control meta-
message is that there is no privileged knowl- policy-making. The bureaucrats provide politi-
edge in the policy process,4 and another is that cians with documentation for any verifiable
the process can only be successful if agreement statement and they substantiate that all relevant
(not only compromise) is reached: then the information has been scrutinized. In turn, the
process has acquired a rationality which serves politicians are controlled by the voters at the
a democratic solution. general elections and by the watchdog function
provided by a free press. So, the rational model
is also to be applied in a pluralistic setting.
The two models compared The model of mutual adjustment is basically
one of interaction, but the number of actors is
The two models are, indeed, adversaries. an open question, dependent on the democra-
Dror explicitly renounced incrementalism, tic procedures of society. It requires a pluralis-
and Lindblom, of course, wrote to warn tic society and a political system that allows
against any belief in the rational model. Dror is various societal interests to enter the policy-
not a rationalist in the classic sense, but his making processes and participate with a
model should be seen as an approximation to prospect to win attention and influence now
rational decision-making. and then. Who exactly will win and when is
The models share an interest for the role of then an open (empirical) question. These
knowledge in the policy-making process. But conditions should be fulfilled in a polity in a
Peter-3383-Chapter-05.qxd 6/8/2006 7:32 PM Page 102
pluralistic society like the USA. But 18 years used the systems approach and wanted to
after his seminal article (Lindblom 1959) was explain the outputs and outcomes of the polit-
published, Lindblom conceded in a much ical system without really analyzing the con-
acclaimed book (Lindblom 1977) that there tents of the policy processes (Sharkansky and
might be a bias in the policy system which Hofferbert 1969). But empirical analyses com-
provided certain actors with more clout than ing closer to the dynamics of the policy process
others – in the American case, big business. In led to the conclusion that focus on the out-
a later book (Lindblom 1990), Lindblom stated comes of single organizations like the legisla-
that, although imperfect, he saw no alternative ture was not really helpful: ideas and principles
to pluralism; instead the challenge was to cope in parliamentary law were often changed
openly with the problems to reduce adverse during processes of adaptation in the executive
consequences as much as possible. branches and in implementing organizations
at the regional and local levels.
Pressman and Wildawsky’s examination of the
fate of a federal program in a local setting is a
TOWARDS NETWORK ANALYSIS
classic example (Pressman and Wildawsky 1973).
One main explanation of the changes was the
Policy analysis has its main roots in American long chain of decision-makers from Washington,
political science – with a little help from DC, to Oakland, CA, which they called the deci-
friends in economics and sociology – and in sion path, and they viewed each decision-maker
public administration, which, however, for as a relatively autonomous actor who could in
most policy purposes itself is rooted in politi- effect block progress. In a second edition in 1979
cal science. Political science developed a strong (Pressman and Wildawsky 1979), Wildawsky
platform in the 1950s and 1960s based on wrote a new chapter on “Implementation in con-
various versions of systems analysis – David text”, and referred to Hugh Heclo’s use of issue
Easton (e.g. Easton 1965) and Karl Deutsch network as a heuristic device to understand how
(Deutsch 1963) are examples of mainstream policies were coordinated.
thinking in the field. And, regardless of the Heclo had coined the term network much
potentials for other ways of doing analysis in, earlier. In a review article on policy analysis he
for instance, Deutsch’s cybernetic ideas, politi- wrote that one should be careful “not to reify
cal scientists focused their interest on organi- collectivities into individual deciders but to
zations within the political systems, often understand the networks of interaction by
conceptualized as institutions, meaning inter- which policies result” (Heclo 1972, 106), and
est organizations, political parties, parlia- he recommended analyzing within programs
ments, the executive, local government and (instead of analyzing organizations). This he
other organizational forms of political life. did himself in Britain, in Sweden and in the
Their aim was to theorize about these compo- USA, research which led him to core concepts
nents of the political system – an example of within policy analysis: policy communities and
such a partial analysis is Sjöblom (1968) on issue networks (Heclo and Wildawsky 1975).
political parties in a multiparty system, Policy communities were more stable interac-
strongly influenced by David Easton and tion patterns among policy interests, issue net-
Anthony Downs (1957). works were mostly ad hoc mode.
However, most policy analysts were not so Heclo was not alone in such research. In a
interested in theorizing about components number of research settings, scholars were
within the political system. The systematic searching for theoretical and conceptual solu-
policy movement started in the second half of tions to their observations of multiple actors
the 1960s (e.g. Ranney 1968) and became a interacting in policy formulation and imple-
thriving field in the 1970s, first of all as policy mentation. Many of them share empirical
implementation research. Many policy analysts observations, but their point of departure in
Peter-3383-Chapter-05.qxd 6/8/2006 7:32 PM Page 103
a large number of interorganizational networks one another for updating and development of
in society (Benson 1982). In Europe, Rod their understanding of the environment.
Rhodes used these and other sources in organi-
zation theory as inspiration for developing his
models of state-local government relationships
SUBSEQUENT TRENDS IN POLICY
(Rhodes 1979; Rhodes 1986) which he contin-
NETWORK ANALYSIS
ued to use for an extensive discussion of policy
networks and policy communities as organizing
factors in British politics and administration Above we inspected some of the roots of policy
(Rhodes 1997). network analysis and its development in the
Another line of inter-organizational research 1970s and early 1980s. We shall now discuss the
had a micro focus and could be said to have subsequent development of analytical pers-
some inspiration from literature dealing with pectives, which include traditionalists, institu-
increased division of labor in society. How tionalism, governance, and trends towards
can such sectors be understood? Most ratio- deliberative discourse analysis. They have devel-
nales in such analysis are based on theories oped historically, of course, so they overlap, and
of public choice, which lead to the conception to some degree they both react to and build
of a service industry (Ostrom and Ostrom upon one another in the sequence of institu-
1977; Ostrom, Parks and Whitaker 1978) – tionalism in the 1980s, governance in the 1990s
which had many common features with a and deliberative analysis in the late 1990s and
policy sector. The basic idea was to counter now under further development in the 2000s.
theories of (large) bureaucratic organization Traditionalists were found all the time – but
by theories of (small) organizational coopera- some of them changed with the currents.
tion, making a case for small-scale govern- What were those trends about? Grossly over-
ment and governmental agencies which would simplifying, one can say that there has been a
pool resources for larger tasks, if necessary. move from system and hierarchy (rational
The basic ideas were developed into game models) towards fragmentation and empower-
theoretical frameworks and applied in various ment (mutual adjustment). Institutionalists
forms of self-government (Ostrom 1990) and were concerned with how political systems fared
in intergovernmental relations in Germany and they worked to re-conceptualize the mod-
(Scharpf 1997). ernistic state apparatus into something less
Both the resource based and rational choice monolithic in processes involving various stake-
models were presented in an often-quoted holders in society. Governance scholars contin-
anthology on inter-organizational policy- ued this work and conceptualized the workings
making in 1978 (Hanf and Scharpf 1978). It of various parts of the systems and helped us
fulfilled at that time the need implementation understand better how network policy processes
scholars had for analytical models, which at took place. Discourse and deliberation scholars
the same time caught interaction among levels cashed in on further changes in society towards
of administration as well as at each level. It was involving citizens in policy processes, and they
to be the first volume of the soon after rapidly also were part of the general movement among
growing literature on the fragmentation of the some social scientists towards social construc-
state apparatus. The fragmentation was due to tivism and pragmatism.
decentralization of powers to lower levels and These three forms constitute some of the
sharing powers with various organizations in “forefront” in research during those years. But
the “gray” zone. This created new and intensi- that is not to say that every one participated.
fied possibilities to exercise influence on sepa- Of course, many policy analysts proceeded in
rate decision-makers. Following this tendency, more traditional veins and challenged the new-
the borders between public and private tend to bees, or approached the new ideas without
become blurred, and the exchanges of infor- buying them wholesale. So, first, we’ll review
mation make the various actors dependent on some of the main arguments among them.
Peter-3383-Chapter-05.qxd 6/8/2006 7:32 PM Page 105
Whatever the case, policy analysts found thus consciously waiving their formal autonomy
that the formal organizational system of govern- in such fields, replacing it by mutual depen-
ment often did not adequately describe the dence. Structuralists likewise asked themselves
patterns of interaction they found in policy about the role of institutional settings, but their
formation and implementation. Moreover, the interest was, more or less, to make plausible a
alternative, “American” behavioral analysis, claim that organizational actors do not decide as
lacked a foundation in or a link to what orga- rational actors, they rather follow standard
nizations meant in political life. If one struc- operating procedures, and normative facets of
tures the field in terms of Richard Scott’s three the organization as they appear in myths, sym-
types of institutional theory – regulative, nor- bols and even rituals – in short, the organiza-
mative and cognitive (Scott 1995), three types tional culture which would define appropriate
of questions interested policy analysts. First, behavior of actors. Therefore, organizational
they found themselves confronted with ques- factors would be important, but in other ways
tions of what systems of rules that might really than traditional political science had used them
apply to the actions of both organizations and in the past (March and Olsen 1989).
individuals. Furthermore, they asked them- The main difference between the two types
selves what social obligations the actors of new institutionalism, then, was rooted in
wanted to adhere to when they set standards opposite hypotheses about the behavior of
for future policies, or when they implemented actors. In addition, there were strong norma-
policies in ways that were not always in close tive differences, in that most rational theorists
accordance with the stated, formal policy prin- did not much care about how services were
ciples. Third, they found a need for identifying provided and therefore might advocate for pri-
norms for proper behavior in networks across vatization; structuralists to a much greater
organizational boundaries – how did actors extent adhered to maintaining the particular
perceive one another, and how did they come democratic values provided by public sector
to terms when their organizational back- organization of services. In policy terms, this
grounds differed? became very visible in normative discussions,
Such questions are to some degree answered e.g. about the pros and cons of new public
by various institutional theories. Scott’s distinc- management (Hood 1991; Barzelay 1992).
tions were not part of the discussions of policy
scholars in the 1980s, at least not explicitly, so
they articulated their institutional theories dif- Governance
ferently. There was a relatively clear cleavage
between scholars working on the basis of ratio- A second major trend came in the 1990s, and its
nal choice theory and those who were more theme was labeled governance. In many ways it
interested in structural analysis. In a way, their was a natural sequel to the focus on institution-
interest was much about the same. They realized alism in the 1980s. There was an enduring com-
that it would be no use only to focus on formal petition between macro- and micro-analytical
organizations like parliaments or bureaucracies approaches to conquer the right to be called new
to analyze policy processes. Rational choice institutionalists (Selznick 1996), and there were
theorists then asked themselves how variations tensions between new and old institutionalists,
in structural conditions would affect various to say nothing about those who still saw formal
types of rational actors. Examples are various organizations as institutions (Aberbach and
ways of organizing the police force (Ostrom, Rockman 1987). So the more the field of
Parks and Whitaker 1978), or ways of organiz- analyzing policy networks developed, the more
ing local governments in metropolitan areas the search for more adequate concepts intensi-
(Oakerson 1987). The rationale behind this was fied. Increasingly, the concept of governance
that small organizations could be effective if gained momentum: it could be seen as some-
they cooperated with other ones about certain thing other than government, and it had a
tasks in a rational way, based on self-interest, processual flavor to it.
Peter-3383-Chapter-05.qxd 6/8/2006 7:32 PM Page 107
that guide the ways people create the shared research practices: “… a close practical and
meanings which motivate them to act.” So: conceptual connection exists between a post-
“Policy analysis and planning are practical positivist policy analysis and today’s decentered
processes of argumentation.” (Fischer and world of governance” (Hajer and Wagenaar
Forester 1993, 1–2). 2003, xiv).
No reference to networks in those sentences, This statement, then, reflects that, in most
but of course the development of policy analy- Western countries, the public sector has been
sis towards networks had unveiled processes opened up for more participation in policy
which were hidden in the organizations of the processes. One may doubt the sincerity of this
1960s and before. So they were closed to the (Miller 2002, vii–viii), but measured on the
type of scholarly scrutiny which was, after all, surface – by the sheer growth in the number of
easier to perform in the networked policy new channels for participation – this is a fact
processes of communication researchers followed (OECD 2001). In policy analysis, this has con-
in the 1970s and 1980s. And, sure enough, sequences for the role and use of expertise
about half of the articles in the anthology (Fischer 1999) which increasingly becomes
discuss various forms of deliberation in the part of an ongoing discourse with less and less
policy process, and hence indirectly network elevated status for policy analysts; instead they
settings. have to make their points of view understood
Discourse thus relates to language, and a by a broader public. Deliberation also means
primer on narrative policy analysis was written that organized interests get more legitimate
by Emery Roe (1994). But there are also roots access to the policy process, but in the light of
in institutionalism: “From this perspective acid the research on institutionalism and gover-
rain is a story-line that, potentially, brings out nance, that is hardly surprising. But the conse-
the institutional dimensions of the ecological quences for the roles of ordinary citizens may
problematique.” (Hajer 1995, 265). This means be more profound, in that citizens get access to
that the author has an analytical interest in participate in ways that earlier on might have
how discourse is structured or embedded in been seen as counterproductive to an efficient
society, while at the same time it structures public sector. Some of the development may be
society – in other words, not unlike Giddens’ conceptualized as empowerment of citizens
ideas of structuration, which has the concept (Sørensen 1997); an interesting research ques-
of institution at the core of the analysis tion is to what degree formal rights to parti-
(Giddens 1984). cipate actually are brought into use for
Speaking metaphorically, the deliberative influencing policy decisions. If that is the case,
policy analysis brings the scholar down from other researchers speak of a strengthening of
the ivory tower to the people. The institution- social capital in society (Putnam, Leonardi and
alists and most governance theorists kept the Nanetti 1993). In more radical versions, one
privileged status of researchers to analyze cur- can say that the citizens decide about the
rents in society and to work for a better theo- future of their communities (Ostrom 1995)
retical understanding of how policy came instead of, for example, relying on a benign but
about. But the 1990s gradually saw changes in bureaucratic welfare state.
the social sciences, which meant that the privi- The research, then, stresses the features of
leged and isolated status of scholars was meant deliberation, dialogue, collaboration and
to be revoked, and their roles to be changed mediation. Much of it should be understood as
from observers to participants in research part of the scientific development towards
processes that stressed dialogue instead of postempiricist social science. Empiricists, or
observation and reporting (Guba and Lincoln traditional policy analysts, have tried to mini-
1989; Erlandson, Harris, Skipper et al. 1993). mize social and interpretative judgments,
Thus, the advocates of change mirrored soci- postempiricists recognize their basic, constitu-
etal developments towards more public partic- tive role in any form of analysis (Fischer
ipation in policy processes in their own 2003, 226). Postempiricist policy analysts do
Peter-3383-Chapter-05.qxd 6/8/2006 7:32 PM Page 109
not speak truth to power, they collaborate with more and more complicated, or at least
power holders and mediate with diverse inter- comprehensive, policy process everywhere in
ests. In that sense, they have become part of the the Western world. The point, then, does not
network society. concern the exact definition of the phenomena
under scrutiny, but the general recognition
that we are observing qualitatively different
policy processes. The challenge is to show how
NETWORK ANALYSIS – A STATUS
they differ from the past, and what measures
then should be taken.
Where is network analysis at now? We have What, then, may we find in common among
followed a historical trajectory from rational the different camps of policy network analysis?
policy analysis to analysis integrating a broader It seems to me that one main distinguishing
conception of the processes involved. So feature of the advanced policy analysts of today
scholars involved in such analysis have all par- is that they apply a new version of pragmatism.
ticipated in a battle between rational decision- The classic Deweyan pragmatist was interested
making and muddling through – maybe not in in theory as a vehicle for promoting change in
the open, and maybe not even as a conscious societal affairs. The pragmatist of today has less
choice by the analyst, but still the theme of faith in theory. To put it crudely, pragmatists of
doing things rationally or not pops up every- today are interested in conceivable practical
where. At the very least, as a pedagogic means to consequences of affirming an idea or taking
illustrate what we are not doing. More seriously, an action – consequences that are satisfying
as a theme that has to be addressed in order to and desirable in the light of power relations
persuade the reader that rationalism is or is not (Cherryholmes 1999, 124–125). They follow a
applicable in this case – and it seems that ratio- pragmatism which is anticipatory and hence
nalism is on the decline.5 That is not to say that inductive and fallible; today’s pragmatists con-
Lindblom’s mutual adjustment is the only struct their reality socially and perform analysis
answer to non-rational demands. But his ideas critically; they are skeptics and hence not believ-
are hovering over many of the solutions we face. ers of a final Truth. They see the world as con-
The outline given above about the changes tingent, and thus they are contextualists. They
in policy analysis stresses a transition towards are holists and reject distinctions like fact/
network analysis, but that does not mean that value, objective/subjective, theory/practice, ends/
former types of policy analysis are gone. means, analytic/synthetic. This credo I will call
“Network” is still a debated term, to say noth- the “new pragmatism.” It certainly covers the
ing of network analysis. Have we not seen it all postpositivists within policy analysis, and to a
before? Of course we have in some sense, our degree it covers many other network analysts –
predecessors in political analysis were not of whom some still subscribe to a distinction of
idiots. In a more narrow sense, things are new, fact and value, and of objective and subjective.
but, as Keith Dowding (1994) has shown, the New pragmatists do not see evidence in the
literature then tends to become bogged down classic sense of getting the data straight, prefer-
because of definitional fights between academic ably in some version of statistical analysis. They
camps. So more energy is used for fights than beg the question of the existence of a network
for sensible analysis of one’s own results as well and involve themselves in processes of argu-
as of the results from colleagues. Christopher mentation and power – resource exchange or
Pollitt’s critique that there are tendencies not, “science” or not. They base their action on
towards ahistorical comprehension, and that it some form of hermeneutic analysis, and many
is hardly proven that networks form a new and of them do not mind using supplementary
better type of democracy, are also worth con- information based on some strand of posi-
sidering (Pollitt 2003, 65–67). That said, I tivism. Nevertheless, they see such evidence as
think that one should interpret the focus on one out of many channels of information for
networks and process as a consequence of a their craft. Hoppe characterizes two types of
Peter-3383-Chapter-05.qxd 6/8/2006 7:32 PM Page 110
Benson, J. K. (1982) A framework for policy Society, eds M. A. Hajer and H. Wagenaar, Theories
analysis. In Interorganizational Coordination: Theory, of Institutional Design, pp. 209–27. Cambridge:
Research and Implementation, eds D. L. Rogers and Cambridge University Press.
D. A. Whetten, pp. 137–76. Ames, Iowa: Iowa Fischer, F. (2003) Reframing Public Policy. Discursive
State University Press. Politics and Deliberative Practices. Oxford: Oxford
Bogason, P. (2000) Public Policy and Local Governance: University Press.
Institutions in Postmodern Society, New Horizons in Fischer, F. and Forester, J. (1993) Editors’ introduc-
Public Policy. Cheltenham: Edward Elgar. tion. In The Argumentative Turn in Policy Analysis
Campbell, J. L., Hollingsworth, J. R. and Lindberg, L., and Planning, ed. F. Fischer and J. Forester,
eds. (1991) Governance of the American Economy. pp. 1–17. London: UCL Press.
Cambridge: Cambridge University Press. Friend, J., Power, J. and Yewlett, C. (1974) Public
Cherryholmes, C. H. (1999) Reading Pragmatism. Planning: The Inter-Corporate Dimension.
New York: Teachers College Press. London: Tavistock Publications.
Colebatch, H. K. (1998) Policy. Buckingham: Open Gage, R. W. and Mandell, M. P. (1990) Strategies for
University Press. Managing Intergovernmental Policies and Networks.
Czada, R. M. and Windhoff-Héritier, A., eds (1991) Praeger: New York.
Political Choice. Institutions, Rules and the Limits of Giddens, A. (1984) The Constitution of Society.
Rationality. Frankfurt am Main: Campus Verlag. Outline of the Theory of Structuration. Cambridge:
Deutsch, K. W. (1963) The Nerves of Government. Polity Press.
Models of Political Communication and Control. Guba, E. G. and Lincoln, Y. S. (1989) Fourth
New York: The Free Press of Glencoe. Generation Evaluation. London: Sage.
Dowding, K. (1994) Policy networks: Don’t stretch a Hajer, M. A. and Wagenaar, H., eds (2003) Deliberative
good idea too far. In Contemporary Political Policy Analysis. Understanding Governance in the
Studies, eds P. Dunleavy and J. Stayrer, pp. 59–78. Network Society, Theories of Institutional Design.
Exeter: Shortrun Press. Cambridge: Cambridge University Press.
Downs, A. (1957) An Economic Theory of Democracy. Hajer, M. A. and Wagenaar, H. (2003) Introduction. In
New York: Harper & Row. Deliberative Policy Analysis. Understanding
Dror, Y. (1968) Public Policymaking Reexamined. Governance in the Network Society, eds M. A. Hajer
Pennsylvania: Chandler Publishing Company. and H. Wagenaar, Theories of Institutional Design,
Dror, Y. (1971) Design for Policy Sciences. New York: pp. 1–30. Cambridge: Cambridge University Press.
American Elsevier. Hajer, M. (1995) The Politics of Environmental
Dror, Y. (1971) Ventures in Policy Sciences. New York: Discourse. Ecological Modernization and the Policy
Elsevier. Process. Oxford: Clarendon Press.
Dror, Y. (1983) Public Policymaking Reexamined. Hanf, K. and Scharpf, F. W., eds (1978) Inter-
New Brunswick, N.J.: Transaction Books. organizational Policy Making. Limits to Coordina-
Easton, D. (1953) The Political System. An Inquiry tion and Central Control, SAGE Modern Politics
Into the State of Political Science. New York: Alfred Series, 1. London: Sage.
A. Knopf. Heclo, H. (1972) Review article: Policy analysis.
Easton, D. (1965) A Framework for Political Analysis. British Journal of Political Science, 2, 83–108.
Englewood Cliffs, NJ: Princeton University Press. Heclo, H. and Wildawsky, A. (1975) The Private
Elmore, R. E. (1979) Backward mapping: Implem- Government of Public Money: Community and
entation research and policy decisions. Political Policy Inside British Politics. London: Macmillan.
Science Quarterly, 94(4), 601–16. Hernes, G., ed. (1978) Forhandlingsøkonomi og bland-
Erlandson, D. A., Harris, E. L., Skipper, B. L. and ingsadministration. Oslo: Universitetsforlaget.
Allen, S. D. (1993) Doing Naturalistic Inquiry: A Hill, M. and Hupe, P. (2002) Implementation.
Guide to Methods. London: Sage. London: Sage Publications.
Fischer, F. (1995) Evaluating Public Policy. Chicago: Hjern, B. and Hull, C. (1982) Implementation
Nelson-Hall Publishers. research as empirical constitutionalism. European
Fischer, F. (1999) Technocracy and the Politics of Journal of Political Research, 10(2), 105–17.
Expertise: Managerial and Policy Perspectives. Hjern, B. and Porter, D. O. (1983) Implementation
London: Sage. structures: A new unit of administrative analysis.
Fischer, F. (2003) Beyond empiricism: Policy analysis In Realizing Social Science Knowledge, eds
and deliberative practice. In Deliberative Policy B. Holzner, K. D. Knorr and H. Strasser,
Analysis. Understanding Governance in the Network pp. 265–77. Wien-Würzberg: Physica-Verlag.
Peter-3383-Chapter-05.qxd 6/8/2006 7:32 PM Page 112
Hjern, B. and Hull, C. (1984) Going interorganisa- Lindblom, C. E. (1990) Inquiry and Change. The
tional: Weber meets Durkheim. Scandinavian Troubled Attempt to Understand and Shape
Political Studies, 7(3), 197–212. Society. New Haven: Yale University Press.
Hjern, B. and Hull, C. (1987) Helping Small Firms Lindblom, C. E. and Cohen, D. (1979) Usable
Grow: An Implementation Approach. Beckenham, Knowledge. Social Science and Social Science
Kent: Croom Helm. Solving. New Haven: Yale University Press.
Hood, C. (1991) A public management for all sea- Lipsky, M. (1980) Street-Level Bureaucracy: Dilemmas
sons? Public Administration, 69 (Spring), 3–19. of the Individual in Public Services. New York:
Hooghe, L. and Marks, G. (2001) Multi-Level Russell Sage.
Governance and European Integration. Boulder: Lowi, T. J. (1964) American business, public policy,
Rowman & Littlefield. case studies, and political theory. World Politics,
Hoppe, R. (1999) Policy analysis, science and poli- 16 (July), 686–715.
tics: From ‘speaking truth to power’ to ‘making March, J. G. and Olsen, J. P. (1989) Rediscovering
sense together’. Science and Public Policy, 26(3), Institutions. The Organizational Basis of Politics.
June, 201–10. New York: Free Press.
Kickert, W. J. M., Klijn, E.-H. and Koppenjahn, J. F. M., Mazmanian, D. A. and Sabatier, P. A. (1983)
eds (1997) Managing Complex Networks. Strategies Implementation and Public Policy. Glenview, IL:
for the Public Sector. London: Sage. Scott & Foresman.
Kooiman, J. (1993) Social-political governance: Miller, H. T. (2002) Postmodern Public Policy.
introduction. In Modern Governance: New Albany, NY: State University of New York Press.
Government-Society Interactions, ed. J. Kooiman, Mills, C. W. (1956) The Power Elite. New York:
pp. 1–6. London: SAGE Publications. Oxford University Press.
Kuhn, T. S. (1962) The Structure of Scientific Oakerson, R. J. et al. (1987) How Fragmentation
Revolutions. Chicago: Chicago University Press. Works – St Louis Style. ACIR: Washington DC.
Lakatos, I. (1974) Falsification and the methodology OECD (2001) Citizens as Partners. Information,
of scientific research programmes. In Criticism Consultation and Public Participation in Public
and the Growth of Knowledge, ed. I. Latakos and Decision-Making. Paris: OECD.
A. Musgrave, pp. 91–196. Cambridge: Cambridge Olsen, J. P., ed. (1978) Politisk organisering. Oslo:
University Press. Universitetsforlaget.
Lehmbruch, G. (1991) The organization of society, Ostrom, E. (1990) Governing the Commons: The
administrative strategies, and policy networks. In Evolution of Institutions for Collective Action.
Political Choice. Institutions, Rules and the Limits of Cambridge: Cambridge University Press.
Rationality, eds R. Czada and A. Windhoff-Héritier, Ostrom, E. (1995) Self-organization and social capital.
pp. 121–58. Frankfurt am Main: Campus. Industrial and Corporate Change, 4(1), 131–59.
Lehner, F. (1991) The institutional control of orga- Ostrom, E., Parks, R. and Whitaker, G. (1978)
nized interest intermediation: A political perspec- Patterns of Metropolitan Policing. Cambridge,
tive. In Political Choice. Institutions, Rules and MA: Ballinger Publishing Company.
the Limits of Rationality, eds. R. Czada and Ostrom, V. and Ostrom, E. (1977) Public goods
A. Windhoff-Héritier, pp. 233–56. Frankfurt am and public choices. In Alternatives for Delivering
Main: Campus. Public Services, ed. E. S. Savas. Boulder, CO:
Lindblom, C. E. (1959) The science of muddling Westview.
through. Public Administation Review, 19, 79–99. Patton, C. V. and Sawicki, D. S. (1993) Policy Analysis:
Lindblom, C. E. (1965) The Intelligence of Democracy. Concepts and Practice. Englewood Cliffs, NJ:
Decision Making Through Mutual Adjustment. Prentice-Hall.
New York: The Free Press. Pedersen, O. K. and Nielsen, K. (1988) The Negotiated
Lindblom, C. E. (1977) Politics and Markets. Economy: Ideal and History. Scandinavian Political
New York: Basic Books. Studies, 11(2), 79–101.
Lindblom, C. E. (1988) Bargaining: The hidden Pennings, J. M. (1980) Interlocking Directorates. San
hand in government. In Democracy and Market Francisco: Jossey Bass.
System, C. E. Lindblom, pp. 139–70. Oslo: Pollitt, C. (2003) The Essential Public Manager.
Norwegian University Press. Maidenhead: Open University Press.
Lindblom, C. E. (1988) Introduction. In Democracy Pressman, J. L. and Wildawsky, A. B. (1973) Imple-
and Market System, C. E. Lindblom, pp. 9–21. mentation. Berkeley, CA: University of California
Oslo: Norwegian University Press. Press.
Peter-3383-Chapter-05.qxd 6/8/2006 7:32 PM Page 113
Pressman, J. L. and Wildawsky, A. B. (1979) Schön, D. A. and Rein, M. (1994) Frame Reflection.
Implementation. Berkeley, CA: University of Toward the Resolution of Intractable Policy
California Press. Controversies. New York: Basic Books.
Putnam, R., Leonardi, R. and Nanetti, R. Y. (1993) Scott, R. W. (1995) Institutions and Organizations.
Making Democracy Work: Civic Traditions in Modern London: SAGE Publications.
Italy. Princeton, NJ: Princeton University Press. Selznick, P. (1957) Leadership in Administration.
Ranney, A. (1968) Political Science and Public Policy. New York: Harper & Row.
Chicago: Markham. Selznick, P. (1996) Institutionalism “old” and “new”.
Rhodes, R. A. W. (1979) Central-Local Government Administrative Science Quarterly, 41, 270–77.
Relationships. London: Social Science Research Sharkansky, I. and Hofferbert, R. I. (1969) Dimensions
Council. of State Politics, Economics and Public Policy.
Rhodes, R. A. W. (1986) The National World of Local American Political Science Review, 63, 867–79.
Government. London: Allan & Unwin. Simon, H. A. (1945) Administrative Behavior.
Rhodes, R. A. W. (1997) Understanding Governance. New York: Free Press.
Policy Networks, Governance, Reflexivity and Sjöblom, G. (1968) Party Strategies in a Multiparty
Accountability. Buckingham: Open University Press. System. Lund: Studentlitteratur.
Rhodes, R. and Marsh, D. (1992) New directions in Sørensen, E. (1997) Democracy and empowerment.
the study of policy networks. European Journal of Public Administration, 75(3), 553–68.
Political Research, 21, 181–205. Stone, D. (1988) Policy Paradox and Political Reason.
Roe, E. (1994) Narrative Policy Analysis. Theory and Glenview, IL: Scott, Foresman and Co.
Practice. Durham, NC: Duke University Press. Streeck, W. and Schmitter, P. C., eds (1985) Private
Rosenau, J. N., ed. (1969) Linkage Politics. Essays on Interest Government. Beyond Market and State.
the Convergence of National and International London: Sage Publications.
Systems. New York: Free Press. Truman, H. S. (1951) The Governmental Process.
Sabatier, P. A. (1986) Top-Down and Bottom-Up Political Interest and Public Opinion. New York:
approaches to implementation research: a critical Alfred E. Knopf.
analysis and suggested synthesis. Journal of Public Van Horn, C. E. (1979) Policy Implementation in the
Policy, 6(1), 21–48. Federal System: National Goals and Local
Scharpf, F. W. (1991) Political institutions, decision Implementors. Lexington, MA: Lexington Books.
styles, and policy choices. In Political Choice. Weimer, D. L., Vining, A. R. and Vining, A. (1998)
Institutions, Rules and the Limits of Rationality, Policy Analysis: Concepts and Practice. Englewood
ed. R. Czada and A. Windhoff-Héritier, pp. 53–86. Cliffs, NJ: Prentice-Hall.
Frankfurt am Main: Campus. Wildawsky, A. (1979) Speaking Truth to Power. The
Scharpf, F. W. (1997) Games Real Actors Play. Actor- Art and Craft of Policy Analysis. London:
Centered Institutionalism in Policy Research. MacMillan.
Boulder, CO: Westview Press. Williamson, O. (1975) Markets and Hierarchies:
Schmitter, P. C. (1974) Still the century of corpo- Analysis and Antitrust Implications. New York:
ratism? The Review of Politics, 36, 85–131. Free Press.
Peter-3383-Chapter-05.qxd 6/8/2006 7:32 PM Page 114
Peter-3383-Chapter-06.qxd 6/8/2006 7:47 PM Page 115
6
Concepts and Theories of Horizontal
Policy Management
B. GUY PETERS
Coordination and coherence are familiar the vertical nature of governing has changed
themes in the discussion of shortcomings of substantially in the UK (6, 2004). The Finnish
public administration and public policy. government has developed an elaborate system
Governments have long sought to discover for managing cross-cutting policy priorities
means of making the policies adopted in one (Peters, 2006). Likewise, the Canadian govern-
department or agency correspond with, or at ment has come to recognize the need for
least not conflict with, those adopted in other enhanced coordination but there has not yet
departments. Likewise, governments have been the level of change desired, despite invest-
sought for mechanisms to ensure that there are ment of a good deal of energy by senior offi-
not major lacunae in their policy regimens, so cials through a variety of organizations and
that all potential clients are served and citizens procedures. We could go on adding examples
do not fall through holes in the various safety of government attempts to create more effec-
nets of government, or escape adequate eco- tive horizontal structures but the general point
nomic regulation. Unfortunately, as Pressman remains that governments are faced with sig-
and Wildavsky (1984) argued some years nificant challenges in creating greater policy
ago, much more has been said about creating coherence.
coordination than actually has been done Asian governments are no different from
about it, and coordination remains a principal those on other continents in confronting the
“philosophers’ stone” in the analysis of good challenge of coherence. Attempting to typify
public administration (Jennings and Crane, government in Asia is in some ways ridiculous,
1994). given the diversity of governing forms and
It is indeed difficult to do anything about experiences, so we will need to examine differ-
coordination, given the way in which govern- ences perhaps more than similarities. For some
ments tend to be structured and given the Asian governments the need to create coordi-
entrenched patterns of thinking about public nation may be more pressing for Asian govern-
policy and about governing. The Blair govern- ments, given that the lower levels of economic
ment in the United Kingdom, for example, has development produce fewer slack resources
made a great ruckus about creating “joined up to be wasted on redundant programs. On
government” but there is no real evidence that the other hand, some governments in Asia
Peter-3383-Chapter-06.qxd 6/8/2006 7:47 PM Page 116
may have greater capabilities for producing within government. Perhaps most importantly,
coordination, given the existence of strong the vertical character of policy and administra-
political parties (China) or strong central agen- tion is maintained by clientele politics, as
cies within the bureaucracy (Japan). Likewise, well as by the desire of government to serve
the absence of strong civil society institutions in particular components of the public. Clientele
most Asian countries (Callender and Johnston, groups that are involved in the public sector
1998) eliminates another source of effective if often believe that they have the right to their
informal coordination of public programs. own organization to serve them. Government
organizations that stray too far from serving
their clientele encounter the risk of losing nec-
MANAGING HORIZONTAL GOVERNMENT essary political support. Sometimes maintain-
ing the separate organizations to serve the
clientele appears important, even if the ser-
Governments need to find mechanisms to man-
vices themselves might be retained, or even
age more effectively in the horizontal as well as
enhanced, in a larger and more heterogeneous
the vertical direction. Reforms such as those
organization.1
associated with the New Public Management
The organizations serving clients have many
have improved vertical management, but if any-
of the same interests in maintaining their
thing have weakened the horizontal manage-
exclusive position as do the clientele groups,
ment capacity of governments (see below).
and perhaps even more. Bureaucratic organi-
There are any number of examples of gover-
zations want to maintain their client base and
nance failures resulting from inadequate policy
their close relationships with their clients in
coordination and a reluctance to expend the
order to maintain their budgets and their
political capital necessary to manage govern-
influence within political circles. In this part of
ment in a more horizontal manner. These
the game of bureaucratic politics clients are a
failures include excessive costs because of dupli-
resource and a source of power. Therefore,
cation of programs, firms having to go to
coordinating programs, and possibly diluting
numerous different regulatory organizations in
that relationship between clients and the orga-
order to get needed permits to go into business,
nization, is usually not considered good poli-
“poverty traps” for less advantaged citizens
tics in these circles.2 Somewhat paradoxically,
resulting from an unwillingness to coordinate
some research has indicated that coordination
taxation and expenditure programs, and citi-
with less-closely allied activities is easier than
zens not receiving needed services because their
with more similar programs – the more
particular set of socio-economic characteristics
remote types of programs are not conceived of
were not covered by a patchwork of laws
as being the potential threat that the more sim-
attempting to provide for all groups within the
ilar programs are. In coalition governments
society. All these failings result in governments
these differences among ministries may be
costing more, and/or providing less high-
accentuated by ministers from different parties
quality services, than they might under a more
controlling ministries that need to cooperate.
coordinated system.
Although it is easy to characterize bureau-
cratic politics in the above totally self-serving
The vertical nature of government manner, we should also remember that most
employees of public programs believe in the
Politically and administratively public policy benefits of their program for clients. The orga-
appears to function better vertically, so that the nizations therefore argue (and often believe)
“stovepipes” that define policy within govern- that they protect their programs as much in
ments are perpetuated and reinforced. It is the perceived interest of the clients as that of
crucial to recognize that this vertical structur- the organization itself. This belief in the cen-
ing of the public sector is as much a function trality and efficacy of their programs by
of political as administrative characteristics employees of public organizations makes
Peter-3383-Chapter-06.qxd 6/8/2006 7:47 PM Page 117
achieving coordination more difficult, given but many of these same vertical patterns emerge
that members of an organization believe that without the peculiar features of American gov-
they are in actuality supplying what the citizen ernment.3 Indeed, the pressing need to manage
needs. Members of a public organization may government in a more horizontal fashion
realize that their clients do require a variety of appears in virtually every country for which
services, but may still believe that the most effi- I can find information. Thus, even in
cient way to provide for those needs is to place Scandinavian countries with corporate pluralist
additional funds and responsibility within the structures and networks surrounding their
one organization. ministries (Olsen, 1987), coordination often is
Further, in addition to the commitments to identified as a significant administrative issue.4
serve their clients, the vertical nature of These systems are cooperative and to some
administrative structures is reinforced by the extent network based in how they administer
different types of expertise held by different programs (Marton, 2000), but still have diffi-
ministries, and the linkage of those organiza- culty in working across ministries, or at times
tions with different professional networks. As working across divisions within the same min-
Martin Painter (1987, 12) put it: istry. On the latter point, one common strategy
… powerful political and intellectual reasons contribute for attempting to improve horizontal manage-
to perpetuating and reinforcing the self-containment of ment has been to create “superministries”
functional and organizational compartments. Each within which numerous closely related pro-
policy sector – health, transport, town planning indus- grams are all included. In many cases all this
trial development, employment and so forth – exists in
a jurisdictional shelter of organizational structures
effort has done is to move the coordination
where actors develop their own partial perspectives and problems within the one ministry, and to make
where specialized relationships of support and opposi- the issue less visible and therefore less likely to
tion develop connecting, bureaucratic and outside be addressed directly.5
interests in distinct arenas of sectoral policymaking.
require deciding what priorities the governing of government. It appears that the highest
system as a whole has, in contrast to the multi- reaches of government can not avoid responsi-
ple priorities that exist with each individual bility for creating and delivering integrated
program and organization. policies, especially if there is to be an attempt
Asian governments may have less capacity to to produce some redefinition of a policy area
depend upon coordination at the bottom than to include the cross-cutting dimensions. That
do governments in European or North having been said, the importance of emerging,
American systems. To depend upon lower level cross-cutting issues, and the absence of as yet
bureaucrats to coordinate, or to engage in what clear definitions of those issues, will require
is to some extent self-effacing behavior, when high levels of interaction between top and
they are not as well integrated as career struc- bottom of organizations.
tures, or may depend upon external sources of Again, the dichotomy between top-down
income to survive economically, is to ask per- and bottom-up being discussed here is to some
haps too much. Descriptions of several of these extent false. In many concrete examples of
bureaucracies (Indonesia, Philippines) indi- coordination, the need for greater cooperation
cates that there is less commitment to public among policy-making organizations at the top
service goals than might be expected for civil first becomes apparent as difficult cases pre-
servants in other political systems. Further, in sent themselves in the field. Likewise, the fail-
some systems (Thailand, India) there is a great ures of coordination at the bottom of the
deal of internal fragmentation in the adminis- structures become apparent as the results of
trative system that may make coordination policy-making and administration are assessed
more difficult. by evaluators at the centers of the political sys-
tems. As well as becoming apparent of the
problems through interactions of top and
Top-down versus bottom-up bottom, it is likely that the problems can only
be addressed through interactions up and
Related to the above question about politics down the hierarchies within government.
and administration is a second question of
whether coordination can best be achieved at
Horizontal or Vertical?
the beginning of the policy process or whether
it should be more focussed at the “bottom” of The coordination question is generally, con-
government. In other words, do the people ceptualized in terms of making government
who actually deliver services know more about function in a more “horizontal” manner, and,
those services, and their clients, than individu- indeed, many of the most important contem-
als at the top of organizational hierarchies? If porary issues concerning coordination are
they do, then it would make sense to have ser- those of working across programs within a sin-
vices coordinated around their targets, gle level of government. These problems of
whether they are individuals, organizations, or managing horizontally are compounded when
areas, rather than impose the coordination the issue of coordination among levels of gov-
from the center of government. If the bottom- ernment is added, especially in federal regimes
up approach to coordination is emphasized such as Canada, Germany, or the United States
then bargaining over coordination would be (Derlien, 1991). Even in unitary regimes, how-
done not by ministers and senior civil servants ever, many of the same inter-governmental
in central agencies but rather by lower level coordination problems among political and
administrators. administrative levels of government arise,
This latter strategy corresponds well to the albeit usually without the political intensity
“empowerment” ideologies now being imple- that can characterize federal-provincial dis-
mented within many governments (Peters, putes in a federal regime (Toonen, 1985).
1996), but it also requires the initial creation of These problems may be confounded to an even
clear policy frameworks at the highest reaches greater extent when, as in most countries,
Peter-3383-Chapter-06.qxd 6/8/2006 7:47 PM Page 120
central government ministries have field United States (Katz, 1995), then there is no
structures that do not correspond to the struc- particular cause for concern. If the central gov-
ture of sub-national government. ernment does care, as most do, then vertical
The fundamental root of the coordination coordination becomes a sine qua non for suc-
problem in federal systems is that most federal cessful horizontal coordination (Derlien,
regimes have evolved in ways that permit all 1991). The reverse is probably also true, and it
levels of government to be involved in almost would be difficult for sub-national implemen-
all policy areas. Thus, to be capable of address- tors to compensate for fundamental design
ing problems of redundancy and incompatibil- errors occurring at the central government
ity requires that all governments agree on some level (Linder and Peters, 1987). Perhaps the
basic approaches to the policy and work to most important point here is that almost all
make their means of service provision more governing now is multi-level governance,
compatible. This agreement is often easier said requiring blending not only the perspectives of
than done, but it has been done. For example, different levels of the public sector, but also
several respondents in a series of interviews on blending different functional policy communi-
coordination in Canada pointed out that the ties existing at all levels of the system.
federal government and the provinces have
come to agreements to coordinate and inte-
grate their efforts at food inspection as a pre-
cursor to further attempts to coordinate PRESSURES FOR HORIZONTALITY
policies among the levels of government. These
agreements may appear to be simple matters –
Coordination and coherence have been
negative coordination only – but should be
problems for as long as there have been gov-
seen as an accomplishment in light of the diffi-
ernments, but the need for improved coordi-
culties involved in achieving those agreements.
nation appears to be more manifest in the early
One respondent pointed out that these agree-
21st century than in the past. This emphasis on
ments were gained only after days of discussion
coordination appears to have arisen for a num-
and then had to be validated at the level of the
ber of reasons. These pressures for horizontal
provincial prime ministers. Further, these verti-
government reflect political demands for a
cal coordination concerns were compounded
range of actors, including changes in political
by the need to coordinate food safety issues
ideologies about the role of public sector. In
across various ministries, such as agriculture,
addition, administrative reforms over the past
trade and industry, and increasingly foreign
several decades have resulted in the need for
affairs. Thus, one of the more common func-
increased coordination interventions in order
tions of government – ensuring that the food
to address new problems that have emerged as
sold in the market – is safe for the consumer
a result of attempts to solve problems of effi-
actually involves a significant level of inter-
ciency and effectiveness in government.
organizational and intergovernmental bargain-
ing to become compatible.6
Further, in most federal systems the central Fiscal pressures
government utilizes sub-national governments
to implement many or most of their policies. Perhaps the dominant pressure for enhanced
This implementation strategy means that, even organizational coordination in the public sector
if policies are effectively coordinated in the comes from the need to save public money. For
national capital, that integration may fall apart both political and strictly financial reasons gov-
once those policies begin to be implemented. If ernments have less money to spend than in the
the central government is not particularly con- past and must attempt to control public spend-
cerned about how their policies are imple- ing as effectively as possible. One way of con-
mented, as appears true under the block grant trolling expenditures is through eliminating
provisions becoming so common in the redundancy and ensuring that services will be
Peter-3383-Chapter-06.qxd 6/8/2006 7:47 PM Page 121
provided in the most cost-efficient manner. The accomplish their goals. These programs are
examples of contradictions and duplications in never easy tasks to implement, and the mecha-
public programs are familiar and make good nisms as yet in place are often inadequate to
copy for the exposes of inefficiency and outright create genuine strategic management. Still,
incompetence in the public sector. Further, for these represent movements in the appropriate
governments intent on saving money, eliminat- direction for enhanced policy coordination. As
ing redundant programs is a way to do so with- governments continue to develop programs of
out necessarily reducing the level of services performance management and stress the
being delivered to the public. results of their programs, then questions of
Although familiar to people inside and out- the contributions of numerous programs to
side of government, the existence of these the achievement of the desired outcomes will
coordination problems and their fiscal impacts become more apparent.
are often exaggerated for political purposes.
These coordination failures certainly do exist,
Issues
but generally represent very minor amounts of
money, compared with the political benefits
The issues that governments must now con-
that the programs involved may generate for
front also call for greater attention to horizon-
their governments. Further, coordination itself
tal management. Governments have invested a
is far from a costless activity. To impose coor-
great deal of effort at finding ways to make
dination on an existing ministerial structure
individual programs work more effectively, but
requires utilizing resources to monitor over-
the policy issues that are emerging tend to
laps and perhaps even more political capital to
transcend the usual boundaries of those min-
impose sufficient political “clout” to eliminate
istries and programs. For example, a principal
those overlaps. Creating coordination may also
economic issue confronting contemporary
require creating new programs to close gaps in
governments has been defined as “competitive-
program structures, and with that will generate
ness”. This is certainly an economic issue but
increased expenditures by government.
also involves education, labor, social policy,
regulation and a host of other policy consider-
ations. Likewise, issues are being defined in
Strategic management
terms of client groups – women, the elderly,
immigrants – whose needs cut across conven-
One positive consequence for coordination
tional boundaries of ministries and require
arising out of the management reforms during
greater integration of existing programs. There
the past several decades is the emphasis on
is every reason to expect issues to continue to
strategic management and the selection of
defy the boundaries of conventional minister-
clear objectives for the public sector. One of
ial structures, so that the needs for coordina-
the first things that any politician or adminis-
tion are not likely to diminish. Likewise, it does
trator engaged in such an undertaking dis-
not seem that simply moving “boxes” in order
covers is that most of the important strategic
to capture a particular set of problems will
objectives for governments cut across con-
solve the problems – the next issue may only
ventional organizational boundaries and
cut across the new administrative boxes that
therefore require working horizontally across
have been created to solve the previous
the “stovepipes” of government. So, exercises
problem.
such as implementing Strategic Results Areas
in New Zealand, Strategic Portfolios in Finland
and (to a more limited degree) the Govern- Earlier administrative reforms
ment Performance and Results Act in the
United States7 require those governments to Although some earlier reforms have empha-
think collectively and horizontally about what sized strategic management and integration of
they want to do, and how they are going to policies, a more common pattern of reform has
Peter-3383-Chapter-06.qxd 6/8/2006 7:47 PM Page 122
been to decentralize, devolve and disaggregate and therefore greater need for horizontal
the public sector (deMontricher, 1999; Peters government structures.
and Pierre, 2000). These reforms may have The impact of international forces on coor-
generated some efficiency benefits for the dination requirements is nowhere clearer than
public sector (although those benefits are often in the European Union. These fifteen countries
difficult to demonstrate) but they have made are finding that they must develop coordina-
governments that were already fragmented tion mechanisms in their national capitals, as
vertically even more diverse and fragmented well as in Brussels, if they are to be successful
institutions. As well as creating autonomous or participants in the game of European policy-
semi-autonomous agencies within the public making (Kassim, Peters and Wright, 2000;
sector, these reforms have created a number of Kassim, Menon, Peters and Wright, 2001).
organizations – referred to as quangos and Although some countries have chosen not to
quagos – that have some characteristics of both invest heavily in coordination activities and to
public and private sector organizations. permit ministries to bargain relatively auto-
Therefore, a subsequent round of reform has nomously in a variety of settings, the typical
been instituted that stresses the need for response to EU membership has been to create
restoring some greater integration among significant structures for coordination. Some,
the component structures of the public such as the United Kingdom, have regarded
sector, and creating greater coherence. These policy coherence vis-à-vis Europe as essential
reforms include strengthening central agen- to their capacity to extract as much as possible
cies, strengthening offices of presidents and from the policy system, and also to be able to
prime ministers (Peters, Rhodes and Wright, defend the national interest effectively in what
2000) and creating “super-ministries” that pull is often assumed by the British to be a relatively
together a range of relevant ministries. In some hostile environment.
instances the simplest response has been to
reinstitute the traditional ministerial struc-
tures and reduce emphasis on decentralization. Cross-cutting concerns
MANAGING HORIZONTAL GOVERNMENT acceptable price level for both. Thus, while
most of the means of producing coordination
that we discuss conventionally require the
The above discussions of the nature of policy
imposition of coordination on government
coordination and the contemporary motiva-
structures by law and authority, the market
tions for enhancing coordination then lead
approach to coordination would substitute
rather naturally to question how this Holy Grail
more indirect and virtually autonomous
of administration can be created in the real
modes of coordination. The actors will be
world of governing. The conventional answer
involved with each other for purposes of
would be to impose greater control from the
mutually advantageous exchange, and their
upper reaches of government. This is certainly
involvement will be episodic and partial, rather
one solution, and often an effective one, but it
than the more comprehensive style of interac-
is by no means the only available solution to the
tion typical of actors entirely within the public
problem. The general style of governing has
sector itself.
changed in the majority of industrialized
In the case of coordination through markets
democracies, with much greater reliance on
within the public sector, the necessary bargain-
indirect instruments and diminished reliance
ing may come about, with money being the
on direct government provision and command
medium of exchange. One clear example is the
and control regulations. As those changes have
creation of internal markets (Jerome-Forget,
occurred so have changes in the ways in which
White and Wiener, 1995; OECD, 1993) as a
coordination is approached. Therefore, we will
means of coordinating actions and imposing
look at markets, networks and hierarchies as
market discipline on organizations that other-
alternative means of achieving coordination.
wise would be governed strictly through hierar-
This classification of social processes has by
chy. The assumption of internal markets is that
now become quite conventional in the social
there are some components of service delivery
sciences (see Thompson, Frances, Levacic and
that can be conceptualized as “selling” their ser-
Mitchell, 1991) but is still a useful scheme
vices, while other actors within government are
through which to approach issues of categoriz-
the “buyers” of those services. These artificial
ing and assessing coordination mechanisms.
markets function as mechanisms of getting
Further, this threefold classification emphasizes
these actors together in the most efficient man-
the extent to which contemporary governance
ner, without having to use formal authority.
ideas have moved away from thinking about
Also, we should remember that, to some extent,
hierarchy as the only, or even the dominant,
the budgetary process has always been some-
approach to coordination (not to mention
thing of a coordination process based on
other aspects of governing (Pierre and Peters,
money, but the increasing reliance on internal
2000)).
markets has made the role of money and
exchange in coordination more explicit than in
Markets conventional patterns of governance.
Likewise, contracts among public organiza-
The economic theory of markets assumes that tions are increasingly being employed as a
coordination will occur almost automatically if means of coordinating their activities, replac-
competitive forces are permitted to function ing previous coordination through hierarchy
without interference. The “hidden hand” that with mutually acceptable “deals” (Fortin,
is assumed to function in other aspects of 2000). Contracts in the public sector combine
market relationships is also expected to oper- some features of markets with an otherwise
ate in the coordination of programs. Indeed, in legally based instrument (Peters, 2002). As
many ways, markets are fundamentally institu- with internal markets the bargaining in inter-
tions for coordination, with the institution nal contracting is also “quasi”, given that it is all
presumed to assure that sellers and buyers will public money and there may be a legal require-
find each other, and find each other at an ment for the delivery of the service.9 Further,
Peter-3383-Chapter-06.qxd 6/8/2006 7:47 PM Page 124
contracting in the public sector may be perhaps the tops of hierarchies responsible for
restricted in the number of possible bidders implementing the law. The very decentraliza-
and in the range of alternative patterns of tion that makes markets so valuable in some
service delivery possible under the law. Despite situations also may limit their effectiveness.
those constraints, contracting is a market- Those limits on markets may be especially
based idea for coordination that may be damaging for policies that contain a strong
applicable to a wide range of service areas. legal or entitlement basis for citizens, and
Finally, the market can be utilized to coordi- hence will not be effective for the vast range of
nate economic policy, especially within the social policies in which government is placed
European Union. The increased linkage of in the position of providing services to clients.
these countries and their fiscal and monetary On the other hand, market-based instruments
policies means that the operations of the such as taxation may be able to replace com-
market places pressures on all the countries to mand and control regulation in a variety of
make their policies compatible (Jacquet, 1998). areas.11
Further, the Maastricht Treaty and the
Amsterdam Treaty place strong demands on
the member governments to coordinate their Networks
fiscal policies. Arguably, some of the same
pressure for coordination is arising, perhaps Networks are another bargained mechanism
less directly than in the EU, in all the industri- for producing policy coordination. Rather
alized democracies as they must compete with than the exchange relationships implied by the
one another in an international marketplace. market, networks themselves are defined much
This coordination may extend not just to mon- as individual organizations would be, though
etary and fiscal policy but also to tax policy patterns of interaction. For organizations
(see Hallerberg and Basinger, 1998) social those interactions occur primarily among
policy (Adema, 1997) and a range of other individuals, while for networks they are among
public policies contributing to economic other organizations as well as individuals.
competitiveness.10 Further, the close financial Networks have many virtues as mechanisms
connections now are necessitating closer coor- for coordination, and to some extent depend
dination of financial and even criminal laws upon natural patterns of interaction that
(Laronche, 2000). emerge among organizations and individuals
In summary, the market does provide some concerned with the same policy issues. These
instruments to promote the coordination of may be “epistemic communities” (Zito, 1999)
policies, but it may not be an all-purpose solu- defined by common intellectual position and
tion. For the market to be effective there must common patterns of training. Networks
be something for the participants to exchange, depend upon the interests and commitment of
and that is not always the case in public pro- individuals and groups to be successful; most
grams. Thus, this approach may work when the of these participants (inside and outside the
programs involve goods and services that are public sector) want to do their jobs as well as
in principle marketable, but almost certainly possible and find networks convenient for
not for all. Not all relationships among multi- enhancing their effectiveness (Chisholm,
ple organizations can be coordinated effec- 1989).
tively through markets and exchange. In some The term “network” is to some extent a
instances there are mutual complementary short-hand term for a variety of patterns of
goals, rather than the somewhat contradictory interactions between state and society, and
goals implied in market exchanges. Further, within the public sector itself (Olsen, 1987;
the goals that might be achieved through Kickert, 1995). In general, these network rela-
mutual adjustment among the interested par- tionships involve the State relinquishing some
ties might be different from those sought of its authoritative powers in order to achieve
by the legislators who wrote the law, or even greater agreement among the interested and
Peter-3383-Chapter-06.qxd 6/8/2006 7:47 PM Page 125
affected parties within a policy area (Marin, rather than organizations in the network
1990). This is a natural mechanism for coordi- solving problems among themselves. This is
nation, given that bargaining tends to create one emerging role for central agencies that
agreement on proposals and tends also to may eschew their traditional role of imposing
involve a wide range of actors. Network bar- coordination in favor of bargaining over issues
gaining can be effective at both the formula- with a range of organizations. Further, certain
tion and implementation stages of the policy public sector organizations have an emerging
process (see Considine, 1992). In both aspects role in creating coordination, particularly
of the process of governing networks can be when utilizing popular ideas and issues, partic-
used to avoid turf-battles over policies and ularly environmentalism, as a means of pro-
clients, and to create a common perception of moting common values (Doern, 1993). Ideas
the policy issues and a common reaction to the like environmentalism, even in the absence of
issues. Networks need not exist only in the an institutional basis, can be used to produce
national capital around the design of policy, coordination across programs.
but often are crucial at the lowest level. That is, Despite their virtues, networks also have
networks organized around clients (classes or some weaknesses intellectually (Dowding,
even individuals) can be the most effective 1995) and as a means for producing public
form of coordination in the social services. For sector coordination. One of these is an analytic
many Asian countries the absence of networks problem; once you have said that a network
of organizations in the civil society that exists in a policy area, what do you say next? It
can provide the external network may be a seri- is difficult to argue that networks do not exist,
ous impediment to using non-governmental but it also appears difficult at times to say much
methods for coordination. Perhaps especially more about them and to use them in any pre-
important for coordination and coherence is dictive manner. This weakness is especially true
the relative absence of social groups that cut given that networks can have rather different
across conventional policy sectors. internal dynamics. For example, Paul Sabatier
Professionalism creates a ready-made net- (1988) conceives of multiple networks existing
work for coordinating some types of public around many policy areas, with the principal
policies. One virtue of the professions in dynamic being conflict over the definition of
their classical definition (see Wolgast, 1992) is the policy problem and over finding the appro-
their function as a reference group for their priate solution to the issue. This clash of ideas
members and, in most instances, professionals is, however, a form of coordination, since it
will have their own network of fellow profes- tends to eliminate conflicting ideas about
sionals that can supplement the networks cre- policy and with that conflicting and probably
ated through the organization itself. Although wasteful duplication. Still, the behavior of net-
in many ways beneficial, professional networks works, especially those structured without a
also can limit coordination. In the first place, central position to government, may be too
these networks are relatively closed to out- indeterminate to permit government to be par-
siders, so that there is less capacity for objective ticularly effective in coordinating programs.
scrutiny of policies than is true for other forms
of decision-making. In addition, each profes-
Interest groups
sion tends to define problems and solutions in
their own terms so that there may be very Political groups advocating the interests of
effective coordination within each profession those segments of society presenting govern-
and therefore (usually) within each single pro- ment with cross-cutting policy issues can func-
gram, but coordination across programs may tion as a means of identifying needs and
actually be more difficult.12 pressing for their solution. In many countries
Other networks could be structured more the target populations for major cross-cutting
vertically, with most interactions being upward policy issues – the elderly, women – are well-
to the relevant government organization, organized and are positive political symbols.
Peter-3383-Chapter-06.qxd 6/8/2006 7:47 PM Page 126
Other populations, however, are less well- encounter greater difficulties in coordinating
organized and in the case of immigrants are policies than will other countries. Civil service
often conceptualized as pariah groups with few systems such as those of the United Kingdom
political rights and resources. Depending upon and other Westminster democracies, in which
the political power of the groups requiring there are relatively frequent movements
service may not be the most effective means of among departments as a civil servant works his
generating coordination and policy coherence. or her way up through the hierarchy during
For example, one of my British respondents their career, should produce a somewhat better
pointed to the recent attempts to coordinate possibility of adequate policy coordination.
and integrate government responses to racial Civil servants who have worked in a variety of
attacks in British cities. In this case the leadership different programs should have a better idea of
had to be from within the bureaucracy itself – the perspectives of other departments, and
the Home Office and the Crown Prosecution tend to have a sense of belonging to a govern-
Service – as opposed to coming in response to ment rather than to a particular organization
the political power of the groups in question. within that government. They should also have
The various immigrant groups and ethnic a better idea of the total range of services deliv-
organizations simply did not have sufficient ered by government and how they could be
legitimacy and political “clout” to provoke the made available to clients.
desired response from the political system Canadian respondents in a set of interna-
themselves. tional interviews about coordination expressed
Interest groups may have many of the same growing concern about the loss of this impor-
problems as political parties in coping with tant informal mechanism for generating coordi-
cross-cutting issues. Many political parties nation within the public sector. They noted that
work with particular definitions of the issue the down-sizing of the public service is reducing
areas that have been functional for them in the the opportunities for movement within govern-
past but which have outlived their utility, or ment so that individuals tend to remain in one
which are not widely shared by other actors post for much longer. Their vision of what gov-
involved in the issue areas. To be successful ernment does and is about has narrowed
they may have to broker deals with other accordingly. Further, the increasing technical
groups with complementary if not contradic- content of most programs means that greater
tory definitions of issues (Sabatier, 1988; Page, expertise is required, and with that individual
2003). This need to compromise and negotiate civil servants may have fewer fungible skills that
often contrasts with their need to serve their can be applied in other settings. While the public
members directly. That service may be ori- service may be becoming more expert it also
ented toward the separation of separate may be lessening the chances of effective coordi-
programs for their constituents rather than nation from within the public service itself.
accepting the interdependencies among poli- The absence of a career, professional civil
cies and issues. service may be an impediment to effective
coordination within government. The prob-
lems are as significant for European and North
The civil service network
American countries that have opened their
Another important network mechanism for senior civil service to non-career appoint-
producing greater coordination within gov- ments as it is for Asian, African, and Latin
ernment is structuring the careers of civil ser- American countries that have not successfully
vants so that they have broad experience and a institutionalized civil service systems. The
broad conception of government and policy. absence of a civil service may be compensated
Countries such as the United States, Finland or for by the connections of the “in and outers”
Norway, in which civil servants spend most with policy networks that can provide them
or all of their careers within a single agency or with both substantive policy information and a
department will, everything else being equal, range of connections within government.
Peter-3383-Chapter-06.qxd 6/8/2006 7:47 PM Page 127
issues and policies are more clearly identified of control and accountability. Again, however,
at the lower echelons of government. to be fully effective, the central agencies must
Central agencies. A more general strategy for be strongly supported by the prime minister.
achieving coordination from the center of gov- Cabinet itself, especially with strong Prime
ernment is to rely upon central agencies. By Minister or Minister of Finance. Cabinet itself is
this term we refer to budgetary, policy, and another locus for the management of cross-
personnel management organizations that cutting policy issues. In some ways it is the
report directly to the chief executive, or which most logical institution to perform this task; all
are assigned principal responsibility for policy the principal actors in policy-making and
coordination and central management of service provision are represented. On the other
issues (Campbell and Szablowski, 1979; Savoie, hand, cabinet may be a place in which the min-
1995b). Examples of central agencies are the isters must protect the interests of their depart-
Treasury in Britain, the Treasury Board ments. Those interests may well not be best
Secretariat in Canada, and Departments of the served (in the short term at least) by excessive
Public Service and Ministries of Finance in a cooperation with other agencies, or by exam-
number of countries. These organizations can ining the broader implications of groups of
be employed to enforce the priorities of the policies. The cabinet can serve as the locus for
chief executive, but they also tend to develop the examination of cross-cutting issues if there
priorities and managerial styles of their own, is adequate leadership, both from the Prime
and to develop substantial power over policy. Minister and from the civil service that serves
Central agencies can play a significant role the cabinet. With that leadership there can be a
in creating coordination, but they also can capacity to redirect the discussion of issues and
generate substantial conflict with the line orga- enhance policy coordination. Even with a
nizations actually providing public services. strong prime minister, a close link with the
These frictions reflect the conflicts between minister of finance appears to be crucial in
“line” and “staff ” organizations that are typical creating coherent government, so that, as
of inter-organizational politics within the Donald Savoie (1999) said, there is a need to
public sector. The former type of organizations ensure that “no light shines between them”.
resent the power exercised by control organiza- Cabinet committees. A cabinet may be too
tions that do not directly serve the public and large an organization to coordinate programs
which, it is argued, know little about the pro- effectively. This is especially true given that
grams being delivered. Staff organizations each minister usually will feel compelled
(including central agencies) tend to believe to defend the interests of his or her own
that line agencies have extremely narrow views department, and this need may make the nec-
on policy and do not understand the need to essary cooperation difficult to obtain. Many
impose overall priorities on government. cabinets have been reduced drastically in size
The role of central agencies has been (Bouckaert, Ormond and Peters, 2000), but
increasing in most contemporary political sys- may still be too large to function effectively as
tems, despite the general Zeitgeist of decentral- a single decision-making entity. In the case of
ization and deconcentration. The political problems that are not well defined and which
dynamic has been that as programs are decen- cut across a range of ministries ministers may
tralized the conventional mechanisms of polit- feel compelled to defend the claims of their
ical control are devalued. The one major department over control of the issue, with
control instrument that remains in place is the some loss of necessary cooperation across
budget, so that ministries of finance in partic- departments.
ular become crucial in the process of control- Most cabinet systems therefore have devel-
ling administration. Further, as performance oped working “inner cabinet” systems, or some
management becomes a central component of committees within cabinet that can establish
managing the public sector, then ministries of collective priorities and coordinate policies
finance become even more central to processes across portfolios (Mackie and Hogwood,
Peter-3383-Chapter-06.qxd 6/8/2006 7:47 PM Page 129
1985). One approach to achieving this goal is to for coordinating a range of other programs,
create an overarching “priorities and planning this is unlikely to receive the same priority as
committee” within cabinet, as in Canada. This running the programs within his or her own
approach can coordinate policies across the department. In the case of a minister without
entire range of public programs, but often will portfolio assigned primary responsibility for
push too many decisions upwards to a few coordinating programs, such an individual
senior officials of government. The alternative may have more time to spend on this activity
approach is to develop a series of cabinet com- but may not have other necessary resources. In
mittees, each responsible for a particular seg- particular, cabinet ministers without a depart-
ment of policy. This approach has the mental power base may not have sufficient
advantage of bringing the relevant departmen- clout within the cabinet to bring his or her col-
tal ministers to the table to coordinate their leagues along if there is a need to coordinate
own activities. This has been particularly evi- their policies. This, however, may be counter-
dent in the budgetary process, with envelope acted by assigning politicians with strong
budgeting in Canada and Sweden as examples. political links to the prime minister these roles.
This approach also has its disadvantages. In Junior ministers. Rather than have a minister
particular, the boundaries between policy areas, accept additional responsibilities and add to an
and therefore between cabinet committees, are already extensive range of duties, governments
not always clear. The boundaries between can instead develop a system of junior minis-
policy areas may be becoming even less clear; ters that can help coordinate their ministries,
for example social policies, labor market policies, and perhaps accept responsibility for services
and even education policies, have become inter- to designated groups or for other special func-
twined through competitiveness concerns to a tions. To some degree junior ministers will
degree not previously experienced. Therefore, have some of the same problems encountered
there may be a proliferation of coordinating by ministers without portfolio. Being desig-
committees, and the consequent need to coor- nated “junior” these officials almost certainly
dinate the coordinators. have less power in government than will min-
Ministers without portfolio, or with additional isters, or probably senior civil servants. If these
coordinative portfolio. Another means of gener- aspiring political leaders are asked to coordi-
ating improved coordination within a cabinet nate a range of services and manage cross-
system is to utilize ministers without portfolio cutting issues controlled by powerful ministers
tasked to coordinate programs within a broad they may have only limited success. Further,
policy area. Another related method would be they may be placed in confrontational posi-
to assign departmental ministers additional tions with the senior ministers and may per-
coordinating portfolios. For example, in the ceive the job as a political detriment rather
Netherlands one minister has been assigned the than a step up the political career ladder.
additional responsibility for coordinating all Ministerial organizations themselves. We
programs being delivered to immigrants, as well have been discussing the need to coordinate
as programs designed to regulate their entry and across cabinet portfolios or their equivalents,
their participation in the labor market. In other but cabinet departments can themselves
cases ministers have been assigned the responsi- develop mechanisms for policy coordination.
bility for integrating the services provided to One that has been tried in a number of coun-
women, or in one case to provide a range of ser- tries is the creation of “superministries” that
vices for the middle class. would incorporate within their own structures
While this system has the advantage of des- a wide range of programs that otherwise
ignating someone to be responsible for coordi- would have to be made compatible across
nation of a policy area, it also has several departmental structures. At one extreme the
important drawbacks. The most obvious is Swiss government is limited to seven govern-
that it can overload an already busy minister. ment departments, so that if their portfolios
Further, although the minister is responsible are relatively homogenous they should be able
Peter-3383-Chapter-06.qxd 6/8/2006 7:47 PM Page 130
services for a specific target population or Task forces, working groups, etc. When
geographical area. At a minimum these organi- government is going to enter a policy area for
zations can act as advocates within govern- the first time, or when there is a great deal of
ment for the interests of those segments of the confusion about the best way to conceptualize
population. Examples of organizations of this a cross-cutting policy issue, a standard
type serving particular demographic groups response is to create a temporary “task force”
are the Administration on Aging in the United or “working group”. These are sometimes given
States, the Ministry for Family and Seniors and executive authority, for example some projets
the Ministry for Women and Youth in Germany, de mission in France or Projektgruppen in
and the former Ministry for the Middle Class in Germany, but generally these organizations are
France. Examples of these organizations serving oriented toward problem identification and
geographical areas are the “regional ministries” clarification – a central need for cross-cutting
in Canada, for example the Atlantic Canada policy issues (Timsit, 1988). A major recent
Opportunities Agency, the Ministry for example is the establishment of major agency,
Macedonia and Thrace in Greece, and the program and expenditure reviews in Canada
Ministry for the Mezzogiorno in Italy. (including one in social policy on pensions).
The development of organizations of this These appear to be very much recognitions of
type does bring attention to the needs of the need to think more broadly about the
demographic or regional groups, but it is far issues facing an aging population and the gov-
from a guarantee that those interests will be ernments that provide them services.
served in the way in which they need to be. If the cross-cutting issues can be “solved” in
These ministries and agencies often are not a limited period of time, or if a clear definition
perceived as central players in government, so of the issues can be developed in the limited
that, even though they may sit at the cabinet time span allowed for most of these special
table, they may not have much influence over organizations, then this approach is perhaps
major players such as the principal large social the most desirable manner to address the coor-
and economic ministries. In addition, these dination problem. They can provide a clear
ministries may provide some services for the focus and perhaps clear answers to a limited
target groups but still must ensure that services problem. If that success is not possible – and
provided from other ministries are compati- that is usually the case – then these organiza-
ble. In other words this may be just another tions either go out of business with little being
version of the division of services among accomplished or become simply another set of
departments. Further, as cross-cutting issues players in the complex network that will sur-
become more significant, the more traditional round these issues.
definitions and limitations characteristic of Another way to think about interministerial
existing programs may not push consideration organizations is as “virtual organizations” –
of the issues ahead quickly enough. organizations that may have no permanent
Interministerial organizations. Another obvi- structure and/or membership. This style of
ous means for coordinating the activities of organization has been advocated by some
existing programs and to explore the needs for Canadian civil servants interviewed in this pro-
new structures to cope with cross-cutting ject as a mechanism for generating coordination
issues is to develop organizations within the without creating yet another permanent struc-
interstices of existing organizations. All gov- ture. The argument is, in part, that creating
ernments have some forms of inter-ministerial another permanent organization with the goal
governance, although they differ in the extent of enhancing coordinate will itself soon require
to which those structures are articulated and additional coordination as issues change and
the power they can exercise over policies. What new patterns of interaction among organiza-
follows here is a brief enumeration of some of tion become the dominant concerns. Still,
those mechanisms. reaching agreements about when and under
Peter-3383-Chapter-06.qxd 6/8/2006 7:47 PM Page 132
what circumstances an organization will cease they could cooperate more effectively with
to exist is not the most pleasant thing for most organizations having similar objectives.
public officials – for them it is thinking about
death.
Interministerial committees. Another flexible Processes
means for attempting to deal with cross-
cutting issues is to employ committees of the The above discussion has centered on the
organizations affected. Almost all governments impacts of structural remedies for creating
use some form or another of these committees. greater effectiveness of coordination activities.
This practice is perhaps best developed in Coordination also could be enhanced by a
France with committees existing at the level of variety of procedures. Procedures may not
officials, ministers (or their cabinets), and appear to be hierarchical in the same way as
finally to coordinate between the Prime structural changes, but do definitely rely on
Minister and the President. Coordinative authority for their capacity to make organiza-
committees of this sort have also been well- tions comply. If nothing else, procedures can
developed in the Antipodean systems. Like all force organizations to consider the implica-
committees formed to link existing organiza- tions of their policy choices for other organiza-
tions, committees of this type will have a diffi- tions and for clients. Just as structures cannot
cult time in advancing the definitions of policy guarantee success in coordination, these
far beyond those that already exist. If, as processes depend upon the commitment of the
argued, there is a need in many policy areas for principal participants to the goals of coordina-
some potentially sweeping redefinition of the tion. Otherwise, the processes may only per-
issues to be considered then these committees petuate or reinforce the independence of
are unlikely to change policies significantly programs and justify that perpetuation as the
(Schon and Rein, 1994). Granted broader pow- result of careful policy analysis.
ers than is usually the case, committees of this Budgeting. Budgeting reflects the priorities
type might be able to advance more innovative of government in dollar terms. Therefore, it
ideas about policy, but would tend to be only can be a central process for improving coordi-
as effective as their most committed member. nation of government priorities and programs.
Coordinating organizations. Another approach Given the tight fiscal constraints under which
to policy coordination is to develop special governments now function, budgeting may be
organizations with the task of ensuring coordi- the most important mechanism for setting pri-
nation for clients. One example of this was the orities and coordinating activities. The goals of
Model Cities program in the United States priority setting and policy coordination can be
which, during the War on Poverty in the 1960s achieved in at least two ways. One would be
and early 1970s, sought to identify the range of through the use of relatively technocratic
services available to residents of poor inner approaches such as those associated histori-
city neighborhoods and to coordinate them in cally with program budgeting. This involves
order to provide the full range of services to assessment of the relative costs and benefits of
clients. The time at which Model Cities was in any expenditures and their relationship with
operation was in many ways very much like the other spending programs. The alternative
present, in terms of the perceived need to approach is “Star Chamber” proceedings in
rethink an area of policy and to attack social which senior political and/or administrative
policy questions differently. For a variety of officials examine expenditure requests, requir-
reasons (financial, bureaucratic among others) ing the advocates of programs to justify their
Model Cities enjoyed only limited success, but expenditures, and then impose some collective
it was one means of incorporating both some priorities on public spending.
rethinking of the problems with service deliv- Budgeting in the contemporary political
ery. In particular, it forced a number of organi- and fiscal environment implies reducing
zations delivering services to think about how spending as well as allocating resources among
Peter-3383-Chapter-06.qxd 6/8/2006 7:47 PM Page 133
competing purposes. This change in the other ends. Evaluation tends to be directed at a
culture of budgeting, in turn, tends to reduce single program rather than at complexes of
the willingness of organizations to invest in programs. Even then, it can point to programs
policy coordination. When there is reduced whose effectiveness is limited by failure to
funding organizations tend to retreat to their coordinate with other programs, or by the
“heartlands” (Downs, 1967) and do not want absence of needed programs. If evaluation can
to use resources to extend their domains or to be oriented around target populations (Rossi
help achieve broader, government-wide, goals. and Freeman, 1989), rather than to specific
The constant threat of reductions and cutbacks programs, it can be a means for pointing to the
makes agencies hunker down and wait for bet- needs for coordination. Further, the definition
ter days. On the other hand, the changing of the relevant target population should, per-
managerial culture in government also means haps, be made external to the program itself if
that it is difficult for public sector organiza- the greatest benefits for managing cross-
tions to avoid demands for greater attention to cutting issues is to be obtained (Schneider and
coordination and policy coherence. Ingram, 1993).
Regulatory review. The Office of Manage- Evaluation will tend to be less useful in the
ment and Budget in the United States exercises case of cross-cutting policy issues because the
regulatory reviews over the activities of the goals and interdependencies within the con-
executive branch (McGarrity, 1991; Stevens, stellation of policies may be less clearly estab-
1995). Whenever an agency wants to issue reg- lished than in more linearly-defined policy
ulations (secondary legislation) OMB reviews areas. Conventional program evaluation may
these regulations in terms of their compatibil- find that a program is working effectively,
ity with the program of the president, their while from a broader, systemic perspective it is
cost, and their relationship to other, existing seriously deficient. Existing social insurance
sets of regulations. This is but one of several programs, for example, may provide pensions
mechanisms that governments use to monitor to the elderly efficiently and effectively but yet
and control secondary legislation to both not address at all the range of services a gray-
ensure the protection of individual rights and ing population requires, nor effectively relate
to coordinate regulations being issued by gov- the skills of aging population to a changing
ernment organizations. labor market.
Central agencies in several other countries “Coordination comments.” In Australia the
also exercise similar forms of regulatory review procedure of “coordination comments” has
and attempts to coordinate the activities of been designed and institutionalized in order to
their bureaucracies (see Pullen, 1994). This prevent members of cabinet from proceeding
control is often not as great a problem, given with departmental policies without adequate
that in cabinet governments a good deal of this coordination with their peers. Cabinet members
regulatory clearance is done at the cabinet level are required to circulate for comment any pro-
itself, or through the Prime Minister’s office. In posals they will bring to cabinet at least several
general, the greater the autonomy granted to days prior to the meeting. Other cabinet sys-
administrative agencies, as is the case for the tems have rules to avoid surprises in cabinet,
Scandinavian countries, the greater will be the but this method in Australia goes the furthest
need to institutionalize some mechanisms of in generating coordination. Although occur-
coordinating issuance of secondary legislation. ring at a lower level, one Canadian respondent
The question then becomes whether eco- in our study pointed out that, in the large
nomic, policy or political criteria will domi- departments created after the 1993 reorganiza-
nate the coordinating decisions, and how will tion, one of the emerging forms of coordina-
the values be structured. tion among sections within some departments
Evaluation. The evaluation of public policies is a formalized comments procedure for the
can be another process for producing coordi- component organizations. Again, we can see
nation, although it is usually directed toward that creating very large structures tends to
Peter-3383-Chapter-06.qxd 6/8/2006 7:47 PM Page 134
induce managerial problems of its own, even if coordination and coherence in government.
the strategy does make some contributions to That is certainly the conventional stance in
enhanced coordination. public administration and it is for the most
Another not unrelated, procedure can be part also the most appropriate stance from my
found in the Finnish government. In that perspective on governing. On the other hand,
system every month the cabinet meets infor- however, there are some questions that should
mally in a so-called “evening school” to discuss be asked about coordination as a value in admin-
collectively a major policy issue. Given the istration, especially about the cost/benefit ratio
informality of that setting, and the absence of that program designed to enhance coordina-
formal decision-taking at the meeting, this tion may have for government. That is, there
procedure promotes full discussion of the are multiple goals that must be pursued
issues and makes substantive coordination through the public sector and coherence may
possible in a less threatening manner. The for- be only one among many.
mal decision-making for government still
must take place within the Council of State,
Is Coordination Always the Answer?
but some of the more difficult issues may be
decided ahead of time in a setting in which
We have been pursuing coordination as if it
informal bargaining rather than defense of
were always the answer to problems facing the
departmental turf can be more possible.
public sector. We should, however, at least
entertain the possibility that enhancing coor-
Summary dination and coherence is not always a positive
contribution to resolving policy problems.
The relatively greater attention to mechanisms There may be some circumstances in which
of hierarchical control in this paper does not competing and incoherent approaches are
imply any greater intellectual appeal of these functional, rather than dysfunctional, for both
mechanisms of coordination. What it does indi- government and the governed. For example,
cate, however, is the relative level of attention of even though government funds a great deal
governments to these as opposed to other forms of scientific research, it almost certainly should
of coordination. Governments continue to not attempt to impose a single line of research
develop and implement means of coordination or establish an orthodoxy (Salbu, 1994).
using their own powers, even when there are Drawing the line between funding “good
network and even market mechanisms operat- science” and establishing such an orthodoxy
ing at the same time. On the one hand this may may be difficult, but also may be necessary. For
appear wasteful, but it does indicate both that example, the “War on Cancer” in the United
governments retain rather old-fashioned com- States demonstrated that settling on a single
mitment to authority and hierarchy and that line of research too quickly can waste resources
they also recognize their legal and political with little result.
responsibilities to provide the best possible gov- In addition to research funding, there may
ernment to their citizens. Their emphasis on law be other policy areas in which coordination is
and authority also indicates that they continue not especially desirable for government. For
to think of the public as citizens as well as example, in many policy areas there is far less
merely economic consumers of services. than certain knowledge (Dror, 1992) about
how to produce desired changes in the behav-
ior of individuals in society. The debates over
welfare policy and criminal justice in the
CAVEATS
United States and numerous other countries
are extreme but not isolated examples of this
To this point in the paper I have been rather knowledge problem. Therefore, government
unambiguous in singing hymns of praise to may be well-advised at times to adopt an
Peter-3383-Chapter-06.qxd 6/8/2006 7:47 PM Page 135
explicitly experimental approach to policy and about accountability in the emerging world of
to minimize coordination and coherence. the more complex delivery of public service,
Some scholars have advocated such an exper- and if not, what principles are better suited for
imental approach (Campbell, 1982; 1988), this task?
and at times even some politicians have
argued that there is insufficient evidence to
make a long-term commitment to any partic- SUMMARY AND CONCLUSIONS
ular policy option. In addition, some program
areas can benefit from redundancy and dupli-
We began by noting that coordination is one
cation and excessive coordination can make
of the major goals of public administration,
the policy area more prone to error (Landau,
and has been so for some time. This paper has
1969).
discussed a number of the issues that arise
when that seemingly simple goal of improving
Accountability and Coordination coordination is advanced. These problems are
intellectual as well as practical, and require
We also must consider the classical question thinking about a number of alternative ways
of public accountability and the impact of of achieving the same policy and administra-
attempts to enhance coordination on the tive goals. Formal mechanisms of coordina-
capacity of governments to enforce account- tion are often the best ways of achieving those
ability. While greater coordination generally goals, but more informal methods also have
should enhance the efficiency and effective- their place, even within the public sector
ness of government programs, there are itself. The changing nature of government,
instances in which it does not, and within a with the emergence of the “New Public
complex multi-organizational policy environ- Management” and a range of other innova-
ment it may be difficult to identify where the tions in public policy and administration,
system has broken down. Accountability, at tend to place a greater emphasis on the infor-
least in its ex post facto sense, depends upon mal and market-based ideas, rather than the
the capacity of politicians and the public to formal.
identify who is responsible for any failures in One of the challenges – analytically and for
a program. The danger with coordination is government itself – is to determine when dif-
that by making everyone to some extent ferent approaches to coordination are most
responsible for programs in the end no one appropriate, and develop an understanding
is actually responsible and it becomes much of the contingencies that may be operative.
easier to evade responsibility. The three approaches we have discussed here
The dangers to accountability may arise reflect different concepts and different loca-
through several of the methods used for tions in the political process. The hierarchical
improving coordination. For example, finan- approach, the most common approach,
cial accountability becomes difficult to enforce appears crucial in program areas that are pecu-
when funds from several departments are min- liarly public, e.g. defense, foreign affairs,
gled in order to create adequate resources for revenue collection and the like. The other two
comprehensive attacks on major cross-cutting approaches appear more applicable to that
policy problems, for example drug enforce- other large range of public programs that are
ment or urban regeneration. How does gov- (at least in principle) marketable or which are
ernment ensure that money is being spent in surrounded by significant networks of private
the ways intended when it was appropriated? If actors. This is as yet a gross generalization, and
it is not, who should be held responsible for substantial refinement is needed to make it at
the misallocation? Perhaps even more funda- all useful. Still, it may point to ways of address-
mentally, are conventional ideas about parlia- ing coordination and improving both the
mentary accountability the best ways to think practice and the analysis.
Peter-3383-Chapter-06.qxd 6/8/2006 7:47 PM Page 136
NOTES REFERENCES
1. For example, President Jimmy Carter attempted to 6. P. (2004) Joined-Up Government in the Western
divide the (then) Veterans Administration by moving the World in Comparative Perspecitve: A Preliminary
educational components into the Department of Literature Review and Exploration, Journal
Education, the health and social service components into of Public Administration Research and Theory
the Department of Health and Human Services, and the 14, 103–39.
residual elements into the Department of Defense. He 6, P. (2001) Joined Up Government Around the
failed rather abjectly at this rationalization and coordina-
World, paper presented at Conference on Joined-
tion in the face of pressure from veterans’ organizations.
2. In a study I did with several colleagues over 25 years
Up Government, Oxford, October 30.
ago we found that social service agencies tended to hang Adema, W. (1997) What Do Countries Really Spend
on to clients and not to refer them for needed services. on Social Policies: A Comparative Note, OECD
This was true even in the context of the Model Cities Economic Studies 28, 153–67.
Program, an attempt to provide more comprehensive ser- Alexander, E. R. (1993) Interorganizational
vices to the residents of poor neighborhoods in American Coordination: Theory and Practice, Journal of
cities. When interviewed agency leaders argued that it was Planning Literature 7, 328–43.
too “dangerous” for them to risk losing clients. Allard, C. K. (1990) Command, Control and the
3. For example, the “Resortsprinzip” in German govern- Common Defense (New Haven: Yale University
ment is another indication of this segmentation of
Press).
government.
4. For example, in recent consulting work for the
Barrett, S. and C. Fudge (1981) Policy and Action:
Finnish government creating policy integration emerged Essays on the Implementation of Public Policy
as a dominant issue. This was true even though this gov- (London: Methuen).
ernment has a long tradition of cooperative management Bogason, P. (1991) Control for Whom?: Recent
and policy development. Adventures on Governmental Guidance and
5. One common pattern has been to create very large Control, European Journal of Political Research
ministries for health and social services. This single min- 20, 189–208.
istry may have several ministers and be responsible for up Bouckaert, G., D. Ormond and B. G. Peters (2000) A
to half of the total public budget. This creates a tremendous New Governance Agenda (Helsinki: Ministry of
managerial overload for the senior public servants, and
Finance).
may do so with little real improvement in coordination.
6. The Canadian Food Inspection Agency was created in
Callender, G. and J. Johnston (1998) Governments
1997 to bring together related activities from Agriculture, and Governance: Examining Social and
Fisheries, Health and Industry ministries. It also absorbed Economic Autonomy in Malaysia and Singapore,
some functions that had been largely provincial. Asian Journal of Public Administration 20, 151–72.
7. The Office of Management and Budget prepares each Campbell, C. and G. Szablowski (1979) The
year a government-wide set of priorities to correspond to Superbureaucrats: Structure and Behaviour in
the more specific priorities coming from the agencies. The Central Agencies (Toronto: Macmillan of Canada).
OMB document tends to be long on pious hopes and short Campbell, D. T. (1982) Experiments and Arguments,
on specifics. Knowledge: Creation, Diffusion, Utilization 3, 327–37.
8. In some ways ministries of foreign affairs have not
Campbell, D. T. (1988) The Experimenting Society,
become central agencies so much as all ministries have
developed their own international sections and interna-
in Campbell, Methodology and Epistemology in
tional political views. This is in contrast to the role of envi- the Social Sciences: Selected Essays (Chicago:
ronmental agencies. University of Chicago Press).
9. In a “real” contracting arrangement the actors involved Chisholm, D. (1989) Coordination Without Hierarchy
would have more of an opportunity to walk away if an (Berkeley: University of California Press).
acceptable deal could not be reached. In the public sector Commission on Social Justice (1994) Social Justice:
the legal mandates restrict the exit option. Strategies for National Renewal (London: Vintage).
10. There is, however, evidence that countries are still Considine, M. (1992) Alternatives to Hierarchy: The
behaving more autonomously than the proponents of Role and Performance of Lateral Structures
globalization might lead us to believe.
Inside Bureaucracy, Australian Journal of Public
11. This can be seen most clearly from the “instru-
ments” perspective on public policy (Salamon, 2000).
Administration 51, 309–20.
12. For example, in my earlier research on the Model Davis, G. (1998) Executive Coordination Mechanisms,
Cities program in the United States, professionals tended in P. Weller, H. Bakvis and R. A. W. Rhodes, eds.,
to refer clients to others in the same profession rather than The Hollow Crown: Countervailing Trends in Core
serving the “whole client” as intended by the program. Executives (London: Macmillan).
Peter-3383-Chapter-06.qxd 6/8/2006 7:47 PM Page 137
Peters, B. G. (2002) Contracts and Resource Savoie, D. J. (1999) Governing from the Centre.
Allocation in Public Governance, In A. Heritier, (Toronto: University of Toronto Press).
ed., Common Goods: Reinventing European and Schneider, A. B. and H. Ingram (1993) Social
International Governance (Lanham, Md: Rowman Construction of Target Populations: Implications
and Littlefield). for Politics and Policy, American Political Science
Peters, B. G. (2006) Horizontal Governance: The Review 87, 334–47.
Search for Coordination and Coherence (Lawrence: Schon, D. and M. Rein (1994) Frame Reflection:
University Press of Kansas). Resolving Intractable Policy Issues (New York:
Peters, B. G. and J. Pierre (2000) Administrative Basic Books).
Reform and the Civil Service (London: Routledge). Smith, G. (1991) The Resources of a German
Peters, B. G., R. A. W. Rhodes and V. Wright (2000) Chancellor, in G. Jones, ed., West European Prime
Administering the Summit: Support for Presidents Ministers (London: Frank Cass).
and Prime Ministers (London: Macmillan). Smith, M. J., D. Marsh and D. Richards (1993)
Pierre, J. and B. G. Peters (2000) Governance, Policy Central Government Departments and the Policy
and the State (Basingstoke: Macmillan). Process, Public Administration 71, 567–94.
Pressman, J. L. and A. Wildavsky (1984) Imple- Stevens, L. N. (1995) Regulatory Reform: How Can
mentation, 2nd ed. (Berkeley: University of Congress Assess the Administration’s Initiatives
California Press). (Washington, DC: General Accounting Office,
Pullen, W. (1994) Eyes on the Prize, International July 18) GAO/T-95–206.
Journal of Public Sector Management 7, 5–14. Taylor, S. (1984) Making Bureaucracies Think: The
Regens, J. L. (1988) Institutional Coordination of Environmental Impact Statement Strategy of
Program Action: A Conceptual Analysis, Inter- Administrative Reform (Stanford: Stanford
national Journal of Public Administration 11, 135–54. University Press).
Sabatier, P. A. (1988) An Advocacy-Coalition Model Thompson, G., J. Frances, R. Levacic and J. Mitchell
of Policy Change and the Role of Policy-Oriented (1991) Markets, Hierarchies and Networks
Learning Therein, Policy Sciences 27, 123–35. (London: Sage).
Salamon, L. M. (2000) Handbook of Policy Instruments Timsit, G. (1988) Les autorites independents admin-
(New York: Oxford University Press). istratives (Paris: Presses Universitaires de France).
Salbu, S. R. (1994) Regulation of Drug Treatments Toonen, T. A. J. (1985) Implementation Research
for HIV and AIDS, Yale Journal of Regulation and Institutional Design: The Quest for
11, 401–52. Structure, in K. Hanf and T. A. J. Toonen, eds.,
Savoie, D. J. (1995a) Globalization, Nation States and Policy Implementation in Federal and Unitary
the Civil Service in B. G. Peters and D. J. Savoie, eds., Systems (Dordrecht: Martinus Nijhoff).
Governance in a Changing Environment (Montreal: Wolgast, E. (1992) Ethics of an Artificial Person: Lost
McGill/Queens University Press). Responsibility in Professions and Organizations
Savoie, D. J. (1995b) Central Agencies: Looking (Stanford: Stanford University Press).
Backward (Ottawa: Canadian Centre for Zito, A. (1999) Making European Environmental
Management Development). Policy (London: Macmillan).
Peter-3383-Chapter-07.qxd 6/8/2006 7:34 PM Page 139
7
Budgeting
IRENE S. RUBIN
Political science has long wrestled with the States has certainly had its share of both of these
question of whether the characteristics of types of solutions, but is not alone in its attempt
decision-making processes affect the outcomes to find a solution for one of the central set of
of those decisions. There are usually many policy decisions for government.
other factors contributing to outcomes besides
processes, which means that any given process
may have different outcomes at different times
BUDGET RULES AS ILLUSTRATION
or in different circumstances. As a result, it has
OF PROCESS
been difficult for political scientists to demon-
strate that decision-making processes impact
outcomes. This chapter examines the question What are budget process rules? What is their
of the impact of process on outcomes, using as function? Budget process rules describe the
an example budget process in the United States order in which budget decisions will be made,
at the national level, with particular attention the timeliness of the decision-making, which
to the post 1998 changes and recent budgetary actors make which decisions, and the con-
outcomes. While this case does not provide straints under which the decisions will be
an airtight design, it does provide a kind of made. These rules provide definitions, for
natural experiment. example, how balance will be defined. Budget
Although the focus will be on the United rules determine what budget claims will be
States, there will be some comparisons with compared with what other claims, what evi-
process reforms in other industrialized democ- dence and whose evidence will be brought to
racies. All the countries in the OECD world, bear on the decision, and how much time there
and indeed in the world more broadly, are will be for deliberation on those decisions.
struggling to find budget processes that produce A good budget process is stable from year to
efficient allocation of funds but yet are not year, it provides necessary definitions that are
dominated by technical considerations. The known to and used by all parties, and it ensures
“ideas in good currency” about budgeting discipline, so that revenues do not exceed
appear to cycle between highly rationalistic expenditures over the long haul. Budgets may
solutions, such as PPBS, to simple, if effective, include borrowing, but are not balanced unless
mechanisms to reduce public expenditure there is a plan and revenue sources to pay off
(Gray, Jenkins and Segsworth, 2002). The United that debt as it comes due. To ensure discipline,
Peter-3383-Chapter-07.qxd 6/8/2006 7:34 PM Page 140
the process normally outlines steps to estimate for all the participants, who must have a chance
(and revise) revenue estimates, and breaks up to learn and use the process. The process must
pieces of the decisions on spending and gives be visible to the public.
them separate limits, so that, when added up, In evaluating the consequences of rules
they do not exceed the total revenue available. changes, one has to look initially at the rules
A good process is reasonably neutral among themselves. They can be described in terms of
policy goals and political interests, providing a how they compare to the idealized process
forum for the articulation, discussion and res- described here. For example, how neutral are
olution of necessary policy issues without bias- they with respect to outcomes? How consistent
ing the results in advance. It combines and predictable are they from year to year?
information coming up from programs and What steps are included in the budget process
professional staff with centrally determined to assure that spending is controlled by a
policy and discipline. Further, in a democracy, revenue estimate? Is there a procedure for
budget process must assure that minority par- prioritization of expenditures? Second, one has
ties are heard, and that decisions are not made to look at the changes. Did new rules supple-
that prevent different factions that might come ment or supplant the old ones? Were the old
to power in the future from also making policy rules kept but their intent evaded? Were the
decisions. rules “gamed”? Was it a minor rule that was
Maintenance of neutral rules allows the avoided or voided, or a major one, was it one
decision-making to proceed in a politically rule or a cluster of them, was the change infor-
charged environment. Without it, conflicts may mal and temporary, or formal and long term?
bring the budget to a standstill, as often occurs Budget process change is a complex variable,
in less stable democratic systems. Without but it must be understood in some detail if its
neutral rules, minorities may feel that they consequences are to be examined.
cannot win or influence decisions, and hence
cease to participate or work at disrupting the
process. Politicians are often tempted to bias
BUDGET CHANGE IN THE
the rules to facilitate some immediate policy
UNITED STATES
goal rather than maintaining neutral rules. For
example, the requirements of the stability pact
in the Eurozone have been used not only to This study focuses on the federal budget
stabilize the budget but also to disadvantage process in the United States, with particular
some social interests (Stolfi, 2006). attention to the period from 1998 to the end of
Some rules can be changed without damag- 2004. However, in order to see a rule change,
ing the collective policy framework necessary one must also examine what budget rules pre-
for democratic decision-making, others ceded the change and what outcomes were
cannot. There must be time to consider legisla- associated with the period in which the pre-
tion, there must be thoughtful prioritization, ceding process was in effect. One also needs to
there must be a role for the minority as well as know the degree to which those prior rules were
majority; there must not be decisions that lock implemented. Thus, the study of budgeting in
in current policy priorities and prevent future the United States compares three periods, one
democratically elected officials from making from 1986 to 1990, when Gramm Rudman
changes. Perhaps one exception to that general- Hollings was in place, one from 1990 to 1998
ization is the use of earmarked taxes and trust (when the Budget Enforcement Act was in
funds as a means of justifying certain types place), and the final one, from 1998 to 2004,
of taxation (Patashnik, 2000). There must be during the decline and termination of the BEA
good quality information provided in a timely and the rise (return?) of ad hoc budgeting.
fashion, to decision makers and to the media Innovations in budget practices in the United
and public. The process must be firmly enough States have been diffused widely in the past,
in place that it is predictable from year to year and some of these same ideas have been used
Peter-3383-Chapter-07.qxd 6/8/2006 7:34 PM Page 141
BUDGETING 141
in other countries to address similar issues. be demonstrated, such that the process addresses
Thus, this one case study is just that but it also issue A and the outcome for issue A changes
illustrates some generalities about budgets and when the process changes, this should be con-
the impact of rules on performance. strued as further evidence supporting a relation-
ship. Finally, if alternative explanations besides
the changes in budget process can be ruled out,
Data Sources
in part or in whole, the case would be even
stronger.
The information on budget process and imple-
A second assumption is that processes can be
mentation comes from laws, journal articles,
more or less formal, with informal processes
newspaper accounts, newsletters from public
easier to change and less durable or predictable.
interest research groups, and reports from think
Informal processes may carry out the intent of
tanks and advocacy groups. Efforts were made
the formal rules or contradict them or supplant
to draw on both sides of the political spectrum
them. Budget rules may be more or less actively
as well as some sources closer to the middle.
enforced, mechanisms of enforcement may
Data also comes from government publica-
exist, but be seldom used, or may be actively
tions, such as the Historical Tables of the U.S.
evaded, through distorted data (gaming the
Budget and budget hearings, and from on-line
system) or definitional changes or various
news magazines focusing on the federal gov-
kinds of budget gimmicks or accounting rules
ernment, such as Government Executive and
changes. The impact of budget rules on out-
The Hill.
comes depends to some extent on the degree
of enforcement in practice. Budget processes
Assumptions do not enforce themselves, they require the
consent and will of the actors to give them life
Several assumptions underlie the analysis. and force. For the purposes of this study, a shift
First, budget processes do not determine out- from rule enforcement to rule non-enforce-
comes, they influence them, making certain ment is considered a change in budget process,
outcomes easier to achieve and others harder. even without a formal, legal change.
This influence may be greater and more visible
at some times than others. Since it is difficult
to either prove or falsify influence or measure The Intervention
it in any precise fashion, the conclusions in this
essay are inferential. This study is about a natural experiment, a set of
The logic of the analysis depends on four changes in budget rules. Budget processes are
kinds of argument. First, there were three phases often overlaid on each other, rather than com-
of budget process change, each associated with a pletely removed and started afresh, but
set of outcomes. If deficits increase during the key features may be added or subtracted. All
first phase, decrease and disappear during later budget processes were overlaid on the 1921
the second, and reappear, almost instantly, in the Budget and Accounting Act, which created
third, even if there are many other things going executive budgeting, so that the president and
on during those time periods, there would be his budget office prepared a budget request for
evidence of a likely relationship between process legislative consideration. Similarly, the changes
and outcomes. Second, if the actors who want to inherent in the 1974 Congressional Budget
accomplish a particular policy goal first change and Impoundment Control Act, in which the
the process and then accomplish their goal, it Congress enhanced its ability to question that
looks as if they believe they had to change the presidential proposal and come up with its
process, formally or informally, to accomplish own if it wished, underlay the later changes.
their goal. Such behavior provides additional From the point of view of budget process, the
strength to the argument for a linkage between central feature of the 1974 Congressional
process and outcomes. Third, if mechanism can Budget and Impoundment Control Act was the
Peter-3383-Chapter-07.qxd 6/8/2006 7:34 PM Page 142
BUDGETING 143
governance processes may require just those Budget Enforcement Act quickly eroded. Spending
types of controls if they are to resist the exceeded the caps by a wide margin. The key
centrifugal pressures within the political system. features of the BEA rules were ignored or
gamed. One common practice was to treat some
routine expenditures as if they were emergen-
CHANGES AND OUTCOMES cies, since emergencies were not required to be
offset by revenue increases or other decreases in
1986 to 1990: Gramm Rudman spending. By calling the expenditure, such as the
Hollings and Deficit Reduction Targets census, an emergency (which it was not except
in political terms) budgeteers did not have to
Gramm Rudman Hollings set annual targets for find revenue to cover the expense, which then
deficit reduction, but those targets were rou- added to the deficit.
tinely missed. Efforts to avoid sequestration, the As budget scholar Allen Schick observed,
technical term for across-the-board cuts, led to
The arrival of a surplus a few years ago triggered a spend-
intense gaming of the system, pushing expendi-
ing frenzy that vitiated the discretionary spending caps
tures into following years to make the budget established by the 1990 Budget Enforcement Act and
look more balanced than it was, and then push- made a mockery of the BEA requirement that increased
ing them back when the next year arrived, for spending be offset by cuts in other spending or by
example. The legislation was not even handed, it revenue increases. In 2000 and 2001, discretionary spend-
ing soared more than $200 billion above the legal limits
approached the issue of deficits from the per-
on annual appropriations. The caps expire at the end of
spective of forcing deep spending cuts, but there 2002 – which will at least enable politicians to be more
was little consensus on which programs to honest about what they are doing. They no longer have to
cut or how deeply to cut them, so the unthink- pretend that the census and other ongoing operations of
able, the sequester process, was supposed to be government are national emergencies. (Allen Schick, The
Brookings Review, Spring, 2002. Vol. 20 No. 2 pp. 45–48.
invoked. The sequester process was supposed
The Deficit That Didn’t Just Happen: A Sober Perspective
to be neutral, in the sense of affecting all pro- on the Budget)
grams, but while it didn’t have a political bias, it
tended to cut some programs disproportionately Huge surpluses turned nearly immediately
because they hadn’t been exempted. The result into large and growing deficits. This process
was illogical, without serious prioritization. The began before 9/11/2001 and hence cannot be
process might have reduced deficits if it had in attributed solely to responses to that event.
fact been implemented, but, with widespread OMB reports in the Historical Tables of the U.S.
evasion of its rules, Gramm Rudman Hollings Budget for 2005 that the on-budget surplus for
proved unable to control deficits. fy 2000 was 86.626 billion dollars, but by 2001
From 1990 to 1998, under the Budget there was a deficit of 33.257 billion (fiscal year
Enforcement Act, the norms of budgetary 2001 spending is controlled by decision-making
balance were strengthened. Spending caps that occurs the previous year). Figures for
were generally treated seriously, forcing a deficits for 2002, 2003, and 2004 were reported
number of difficult trade-offs, many in full as 317.456 billion, 536.128 billion, and 674.766
public view. The always present temptation to billion. These numbers are so huge and growing
reduce taxes was curtailed; taxes were actually so rapidly as to suggest they are out of control.
increased to help balance the budget. By 1998, Deficits are listed in Table 7.1, along with the
with the help of a growing economy, the budget process in effect at the time.
budget was more or less balanced; surpluses As former CBO staffer Phil Joyce observed,
were projected for many years in the future. there was insufficient consensus after 1998 to
Public debate focused on what to do with the come up with congressional targets and adhere
surpluses, whether to reduce taxes or fix social to them. The lack of agreement meant that
security, or buy down the national debt. the budget process was operating without a
After 1998, because of the surpluses, the con- notional budget constraint. As a result, Joyce
sensus to adhere to the spending caps in the claimed “no one knows how much is enough – or
Peter-3383-Chapter-07.qxd 6/8/2006 7:34 PM Page 145
BUDGETING 145
Table 7.1 Deficits/surpluses, on budget, from wars in Afghanistan and Iraq were paid for
1980 through 2004 (millions) primarily through supplemental appropria-
Budget process in place year deficit/surplus tions, which were not offset by reductions in
Congressional budget 1980 −72,710 spending elsewhere, and added to the deficit,
and Impoundment Control 1981 −73,948
but they were in the range of 60 to 80 billion a
1982 −120,040
1983 −208,014 year, nowhere near the total of the deficit.
1984 −185,629 Three sources are often cited as the cause of
1985 −221,671 deficits, a slow growing economy, increases in
Gramm Rudman Hollings 1986 −237,946 spending, and reductions in tax rates. Of these,
1987 −169,298
the slow growing economy had the least impact,
1988 −193,951
1989 −205,910 because while revenue growth may have
Budget Enforcement Act 1990 −277,786 slowed down some, so did interest rates, which
1991 −321,525 lowered the costs of government borrowing.
1992 −340,463 Expenditures did increase, but not just for the
1993 −300,434
wars and intelligence efforts; entitlements
1994 −258,904
1995 −226,387 also grew rapidly, especially for Medicare and
1996 −174,061 Medicaid. Other spending increased as well.
1997 −103,322 Overall, spending increased from 18.4 percent
1998 −29,982 of GDP in fiscal year 2000 to 19.8 percent
BEA key provisions not 1999 +1,873
in 2004, a startling increase in the size of gov-
enforced 2000 +86,626
2001 −33,257 ernment spending. But this increase would
BEA expires 2002 −317,456 have been more than offset if revenues had
2003 −536,128 been allowed to remain the same percent of
2004 −674,766 (est) GDP they had been in 2000. If taxes had not
Source: Historical tables of the U.S. Budget, 2005, been reduced as a proportion of the economy,
table 1–1. there would have been no deficit in 2004,
instead there would have been a surplus of more
than 100 billion dollars (Center for American
too much – spending. And nobody knows – or Progress, An analysis of the recent deterioration
everybody knows, but nobody agrees – when of the fiscal condition of the U.S. Government.
the deficit is too large or the surplus too small” Washington, DC September 2004, p. 9, accessed
(Philip Joyce “Federal Budgeting After online at http://www.americanprogress.org).
September 11th: A Whole New Ballgame, or is Increases in entitlement spending for Medicare
it Déjà vu All Over Again?” in Public Budgeting and Medicaid should not be laid at the feet of
& Finance (Winter, 2005). the budget process, nor probably should the
To the extent that the increase in deficits increases in spending on the military post
since 1998 resulted from 9/11 and terrorism 9/11; but the reductions in taxation were made
related expenditures, one could argue that they decidedly easier by the removal of the con-
are not related to the budget process and would straint that required either spending reduc-
have occurred no matter what the budget tions or alternative revenue increases to
rules were. However, as Brian Reidl from compensate for tax reductions. These tax
the conservative Heritage Foundation has reductions contributed in a major way to the
argued, most of the spending increase since rise of deficits.
2001 was not for terrorism related expenses. While some of the spending increases were
The recent round of deficits was the result of not related to the budget process and would
tax cuts that were not offset by reductions in have occurred even if there had been no changes
spending, and by spending increases not only in the budget process, some of it may well
for the war in Afghanistan, and later in Iraq, have been related to the changing process. Many
but also for domestic spending, including observers noted the increase in the amount of
pork, which began to surge after 1998. The pork type spending, or earmarks after 1998.
Peter-3383-Chapter-07.qxd 6/8/2006 7:34 PM Page 146
Brian Reidl of the Heritage foundation reported “People say earmarks are part of the grease that makes
the number of pork projects skyrocketed from government work, and to some extent, that’s true,” said
John Cogan, a Stanford University economist and
under 2,000 five years ago [1998] to 9,362 in the former OMB deputy director who headed the initial
2003 budget. Total spending on pork projects drafting of Bush’s budget blueprint as a member of the
has correspondingly increased to over $23 transition team. “But 6,183 earmarks is an awful lot of
billion. This trend continued in the fiscal year grease.” Government Executive Magazine – 2/13/01
(FY) 2004 appropriations bills, which include Bush team eyes earmarks, may push biennial budget
plan, February 13, 2001, James Barnes, National Journal
approximately 10,000 earmarks (Another 2002: Expiration of the BEA and Ad Hoc Budgeting
Omnibus Spending Bill Loaded with Pork. The
Heritage Foundation by Brian M. Riedl When the budget enforcement act of 1990
WebMemo #377 December 2, 2003). expired in 2002, it was not renewed. Congress
Reidl described what he saw as the legions of could have fallen back on the 1974 process, but
interest group lobbyists asking for these funds, with lack of consensus on balance, tax cuts,
and a shift from program funds for general and offsets, it was difficult to put a budget
programmatic purposes, often awarded on the resolution together that both houses and both
basis of technical merit through open competi- parties could agree on. The result can best be
tions, to these earmarked projects, which he described as ad hoc budgeting, unpredictable,
argued went to the highest bidder. To finish up marked by stalemates, and noted for huge
the process, he argued, Congress then had to omnibus appropriations that no one could
appropriate more money for the initial program read in the time allotted. The omnibus appro-
and then in turn earmarked more of that money. priations were laden with pork and special
During the period of the Budget Enforcement interest legislation. War was funded by supple-
Act, while the caps were closely observed, appro- mental appropriations that added to the
priations subcommittees had to make painful deficits, and obscured the fact that the budget
trade-offs for each item of spending. Pork pro- as passed seriously and systematically underes-
jects were obvious and required a publicly timated expenditures and the size of the
acknowledged cut in some other worthy project. deficit. Deficits continued to grow, despite
Because the caps were binding, the amount of promises to curtail them.
pork was low during this period. When pork The quality of budgetary decision-making
spending no longer required an explicit trade- deteriorated in this environment. The admin-
off, the budget process made it easier to include istration supported new budget rules that were
them. Moreover, policy deadlocks, particularly one sided, that required offsets for new expen-
over whether the rules still required offsets to tax ditures or program expansions, but not for tax
reductions, delayed appropriations, resulting in reductions. This proposed rule would make it
omnibus appropriations, easy vehicles to slip easy to reduce taxes almost at will and without
pork into. The omnibus appropriations bills regard for budgetary consequences. This pro-
were often so large and so quickly passed that posed rule was attached to the House version
legislators did not have time to read them, let of the budget resolution for the 2005 fiscal year
alone pull back the pork projects from them. budget. It was not only biased toward one
The successes of the BEA exacerbated the policy, it was unlikely to actually reduce or
demand for pork, but the failures of budget control the burgeoning deficit since both
process after 1998 facilitated this growth in revenues and expenditures need to be con-
spending and moved the budget more in the trolled to reduce or eliminate deficits. The
direction of short-term, personal or partisan Senate refused to go along with the House
benefit and away from collective benefit. rules proposal and hence was unable to pass
The growth in earmarks was noted by many the budget resolution. Consideration of
other observers, and had begun well before budget issues in the Senate went forward with-
9/11. In fiscal year 2000, OMB counted 6,183 out a resolution but few of the appropriations
legislative earmarks. That number represented bills were passed by the end of the fiscal year,
a considerable increase from prior years. necessitating an omnibus appropriation of the
Peter-3383-Chapter-07.qxd 6/8/2006 7:34 PM Page 147
BUDGETING 147
remaining appropriations bills. The omnibus rules, they would only have to count a few years
legislation was huge, no one was given a chance of the consequences, even though the financial
to read it before passage, violating some basic impact increased after the end of that period.
democratic values and process. Moreover, Then, a couple of years later, the administration
there is little indication that the omnibus legis- argued that the tax cuts should be treated as
lation represented considered evaluation of permanent, so that an extension of the tax cuts
competing projects, programs, or budgetary would not be scored as having any increased
claims. The pork that was stuffed into this costs. The administration sought to dodge the
mammoth appropriation act prompted mock- implications of the tax cuts first one way, and
ing editorials in newspapers. then serpentined back to dodge the implica-
Under the 1974 act, both houses of Congress tions the other way (Budget Rule Change
agree to not only totals in the budget resolution, Would Make The Cost Of Extending The Tax
and revenues necessary to fund those totals, but Cuts Disappear, 2/27/04 Robert Greenstein
also to how those totals will be allocated to dif- and Joel Friedman The Center for Budget and
ferent functions of government. The commit- Policy Priorities, on line at http://www.cbpp.
tees work within these targets, so they will not org/2-27-04bud4.htm).
have to do their work only to find out at the end It was not only the executive branch that
that their decisions and carefully worked out engaged in creative scoring and rules changes
compromises would be undone by further cuts to make deficits look smaller or vanish in order
to come in under limits. By allocating targets to justify the policy of tax cuts without replace-
up front, the process provides some financial ment of the revenues or compensating reduc-
discipline. In addition, Congress collectively tions in spending. House Republicans adopted
prioritizes spending at the macro rather than what has been called dynamic scoring as a way
the micro level. The subcommittees make prior- of measuring the impact of revenue cuts.
itization decisions within their areas, deciding Dynamic scoring tries to take into considera-
between program requests as long as they stay tion the increase in economic activity, and
within their allocations. Prioritization is the hence the increase in federal revenue, that
heart of budgeting; when Congress does not would result from any tax cut. The problem is
come up with a budget resolution or automati- that no one knows how big that effect might
cally accept the president’s proposal, the com- be, and the temptation is to make it as big as
mittees may go over the totals (which may not necessary to wipe out the apparent negative
even be known or agreed to in advance) and effects of a tax cut. CBO, the Congressional
then make across-the-board cuts later, to get Budget Office, has not adopted dynamic scor-
down to the maximum which has by then been ing, and neither did the Senate, meaning that
determined. This process allows for a minimum budget estimates are likely to differ between
of prioritization. This is in fact what happened committees and between houses.
in 2004, for the fy 2005 budget. In fact, Congress
failed to pass a budget resolution in three of the
last five recent years.
INFLUENCE OF BUDGET PROCESS
The deficits were largely a function (though
(AND DETERIORATION)
not exclusively) of the tax reductions ardently
desired by the president. The president wanted
the tax cuts, but did not want to take responsi- Budget processes do not stand alone, they do
bility for either programmatic reductions or the not reduce spending, or increase deficits, they
deficit. The result was a series of rules that would are a set of rules used by people making
make it look as if the tax reductions cost less than decisions. When the consensus that underlies
they actually do. Budget observers call these the rules evaporates, the rules may remain in
ad hoc accounting rules “gimmicky scoring.” place while implementation erodes. When that
The administration first argued that its tax happens, the effect of the rules is either watered
cuts would be temporary, so that, under the down or eliminated. Gramm Rudman Hollings
Peter-3383-Chapter-07.qxd 6/8/2006 7:34 PM Page 148
did not eliminate or reduce deficits in the manner gamed; deficit reduction targets were missed.
predicted by the rules, in large part because BEAs caps and offset requirements were imple-
Gramm Rudman Hollings was not enforced. mented and, even though they did not control
Similarly, the Budget Enforcement Act had major the cost increases due to existing legislation,
impact on holding down expenditure growth they provided a kind of discipline and con-
and reducing the amount of pork in the budget, sciousness of the impact of spending decisions.
as long as there was consensus that the process BEA also forced trade-offs, making earmarks
was needed and should be enforced. When that not only visible but politically expensive, as
consensus evaporated in 1998, the rules were no legislators had to say what they would cut (and
longer enforced and spending surged, tax who else’s earmarks they would cut) in order
reductions were passed, and deficits grew at a to fund their constituency-based projects.
very fast pace. When decision-makers agree that When these constraints ended, spending,
they want a particular outcome, particular fea- including for earmarks, surged. BEA required
tures of budget processes can make it easier to explicit offsets for revenue reductions; when
bring about those outcomes. the BEA fell into disuse, the president pro-
This chapter has argued that there are two posed and Congress passed large and repeated
classes of outcomes of the budget process. The tax breaks. House efforts to reform the budget
first is this indirect impact or influence on the process by asking for offsets for increased
size of the deficit, the growth of expenditures spending, but not for tax reductions, suggests
(as a measure of the size of government) and that they felt such rules facilitated their policy
the degree to which spending is constituent goals of reducing tax levels. There would be no
related (to satisfy requests from either individ- procedural obstacles to doing so.
ual or companies in a legislator’s district) or Are there any rival hypotheses that could
for narrow partisan purposes. The second is a explain these results other than the influence
direct effect on the quality of decision-making of budget process? There is little doubt that the
and on democratic decision-making. growth of the economy in the later 1990s helped
The chapter makes the case that deficits and to balance the budget, but the growth of the
spending respond to budget rules, since deficits economy was not sufficient in and of itself to
grew during GRH and, after a couple of years reach balance. In addition, spending rates were
of growth, began to and continued to decline controlled. The two together helped restore
under the Budget Enforcement Act, and then balance. Similarly, the slow down of the econ-
when BEA was no longer enforced and allowed omy in the early 2000s had some effect on reduc-
to lapse, and the features of the 1974 Congres- ing revenue, but tax reductions and spending
sional Budget and Impoundment Control Act increases had a much larger impact. A second
were only used episodically, spending surged possible rival hypothesis is that the amount of
and deficits rebounded. The effectiveness of the pork spending was increasing anyway, possibly
BEA in controlling earmarks may have built up due to a changeover in leadership and determi-
demands to the point that, when the controls nation to build new coalitions of supporters in
were relaxed, earmarking too took a major many tightly contested races. However, the long
upswing. term trend was a reduction in geographically
It is not only the case that these outcomes cor- based spending, and although there have always
related with the periods when various budget been some earmarks in the budget, BEA seemed
process rules were in place, and implemented to hold down the level, which then jumped after
or not implemented, but also that the specific the BEA constraints were over. It may well be
features of the process that were or were not that the reduction in the number of programs
being implemented spoke directly to the that can be used distributively has increased the
results. In other words, there is a mechanism demand for pork, but, even within this model,
linking the budget process with these three BEA spending controls and explicit and public
outcomes. The spending controls in GRH were trade-offs held it down for a while. In the post
not implemented fully and were frequently BEA world, trade-offs, when they occur, are
Peter-3383-Chapter-07.qxd 6/8/2006 7:34 PM Page 149
BUDGETING 149
invisible; tax reductions occur seemingly without aware that there are minimum requirements for
the necessity of cutting spending programs. a budget process, that collective rules are neces-
The effect of budget process rules on the sary for functioning, that they must be agreed
quality of decision-making is less inferential on by all parties, and be perceived as fair. Both
and there is no need to rule out alternative Democrats and Republicans have become more
hypotheses. The process is the outcome. The ideological in recent years, the differences
deterioration of scoring rules, altering them to between them have become more acute. The
gain short term policy advantages, and then resulting policy quarrels need to be fought on
changing them back again, to get additional a level playing field so the public trusts and
benefits in the short run, the adoption of accepts the outcomes. If the process does not
accounting rules that obscure present and seem fair, legitimate, or neutral with respect to
future costs, the huge and quick omnibus partisan policy outcomes, or if it is not open to
spending bills that bypass the committee scrutiny, public mistrust grows, participation
process and that legislators are unable to read, continues to decline, and democracy itself
the inability to implement the budget resolu- erodes. In this sense, budget process not only
tions that force some prioritization, all suggest a matters, it may be fundamental to governing in
deterioration of the quality of decision-making. the United States. That having been said, how-
Perhaps most seriously, the budget process ever, rules are but one part of the equation of
has become less democratic. The processes have budget control in most industrialized democra-
become ad hoc, to produce a budget in the cies, and a range of factors must be considered
face of policy stalemates. Rules are made and for a more complete understanding.
remade, often in the service of particular imme-
diate policy goals, not known in advance, not
neutral, and not agreed to by all parties. They do
not allow for a forum in which disagreements REFERENCES
can be discussed and worked out, exaggerating
the tendency toward stalemate, contributing to Albert-Roulhac, C. (1998) The Influence of EU
the lack of budget resolutions. The administra- membership on Methods and Processes of
tion is not content merely to lower taxes, but Budgeting in Britain and France, Governance 11,
also seeks to lock in those reductions for years to 209–30.
come, and has increased borrowing to an extent Gray, A., W. I. Jenkins and B. Segsworth (2002)
Budgeting, Auditing and Evaluation (New
that will burden the next generation, and reduce
Brunswick, Transaction).
its ability to prioritize. HM Treasury (2005) Public Expenditure Planning
With the end of divided government as a and Control in the United Kingdom (London: HM
result of the elections in 2004, policy stalemates Treasury).
should be reduced. It should be easier to pass Meyers, R. (1992) Federal Budgeting and Finance;
budget resolutions and hence to prioritize the The Future is Now, Public Budgeting and Finance,
budget because the administration will control 4, 2–14.
a majority of both Houses of Congress. It should Palazzolo, D. (1999) Done Deal?: The Politics of the
be easier to pass appropriations bills, and hence 1997 Budget Agreement (New York: Chatham
the need for huge, last minute omnibus appro- House).
priations should be reduced. The quality of Patashnik, E. M. (2000) Putting Trust in the US Budget
(Cambridge: Cambridge University Press).
decision-making should therefore improve and
Stolfi, F. (2006) Reforming the Italian State:
the process should become more predictable Administrative Modernization and Fiscal
and open. But when the same political party Management (Unpublished Ph.D. Dissertation,
controls the White House and both Houses of Department of Political Science, University of
Congress, there is an increased chance of pass- Pittsburgh).
ing budget rules with marked policy biases. As Wanna, J., L. Jensen and J. de Vries (2003) Controlling
yet, there is no evidence that political actors are Public Expenditure (Cheltenham: Edward Elgar).
Peter-3383-Chapter-07.qxd 6/8/2006 7:34 PM Page 150
Peter-3383-Chapter-08.qxd 6/8/2006 7:51 PM Page 151
8
Implementation
SØREN C. WINTER
Although the field of implementation research is stage for later implementation research. Most
barely 30 years old, implementation has already implementation research has focused on
been analyzed from many different perspectives implementation problems, barriers, and fail-
representing different research strategies, evalua- ures, and this pessimistic view of implementa-
tion standards, concepts, focal subject areas, tion was already reflected in the subtitle of this
and methodologies. This chapter first performs seminal work, ‘How great expectations in
a critical review of some of the major contribu- Washington are dashed in Oakland; or, Why it’s
tions to the literature. This examination will amazing that federal programs work at all…’.
follow the development of the field over three This case study focused on the local imple-
generations of research (Goggin, 1986). mentation of a federal economic development
Second, based on a critical examination of the program to increase employment among
development and status of the research field, the ethnic minority groups in Oakland in
chapter will suggest promising ways of moving California. Its guiding research questions were:
ahead. It claims that implementation research ‘How well was this authoritative mandate (law,
can be improved by accepting theoretical diver- regulation, program, official pronouncement)
sity and partial theories and hypotheses, rather implemented?’ and ‘How might it have been
than looking for one common and general better implemented?’ Later research redefined
theoretical framework, seeking conceptual clari- the question to focus on achieving the explicit
fication, including focusing on output (perfor- or implicit values in a given mandate rather
mance of implementers) as well as outcomes as than its prescriptive details (Bardach, 2001).
dependent variables in implementation research, Accordingly, goal achievement has been the
and applying more comparative and statistical dominating standard and dependent variable
research designs, rather than relying on single for implementation research since the 1970s.
case-studies in order to sort out the influence of Pressman and Wildavsky focused on the
different implementation variables. ‘complexity of joint action’ as the key imple-
mentation problem. In their Oakland economic
development case – as in many others – federal,
THE PIONEERS
regional, state, and local government actors,
courts, affected interest groups, private firms,
In several respects the book, Implementation, and media had a role and stake in policy
by Pressman and Wildavsky (1973) sets the implementation. Implementation problems
Peter-3383-Chapter-08.qxd 6/8/2006 7:51 PM Page 152
were amplified not only by the many actors but SECOND GENERATION MODEL
also by the many decision and veto points, BUILDERS: TOP-DOWN, BOTTOM-UP
which must typically be passed during the AND SYNTHESES
implementation process. Although they proba-
bly over-emphasized the lack of conflict in Second generation implementation studies began
their case, Pressman and Wildavsky convinc- in the early 1980s. While the first generation
ingly showed that merely slightly different per- studies had been explorative and theory generat-
spectives, priorities, and time horizons among ing, the ambition of the second generation was to
multiple actors with different missions in take a next step in theory development by con-
repeated and sequential decisions could cause structing theoretical models, or rather frame-
delay, distortion, and even failures in policy works of analysis, which could guide empirical
implementation. analysis. Some of these studies had more opti-
However, the two authors also demon- mistic views on successful implementation.
strated that failures are not only caused by bad The construction of models and research strate-
implementation but also by bad policy instru- gies, however, immediately led to a major con-
ments. Many of the problems in the Oakland frontation between the so-called top-down and
case would have been avoided had policy makers bottom-up perspectives on policy implementation.
chosen a more direct economic instrument The predominant top-down researchers
that would ex post have tied spending of public focused on a specific political decision, nor-
expenditures to the actual number of minority mally a law. On the background of its official
workers employed, rather than relying on end- purpose they followed the implementation
less ex ante negotiations with affected parties down through the system, often with special
and authorities. interest in upper level decision-makers. They
Pressman and Wildavsky are good represen- would typically assume a control perspective on
tatives for the first generation of implemen- implementation, trying to give good advice on
tation studies, which were typically explorative how to structure the implementation process
and inductive case studies with a theory- from above in order to achieve the purpose of
generating aim. Very few central theoretical the legislation and to minimize the number of
variables were in focus, in this case the number decision points that could be vetoed.
of actors and decision points and the validity The best-known and most frequently used
of the causal theory. Another outstanding (Sabatier 1986) top-down analysis framework
example is Eugene Bardach’s (1977) The was developed by Mazmanian and Sabatier
Implementation Game, which placed more (1981). It contains 17 variables placed in 3 main
emphasis on the aspects of conflict in imple- groups concerning the tractability of the prob-
mentation, seeing implementation as a contin- lems addressed by the legislation, the social
uation of the political game from the policy and political context, and the ability of the leg-
adoption stage, although partly with other islation to structure the implementation
actors and other relations among actors. process. This structuring can be made by
Bardach analyzed the types of games that means of, for example, hierarchy, appointing of
various actors apply in the implementation authorities and staff with a positive attitude
process in order to pursue their own interests. towards the legislation/program, and use of
However, these games tend to distort imple- incentives including competition among
mentation from the legislative goals. Among providers. By adding a long-term perspective
other representatives from what has later been of 10–15 years to implementation, the authors
called the first generation of implementation show that, over time, start-up problems are
research we find Erwin Hargrove (1975), who often ameliorated by better structuring of the
called implementation research ‘the missing implementation by policy advocates (see also
link’ in the study of the policy process, and Kirst and Jung, 1982). This gave rise to much
Walter Williams and Richard Elmore (1976). more optimistic views of implementation in
Peter-3383-Chapter-08.qxd 6/8/2006 7:51 PM Page 153
IMPLEMENTATION 153
contrast to the pessimism introduced by public policies, their similar working conditions
Pressman and Wildavsky (1973) and joined by make them all apply similar behavior. This
most implementation analysts. means that street-level bureaucrats even
Mazmanian and Sabatier’s framework was met across policy-types, tend to apply similar types
by two different kinds of criticism. According of practices, whether they are teachers, police-
to one strand, the model was naive and unreal- men, nurses, doctors, or social workers. It also
istic because it overemphasized the ability of means that their individual attitudes are not
policy proponents to structure implementa- expected to have important implications for
tion, thus ignoring the ability of policy oppo- their behaviors.
nents to interfere in this structuring process Although trying to do their best, street-level
(Moe, 1989). Often policy opponents are able bureaucrats experience a gap between the
to make policy goals less clear and to increase demands made on them by legislative man-
their own long-term influence in the imple- dates, managers and citizens on one side and
mentation process in order to avoid some of their high workload on the other. In this situa-
the effects intended by policy proponents. tion they all apply a number of coping mecha-
Conceptually, the model ignored the politics of nisms that systematically distort their work in
policy-formulation and policy design (Winter, relation to the intentions of the legislation. They
1986b; May, 2003). ration services, make priorities between their
Another strand of criticism came from the tasks, apply simple standardized processing of
bottom-up researchers who took special interest clients, which tends to be in the favor of more
in ‘the bottom’ of the implementation system, easy cases and resourceful clients (creaming),
the place where public policies are delivered to they seek to control clients, and over time street-
citizens or firms. They all emphasized the influ- level bureaucrats develop more cynical percep-
ence that front-line staff have on the delivery of tions of clients and modify the policy objectives
policies such as social services, income transfers that are the basis of their work. According to
and law enforcement in relation to citizens and Lipsky, increasing staff resources is no cure
firms. Front-line workers are crucial decision- to coping as more resources will merely lead to
makers in these studies that emphasize the more demand for services.
weak control that politicians and administrative Other bottom-up researchers go the whole
managers have of control front-line staffs. length, rejecting the objective of policy man-
Like top-down researchers (and also most eval- dates as an evaluation standard. Instead their
uation researchers), some bottom-up researchers analysis departs from a specific problem
use the official objectives of a given legislation as (Elmore, 1982) such as youth unemployment
the standard of evaluation (Lipsky, 1980; Winter, (Elmore, 1982) or small firms’ conditions of
1986a). Michael Lipsky (1980) developed a growth (Hull and Hjern, 1987). In practice it is
theory on ‘Street-level Bureaucracy.’ It focuses the researcher himself who in most cases
on the discretionary decisions that each front- defines the problem and thereby his evaluation
line worker – or ‘street-level bureaucrat’ as standard. This is acceptable if done explicitly,
Lipsky prefers to call them – make when deliv- and it can be fruitful if the researcher is able to
ering policies to individual citizens. This discre- convince his readers about the appropriateness
tionary role in delivering services or enforcing of his problem definition.
regulations makes street-level bureaucrats The next task in Hull and Hjern’s (1987) bottom-
essential actors in implementing public poli- up approach is to identify the many actors that
cies. Indeed, Lipsky (1980) turns the policy are affecting the problem in question and to
process upside-down by claiming that street- map relations between them. In these network
level bureaucrats are the real policy makers. analyses – using a ‘snowball method’ – both
However, one ironic aspect of the theory is that, public and private actors become essential,
although Lipsky emphasizes the individual role and the analyses often include several policies
of street-level bureaucrats in implementing that affect the same problem. In this way,
Peter-3383-Chapter-08.qxd 6/8/2006 7:51 PM Page 154
the analysis maps the informal, empirical one approach might be more relevant than the
implementation structure around a given other. Sabatier (1986) claims that the top-down
problem, while top-down research tends to look perspective is best suited for studying the
at the formal implementation structure related implementation in policy areas that are domi-
to one particular policy program. According to nated by one specific legislation, limited
Hull and Hjern, empirical implementation research funds, or where the situation is struc-
structures tend to be far less hierarchical than tured at least moderately well. Bottom-up per-
formal ones, and they often cross organizational spectives, on the other hand, would be more
borders in forming collaborative networks at relevant in situations where several different
the operational level. The bottom-up analyses policies are directed towards a particular
by Hjern and associates are important in draw- problem, and where one is primarily interested
ing attention to implementation activities and in the dynamics of different local situations.
structures at the local operational level, but the Attempts were also made to synthesize the
perspective has more the character of guidelines two models. Richard E. Matland (1995) suggests
for an inductive research strategy and method- that their relative value depends on the degree of
ology than a development of theory and ambiguity in goals and means of a policy and the
hypotheses that can be empirically tested. A degree of conflict. Traditional top-down models,
recent example of the bottom-up approach is based on the public administration tradition,
Bogason’s (2000) study of local governance. It is present an accurate description of the imple-
inspired by Hull and Hjern but adds elements mentation process when a policy is clear and the
of institutional and constructivist analyses and conflict is low. However, newer top-down
points to the fragmented character of the mod- models, such as the Mazmanian-Sabatier frame-
ern state in policy making and implementation. work, are also relevant when conflict is high and
ambiguity is low, which makes the structuring of
the implementation particularly important. In
Suggested Syntheses contrast, bottom-up models provide an accurate
description of the implementation process when
The top-down and bottom-up perspectives were the policy is ambiguous and the conflict is low.
useful in drawing attention to the fact that When conflict as well as ambiguity is present,
both top and bottom play important roles in both models have some relevance according to
the implementation process, but in the long Matland.
run the battle between the two approaches was Other attempts at synthesizing the two
not fruitful. Each tended to ignore the portion approaches were made by the former main
of the implementation reality explained by the combatants. The bottom-uppers, Hull and Hjern
other (Goggin et al., 1990: 12; Hill and Hupe, (1987), proposed a synthesis – called ‘an induc-
2002). Elmore (1985) actually recommends tive approach to match outcomes of politics and
using both forward mapping – which is essen- their intentions,’ which calls for systematic
tially a top-down analysis – and backward interview analysis of relevant actors from the
mapping for policy analysis because each tends bottom to the very top. The approach would
to offer valuable insights for policy makers. He require immense research resources, and I am
claims that policy designers need to consider not aware of any such study performed in prac-
the policy instruments and the resources they tice. In addition, their proposed synthesis suffers
have at their disposal (forward mapping), as from being methodological recommendations
well as the incentive structure of the target rather than theoretically based expectations,
group and street-level bureaucrats’ ability to which can be tested systematically.
tip the balance of these incentives in order to Also Sabatier (1986) has suggested a
affect the problematic situation of the target synthesis – the so-called Advocacy Coalition
group (backward mapping). Framework (ACF). However, although making
Other scholars have tried to solve the con- an important contribution to the public policy
troversy by specifying the conditions where literature, Sabatier and his later associate,
Peter-3383-Chapter-08.qxd 6/8/2006 7:51 PM Page 155
IMPLEMENTATION 155
relatively automatic and easy to collect, such as can produce and deliver implementation
an environmental tax per liter gasoline sold, outputs independently of each other, can pro-
while others require a substantial staff for vide relatively good implementation results,
inspection and enforcing, for example, taxing although the coordination may not be optimal.
diffuse pollution. O’Toole (2003) and May (2003) show how inter-
It is important to understand that ineffec- organizational coordination problems can be
tive policy designs are not always due to lack of reduced by using policy design to increase com-
knowledge on the part of the policy designers. mitment, build and use a common interest, and
Policy design of instruments and organiza- facilitate cooperation via exchange.
tional structure is first of all a political process, The behaviors of street-level bureaucrats are
in which political actors – both policy propo- also crucial for the implementation of most
nents and opponents – try to maximize their policies, and Lipsky’s (1980) insights above on
interests, including selecting an organizational ‘street-level bureaucracy’ are included in the
structure, which will allow themselves to max- Integrated Implementation Model. Street-level
imize long-term control of the implementa- bureaucrats are making important discre-
tion process (Moe, 1989). tionary decisions in their direct contact with
The next set of factors of the model focuses citizens and firms. Because such bureaucrats
on how the implementation process affects the work in situations characterized by many
results. Implementation processes are charac- demands and limited resources, they respond by
terized by organizational and interorganizational resorting to coping behaviors. These short-cuts
behaviors representing different degrees of com- systematically bias the delivery behavior in
mitment and coordination. Interorganizational relation to the policy mandates. While Lipsky’s
implementation settings seem to become ever contribution was important for understanding
more important (O’Toole, 2003). As mentioned implementation, the theory needs more specifi-
above, Pressman and Wildavsky (1973) focused cations of the causal mechanisms that can
on the ‘complexity of joint action,’ according to explain variation in coping-behaviors and their
which successful implementation is likely to be consequences (Winter, 2002; Heinesen et al.,
negatively related to the number of actors and 2004), cf. below. For a review of the literature on
decision and veto points. street-level bureaucrats, see Meyers and
However, as shown by O’Toole and Montjoy Vorsanger, 2003).
(1984; O’Toole, 2003) this insight only applies to According to the Integrated Implementation
certain kinds of interorganizational implementa- Model also target groups of public policies,
tion settings. Decision points are not indepen- that is, citizens or firms, play important roles,
dent of each other, but successful implementation not only on the effects of the policy, but also
results can be stimulated by an early agreement in affecting the performance by street-level
on basic understandings, which can promote bureaucrats through positive or negative
‘bandwagon effects’ in later decisions, and deci- actions in co-producing public services and
sions can be merged by crafting ‘package deals.’ regulation (Hill and Hupe, 2002: 134–36).
The implementation prospects also depend Finally, socio-economic contexts form impor-
on the type of resource-dependency among tant framework conditions for implementa-
participating organizations. The ‘complexity of tion. For example, in employment policies
joint action’ best applies to a chain of sequen- delivery behavior (the types of employment
tial relations, in which one organization offers) and effects depend heavily on ups and
depends on outputs from another as input for downs in the business cycle.
its own contribution to implementation. The Integrated Implementation Model is
However, reciprocal relations, in which two obviously not a model in the strict sense of a
organizations depend on each other for inputs simple causal model. It is rather a framework
can decrease the likelihood of veto points of analysis presenting key factors and mecha-
because both have incentives to cooperate. nisms that affect implementation outputs
Pooled relations, where multiple organizations and outcomes. For each set of factors a number
Peter-3383-Chapter-08.qxd 6/8/2006 7:51 PM Page 157
IMPLEMENTATION 157
of more specific hypotheses can be developed Later, Lester and Goggin (1998), in making a
(Winter, 1990, 1994, 2003a; Jensen et al., 1991; status for implementation research, called for
May, 2003; O’Toole, 2003; Meyers and the development of ‘a parsimonious, yet com-
Vorsanger, 2003). plete, theory of policy implementation.’ They
suggested that such meta-theory might be devel-
oped by combining the insights of communi-
cations theory, regime theory, rational choice
THIRD GENERATION: QUANTITATIVE
theory (especially game theory), and contin-
RESEARCH DESIGNS
gency theories. As dependent variable for imple-
mentation studies they proposed to focus on
While the first and second generations of implementation processes rather than outputs
implementation studies have been helpful in and outcomes.
directing attention to implementation prob-
lems and identifying implementation barriers
and factors that might ease implementation,
THE NEED FOR A NEW
the research had not succeeded in sorting out
RESEARCH AGENDA
the relative importance of the explanatory
variables. A substantial part of the studies
could be criticized as merely presenting – often While agreeing with Goggin’s (1986) call for
long – checklists of variables that might effect more comparative and statistical research
implementation. designs based on quantitative methods, I dis-
Malcolm Goggin (1986) pointed out that, agree with several of the later methodological
because implementation research had been and theoretical recommendations made by
dominated by single case studies, it was plagued him and his colleagues. As recognized by
by the problem of ‘too few cases and too many one of these authors, O’Toole (2000), to follow
variables’ or by ‘overdetermination,’ where two the methodological suggestions given by
or more variables explain variation in the Goggin, Bowman, Lester, and O´Toole (1990)
dependent variable equally well. The fact that would involve at least outlining a research
the single case-study approach does not allow career’s worth of work. This work would
for any control of third variables had hampered require applying research designs that involve
the development of implementation theory, numerous variables, across different policy
according to Goggin. He therefore called for types, across 50 states, over at least 10 years, as
a third generation of implementation studies well as measuring the relevant variables by a
that would test theories on the basis of more combination of content analyses, expert pan-
comparative case-studies and statistical research els, elite surveys, and expert reassessment of
designs, which could increase the number of the data from questionnaires and interviews.
observations. As such a research strategy is too demanding;
Goggin followed up on these recommenda- less demanding research strategies, which can
tions with his associates (Goggin, Bowman, still secure a sufficient number of observa-
Lester, and O´Toole, 1990) in a study, which was tions, would be more realistic.
based mainly on a communications theory per- My suggestions for further development
spective on intergovernmental implementation, of implementation research can be summarized
but also included many variables from previous in six points: (1) providing theoretical diversity,
top-down and bottom-up research. The study (2) focusing on partial rather than general
focused especially on variation among states in implementation theories, (3) seeking conceptual
the way and extent they implement federal poli- clarification, (4) focusing on the implementa-
cies in three different social and regulatory poli- tion output (performance of implementers) as
cies. The authors tried to encourage further a dependent variable, (5) including studies of
research involving multiple measures and mul- outcomes, and (6) using more comparative and
tiple methods, including quantitative methods. statistical research designs (Winter, 1999).
Peter-3383-Chapter-08.qxd 6/8/2006 7:51 PM Page 158
While the last point has been developed above, I Lynn, Heinrich and Hill (2001; Heinrich and
will elaborate on the other ones in the following Lynn, 2000). They have proposed a ‘Logic of
and illustrate them by some of my recent Governance.’ They define governance as ‘regimes
research on implementation of Danish agro- of laws, rules, judicical decisions, and adminis-
environmental regulation with Peter May and of trative practices that constrain, prescribe, and
Danish integration policy towards refugees and enable the provision of publicly supported
immigrants. goods and services.’ The logic can be expressed
as a basic ‘reduced form’ according to which
outputs/outcomes are a function of environ-
Theoretical Diversity and mental factors, client characteristics, treatments,
Partial Theories structures and managerial roles and actions.
A more complex form allows for interrelations
Given the many exploratory variables, which between the explanatory factors. This frame-
have already been identified by various imple- work broadens the scope of traditional
mentation scholars, the suggestion by Goggin implementation research. It focuses on the inter-
et al. of developing a ‘parsimonious, yet complete relations and complexities of administrative
implementation theory’ by combining theoreti- phenomena, and calls for sound empirical
cal elements from at least four different theories, yet theory-driven research. While its main theo-
appear to be a contradictio in adjecto and is more retical inspiration is political economy, other
likely to lead to theoretical mismatch. Rather theoretical perspectives are welcomed.
than looking for the overall and one for all The main contribution of the framework is
implementation theory, as has been the utopian its emphasis on systematic empirical testing of
objective for many implementation scholars, we theory driven hypotheses and for generating a
should welcome diversity in both the theoretical framework that allows many studies to talk to
perspectives and methodologies applied. Such each other. The authors claim that both top-
diversity will give us new insights. It also down and bottom-up considerations are
strikes me as unrealistic – and probably not very included, but the top-down control orienta-
fruitful – that many scholars could agree on tion seems to be stronger than bottom-up per-
applying one common theoretical framework. spectives. The motivations of street-level
Although the general implementation frame- bureaucrats have not been conceptualized as
works presented by model builders so far have part of the framework, which focuses more on
been helpful in giving an overview of some cru- managerial behaviors.
cial implementation variables, the generality of A related framework and research program
such models may in fact be an obstacle for on management and performance has been
further development of our understanding established by Meier and O’Toole (2004; 2006).
of implementation. This is due to the fact that
generality inhibits precise specification of vari-
ables and causal mechanisms (May, 1999). Need for Conceptual Clarification and
Consequently, it seems more fruitful to use Focusing on Implementation Outputs as
research resources on developing partial a Dependent Variable
theories and hypotheses about different and
more limited implementation problems and on As pointed out by Peter May (1999) most con-
putting these to serious empirical tests. ceptual frameworks in the implementation
Some of the different implementation per- literature are weakly developed, lacking ade-
spectives may be integrated into broader analyt- quate definitions of concepts and specification
ical frameworks or models (Mazmanian and of causal mechanisms. The most important
Sabatier, 1981; Winter, 1990; Goggin, Bowman, issue for the development of implementation
Lester and O’Toole, 1990). A new promising research may be to reconsider what constitutes
research collaboration around a common the object of the study. There has been some
analytical framework has been initiated by disagreement in the literature on the term of
Peter-3383-Chapter-08.qxd 6/8/2006 7:51 PM Page 159
IMPLEMENTATION 159
from one policy area to another. In order to 2004) However, in addition, the dimension of
stimulate theory building we should avoid professional distance (versus close personal
concepts that are very policy specific, because involvement with clients) is included, which
generalizations based on these would have a seems to be particularly relevant in social
rather narrow sphere of application. policy implementation.
At the other extreme, Lipsky’s (1980) street- One advantage of creating such conceptual-
level bureaucracy theory represents an ambi- ization of the behavior of implementers is that
tious attempt to offer a universally applicable it is well suited for testing hypotheses for
set of concepts for describing the coping explaining variation in implementation behav-
behavior of street-level bureaucrats in all ior across time and space. Variables from
policy areas. Although Lipsky’s coping behav- implementation theory characterizing aspects
iors might seem to have a better fit for social of the implementation process would be an
policies with weak clients, it has been demon- important basis for the development and test
strated that they are also relevant for a regula- of such hypotheses. However, another advan-
tory policy with strong clients in a study of tage of focusing on outputs as a dependent
Danish agro-environmental and integration variable in implementation research is that we
policies (Winter, 2002). However, coping can integrate the study of implementation
focuses on dysfunctional behaviors, and we much more with theory on bureaucratic poli-
need more concepts for characterizing agency tics and organization theory. Implementation
and front-line staff behaviors adequately. research can thereby gain inspiration from
Other concepts have been developed for these research fields that have a long tradition
classifying the behavior of implementers in of studying the behavior of agencies and
almost any kind of social regulation policy bureaucrats (see also Lynn, Heinrich and Hill,
(Kagan, 1994). May and Winter (1999; 2000; 2001). In return, these sub-disciplines can
Winter and May, 2001) have developed con- benefit from implementation concepts that are
cepts for regulatory enforcement at both much more policy relevant than those behav-
agency and individual street-level bureaucrat ioral variables applied in most bureaucracy
levels. Agency enforcement choices are concep- and organization theory.
tualized as (1) tools (use of different enforce- As an example, Winter (2003b) analyzes the
ment measures: sanctions, information, and discretion of street-level bureaucrats in imple-
incentives), (2) priorities (whom to target and menting agro-environmental regulation and
what to inspect for), and (3) effort (use integration policy in Denmark by applying a
and leveraging of enforcement resources). The modified principal-agent perspective and its
enforcement style of individual inspectors is notion of information asymmetry in examin-
defined as the character of the day-to- ing the extent to which local politicians control
day interactions of inspectors with the target their street-level bureaucrats (Moe, 1984;
group. May and Winter expect and verify, in Brehm and Gates, 1999). Regression analyses
a study of agro-environmental regulation in of 216 local inspectors and 388 social case-
Denmark, that enforcement style has two workers show that local politicians’ policy pref-
dimensions comprised as the degree of formal- erences have very little direct impact on the
ity of interactions and the use of threats and behaviors of street-level bureaucrats. However,
other forms of coercion (May and Winter, to some extent the politicians do control rela-
2000; see also May and Burby, 1998; May and tively visible kinds of performance, such as the
Wood, 2003). number of inspections and the number of
In a study of the implementation of Danish timely processed cases, through funding capa-
integration policy for refugees and immigrants city for implementation. On the other hand,
the same two implementation style dimen- when it comes to less transparent front-line
sions have been found relevant for examining behaviors – such as the implementation styles
a social policy (Winter, 2003b; Heinesen et al., and the strictness front-line staffs apply in
Peter-3383-Chapter-08.qxd 6/8/2006 7:51 PM Page 161
IMPLEMENTATION 161
bureaucrats can influence and build on these has found that street-level bureaucrats’ use of
incentives. For example, in examining Danish coping as well as various implementation
farmers’ compliance with environmental regu- styles affect the outcome in terms of labor
lations, Winter and May (2001) map the regula- market integration of refugees and immi-
tees’ action model. In multiple regression grants. However, the effects depend on the
analyses of survey data of 1,562 farmers, they context (Lynn, Heinrich and Hill, 2001; Hill
show that compliance is affected by farmers’ (1) and Hupe, 2002). In municipalities with a dif-
calculated motivations based on the costs of ficult integration task – due to the composition
complying and the perceived risk of detection of of immigrants and labor market conditions –
violations (while the risk of sanctions, as in coping and professional distance increases
most other studies, had no deterrent effect), (2) employment, while in municipalities with an
normative sense of duty to comply and (3) easy integration task formalism decreases
social motivations based on adaptation to employment (Heinesen et al., 2004). As coping
expectations from significant others. Inspectors is normally assumed to bias the implementa-
signal such expectations through their style of tion in a dysfunctional way, it is remarkable
interacting with target groups. Inspectors’ for- that it here has positive outcomes, at least in
malism increases compliance up to a point by the short run, which may be due to creaming
providing greater certainty of what is expected of the most resourceful immigrants. The find-
of regulatees, while coercive styles with threats ings from the two studies illustrate that deliv-
of sanctions backfire for regulatees who are not ery performance variables can be constructed
aware of the rules. Willingness to comply is not that are fruitful both as dependent variables in
enough if the ability to comply is not there. explaining implementation outputs and as
Thus, awareness of rules and financial capacity independent variables in explaining outcomes.
increase farmers’ compliance. As the most relevant criterion for evaluating
An understanding of the motivations and the relevance of a policy is its outcomes, some
incentives of target groups is essential for speci- implementation scholars are tempted to skip
fying causal links between implementation focusing on outputs and go directly to examining
behavior and target group responses. Further the relation between implementation processes/
research along this line has shown that inspec- structures and outcomes (Lynn, Heinrich and
tors not only affect farmers’ compliance directly Hill, 2001; Heinrich and Lynn, 2000; Meier and
through social motivation. They can also do so O’Toole, 2004). However, if implementation is
indirectly by using deterrence, because frequent important to outcomes, it is likely to work
inspections increase farmers’ perceived risk of through outputs, and we will not get a full under-
being caught if violating the rules – or they can standing of the causal links between implemen-
use information provision for increasing regula- tation and outcomes, unless we understand how
tees’ awareness of rules. Affecting their norma- implementation structures and processes shape
tive commitment to comply is much trickier. outputs, and how outputs shape outcomes. Often
Inspectors try often to do so, but are unlikely a given implementation structure – for example
to succeed because farmers do not trust them specialization of front-line staffs – affects several
enough. In contrast, advice from credible front-line practices with opposite effects on out-
sources – such as farmers’ own professional comes. Outcomes are also often affected by other
trade-organizations and consultants – is much factors than implementation processes, struc-
more effective in fostering a sense of duty to tures and outputs (Heinesen et al., 2004).
comply. This demonstrates an important role for Therefore, we need implementation research that
third parties – including interest groups and seeks to explain variation in outputs as well as in
consultants – as intermediaries in affecting outcomes by focusing on the links between out-
policy outcomes through information provision puts and outcomes.
and legitimization (Winter and May, 2002). I have argued for an implementation
By the same token, a study on the research agenda that involves theoretical diver-
implementation of Danish integration policy sity, partial theory and hypothesis building and
Peter-3383-Chapter-08.qxd 6/8/2006 7:51 PM Page 163
IMPLEMENTATION 163
of the century, articles in sector journals made Brehm, John and Scott Gates (1999) Working,
up no less than 84 percent of all articles with Shirking, and Sabotage: Bureaucratic Response to a
the word ‘implementation’ or ‘implementing’ Democratic Public. Ann Arbor: University of
in their titles, whereas core journals of political Michigan Press.
Dye, T.R. (1976) What Governments Do, Why They
science, public administration, and public
Do It, and What Difference It Makes. Tuscaloosa,
policy only accounted for 16 percent. The
Al.: University of Alabama Press.
highly fragmented character of implementa- Easton, David (1953) The Political System. New
tion research is not very conductive to theory York: Alfred A. Knopf.
accumulation (Sætren, 2005). Elmore, Richard F. (1982) ‘Backward Mapping:
In making a status and reconsideration of Implementation Research and Policy Decisions,’
implementation research, I suggest that the pp. 18–35 in W. Williams, R. F. Elmore, J. S. Hall
most promising research program would be et al. (eds.), Studying Implementation. Chatham
dropping the ambition of constructing a general N.J.: Chatham House Publ.
theory of implementation in favor of develop- Elmore, Richard F. (1985) ‘Forward and Backward
ing and testing different partial theories and Mapping: Reversible Logic in the Analysis of
Public Policy,’ in K. Hanf and T. A. J. Toonen (eds.),
hypotheses; providing more conceptual clarifi-
Policy Implementation in Federal and Unitary
cation in specification of causal relations,
Systems. Dordrecht: Martinus Nijhoff, pp. 33–70.
including focusing on explaining variation in Elmore, Richard F. (1987) “Instruments and
implementation outputs (delivery perfor- Strategy in Public Policy”, Policy Studies Review
mance) as well as outcomes; and giving higher 7(1): 174–86.
priority to comparative and statistical research Goggin, Malcolm L. (1986) ‘The ‘Too Few Cases/Too
designs, which can, better than single case-stud- Many Variables’ Problems in Implementation
ies, control for third variables and form a better Research,’ The Western Political Quarterly, 39:
basis for generalizing research findings. 328–47.
Returning to the classical questions of public Goggin, Malcolm L., Ann O’M., Bowman, James P.,
policy research (Dye, 1976), implementation Lester, and Laurence J., O’Toole, Jr., (1990)
Implementation Theory and Practice: Toward a
outputs is policy at its most operational level.
Third Generation. New York: HarperCollins.
Policy design and the implementation processes
Gunningham, N., R.A. Kagan and D. Thornton
are important causes of policy-outputs, as (2003) Shades of Green: Business, Regulation and
outcomes are the consequences. Consequently, Environment. Stanford: Stanford University Press.
we should seek to explain outputs as well as Hargrove, Erwin (1975) The Missing Link: The Study
outcomes. of the Implementation of Social Policy Washington,
D.C.: The Urban Institute.
Heinesen, E., S. C. Winter, I. R. Bøge and L. Husted
REFERENCES (2004) Kommunernes integrationsindsats og inte-
grationssucces. Copenhagen: AKF Forlag.
Heinrich, Carolyn and Laurence E. Lynn, Jr. (eds.)
Bardach, Eugene (1977) The Implementation Game. (2000) Governance and Performance: New
Cambridge, Mass.: MIT Press. Perspectives. Washington D.C.: Georgetown
Bardach, Eugene (2001) ‘Implementation, Political,’ in University Press.
N. J. Smelser and P. B. Baltes (eds.), International Hill, Michael and Peter Hupe (2002) Implementing
Encyclopedia of the Social & Behavioral Sciences. Public Policy: Governance in Theory and Practice.
Amsterdam: Elsevier Science Ltd. London: Sage Publications.
Bloom, Howard, S., Carolyn J. Hill and James A. Hull, Christopher J. with Benny Hjern (1987)
Riccio (2003) ‘Linking Program Implementation Helping Small Firm Grow: An Implementation
and Effectiveness: Lessons from a Pooled Sample Perspective. London: Croom Helm.
of Welfare-to-work Experiments.’ Journal of Jensen, T. P., S., Winter, J. Manniche, and P. D. Ørberg,
Policy Analysis and Management 22(4): 551–576. with Thaulow I. (1991) Indsatsen for langtid-
Bogason, Peter (2000) Public Policy and Local sledige – en undersøgelse af administration og effekt
Governance: Institutions in Postmodern Society. af arbejds- og uddannelsestilbud. Copenhagen:
Cheltenham: Edward Elgar. AKF Forlag.
Peter-3383-Chapter-08.qxd 6/8/2006 7:51 PM Page 165
IMPLEMENTATION 165
Kagan, Robert A. (1994) ‘Regulatory Enforcement,’ May, Peter J. and Søren Winter, (2000) ‘Reconsidering
in David H. Roosenbloom and Richard D. Styles of Regulatory Enforcement: Patterns in
Schwartz (eds.), Handbook of Regulation and Danish Agro-Environmental Inspection,’ Law &
Administrative Law. New York: Marcel Decker, Policy, 22 (2): 143–73.
pp. 383–422. May, Peter J. and Robert S. Wood (2003) ‘At the
Keiser, L.R. and Soss, J. (1998) ‘With Good Cause: Regulatory Frontlines: Inspectors’ Enforcement
Bureaucratic Discretion and the Politics of Child Styles and Regulatory Compliance,’ Journal of
Support Enforcement.’ American Journal of Public Administration Research and Theory,
Political Science, 42 (4): 1133–56. 13: 117–39.
Kirst, M. and R. Jung (1982) ‘The Utility of a Mazmanian, Daniel A. and Paul Sabatier, (eds.)
Longitudinal Approach in Assessing Imple- (1981) Effective Policy Implementation. Lexington:
mentation: A Thirteen Year View of Title 1, ESEA,’ Lexington Books.
in W. Williams, R. F. Elmore, J. S. Hall R. Jung, M. Meier, Kenneth J. and Laurence J. O’Toole, Jr. (2004)
Kirst, S.A. MacManus, B.J. Narver, R.P. Nathan ‘Conceptual issues in Modeling and Measuring
and R.K. Yin (eds.), Studying Implementation. Management and Its Impact on Performance,’ in
Chatham N.J.: Chatham House Publ, pp. 119–48. P. Ingraham and L.E. Lynn, Jr., The Art of
Lasswell, H.D. (1936) Politics: Who Gets What, Governance: Analyzing Management and Admini-
When, How. New York: McGraw Hill. stration. Washington DC: Georgetown University
Lester, James P. and Malcolm, L. Goggin (1998) Press.
‘Back to the Future: the Rediscovery of Meier, Kenneth J. and Laurence J. O’Toole, Jr. (2006)
Implementation Studies,’ Policy Currents – Modeling Public Management: Empirical Analysis
Newsletter of the Public Policy Section of the of the Management-Performance Nexus. Paper
American Political Science Association, 8(3): 1–9. for delivery at the X International Research
Linder, Stephen H. and B. Guy, Peters (1989) Symposium on Public Management (IRSPM)
‘Instruments of Government: Perceptions and in Glasgow, 10–12 April 2005. Texas A&M
Contexts,’ International Public Policy, 9: 35–58. University and University of Georgia.
Lipsky, Michael (1980) Street-Level Bureaucracy: Meyers, Marcia K., Glaser, Bonnie, and MacDonald,
The Dilemmas of the Individual in Public Services. Karin (1998) ‘On the Front Lines of Welfare
New York: Russel Sage Foundation. Delivery: Are Workers Implementing Policy
Lynn, Laurence E. Jr., Carolyn J. Heinrich and Reforms?’ Journal of Policy Analysis and
Carolyn J. Hill (2001) Improving Governance: A Management, 17(1): 1–22.
New Logic for Empirical Research. Washington Meyers, Marcia K. And Susan Vorsanger (2003)
D.C.: Georgetown University Press. ‘Street Level Bureaucrats and the Implementaion
Matland, Richard E. (1995) ‘Synthesizing the of Public Policy’ in B. Guy Peters and Jon Pierre
Implementation Literature: The Ambiguity- (eds.), Handbook of Public Administration.
Conflict Model of Policy Implementation,’ London: Sage Publications, pp. 245–55.
Journal of Public Administration Research and Moe, Terry M. (1984) ‘The New Economics of
Theory, 5(2): 145–74. Organization,’ American Journal of Political
May, Peter J. (1999) Toward A Future Agenda Science, 28: 739–77.
for Implementation Research: A Panelist’s Notes. Moe, Terry M. (1989) ‘The Politics of Bureaucratic
Prepared for the annual meeting of the Western Structure,’ in John E. Chubb and Paul E. Peterson
Political Science Association in Seattle. Department (eds.), Can the Government Govern? Washington
of Political Science, University of Washington. D.C.: Brookings.
May, Peter J. (2003) ‘Policy Design and Nielson, Vibeke Lehmann (2002) Dialogens Pris.
Implementation’ in B. Guy Peters and Jon Pierre Uformelle spilleregler, ressourceasymmetri og
(eds.), Handbook of Public Administration. London: forskelsbehandling i offentligt tilsyn. Aarhus:
Sage Publications, pp. 223–33. Politica.
May, Peter J. and Raymond J. Burby (1998) ‘Making Nørgaard, Asbjørn S. and Thomas Pallesen (2003)
Sense Out of Regulatory Enforcement,’ Law and ‘Governing Structures and Structured Governing:
Policy, 20: 157–82. Local Political Control of Public Services in
May, Peter J. and Søren Winter, (1999) ‘Regulatory Denmark.’ Journal of Public Administration
Enforcement and Compliance: Examining Research and Theory 13(4): 543–61.
Danish Agro-Environmental Policy,’ Journal of O’Toole, Laurence J., Jr. (2000) ‘Research on Policy
Policy Analysis and Management, 18(4): 625–51. Implementation: Assessment and Prospects,’
Peter-3383-Chapter-08.qxd 6/8/2006 7:51 PM Page 166
Journal of Public Administration Research and Employment and Training. New York: The Falmer
Theory, 10: 263–88. Press, pp. 109–38.
O’Toole, Laurence J., Jr. (2003) ‘Interorganizational Winter, Søren (1986b) ‘How Policy-Making Affects
Relations in Implementation’ in B. Guy Peters Implementation: The Decentralization of the
and Jon Pierre (eds.), Handbook of Public Danish Disablement Pension Administration,’
Administration. London: Sage Publications, Scandinavian Political Studies, 9: 361–85.
pp. 234–44. Winter, Søren (1990) ‘Integrating Implementation
O’Toole, Laurence J., Jr. and Robert S. Montjoy (1984) Research,’ in Dennis J. Palumbo and Donald J.
‘Interorganizational Policy Implementation: A Calista (eds.), Implementation and the Policy
Theoretical Perspective,’ Public Administration Process. New York: Greenwood Press, pp. 19–38.
Review, 44(6): 491–503. Winter, Søren (1994) Implementering og effektivitet.
Pressman, Jeffrey L. and Aron Wildavsky, (1973) Aarhus: Systime.
Implementation. Berkeley: University of California Winter, Søren (1999) ‘New Directions for Imple-
Press. mentation Research,’ Policy Currents – Newsletter
Rossi, Peter H. and Howard E. Freeman, (1989) of the Public Policy Section of the American
Evaluation, 4th ed. Newbury Park, London and Political Science Association, 8(4): 1–5.
New Delhi: Sage. Winter, Søren C. (2002) Explaining Street-Level
Sabatier, Paul A. (1986) ‘Top-Down and Bottom- Bureaucratic Behavior in Social and Regulatory
Up Approaches to Implementation Research: A Policies. Paper prepared for the Annual Meeting
Critical Analysis and Suggested Synthesis,’ Journal of the American Political Science Association in
of Public Policy, 6(1): 21–48. Boston, 29 August–1 September 2002. Danish
Sabatier, Paul A. and Hank C. Jenkins-Smith, National Institute of Social Research
(eds.) (1993) Policy Change and Learning: An Copenhagan.
Advocacy Coalition Approach. Boulder: Westview Winter, Søren C. (2003) ’Implementation’ in B. Guy
Press. Peters and Jon Pierre (eds.), Handbook of Public
Salamon, Lester M. (ed.) (2002) The Tools of Administration. London: Sage Publications,
Government: A Guide to the New Governance. pp. 205–11.
New York: Oxford University Press. Winter, Søren C. (2003) Political Control, Street-
Sætren, Harald (2005) ‘Facts and Myths about Level Bureaucrats and Information Asymmetry in
Research on Public Policy Implementation: Out Regulatory and Social Policies. Paper prepared for
of Fashion, Allegedly Dead, But, still very much the Annual Meeting of the Association for Policy
Alive and Relevant’, The Policy Studies Journal, Analysis and Management held in Washington
33(4): 559–82. D.C. 6–8 November 2003. Danish National
Scholz, J. T. and Wei, F. H. (1986) ‘Regulatory Institute of Social Research, Copenhagen.
Enforcement in a Regulatory System,’ American Winter, Søren C. and Peter J.May, (2001) ‘Motivations
Political Science Review, 80: 1249–70. for Compliance with Environmental Regulations,’
Vedung, Evert (1995) ‘Policy Instruments: Journal of Policy Analysis and Management,
Typologies and Theories,’ in M. -L. Bemelmans- 20(4): 675–98.
Videc, R. Rist and E. Vedung (eds.), Policy Winter, Søren C. and Peter J. May, (2002) ‘Infor-
Instruments and Evaluation. New Brunswick, NJ: mation, Interests, and Environmental Regulation,’
Transaction Books, pp. 21–58. Journal of Comparative Policy Analysis, 4(2):
Williams, Walter and Richard F. Elmore, (eds.) 115–42.
(1976) Social Program Implementation. New York: Yin, Robert K. (1982) “Studying the Imple-
Academic Press. mentation of Public Programs,” in W. Williams,
Winter, Søren (1986a) ‘Studying the Imple- R. F. Elmore, J. S. Hall, R. Jung, M. Kirst, S. A.
mentation of Top-Down Policies From the MacManus, B. J. Narver, R.P. Nathan and R. K. Yin,
Bottom-Up: Implementation of Danish Youth Studying Implementation: Methodological and
Employment Policy,’ in Ray C. Rist (ed.), Administrative Issues. Chatham N.J.: Chatham
Finding Work: Cross National Perspectives on House Publ., pp. 36–72.
Peter-3383-Chapter-09.qxd 6/8/2006 7:59 PM Page 167
Section Two
Substantive Policy Areas
Peter-3383-Chapter-09.qxd 6/8/2006 7:59 PM Page 168
Peter-3383-Chapter-09.qxd 6/8/2006 7:59 PM Page 169
9
Constitutions and Rights
JOHN UHR
of citizenship as part of the ongoing struggle of to find practical solutions to policy problems.
political activity. For the purposes of this chapter, we suggest that
The second story is about how constitutions those telling the story are in some sense part of
frame policy disputes over rights. Core rights the story, drawn to the study because of their
like that of citizenship can be embedded in fun- interest in improving policy processes and out-
damental laws, such as a constitution, where comes. This situation gives rise to a number of
they are protected against government tempta- ethical issues about the rights and responsibili-
tions to tamper with them or reduce their scope. ties of policy analysts, particularly when deal-
But other rights are creatures of statute, made ing with policy debates over equality rights. We
by the legislature and able to be remade by a can demonstrate that these ethical issues have
subsequent legislature. Many policy rights fit been with the discipline of public policy analysis
this model, as, for example, laws on social secu- since its foundations by reviewing a neglected
rity which define entitlements, or laws on public aspect of the well-known Friedrich-Finer
health insurance which define rights of access to debates from the 1940s over the democratic
medical facilities and services. Constitutions can norms of administrative discretion in the
protect the rights of legislatures to make laws policy process – Friedrich defending the ‘inner
but they need not protect these legal rights from checks’ of public service professionalism against
later alteration or modification by the same leg- Finer’s call for external political accountability.3
islatures when dominated by different legisla- As we will show, this debate brings together
tors. Still another class of rights derive from old concerns for democratic rights and new
discretionary recognition by executive govern- concerns for policy analysis. Friedrich’s case is
ments which, unless the legislature otherwise part of a wider justification of the innovation
provides, can set their own terms and condi- of academic public policy analysis, designed to
tions on rights of access to many government increase bureaucratic capacities for responsible
programs. But rights extended by one govern- policy work. For our purposes, Friedrich and
ment can be withdrawn by a successor govern- Finer mark out two analytical perspectives on
ment. A final class of rights are judge-made the study of rights in the policy process.
rights, where courts or tribunals can recognise Friedrich represents an applied and quite per-
rights flowing from ordinary laws or constitu- sonal perspective, aiming for a practical impact
tions or even from such traditional sources as through improvements in government capaci-
‘the common law’ (meaning other judges, for ties to promote citizen rights as beneficiaries of
the most part). How each or any of these rights- government programs. Finer represents a valu-
determination processes operates will depend able and equally personal skepticism about the
on the precise constitutional circumstances ambitions of public policy analysis, based on
of particular policy systems. The important his conviction that democratic rights are best
point is that constitutions affect rights-makers protected through political rather than admin-
(among the various branches of government) as istrative institutions, operating in the political
well as rights-takers (among the citizenry).2 process with its traditional constitutional safe-
Time now to get personal. Our third story is guards against misguided rights enthusiasts,
about the story-teller, in this case the practi- including well-intentioned policy activists
tioner of public policy analysis. The study of within government. The work of most policy
public policy is not easily divorced from the analysts falls between these two poles, but it is
private politics of the person undertaking the important to note this foundational debate
study. Few policy analysts admit to thorough- over the rights and responsibilities of policy
going neutrality in their work and this personal analysts in and out of government. Thus, this
imprint is probably clearer in the study of challenging topic of constitutions and rights
rights-policies than other fields. Despite its impacts on ‘our’ rights as students of public
theoretical ambitions, much of the best policy policy as much as on the rights and wrongs of
analysis is applied research, where analysts want those policy actors whom we study.
Peter-3383-Chapter-09.qxd 6/8/2006 7:59 PM Page 171
the last half century we have also seen that BACK TO THE BEGINNING
nation-states can agree to submit to various
international constitutions, either loose arrange- In many ways, the intellectual foundations of
ments like the United Nations or tighter the academic study of public policy rest on
arrangements like the European Union. At rights or at least a view about how rights can
whatever level, constitutions and rights can be promoted through public policy. An exam-
restrain the policy process when policy partic- ple of this view is evident in the establishment
ipants feel constrained to modify their conduct of the first academic journal to be called Public
and their policy ambitions consistent with the Policy, co-edited by Carl Friedrich of Harvard.9
legal norms associated with constitutions and This famous journal is the original predecessor
rights. In this sense, constitutions and rights of the current Journal of Policy Analysis and
are thought to be ‘off limits’ to the policy Management, which contains many inter-
process, providing the boundaries of accept- national reports on policy debates over rights.
able policy for those engaged in the political The aim of the founding editors of Public
contest over policy options. At yet another Policy was to devise an analytical perspective
level, the democratic policy process is never on ‘public action’ that would overcome tradi-
really locked down by unchangeable legal con- tional barriers of specialisation among the
straints and is best understood as a work in social sciences and that most solid of all
progress, with few limits to the policy implica- analytical barriers: the separation of theoreti-
tions capable of being derived – by executive, cal studies of ‘public action’ from the practical
legislative or judicial institutions – from con- applied studies common to the academic study
stitutions and rights. Even in the absence of of public administration. Their aim was to
formal change, constitutions and rights are bring greater cohesion to ‘the analysis of the
used in new ways to reflect changing commu- conditions which determine the success or
nity standards by policy actors across the failure of whatever modern government
system of government.7 undertakes’.10 The decade leading up to this
This chapter provides an overview of innovation had seen a revolution in ‘whatever
debates over policy implications of constitu- government undertakes’, with social-democracy
tions and rights. Our theme is that rights are modifying earlier liberal-democracy. The New
best considered as claims – particularly claims Deal in the US had overhauled the regulation
on government for policy recognition – and of the political economy of the world’s most
that different constitutional arrangements wealthy industrial nation-state, ushering in
shape the political process for making and new forms of state intervention in what had
determining competing claims.8 Drawing traditionally been unregulated aspects of social
examples from each of the three levels just out- life. A welfare state had arisen, with more gen-
lined, we provide a stylised account of democ- erous social protection for employees and
ratic policy arguments, including arguments stricter regulation of the social obligations of
about the place of constitutions and rights in private enterprise. Friedrich argued in his own
public policy analysis. We begin by recounting contribution to the first issue of Public Policy in
an episode from the foundations of the intel- 1940 that the new discipline of public policy
lectual history of the modern academic disci- was an attempt to modernise the study of
pline of public policy, revealing a fascinating public administration by bringing a wider
scholarly debate over the place of constitutions range of social sciences to the analysis of public
and rights in the practice as well as the theory decision-making.11 The emergence of this new
of public policy. We return to this foundational journal marked the arrival of the larger acade-
debate in the conclusion to this chapter, mic enterprise of public policy as a multidisci-
emphasising just how far the discipline has plinary effort to marshall social science for the
traveled in the days since its foundations, purposes of better government, based on more
despite the apparently permanent uncertainty intelligent use of evidence and argument by
over constitutions and rights. government officials.
Peter-3383-Chapter-09.qxd 6/8/2006 7:59 PM Page 173
Looking back now and applying contempo- to the people. The traditional constitutional
rary terms, we can see here the beginnings of arrangement was one where the undoubted
evidence-based policy-making. What has not power of state officials was held in check by
been properly appreciated is the debated place their accountability to elected representatives of
of constitutions and rights in this novel public the people. With the emergence of the public
policy orientation. We can take Friedrich as an policy movement, this constitutional arrange-
exemplary exponent of the new discipline. His ment, with its neat divisions between elected
foundational article in the initial issue of and unelected officials, and between policy and
Public Policy is something of an intellectual administration, is challenged in the name of a
charter for the emerging discipline of public more effective and responsible form of democ-
policy. His theme was the importance of ratic government. In Friedrich’s orientation, the
‘administrative responsibility’, by which he new discipline of public policy will not only
meant the constructive contribution to policy- increase the analytical power of state bureau-
making made by those holding administra- cracy but also increase its capacity to act as a
tive as distinct from political power. When responsive institution of democratic govern-
Friedrich’s article is seen against his contribu- ment. Friedrich locates this version of democra-
tion to the famous Friedrich-Finer debates tic accountability in the professional sense of
over the competing forms of democratic public service responsibility cultivated by public
accountability,12 we can see the importance of officials as a condition of their discretionary
his reordering of traditional relationships powers. These powers are partly policy, partly
between constitutions and rights. Traditional administrative – just as the emerging science of
approaches to the policy process rest on dis- public policy seeks to train officials in the new
tinctions between a political realm of ‘policy’ skills of policy analysis as well as the old skills of
and a non-political realm of ‘administration’. administrative competence.
In Finer’s hands, this older approach is associ- Friedrich’s championing of public policy
ated with a larger political theory about consti- analysis had major implications for the consti-
tutional government, with elected politicians tution of rights. A state bureaucracy empow-
responsible to voters for public policy and ered by new forms of public policy analysis
unelected officials responsible – primarily to could act as a responsive instrument of govern-
elected politicians – for public administration. ment services by satisfying public rights to
Friedrich distanced himself from this tradi- services in ways not contemplated by tradi-
tional constitutionalism by blurring these dis- tionally unresponsive state bureaucracies. In
tinctions in ways that left unelected public Friedrich’s formulation, this was a vision of the
officials with substantial discretionary power welfare state, with more than a hint of policy
over policy as well as administration. This paternalism. The rights-attentive bureaucracy
model of social democracy had great potential was not envisaged as directly responsive to cit-
for official regulation of industry as well as izens, as with later developments we will exam-
welfare. Friedrich’s aim was not to take power ine below. Instead, in this foundational era of
away from elected politicians but to recognize public policy, the welfare bureaucracy was seen
the inevitable power of government bureau- as taking responsibility itself for the identifica-
crats in modern democratic systems and to try tion of a rights agenda, and of being responsive
to devise ways in which that power can be to the needs, if not the actual claims, of rights
increased to serve democratic policy ends – to government services.
including the protection of economic, social As Finer suspected, this new model of
and political rights. responsiveness departed from many of the
Traditional approaches held that admini- traditional norms of responsiveness to governing
strative power ideally serves policy ends by authorities. By contrast, Friedrich’s confidence
confining itself to questions of policy imple- reflected the growing faith in policy analysis:
mentation, leaving policy-making in the hands of he shared the hope that the innovation of
elected politicians who are directly responsible policy analysis would strengthen responsible
Peter-3383-Chapter-09.qxd 6/8/2006 7:59 PM Page 174
modern democratic countries – a range of policy process has been transformed by the
indigenous peoples, a bicultural social tradi- court’s growing policy arbitration of interest
tion with two settler peoples (French and group conflicts that have become congested in
English), intergovernmental complexity result- the other two branches of government. The
ing from well over a century of experience of 1982 Charter of Rights and Freedoms is not so
federal government, widely diverse ethnic pop- much the cause of this transformation as the
ulations from extensive recent immigration, vehicle driven by the real cause – policy
and multiple perspectives on the national activists who have seized on the new policy
government’s internationally pioneering potential of the courts to settle rights-claims
programmes of multiculturalism. unresolved by legislatures and governments.
As part of this remarkable mosaic of diverse We acknowledge, of course, that in other
identities, Canadian political values responded nations with no such explicit rights-charter,
positively to the prospect of a formal bill of the judiciary can also get involved in what can
rights to lay down some core principles of fair be called ‘judicial policy-making’ by discover-
treatment of citizens by governments. What ing implied if not explicit constitutional rights
began as a parliamentary ‘bill of rights’ statute that governments must respect: the high courts
in 1960 gathered institutional strength to of France, Israel and Australia are all exam-
emerge as the national Charter of Rights and ples.14 But we see the Canadian policy process
Freedoms in 1982, resulting in an increased as nicely illustrating contemporary democratic
policy role for the Canadian courts as a central policy-making writ-large, with the policy
forum for resolving policy conflicts over com- process charged with competing claims against
peting rights. Some of the rights claims have government, at one point demanding the
been traditional civil liberties claims against courts to restrain government from undue
government interference with the liberties of interference with liberties and at another point
citizens (e.g. restrictions on police evidence in demanding the courts to direct governments
court proceedings). Others have been more to interfere to guarantee equality of opportu-
contemporary equality claims against govern- nity for neglected groups. We see the Canadian
ment failures to provide services to discrimi- policy process as containing an armoury of
nated groups (e.g. prisoners’ voting rights). arguments over relationships between consti-
These two categories of claims illustrate the tutions and rights – with a growing case load
broad range of rights preoccupations: at one of public policy dispute over such contempo-
end, ‘negative’ rights to protect individual liber- rary challenges over rules relating to disability,
ties held in common by all citizens and, at the discrimination, employment, family status,
other end, ‘positive’ rights to promote equality hate propaganda, race, religious and sex dis-
for neglected groups who have been denied crimination, and sexual harassment and sexual
access to services necessary for full citizenship. orientation.
Both ends address citizenship claims, but they
do so in different ways with different balances of
the two core values of liberty and equality.
MODELLING RIGHTS DETERMINATION
One particular relevance of the Canadian
story is that it shows how the formal provisions
of a nation’s constitution can be interpreted by We will return to the Canadian story below,
different rights-claimants in different ways, when trying to identify parallels with the story
sometimes elevating individual liberties over in other polities. But before we enter the
equality for neglected groups and at other thicket of rights and wrongs, we will try to
times elevating group equality over individual clarify the general nature of rights resolution
liberties. Constitutions manage the policy in the policy process. Our contention that
process through their allocations of powers rights can be usefully understood as claims and
and responsibilities among legislative, execu- analysed by reference to ‘claims systems’ draws
tive and judicial institutions. The Canadian on Peattie and Rein who in turn draw on the
Peter-3383-Chapter-09.qxd 6/8/2006 7:59 PM Page 176
clash of rights associated with feminism and in the wake of the military defeat of the slave
the women’s movement.15 Their influential states did the claims of the opponents of
study focuses largely on policy battles in the slavery emerge victorious. But, even then,
United States, where so much of public policy reformers had to institute a formal constitu-
is argued out by reference to the right to ‘equal tional amendment to put their preferred policy
protection of the laws’ contained in the US beyond legal doubt.
Constitution. This constitutional provision is Consistent with this, Peattie and Rein
itself a good example of rights as claims, since acknowledge that their approach to ‘claims sys-
it is contained in an amendment to the tems’ puts more emphasis on politics than on
Constitution that arose from the Civil War law. Legal battles emerge as means towards
struggles over slavery. This provision reflects political ends, with rights established only
the commitment to equal rights of citizenship where rights proponents win the political
for blacks and whites that only came about battle over the agenda of public policy. The
because of the military victory of the Union claims Peattie and Rein deal with are claims on
over the secessionist South. In this case, tensions government – articulated by groups demanding
within the ‘claims system’ could only be provision of government services ‘as of right’.17
resolved through a momentous civil war. In Claims tend to develop through a structured
extreme cases, might can bring right. But process of steadily increasing demand. Claims
extreme cases also illustrate the might that on government tend to begin as pleas for assis-
suppresses right: in this case, the might of tance, turn into requests for closer considera-
social groups with vested interests in slavery. tion, before becoming demands for urgent
The right to ‘equal protection of the laws’ action. We can think of this as a three-fold
sounds quite basic but it was denied to sub- process involving all three branches of govern-
stantial racial minorities by established social ment, beginning with routine requests to the
groups, who benefited economically from slav- executive government to modify aspects of
ery. The struggle to right the wrongs of slavery public administration, followed if necessary by
is an awesome example of the war-like poten- political pressure on legislators to write new
tial of the policy process in modernising laws to authorise preferred practices, and
nations, and of the lengths that democratising finally followed if necessary by legal challenges
systems may have to go to resolve disputed against the government to compel compliance
claims about central values of citizenship. with the claimed right.
The civil war over equal citizenship rights is Claims systems vary, with the constitutional
a good example because it illustrates a default order standing out as the largest domain
position not uncommon in democratic soci- within modern nation-states. As we argue later,
eties, where competing social interests have many contemporary rights disputes involve
fundamentally opposed views on rights- claims on international authorities (such as the
claims, even to the point that a head of gov- human rights committee of the United
ernment like President Lincoln understood Nations) to take action against recalcitrant
that his constitutional duty was not simply to national governments. But within the nation-
erase slavery but to overturn what we might state, the battle is typically over domestic
now term judicial policy-making – which had public policy priorities, which can be won even
justified slavery by rejecting minority claims to in the face of determined opposition by
equal rights. Not for nothing does Bernard executive government. Although victory in this
Crick use the drama of Lincoln’s presidency as battle has enormous implications for govern-
his primary example of democratic policy- ments, the claimants know that victory can
making.16 To oversimplify: opponents of slavery come through the acceptance of their claim by
claimed that the US constitutional order legislative or judicial institutions as much as by
recognised the rights of African–Americans as the institutions of executive government. Each
citizens; proponents of slavery countered that branch of government has its distinctive
there were no such constitutional rights. Only claims-system. At the heart of government is
Peter-3383-Chapter-09.qxd 6/8/2006 7:59 PM Page 177
the executive administration with its ‘realm of interests driving purposes’. Many rights-advocates
interpretation’, causing claims advocates to justify their claims in terms of benefits to the
appeal to revised ‘interpretations’ of the rules general interest in equal citizenship rights rather
and regulations administered by the offending than to neglected sectoral interests. While it is
government.18 As alternatives, claims advocates true that many conventional sectoral interests
can pursue their cause through the courts in disguise their specific interests under the rhetor-
order to obtain a definitive legal interpretation ical banner of general interests, it is a distinctive
in favour of their rights; or they can mobilise feature of many rights-claimants that they
support so that the political assembly can pass shape the policy process to focus on arguments
new legislation to superimpose a favourable about citizenship and ‘the idea of the public
definition over less sympathetic interpreta- interest’ with claims based on ‘a rationale of
tions from either executive government or the collective purpose’.19
judiciary.
Rights-advocates often present rights as
non-negotiable. But even this strategy is part
RIGHTS AND POLICY-MAKING
of a larger process of social negotiation over
public policy priorities. Established rights
might appear to be part of the natural order of We return now to the Canadian story, hoping
social arrangements, but almost every right to identify links to developments in the policy
had at some point to be fought for through a process of other nations. No one nation stands
social process of ‘requests, bids, bargins or out as exemplifying contemporary practices in
negotiation’. As Peattie and Rein map this out, public policy. But Canada is relevant because
we see that competing interests engage in over recent years it has generated a remarkable
political competition to ‘demand, extract, public dialogue over the competing norms and
request, enforce their bids for resources’. principles of democratic policy-making. Our
Calling this ‘political’ might seem to suggest interest is not so much in what Canadian
that rights can be reduced to the social and policy actors have done as what they have said
economic interests characteristic of their polit- about what needs to be done, including what
ical advocates. While partly true, such a view needs to be done about aligning constitutions
misleads us by taking attention away from the and rights.
important role of ideas and argument in the The 1982 Canadian Charter of Rights and
policy process. Freedoms begins by declaring that its rights of
Rights policies are classic expressions of the citizenship are subject to only those ‘reason-
power of ideas in the public sphere. Peattie and able limits’ which can be ‘demonstrably justi-
Rein argue that rights policies illustrate the fied in a free and democratic society’.20 This
neglected importance of ‘a struggle of ideas, a clearly deals the courts into the policy process
struggle over the extension and application of as the arbiters of claims to ‘reasonable limits’
various principles or claims rationales’ in the by the two other political branches of govern-
policy process. They contend that ‘one of the ment: parliament and the political executive.
basic elements in the claiming process is an But, just as importantly, the Charter reinforces
intellectual one: a definitional element’, usually the power of the courts by providing a lengthy
involving the ‘development of a new political list of rights with which Canadians may expect
language’ as illustrated by the deliberately con- governments to comply. The Charter defines a
frontational rhetoric about politics of the per- set of so-called ‘fundamental freedoms’, deal-
sonal associated with the women’s movement. ing with free thought, free expression and free-
Finally, it is useful to remember that rights- dom of assembly. These fundamental freedoms
claims do not necessarily reflect economic or are then followed by an extensive articulation
material interests on the pattern of so much tra- of more specific rights – such as a list of core
ditional interest-group politics. As Peattie and ‘democratic rights’ (primarily the right to
Rein note: ‘purposes create interests as well as vote), ‘mobility rights’ (very relevant to a
Peter-3383-Chapter-09.qxd 6/8/2006 7:59 PM Page 178
federal polity with a history of division exaggerate the apparent victory of leftish
and separatism), ‘legal rights’ (eight detailed causes in using judicial review to promote their
provisions relating to criminal procedure), preferred policy agenda. Constitutions can just
‘equality rights’ (a combustible combination as easily favour rightist causes, as shown in
of anti-discrimination and affirmative action earlier eras, and not only in Canada, when con-
declarations), and finally language and language- servative forces used prevailing legal doctrines
of-choice education rights (at least for French of ‘judicial restraint’ to reinforce a conservative
and English speakers). The traditional ‘treaty hegemony. To use the old adage: the constitu-
rights’ of Aboriginal peoples of Canada are tion is what the judges say it is. Therefore, what
separately identified (including new rights to matters is not so much the internal text of the
participate in constitutional change processes). constitution as the external political context
This 1982 change to the Constitution of of judicial interpretation, including the evo-
Canada has given the Supreme Court of lution of interpretative doctrines about judi-
Canada a prominent role in arbitrating rights- cial policy-making. There are always powerful
claims against Canadian governments. Policy entrenched interests favouring judicial
activists have used the new rights system to restraint and always vocal outsiders favouring
‘constitutionalize policy preferences that could judicial activism. The power of the judiciary is
not easily be achieved through the legislative just as crucial to policy outcomes in eras of
process’.21 In the first 16 years of Charter juris- restraint, when courts decide not to intervene,
diction, the Court struck down nearly 60 fed- as in eras of activism, when courts decide to
eral and provincial statutes as incompatible intervene. That is, non-decisions (i.e. decisions
with the new Charter. While the Court might not to intervene) are decisions too. Second, and
not have become a policy-making body, the building on this, we note that the new rights
courtroom has become ‘a policy-making arena’ system in Canada has not established a con-
because policymaking has become ‘judicialized, sensus about preferred policy rights or pre-
legalized, and conducted in the vernacular of ferred policy outcomes. Even those policy
rights talk…’.22 We repeat our warning that this activists favouring judicial intervention are
new policy-making arena is one that, in princi- divided about the policy priorities deserving
ple, can benefit the rights of social establish- judicial intervention. For instance, the rights-
ments as much as the rights of the socially camps of negative liberty (e.g. civil libertari-
excluded. It all depends on the policy responsi- ans) and positive equality (e.g. feminists)
bilities and interpretative legal doctrines enter- continue, as before, to struggle against one
tained by the courts themselves. What one another – but now before the courts rather
generation of judicial officers might favour can than before the legislature.
be overturned by a succeeding generation. An example of the new policy activism
Canadian commentators have noted that con- comes from the Charter’s equality rights (s15),
temporary rights-advocates have turned to the where policy activists opposed to various
courts in ways that are reminiscent of the strat- forms of systemic discrimination – ‘women,
egy of earlier ‘business elites … (who) used visible and religious minorities, the mentally
litigation to slow the advent of the emergent and physically disabled, and the elderly’,
welfare state’. This potential variability in judi- together with gays and non-citizens – have
cial policy-making postures has provoked pursued ‘judicial revision of legislative
renewed attention in Canada to the policy decision-making’.24 Unlike civil libertarian
implications of appointments processes for activists, these new social movement activists
judges, with hopes that greater public trans- have sought more rather than less government
parency might weaken the power of serving regulation of social affairs, with governments
governments to stack courts with policy being called on to resolve inequalities of access
favourites.23 to an increasing range of private services (e.g.
We note two aspects of this new rights- employment terms and conditions) as well as
responsive policy process. First, we should not public services (e.g. employment training).
Peter-3383-Chapter-09.qxd 6/8/2006 7:59 PM Page 179
Where is ‘the government’ in the larger story policy ‘dialogue’ between governments and
of rights struggle? Are governments simply the courts. This debate reflects widespread uneasi-
plaything of judicial arbitration of contested ness over the political powers exercised by the
rights? Public policy is often effected by gov- courts in the policy process, with various reform
ernments who pride themselves on getting options to limit the policy clout of courts to
ahead of the prevailing political demands. ‘dialogue’ about, rather than determination of,
Particularly in the area of citizenship rights, policy outcomes. An example of this option is
governments can aspire to policy leadership the new role of British courts in ‘declaring’
that puts demands on themselves shaping incompatibility of suspect British legislation
rather than reflecting rights-claims. We con- with the European Human Rights Convention,
clude this section by noting that governments thereby calling on government to participate in
can shape social movements as much as social public ‘dialogue’ with courts over the necessity
movements can shape governments. Again, for the suspect legislation. In this model, final
Canada is a good example. For instance, it was power rests with executive government to man-
only after the Trudeau government in 1971 age the process of resolving instances of ‘incom-
proclaimed a national policy on multicultural- patibility’. This example of course relates to
ism that government bureaucracies actively judicial views about incompatible legislative
engaged in programmes of ‘social animation’ schemes rather than public policies as such. But
to bring into existence funded bodies repre- it is held out as an example of judicial restraint
senting relevant ethnic groups.25 This is a case worthy of consideration by policy analysts.
of how the family of identity rights, to take but Another example is the recently established Bill
one prominent example, can be promoted by of Rights in the Australian Capital Territory,
executive governments exercising a form of which is Australia’s first statutory bill of rights:
policy leadership that owes very little to interest one which allows the legislature to override
group advocacy. But, once formed, rights court ‘declarations’, but only after a process of
groups can take publicly-funded government deliberate public ‘dialogue’ between govern-
programmes off in directions unanticipated or ments and courts over the policy merits of the
even undesired by sponsoring governments – suspect legislation.28 These examples illustrate
possibly playing off courts against the other in quite practical ways the potential for courts to
branches of government. The larger point is operate as mediating rather than determinative
that, despite our focus here on rights policy as institutions, with their policy powers limited
determined by the courts, rights-claimants are to generating greater public accountability (or
‘often as much the result of key public policies better, perhaps, ‘answerability’) in the policy
as the source of those policies’.26 It is important process.
to remember that governments can govern and The ‘dialogue’ metaphor arose from US
that a vital part of democratic governing is set- commentary on judicial review but has now
ting and seeding the policy agenda – if other expanded around the globe to inform wider
governing institutions will permit, an issue to debates about overlaps between the policy
which we now turn. process and the judicial process. Concepts of
‘dialogue’ circulate as ways that supporters of
judicial review of government action can
Institutional implications describe the ideal form of judicial participa-
tion in the policy process. This ideal distin-
The Canadian experience has fed into wider guishes the legal role of courts in exercising
international policy debate over the appropri- judicial power from the more limited role of
ateness of judicial policy-making to norms of courts in ‘dialoguing’ with the political
democratic governance.27 We have discussed branches of government about the legal form
tensions between political executives and the of policy programmes. Critics of judicial
courts and here we should note a growing inter- review fear that ‘dialogue’ does not properly
national debate over theories and practices of describe the considerable policy impacts that
Peter-3383-Chapter-09.qxd 6/8/2006 7:59 PM Page 180
disputes about legal form can have on the Second, the debate over ‘dialogue’ revives
substance of public policy. There are exagger- older constitutional debates over separation of
ated positions at both ends of this debate powers. There is little scope for serious discus-
over constitutions and rights: supporters of the sion of policy rights in the absence of discus-
rights of elected representatives to determine sion of the web of political powers used to
policy priorities tend to revive the old policy/ manage rights claims. The competing perspec-
administration distinction to limit the policy tives on the relative balance of governments
reach of judicial institutions; and advocates and courts draw on deeper, more fundamental
of ‘dialogue’ claim that judicial power can principles of constitutional design. This is
improve the policy process by drawing govern- recognised by those participants who acknowl-
ments into public debate over the public edge the place of ‘liberal constitutionalism’ in
merits of suspect policy initiatives. There is no conditioning the relationships among different
agreed best answer, in part because the range constitutional organs which are not simply
of permissible forms of democratic govern- autonomous in their choice of public roles.31
ment is broader than the range of arguments Constitutional doctrines about the separation
about the relative merits of governments and of powers provide the scholarly backdrop
courts in determining the rights of those against which the ‘dialogue’ debate is acted out.
affected by public policies.29 But, when trying to locate the rights at issue in
Three aspects of this growing international this policy debate, much can be gained by an
debate are of importance here. First, the focus explicit examination of the range of ways that
on ‘dialogue’ highlights the extent to which governmental powers can be constitutionally
rights disputes depend for their resolution on separated and re-arranged. The ‘executive
extensive public discussion rather than simply government’ and the ‘courts’ are not the only
decisive action by governments or courts. The constitutional actors: formally one has to add
effectiveness of most public policies is the ‘legislature’ and more substantively one has
increased when they have widespread commu- to add ‘the people’ on whose consent legitimate
nity support, and elected governments in democratic government ideally rests. Just as
particular have leadership roles in trying to the people have rights recognised by democra-
build public support for policy initiatives. tic constitutions, so too the three branches of
Democratic governments rest on the consent of government have constitutional rights: the
the governed, and one of the important rights legislature exercising legislative powers, execu-
of democratic citizens is the right to informed tive governments exercising executive powers
consent. Public authorities have to respect the and the judiciary exercising judicial powers. As
community’s reasonable expectation that their we have seen, rights-claims can be complaints
consent should, ideally at least, not be forth- about powers or pleas for more powers. Rights-
coming until those in authority have convinced claims can never be resolved or properly
them of the merits of policy proposals. managed without close attention to the consti-
Elections are one but not the only means tution of public powers framing them.
through which informed consent can be tested. Third, the ‘dialogue’ debate raises questions
Open debate and public argument over policy about which particular institutions are best
options are just as important a means of testing placed to participate in public argument over
informed consent. This rolls into wider consid- contested rights and related policy priorities.
eration of features of deliberative democracy We suspect that the legislature is usually hon-
(such as opportunities for oppositions to ques- oured in a tokenistic fashion. Opponents of
tion and scrutinise governing parties under fair judicial activism advance arguments in defence
and impartial procedures accepted by both) of the rights of elected political assemblies to
that might be used to reshape the everyday determine policy priorities. But, more often
practices of democratic policy-making to bring than not, they really mean to defend the rights
about more evenhanded consideration of com- of the political executive to use their party-
peting policy perspectives.30 political influence to manage legislative
Peter-3383-Chapter-09.qxd 6/8/2006 7:59 PM Page 181
assemblies along policy pathways determined seek to develop citizenship but in quite different
by the executive. In this scheme, the legislative ways, reflecting contrasting interpretations of
assembly plays its part by granting formal con- democratic rights.
stitutional approval to policy choices originat- There are also models in between these two
ing in the political executive.32 This confidence alternatives, as this section attempts to show.
in the ability of executives to mobilise the One of the most prominent is the rise of
power of legislative assemblies reflects a very government-sponsored charters of responsibili-
traditional orientation to parliamentary poli- ties and obligations to balance the rights orien-
tics that is increasingly at odds with the institu- tation favoured by so many policy activists. A
tional independence open to parliaments. Our good example is the framework of ‘mutual
point here is that the ‘dialogue’ metaphor obligation’ introduced by the conservative
suggests two players, but this is at odds with Howard government elected in Australia in
the underlying dynamics of the three separated 1996. This framework was devised to modify
constitutional powers common in modern the notion of welfare rights advocated by many
democratic systems. Our conclusion is that policy activists and adopted by beneficiaries of
‘dialogue’ debates will be unproductive if they public services. The concept of mutual obliga-
persist in ignoring the potential for indepen- tion was intended to establish a contract of sorts
dent action by legislatures as well as political between service providers and service con-
executives and courts. Or, to put this otherwise: sumers on the basis that consumers had obliga-
rights at risk in democratic policy systems can tions to payback or return some of the
be addressed from three rather than two insti- investment that ‘the community’ was investing
tutional perspectives, as students of presidential in them. This payback might be something as
systems of government have long appreciated. simple as the voluntary acceptance of terms and
conditions of reporting work-availability while
on income assistance; or it might be more oner-
ous acceptance of training programmes to get
RECONSTITUTING RIGHTS
beneficiaries job-ready and off welfare at the
earliest opportunity. The policy framework was
One final issue merits attention. We argued not a throwback to earlier eras when welfare
that the public policy movement was origi- was considered a privilege. The innovation was
nally associated with a rights-attentive policy intended to replace the notion of rights to wel-
agenda, ideally implemented by a responsible fare with a more balanced picture of rights and
bureaucracy armed with the new skills of responsibilities – or ‘obligations’ to use the more
policy analysis. Building on the Friedrich for- formal language of the Howard government.
mulation, we suggested that this could be Mutual obligation is a conservative response
seen as a form of policy paternalism, in that to the call for welfare rights. As a policy con-
the rights agenda in question was one devel- struct about reciprocity, it is not opposed to
oped by governing officialdom and not, as in welfare as such and draws explicitly on pro-
the Canadian examples discussed above, fre- gressive policy analysis of mutual exchange
quently by governing bodies in direct response arrangements.33 This policy construct explic-
to rights-claims articulated by policy activists. itly addresses ‘participation support’ rather
We have termed this later approach a rights- than ‘welfare support’. This reflects the policy
responsive model. There is an interesting con- aim, which is to get individuals participating
trast here between an original public policy actively in society rather than passively as
model of administrative obligations to keep recipients of welfare – and as responsible citi-
alive the operational requirements of citizen- zens returning services to the community
ship, and later public policy models of govern- assisting them. The language of responsibility
ment obligations to take note of changing does not replace that of rights but balances it:
citizenship requirements as determined by ‘The right to support should be balanced by a
courts and other adjudicating bodies. Both responsibility to the community that provides
Peter-3383-Chapter-09.qxd 6/8/2006 7:59 PM Page 182
that support’.34 The Howard policy framework bring citizenship to deserving communities.
is flexible enough to justify liberal or conserv- Half a century later, the skills of policy analysis
ative experiments in welfare reform: declara- are more generally available, illustrated in the
tions to the effect that ‘the whole of society has reliance of the conservative Howard govern-
an obligation to provide assistance to those in ment for its rethink of welfare rights on
most need’ can justify the ‘leadership role of sympathetic elements of the ‘not for profit’
government’ in taxing for welfare; or it can sector – initially as policy advisors and later as
justify government deferring to private forms programme providers.
of welfare assistance. An example of the defer- There is a link here between contemporary
ment option flows from the application of conservatism and the fears of Finer reported
mutual obligations to immigrants, who can on earlier in our chapter: they are comparable
expect to be ‘subject to the two-year waiting small-government perspectives with a distrust
period for eligibility for income support’.35 of bureaucracy and of welfare-dependency. We
The Howard government framework targets began this chapter with a review of the rise of
the responsibilities of employers as well as public policy analysis, noting early fears that
employees: businesses must ‘look beyond the increased analytical competence at the disposal
interests of their shareholders or profit margin’. of government might harm rather than help
The stated policy is to ensure that individual the cause of rights.
and corporate citizens ‘behave in ways that are Taking Finer as our exemplary sceptic, we
not purely self-centred’. The administrative encountered an explicit constitutional argu-
concept of mutual obligation is ‘underpinned ment against rights-attentive policy analysis
by the concept of social obligations’ which performed by unelected public administrators.
refers to the ‘web of mutual expectations’ hold- This was not so much an argument against the
ing society together. But even a free society emerging discipline of public policy as a
requires government leadership. The power of restatement of the policy/administration
compulsion is often necessary to enforce these distinction designed to protect the integrity of
mutual obligations, with government working elective politics against misplaced enthusiasm
with and, in some cases, on businesses and for bureaucratic rule. Seen in another light,
individuals to participate as full citizens, this was an early version of the constitutional
including honouring obligations to act in case for limited government, with political
socially responsible ways by providing access checks and balances against the regulatory rule
to jobs (business obligations) and being avail- of big government.
able for training schemes (individual obliga- Finer was understandably wary about rights:
tions). Government regulation is required to his preference for an older constitutionalism
help some individuals ‘achieve their own goals with norms of limited government is consis-
for participation and self-reliance’.36 tent with much of contemporary public choice
This approach is shared by other conserva- theory, which is suspicious of rights that are
tive governments, such as the Bush adminis- not embedded in a constitutional framework,
tration, and to a not inconsiderable extent by with checks and balances against discretionary
the former Clinton administration. How does rights managed by executive government.37
this international turn to concepts of responsi- Rights claims are raised by every politically-
bility relate to the original ‘administrative ambitious interest group and Finer’s scepti-
responsibility’ framework of Friedrich? We cism is still valuable in reminding us of the
have suggested that Friedrich’s model was one merit of assessing rights-claims against explicit
of policy paternalism and there are clearly criteria of public choice – constitutionally
many elements in common with the responsi- understood in terms of duly processed popular
bility model of mutual obligation. In different consent, as distinct from bureaucratically
ways, both models rests on concepts of citizen- processed government consent. Claims of
ship. Friedrich looked to intelligent govern- rights bring with them risks of capture of the
ment, wielding the power of policy analysis, to political system by rights-advocates in and out
Peter-3383-Chapter-09.qxd 6/8/2006 7:59 PM Page 183
of government, and Finer’s voice remains In conclusion, we can identify three main
relevant when trying to get some distance from implications of these related stories. First, our
government sympathy with rights policies. argument has been that the policy process in
modern democratic regimes involves political
contests over the rights of citizenship. All poli-
CONCLUSION tics involves contests over benefits that govern-
ments can confer, and often this is part of a
power struggle between competing elites
Recapitulating the main themes in this account
occurring behind the scenes. The secrecy of
of constitutions and rights, we can summarise
such power struggles is one indication that
the three related stories presented here.
they are usually more concerned with interests
(1) Rights are claims on government for than rights. Rights as we present them here
policy initiatives to benefit particular derive from norms of citizenship grounded in
classes or groups of individuals. We high- a set of constitutional principles which are,
light the importance of one particular set ideally at least, acknowledged by all citizens. By
of rights – equality rights – for modern contrast, interests are inevitably partial, reflect-
democracy. Most rights claims call for ing a narrower range of considerations that cut
equality of access to policy benefits for across the common field of citizenship. This is
traditionally excluded or marginalised not to suggest that democratic nations will
classes or groups. We argue that many of or should display a social consensus on all or
the most prominent and enduring policy indeed any policies, including rights policies.
disputes in democratic systems arise from Democracy involves disagreement, including
disagreement over the most fundamental disagreement about the core value of equality.
of all of policy applications of equality Interpretations of democratic citizenship are
rights – the rights of citizenship. interpretations of equality and of its political
(2) We argue that constitutions are important and policy implications. At the level of political
devices for framing the public manage- theory, interpretations differ according to
ment of policy disagreement: sometimes philosophical orientation; and at the level of
by clarifying the meaning of core rights, political practice, interpretations differ from
but more generally by clarifying the proce- nation to nation and across nations according
dures for resolving policy disputes, includ- to political and policy preferences.
ing disputes over rights policies. But even Second, we have argued that constitutions
constitutions can become the object of are valuable to public policy because they for-
policy disputes, with contest over the mat, or structure or, as the name suggests, ‘con-
appropriate policy roles of executive, leg- stitute’ the policy process so that rights-claims
islative and especially judicial institutions. are considered as fairly as possible. There is no
(3) Finally, we argue that debate over rela- one preferred model of a democratic constitu-
tionships between rights and constitu- tion. Some modern democracies, like the
tions can be traced back to the origins of United Kingdom, have no single constitutional
public policy analysis, where Friedrich and text or ‘purpose-built’ foundational document.
Finer mapped out contrasting approaches Regardless, they manage to generate and regu-
that help plot the range of analytical pos- late a healthy policy process through a variety
sibilities open to contemporary policy of other means, attributing privileged status to
analysts interested in the place of consti- some historic declarations of public right, such
tutions and rights in democratic policy as the Bill of Rights of 1689. Other democratic
processes. Few policy analysts fit either nations have written constitutions sourced
extreme position; most fall between the from their nation-building period. In some
two poles; but many move between the cases, the text of these documents is
two positions, depending on the policies unchanged from that founding period, with
under analysis. little or no explicit recognition of equality
Peter-3383-Chapter-09.qxd 6/8/2006 7:59 PM Page 184
rights. In other cases, written constitutions are institutions of ‘the government’ and spread out
relatively flexible by virtue of amending across many institutions of civil society, some
processes, enabling changing community stan- with a socialist rights agenda of social justice
dards to shape the ongoing development of the protected by big government and others with a
original set of constitutional provisions. In still libertarian rights agenda of individual liberties
other cases, one or more of the core constitu- protected by limited government. Rights are
tional institutions can use its power (as in the better kept alive by open public dialogue
case of the judiciary) to revise the authorised among competing interests than by either the
interpretation of provisions; or (as in the case Friedrich tendency towards a government
of executive governments) to reshape adminis- monopoly of ‘administrative responsibility’ or
trative practices; or (as in the of legislatures) to by the Finer tendency towards the black letters
fill out the core provisions with supplementary of a constitutional provision. Contemporary
provisions, including legislating for new rights. democracy has no agreed model of the ideal
Rights are thus affected by both original con- relationship between constitutions and rights,
stitutional provisions (with or without explicit and one of the enduring strengths of public
declarations of rights) and by later develop- policy analysis is this open dialogue, over which
ments by all three branches of government. social interests deserve the attention of policy
Democratic policy processes are shaped by the analysts, within and without government.
changing institutional relationships among the
many clusters of institutions across the three
branches of government. Rights-claims vary in NOTES
the institutions of government they target; and,
when most contentious, rights-recognition can 1. A. Halpin, Rights and Law: analysis and theory. Hart
involve all three branches of government in the Publishing 1997; see also A Phillips, ‘Defending Equality of
resolution of policy disagreement. Outcome’, The Journal of Political Philosophy, 12/1, 2004,
Finally, we note some implications of our 1–19.
2. V. Bogdanor, ed, Constitutions in Democratic Politics.
story about policy story-telling itself. Just over Gower 1988; J-E Lane, Constitutions and Political Theory.
50 years ago, the emerging discipline of public Manchester University Press 1996; L Alexander,
policy found itself in the middle of a contest Constitutionalism: philosophical foundations. Cambridge
over rights. The polarised positions adopted by University Press 2001.
Friedrich and Finer left open plenty of room 3. Consider T. L. Cooper, The Responsible Administrator:
an approach to ethics for the administrative role. 4th edition.
for more practical perspectives, with different Jossey-Bass 1998, 132–140, 153–162; M. M. Harmon and
analytical approaches to constitutions and R. T. Mayer, Organization Theory for Public Administration.
rights. We have sketched in some of this middle Little Brown, 1986, 334–5.
ground by using a pluralist constitutional 4. M. Bevir, R. A. W. Rhodes and P. Weller, ‘Comparative
Governance: Prospects and Lessons’, Public Administration,
model where rights are treated as claims, usually
81/1, 2003, 191–210.
sorted through the policy process as a matter of 5. See e.g. J. K. M. Gevers and G. Pinet, The Rights of
political contest. Purists might prefer that Patients in Europe: a comparative study. Kluwer 1993.
rights be managed either by a rights-attentive 6. J. Cohen, ‘Minimalism about Human Rights: the
bureaucracy (e.g. human rights commissions most that we can hope for?’, The Journal of Political
Philosophy, 12/2, 2004, 190–213.
inspired by Friedrich’s moral mission) or
7. Consider E. Giglio, Rights, Liberties and Public Policy.
by legislatures with limited legal powers of Avebury 1995.
rights instrusion (e.g. responsible legislatures 8. L. Peattie and M. Rein, Women’s Claims. Oxford
restrained by Finer’s constitutional norms). University Press 1983.
The world of modern government has devel- 9. C. J. Friedrich and E. S. Mason, eds, Public Policy:
Yearbook of the Graduate School of Public Administration,
oped quite practical policy processes that fall
Harvard University, 1940. Harvard University Press 1940.
short of the purity of either of these theoretical 10. C. J. Friedrich and E. S. Mason, ‘Introduction’, Public
perspectives. More importantly, the discipline Policy: Yearbook of the Graduate School of Public
of public policy has escaped the captivity of Administration, xi–xii.
Peter-3383-Chapter-09.qxd 6/8/2006 7:59 PM Page 185
11. C. J. Friedrich, ‘Public Policy and the Nature of Thorton, ibid, 529–36. Consider also P. W. Hogg and A. A.
Administrative Responsibility’, Public Policy: Yearbook of Thornton, ‘The Charter Dialogue between Courts and
the Graduate School of Public Administration 3–24. Legislatures’, and F. L. Morton, ‘Dialogue or Monologue?,
12. See e.g. the contributions of Friedrich and Finer Policy Options, April 1999, 19–26; Compare J. Hiebert, ‘A
in P. Woll, ed, Public Administration and Policy. Harper relational approach to constitutional interpretation: shared
Torchbooks 1966, 221–275. legislative and judicial responsibilities’, Journal of Canadian
13. Friedrich, ‘Public Policy and the Nature of Studies, 35/4, Winter 2001, 161–72. See also J. Rohr,
Administrative Responsibility’, 19–24. ‘Administrative Law and Normative Dialogue’ in Founding
14. F. L. Morton and R. Knopff, The Charter Revolution Republics in France and America, 207–251.
and the Court Party. Broadview Press 2000, 23–24. See also 28. Consider ‘A Dialogue Model’ in Report of the ACT Bill
J. Rohr, Founding Republics in France and America. of Rights Consultative Committee, May 2003, 61–3; and
University of Kansas Press 1995; and more generally G. Williams, ‘Constructing a Community-Based Bill of
C. Epp, The Rights Revolution. University of Chicago Press Rights’, in T. Campbell, J. Goldsworthy and A. Stone, eds, Pro-
1998; and S. Walker, The Rights Revolution: rights and com- tecting Human Rights, Oxford University Press 2003, 247–62.
munity in modern America. Oxford University Press 1998. 29. The wider debate is documented in F. L. Morton, ed,
15. Peattie and Rein, Women’s Claims, particularly Law, Politics and the Judicial Process in Canada. University
16–36. For more recent accounts, see L. Chappell, of Calgary Press 2002, Chapter 13, 571–625. cf Waldron,
‘Feminist Engagements with Federal Institutions’, in The Dignity of Legislation, Cambridge University Press
D. Laycock, ed, Representation and Democratic Theory. 1999, 63–91.
UBC Press 2004, 65–89. 30. J. Parkinson, ‘Why Deliberate?’, Public Administration,
16. B. Crick, Democracy: A very short introduction. 82/2, 2004, 377–395. See more generally H. S. Richardson,
Oxford University Press 2002, 69–71. See also B. Crick, In Democratic Autonomy: public reasoning about the ends of
Defence of Politics, 2nd edition, Penguin 1982, 151–160. policy. Oxford University Press 2002, 179–241.
17. Peattie and Rein, Women’s Claims 19. 31. See e.g. Manfredi, Judicial Power and the Charter.
18. Peattie and Rein, Women’s Claims 88. 32. See J. P. Hiebert, Charter Conflicts: what is
19. Peattie and Rein, Women’s Claims 25, 109, 111, 116, 134. Parliament’s role? McGill-Queen’s University Press 2002.
20. Quoting the text as in R. Knopff and F. L. Morton, See more generally Jeffrey Goldsworthy, The Sovereignty of
Charter Politics. Nelson Canada 1992, 385–400. Parliament: history and philosophy. Oxford University Press
21. Morton and Knopff, The Charter Revolution, 25. 1999; Waldron, The Dignity of Legislation; and Rohr,
Also relevant is M. Glendon, Rights Talk: the improverish- Founding Republics in France and America, 93–137.
ment of political discourse. Free Press 1991. 33. Participation Support for a More Equitable Society
22. Morton and Knopff, The Charter Revolution 21. (The McClure Report), Final Report July 2000, 4. See also
Compare P. Hogg and A. Bushell, ‘The Charter Dialogue A. Yeatman, ‘Mutual Obligations’, in S. Shaver and
between Courts and legislatures’, Policy Options, 20/3, P. Saunders, eds, Social Policy for the 21st Century. Sydney:
1999, 19–23. Social Policy Research Centre, 1999.
23. Morton and Knopff, The Charter Revolution 17, 27, 51. 34. McClure Report, 40. Compare M. Dean,‘Administering
But note M. Mandel, The Charter of Rights and the Legalization Asceticism’, in M. Dean and B. Hindess eds, Governing
of Politics in Canada. Thompson Educational, 1994. Australia. Cambridge University Press 1998, 87–107.
24. Morton and Knopff, The Charter Revolution 68 35. McClure Report, 34–5.
25. Morton and Knopff, The Charter Revolution 89–90 36. McClure Report, 32, 40. See also J. Pixley, ‘Welfare,
26. Morton and Knopff, The Charter Revolution 92. See Poverty and Social Inequality’, in P. Boreham, G. Stokes
also D. Schneiderman and K. Sutherland, eds, Charting the and R. Hall, eds, The Politics of Australian Society. Longman
Consequences. University of Toronto Press 1997. 2000, 286–301.
27. See e.g. C. P. Manfredi, Judicial Power and the Charter. 37. E. C. Page, ‘The Civil Servant as Legislator’, Public
University of Oklahoma Press 1993; C. P. Manfredi and J. B. Administration, 81/4, 2003, 651–79; and see more generally
Kelly, ‘Six Degrees of Dialogue’, Osgoode Hall Law Journal, G. Brennan and A. Hamlin, Democratic Devices and Desires.
37/3, Fall 1999, 513–27; and ‘Reply’, by P. W. Hogg and A. A. Cambridge University Press 2000.
Peter-3383-Chapter-09.qxd 6/8/2006 7:59 PM Page 186
Peter-3383-Chapter-10.qxd 6/8/2006 7:35 PM Page 187
10
Social Policy: Pensions
HELEN FAWCETT
THE GROWTH OF THE WELFARE STATE: the development of pension policy also confirms
STRUCTURAL FUNCTIONALIST VS. the importance of social democracy, and does
POLITICAL EXPLANATIONS OF WELFARE not support the structural-functionalist hypoth-
STATE DEVELOPMENT esis. Research conducted by Myles questioned
the impact of demography on pension develop-
ment (Myles, 1984). He argued that the effect of
State pension policy has formed one of the key
increases in the elderly population is the oppo-
components of the welfare state. Political scien-
site of that predicated by structural functionalist
tists have been concerned to explain its role in
theory. Rather than encouraging the improve-
the development of the welfare state. In the
ment of public pension policy, a large elderly
nineteenth century retirement did not exist – the
population appears to discourage the develop-
old continued to constitute a part of the work-
ment of high quality pensions. In the absence of
force, and numerous studies found that
a political system dominated by the left, it would
the elderly were likely to be in poverty. As a
seem that the characteristic response to a large
result, structural functionalists have argued that
elderly population is to narrow the entitlements
pension provision was a product of the
available to the elderly. Using the various indica-
processes of modernisation, industrialisation
tors of social democratic influence (the size of
and demographic change (Wilensky, 1975). As
the vote for the left party, the number of years
family structure evolved with changing patterns
the left party is in power, union membership,
of work, supporting and caring for the elderly
and union centralisation), Myles (1984) found
was no longer a private concern. While the
the most convincing correlation between these
advantage of structural functionalist explana-
variables and his index of pension quality. The
tions of welfare state development is their focus
results suggest that a well-organised and cohe-
on the important pre-condition of state inter-
sive labour movement that is able to achieve
vention, such as a nation’s economic level of
Government office for extended periods of time
development, political explanations tend to
is the best guarantor of high quality pensions.
highlight the role of strategic choice between
However, as will be discussed later in this
competing alternatives. There are significant dif-
chapter, these results refer to research conducted
ferences in the timing of government interven-
specifically on the issue of welfare state expan-
tion and the manner in which they chose to
sion not retrenchment. In addition, this discus-
intervene. However, there are many important
sion has focused on the nations that reached a
political explanations of welfare state develop-
high level of economic development in the late
ment which highlight a wide range of political
nineteenth and early twentieth century.
variables, such as the rise of universal suffrage
Provision for the elderly is also an important
and how increased democratisation influenced
issue for public policy in countries with fewer
the growth of public pension provision
resources – those in the medium to low income
(Uusitalo, 1984). Other commentators have
category (World Bank, 1994) and those transi-
highlighted the role of political learning (Heclo,
tional democracies of Central and Eastern
1974). Building on explanations of welfare state
Europe. All of these countries are, of course,
development, it is important to explain the vari-
attempting to develop pension systems at a dif-
ations in the level and types of provision pro-
ferent historical point, and under very different
vided by the state. After the Second World War
circumstances, in comparison to those which
most countries amongst the advanced industrial
have developed in Western Europe since the
democracies had consolidated pension provi-
nineteenth century.
sion and, as a result, political scientists were con-
cerned to explain variations in character of
public pensions amongst these countries. Many
THE KEY GOALS OF PENSION POLICY
commentators have highlighted the importance
of left wing parties, trade unions and the role of
social democracy (Esping-Andersen, 1990). Like any other area of welfare provision, the
Much of the comparative research examining characteristics of national pension systems are
Peter-3383-Chapter-10.qxd 6/8/2006 7:35 PM Page 189
subject to considerable variations. The system Furthermore, if wage increases are out-pacing
design of public pensions is based on a variety those of prices and the standard of living in
of different factors. Despite the contemporary society as a whole is rising, a pension which
focus on the cost of public pensions, it is reflected the earnings of the previous decade will
important to remember the other key objec- appear to have depreciated. If pension increases
tives of pension provision in system design. If are linked to wage increases, pensioners will
the arrangements for pension delivery fail, the benefit from real growth in the economy. With
old may have to rely on means-tested assistance price indexing, the value of the pension is pre-
or personal saving. This section explores the served but the pensioner does not share in eco-
key factors that need to be taken into account in nomic growth. Whichever option is selected,
system design. Myles (1984) outlines six factors it is vital to ensure that indexing takes place at
which play a critical role in determining the regular intervals.
character of state pension provision.
Eligibility
Pension Level
Some societies provide retirement pensions as
The first factor that influences the character of of right to all citizens. However, most base
a state pension system is the value or level of their arrangements on the insurance principle,
the retirement pension. Pension level can be which means entitlement is based on the con-
evaluated in two ways: firstly, the value of the tributions made during working life.
pension in comparison to the pensioner’s
earnings during working life; and secondly, the
Security
value of the retirement pension in comparison
to the standard of living enjoyed by the working
When the individual retires from the labour
population.
force s/he will normally experience a reduction
in income. In order to maintain the standard
Pension Structure of living which is comparable to that enjoyed
during working life, the retirement pension
The structure of a pension scheme can range must replace 60–80% of pre-retirement earn-
from the provision of flat-rate uniform bene- ings. The concept of income security refers to
fits granted to all pensioners on the basis of the level of income which allows the pension-
citizenship, without any reference to their pre- ers to maintain their pre-retirement living
vious earnings or contributions, to the provi- standards. Irrespective of the pension struc-
sion of benefits which are strictly related to the ture, the elderly will not receive inadequate
contributions which have been paid during financial support if the value of the
working life. As soon as pension arrangements pension is low in comparison to the standard
depart from the flat-rate principle and become of living of society as a whole.
related to lifetime earnings the wage inequali-
ties which exist among the working population
Adequacy
will be extended to the retired.
Income security will not protect the poorest
Income Stability members of society from poverty in old age. If
the pension system is based on the principle of
This is one of the most important features of linking earnings to benefits the poor will not
any pension system because it refers to the have earned enough during working life to
mechanisms which ensure the pension will allow their income to be reduced on retire-
retain its value after the date of retirement. In ment. To avoid this problem the replacement
order to retain its value, the pension must be ratios must be higher for the poor than for the
increased so as to keep pace with inflation. well-paid members of society.
Peter-3383-Chapter-10.qxd 6/8/2006 7:35 PM Page 190
regarded these schemes as a way of attracting contrast the UK public pension system was far
and retaining labour. However, in recent years, less generous, replacing a far smaller propor-
as will be discussed later, these schemes have tion of average male earnings. When the
undergone reform, and their generosity has Swedish pension scheme was reformed, the cut
declined. The funding requirements have tight- backs meant that it was still replacing a signifi-
ened and many have moved to a less generous cant percentage of average male earnings, and
defined contribution system. As a result, as a result continued to provide good quality
employees do not know what pension entitle- provision. In the UK, the base point from
ments they will have upon retirement because which reform took place was already barely
the final pension depends on contributions providing an adequate pension.
and investment. A second common form of If the design of the pension scheme is impor-
private pension saving is Personal Pensions, tant to reform efforts, so too are political institu-
sometimes known as individual accounts. tions, the configuration of societal and political
These plans are fully funded and are based on interests and interest groups formation. If
defined contributions. Again, this form of per- welfare reform efforts retrench benefits in such
sonnel saving is often aided by tax incentives. a way that certain individuals lose out, then
The size of the final pension is contingent on there will, of course, be political consequences.
stock market performance, so the individual The issue is how to implement reform measures
bears the risk. However, in recent years, some that avoid blame being attributed to the govern-
countries have made these types of individual ing party. In some countries the progress of
pensions mandatory. Sometimes government reform has been relatively easy because the gov-
has managed these mandatory schemes and in ernment of the day has not faced effective oppo-
others they have been arranged by the private sition, or been forced to deal with institutional
sector. obstacles. The best example of this was the case
of the UK. However, in continental Europe,
while reform has taken place, the governments
have faced opposition from entrenched inter-
PENSION REFORM
ests, which has slowed the progress of reform
and also conditioned the character of the
From the 1980s and throughout the 1990s, public reforms that have been attempted. In Germany,
pension systems have undergone a period of attempts to reform the pension system began
transition and reform. In many countries under Helmut Kohl. The problem with the
public pensions constitute the largest segment German pension scheme was that high contri-
of the social security budget, which has meant butions and high rates of income replacement
there have been powerful economic reasons to discouraged growth in employment. High wage
either curtail the state’s generosity or to priva- costs made employers cautious about expand-
tise elements of provision, thus reducing the ing their work force and high benefits encour-
state’s obligations. However, efforts to reform aged early retirement. The public pension
public pensions have not followed a uniform scheme was un-funded, which left government
path, nor have they had uniform conse- responsible for a greater share of the cost (Cox
quences. As Bonoli notes, the pension reform and Fawcett, 2000). However, reform efforts
trajectory is based on an interaction between were difficult in Germany because of the con-
the structure of political institutions and the tentious nature of the issue: there has been
design of the pension schemes operating in disagreement between the governing parties,
each country (Bonoli, 2000). As a result, the opposition from the social partners, and institu-
impact of retrenchment is subject to marked tional obstacles in the form of the bi-cameral,
variations. In the 1990s, the Swedish welfare federal system of government. In addition, there
state offered extensive coverage of the work- was great public loyalty to the “conservative-
force and very generous pensions replacing a corporatist” welfare state, in which citizens had
high percentage of average male earnings. By become accustomed to a state which provided
Peter-3383-Chapter-10.qxd 6/8/2006 7:35 PM Page 192
income security. This was a marked contrast to it is in payment can also be altered. Some
a liberal welfare state such as the UK, in which pensions systems have moved from indexation
there are strong cross-class divisions in welfare. that was based on an annual re-valuation, based
In the 1970s and 1980s, around half the popula- on increases in average earnings, to one which
tion were covered by private occupational pen- increased the pension in line with prices. As a
sions and were used to market welfare. The state result, the value of the basic state pension fell
pension had never reached the generous levels progressively over the years (Fawcett, 1995).
of other continental European nations, and A further method is to encourage the
therefore was easier to reform. growth of private sector pensions, by increas-
The German pension reform of 2001 was ing tax incentives or offering other forms
interesting because it might be termed a of subsidy. Governments attempt to shape
“direction-setting or pioneering law” which individual preferences by reducing the value
changed the course of the German pension of the state pension while simultaneously
system set in place in 1957, appearing to intro- encouraging private sector growth. Increasing
duce minor changes in the first instance private sector coverage overall expands the size
(Lamping and Rub: 2004). The law introduced of funded pension schemes. In the case of the
a public private mix in pension provision, with UK, many of these measures were successfully
a new private tier of provision, subsidised by implemented during the 1980s in combina-
the state and based on defined contributions. tion. The basic state pension was reduced in
However, the public PAYG pillar remained value by altering the indexation criteria. The
the dominant component of provision. The State Earnings Related Pension was reduced in
law also introduced a new formula aimed at generosity and the government encouraged the
slowing down or even stopping increases in growth of Personal Pensions with generous
the contribution rate.1 Contributing to private incentives.
pensions is voluntary. The plan had been However, the UK experience also draws our
to make it compulsory in 1999, but now attention to another aspect of how the state is
employees are encouraged to contribute via tax involved in pension provision: the regulation
incentives. of the private sector. During the 1980s, the
There are a variety of reforms measures government took steps to de-regulate occupa-
available when attempting to contract expen- tion pensions, allowing contracting out of the
diture on public pensions, and generally they state scheme if they offered pensions based
are associated with attempts to reduce the on defined contributions rather then defined
scope and generosity of the state scheme. First benefit; as a result the number of occupation
of all, governments can increase the age of schemes increased (Fawcett, 1995). In this way,
retirement. In addition they can reduce the the government re-orientated the UK system,
quality of the pension by increasing the eligi- encouraging the growth of private sector
bility and qualifying criteria. For example, in pensions in order to reduce the size and the
the past pension levels were often based on cost of public sector pensions.
something like the “best twenty years” of earn-
ings. This can be extended to the whole of
individuals’ working life. Equally, the number
REFORMING PENSION SYSTEMS:
of qualifying contribution years may be
ISSUES OF PATH DEPENDENCY
extended. Finally, the proportion of earnings
replaced by each year’s pension contribution
can be made less generous (Fawcett, 1995). From the 1980s, many governments have sought
Another method of reducing costs is to target to pursue a welfare reform agenda. However,
benefits. This has generally been applied in despite a number of governments committing
countries that had a basic flat-rate state themselves to change, many commentators
pension, such as Australia and New Zealand. observed that the welfare state seemed somewhat
In addition, the indexation of the pension once impervious to reform attempts, in particular that
Peter-3383-Chapter-10.qxd 6/8/2006 7:35 PM Page 193
it appeared difficult to reduce the overall levels way in which previous political choices effect
of spending on state welfare. This gave rise to a the current political agenda and the options
number of theories attempting to explain wel- open to political parties. Firstly, it is argued that
fare state retrenchment, one in particular public policies give rise to structures and
focuses on the notion of path dependency. resources that either encourage or inhibit the
This appeared highly relevant to the study of formation of interest groups. In this sense, as
pension reform because, of all the component Schattschneider wrote, “new policies create a
parts of the welfare state, pensions seem highly new politics” (1935: 288). Policies themselves
susceptible to path dependent effects. If we create a specific constellation of interests and
consider the basic characteristics of a pension interest groups, and generate incentives which
scheme we can immediately observe certain either facilitate or inhibit their expansion.
difficulties in reform measures. The contribu- Secondly, existing policies influence the percep-
tory nature of pension schemes means that tions and agendas of key political actors. Past
pension rights are accrued over a long period policy plays an important role in framing polit-
of time. In addition, there are significant prob- ical actors’ perceptions of policy alternatives.
lems in the transition from one pension When we examine the various attempts to
scheme to another, when pensions are already reform pension policy since the 1980s we can
in payment under one scheme and a high see that these concepts have some purchase. For
proportion of workers will not have time to example, many continental European countries
contribute to a new scheme. The public pen- lacked a well-developed private sector which
sion system of most advanced industrial coun- was ready to take over the burden of public
tries means that governments have made long pension provision. Indeed, in some countries,
term pension commitments to large sections such as Germany and Sweden, the majority of
of the electorate. In addition, those with a wage earners relied on state provision for all or
stake in existing programmes and policies most of their retirement income. As a result,
have a high level of influence over the reform there was a substantial cross-class coalition in
measures that can be attempted. Since aboli- support of public provision which was difficult
tion or cuts in provision have significant polit- to de-stabilise.
ical risks and may reflect negatively on the By the same token, the example of the UK
government of the day, is it possible to avoid shows how the policy choices of political actors
blame? shaped the behaviour of ordinary citizens.
The most common technique that has been 11 million workers joined private sector occu-
used in pension reform is introducing policies pational schemes in the 1960s because the UK
which defer the impact of reform until some lacked a good earnings-related second tier of
future date. Another technique that has been state pension provision. Future attempts to
used is to contract or re-shape a pension improve public provision in the late 1960s and
system by making a variety of small scale 1970s were constrained by these choices. The
reform measures that erode the value of bene- growth of the private sector created new inter-
fits over time. Pension policy is highly complex ests: insurance companies, white collar trade
and it has been shown that consumers have unions, and of course, members of the public
imperfect knowledge of the options available who had joined private sector schemes.
(Mayhew, 2002). As a result, governments can The design of a state earnings-related scheme
enact highly technical changes that may have a in 1975 was based on the notion of partnership
very big impact on the long-term value of rather than competition with the private
these commitments. Often the implications of sector. Equally, any attempt to regulate the
these changes do not become apparent for quality of pensions offered by the private
some time (Fawcett, 1995). sector in order to improve quality was likely to
The path dependency argument, associated provoke opposition from pension providers
with the new institutionalist school of political and resistance from those who had become
science, refers to the legacies of past policy: the clients of the private sector. Whereas it had
Peter-3383-Chapter-10.qxd 6/8/2006 7:35 PM Page 194
been possible to think of launching an sector and created a new market in private
ambitious state scheme in the 1950s and early sector defined contribution Personal Pensions.
1960s, when private pensions accounted for a The Personal Pension market was stimulated
relatively small sector of the market, and the by generous tax incentives. In the 1970s, about
existing state pension was set at a basic level, a half of the UK work force was covered by
there were now severe political costs to be private pensions. By the end of the 1980s,
taken into consideration. (Fawcett, 1995, 1996) around two-thirds of the workforce was covered
As Pierson writes: by the private sector. The state sector was in
long-term decline, and as its benefits became
“Policies may create incentives that encourage the emer-
gence of elaborate social and economic networks,
less competitive and the numbers covered less
greatly increasing the cost of adopting once possible substantial, this paved the way for further
alternative and inhibiting exit from a current policy retrenchment, eventually leading to the greater
path. Individuals make important commitments in reform efforts (Fawcett, 1995).
response to certain types of government action. These The fundamental issue here is that a quanti-
commitments in turn may vastly increase the disruption
caused by new politics, effectively locking in previous
tative change can produce a qualitative change.
decisions.” (Pierson, 1994, 41–3) The Conservatives wished to retrench public
provision and encourage private provision. The
However, the importance of path dependency strategy they pursued had a number of compo-
and the feedback effects of previous policy nent parts which produced the same result in a
choices should not be exaggerated. Some of the very short time period. As a result, the state
literature in this school has emphasised these pension system in the UK underwent retrench-
factors as a reason why it was difficult to ment; however, it did so by a rather different
retrench the welfare state during the 1980s. route than all-out abolition at one particular
The fundamental problem with this literature time. Hence, the way in which we understand
is the specification of the dependent variable the specification of retrenchment and how we
“welfare state retrenchment”. What do we chose to measure it, is the issue here.
understand by change and how best to In fact, looking back over the history of the
measure the degree of change that has taken 1980s and 1990s, many governments pursued
place? For example, in the mid-1980s, the UK similar courses of action, which we might
Conservative government under Margaret describe as “path-shaping” (Cox and Fawcett,
Thatcher considered abolishing the State 2000). The issue of “blame avoidance” is
Earnings Related Pension Scheme. The fact important in devising retrenchment policies,
that the government chose not to pursue all- because restructuring can lead to important
out abolition has been cited as an example that sectors of the population suffering the negative
shows how difficult it was to retrench the welfare consequences of retrenchment, which may cre-
state and as proof of the path dependency argu- ate electoral losses for politicians. There is
ment (Pierson, 1996). In fact, the State Earnings some benefit in pursuing a retrenchment strat-
Related Pension scheme was not defended by egy that proceeds in a gradual and incremental
any significant interests. The then Chancellor manner because it assists in a blame avoidance
of the Exchequer, Nigel Lawson, did not see strategy by defusing opposition from groups
immediate benefit in terms of the administra- who feel they will lose out because of reform
tion’s priorities and opposed the plan. efforts.
However, the strategy that was chosen was However, the ability to reform public sector
equally effective but far more politically astute. pensions varies from country to country. These
The government pursued a dual strategy of differences can be explained by the institu-
reducing the generosity of the second tier of tional and political veto points which either
earnings-related state pensions, while chang- facilitate or inhibit the progress and the possi-
ing the way in which the basic pension indexed bility of reform measures. Bonoli points out
so it rose in line with prices. At the same time, that the political limits to reform are specific to
the government de-regulated the occupational individual countries, based both on the design
Peter-3383-Chapter-10.qxd 6/8/2006 7:35 PM Page 195
of the pension system and the political and pension provision. However, in some countries
institutional structures in place (Bonoli, 2000). the pension system was designed from the out-
His argument is based on new institutionalist set to operate on the basis of a public-private
focus on the capabilities of government and mix of provision. From 1946, the Dutch pen-
the fact that the degree of government control sion scheme combined public sector pensions
over policy-making is subject to significant with private occupational pensions, and occu-
variation. In any political system there may be pational coverage expanded over time so that
institutional structures that have to be ninety percent of the population was covered.
managed in order to implement policy change. However, in recent years, we have seen a trend
Equally, there might be coalitions of actors that towards increasing the extent of private provi-
can obstruct the policy-making process. These sion as a way of reducing the state’s commit-
“veto points” can act as a major limitation on a ments to public pensions. This has been the
government’s ability to control policy out- case in the German pension reform of 2001, the
comes. By contrast, the absence of veto points, American debate on social security reform, and
combined with strong concentrations of the Swedish case.
power, gives government a much stronger abil-
ity to bring about change in policy-making.
The types of factors that impact on govern-
PRIVATE SECTOR: SOLUTION OR
ment capacities are systems based on a separa-
PROBLEM
tion of powers; bi-cameralism; proportional
representation; referendums; and balanced
dual executive. Policy-makers in the UK, a uni- Does the shift to the private sector mean less
tary state with a first-past-the-post-electoral adequate income security in retirement? Rein
system, and a weak second legislative chamber, and Schmahl (2004) argue that, to answer this
faced very few institutional obstacles when question, we have to examine the overall
passing their reform legislation – especially mix of the state and private sector income
since the governing parties often enjoyed large package combined. In many countries there
majorities. Other countries have had to negoti- seems to be no immediate cause for concern.
ate and frame reform in a far more complex In two countries in particular, the UK and the
and cautious manner. USA, there is high-income inequality and high
levels of pensioner poverty.
The first point to note in a consideration
of the private sector is that the state continues
ISSUES RELATING TO THE PRIVATISATION
to be involved in a variety of important ways,
OF PENSION POLICY: THE PUBLIC
especially as a public sector employer and as a
PRIVATE MIX
regulator. There are three principle sources
of regulation: trust law, taxation policy, and
An important part of the welfare reform efforts regulatory authorities. In addition, if the
that have taken place over the last two decades mechanisms for privatisation are based on
has been the attempt to introduce increase pri- contracting out of the public sector scheme,
vate pensions coverage. However, some coun- the state regulates the conditions of pension
tries have always based their retirement quality under which occupational schemes are
provision on a mix of public and private provi- allowed to contract out. Thus it is still the
sion. In many cases, funded private pensions state’s role to stipulate the quality of pension
have been small in scale aimed at professional the private sector provides. The increase in
well-paid wage earners. In other countries, pri- private provision may have reduced some of
vate provision was extensive and well-institu- the obligations on the state. However, it raises
tionalised after the Second World War. For most the issue of whether private pension providers
of the post-war period, private occupational will be able to deliver a high quality of provi-
pension schemes existed as supplements to state sion for future pensioners. A highly privatised
Peter-3383-Chapter-10.qxd 6/8/2006 7:35 PM Page 196
configuration of pension arrangements forces heavily reliant on the private sector, and both
politicians to consider the issue of regulation, the Labour and Conservative parties have
and the dilemmas posed in trying to balance encouraged further increases in private provi-
risk and regulation. The UK now has one of sion. Hence the “regulation/risk” dilemma is
the least expensive public pension schemes in acute in the British context. The greater the
Europe. However, this has come at the price of population covered by the private sector, the
de-regulating the private sector, and then more important it is to deliver high quality and
being forced to re-regulate to restore public reliable benefit systems. However, this reliance
confidence in the aftermath of a variety of on the private sector creates a powerful interest
scandals (Blake, 1995). These scandals group. The development of regulation is a matter
increased public awareness of the degree of for negotiation rather than imposition. In
risk associated with market provision and a practical terms, the UK has established a rela-
re-evaluation of the regulatory regime. Indeed, tionship of dependency on the private sector,
since de-regulation, pressure has mounted on and, as a result, regulation is viewed in terms of
this issue, because it was becoming increasingly what is perceived as feasible for the private
clear that (for a variety of reasons) funded sector to provide. The UK case is an extreme
pensions might not deliver the appropriate example, perhaps because of a mistaken belief
level of benefits at the point of retirement. In that expanding private pensions is the solution
addition, since 1998, two-thirds of final salary to the problem of financing public pensions.
schemes closed to new members and fund In fact, it may be a cautionary tale, demon-
deficits are estimated to be £100 billion since strating what can happen when the state
the stock market slump of 2000. The govern- becomes over-reliant on the private sector.
ment seems caught between the proverbial
“rock and a hard place” because increasing reg-
ulation is likely to provoke a backlash from the
CONCLUSION
private sector – failing to act is likely to further
damage public confidence. However, while it
has been important to develop a structure for During the affluence of the post-war years,
regulation, as has been done since the 1990s, it pension provision was consolidated and
is equally important to regulate the quality of expanded, but by the 1970s pension policy was
pension provision for future pensioners. In perceived to be a key part of the “crisis” of the
this sense, the current crisis in British pensions welfare state. In the aftermath of the oil crisis of
has been the result of the liberalisation of the 1973, stagflation and the pressure to retrench
quality conditions under which the private public expenditure, spending on pension policy
sector has been allowed to operate. By moving became of enormous concern in the political
to defined contribution rather than defined and policy debate. At one time, relatively uncon-
benefit schemes, Blake has shown that the troversial, fears of a demographic crisis –
overall level of contributions has fallen (Blake, namely a growing number of older citizens
2001). Indeed, increased regulation has come living longer – was placing the pension systems
at the price of a decline in pension quality. In under pressure. This was combined with
order to get the private sector to contribute to an unfavourable ratio of workers to retired –
a Pensions Protection scheme, the government namely, the size of the working population
allowed a reduction in inflation proofing, contributing to the finance of the pension
which could have serious implications for systems – was declining in comparison to the
future pensioners. rising number of the elderly.
The provision of private pensions is a Since the 1970s, therefore, the succeeding
voluntary activity. If the burden of regulation decades of welfare retrenchment and reform
becomes too onerous, the incentive to provide have focused heavily on the issue of pension
private provision becomes less attractive. policy. Policy makers have looked for ways to
However, the UK’s pension arrangements are make their pension systems more affordable.
Peter-3383-Chapter-10.qxd 6/8/2006 7:35 PM Page 197
This has usually focused on two aspects of The Pensions Institute, Birkbeck College, University
system design: first of all reforming public of London (Revised 2002)
pensions to make them less generous and Bonoli, G. (2000) The Politics of Pension Reform:
to reduce the obligations on the state; secondly, Institutions and Policy Change in Western Europe.
Cambridge: Cambridge University Press.
encouraging the privatisation of provision.
Borsch-Supan, A. and Miegel, M. (eds.) (2001)
Reconstituting pension systems in a way in
Pension Reform in Six Countries: What we can
which the private sector became responsible learn from each other. London: Springer.
for some proportion of pension provision Clark, G. L. (2000) Pension Fund Capitalism.
would seemingly diminish the obligations on Oxford: Oxford University Press.
the state. However, in the current economic Clayton, R. and Ponthusson, J. (1998) ‘Welfare State
climate, the way in which this has been done Retrenchment Re-Visited: Entitlement, Cuts,
can produce lower quality provision. As a Public Sector Restructuring, and Inegalitarian
result, a new problem may be on the horizon, Trends in Advanced Capitalist Societies’, World
namely a retirement income package that fails Politics, 51(October), pp. 67–98.
to produce financial security in old age. In Cooper, D. R. (2000) A Tale of Two Pension Systems,
Employer Relations, 22(3): 286–292.
addition, the private public mix in pension
Cox, R. H. (1998) ‘From Safety Net to Trampoline:
provision raised other important issues for
Labour Market Activation in the Netherlands and
public policy, namely the regulation of the Denmark’, Governance, 11(4), pp. 397–414.
private sector, as well as their ability to deliver —— (1997) ‘The Consequences of Welfare
and guarantee high quality pensions. Retrenchment in Denmark’, Politics and Society,
25(3), pp. 302–325.
Cox, R. H. and Fawcett, H. (2000) Same Idea,
NOTE Different Result: Welfare Reform and the Third
Way in Britain and Germany, paper presented to
the research Committee 27 of the International
1. According to Lamping and Rub the contribution rate
Political Science Association, Bangalore, India.
is set so that it should not exceed 20 percent in 2020 and
Davis, B. (1993) Better Pensions for All. London:
subsequently should remain below 22 percent. The result
of the new alteration in the pension formula is that pen- IPPR.
sion level should average 67 percent of average net earn- Davis, E.P. (2000) Regulation of Private Pensions:
ings minus the assumed saving rate for private old age A case study of the UK, Discussion Paper PI-0009.
provision. The Pensions Institute, Birkbeck College, University
of London.
Deakin, N. (1994) The Politics of Welfare: Continuity
REFERENCES and Change. Hemel Hempstead: Harverster
Wheatsheaf.
Department of Social Security (1998) New Ambitions
Anderson, K. M. and Meyer, T. (2003) ‘Social for our Country: A New Contract for Welfare. Cmnd
Democracy, Unions and Pension Politics in 3805.
Germany and Sweden’, Journal of Public Policy, Department of Social Security (1998) A New
vol. 23, no. 1, pp. 23–54. Contract for Welfare: Partnership in Pensions, Cm
Atkinson, A. B. (1991) The development of state pen- 4179.
sions in the UK (Discussion Paper WSP/58: LSE). Department of Social Security and HM Treasury
Baldwin, P. (1990) The Politics of Social Solidarity: (2000) Security for Occupational Pension Schemes:
Class Bases of the European Welfare State 1875–1975. A Consultation Document, London, September.
Cambridge: Cambridge University Press. Department of Social Security and HM Treasury
Blake, D. (1995) Pension Schemes and Pension Funds (2001) Security for Occupational Pensions: The
in the United Kingdom. Oxford: Clarendon. Government’s Proposal, London, March.
Blake, D. (2000) Two Decades of Pension Reform Department of Work and Pensions (2002)
in the UK: What are the implications for occupa- Simplicity, Security and Choice: Working and
tional pension schemes? Employee Relations, Saving for Retirement, Cm 5677.
22(3): 223–245. Department of Work and Pensions (2003) Working
Blake, D. (2001) The United Kingdom Pension and Saving for Retirement: Action on Occupa-
System: Key Features, Discussion Paper PI-0107, tional Pensions, London, June.
Peter-3383-Chapter-10.qxd 6/8/2006 7:35 PM Page 198
DHSS (1969) National Superannuation and Social Johnson, P. (1994) The Pensions Dilemma. London:
Insurance: Proposals for Earnings Related Social IPPR.
Security, Cmnd. 3883. Labour Party (1957) National Superannuation:
DHSS (1974) Better Pensions: Proposals for a New Labour’s Policy for Security in Old Age.
Pension Scheme Fully Protected Against Inflation, Lamping, W. and Rub, F. W. (2004) From the
Cmnd. 5713. Conservative welfare state to an ‘uncertain some-
Emmerson, C. and Tanner, S. (2000) A Note on the thing else’: German pension politics in compara-
Tax Treatment of Private Pensions and Individual tive perspective, Policy and Politics, vol. 32, no. 2,
Savings Accounts, Fiscal Studies, 21(1): 65–74. pp. 169–91.
Esping-Andersen, G. (1990) The Three Worlds of Levy, J. (1999) ‘Vice into Virtue? Progressive Politics
Welfare Capitalism. London: Polity. and Welfare Reform in Continental Europe’,
Fawcett, H. (1995) The Privatisation of Welfare: The Politics and Society, 27(2), pp. 239–273.
Impact of Parties on the Private/Public Mix Lynes, T. (1969) Labour’s Pension Plan (Fabian
in Pension Provision. West European Politics: Tract 396). London: Fabian Society.
18(4), 150–69. Lynes, T. (1997) The British Case, in Rein, M. and
Fawcett, H. (1996) The Beveridge-Straight Jacket: Wadensjo, E. (eds), Enterprise and the Welfare
Policy Formation and the Problem of Poverty State. Cheltenham: Edward Elgar, 309–351.
in Old Age. Contemporary British History, Mayhew, V. (2002) Public Attitudes to Pensions,
10(Spring): 20–42. Department of Work and Pensions.
Field, F. (2000) The State of Dependency: Welfare Minns, R. (2000) The Cold War in Welfare. London:
under Labour. London, Social Market Foundation. Verso.
Financial Times (2002) “Unscrambling the Pensions Myles, J. (1984) Old Age in the Welfare State: The
Mess”, March 4th, p. 20. Political Economy of Public Pensions. Kansas:
Financial Times (2002) “FRS 17: Little Name, enor- University of Kansas Press.
mous impact” (Michael Peel and Patrick Jenkins), Myles, J. (1996) When Markets Fail: Social Welfare in
March 3rd. Canada and the United States. In Welfare States in
FT.com, 12 June 2003. Transition: National Adaptations in Global
Francis, J. (1993) The Politics of Regulation: a com- Economies, G. Esping-Andersen (ed.). London: Sage.
parative perspective. Oxford: Blackwell. Myles J. and Quadagno, J. (eds.) (1991) States,
Guardian, 2003, Safety Net Plan for Pension Labour Markets, and the Future of Old-Age Policy.
Schemes, 12th June. Philadelphia: Temple University Press.
Hannah, L. (1986) Inventing Retirement: the devel- Myners, P. (2001) Institutional Investment in the
opment of occupational pensions in Britain. United Kingdom: A Review. HM Treasury,
Cambridge: CUP. London, March.
Heclo, H. (1974) Modern Social Polices in Britain Pickering, A. (2002) A Simpler Way to Better
and Sweden. New Haven: Yale University Press. Pensions. London: HMSO.
Hills, J. (1998) Thatcherism, New Labour and the Pierson, P. (1994) Dismantling the Welfare State?
Welfare State, Centre for the Analysis of Social Reagan, Thatcher and the Politics of Retrenchment.
Exclusion, CASE 13, London, The London School Cambridge: Cambridge University Press.
of Economics. Pierson, P. (1996) The New Politics of the Welfare
HMSO (1993) Pension Law Reform, Report of State, World Politics, 48: 143–79.
the Pension Law Review Committee (Goode Pierson, P. (ed.) (2001) The New Politics of the
Committee) Cm 2342, London: HMSO. Welfare State. Oxford: OUP.
Hood, C., James, O., and Scott, C. (2000) Regulation Rake, K., Falkingham J., and Evans, M. (1999)
of Government: Has it Increased, Is it Increasing, Tightropes and Tripwires: New Labour’s
Should it be Diminished?, Public Administration, Proposals and Means-testing in Old Age, Centre
78(2): 283–304. for the Analysis of Social Exclusion, CASE paper
Hood, C., Rothstein, H., and Baldwin, R. (2001) The 23, London: The London School of Economics.
Government of Risk: Understanding Risk Regulation Rein, M. and Wadensjo, E. (eds.) (1997) Enterprise
Regimes. Oxford: Oxford University Press. and the Welfare State. Cheltenham: Edward Elgar.
House of Lords (1995) Hansard, February 7th, Rein, M. and Schmahl, W. (ed.) (2004) Rethinking
col. 109. the Welfare State: the Political Economy of
Hughes, G. and Stewart, J. (1999)The Role of the Pension Reform: Cheltenham: Edward Elgar.
State in Pension Provision: Employer, Regulator, Rose, R. and Davies, P. L. (1994) Inheritance in
Provider. London: Kluwer Academic Press. Public Policy. New Haven: Yale University Press.
Peter-3383-Chapter-10.qxd 6/8/2006 7:35 PM Page 199
Ross, F. (2000) Beyond Left and Right: The New Wilensky, H. L. (1975) The Welfare State and Equality:
Partisan Politics of Welfare, Governance, Structural and Ideological Roots of Public
pp. 155–183. Expenditures, Berkeley, University of California
Schattschneider, E. E. (1935) Politics Pressures and Press.
the Tariff. New York: Prentice Hall. Wilsford, D. (1994) Path Dependency, or Why
Uusitalo, H. (1984) ‘Comparative research and the History Makes It Difficult but Not Impossible to
determinants of the Welfare State: the state of the Reform Health Care Systems in a Big Way, Journal
art’, European Journal of Political Research, 12, of Public Policy, 14(3): 251–283.
pp. 412–13. World Bank (1994) Averting the Old Age Crisis:
Walker, A. (1999) The Third Way in Pensions (by Policies to Protect the Old and to Promote
way of Thatcherism and avoiding today’s Growth. Oxford: OUP.
pensioners), Critical Social Policy, 9(4): 511–527.
Waine, B. (1999) The Future is Private, New Internet Sources
Economy, 7–10. http://opra.gov.uk
White, S. (ed.) (2001) New Labour. Basingstoke: http://dss.gov.uk
Palgrave. http://dwp.gov.uk
Wilding, P. (1992) The British Welfare State:
Thatcherism’s Enduring Legacy, Policy and
Politics, 20(3): 201–212.
Peter-3383-Chapter-10.qxd 6/8/2006 7:35 PM Page 200
Peter-3383-Chapter-11.qxd 6/8/2006 7:35 PM Page 201
11
Social Policy: Is There A Crisis of
the Welfare State?*
HAROLD L. WILENSKY
The essence of the welfare state is government- illness, shifts in family composition, or other
protected minimum standards of income, random shocks (wars, depression, recessions).
nutrition, health and safety, education, and Because the welfare state is about shared risks
housing assured to every citizen as a social cross-cutting generations, localities, classes,
right, not as charity (Wilensky, 1965, p. xii). In ethnic and racial groups, and educational
the abstract this is an ideal embraced by both levels it is a major source of social integration
political leaders and the mass of people in in modern society. Because it lends a measure
every affluent country, but in practice, and at of stability to household income, it has also
high levels of development, it becomes expen- been an important stabilizer of the economy in
sive enough and evokes enough ambivalence the downswings of the business cycle, especially
to become the center of political combat since World War II.
about taxes, spending, and the proper role The welfare state is at once one of the great
of government in the economy. In public structural uniformities of modern society
expenditures, the welfare state is about two- and, paradoxically, one of its most striking
thirds to three quarters of what modern gov- diversities. In the past century the worlds 22
ernments do. The core programs of the welfare richest countries (our 19 rich democracies and
state, often subsumed under the general head- three countries that became communist,
ing of “social security,” have taken the form of Czechoslovakia, East Germany, and Russia),
social insurance against the basic risks of mod- although they vary greatly in civil liberties and
ern life: job injury, sickness, unemployment, civil rights, have varied little in their general
disability, old age, and income lost due to strategy for constructing a floor below which
*Most of this chapter is excerpted from Harold L. Wilensky, Rich Democracies: Political Economy, Public Policy, and
Performance (Berkeley: University of California Press, 2002), pp. 211, 221–232. It draws on my remarks at an OECD
Conference of Finance Ministers and Social Affairs Ministers (Wilensky, 1981a) and subsequent research. Chapter and table
cross-references are to data and evidence in Wilensky, 2002, which also contains a 90-page bibliography. Among those who
share my view of the alleged welfare-state “crisis” and provide empirical evidence casting doubt on the idea are Alber
(1988), Klein and O’Higgens (1985), Marmor and Mashaw (1988), and Lindert (2004, two vols.). Gene Park assisted in
updating the cases.
Peter-3383-Chapter-11.qxd 6/8/2006 7:35 PM Page 202
no one sinks. The richer these countries government spending or confine analysis to
became, the more likely they were to broaden social spending. A thorough study of gross
the coverage of both population and risks. At debt-to-GDP ratios from 1961 to 1990 among 12
the level of economic development they EC countries shows that the ratio of government
achieved in the past 40 years, however, we find expenditures to GDP is unrelated to the gross
sharp contrasts in spending, taxing, and the debt ratio and that rising deficits are not the
organization of the welfare state and, of result of growing expenditures (von Hagen,
course, in the politics of the welfare state.1 1992, pp. 12–13; cf. Cameron, 1982). For
instance, during 1986–90 among welfare-state
leaders the average debt ratio ranged from 128
percent of GDP for Belgium and 110 percent for
CRISIS OF THE WELFARE STATE?
Italy to 44 percent for Germany and 35 percent
for France, while among the welfare-state lag-
In 1975 I wrote that “the welfare state in its gards the ratio ranged from 110 percent for
wondrous diversity has proved hard to shoot Ireland to 50 percent for the United Kingdom
down;” it marches on through thick and thin (the U.S. ratio in 1994 was 70 percent). Similarly,
(Wilensky, 1975, p. xvii). Since then there’s the annual net government deficit for 1993–94
been recurrent talk of a “crisis of the welfare among big spenders ranged from Sweden’s 8.3
state.” Politicians, finance ministers, and even percent of GDP (down to 5.2 percent in 1996),
some scholars complain that accelerating Belgium’s 3.7 percent, Netherlands’s 1.2 percent
social spending accounts for a rising burden of and Norway’s 2.4 percent surplus, while among
debt and deficits; that public support for the welfare-state laggards the annual deficits ranged
welfare state has eroded, another reason for from Japan’s 6.0 percent and the United States’s
cutting it down; and that social spending is a 4.1 percent to Ireland’s 1.6 percent (OECD
drag on economic growth, is inflationary, or National Accounts). Debt ratios depend not
contributes to unemployment. We should put upon government spending but upon what else
these complaints in historical perspective. these countries do – whether they tax enough to
Crisis talk has been a feature of welfare-state pay for the services their citizens demand, their
politics for almost a century. In Germany, economic performance, the structure of their
when Bismarck’s social legislation reached a government (e.g., the strategic dominance of the
cost of 1.4 percent of GDP in 1905, it triggered prime minister or finance or treasury minister
heated debate over its backbreaking economic over the spending ministers), and the structure
burden and threat to civic morals. In recent of the budgetary process. For example, experi-
decades crisis-mongering has escalated, aided ence with budget norms in the United States
by the rise of the broadcast media, which shows that they are ineffective in the long run
amplify extreme views. for two reasons: first, the decentralized and
If by “crisis” we mean rapidly accelerating divided structure of government means that
social spending among the 19 rich democra- Congressional spending committees and govern-
cies, by any measure social spending as a frac- ment agencies can maneuver to increase spend-
tion of GDP has evidenced slower growth since ing throughout the budgetary process by a
1975 or 1980, in some cases leveling off. principle of reciprocity; second, the states,
Moreover, the burden of social spending in whether they have spending and taxing limits or
these countries varies both in level and trend. not, lavishly issue long-term bonds, resort to
Thus, if by “crisis” we mean accelerating social creative accounting tricks, and conduct one-time
spending, there is no crisis common to all. sales of assets to meet legal requirements for a
If the welfare-state crisis means that the social balanced budget (von Hagen, 1992, pp. 38ff.).2
budget is heavy and growing, that the welfare In short, there is such great variation in the
state is the root of public deficits, and deficits are depth and duration of fiscal stress and debt,
dangerous, then again there is no general crisis. and so much evidence that the welfare state is
Results are the same whether we consider total not the culprit causing changes in debt, that it
Peter-3383-Chapter-11.qxd 6/8/2006 7:35 PM Page 203
makes no sense to talk about a general welfare- Recent literature on the impact of the
state crisis. welfare state confirms these conclusions. In a
If the meaning of the welfare-state crisis is careful review of the mixed and weak findings
that there has been a withdrawal of mass support of nine studies, A. B. Atkinson (1995) con-
for social programs, there is no evidence of it. cludes that none of them shows that the wel-
The most remarkable and solid finding of public fare state is bad for economic growth,
opinion research on taxing and spending – both employment, or productivity. Similarly, recent
over time and across countries – is the stability research by economist Peter Lindert (2004)
of issue-specific opinion about social programs provides the best overview of growth and
and the taxes to finance them. Since World War retrenchement from the Poor Laws of the 18th
II, pensions and national health insurance century to 2000; he confirms the point. Using
remain overwhelmingly popular, most family measures and econometric methods different
policies retain a majority, while public assis- from mine he also finds that since 1980 among
tance to the non-aged, non-working poor modern democracies the net impact of social
remains stably unpopular. Similarly, consump- spending on the economy (i.e. GDP/person
tion taxes and social-security payroll taxes evoke and growth) is close to zero (2004, vol. 1,
no sustained mass hostility, while property taxes ch. 10; vol. 2, chs. 18 and 19).
and income taxes arouse the most persistent
resentment. The rank order of enthusiasm
regarding both spending and taxing is similar
RETRENCHMENT OF THE
across countries and over time. Chapter 10
WELFARE STATE?
[Backlash, pp. 369–373] reviews the evidence
and explains why the relative uniformity
of public opinion about the welfare state cannot If “crisis” talk, however misleading, is univer-
explain great national differences in the sal; if anti-taxing, anti-social spending, anti-
electoral success of tax-welfare backlash move- bureaucratic themes have helped candidates to
ments and parties like those of Mogens Glistrup win some elections (see Chapter 10), has the
in Denmark, Ronald Reagan in the United action of governments matched the rhetoric of
States and Margaret Thatcher in the United campaigns? What cutbacks have actually
Kingdom. occurred in the period of austerity after 1975 or
If the “welfare state crisis” is not an inevitably 1980 when economic growth and productivity
accelerating rate of social spending, not the growth slowed down? With some exceptions, the
withdrawal of mass support for social spending, core programs of the welfare state – pensions,
and not the inevitability and dangers of public disability insurance, and national health insur-
debt, then surely it means that the burdens of ance, programs that have generally outpaced
the welfare state universally subvert good eco- GDP growth – have proved most resistant to
nomic performance. My Chapter 12 shows that real cuts in benefits per capita or even in their
the evidence is overwhelmingly to the contrary. GDP shares (on disability insurance see
Aggregate social spending up to 1973 was a pos- Chapter 16, Environment). Most vulnerable to
itive contribution to the combination of low-to- real cuts or at least spending restraint have been
moderate inflation, good real GDP growth per education, family allowances, social assis-
capita, and low unemployment; since the oil tance, and unemployment compensation. For
shocks of the 1970s it has been on average neu- instance, in the period 1975–1981, ten of the
tral. In no period and for no measure of perfor- 16 of our rich democracies studied by Jens
mance is social spending a significant drag. Alber (1988, pp. 190–191) reduced the GDP
Much depends on the mix of social and labor share of education. Only Sweden, Italy, and
market policies a nation adopts, how it finances Ireland increased as much as a point or more.
and manages the welfare state, and, more In contrast, the GDP share of health spending
important, what economic and industrial rela- increased almost everywhere (only Canada,
tions policies it pursues. Germany, Denmark and Norway showed
Peter-3383-Chapter-11.qxd 6/8/2006 7:35 PM Page 204
popular program. In contrast, at the outset of that the United Kingdom, with by far one of
his administration, Reagan proposed to cut the the cheapest, most-accessible medical-care sys-
minimum pension benefit for low-earnings tems, made the most radical reforms, while the
workers and to delay a cost-of-living adjust- United States, with by far the most expensive
ment three months; immediately 96 Senators and least accessible system, failed either to con-
voted “No,” zero “Yes.” This episode triggered a trol spending or to increase access while it suc-
big revival of an elderly lobby (the National ceeded in increasing the number of families
Committee to Preserve Social Security and with no medical insurance at all (by 2005 more
Medicare was formed in 1982). Later, in 2005, than 43 million). It is another paradox, which
President Bush II, with dominance in both runs through all policy areas (see chapter 8
Executive and Congress, spent a solid year try- [Welfare Mess], 15 [Environment], and 16
ing to sell the privatization of Social Security [Health Care]), that the most ideologically-
with no success. committed free-marketeers, Prime Minister
It is possible that Thatcher could succeed in Thatcher and Presidents Reagan and Bush I,
modest steps where Reagan and Bush II failed, created the most intrusive regulatory regimes.
not only because a more centralized parlia- In U.S. health-care reforms it was the Reagan –
mentary system with party discipline gives an Bush regulations of hospital and physician
ideologically-committed prime minister more payments – rate-setting, prospective payment
power than her counterpart in a divided plans based on diagnostic-related groups, and
Presidential system, but also because of differ- resource-based relative value scales (Brown,
ences in the degree of consolidation of the 1991; Ruggie, 1992). The payment reforms did
pension system. As Pierson (1994, pp. 69–73) modestly reduce Medicare costs through shift-
suggests, Britain’s SERPS (State Earnings- ing the burden to private insurers, who then
Related Pension Scheme), the target of raised premiums and gave employers and
Thatcher’s reform, was a young program first providers strong motive to reduce coverage
implemented by the Labour government in and services, shifting the burden of coverage
1974; America’s Social Security law had four back to the public sector – a typical cost-
additional decades to become institutionalized shifting game in the uniquely commercialized
and therefore more resistant to change. As U.S. public/private system. A second game is
always, system maturity counts. risk selection: commerical providers also save
That political consenus is esssential for suc- money by skimming off the younger, healthier
cessful reforms is shown by the enactment in patients while dumping older, sicker, costlier
the United States in 1983 of the pension-reform patients onto the public sector. The net effect
recommendations of a bipartisan commission of all these measures in two decades was to
on social security. Rather than cutting benefits increase administrative complexity and cost
they raised payroll taxes slightly, taxed benefits (by 2005 more than 15 percent of GDP),
for the first time, brought new federal employees reduce medical coverage, enrage doctors, and
into the system to broaden the payroll tax base, put the U.S. below the average on a wide range
and trimmed future costs by slowly phasing in of real health indicators.
the higher retirement age from 65 to 67 by the Because the recent British restructuring of
year 2027 – sensible steps, which “fixed” the medical care is so sweeping, a brief account is
system for 47 years, hardly a revolution. relevant as we consider “cuts” in welfare-state
Regarding cuts in health-care spending and spending. A popular, universally-accessible, tax-
benefits, Reagan, Thatcher and their successors financed, already low-priced health-care system –
(Bush, Clinton, and Major) all attempted the National Health Service (costing only 7.1
reforms designed to reduce the rate of growth percent of GDP in 1992 compared to 8.7 percent
of government health-care spending; Clinton, for Germany and 13.6 percent for the United
in addition, made a failed attempt to over- States) – was radically reformed in 1991 by
come American exceptionalism and establish Margaret Thatcher on both cost-containment
national health insurance. It is a curious paradox and ideological grounds.6 In 1989 she released
Peter-3383-Chapter-11.qxd 6/8/2006 7:35 PM Page 206
Working for Patients, the product of a closely- of capitation payments (originally designed to
held review by her trusted lieutenants plus advis- overcome the overserving tendencies of fee-for-
ers from neo-liberal think tanks – with no service medicine), fundholders will receive a
consultation with provider groups (the BMA, fixed sum out of which they will buy hospital
the Royal Colleges). Answering critics of her services; like the self-governing hospitals, they
stewardship of the NHS who accused her of can retain surpluses but must balance deficits
over-zealous starving of health care through from their budgets (Döhler, 1991, pp. 264–265).
budget stringency, the report assumed that the Finally, in this war on the autonomy of
NHS as structured subverts the free choice of physicians, the Thatcher reforms increased
individuals in the medical market; it claimed the powers of managers of the NHS and
that doctors – so dominant in NHS policy- governing Ministers over practitioners: the
making and implementation – lack the incen- trusts specify doctors’ job contracts and review
tives to respond to the patient-consumer; that them annually; they have increased power to
doctors’ monopolistic, self-aggrandizing behav- hire and fire hospital personnel. Both profes-
ior (e.g., restricting entry to the trade) increases sionals and local governments are barred from
rigidity and inefficiency (e.g., wide variation in serving on controlling boards. The Major
patients treated and referrals per doctor) and government continued to implement these
accounts for long waiting lists for elective 1991–92 reforms.
surgery and ward closures in hospitals that have Britain stands alone in the rapidity, depth, and
used up their NHS budget; and that all of this is implementation of reform, although New
a drag on the economy. An underlying ideolog- Zealand recently tried to match the pace of
ical theme of all of Thatchers reforms was that British change. It is too difficult to judge whether
the autonomy and monopoly power of profes- this combination of managerialism, statism, and
sional groups, including doctors, must be market theory either saves money or improves
checked by creating free markets. The power of access, innovation, equity or quality of care, let
physician groups, she felt, threatens not only the alone real health (Klein, 1995, pp. 230–237).
sovereignty of the consumer but the proper Besides, some of the Tory reforms have been
authority of the state. Although the Thatcher abandoned. Because there was no consensus
government did not dare to abandon the found- on reform or even what problem it was sup-
ing principle of the NHS – to provide a univer- posed to solve, there is no consensus on crite-
sal, comprehensive, tax-financed health-care ria of evaluation. There is consensus on four
service to the entire population – the reforms points, however:
did introduce an “internal market” into the state
system. Purchasers (District Health Authorities) 1. Reform has not reduced costs. Spending
were separated from providers (hospitals and on health care was 6 percent of GDP in
their consultant-specialists); purchasers would 1989; in 1992, after reform, it had gone up
no longer manage hospitals but instead buy ser- to 7.1 percent, a more rapid growth than
vices for their localities by contracting with that of Germany and Netherlands, while
NHS hospitals or private hospitals, hopefully Sweden and Denmark actually cut the
the latter. Hospitals could choose to be self- health-care share of GDP.
governing trusts. By 1994 “more than 400 2. It inflated both the number and salaries of
providers accounting for about 95 percent of the “bureaucrats.” While 50,000 nursing jobs
NHS’s activities had become self-governing and 60,000 beds have disappeared since
Trusts” (Klein, 1995, pp. 204–205). A new struc- 1990, there has been an increase of 20,000
ture – General Practitioner Fundholders senior managers in the NHS (New York
(GPFH) comprised of larger GP practices – was Times, August 6, 1996). There is dispute
created as another group of presumably cost- about how much of a managerial explo-
conscious purchasers. Fundholders would nego- sion there is but no one doubts that there
tiate contracts with providers to offer care more has been a large increase in administrative
efficiently without massive extra funds. Instead costs.
Peter-3383-Chapter-11.qxd 6/8/2006 7:35 PM Page 207
3. It has accelerated a 1980s trend toward Chapter 16 (2002) assesses the connection
co-payments in dental, pharmaceutical, between the organization and financing of
and ophthalmic services, with exceptions health-care systems and national health
for the poor. performance.
4. No great change has yet occurred in the
behavior of physicians or of health author-
ities; for instance, they typically refuse to The Netherlands
explicitly restrict the menu of services. But
the reforms have reduced the power of The Christian Democratic and Liberal cabinet
hospital specialists and secondary-care of 1982 cut social benefits, education, and health
providers, while enhancing the power and care by three percent and froze a number of ben-
status of general practitioners and pri- efits at that level. In 1986 the link between wage
mary-care providers, especially the “fund- increases in the private sector and social benefits
holders” who now have to pay the bills changed so that the Minister of Social Affairs
within a fixed budget. This may be the could consider the ratio of workers to beneficia-
major achievement of reform. However, ries in determining benefits each year. Later the
even with one of the lowest-cost systems generous early retirement benefit was reduced.
among rich democracies, and in the face of In 1991 and 1992 sickness and disability benefits
increased demands, the Conservative gov- were lowered. The social partners, however,
ernment of the mid-1990s required rebelled by restoring the benefits through collec-
annual cuts of three percent at every level. tive bargaining. Despite the unusual Dutch level
Continued budget stringency put an of abuse of these generous programs (Chapter
already lean system under increased strain; 15) sickness benefits by 1995 were back to their
many emergency rooms and intensive- former level at 80–100 percent of previous
care units had to impose long waits before wages while most of the disability benefits were
any treatment (New York Times, January back to their former 80 percent level. An uphill
30, 1997.) effort to reexamine eligibility procedures was
begun in 1995–96. The employer is now respon-
When the Labour Party swept to power partly on sible for paying the first two to six weeks of sick
the promise to reverse the Thatcher/Major leave before the sickness fund kicks in; employ-
health-care ideology, improve service, and spend ers also have more responsibility for disability
more on the NHS, Prime Minister Tony Blair benefits, a reform that gives them an incentive to
increased NHS funding modestly in his first two get rid of less-than healthy workers, thereby
years (e.g. an extra $2.4 billion in the 1997–98 shifting costs to unemployment compensation
budget.) He also reversed some of Thatcher’s or public assistance. (P.K. Keizer, 1996, pp. 4–5,
reforms, especially in the “internal market 15–16.) However, the number on partial or full
system.” In July 2000 Blair promised to accelerate disability did drop from about 985,000 at the
increases. He presented an ambitious 10-year end of 1993 (14 percent of the labor force) to
plan to overhaul the NHS and deliver by 2004 an 735,000 at the end of 1996 (Wall Street Journal,
extra 20,000 nurses, 7,500 consultants, 7,500 December 26, 1996). Recent governments have
more specialists and 2,000 more GPs, plus 6,750 emphasized the need to increase the rate of Dutch
more therapists and other professionals. The labor-force participation and de-emphasize the
plan will be funded by an increase in taxes of 5 replacement of income for the unemployed and
billion pounds to help pay for both health disabled (Chapter 15).
service and anti-poverty programs, themselves The Netherlands story is one of many small
important sources of enhanced health. reductions in a wide array of social benefits
Substantial increases have, infact, occurred. This that, over 10 or 15 years, add up. These cumu-
is a dramatic reversal characteristic of big party lative adjustments and the effort to freeze some
shifts in a polarized, majoritarian political spending provoked older voters to form two
system. new parties that won seven seats in May 1994.
Peter-3383-Chapter-11.qxd 6/8/2006 7:35 PM Page 208
In the late 1990s through 2003 German structural reform was introduced in 1984 but
governments of both center-right and center- it took 10 years to be fully implemented, at
left continued efforts to reform the welfare state which point it provoked a storm of protest
with only small results. The SPD (Social (Ferrera, 1994). For example, in December
Democratic Party) in 2003 accepted Schroeder’s 1993 a three-tier system was introduced (users
“2010 Agenda” aiming to cut non-labor costs by with no charge, users with limited co-payments,
reducing benefits for the long-term unem- users with deductibles). Health-care entitle-
ployed, liberalize the pharmaceutical market ments for higher-income families were drasti-
and increase co-payments by patients. It also cally cut back (the only “free” service
calls for pension reforms. Once more, massive remaining was hospitalization); for families
protests greeted the proposals; only cuts in earning less than the threshold, co-payments
unemployment benefits were adopted. Finally, were increased; while the poor elderly were
the German election of 2005 was a standoff exempt, a new voucher system put a ceiling on
between Angela Merkel, a free marketeer, and all services, after which they would bear the
Gerhard Schroeder, a Social Democrat who full cost. The revolution lasted only a few
defended the “social market” enonomy. What is months. The affluent were incensed at their
obvious is that despite three decades of bold talk exclusion; the poor and near-poor complained
about big cutbacks of the big welfare-state pro- that the voucher system, like the means-testing
grams, the action has been severely limited by that preceded it, was cumbersome and humili-
mass resistance. ating; and everyone complained about
increased out-of-pocket costs. Opposition par-
ties and labor unions gave voice to these com-
Italy plaints. Income-testing and means-testing
created a nightmare for administrators, exacer-
Cutbacks, begun in 1978 under the unity gov- bated by the ineffectiveness of the tax system.
ernment with Communist support, acceler- Income testing encouraged even more tax
ated in 1981–1983 (e.g., rules on public evasion as benefit-seekers just above the income
assistance were tightened, child allowances for threshold rearranged their official labor-force
children of high income families were participation to avoid co-payments.
reduced). But Italy nevertheless joins Sweden, It did not help public acceptance of the 1993
Denmark, Belgium, and France as the standout revolution that in the same year the Head of
cases of welfare-state expansion that substan- Pharmaceutical Services was arrested for
tially exceeds their GDP growth from 1975 to accepting huge bribes from drug companies
the early 1980s (Alber, 1988, pp. 188–189). and the Minister of Health and several of his
Major reforms, driven by cost considerations, associates were indicted on corruption charges.
were made in the 1980s and especially the early In 1994 the new system was abandoned.
1990s.9 A major 1978 reform undertaken by Income selectivity was replaced by risk selectiv-
the National Solidarity Coalition changed ity: persons with one of a long list of illnesses
health-care entitlements from occupation- (mostly chronic), pregnant women, the dis-
based funds to all citizens, modeled on the uni- abled, children up to the age of 10 and everyone
versalism of the British National Health older than 60 were exempted from cost-sharing,
Service – a principle of free and equal benefits regardless of income. Reforms throughout the
to everyone. Severe problems of corruption decade followed a stop-reverse-go pattern.
(parties controlled every component of the Italy has long had one of the most generous
system), concomitant cost explosions, and paid maternity leave policies. Even hints of
deterioration of services resulted in incremen- possible cutbacks provoke a political uproar;
tal revisions in the 1980s and 1990s. These the system remains intact.10
reforms include gradual increases in the Meanwhile the efforts to restructure the
amount of co-payments and their extension to state by decentralization of power also shaped
new areas – e.g., prescriptions, laboratory the welfare state. New laws of 1991 regulated
tests, and out-patient specialist care. A major the financing of non-profit and voluntary
Peter-3383-Chapter-11.qxd 6/8/2006 7:35 PM Page 210
agencies in the social services. These reforms The vigor and breadth of demonstrations
accented self-governing municipalities, the and strikes to protest cuts of benefits in the
merging of small towns and the setting up of core programs of the welfare state are greater
consortia for the management and delivery of in France than Italy, largely because Italians are
services; the creation of metropolitan areas for aware of the corruption of major programs
big cities like Rome, Milan, Turin, Naples (they and the need for reform. But Italian voters are
are now accorded the powers of provinces). moving in the direction of the French in their
Municipalities are acknowledged as main organized resistance to erosion of benefits.
actors in all areas of social services, except for For instance, Prime Minister Romano Prodi’s
the organization and planning of health ser- center-left majority coalition of 1998 split over
vices. Two laws of 1991 aim to regulate and $14.5 billion in proposed spending cuts,
help finance voluntary associations and social specifically the $2.9 billion reduction in
cooperatives to which public administrators pensions and health-care programs. Prodi
may contract out services. Relative to market resigned. Renegotiation led to $291 million less
actors, these cooperatives have a privileged in pension cuts but total spending cuts were
position as contractees. None of this urge to reduced to almost one-third of those originally
decentralize reduced bureaucratic bloat or proposed and in return the left demand for a
total spending but it did enliven the blame 35-hour workweek was accepted.
game: The central government could point to
regional and local governments as the source
of trouble and vice versa. In general, however, France
the structural reforms were a laudable attempt
to reduce political clientelism by strengthening A glimpse of French policy and protest comes
the power and responsibilities of administra- from the Spring of 2003 when the conservative
tors and professionals within local agencies Chirac/Raffarin government proposed a broad
and especially within the National Health retrenchment of the welfare state – curbing
Service. state spending for pensions and health bene-
Italy is an extreme case, both in acceleration fits. An estimated one to three million people
of spending and repeated efforts to cut spend- demonstrated against these reforms in May
ing, only a few of which were successful. It 2003. Protests included strikes by about two-
uniquely combines political corruption and thirds of national railway workers, as well as
inefficiency in the administration of both tax- massive street action. In the end, in July, the
ation and social spending (even including pen- government got Parliament to approve only a
sions, health-care and disability insurance), modest pension reform: civil servants would in
strongly alienated voters, party realignment 2008 be asked to work an additional two and a
and dealignment, and very disorderly finances half years before receiving a full pension after
(a uniquely wide gap between revenues and 40 years service so they will be equal to the cur-
expenditures over long periods). This helps to rent system for all private-sector workers.
explain big swings in social policy and the (Wall Street Journal July 24, 2003.)
limited modifications in the last 10 years – Again, in late 1995, the conservative Chirac/
some shifts toward a more selective universal- Juppe government’s proposed cuts triggered a
ism in welfare-state design, some actual cuts in strike movement that almost shut down the
benefits, and ultimately a major brake on total French economy for three weeks. It had a
social spending. After negotiations between wider base of public sympathy than les grands
unions and the Dini government, even a événements of 1968. It was presented by the gov-
broad-pension reform was passed in August ernment with almost no consultation or partici-
1995. It begins a shift from the old very gener- pation, as a non-negotiable policy package to
ous “defined benefit” formula to a new, some- save the French welfare state, restore French
what less generous “defined contribution” competitiveness, and meet requirements for
formula to be phased in very gradually. joining the EMU (European Monetary Union).
Peter-3383-Chapter-11.qxd 6/8/2006 7:35 PM Page 211
The changes were sudden and radical. Strikers the VAT and, in its most unpopular move,
concentrated on protecting pensions, health care, changed property tax valuations (see Chapter 10)
job security, traditional job rights and working while it made modest reforms to prevent waste
conditions. The government withdrew its pro- and abuse in social programs, especially in the
posal to reform public sector pensions (and was replacement rate of unemployment insurance
subsequently forced to accept a full pay pension and sick pay. For instance, sickness cash benefits
for truck divers at age 55); but did not give up its were calibrated according to reduced capacity to
ambition to drastically reform the entire social work, with a doctor’s certificate required after
security system. In July of 1997 Chirac/Juppe the eighth day. In 1992, the Bildt centrist gov-
were defeated by Lionel Jospin of the center-left. ernment agreed with the Social Democratic
The French rejection of the EU constitution in opposition on a “crisis” package: The maximum
2005 was partly inspired not only by high unem- sickness cash benefit was reduced to 90 percent
ployment but also by Chirac’s current effort at of lost income. Only 65 percent is paid for the
social cutbacks and the public image of EU poli- first three days; 80 percent from fourth to 90th
cies as imposing an Anglo-American economic day. The cost for the first 14 days of claimed
model on Europe that would threaten social sickness was shifted to the employer instead of
protections and labor standards. the government. Since 1993 no sickness cash
In 1995, compared to other rich democra- benefit has been paid for the first day of illness
cies, France had a low “normal” retirement age (Denmark also introduced a waiting period
of 60, a very low labor-force participation rate recently) and other minor cuts were made. The
among men aged 55–64 and a high pension result: a sharp decrease in the high rate of
cost/GDP (almost 11%) (Myles and Pierson, absenteeism; Swedes who used to report in sick
2001, p. 309). In short, its pension problems because of a Monday-morning hangover now
were severe, so it is not surprising that its gov- show up to work.
ernment tries to retrench in the face of mass Similarly, in reforms of a second program
resistance. Neither is it surprising that most vulnerable to abuse, job injury insurance,
Napoleonic efforts and style yield so little. a stricter definition of job injury was adopted in
the early 1990s and the insured were required to
pay a larger portion of the cost themselves.
Sweden11 Deficits in both these programs were turned
into surpluses by 1994, although they remain
Starting in 1980–1982 under a “bourgeois” costly. In addition, small downward adjust-
coalition government, there were minor cuts in ments were instituted in calculating pension
already very generous health, housing, and even benefits; and unemployment benefits were
pension benefits (e.g., reducing compensation reduced to 80 percent of previous earnings (still
for part-time pensions) while unemployment high by any standard).
benefits and child allowances were extended In January 1996 the present Social
(see 2002, Chapter 7). During subsequent years Democratic government decided on a second
of return to Social Democratic rule, 1982–1991, round of austerity – a further tightening of
there were both incremental cutbacks and sick pay, reduced child allowances, and a reform
incremental expansions of social programs. The of family policy: the parental leave benefit was
unprecedented annual deficits run up by a series reduced from 90 percent of income loss to 80
of center-right coalitions from 1976 to 1982 (13 percent (two months of the 12 to 18 months of
percent of GDP when the Social Democrats parental leave will remain at 90 percent, one
returned) were reduced not mainly by these month for each parent, use it or lose it, then 80
1980s spending changes but by currency deval- percent); compulsory counseling will be
uations, economic growth, and tax reforms. required for the divorced or separated; and child
On the eve of a recession, in the two years care allowances were cut (in 1994 dollars, to a
leading up to their defeat in the election of 1991, maximum of about $267 per month up to 3
the Social Democratic government increased years of age); families benefiting from municipal
Peter-3383-Chapter-11.qxd 6/8/2006 7:35 PM Page 212
subsidies for a daycare center get a reduced child One reason for Sweden’s above-average cost
allowance, depending on the hours. of medical care – aside from the strong organi-
Not all of the 1990s reforms involve reduced zations of provider groups and the unusual
benefits and services. The Swedes have long percentage of the aged – is that the system is
led in rehabilitation programs to reintegrate the biased toward hospitals; there is no concept of
physically or mentally disabled into the labor the general practitioner (Ibid., p. 4). Primary
force and promote independent living (for care is organized around local health centers
details see 2002, pp. 550ff.). In 1994 this effort to staffed with salaried doctors, nurses, and
enable the handicapped to live an independent specialists acting as teams. There are signifi-
life was extended via a legal right to benefits, cant user charges – co-payments for prescrip-
including counseling, support for individuals as tions, visits to doctors, and other services
well as parents with disabled children, escort (Ibid., p. 5), with no exception but an upper
services, housing with special service, support limit of 1600 SEK or $272 per year in 1990
for minor handicaps, interpreter services for the dollars for all publicly-provided health care.
deaf and blind, and more. At the same time, In response to economic austerity and the
eligibility standards were tightened. need for increasing efficiency and consumer
Finally, after 15 years of debate about pension responsiveness, parties of left and right agreed
reform, the 1996 proposals will result in real in the mid-1980s to replace fee-for-service with
reductions, especially for middle and upper capitation payments from the national insur-
income earners – e.g., in the pension benefit for ance fund directly to counties, which would
married pensioners – and a gradual rise in the decide on reimbursement of practitioners.
retirement age. The accrual rate at retirement Most counties are led by Social Democrats.
will be indexed to the average life expectancy of There are “almost as many reform programs as
the retiree’s cohort, so more years in the labor there are [county] councils” (Ibid., p. 8).
force will be required to receive benefits equal to “Patients still have full freedom of self-referral”
those of earlier cohorts. Two percentage points (Ibid., p. 6). The net effect of these reforms is
of the 18.5 percent contribution rate will be that waiting time and lists for non-emergency
allocated to a “premium reserve” account to surgery and out-patient visits have been
pay benefits on the basis of return on invested reduced, especially in the Stockholm district,
capital – a small step away from pay-as-you-go and the GDP share is down from the 1980 peak
financing. Sweden leads in a continuing effort to of 9.4 percent. In fact, Sweden is the only
make the pension system flexible by increasing a country among our 19 rich democracies that
choice of various combinations of work and reduced the share of total domestic expendi-
pensions. Sweden is also the world’s “oldest” ture devoted to health care from 1980 to 1991
country; in 1995 17.4 percent of its population (OECD, 1993, Vol. 1, Table 1).
was 65 or older compared to 12.7 percent for the With all of these reforms, and despite levels of
United States. unemployment and annual deficits unprece-
What about health-care reform? In 1990 the dented for Sweden, the contours of the most
Swedish health-care system spent 8.6 percent of generous welfare state in the world remain
GDP – less expensive than the United States intact; the adjustments may actually make the
and Canada (despite Sweden’s much older social-spending share of GDP level off or even
population) but more costly than the rest of our move slightly downward. In per capita benefits
19 rich democracies. It delivers “care of high the Swedes moved in the 1990s from lavish to
quality, equity in access and local accountability merely very-generous social expenditures.
unrivaled in other countries” and a high level
of patient satisfaction (Glennerster and
Matsaganis, 1992, pp. 3, 5). (In a recent Swedish Finland: Balancing Equity and Cost
national survey 85 percent thought the quality
of medical care in Sweden was either very good Finland’s recent welfare-state reforms are a
or good.) model of what corporatist bargaining among the
Peter-3383-Chapter-11.qxd 6/8/2006 7:35 PM Page 213
social partners (labor, professions, management, age 18–52, to 4.5% at 63–68 – a strong incentive
and other relevant and inclusive interest for continued work. Various grandfather clauses
groups interacting with government and polit- are included to prevent transitional shocks to
ical parties) can do to forge consensus on older workers. Finally there will be provision for
major pension and related programs. It is a unpaid periods: periods of earnings-related
lesson in how a consensual democracy, espe- maternity, paternity, and parental leave; days
cially if it integrates labor and the left into high with earnings-related unemployed allowance,
policy, can reform the core programs of the periods of study; caring for a child under 3, and
welfare state to cope with issues of both equity job alternation leave – all with an accrual rate of
and cost. 1.5%. A wage earner can take a job alternation
Here is a broad outline of a major reform of leave by agreement with the employer – stop
pension and related programs that was adopted working for a defined period, e.g., a year, and
by parliament in 2003 after agreement among receive partial compensation for lost earnings. It
the “labor-market parties” and after a thorough is like a publicly financed sabbatical.
government assessment of long-run costs and Funding will be increased from 2003 on, but
benefits. Most took effect on Jan 1, 2005.12 savings from the flexible retirement provisions
The major aims: postpone the average effec- and later retirement are expected to reduce the
tive retirement age by 2–3 years; adapt the growth of pension expenditure as a percent of
pension scheme to an increased average life wages from 2005 to about 2030 and thereafter
expectancy and other demographic shifts; to level off. Combined with real cuts made in
achieve unification and simplification of pri- the 1990s – a product of the same consensual
vate sector earnings-related pensions. The new bargaining process among the social partners –
system reforms cover three types of pensions: the current reforms mean that pension expen-
old age, disability, and part-time. (Survivors ditures and premiums will grow much less
pensions were left intact.) It changes the pen- than they would have without reforms.13
sion index (so consumer prices weigh 80%, All rich democracies have experienced twin
earnings increases 20%). It establishes specific trends that pose a serious issue for policy plan-
rights to vocational rehabilitation “if illness, ners: a century-long decline in the age of exit
handicap, or injury poses a threat to work of men from the labor force; and an increase
capacity within about 5 years” and similar in healthy older populations. At the root of
rights within the earnings-related pension the problem is management and union desire to
system, with appeals procedures. Changes in ease out older workers. Managers prefer
the disability pension were necessary because younger, cheaper men and women and middle-
some early retirement schemes were abolished aged women and, if the state can pick up the tab,
(individual early retirement pension and they will help older workers into an early retire-
unemployment pension). The tax-financed ment; unions go along because they want to
unemployment insurance program remained. reduce unemployment and make way for
The flat-rate pension was changed to reflect younger members. In trying to contain explod-
the increased retirement age of the earnings- ing costs of pensions while they cope with an
related mandate. oversupply of healthy displaced older workers
To further enhance flexible retirement and who prefer to work at least part time, many
encourage work, changes in both part time governments have tried to devise flexible retire-
and full pensions are substantial. The part- ment systems. Surely it is good public policy to
time pension age is increased from age 56 to transform the healthy aged who want to work
58, with an accrual rate of 1.5% of earnings into taxpayers, part-time workers, and partial
and 0.75% of the reduction of earnings due to pensioners rather than pressuring them to retire
part-time reitirement. Early old-age pensions fully. But it is extraordinarily difficult both tech-
start at 62. At 63 one can retire on the old- nically and politically to craft social-security
age accrual rate, which begins on earnings after systems that would reverse the long-term slide
age 18 and climbs in three steps, from 1.5% at in the age of exit from work. One obstacle is the
Peter-3383-Chapter-11.qxd 6/8/2006 7:35 PM Page 214
reforms, in adopting and financing them, and long-term care for the aged, a program that enjoys broad
implementing them. popular support. Finally, the measure – percentage of total
social spending – yields results different from those of GDP
shares or per capita benefits. Cuts in real benefits in means-
tested public assistance from 1980 to 1995 were substantial
NOTES in both countries (see Chapter 6. Cf. Bawden and Palmer,
1984; and Alber, 1996).
6. This account of NHS reforms is based on Klein, 1995;
1. In 1966 East Germany and Czechoslovakia ranked Giaimo, 1995, and Döhler, 1991.
eighth and ninth of these 22 in social-security spending 7. One estimate suggests that by 2030, 38 percent of
while the USSR lagged at 16th. They fit a convergence theory Germans will be over 60, 16 percent under 20, Economist,
because their estimated GNP per capita at factor cost puts February 1994. Assuming no change in retirement age, a
the USSR quite a bit below the two more modern slightly different measure for 1995 and 2025 – the ratio of
Communist regimes, both of which got into the welfare- retirement-age persons to 100 working-age adults – yields
state business earlier than the former USSR and have older similar results: in 1995, 29.6, in 2025, 54.1. That is almost
populations (Wilensky, 1975, pp. 121–128). Because so as burdensome as Austria’s 36.7 and 61.4. Comparable
many of the data on these three countries are of dubious figures for the United States, where the attack on “entitle-
comparability, we dropped them from further analysis. The ments” is most hysterical, are only 22.3 and 35.1.
19 rich democracies were by the mid-1960s, in the upper 8. The long established system of tripartite self-governance
one-sixth of the world’s GNP/capita and remain there. I involves government, patients, and providers, with some
eliminated democracies with less than 3 million population. input from other groups. Non-profit sickness funds based
2. Because of variation in what is off-budget, in one- on occupation or region combine in national associations
time asset sales, and in other obscurities, measures of debt and must bargain with physicians associations; they are gov-
and deficits are not included in my analysis of economic erned by national framework legislation. Since 1977 they
performance, chapter 12; cf. Pashnik, 2000. must take account of annual recommendations of the
3. Heavy means-testing in these programs creates polit- Concerted Action in Health Care, a roundtable which
ical resistance to funding them at an adequate level. The includes labor unions, employers, officials at various levels
distinction between divisive highly-visible means testing of government, and various provider groups. See Chapter 16.
and simple quiet income testing is important. By “means- 9. The following account of recent reforms of Italian
testing” I mean (1) noncategorical benefits targeted to the social programs is based on Maurizio Ferrera, 1994; Chiara
poor via a stiff income- and/or assets-test, (2) applied by Saraceno and Nicola Negri, 1994; MacFarlan and Oxley,
welfare administrators with substantial discretion, (3) with 1996 and my interviews.
a high probability of stigma. “Income testing” is the oppo- 10. At this writing, mothers are still required to take off two
site. It is categorical as a social right with co-payments months before and three months after childbirth. For risky
graded by income bracket and, because it is private and pregnancies, they can take off the entire period of pregnancy.
invisible, has no stigma. Means-testing is characteristic of (Many women claim “risky” pregnancies from day one, with
Britain and other decentralized and fragmented political the acquiescence of their physicians.) Most working women
economies (USA, Canada, Ireland); democratic corpo- receive full salaries during leave, 80 percent covered by social
ratist countries, especially those with cumulative left party security, 20 percent by employers. Chapter 7 compares an
power, avoid means testing and rely much more on income array of family policies in 19 countries, including parental
testing (Wilensky, 2002, pp. 321–332.) leave. It shows that corporatist countries with the combina-
4. The exception is a country with a Westminster elec- tion of strong Catholic and left parties score medium on an
toral system – first-past-the-post, winner-take-all. That index of innovative and expansive family policies, reflecting
permitted Thatcher to rule for 12 years with about 40% of the ambivalence of Catholic parties toward women’s place
the popular vote and a steadily rising majority hostile to and the offset of both corporatism and left power. But on
her social policies (Wilensky, 2002, p. 374). maternal leave Italy ranks high and remains there.
5. Pierson concludes that Thatcher had more success 11. Based on my interviews; publications of the
than Reagan in increasing the share of total social expendi- Ministry of Health and Social Affairs and the Swedish
tures that are means-tested (Table 6.3, p. 145). But the com- National Insurance Board, various years; Niemelä and
parison is misleading. He excludes U.S. state expenditures Salminen, 1994; Huber and Stevens, 1996; Kuhnle, 1996;
from the programs he covers (Medicaid, food stamps, public Palme and Wennemo, 1998.
assistance, and Supplemental Security Income). Moreover, 12. This account is based on documents supplied by the
the aggregate figure lumps together visible, stigmatized ben- research scholar who directed the calculations and analysis
efits (AFDC and food stamps) with invisible programs tar- carried out at the Finnish Centre for Pensions, Hannu
geted to the poor (EITC), which lack stigma and can grow Uusitalo. I am grateful for his comments.
more easily (see Chapter 8); SSI, where the aged population 13. With reform, as % of GDP, pension spending is
is the target and where stigma is less; and Medicaid, where 11.7% in 2003 and with reform is estimated at 14.5% in
provider and administrative costs soar and where much of 2030 (15.8% if no reform had been made), 13.8% in 2050
the money comprises the only government support for (15.7% if no reform).
Peter-3383-Chapter-11.qxd 6/8/2006 7:35 PM Page 216
Saraceno, Chiara and Nicola Negri (1994). “The Expenditures. Berkeley, CA: University of California
Changing Italian Welfare State.” Journal of European Press.
Social Policy 4, 1: 19–34. Wilensky, Harold L. (1981a) “Democratic Corpo-
Seeleib-Kaiser, Martin (2002). “A Dual Trans- ratism, Consensus, and Social Policy: Reflections
formation of the German Welfare State.” West on Changing Values and the ‘Crisis’ of the Welfare
European Politics, 24, 4 (October): 25–48. State.” In The Welfare State in Crisis: An Account of
Von Hagen, Jürgen (1992). Budgeting Procedures and the Conference on Social Policies in the 1980s.
Fiscal Performance in the European Communities. Paris: Organization For Economic Cooperation
University of California at Berkeley Center for and Development, 185–195.
German and European Studies, Political Economy Wilensky, Harold L. (1981b) “Leftism, Catholicism,
of European Integration Research Group, Working and Democratic Corporatism: The Role of
Paper 1.9. Political Parties in Welfare State Development.”
Wilensky, Harold L. (1965) “The Problems and In The Development of Welfare States in Europe
Prospects of the Welfare State.” In Industrial and America, ed. Peter Flora and Arnold J.
Society and Social Welfare, by Harold L. Wilensky Heidenheimer, 345–382. New Brunswick, NJ:
and Charles N. Lebeaux. v-lii. Enlarged Edition. Transaction Books.
New York: Free Press-Macmillan. Wilensky, Harold L. (2002) Rich Democracies:
Wilensky, Harold L. (1975) The Welfare State and Political Economy, Public Policy, and Performance.
Equality: Structural and Ideological Roots of Public Berkeley: University of California Press.
Peter-3383-Chapter-11.qxd 6/8/2006 7:35 PM Page 218
Peter-3383-Chapter-12.qxd 6/8/2006 7:35 PM Page 219
12
Health Policy
MICHAEL MORAN
political science were already entrenched in and by the need to make some practitioner
faculties and degree programmes. The result relevant sense of how to allocate this huge flow
has been that individuals interested in policy of resources. The result, unsurprisingly, is a
studies have tended to find a home within the field of study which is dominated by substan-
pre-existing institutional shelters – established tive issues to do with the health field itself.
social science departments – or, as we shall see The reader might respond that it is hardly
in a moment, have tried for differentiation by surprising that students of health policy are
stressing distinct, substantive policy domains. focused on substantive issues to do with
(It is of course interesting to ponder why health. But this apparently obvious feature,
policy studies was not able to achieve what was replicated across substantive policy domains, is
accomplished by other new quasi-disciplines, exactly what helps explain so much of the frag-
like cultural studies, which was to use the inde- mented nature of policy studies as a field. And,
pendent power of student demand to acquire a as we shall see in a moment, the historical con-
secure, independent institutional base.) juncture between the rise of health policy as a
Part of the history of the field of policy stud- specialised field of study and the rise of health
ies is therefore shaped by the timing of its as a major consumer of societal resources has
emergence as an activity in higher education. had important consequences for the way
A second critical aspect of the timing of intel- health policy is framed as a field of study. As in
lectual development is particularly relevant to all framing exercises, the results privilege parti-
health policy. The study of health policy on any cular approaches, questions, and even answers,
systematic scale is closely associated with, and, over others.
as we shall see, has in part been prompted by
the rise of health as a significant policy sector
in the state. The most obvious indicator of that
FRAMING HEALTH-CARE POLICY:
rise is money. In the three decades or so after
THREE DOMAINS OF HEALTH CARE
1945 – the period roughly corresponding to
the ‘long boom’ which was the most important
feature of the leading capitalist economies in Policy domains are not ‘given’; they are con-
the middle decades of the twentieth century – structed. The practical meaning of ‘health
societal spending on health care rose substan- policy’ famously varies from jurisdiction to
tially, and public spending more or less jurisdiction. These variations are the product
matched this rise. (For a summary of policy of entirely unsurprising considerations: of cul-
history, Moran 1999: 2–3; and OECD, various tural understandings of what constitutes the
dates, for the underlying data collection.) The concept of health; of administrative and his-
observation holds virtually right across the torical conventions; of bureaucratic politics
advanced capitalist world. And the three which in virtually all jurisdictions constantly
decades since the end of the long boom have reshape the ‘turf ’ of state institutions. (For
been dominated by attempts to cope with the some illustrations, Kelman, 1975; Wildavsky,
consequences of that first great historic rise in 1977; Evans and Stoddart, 1990.) A simple
spending: if the thirty years after the Second example of this point at work in practice can
World War were the age of health-care expan- be provided by the nightmare of trying to con-
sion, the next thirty were the age of struct accurate comparative cross-national
(attempted) cost containment (for examples: estimates of health-care spending, where the
Levit et al., 1994; Saltman and Von Otter, 1995; results are hugely sensitive to the classification
Altenstetter and Björkman, 1997; Freeman, conventions allocating particular programmes
2000). The study of health policy has been to particular domains. For instance, Swedish
shaped by this extraordinary history: by the health-care spending has looked consistently
attempt to make sense of those early boom high, viewed comparatively, because, unlike
decades; by the attempt to make sense of what many other nations, care of the aged in Sweden
has happened since the end of the long boom; has been allocated to the health care, rather
Peter-3383-Chapter-12.qxd 6/8/2006 7:35 PM Page 221
than the social security, budgets. Beneath this the policy problems arising from the exercise
administrative variation in classification lies a of those entitlements. The shape of the health
conceptual jungle, notably the tangled issues of policy arena is thus largely determined by the
what constitutes ‘health’ itself. wider welfare state arena. This conceptualisa-
This does not mean that our understanding tion of health policy as health-care policy then
of health or of health policy is an arbitrary feeds into the definition of both policy prob-
matter. On the contrary: it has been shaped lems and policy issues. Thus, the account of
intensely by national cultural patterns, by the the last three decades or so is seen largely as an
workings of economic interests, by the age of retrenchment, parallel to the wider pres-
manoeuvrings of bureaucratic politics – and sures experienced by the welfare state (for
by recent historical experience. The last of instance, Ferrera and Rhodes, 2000). This
these takes us to the heart of the way health approach, we shall see later in the chapter, has
policy has been framed by both policy makers also helped frame the definition of the most
and researchers in the last generation. It is a famous and most intensely studied puzzle
simplification, but a serviceable simplification, in health policy: the puzzle of what is com-
to say that health policy can be framed in monly called American exceptionalism, to
one of three ways: as a species of social policy; which I later return. As I shall show, American
as a species of labour market policy; and as ‘exceptionalism’ is indeed only ‘exceptional’
a species of industrial policy. The three are not when health policy is framed in these ‘welfarist’
mutually exclusive. On the contrary, they arise terms. And when health policy is framed in
out of historically embedded linkages between alternative ways – notably as a form of indus-
different parts of health-care systems. trial policy – the United States, far from looking
Converted into influences over how we can like a deformation, looks state of the art.
think about health policy, however, they truly Approaching health-care policy through the
represent (mostly unselfconscious) framing recent historical experience of the welfare
exercises: that is, the different ways of framing state, and through the vocabulary of social
put different parts of health-care systems at the policy, has some clear advantages. It reflects an
centre or the margins of the policy picture. obvious, important facet of the historical expe-
Some modes of framing are more popular rience of the last sixty years, since plainly the
than others, which is where the policy experi- ‘golden age of the welfare state’ and the golden
ence of health care in recent decades becomes age of health-care provision did indeed closely
relevant. The boom in health-care spending in coincide. It also helps counter one of the key
the decades of the long boom, and the turn to features of the study of public policy generally
efforts at cost containment in the decades since identified at the start of this chapter – the ten-
the mid-1970s, plainly mirrors a wider policy dency towards balkanisation – since it links
history: that tracking the golden age of the wel- themes in health policy to important sets of
fare state, and the decades of structural reform wider themes in welfare policy.
and comparative austerity that have followed But this particular framing device, like any
the end of the golden age. (Definitively other mode of framing, has its costs. It focuses
anatomised in Esping-Andersen, 1996). attention overwhelmingly on issues of health-
Framed thus, health policy is a sub-set of wel- care consumption: crudely, in the era of the
fare policy. The way to tell health policy stories long boom on how it could be extended, and in
is in the vocabulary, and the dramaturgy, of the era since the end of the long boom on how
stories about the achievements and problems it can be constrained. Health-care policy gains
of welfare states. When the health arena is from being linked to a wider policy domain –
framed in these ‘welfarist’ terms, the dominant that of welfare policy – but is heavily shaped by
vocabulary is a vocabulary of consumption the consumption concerns of that wider
and of social citizenship. It focuses on entitle- domain. If we push our historical vision back-
ments to health care, on the conditions under wards, beyond the beginning of the golden age
which those entitlements can be exercised, and of the welfare state, we can immediately see
Peter-3383-Chapter-12.qxd 6/8/2006 7:35 PM Page 222
other framing possibilities. In the half century labour market policy. That in turn ties
or so before the beginning of the First World health-care policy to the wider imperatives of
War, in both North America and Western labour market policy: to issues such as labour
Europe two critical features of modern health- supply, training and the discipline of the
care systems were established. Both concerned labour force. One very important connection
the delivery rather than the consumption of care, with these wider economic policy issues is pro-
and both have left huge imprints on modern vided by the need to manage medical labour
health-care systems: in summary, they help forces which are increasingly globalised, or at
explain why the delivery of modern health- least internationalised: issues of the circulation
care is a labour intensive process, in which the of medical labour forces across national
labour is highly professionalised, and why for boundaries, of the supply and training of
over a century it has involved the mobilisation labour, and of the place of medical labour in
of an elaborate apparatus of high technology. the international political economy of labour
These two features of health-care delivery are all central (Moran and Wood, 1996; Alsop
immediately introduce us to our two alterna- and Saks, 2002).
tive ways of framing health policy. In the Framed thus, the appropriate vocabulary of
second half of the nineteenth century doctors health policy becomes subtly, but importantly,
emerged as the dominant actors in health-care different from the vocabulary offered by a ‘wel-
systems, separating themselves from a long his- farist’ framing. For now, health policy is tied to
torical association with lesser occupations. In the political economy of labour markets, and
the United States, for example, the process the stories about policy that need to be told
involved the suppression of a wide range of involve in particular one part of labour
‘folk’ medicine with its roots in a democratic markets – that marked by competition for pro-
culture (Starr, 1982). In the United Kingdom, fessional power and privilege. In terms of
and in other European countries, it involved research agendas, too, the study of health
the separation of doctors from a long trail of policy framed thus becomes linked to the
more ‘artisan’ occupations, like barbering, and kinds of issues of occupational control central
a process of social closure involving all the to the classic literatures on professionalism
classic forms of professionalisation: suppres- (Johnson, 1972).
sion of price competition, control of entry, Framing health policy as a process involving
regulation of both clinical and ethical practices service delivery through highly organised and
(Freidson, 1970; Berlant, 1975; Waddington, competing institutions of professionalism
1984). These developments gave to the organi- already ties health care to the international
sation of medical care a highly distinctive political economy of the advanced capitalist
imprint, of which two are particularly impor- world. Our second historical ‘take’ on delivery
tant: the almost universal domination of produces a way of framing policy that makes the
health-care delivery, at least in the advanced link to a global political economy even more
capitalist world, by doctors; and the long term direct. In the decades before the First World
emergence of a wide range of specialised pro- War, when the medical profession was estab-
fessions attempting to emulate the success of lishing its ascendancy in health-care delivery,
doctors in establishing their occupational there also occurred a great change in the tech-
autonomy. nology of medicine. Hitherto a species of
An important consequence of these ‘bodging’, or the domain of the low level arti-
developments – labour intensive service deliv- san, medical technology was subjected to the
ery, an elaborate division of labour based on full power of laboratory science (Reiser, 1978).
professionalisation, the historical dominance of In the half century or so from about 1870 there
one particular profession, doctors – is to alter emerged two linked industries which became,
the ‘frame’ of health policy. Viewed in this light, and remain, central to health-care delivery:
much of health policy is an aspect, not simply of the pharmaceutical industry, based on labora-
the consumption of welfare type services, but of tory developments in immunology; and the
Peter-3383-Chapter-12.qxd 6/8/2006 7:35 PM Page 223
medical devices industry, which at its most Health care policy, like other policy
highly capital intensive was spurred by, first, domains, does not ‘frame itself ’. How it is
developments in electronics, and latterly by IT framed is not an arbitrary business, but it is an
related innovations in digitalisation (Brown, uncertain business. The purpose of sketching
1979; Pickstone, 1992; Foote, 1992). These the three frames above is to show that how the
developments have by the twenty first century meaning of health policy is conceived depends
created a large, globalised industrial economy heavily on which route we approach health
of health care. The technologically based domains through. Is it the world of the health-
health-care industries are central to the indus- care consumer, in which case we instinctively
trial economies of many of the most advanced frame health policy as a facet of welfare policy?
capitalist economies. They are also often symbi- Is it the world of professionalised medical
otically connected to other industrial sectors – delivery systems, in which case we instinctively
a striking example being the link between see health policy ‘framed’ inside labour market
digitalised health technology and the wider policy, especially policy about professionalised
IT sector (for example Trajtenberg, 1990). An labour markets? Or is it the world of medical
equally striking link existing historically is technology, in which case we frame health
between innovation in the technology of war policy as part of the wider management of the
and innovations applied in the health-care modern industrial state?
sector. Much of the highly sophisticated The way policy is framed is plainly impor-
imaging technology used in medical care, for tant for how we study it, but in the case of
example, has its origins in innovations like health policy it is important also for a more
ultrasound used for hunting and detection in substantive reason. The way ‘framing’ happens
sea warfare; many modern developments in depends crucially on the role of states. States
fields like surgery are traceable back to innova- are key actors in all three domains of health-
tions in battlefield surgery designed to cope care policy sketched here – consumption of
with the horrific injuries of the great total wars care, delivery of that care as a personal service,
of the twentieth century; and the development application of technology to care. But, as we
of penicillin was greatly assisted by innovation will shortly see, they are actors in very different
under the demand of war (for examples, Foote, ways. The state’s role in the three domains in
1992; Temin, 1980). effect helps us identify what kind of health-care
The three domains of health policy identified system we are looking at. All states are in differ-
here obviously overlap in important ways. For ent ways ‘health-care states’, in the sense that all
instance, health care as a form of welfare con- health-care institutions are embedded in the
sumption intimately involves the profession- surrounding state system. That innocent image
alised world of health-care labour markets; and of ‘embeddedness’ is suggesting two distinct,
the twin roles of the health-care technology complex processes, The first is obvious: that
industries as suppliers of goods central to the states have a shaping and continuing influence
consumption of modern health care and as key over the character of health-care institutions –
sectors of the modern industrial economy over consumption, over professionalised labour
hugely complicates the management of both markets, and over medical technology. Indeed,
health-care consumption and the management in many cases they have been historically
of national economies. But ‘framing’ is a par- responsible for constituting these markets, and
ticularly appropriate image to describe how we everywhere remain critical to their functioning.
think about three domains of health policy. But a second facet of embeddedness is less
When we frame in one way – whether in ‘wel- obvious: health-care institutions also have a
fare policy’ or ‘industrial policy’ terms – we shaping influence on their surrounding envi-
decide what is put into the centre of the picture ronment. They amount to large scale concen-
and what is pushed to the margins, even when trations of interests, economic activity and,
we do not acknowledge, or are perhaps even often, electoral muscle, and in these guises
not aware, of what is going on. intervene in the wider shaping of policy.
Peter-3383-Chapter-12.qxd 6/8/2006 7:35 PM Page 224
To try to make these sweeping generalisations the United States, we shall see in a moment, is
clearer I next show, by some examples from rather different: in a nutshell, the United States
real existing states, how and why ‘health-care is sui generis, but a unique case of overwhelm-
states’ do indeed exhibit different policy pat- ing importance for the global system of health-
terns, and encounter different policy problems. care policy.
Space imposes drastic selection in this exercise, Narrowing the focus on these cases plainly
but the highly selective approach is justified has intellectual costs. Two should be high-
because of the point of the exercise. My pur- lighted. First, the cases focus us mostly on
pose is not to offer a systematic classification of health care in the advanced capitalist world.
health-care systems, but to use some critical My defence for this is simply pragmatic: that,
examples to show how the varying roles of beyond noting some of the global characteris-
states help constitute the character of some of tics of the political economy of health care,
the most important kinds of health systems. In space just does not allow analysis of health
this way, we will see how the manipulation of policy in the impoverished nations of the
the framing exercise critically determines how world. Second, though the United Kingdom
we compare systems and how we view policy and Germany are here to do work on behalf of
problems and patterns. wider groups of health-care systems, the
implicit classifications which they represent
are not exhaustive. In particular, we should
note the exclusion of Mediterranean members
STATES, HEALTH-CARE STATES AND
of the European Union, with their own dis-
POLICY PATTERNS
tinctive histories of clientelist politics, which
have played an important part in shaping pat-
States are central to the practice of health policy, terns of health policy in those countries; and
and how states ‘frame’ health policy is the product the health systems of the rising economies of
of a complex mix of historical factors and insti- East Asia which, like those economies, exhibit
tutional constraints. And the most immediately highly distinctive patterns of both state inter-
observable feature of states are the striking vari- vention and policy development (for examples
ations in the way states behave in health-care of the former, Ferrera, 1996; Guillén and
systems. Even if we confine our observations to Cabada, 1997).
the advanced capitalist nations, we soon see very If the case for focusing on the United States
striking variations. is its uniqueness, in what does that uniqueness
In this section I try to make these points consist? In a nutshell, the uniqueness is three-
concrete by taking three examples of national fold: it is analytically unique; it is unique in its
health-care systems: the United States, the scale; and it is unique in its impact on the
United Kingdom and Germany. These systems wider global political economy of health.
have a powerful substantive importance. Of these three species of uniqueness, the
Simply: by most measures they are big, with first has dominated the comparative health
the United States by any measure the largest policy literature. Every comparative observer
national health system on Earth, and Germany of health-care policy very quickly notices a
by most measures the largest in Europe. But striking American anomaly, which has been
their choice does not depend on these substan- well expressed by Rothman as follows:
tive measures. They stand as critical cases. In There are some questions that historians return to so
part – in the case of the United Kingdom and often that they become classics in the field, to be
Germany – they exemplify larger families of explored and reexplored, considered and reconsidered.
health-care systems, and the justification for No inquiry better qualifies for this designation than the
question of why the United States has never enacted a
focusing on them alone is that, in the space
national health insurance program. Why, with the
available, this is the most economical way of exception of South Africa, does it remain the only
illustrating with convincing detail the traits of industrialized country that has not implemented so fun-
these larger families. The case for focusing on damental a social welfare policy? (1994: 11).
Peter-3383-Chapter-12.qxd 6/8/2006 7:35 PM Page 225
Over a period of nearly eighty years to the 1990s single predictor – of the level of health-care
there were six major attempts to create some spending is some standard measure of national
system of comprehensive health insurance in wealth, such as per capita income. But any
the United States: they spanned the Progressive examination of the trend line of the distribu-
campaign for compulsory insurance that origi- tion points on a graph mapping the health
nated before the First World War, to the failed spending/wealth correlation soon makes clear
Clinton reforms of the 1990s. All failed. that the United States is well above the trend:
The particulars of failure have been exten- in short, it spends far more on health-care than
sively trawled over in a rich series of policy we should ‘expect’ given its wealth. The best
histories (for an anthology of failure, Morone summary of why this state of affairs obtains
and Belkin, 1994; for the most recent compre- links it to the historical development sketched
hensive failure, of the Clinton reforms, above, and helps explain the twin anomalies of
Skocpol, 1996). Not surprisingly, these individ- the failure to develop comprehensive health
ual policy histories all stress their own special insurance and the unusually high levels of
contingencies. Behind all these contingencies spending on health care. It has been offered by
lies broad structural features of the relations Jacobs:
between the state and health care in the United the general sequence and form of health policy in the
States. The most important of these relate to United States diverge from those of all other industrial-
the historical timing of the development of the ized nations. The U.S. government’s first and most gen-
system. The American federal state emerged erous involvement in health care focused on expanding
the supply of hospital-centered, technologically sophis-
late – or at least emerged onto the health-care ticated health care … . In contrast to the United States,
scene as a significant actor after the rise of however, other Western countries have made the expan-
other powerful interests in health-care policy. sion of access their first and primary priority; govern-
Of the latter, two were especially important. ments have accelerated the expansion of supply in
First, there were the interests embodied in the response to widening access and growing demand for
care. (1995: 144–5).
medical profession, who by the early twentieth
century had organised into a powerful regula- This quotation catches exactly the important
tory group, and a powerful, well organised feature. The state in the U.S. has been by no
national lobby in the form of the American means an unimportant actor in the health-care
Medical Association. The second was a nascent system. But in terms of the ‘three worlds’
‘medical industrial complex’: the creation of a sketched earlier in this chapter the contents of
high quality, well organised research based its interventions have been most heavily shaped
complex linked to important corporate inter- by the interests represented by the two linked
ests, and united in the pursuit of advances in worlds of health-care suppliers – suppliers of
high technology medicine. The upshot was professionalised medical care and suppliers of
that, from the beginning the shape of health high technology medical care. It has, in sum-
policy in the United States has been dominated mary, been a supply state. That is, it has been
by private corporate interests, and particularly dominated by actors in the second and third
by well organised suppliers of medical care domains identified in our preceding section, the
(Wohl, 1984, on the origins of the ‘complex’). organised corporate suppliers of care.
This history also helps explain the second The mixture of sheer scale and the impor-
unique feature of the US system. The country tance of suppliers of care – notably suppliers of
almost unique in not offering some kind of high technology medicine – also lies at the base
comprehensive health coverage is also the of the third unique feature of the American
country with by far the highest level of spend- system noted earlier: its unique global impact.
ing on health care. A word should be said about Two remarkable signs of this should be high-
this correlation. Health spending has many of lighted. First, some of the most important inno-
the characteristics of a ‘luxury’ good, not least vations in care, many stimulated by the search
that, across nations, spending rises with wealth. for cost containment, have been diffused from
One of the best predictors – perhaps the best the United States. One reason for this is the
Peter-3383-Chapter-12.qxd 6/8/2006 7:35 PM Page 226
who provided primary care from ‘high street’ The United Kingdom’s command and
surgeries. The general practitioners were key control system shared one important charac-
‘gatekeepers’ to the wider health-care system. In teristic with the greatest of all the corporatist
a famous summary, Klein characterised the systems, that developed in Germany: they were
British system as ‘the politics of the double bed’: based on an attempt to suppress market forces.
And, as in the UK case, we shall see, this
On the one hand the state became a monopoly
employer: effectively members of the medical profession attempt at suppression has been a source of
became dependent on it not for only for their own immense policy problems. At the core of these
incomes but also for the resources at their command. corporatist arrangements lay two features: the
On the other hand the state became dependent on the institutional arrangements for financing
medical profession to run the NHS and to cope with the
health care; and the institutional arrangements
problems of rationing scarce resources in patient care
(Klein, 1990.) for its delivery. Financing of health care is
extensively ‘franchised’ to independent health-
There was also a critical intersection with the care insurance funds, typically occupationally
‘third world’ of health-care policy: with that cre- based. The delivery of health care, both pri-
ated by the medical technology industries. mary care and hospital based care, is likewise
Although the ‘command and control’ system in extensively franchised to associations of doc-
health care mimicked some of the features of tors, who typically control entry to, and man-
a command and control economy, it is this point agement of, the profession. The financing of
of intersection that reminds us that it was an health-care delivery is largely dependent on
island of socialism embedded in a large market independent negotiations between the insurers
economy. For over a century the market econ- and the health-care deliverers (for overviews,
omy has been a powerful driver of medical Alber, 1992; Burau, 2001). A systematic
technology – and a driver increasingly organised attempt is thus made to lock together two of
on a global scale. Not only is this system of inno- our three domains of health-care policy: the
vation largely out of the control of the state – and domain of consumption and the domain of
in particular out of the control of any ‘command service delivery. One persistent theme of
and control’ arrangement. It is a global economy health-care policy in corporatist systems such
with a particular bias: it is dominated, both as far as this is the struggle for power and resources
as production and markets are concerned, with between the key actors in the two domains: the
the ‘supply driven’ American health-care system various insurance funds who pay for con-
that we described earlier. The British medical sumption, and the key dominant interest,
technology industries are locked into this global doctors, in the domain of delivery.
system, and, especially in the case of the phar- Two sets of interests do not have an obvious,
maceutical industries, are comparatively success- settled place in this corporatist system. First,
ful parts of the competitive economy. The effect the key actors in the third domain of medical
of this intersection between the two domains of technology – the big corporatist interests who
health care – the domain of consumption, and dominate the production and marketing of
the domain of technology production – was medical innovation – are in effect external to
to produce a fatal contradiction at the heart of this system. In the particular case of Germany,
the health policy governing system: governments for well over a century, since the great German
simultaneously needed to inhibit the medical inspired breakthroughs in laboratory based
technology industries in order to keep the lid on medical research in fields like immunology, the
costs, and to promote this vital part of the econ- German medical technology industries have
omy as part of the management of industrial been world leaders, both in the sphere of phar-
policy. maceuticals and in ‘big ticket’ technologies
The intersections between the different based on innovations in electronics. Second,
domains of health care in the command and the role of the state in these arrangements was
control systems have, we shall see later, shaped potentially problematic: it was a ‘steerer’ of the
some key problems in health-care policy. system, but this left open what steering
Peter-3383-Chapter-12.qxd 6/8/2006 7:35 PM Page 228
amounted to (Döhler, 1995; Burau, 2001.) As In a nutshell: the system excludes millions
we shall see, in recent decades this has proved from health insurance coverage, and those mil-
a critical difficulty for corporatist health-care lions tend to be among the most vulnerable of
systems. Americans. At the same time, in the last gener-
ation, the American system has been disastrous
at cost containment. The wider political
system has proved incapable of producing
ISSUES IN HEALTH-CARE POLICY
systemic reform capable of addressing either of
the issues of coverage and costs. One result of
I have argued throughout that health policy is the latter failure is that policy struggles in the
essentially a ‘constructed’ concept, and that the United States now involve a constant, complex
way we understand it depends heavily on which battle of attrition between providers and third
domain we view it from: from the domain of con- party payers, both in the public sector and in
sumption, from the domain of service delivery, the commercial insurance industry.
or from the domain of technology. This thought The achievements of command and control
also shapes our understanding of issues and systems are well summed up in the British expe-
problems in health-care policy. The simple rience. They have succeeded in integrating
typology of health-care states that informed dis- health care into the domain of the welfare state,
cussion in the preceding section amounts to a and thus of extending free, or nearly free, care to
kind of classification of ‘families’ of health-care whole populations as a right, or virtual right, of
states: the supply, the command and control citizenship. In the British case the system also
and the corporatist. Health-care families, like proved highly successful at the containment of
human families, tend to be functional in their global costs. But these achievements depended
own special ways, and dysfunctional in their fundamentally on systems of administrative
own special ways. To put it slightly more for- rationing. Virtually all the command and con-
mally: there is no unique set of health-care trol systems have tried, with varying success, to
issues ‘out there’. Each state formation has great replace administrative rationing with some
achievements, and each has weaknesses, and the systems that at least mimic markets in recent
achievements and weaknesses are inscribed in decades. The problems of administrative
their natures. rationing have been threefold. First, the fact that
The American supply state has huge the third domain of health policy – the technol-
achievements to its credit. For the section of ogy innovation system – has been external to
the population that has comprehensive health the command system has meant that a constant
insurance – and that amounts to the majority stream of innovations has flooded onto mar-
of Americans – it delivers care that is highly kets, widening the range and effectiveness of
sensitive to the demands of payers, and that health care, and widening also the cost base
makes accessible the very latest innovations of systems. Second, the actual administration of
from a hugely creative system of high technol- rationing has depended heavily on the medical
ogy innovation. Its problems are also well profession acting as ‘gatekeepers’ – and doctors’
documented. As we have already seen, there is willingness in turn to perform this role has
one task beyond the system which has been depended on a fragile political compact with the
accomplished by a wide range of much less state. Finally, command and control in health
affluent nations: universal coverage. A large care has had to survive in an increasingly alien
and growing number of Americans have no, or environment: one where consumption is organ-
inadequate, health insurance cover, and ised predominantly according to demands
because the most effective systems of health expressed through markets, rather than through
insurance are occupationally based, this also queues. The political sensitivity of health-care
largely means Americans who are already either queuing – expressed, for example, in the high
excluded from work or are in sections of the political salience of health-care waiting lists – is
labour market where employment is precarious. a sign of this difficulty.
Peter-3383-Chapter-12.qxd 6/8/2006 7:35 PM Page 229
Ferrera, M. (1996). ‘The “southern model” of welfare for the future. Durham, NC.: Duke University
in Europe’, Journal of European Social Policy, Press.
6: 17–37. OECD (various dates). Health Data, cd rom. OECD:
Ferrera, M. and Rhodes, M. (eds) (2000). Recasting Paris.
European Welfare States. London: Cass. Pickstone, J. (ed.) (1992). Medical innovations in
Foote, S. (1992). Managing the medical arms race: historical perspective. London: Macmillan.
public policy and medical device innovation. Reiser, S. (1978). Medicine and the reign of technol-
Berkeley: University of California Press. ogy. Cambridge: Cambridge University Press.
Freeman, R. 2000. The politics of Health in Europe. Rothman, D. (1994). ‘A century of failure: class bar-
Manchester: Manchester University Press. riers to reform’, in J. Morone and G. Belkin, eds.,
Freidson, E. 1970. Profession of Medicine: a study in The politics of health care reform: lessons from the
the sociology of applied knowledge. New York: past, prospects for the future. Durham, NC: Duke
Harper and Row. University Press, pp. 11–25.
Guillén, A. and Cabada, L. (1997). ‘Towards a Saltman, R. and von Otter, C. (1995). Planned markets
National Health Service in Spain: the search for and public competition: strategic reform in Northern
equity and efficiency’. Journal of European Social European Health Systems. Buckingham: Open
Policy, 7: 319–36. University Press.
Jacobs, L. (1995). ‘Politics of America’s supply state: Skocpol, T. (1996). Boomerang: Clinton’s health secu-
health reform and technology.’ Health Affairs, rity effort. New York: Norton.
14:2, 143–57. Starr, P. (1982). The social transformation of American
Johnson, T. (1972). Professions and power. London: medicine. New York: Basic Books.
Macmillan. Temin, P. (1980). Taking your medicine: drug regulation
Kelman, S. (1975). ‘The social nature of the defini- in the United States. Cambridge, Mass.: Harvard
tion problem in health.’ International Journal of University Press.
Health Services, 5:4, 625–42. Trajtenberg, M. (1990). Economic analysis and product
Klein, R. (1990). ‘The state and the profession: the innovation: the case of CT scanners. Cambridge,
politics of the double bed’. British Medical MA.: Harvard University Press.
Journal, 301: 700–02. Waddington, I. (1984). The medical profession and
Levit, K., Cowan, C., Lazenby, H., McDonnell, P., the Industrial Revolution. Dublin: Gill and
Sensenig, A., Stiller, J., Won, D. (1994). ‘National Macmillan.
health spending trends, 1960–93’, Health Affairs, Webster, C. (1988). The health services since the war.
13:5, 14–31. Vvolume 1, problems of health-care: the National
Moran, M. (1999). Governing the Health Care State: Health Service before 1957. London: HMSO.
a comparative study of the United Kingdom, Weisbrod, B. (1991). ‘The health care quadrilemma:
the United States and Germany. Manchester: an essay on technological change, insurance,
Manchester University Press. quality of care, and cost containment’. Journal of
Moran, M. and Wood, B. (1996). ‘The globalisation Economic Literature, 29:1, 523–52.
of health care policy’, in P. Gummett, ed., Wildavsky, A. (1977). ‘Doing better and feeling
Globalisation and public policy. Cheltenham: worse: the political pathology of health policy’
Elgar. Daedalus, 106:1, 1–5–23.
Morone, J. and Belkin, G. eds., (1994), The politics of Wohl, S. (1984). The medical industrial complex.
health care reform: lessons from the past, prospects New York: Harmony Books.
Peter-3383-Chapter-13.qxd 6/8/2006 7:37 PM Page 231
13
Education Policy
SUSAN MARTON
institutional structure of the polity and the THE IMPORTANCE AND UNIQUENESS
welfare system itself shapes the domestic policy OF EDUCATION POLICY
response to globalization. In order to study
such responses, one must focus on the impor- John Dewey’s declaration from 1958 reminds us
tance of institutions as organizational struc- that education is “essentially about the develop-
tures and as the embodiment of normative ment of democratic communities in which
orientations. everyone can feel free and capable of participat-
Recent policy studies in education restruc- ing” (Taylor et al., 1997:19). Education obvi-
turing also reflect this dichotomy between ously has an individual role (increasing one’s
globalization as primarily constraining versus knowledge levels) but education also has a soci-
globalization as enabling. It is beyond the etal role where schools function as a mechanism
scope of this chapter to address this debate in to transfer ‘values into future citizens’ and to
length, but let it suffice to say that demands for socialize them in preparation for their future
restructuring state education in order to meet voting rights (Heidenheimer et al., 1983). It is
these globalization pressures are evident across the nature of this societal role that is often mis-
the industrialized, and even the developing understood in education policy. Citing
nations, albeit with different timing patterns. Bernstein (1971), Whitty emphasizes that,
The advanced industrial nations went through “How a society selects, classifies, distributes,
restructurings primarily in the 1980s, whereas transmits and evaluates the educational knowl-
the Asian and Arab countries changed more in edge it considers to be public, reflects both the
the 1990s. A new language has appeared in the distribution of power and the principles of
sphere of education policy across the globe, social contract” (2002: 9). Educational reforms
incorporating terms such as: excellence, com- depict widespread social changes, and are there-
petitiveness, efficiency, accountability and fore not just a matter of pedagogy or didactics,
devolution. Nonetheless, it is important to but rather social reforms that should be ana-
point out that “the considerable convergence at lyzed as such (Husén, 1998; Kogan, 1973).
the policy rhetoric and general policy objec- Certainly, one of the most significant social
tives may not satisfactorily explain the compli- reforms has been the transformation of school-
cated processes of changes and the dynamic ing as a private activity for the few (often reli-
interactions between global-regional-local giously organized and charity-based), to that of
forces that shape education policy-making in education as a right to be provided based on
individual countries” (Mok & Welch, 2003: 25; public expense for the masses. As the modern
see also Dale, 1999; and Green, 1999). nation-state developed, two goals for the educa-
Educational restructuring may not only be tion system were delineated: first, to create a com-
a response to global competition and human mon culture and, second, to educate the future
capital needs, but also to a state’s particular generations of civil servants (Heidenheimer et al.,
financial situation and management capa- 1983). A common culture was seen as a necessity
bilities, as well as communitarian and post- in promoting social solidarity. The state was to
modernist demands (Daun, 2002). The forms treat its citizens based on ability; not on class,
that these restructuring demands acquire are gender or race. The notion of equality of oppor-
not unilateral, but rather are embedded in the tunity was an important component of this com-
specific political circumstances and structures mon culture, with the bureaucracy playing a key
(such as the relationship between church and role in promoting and implementing the idea of
state), and the cultures and histories of the meritocracy. Given the second goal of an edu-
particular nations. Thus, there is no “mono- cated administrative corp, the state could ensure
lithic strategy” at the state level, but rather a its ability to steer effectively and to attract the
complex interweaving of internal and external loyalty of its citizens (Castles, 1998).
dynamics that contribute to the specific con- As the role of the state in providing educa-
text for restructuring education policies (Mok tion grew, education as a national policy con-
& Welch, 2003). cern augmented in importance. Policy issues
Peter-3383-Chapter-13.qxd 6/8/2006 7:37 PM Page 233
such as the extent of professional autonomy principles of equality and meritocracy. Mok &
versus bureaucratic control, and the proper Welch (2003) argue that Weber’s classic ideal
form of financing (local versus state-wide) type of bureaucracy is being deconstructed as
have characterized the uniqueness of this control of the public service is being handed
policy area since the 1950s. The role of the over to non-state actors. The nature of govern-
teaching profession in carrying out the goals of ment has changed from service providers to
public education has become even more com- regulators – with the regulation of quality as a
plex given the advance of globalization. In high priority. Politicians, in the name of decen-
comparison to other policy areas, the study of tralization and choice policies, have delegated
education policy includes the need to review management of schools to parental and/or
the profession’s claims for autonomy in the community control. One can ponder to what
implementing process. As Hargreaves has extent this impacts upon the role of the bureau-
pointed out, there is a difference between cratic experts – can they maintain their influ-
restructuring as bureaucratic control “where ence in the face of what may be unreasonable
teachers are controlled and regulated to imple- state/central policies?
ment the mandate of others” versus restructur- Returning to John Dewey’s statement, we
ing based on professional empowerment need to ask how education policies are con-
“where teachers are supported, encouraged, tributing to the development of democratic
and provided with newly structured opportu- communities. The democratic view of educa-
nities to make improvements on their own” tion emphasizes the “collective, common and
(cited in Brown et al., 1997: 23). Today it community purposes, and it recognizes that
appears that some restructuring policies the political values of freedom and equality are
threaten teachers’ collective professional interdependent rather than antithetical” (Reid,
autonomy and responsibility. One can also 2002: 573). Castles (1998) argues that it is
question whether rewards based on individual doubtful whether the state will retreat from its
performance, merit pay and school perfor- responsibility for public education given that
mance are legitimate when student achieve- the state is not willing to surrender its role in
ment is the result of collective efforts over the development of human capital and politi-
many years, and can not be seen as the product cal socialization. However, in analyzing the
of one teacher (see Torrance, 1997). recent educational reforms in New Zealand,
Furthermore, new forms of decentralization Peters (2003) describes a situation where edu-
threaten teacher autonomy (Brown et al., cation as a social good has disappeared,
1997). New financing systems have ended the replaced by an ideology where education is an
previous system of zoning principles and legis- economic good for which one must pay. Reid
lated “catchment areas”, where teachers had a (2002) reports similar findings for the case of
“guaranteed” supply of students to teach, and a Australia, where the collective community
“guaranteed” job regardless of their classroom purposes of public education are at risk in the
performance. These new decentralized financ- face of neo-liberal ideas of individualism and
ing systems, along with decentralized employ- competition. Reid argues that the classic view
ment arrangements, can be used as a way to of the public good must be revised to meet the
weaken the power of organized teacher trade challenges related to globalization. We will
unions on the central level. National, or state- return to Reid’s suggestion for a deeper prob-
wide, employment conditions may no longer lematization of the concept of the “public
be relevant with decentralization of employ- good” at the end of this chapter.
ment policies down to the local school level. Using state theory, Carnoy (1992: 159) has
A further aspect requiring consideration in outlined the two extreme positions for the
education policy is the relationship between the state regarding its role in providing education.
bureaucrat and the politician. Today the role of The first position argues that the state is inde-
the education bureaucrat is being challenged, pendent from civil society, and thus the state
given an uneasiness regarding the Weberian can be seen as “interfering” in the workings of
Peter-3383-Chapter-13.qxd 6/8/2006 7:37 PM Page 234
a “social welfare maximizing free market”. The citizenship. In the final section of this chapter,
second position views the state as reflecting possible solutions to bridge this gap will be
and reinforcing “the power relations that reviewed.
derive from those free market economic rela-
tions”. Using an institutionalist view of the
state for the second position, the state can
RESTRUCTURING STATE EDUCATION –
potentially act as a “good cop”, using democra-
GENERAL TRENDS AFTER 1950
tically represented interests to ensure equal
opportunities and outcomes.
Between these two extreme positions, one The post-war period can be described as an
can place various ideologies regarding educa- “era of optimism” in terms of the ability of the
tion. Focusing first on the neo-liberal ideology, state to implement social reform. Provision of
there is an emphasis on individual motivation, and access to education expanded in the 1960s
competition, decentralization, fiscal restraint, and 1970s in all industrialized countries as part
freedom of choice, and private consumption of the state-driven modernization process.
(Griffin, 2002; Wells, 2002). Forcing public National education policy at that time was “an
sector institutions to act more like private sector important site of struggle over class formation
businesses with the initiation of “quasi- and social distribution” within the nation-state
markets”, the neo-liberals believe that quality is (Marginson, 1999: 27). Human capital theory
produced through competition (Lauder, 1997). combined with pressures for the democratiza-
In this light, families should be encouraged to tion of the education system in order to spur
see choices regarding schooling in a similar way the expansion of the system. Political debates
as decisions regarding private consumption. It is focused on spending levels, types of facilities
argued by the neo-liberals that the time has and structural barriers to equality, but the
come for consumer interests to lead the way, guiding principle was “collective provision
and that “producer-capture” in education by elected bodies with a mandate to cater for
should end (Brown et al., 1997; Demaine, 1999). the needs of the whole population” (Whitty,
More reliance on market forces, voluntary ser- 2002: 85).
vices and individual demands are seen as the Yet, in the 1980s, doubts about the states’ abil-
way to achieve social goals. ity to maintain welfare provisions, in combina-
On the other side of the ideological spec- tion with challenges to Keynesian economic
trum, the social-democratic ideologies regard policy, have led to a restructuring of the state.
these neo-liberal policies as privileging individ- The provision of state services is now more
ual rights at the expense of social justice and do dependent on individual behavior and institu-
not contribute to a more equal society, but tions of civil society than previously. New steer-
rather to one that is more “acceptably unequal” ing mechanisms are in place, such as new
(Whitty, 2002 citing Connell, 1993). Social jus- regulations, incentive programs and sanctions
tice concerns have unfortunately been put to for “autonomous individuals and autonomous
the side, the social democrats would argue, quasi-governmental and non-governmental
since there is no longer any aspiration towards institutions”, which Marginson refers to as
equality of outcome between social classes or “steering from a distance” (1999: 25); a concept
ethnic groups. Yet the counter argument from developed in the work of Kickert (1995) on gov-
the neo-liberals would be that diversity in ernance in the Netherlands.
the provision of schools is actually a way of In the old system, the publicly managed
opening a democratic society, so that it can be education system provided for a scientific
more responsive to the needs of special com- form of competition with curricula, pedagogy
munities with multiple identities. It is believed and modes of assessment that were “mono-
that such diversity is more attractive to these com- cultural in process and culturally homoge-
munities than the previously one-dimensional nizing in effect” (Marginson, 1999: 28). Today,
notions of comprehensive schooling and contemporary governments emphasize a
Peter-3383-Chapter-13.qxd 6/8/2006 7:37 PM Page 235
welfare-ethic based on “self-responsibility” and Kogan (1993). This will allow for some degree
the “work-ethic”, and the teacher-student rela- of synthesis as well as further specification.
tionship is more individualized (Tomlinson, These models will later be used to analyze the
2001). Diversity, both ethnically and reli- case studies of England and Sweden in the later
giously, is now one of the major policy chal- section of this chapter “Policy change in
lenges, with questions of access focusing on the National Contexts”.
needs of various cultural groups and the cul- First, in the “competitive market model”, the
tural dimensions of curricula and achieve- school is viewed primarily as a business operat-
ment. The relationship between the state and ing in a commercial market place with a high
its political constituencies (including teacher degree of autonomy from government struc-
organizations, school boards and parental tures. The school competes for both pupils and
groups) is further complicated by these “poli- funding within a “competitive area” with nearby
tics of difference and shifting identities” schools. Cordingley and Kogan referred to this
(Taylor et al., 1997: 80). as the “individualistic model” where the indi-
vidual client acts as purchaser of a product.
In the second model, that of “school
empowerment”, the focus instead is on the
FROM GOVERNMENT TO GOVERNANCE
school itself and how freedom and choice,
IN EDUCATION POLICY
often along with managerial ideas on allocat-
ing decision making to the shop-floor, can
The new forms of governance in education empower the school. Ideas regarding participa-
policy are putting the following issues at the tion are also different from the “competitive
forefront of policy debates: centralization versus market” model due to the emphasis here on
decentralization, public versus private, auton- ideas of partnership and community.
omy versus accountability, and equality versus Model three, the “local empowerment”
quality. Before reviewing these issues in more model, is quite different from the “school
detail, the work of Dale (1997) will help in clar- empowerment” model, given that the focus is
ifying how governance in education policy can instead on the devolution of previously central
be understood. First, we can identify three sepa- government powers to a local authority, rather
rate governance activities regarding education: than to the schools. The school is envisioned as
funding, provision and regulation. Second, we part of a local educational system, where there
can identify three coordinating institutions are reciprocal rights and obligations to be car-
which may carry out these activities: state, ried out. The distinguishing point of this
market, and/or community. Public education model is that more power is given to the local
exists when all three of the activities are pro- authorities than to the schools. Yet, in the work
vided by the state. Dale argues that we should be of Cordingley and Kogan, the school empow-
more aware of the complexities involved in the erment and local empowerment models may
shift from state control to other governance be mixed, allowing for a “professional-electoral
forms and that the distinction between “public” model”, where power is to be shared between
and “private” is sometimes inadequate. the principles and teachers in the school and a
Carrying the discussion further, various locally based intermediary body. In addition,
models of education governance have been the role of government in relating to the “indi-
outlined in the literature. A recent OECD vidual school boards” may vary, with one type
report presented four ideal-type models based on a stronger role for the central govern-
for education governance: (1) “competitive ment in relating to the school board, whereas
market”, (2) “school empowerment”, (3) “local in another type, there is more anchorage in the
empowerment”, and (4) “quality control” local community through a mixture of locally
(OECD, 2003). Below we will combine the dis- appointed and elected members.
cussion of the models outlined in the OECD In the fourth model, that of “quality con-
report with those reported in Cordingley and trol”, the government aims for control over
Peter-3383-Chapter-13.qxd 6/8/2006 7:37 PM Page 236
both school processes and products, relying accountability through their responsibilities to
heavily on bureaucratic procedures of rules stakeholders, and here one could say that the
and requirements. This is in line with the “local empowerment” model is very responsive,
Cordingley and Kogan model of “funding like a “community forum”, whereas the “school
councils” which would provide funds to insti- empowerment” model has dual accountability
tutions (i.e. schools and/or school districts) to both professionals and non-professionals.
that meet their contractual demands for quality The “quality control” model also represents a
levels, costs and volume. contractual type of accountability, which can be
The four main models outlined above will characterized as “hierarchical”, since a higher-
naturally imply different usages for concepts level power decides over the definitions in the
such as school autonomy and accountability. contract.
Starting with school autonomy, one can ask: Working from these various governance
“Autonomy for whom, and autonomy over models, efforts have been made to study the
what?”As Bullock and Thomas (1997) point out, linkages between the models and school effi-
autonomy is a relative concept, depending on if ciency and student learning, but the relation-
it is viewed from the position of the individual ship certainly is complex and is not automatic
learner, the educator or the institution (cited in (Cohen & Spillane, 1992). Positive school reform
OECD, 2003). Viewing autonomy from the has been found in models where the leaders
point of the institution to the higher authority and teachers involved are empowered and share
on a continuum ranging from “full autonomy; decision making through a supportive structure
decisions made after consultation with another and where the professional community has a
authority at an adjoining level; and those made shared normative base (OECD 2003, citing
within guidelines set by another authority, gen- Silins and Mulford, 2002; Goddard et al., 2000;
erally at the top”, the OECD report related the and Heck, 2000). But others have reported that
governance models to the types of autonomy. many managers are following the principle of
The “competitive market” model depicted “sub- “do things right rather than doing the right
stantial” autonomy; the “school empowerment” thing” (Sizer, 1984 cited in OECD, 2003: 78).
model entailed “devolved” autonomy since the Regarding school-based management, both
school is still part of an overall system; the “local Hannaway (1993) and Elmore (1993) have
empowerment” model emphasized “consulta- found little effect on student achievement
tive” autonomy since the school is a member of (cited in Daun, 2002).
a network of institutions; and the “quality con-
trol” model described “guided” autonomy since
there is a pronounced role for an authority at a Decentralization, Multi-Level
higher level. Governance and Democracy
Using terminology developed from Halstead
(1994), the OECD 2003 study also evaluated As the section above described, there are various
accountability in the models based on “con- models for governing education policy, and the
tractual” versus “responsive” accountability. ramifications of some of these on decentraliza-
Contractual accountability deals with fulfilling tion and democracy will now be discussed.
the expectations of particular audiences by McGinn (1992: 163) poses two major questions
providing them with standards and outcomes. regarding the effect of decentralization policies
Responsive accountability is based on decision on democracy: “Do they reflect a genuine shift
making by educators, after listening to the in power toward those who believe in popular
desires of relevant stakeholders. Clearly the democracy? Would decentralization improve
“competitive market” model relates to a contrac- education in an equitable fashion?” Three
tual, “consumerist” view of accountability, since types of decentralization are outlined: decon-
this is analogous to commercial behavior. The centration, decentralization and devolution.
“school empowerment” and the “local educa- Deconcentration entails a territorial sense of
tion” models are both demonstrating responsive decentralization, with government authority
Peter-3383-Chapter-13.qxd 6/8/2006 7:37 PM Page 237
shifting from central level to local level by weakening the voice of members of society who
moving the offices of government officials. may have a disproportionate share of resources,
Decentralization involves the “delegation of both economically and socially. Privatization,
authority for the control and governance however, is a very complex concept, and “is not
aspects of education to another organization, necessarily driven by market forces nor does it
generally at a lower level in the hierarchy of size” necessarily follow market principles and prac-
whereas devolution means “the more central tices” (Daun, 2002: 95). Many different factors
authority turns over all responsibility, including may be involved in the usage of the term
that of funding, to a more local organization” “private school”, with various “combinations
(McGinn, 1992: 164). of ownership, governance, management, degree
Karlsen (2000), in his study of Norway and of subsidization, type of control and type of
British Columbia, introduces the concept of students” (Daun, 2002: 95; see also Boyd and
“decentralized centralism”. This term allows Cibulka, 1989). Nonetheless, as diversity is
him to discuss the dynamic interactions in the encouraged by market solutions, the role of the
decentralization process when a strategy of education system to form shared political values
bottom-up innovation is present. He found that may put collective action at risk.
decentralization of tasks and administration The debate on these issues is certainly not
responsibility did not necessarily mean a shift of settled. Moe (1994) and Pollard (1995) argue
power from a higher to a lower level and that the instead that the introduction of quasi-markets
decentralization of authority could have a legit- has given deprived children a chance for better
imating role for the central level. Bache (2003) educational opportunities. Tooley (1995)
found similar results in his analysis of the local argues that the positive effects of deregulation
education authorities in England. Although and school choice have not yet fully material-
much of the deregulation seemed to be in line ized because the policies have not gone far
with New Public Management ideas, the decen- enough. Yet Henig discusses the “sad irony of
tralizing could be viewed as part of the state’s the current education/reform movement” in
strategy to retain policy control (Pierre, 2000). that the heavy emphasis on market-based ideas
Using a multi-level governance and policy net- overshadows “the healthy impulse to consider
works perspective, Bache found that the state radical reforms to address social problems”
retained its power to control policy and to and instead may bring about “initiatives that
achieve its policy goals more effectively. This further erode the potential for collective delib-
leaves us, however, with the important question eration and collective response” (Henig, 1994:
of the role of local authorities as institutions of 222, cited in Whitty, 2002: 21). Given recent
local democratic accountability when such changes in the governance of education, Dale
central-level control is exercised. argues that Western governments must face
McGinn (1992) highlights two forms of two major issues in the future: first, “the ques-
decentralization that reduce the vitality of tion of how equality of resources and out-
mechanisms for democratic participation in the comes can be achieved under a complex
education system; professionalization and pri- system of governance in which particularism
vatization. First, professionalization refers to the rather than universalism is an important guid-
professional management of schools by educa- ing factor in the provision of education” and
tion administration specialists. There is a risk second, “the question of how effective democ-
that this type of development could alienate ratic accountability can be introduced into the
participation by parents, or even by teacher system” (Dale, 1997: 281).
groups, if they are challenging the “expert” solu-
tions put down by the professionals. Second,
SOCIAL JUSTICE ISSUES
privatization can also limit possibilities for
democratic participation by reducing the input
from existing political organizations that were The predominant role of government after the
created to express the popular will and by post-war period had been to ensure access to
Peter-3383-Chapter-13.qxd 6/8/2006 7:37 PM Page 238
education – the expansion of the school system rights to education over citizens’ rights, a vacuum
and compulsory education for all were key exists as to who has responsibility for the
components of this policy. When these levels of whole. Important issues regarding the need to
participation were achieved, it was believed that build social capital versus the need to meet the
governments had accomplished as much as was demands for diversity come to the forefront,
in their power. During the 1960s this view was especially in terms of faith-based schools that
challenged by broad-based popular movements may be culturally and religiously exclusive.
such as the civil rights movement in the US Relating the education system to the societal
and the feminist movement that was gaining sphere can bring about contradictory demands
strength world-wide. The role of government in regards to culture and linguistic sensitivity.
was increasingly understood to include more Research on the topic of school choice and
than just access policies, and thus the causes of deregulation discusses the risks of horizontal
educational disadvantage came to the forefront diversity in schooling, meaning that no greater
of policy discussions (Taylor et al., 1997). diversity in the provision of educational ser-
Four main approaches to improving social vices is achieved. Instead, the previously exist-
justice have been tried in the past, and some ing hierarchy of schools can be strengthened
continue to be used today. These include: (1) based on academic test results and social class
meritocracy, the idea that social advancement (Glatter et al., 1997). Difficulties arise in the
will take place based on individual merits and way schools are judged, as evaluations may be
an educational system that is able to recognize based on social grounds or narrow academic
such merits; (2) compensatory mechanisms, criteria that ignore the overall performance of
such as free meals in school for low-income the school. Furthermore, the funding system
families; (3) specific intervention programs, can make it extremely difficult for schools in
such as the Head Start pre-school programs disadvantaged neighborhoods to break out of
for children from disadvantaged families; and the cycle of decline (Goldhaber, 1999).
(4) school improvement projects, which allow Certainly, when discussing equity in educa-
the teachers, staff and school administrators to tion systems, the gender aspect is also relevant.
decide how to improve their institutions Most studies of gender politics and school
(Whitty, 2002: 108–109). However, although reform concentrate on (a) the gendered nature
there are individual success stories (Maden of teaching as a profession, (b) the divergence
and Hillman, 1996), the problem remains that of career opportunities among men and women
the long-term “patterning of educational teachers, and (c) the “counter-hegemonic, anti-
inequality has been strikingly consistent sexist efforts of feminist teachers” (Datnow,
throughout the history of public education” 2000: 133). In addition, studies on parental
(Whitty, 2002: 110; see also Wilson, 1997 and choice and market forces in education (David,
Ogbu, 1997). 1999) are focusing on the role of families, and
It may, however, be necessary to admit that in particular mothers, in a child’s schooling
the demands placed on the educational system decision (an issue which is increasingly impor-
at times have been much broader than the tant as politicians place more responsibility on
system’s capacity to handle such problems. The the family for schooling).
negative correlation between most forms of
social disadvantage and school achievement is
a multi-faceted problem that is not easily fixed.
POLICY CHANGE IN NATIONAL CONTEXTS
Social disadvantage impacts on schooling both
directly and indirectly, for example through
poor health conditions and housing arrange- In the brief review below of national policy
ments. As Halsey pointed out over twenty years changes in England and Sweden, the intercon-
ago, “the teacher cannot reconstruct the com- nectedness of governing models and ideologies
munity unaided” (cited in Whitty, 2002: 116). of education is evident. However, the cases
But, as policies are encouraging individuals’ represent policy movements in nearly opposite
Peter-3383-Chapter-13.qxd 6/8/2006 7:37 PM Page 239
directions, raising interesting questions about policy gave individual schools much more
the significance and consequences of decen- power over budgets and daily management
tralization and centralization. The English case issues. Funds were distributed based on the
represents a movement from a highly local, number of students a school could attract,
decentralized system to one where significant which helped to create a “quasi-market” for
powers over provision and regulation have education. Sexton (1987) refers to this as the
been centralized. The Swedish case on the “virtual voucher” system (cited in Whitty,
other hand reflects a transformation of a 2002). Thus parents were defined as the main
highly centralized system, where education was consumers, the customers of education, and
a core element of the Swedish welfare state, to seen as most able to decide what is best for
a largely decentralized, fragmented system. their children.
Ideologically, the movement in England to
allow for local school empowerment was not
England only a right-wing movement, but the new left
was also attracted by the potential of the LMS
The provision of education prior to the 1988 for turning around negative bureaucratic
Education Reform Act was primarily con- effects and poor professional practice. It was
ducted in state schools under the political and believed that this framework also could allow
bureaucratic control of a local education for social justice and increased stakeholding –
authority (LEA) which was democratically with new co-operative efforts for the local and
elected. When the Conservatives came to national agencies (Demaine, 1999).
power, the LEA represented a monopoly over In addition to the focus on the management
state schooling that was subsequently to be of the schools, the reform policies also heavily
broken through various Education Acts passed emphasized quality and assessment in the
throughout the 1980s and early 1990s. School school system. In a detailed account of these
self-management became a primary concern, assessment policies, James (2000) describes the
with the leading idea that “the proper manage- process from the Task Group of Assessment and
ment of schools is best achieved through part- Testing appointed under the Conservative gov-
nership and the co-operative participation of ernment in 1987, to the latest policy from the
parents, teachers, local politicians and commu- Department for Education and Employment.
nity representatives, within the context of Every year, the Office for Standards in
national policy” (Demaine, 1999: 17). Education (OFSTED) provides national sum-
The marketization of the public schools was mary results and national benchmark data to
largely conducted through two initiatives. The all schools, with detailed accounts of how
first is known as “grant-maintained schools” schools rank against each other. Furthermore,
which allowed existing LEA schools to opt-out the quality of the teaching staff has been under
from the LEA structure if a parental ballot debate, and recently a General Teaching
voted in favor (these were later renamed Council, somewhat like a professional honor
“foundation schools” by the New Labour gov- society, has been arranged to help raise stan-
ernment). Funding was provided directly from dards for teachers.
the central government, thus bypassing the Although their assessment strategies have
LEA. The second initiative was the establish- been similar, James (2000) argues that the
ment of the City Technology Schools, which Conservative and Labour governments origi-
were a new type of secondary school with a nally had different goals with their assessment
focus on science and technology studies for the policies. The Conservative governments under
inner cities. Cooperation was to be attained Thatcher and Major were primarily concerned
from business sponsors and the school would in fostering economic growth, with increased
be run by independent trusts. For those competition in the “market of education” con-
schools that remained under LEA provision, tributing to this. The Labour government was
the Local Management of Schools (LMS) also concerned with the competitiveness of the
Peter-3383-Chapter-13.qxd 6/8/2006 7:37 PM Page 240
economy, but their emphasis was more on their school to have the status of “free”, with
transforming the old manual-skills labor funding following the pupil to the “free school”
supply to a labor force ready to compete in the from the municipality (like a system of “vouch-
knowledge economy. Issues of broader access ers”) and thus a quasi-market in education was
were important to Labour, but raising the introduced, now enrolling approximately four
quality of education and training was more percent of the student population. (A private
pronounced. Although meritocracy as a way of education company operates a very small
selecting the best and the brightest is still amount of these “free schools” – but the law
actively promoted, James concludes that prohibits the charging of tuition fees.) The
Labour recognizes that all social classes should Conservatives also increased autonomy for the
have an equal opportunity. schools in relation to the national curriculum,
Thus, the Assisted Places Scheme (started making the curriculum more flexible and less-
under the Thatcher government to provide regulated by the national authorities (Daun and
public funding for academically talented but Benincasa, 1998). The state however did impose
economically disadvantaged students to attend some new forms of control, through national
elite private schools) was later cancelled by the tests and a new national grading system, as well
Labour government, which argued that it was as the introduction of a system of quality audits
better to benefit the many and not the few. in 1997.
Instead Labour introduced “Education Action The reforms were put in place by both Social
Zones” (EAZs), based on the idea of business, Democratic and Conservative-coalition parties,
schools and local education authorities and without any major swings in policy in relation
parents working together to modernize to government changes. (The “free-school”
schools in deprived areas. As Whitty (2002) reform remains in some ways ideologically dif-
explains, the policy seemed to entail both a ficult for the Social Democrats to accept, but
private market, managerial feature (with the there has been no successful overturn of the
input from the business leaders) plus a com- policy.) In trying to explain this policy consen-
munity emphasis that equates with collective sus, Lundahl states that the politicians in
responsibility for education. Studies analyzing Sweden (on both the Left and the Right) had
the results of EAZs show so far mixed success the same view on the necessity of change. The
(see Dickson et al., 2001). following three justifications were provided for
the policy changes: (1) it was no longer possi-
ble for the state to steer over education and
Sweden regulate the schools in detail given the pace
and complexity of change in today’s society;
In the late 1980s, the Social Democrats decen- (2) the variance between regional needs and
tralized the state controlled education system, municipalities’ needs was too broad to handle
and transferred significant powers to the munic- with detailed state steering; and (3) schools
ipalities. Although the overall funding responsi- had lost their authority over the years as the
bility remained with the central state, the usage parents’ own level of education rose, thus mak-
of funds was significantly more flexible for the ing it necessary for the schools and teachers “to
municipalities after the reform. The policy was earn their legitimacy” (Lundahl, 2002: 691).
now one of “lump-sum funding” provided Recent Social Democratic reforms have
directly to the municipalities rather than the included the adding of a third year of education
specified, earmarked funding of the past. When to the vocational training programs, thus
the Conservative-coalition government came to encouraging a stronger academic emphasis. The
power in 1991, they introduced subsidies for main purpose of this reform was to increase
what were to be called the “free schools” – basi- individual flexibility and broaden students’
cally a similar idea to that of the “charter” possibilities to enter the university from the
schools in the US. Any group could apply for upper-secondary education programs. These
Peter-3383-Chapter-13.qxd 6/8/2006 7:37 PM Page 241
reforms have been motivated by both a goal of significant, hierarchical follow-up of quality
broader access to higher education as well as through control and assessment programs, and
economic concerns regarding the changing con- the contractual accountability which the system
ditions in the labor market. However, Lundahl entails, one can also find an explanation based on
questions to what extent equality was a goal in normative orientations of the English polity –
the reforms of the 1990s and concludes that, for a strong public discourse regarding the failure
Swedish Social Democrats, “the belief that edu- of the schools and an acute awareness for the
cation could play a significant role in reducing need for increased competition to solve the
class and other social differences had become problem. Thus a strong, central-level authority
seriously weakened” (Lundahl, 2002: 696). for quality control could be established where
one previously did not exist.
In the Swedish case, we see similar “global”
Comparing Sweden and England: trends as those in the English case, yet rather
Governance Models and Institutional different governance models are adopted
Explanations because of different institutional conditions.
It is important to note that the “school-
Although we see similar “global” trends in the empowerment” model was not adopted in
two cases above, in order to understand what Sweden. Instead, local empowerment in the
shapes the domestic policy response and choice Swedish case meant power to the local politi-
of governance models in each country, we need cians, not to the school boards, or to the pro-
to look at institutions as both organizational fessionals, or to the parents. And even if we see
structures and as the embodiment of normative some elements of the “competitive market”
orientations. model in the Swedish case (given the policy
In the English case, we find elements of both a initiatives of “vouchers” and the “free”
“school-empowerment” model and a “competi- schools), it is important to note that these
tive market” model. The “school-empowerment” schools are not able to “opt-out” from local
model is evident in such policies as the government control as in the English case. By
Education Action Zones and the City studying historical institutional structures, this
Technology Schools, emphasizing both the role difference in the Swedish case can be more thor-
of the school in its community along with oughly explained. For example, in Sweden there
the role of businesses to participate, even at the continues to be a strong Social Democratic con-
funding level. With these policies, the role of the sensus, which combines with strong teacher
local community is very important, yet the role trade unions to prevent parental influence
of the local politicians seems to be sidelined. (Lindbom, 1995). In addition, the recent estab-
Such a policy emphasis may very well be lishment of parental advisory boards has not
explained by a desired break in the institutional been enacted in the name of increasing
legacy of the LEAs (a very favorable develop- parental influence over school policy, but
ment in the view of central-level policy-makers rather in the name of furthering democratic
after the many years of undesirable local politi- education and participation (Jarl, 2004). And
cal control). There is also an explanation to be although Sweden has adopted some aspects of
found in the normative orientations, which in the “competitive market” model, only very
the English case can be said to reflect a belief in limited business interests are allowed. This can
the ability of business in helping to solve societal be partly explained by the dominant norma-
problems. The Thatcher ideological legacy is tive orientations that are quite critical to busi-
also found in the policies that relate to the ness interests in the provision of public goods
“competitive market” model; such as the run- (seen also in the cases of health care and higher
ning of schools by independent trusts (some- education provision). Regarding quality control
times in co-operation with business) and the and assessment, the situation is more mixed in
“virtual” voucher system. In reviewing the Sweden than in England. Elements of both
Peter-3383-Chapter-13.qxd 6/8/2006 7:37 PM Page 242
and higher education institutions operate in the question of whether another type of
an open market to provide services to clients “framing” of education policy is needed. What
(thus autonomy is at risk). would happen if “education is conceptualized
Comparing higher education reforms over as part of the cultural rather than the eco-
time and across nations, these models help to nomic domain?” (Taylor et al., 1997: 77). This
depict the changes taking place. Empirically is not to say that the important role of educa-
applied to cases in England, Norway and tion in producing talent for the labor market
Sweden it has been shown that, although all should be neglected, but rather that the more
countries started at a different position in the humanizing effects of education could also
1960s, they all matched the “invisible hand/ come to the forefront of policy discussions
market model” in the 1990s (Kogan et al., 2000). (Apple, 1994; Cox, 1995). Critical questions
In a separate study, Musselin (1999) described about citizenship and identity in multi-
the governance changes in France in the 1990s cultural societies could be addressed.
as one of “negotiated contract”. The state Therefore, we need to contemplate to what
maintained primary control over the system, extent the democratic ideals and notions of the
but the provision of higher education services education system as an important democratic
were “contractualized” which placed the state institution can be upheld. As test scores, cur-
in the role of “consumer”. riculum content, and assessment exercises
Given the increased marketization of higher dominant the scene, some would argue we
education, evaluation as a tool for state moni- have reached the point of an “educational
toring and control was a dominant policy war of all against all” (Hargreaves & Fullan,
reform in the 1980s and 1990s. Neave and 1998). How are we going to maintain a healthy
van Vught (1991) have labelled this as a change democracy if our state school system is not
from “process control” to “product control”. The serving the “public good”? As society becomes
previous bureaucratic and uniform-across- more ethnically and religiously diverse, trans-
system regulations are being replaced by more forming values to future citizens becomes
diverse and discretionary quality audits and more complex.
assessment exercises. In studying evaluation sys- Thus, we should investigate what type of
tems in Europe, Brennan found that “much institutional structures in the education
quality assessment within individual institu- system could enable citizens to engage in civic
tions appears to be unrelated to external quality life. Hargreaves & Fullan (1988) argue that
assessment requirements of national agencies” what is needed are “democratic communities”
(1999: 231). Although no exact explanation for which value participation, equality, inclusive-
this finding is given, references are made to the ness and social justice (citing Mertz & Furman
potential conflicts of interest that arise with (1997)). Whitty (2002) suggests new forms of
evaluation procedures. In a study of the fifty “collective association” which would exist
American states, Nettles and Cole (2001: 216) between the state and the “marketized” civil
also found tension between the actors involved society and would be based on a new concept of
in the assessment exercises, since the states and citizenship, focusing on citizen rights instead
the public institutions did not necessarily “share of consumer rights. Such a form of association
the goal of accountability”. could reassert a collective responsibility for
education policy, but without returning to the
bureaucratic machinery that contributed to
previous shortcomings in the system. Reid
POLICY CHALLENGES FOR THE FUTURE
argues, “there is a need to reclaim the public
sphere and reinvest community discourse with
As discussed previously, globalization pressures notions of social capital. Places are needed
have contributed to the framing of education where people can share and understand differ-
policy in terms of economic competitiveness. ences, and where they can demonstrate a col-
Given this background, it is worthwhile to pose lective concern for all members of society”
Peter-3383-Chapter-13.qxd 6/8/2006 7:37 PM Page 244
(Reid, 2002: 578). But in advocating an diversity and pluralism. Such a reform could
exchange of “collective concerns”, Reid is care- take place if the polity agreed and supported a
ful to point out that this is not the same as a new set of principles for education that would
“search for consensus”. The notion that groups apply to both public and private schools and a
need to share some “common ground” in order funding system that would be dependent on
to talk with one another is not incompatible whether the schools operated according to these
with diversity, but rather is crucial to it. principles. Principles viewed as particularly
Hargreaves and Fullan (1998) believe that important include: (1) a school curriculum
these types of communities can start in the emphasizing skills important for a pluralist
classroom where pupils would share responsi- moral democracy, (2) a democratic school
bility for their own learning and pupils and management system with a pedagogy reflect-
parents would be involved in decision making ing democratic principles, (3) an emphasis
on an array of topics from teaching and learning on diversity, which includes promoting
to assessment of achievement. Whitty (2002) inclusiveness – not based on homogenous com-
and Reid (2002) argue, however, that the asso- munities or notions of assimilation – but based
ciation should be democratically accountable, on a sense of shared community; and (4) the
not just accountable to the school or the class- non-discrimination of the students in accep-
room. Given that the power of market forces tance procedures. Reid admits that further
can benefit those with resources already, a discussion is needed regarding the political
more democratic forum to counterbalance structure to carry out these tasks as well as the
structural inequalities is required. Referring to details for the funding mechanisms; however
Apple and Beane (1996), Whitty concludes the principles are needed in order to begin to
that “if we want students to learn democratic rethink the relationship between democracy
citizenship we need to put in place structures and public education in a global world.
that embody those principles” (2002: 104). In other works discussing the new politics of
Thus a major issue for the future of educa- education, there is a similar effort to identify
tional systems is what type of democratic new institutional principles along with a ques-
structures can be representative and legiti- tioning of whether the traditional “school
mate? Reid urges us to “retheorize the role of system” is the adequate structure for learning
public schools in building and sustaining a today. Halpin, citing the work of Bentley (1998),
pluralist moral democracy” (2002: 581). This is writes “that education should strive, not so
based on the view that a democratic society much for equality of educational opportunity,
needs a democratic education system, and vice but the provision of opportunities for learning
versa. Reid outlines at least four policy impli- that foster autonomy, responsibility and creativ-
cations which can be derived from this retheo- ity” (1999: 358). Halpin argues that this may
rizing of the role of public schools: (1) policy best take place in “contexts where knowledge is
should not be designed at the “top” by a level of actually used and valued” and thus structures
experts, but rather with a range of participants such as “overlapping networks of learning”
at all levels; (2) the notion of the “nation- may be better suited for the task. Such net-
building” role of education should be revised works would include not just the schools, but a
to incorporate the more relevant global ele- “public knowledge infrastructure” based on
ments as well as the local (see also Lödén, “homes, libraries, firms, universities, commu-
2002); (3) the notion of “choice” needs to be nity centers and voluntary organizations as
re-positioned from the emphasis on individual well” (Halpin, 1999:358). New ideas are also
choice to pursue self-interest to a more collec- coming to the table in New Zealand and
tive notion which emphasizes the privilege to England regarding “Community Education
make choices for the future of society; and (4) Forums”. The exact role and institutional pow-
the concept of “public good” needs to be ers such forums will possess is still unclear,
understood in broader terms, allowing for but certainly worthy of further investigation.
Peter-3383-Chapter-13.qxd 6/8/2006 7:37 PM Page 245
However, one issue does seem to have been Brown, P., Halsey, A. H., Lauder, H., & Wells,
resolved: a common school system for all A. (1997). The Transformation of Education and
students will not meet the demands for diver- Society: An Introduction. In A. H. Halsey, H.
sity in today’s modern society. Other arrange- Lauder, P. Brown & A. Wells (eds.), Education:
Culture, Economy and Society. Oxford: Oxford
ments still need to be encouraged.
University.
Brown, P., & Lauder, H. (1997). Education,
Globalization, and Economic Development. In
ACKNOWLEDGMENTS A. H. Halsey, H. Lauder, P. Brown & A. Wells
(eds.), Education, Culture, Economy, Society.
Oxford: Oxford University.
The author would like to acknowledge financial
Bullock, A., & Thomas, H. (1997). Schools at the
support provided for this project from STINT Centre? a study of decentralisation. London:
(The Swedish Foundation for International Routledge.
Cooperation in Research and Higher Education). Carnoy, M. (1992). Education and the State: From
In addition, many thanks are extended to the Adam Smith to Perestroika. In R. F. Arnove (ed.),
Political Economy Research Centre (PERC) at Emergent Issues in Education. Ithaca: State
Sheffield University and the Center for Public University of New York.
Sector Research (CEFOS) at Göteborg University Castles, F. (1998). Comparative Public Policy. Patterns
for providing a generous research environment of Post-war Transformations. Cheltenham: Edward
during the completion of this project. Elgar.
Clark, B. (1983). The Higher Education System:
Academic Organization in Cross-National Persp-
ective. Berkeley: University of California.
REFERENCES Clark, B. (1998). Creating Entrepreneurial Universities.
Organizational Pathways of Transformation.
New York: Pergamon.
Apple, M. (1994). Official Knowledge. New York:
Cohen, D., & Spillane, J. P. (1992). Policy and
Routledge.
Practice: The Relation Between Governance and
Apple, M., & Beane, J. (eds.). (1996). Democratic
Instruction. Review of Educational Research, 18.
Schools. Buckingham: Open University Press.
Connell, R. W. (1993). Schools and Social Justice.
Bache, I. (2003). Governing through Governance:
Toronto: Our Schools/Our Selves Education
Education Policy Control under New Labour.
Foundation.
Political Studies, 51(2), 300–315.
Cordingley, P., & Kogan, M. (1993). In Support of
Becher, T., & Kogan, M. (1992). Process and Structure
Education: Governing the Reformed System.
in Higher Education. London: Routledge.
London: Jessica Kingsley.
Bentley, T. (1998). Learning Beyond the Classroom.
Cox, E. (1995). A Truly Civil Society. Sydney: ABC
London: Routledge.
Books.
Bernstein, B. (1971). On the classification and fram-
Dale, R. (1997). The State and the Governance of
ing of educational knowledge. In M. F. D. Young
Education: An Analysis of the Restructuring of
(ed.), Knowledge and Control. London: Collier
the State-Education Relationship. In A. H. Halsey,
Macmillan.
H. Lauder, P. Brown & A. Wells (eds.), Education:
Boyd, W. L., & Cibulka, J. G. (eds.) (1989). Private
Culture, Economy and Society. Oxford: Oxford
Schools and Public Policy. International Perspectives.
University.
London: Falmer.
Dale, R. (1999). Specifying globalization effects on
Brennan, J. (1999). Evaluation of Higher Education
national policy: a focus on the mechanisms.
in Europe. In M. Henkel & B. Little (eds.),
Journal of Education Policy, 14(1), 1–17.
Changing Relationships between Higher Education
Datnow, A. (2000). Gender Politics in School Reform.
and the State. London: Jessica Kingsley.
In N. Bascia & A. Hargreaves (eds.), The Sharp Edge
Brown, P. (1997). The ‘Third Wave’: Education and
of Educational Change. London: Routledge Falmer.
the Ideology of Parentocracy. In A. H. Halsey,
Daun, H. (ed.) (2002). Educational Restructuring in
H. Lauder, P. Brown & A. Wells (eds.), Education,
the Context of Globalization and National Policy.
Culture, Economy, Society. Oxford: Oxford
New York: Routledge Falmer.
University.
Peter-3383-Chapter-13.qxd 6/8/2006 7:37 PM Page 246
Daun, H., & Benincasa, L. (1998). Restructuring Fulfill the Promise? San Fransisco: The Jossey-Bass
Education in Europe. Four Country Studies. Publications.
Stockholm: Stockholm University, Institute of Hargreaves, A., & Fullan, M. (1998). What’s Worth
International Education. Fighting for in Education? Buckingham: Open
David, M. (1999). Educational Reforms, Gender and University.
Families. In J. Demaine (ed.) Education Policy and Heck, R. (2000). Examining the impact of school
Contemporary Politics. Basingstoke: Palgrave. quality on school outcomes and improvement: A
Demaine, J. (ed.) (1999). Education Policy and value-added approach. Educational Administration
Contemporary Politics. Basingstoke: Palgrave. Quarterly, 36(4), 513–552.
Dickson, M., Halpin, D., Power, S., Telford, D., & Heidenheimer, A. J., Heclo, H., & Adams, C. T.
Whitty, G. (2001). Education Action Zones and (1983). Comparative Public Policy: the politics
Democratic Participation. School Leadership and of social choice in Europe and America (2nd ed.)
Management, 21(2), 169–181. New York: St. Martin’s.
Elmore, R. F. (1993). School Decentralization: Who Held, D., McGrew, A., Goldblatt, D., & Perraton, J.
Gains? Who Looses? In J. Hannaway & M. Carnoy (1999). Global Transformations: Politics, Economics
(eds.), Decentralization and School Improvement. and Culture. Cambridge: Polity.
Can We Fulfill the Promise? San Fransisco: The Henkel, M., & Little, B. (eds.) (1999). Changing
Jossey-Bass Publications. Relationships between Higher Education and the
Etzkowitz, H., & Leydesdorff, L. (eds.) (1997). State. London: Jessica Kingsley.
Universities and the Global Knowledge Economy: A Henig, J. R. (1994). Rethinking School Choice: Limits
Triple Helix of University-Industry-Government of the Market Metaphor. Princeton: Princeton
Relations. London: Pinter. University Press.
Glatter, R., Woods, P. A., & Bagley, C. (eds.) (1997). Hirst, P., & Thompson, G. (1996). Globalisation in
Choice and Diversity in Schooling: Perspectives and Question: The International Economy and the
Prospects. London: Routledge. Possibilities of Governance. Cambridge: Polity.
Goddard, R., Hoy, W., & Hoy, A. (2000). Collective Husén, T. (1998). The Swedish School Reforms:
teacher efficacy: its meaning, measure, and Trends and Issues. In A. Tjeldvoll (ed.), Education
impact on student achievement”. American and the Scandinavian Welfare State in the Year
Educational Research Journal, 37(2), 479–507. 2000. New York: Garland.
Goldhaber, D. D. (1999). School choice: an examina- James, M. (2000). Measured lives: the rise of assess-
tion of the empirical evidence on achievement, ment as the engine of change in English schools.
parental decision-making, and equity. Educational The Curriculum Journal, 11(3), 343–364.
Researcher, 28(9), 16–25. Jarl, M. (2004). En skola i demokrati? Föräldrarna,
Green, A. (1999). Education and globalization in kommunen och dialogen (A school in democracy?
Europe and East Asia: convergent and divergent Parents, municipalities and dialogue). Göteborg:
trends. Journal of Education Policy, 14(1), 55–71. Department of Political Science, Göteborg
Griffin, R. (2002). Education in Transition: Inter- University.
national Perspectives on the Politics and Processes Karlsen, G. E. (2000). Decentralized Centralism:
of Change. Oxford: Symposium Books. Framework for a Better Understanding of
Halpin, D. (1999). Utopian realism and a new poli- Governance in the Field of Education. Journal of
tics of education: developing a critical theory Education Policy, 15(5), 545–538.
without guarantees. Journal of Education Policy, Kickert, W. (1995). Steering at a Distance: A New
14(4), 345–361. Paradigm of Public Governance in Dutch Higher
Halsey, A. H., Lauder, H., Brown, P., & Wells, A. S. Education. Governance, 8(1), 135–157.
(eds.) (1997). Education, Culture, Economy, Kogan, M. (1973). The Politics of Education. London:
Society. Oxford: Oxford University. Penguin.
Halstead, M. (1994). Accountability and values. In Kogan, M., Bauer, M., Bleiklie, I., & Henkel, M. (2000).
D. Scott (ed.), Accountability and Control in Transforming Higher Education : A Comparative
Educational Settings. London: Cassell. Study. London, Philadelphia: J. Kingsley Publishers.
Hannaway, J. (1993). Decentralization in Two Lauder, H. (1997). Education, Democracy and the
School Districts: Challenging the Standard Economy. In A. H. Halsey, H. Lauder, P. Brown &
Paradigm. In J. Hannaway & M. Carnoy (eds.), A. Wells (eds.), Education: Culture, Economy and
Decentralization and School Improvement. Can We Society. Oxford: Oxford University Press.
Peter-3383-Chapter-13.qxd 6/8/2006 7:37 PM Page 247
Lindbom, A. (1995). Medborgarskapet i välfärdsstaten: Nowotny, H., Scott, P., & Gibbons, M. (eds.)
Föräldrainflytande i skandinavisk grundskola (2001). Re-Thinking Science: Knowledge and
(Citizenship in the Welfare State: Parental influence the Pubic in an Age of Uncertainty. Cambridge:
in the Scandinavian compulsory school). Stockholm: Polity.
Almqvist & Wiksell. Ogbu, J. (1997). Racial Stratification and Education
Lundahl, L. (2002). Sweden: decentralization, dereg- in the United States: Why Inequity Persists. In
ulation, quasi-markets – and then what? Journal A. H. Halsey, H. Lauder, P. Brown & A. Wells
of Education Policy, 17(6), 687–697. (eds.), Education: Culture, Economy and Society.
Lödén, H. (2002). Skilda världar? Om nationell och Oxford: Oxford University.
global identitet (Separate worlds? On national Organisation for Economic Co-Operation and
and global identity. Utbildning & Demokrati Development (OECD). (2003). Networks of
(Education & Democracy), 11(1), 37–55. Innovation: Towards New Models for Managing
Maden, M., & Hillman, J. (1996). Lessons in Success. Schools and Systems. Paris: OECD.
In National Commission on Education (ed.), Peters, M. (2003). The New Zealand Education
Success Against the Odds. London: Routledge. Experiment: From Democratic Participation to
Marginson, S. (1999). After Globalization: The Self-Management, and From Universal Welfare
Emerging Politics of Education. Journal of Entitlement of Private Investment. In K. Mok &
Education Policy, 14(1), 19–31. A. Welch (eds.), Globalization and Education
Marton, S. (2000). The Mind of the State: The Politics Restructuring in the Asia Pacific Region. Basingstoke:
of University Autonomy in Sweden 1968–1998. Palgrave.
Göteborg: BAS. Pierre, J. (ed.) (2000). Debating Governance: Authority,
Mertz, C., & Furman, G. (1997). Community Schools: Steering and Democracy. Oxford: Oxford University
Promise & Paradox. New York: Teacher’s College Press.
Press. Pollard, S. (1995). Schools, Selection and the Left.
McGinn, N. (1992). Reforming Educational Gover- London: Social Market Foundation.
nance: Centralization / Decentralization. In R. F. Reid, A. (2002). Public education and democracy: a
Arnove (ed.), Emergent Issues in Education. changing relationship in a globalizing world.
Ithaca: State University of New York Press. Journal of Education Policy, 17(5), 571–585.
Moe, T. (1994). The British battle for choice. In K. L. Sexton, S. (1987). Our Schools – a Radical Policy.
Billingsley (ed.), Voices on Choice: The Education Warlingham: IEA Education Unit.
Reform Debate. San Francisco: Pacific Institute for Silins, H., & Mulford, B. (2002). Leadership and
Public Policy. school results. In K. Leithwood & P. Hallinger
Mok, K., & Tan, J. (2004). Globalization and Market- (eds.), Second International Handbook of Educa-
ization in Education: A Comparative Analysis of tional Leadership and Administration. Dordrecht:
Hong Kong and Singapore. Cheltenham: Edward Kluwer.
Elgar. Sizer, T. (1984). Horace’s Compromise. Boston:
Mok, K., & Welch, A. (eds.) (2003). Globalization Houghton-Mifflin.
and Educational Restructuring in the Asia Pacific Sporn, B. (1999). Adaptive University Structures.
Region. Basingstoke: Palgrave Macmillan. London: Jessica Kingsley Publishers.
Musselin, C. (1999). State/University Relations and Strange, S. (1996). The Retreat of the State: The
How to Change Them: The Case of France and Diffusion of Power in the World Economy.
Germany. In M. Henkel & B. Little (eds.), Cambridge: Cambridge University.
Changing Relationships between Higher Education Swank, D. (2002). Global Capital, Political Institutions,
and the State. London: Jessica Kingsley. and Policy Change in Developed Welfare States.
Neave, G., & van Vught, F. (eds.) (1991). Prometheus Cambridge: Cambridge University.
Bound. The Changing Relationship Between Taylor, S., Rizvi, F., Lingard, B., & Henry, M. (1997).
Government and Higher Education in Western Educational Policy and the Politics of Change.
Europe. Oxford: Pergamon Press. London: Routledge.
Nettles, M., & Cole, J. (2001). A Study in Tension: Tomlinson, S. (2001). Education in a Post-Welfare
State Assessment and Public Colleges and Society. Buckingham: Open University.
Universities. In D. Heller (ed.), The States and Tooley, J. (1995). Markets or democracy? A reply to
Public Higher Education Policy. Baltimore: Johns Stewart Ranson. British Journal of Educational
Hopkins University. Studies, 43(1), 21–34.
Peter-3383-Chapter-13.qxd 6/8/2006 7:37 PM Page 248
Torrance, H. (1997). Assessment, Accountability, Wells, A. S. (2002). Where charter school policy fails: the
and Standards. Using Assessment to Control the problems of accountability and equity. New York:
Reform of Schooling. In A. H. Halsey, H. Lauder, Teachers College Press.
P. Brown & A. Wells (eds.), Education, Culture, Whitty, G. (2002). Making sense of education policy:
Economy, Society. Oxford: Oxford University. studies in the sociology and politics of education.
van Vught, F. (1989). Governmental Strategies and London: Paul Chapman.
Innovation in Higher Education. London: Jessica Wilson, W. (1997). Studying Inner-City Social
Kingsley. Dislocations: The Challenge of Public Agenda
Weiss, L. (ed.) (2003). States in the Global Economy: Research. In A. H. Halsey, H. Lauder, P. Brown &
Bringing Domestic Institutions Back In. Cambridge: A. Wells (eds.), Education: Culture, Economy, and
Cambridge University. Society. Oxford: Oxford University.
Peter-3383-Chapter-14.qxd 6/8/2006 8:06 PM Page 249
14
Environmental Policy
CHRISTOPH KNILL
countries, often beyond the borders of the As a consequence, we have to look for
Western industrialized world. For instance, both alternative explanations. In the literature, a vari-
OECD and Central and Eastern European ety of potential sources of cross-national policy
countries have progressively adopted similar convergence is discussed. Notwithstanding this
policies and administrative structures in the diversity, three analytically distinct factors can
areas of air and water protection, as well as be identified, namely regulatory competition,
waste management from the 1950s onwards international cooperation and transnational
(Kern, Jörgens and Jänicke 2000). However, communication (Hoberg 2001; Bennett 1991;
these developments are not restricted to the Holzinger and Knill 2004). In the following
initial establishment of basic legislation and sections I will investigate the relevance of each
specific institutions in this relatively new of these factors with respect to environmental
policy area. A more recent parallelism refers to policy. The central conclusion which can be
the shift from sectorally fragmented media- drawn from this analysis is that none of the
specific approaches to integrated policies three factors alone, but only the investigation
which seek to tackle environmental problems of their combined effects, provides a satisfac-
from a cross-media perspective (Lenschow tory explanation for the striking parallelism in
1999). Moreover, we observe a general trend national regulatory patterns.
towards the introduction of new policy instru-
ments which emphasize economic incentives,
public participation and information, as well
REGULATORY COMPETITION: THE
as voluntary agreements and private self-
EFFECTS OF INTERNATIONAL
regulation (Héritier 2002; Holzinger, Knill and
ECONOMIC INTEGRATION
Schäfer 2003).1
With regard to these developments it is the
central objective of this article to investigate A first factor that might explain the observed
potential causes for this parallelism in national pattern of increasingly similar environmental
environmental policies. How can the broad con- policy arrangements across countries refers to
vergence of regulatory patterns be explained? To the effects of regulatory competition. With the
be sure, one plausible explanation could be that increasing integration of European and global
governments throughout the world are reacting markets and the abolition of national trade bar-
independently, but similarly to similar environ- riers, there is a certain potential that the inter-
mental problem pressures. As Simmons and national mobility of goods, workers and capital,
Elkins put it: ‘Just as individuals open their puts pressure on the nation states to redesign
umbrellas simultaneously during a rainstorm, domestic market regulations in order to avoid
governments may decide to change their policies regulatory burdens restricting the competitive-
in recession, in the presence of capital flight, or ness of domestic industries (Goodman and
in the presence of high world interest rates’ Pauly 1993; Keohane and Nye 2000).
(2003, 275). Such a constellation implies that As a consequence, national governments com-
there is ‘one best solution’ to a problem that is pete over the optimal design of domestic regula-
common to all countries. This scenario, how- tions in order to attract foreign capital and to
ever, will hardly apply for the comprehensive improve the competitive position of their econ-
area of environmental protection in general, as omy. The presence of mobile capital can induce
different geographical, biological, economic and governments to attract capital from elsewhere by
social conditions will lead to highly varying pres- lowering environmental standards on the one
sures on national policy-makers. Although par- hand, and, on the other, domestic capital can
allel problem pressures might indeed constitute threaten to exit and this way exert pressure on the
an important factor to account for similar policy governments to lower the level of environmental
developments in certain cases, this approach is regulation, hence implying cross-national policy
hardly sufficient to explain the broad conver- convergence (Hoberg 2001, 127; Simmons and
gence patterns that are observed empirically. Elkins 2003; Drezner 2001, 57–59).
Peter-3383-Chapter-14.qxd 6/8/2006 8:06 PM Page 251
The central and already classical research aspects dominate, stricter standards will
question in the public policy literature refers to constitute a competitive advantage, hence
the consequences of this development on the implying a race to the top (Scharpf 1997, 523).
level of environmental standards. Can we Second, downward pressures can be avoided if
expect a race to the bottom, implying the grad- trade rules allow individual countries to erect
ual weakening of environmental regulations? exceptional trade barriers for products which
Or is it even conceivable that regulatory com- do not comply with national environmental
petition coincides with the opposite scenario standards. Such measures are, for instance,
of a mutual strengthening of national stan- possible within the trade regimes of the World
dards (race to the top) (Holzinger 2002, Trade Organization (WTO) and the EU.
62–63)? Theoretical work suggests that there David Vogel (1995) shows that the erection
are a number of conditions, which may drive of trade barriers might not only avoid a race to
policy in both directions (Vogel 1995; Scharpf the bottom, but can even induce an upward
1996, 1997; Kern, Jörgens and Jänicke 2000; dynamic between national regulations. He
Drezner 2001; Holzinger 2002, 2003). In this observed this development for the regulation
context, particular emphasis is placed on the of car emission standards in the United States
distinction of different types of environmental (US). When California raised its emission
regulation, namely product standards and standards, most US states followed quickly
process standards. for two reasons. First, California was permitted
to apply its standards to foreign car producers.
Second, since licensing procedures for cars are
Product versus Process Standards very expensive, car producers wanted to avoid
multiple arrangements and hence demanded
Product standards define regulatory require- harmonized requirements throughout the US.
ments for the quality and specific characteris- The described mechanisms are less relevant,
tics of traded goods. Process standards, by however, when it comes to process standards.
contrast, refer to the conditions under which First, competition over product quality has no
certain goods are produced. While for product effects on the conditions under which the
standards several factors might inhibit a race to goods of similar quality are produced. For
the bottom and even trigger a race to the top, instance, for consumers of energy it makes no
we find a widely shared expectation that policy difference whether the producing company had
convergence will occur at the lowest common to cope with strict environmental standards or
denominator in the case of process standards not. Second, in the case of process standards,
(Drezner 2001; Holzinger 2002, 2003). Typical governments are not allowed to erect excep-
examples of process standards are sulfur diox- tional trade barriers. No member state of the
ide or nitrogen oxide emission standards for EU or WTO can restrict the import of foreign
large combustion plants. Strict standards products, claiming that the foreign process reg-
demand filters, which raise production costs. ulations applying to these goods were below
Then the domestic steel industry, for example, their domestic standards. In the case of process
suffers from a competitive disadvantage standards, there is thus a higher probability that
against the steel producers abroad, if the latter regulatory competition between countries
need not apply the same strict standards. In implies a race to the bottom (Scharpf 1999, 92).
order to avoid such a disadvantage govern-
ments may want to decrease their standards to
the level of other countries. Other Factors
In the case of product regulation, two con-
ditions can avoid downward dynamics of Notwithstanding these rather clear theoretical
national standards. First, competition between expectations, the empirical literature generally
products might not only be based on their finds ‘lack of empirical support’ (Drezner
price, but also on their quality. If quality 2001, 75) for the hypothesis that regulatory
Peter-3383-Chapter-14.qxd 6/8/2006 8:06 PM Page 252
competition leads to convergence ‘at the bottom’. sufficient explanation for this general tendency.
There is case study evidence for races to the top In the following sections, the focus is therefore
but no systematic confirmation of a race to the on additional factors which might fill this
bottom (Tobey 1990; Vogel 1997; Jänicke 1998; explanatory gap.
Beers and van der Bergh 1999; Kern 2000).
These findings suggest that the extent and level
of environmental policy convergence is
INTERNATIONAL COOPERATION
affected by a number of additional factors.
First, it has to be emphasized that the distinc-
tion of product and process standards covers Environmental policy convergence might not
only parts of environmental policy, namely only be the result of competitive pressures
those regulations which affect competition related to economic integration. It can also
among national industries. By contrast, no con- emerge from deliberate activities of national
vergence can be expected for environmental governments to reduce such pressures through
policies that are not directly related to products regulatory cooperation at the supranational or
or production processes, such as ambient qual- international level (Holzinger and Knill 2004).
ity standards or the protection of nature. This pattern of ‘obligated transfer’ (Dolowitz
Second, pressures to adjust national standards and Marsh 2000, 15) or ‘convergence through
can be low, even if trade-related policies are harmonization’ (Bennett 1991, 225) refers to
concerned. This is the case, for instance, if the constellations, in which national governments
level of environmental regulation has no signif- are legally required to adopt policies and pro-
icant impact on production costs (Vogel 1995; grams as part of their obligations as members
Jänicke 1998). The same holds true if the rela- of international institutions. In other words,
tive market shares of the countries involved in national policies converge because of corre-
competition are rather small. Third, the extent sponding legal obligations defined in interna-
and direction of national adjustment is affected tional or supranational law.2
by the political and social context in which reg- International cooperation as a source of
ulatory decisions are made. In this respect, the environmental policy convergence is generally
political strength and presence of other interests traced to the existence of interdependencies or
than business in national politics, such as envi- externalities which push governments to
ronmental groups or green parties, plays a resolve common problems through coopera-
central role. Depending on the power and polit- tion within international institutions, hence
ical support of these actors, the opportunities sacrificing some independence for the good
for regulatory adjustments to strengthen the of the community (Drezner 2001, 60; Hoberg
competitiveness of national industries might 2001, 127). Once established, institutional
thus be quite restricted (Holzinger 2002). arrangements will constrain and shape the
However, even if we accept that these factors domestic policy choices, even as they are con-
play a role in avoiding an overall pattern of a stantly challenged and reformed by their mem-
constant downgrading of environmental stan- ber states. This way, international institutions
dards, we are still left with two unresolved are not only the object of state choice, but at
questions. First, it is unclear from the theoreti- the same time consequential for subsequent
cal considerations presented so far why there governmental activities (Martin and Simmons
should be convergence, even for those environ- 1998, 743). However, as member states volun-
mental policies that neither refer to product tarily engage in international cooperation and
nor to process standards. Second, it is puzzling actively influence corresponding decisions
that there is only clear empirical support for and arrangements, the impact of international
races to the top but not for the opposite legal obligations on national policies consti-
scenario of races to the bottom. Although the tutes no hierarchical process; it can rather be
factors mentioned above can account for this interpreted as ‘negotiated transfer’ (Dolowitz
pattern in some cases, they hardly constitute a and Marsh 2000, 15).
Peter-3383-Chapter-14.qxd 6/8/2006 8:06 PM Page 253
requirements for the quality and market access hardly constitutes a feasible alternative. In view
of products would complicate international of the high political priority attached to envi-
trade and hence imply economic disadvantages ronmental protection in rich countries, any
for the industries in all countries. As a conse- attempt to downgrade regulations will meet
quence, all member states share a common strong political opposition and hence be diffi-
interest in international or European harmo- cult to achieve. As a consequence, rich coun-
nization; i.e., all states prefer an international tries will generally prefer the status quo of
agreement to the status quo of different national different national regulations to international
regulations (Holzinger 2002, 69). Interests dif- harmonization at the level of the low-regulating
fer, however, with respect to the level of harmo- countries (Scharpf 1996, 119–120).
nization, with poor countries preferring lower The same constellation applies, albeit for
standards than rich countries. different reasons, also to poor countries. On
In this context, however, member states the one hand, strict international standards
interested in higher standards are generally in would imply massive problems for their less
a more powerful bargaining position in order developed industrial sectors. On the other
to realize their objectives. This can be traced to hand, even harmonization at a lower level
two aspects. First, they have the opportunity to would imply high economic disadvantages, as
erect exceptional trade barriers, denying domestic industries would face strong compe-
market access to products which do not com- tition from their more productive counter-
ply with the domestic regulatory requirements. parts in rich countries. Similar to rich
For the EU, this option is specified in article 30 countries, poorer states therefore prefer the
of the European Treaty. Second, article 95 of the status quo of different national standards to
Treaty allows member states under certain con- international or European harmonization
ditions to keep or introduce stricter domestic (ibid.). Figure 1 summarizes the different
regulations, even if corresponding harmoniza- interest constellations of rich and poor coun-
tion at the European level has already been tries for product and process regulation.
agreed (Stewart 1993). Although the erection In sum, this theoretical perspective suggests
of exceptional trade barriers based on articles that international agreements on strict process
30 or 95 is confined to specific conditions, standards are rather unlikely. In view of the
these rules nevertheless allow rich member underlying constellations of interests, the
states to unilaterally push through their strict establishment of strict international regula-
standards at the European level (Scharpf 1996, tions can only be expected if rich states are able
118). As a consequence, there is a high probabil- and prepared to financially compensate the
ity that harmonization for environmental prod- poor countries for potential economic disad-
uct standards occurs at the top rather than vantages emerging from a high level of inter-
bottom level of national standards. national regulation. By contrast, the chances
The conditions for international coopera- for international cooperation look much
tion are less favorable, by contrast, for the har- better for product regulation, given the har-
monization of process standards. From the monization advantages for both rich and poor
perspective of rich countries, their high level of countries.
regulation implies economic disadvantages, As mentioned earlier, however, empirical
because they cannot restrict market access for evidence does not confirm the pessimistic
products from countries with lower process expectations on the occurrence and level of
standards. Rich countries, therefore, have a international harmonization of process stan-
strong interest to establish their strict regula- dards. The EU, for instance, in many areas (reg-
tions as international standard. If this objective ulation of chemicals, air and water pollution)
cannot be achieved, the second-best solution adopted process standards that are far beyond
would be harmonization at a lower level in the ‘lowest common denominator’ of national
order to avoid economic disadvantages for the regulations (Jachtenfuchs 1996; Knill 1998).
domestic industry. This solution, however, Moreover, member states often accept European
Peter-3383-Chapter-14.qxd 6/8/2006 8:06 PM Page 255
•P
•R
SQ SQ
•P
•R
Figure 14.1 Benefits of possible agreements compared to the status quo (SQ) of different national
standards for product and process regulation (P = regulation at the level of the poor countries,
R = regulation at the level of the rich countries). Agreements above SQ improve the position of poor
countries; agreements on the right of SQ improve the position of the rich countries (compared to the
status quo)
(Source: Scharpf 1996, 121).
policies, although their implementation entails interest position of the member states in nego-
high costs of economic and institutional adap- tiations at the European level. Rather regula-
tation (Héritier, Knill and Mingers 1996; tory competition, in the sense of systems
Holzinger 1994; Jordan 1999). As will be shown competition between member states, is over-
in the following, the theoretical considerations lapped by a second pattern of regulatory com-
presented so far require further differentiation petition which is characterized by a highly
in order to better account for the empirical different logic. Central to this argument is the
developments. In particular, it has to be consid- empirical observation that member states pur-
ered that the level of international regulation is sue their interests not only through the adjust-
not only affected by economic factors, but also ment of national standards (as suggested by
by the objective of national governments and classical theories of regulatory competition). At
administrations to avoid costs of institutional the same time, they strive to make the regula-
adaptation. tory design of European policies as much as
possible consistent with their domestic arrange-
ments; i.e., they want to establish their approach
Regulatory Competition and as the respective European policy.
International Cooperation – II As a consequence, member states perma-
nently compete for influence in European
How can we explain the rather progressive policy-making. They engage in a regulatory
development and level of international coop- contest to shape European policy, to champion
eration in the environmental field, which is their interests and push through their policy
rather surprising, from the perspective of reg- concepts. All member states are interested to a
ulatory competition in integrated markets? greater or lesser degree in putting their stamp
Research on the formulation of EU environ- on Community policy. As shown by Héritier,
mental policy shows that this form of compe- Knill and Mingers (1996), the big three member
tition is not the only factor determining the states Britain, France, and Germany, have been
Peter-3383-Chapter-14.qxd 6/8/2006 8:06 PM Page 256
the major protagonists in a regulatory contest to that are not in line with existing regulatory
introduce their philosophies and practices in traditions and approaches require fundamen-
combating industrial emissions at the European tal institutional changes at the level of the
level, whilst other countries in the Union have member states (Knill and Lenschow 1998;
acted as coalition partners, negotiating specific Knill 2001). The attention of member states is
exchange deals for their concessions. Similar pat- therefore not so much directed at the level of
terns were also observed in other fields of envi- environmental standards to be decided at the
ronmental policy both at the European European level, but at the extent to which the
(Andersen and Liefferink 1997) and interna- underlying approaches and instruments of
tional level (Jänicke 2003). the respective policy can be integrated into the
What are the reasons for this phenomenon? domestic regulatory framework without trig-
Why should member states be interested in gering major institutional changes.
influencing European policy-making in line The fact that this problem is far from trivial
with their regulatory approaches and instru- is demonstrated by the widespread implemen-
ments? In this context, two factors are of tation deficit of EU environmental policy. A
importance: institutional and economic inter- large part of these implementation problems
ests of the member states (1) as well as the spe- are the result of a lack of fit European policies
cific advantages for those member states with national regulatory institutions (Knill
relying on the strategy of the first move (2). and Lenschow 2000). In many member states,
The interests of the member states in the con- for instance, the state tradition with a strong
text of EU policy-making refer to two basic emphasis on the rule of law pre-structures the
aspects. First, member states concentrate on administration’s behavior in the environmen-
strengthening the competitive position of their tal field; European policies implying adminis-
industries. In so doing, their focus is not only on trative discretion or a transparent relationship
the protection of domestic markets against with the public are perceived as misfits in this
foreign competition, but also on the opening- regulatory framework and hence resisted
up of new markets for domestic producers of (Kimber 2000).
environmental protection technology. While To be sure, it could still be argued that
this orientation is basically in line with the national interests to avoid institutional adjust-
underlying assumptions of classical theories of ment must not necessarily imply a constellation
regulatory competition, this no longer holds where member states compete over influencing
true for a second issue that is of crucial rele- the design and content of European policy pro-
vance for member states in European negotia- posals. Why do national governments not just
tions, namely, the objective to minimize costs of simply rely on the option to block policies
institutional adjustment emerging from poten- which are not in line with their interests? How
tial incompatibilities between EU requirements can we explain the proactive strategies of many
and domestic regulatory arrangements. countries to influence the development of
The particular importance of this latter European policies already at very early stages?
aspect can be traced to the fact that, in many The central factor accounting for this devel-
instances, European policy content and insti- opment refers to strategic advantages of the
tutional requirements for the effective imple- first move. Environmental leader or pioneer
mentation of these policies are closely related. states have better chances to successfully influ-
While being aware of the fact that the degree to ence European policy decisions in line with
which policy contents and institutional impli- their interests as it is the case for states that rely
cations are coupled may vary from policy to on more passive strategies. This holds true,
policy and from sector to sector, it cannot be especially for the case of the EU, as the possi-
ignored that the growing importance of bilities to veto environmental policy decisions
European policies leaves its mark on domestic have been subsequently reduced through the
administrations and institutions. Accordingly, introduction of qualified majority voting as a
it is well conceivable that European policies decision rule in the Council (Andersen and
Peter-3383-Chapter-14.qxd 6/8/2006 8:06 PM Page 257
Liefferink 1997; Héritier, Knill and Mingers through their interests in earlier stages of the
1996; Jänicke and Weidner 1997). policy process now have a final chance to
First mover advantages can be observed in achieve their objectives. In general, the extent
particular during the stages of problem defini- to which the final decision deviates from the
tion and agenda-setting at the European level. initial proposal is strongly affected by the
The strategy of the ‘first move’ gives the initiator underlying constellation of national interests
the opportunity to participate in determining in the Council.
which problems require action from the Empirical evidence indicates four possible
European perspective, and thus to influence the patterns in this context, systematized from the
European decision-making agenda. However, it perspective of the first mover in the question.
succeeds only with the support of the European In making the first move, what success does the
Commission, which is the ‘gatekeeper’ in this initiator country have in shaping the measure
process, given its formal monopoly to initiate concerned? First, the first mover succeeds in
European policies. The policy concepts pushing through its regulatory proposal in all
advanced by the first mover therefore have to be phases up to policy formation with respect to
in line with the ideas and interests of the both the basic regulatory approach and sub-
Commission. stantive provisions. This was the case, for
If in this regulative contest the first mover is instance, for the development of the directive
able to win the support of the Commission for on free access to environmental information
its proposal, it gains the opportunity to add an which was strongly influenced by correspond-
item to the European agenda for which it has a ing policies in Britain (Kimber 2000). The
relatively precise concept. If the proposal is second path is taken where the initiator’s pro-
compatible with the Commission views on posal is complemented and expanded by addi-
the issue and how to tackle it, the initiator tional requirements. Here, too, the first mover
country has a good chance of seeing its attains its regulatory goal, which is, however,
national approach become the received view of amplified and augmented by the Commission
the issue for the ensuing drafting phase. The and other member states until a comprehen-
strong interest in the policy proposal generally sive piece of legislation on the issue is adopted.
goes hand in hand with a higher degree of In a third scenario, opposition from other
national expertise in the area to be regulated. member states implies that the initial proposal
Both aspects enhance the first mover influence. is scaled down so that countries opposed to the
Once an actor has formulated a proposal and – measure can see their way to giving their
in cooperation with the Commission – has approval. A fourth pattern includes constella-
occupied the problem-solving terrain in a spe- tions in which the initiator comes up against
cific manner, and has moreover collaborated in the decided opposition of other countries and
drawing up the agenda (profiting from the might have to accept far-reaching changes that
inactivity of those concerned) the others are call into question the fundamental regulatory
necessarily in less favorable starting positions. approach. The reason may be that either
Initially, where it is a matter of developing this another high-regulating country prefers a dif-
specific proposal, other member states can ferent approach and wants to see it receive
only react to the initiator’s project. In other equal consideration, which results in the typi-
words, whoever takes the initiative and makes cal policy mix or leads to a new, third solution
the proposal calls the tune in this phase of being worked out jointly (Héritier, Knill and
defining the problem and setting the agenda Mingers 1996, 332–334; Eichener 1997, 604).
(Héritier, Knill and Mingers 1996, 13). In some cases, this constellation implies that
However, if and to what extent the advan- individual policies are characterized by differ-
tage of the first move actually results in corre- ent, partially contradictory regulatory tradi-
sponding policy decisions, crucially depends tions of different member states. A case in
on the negotiations in the Council. Member point is the directive on integrated pollution
states which were not successful in puting and prevention control which combines both
Peter-3383-Chapter-14.qxd 6/8/2006 8:06 PM Page 258
the German orientation on best available at point sources (emission standards), the control
technologies and the British emphasis on bal- or construction of products (product standards),
ancing environmental quality and economic limits for all emission irrespective of origin
costs of pollution control (Weale et al., 2000). (emission ‘bubbles’), the application of control
In sum, it becomes apparent that it is only technologies (technological standards); (2) pro-
in the fourth scenario that the advantage of cedural requirements concerning the authoriza-
the first move is completely lost during the tion of industrial processes or other projects as
decision-making stage. In all other cases that well as public participation, and (3) voluntary
can be observed empirically, pioneer countries agreements and industrial self-regulation (Haigh
more or less succeed in influencing the final 1996; Knill 2003, 39–72).
policy decision in line with their interests. The This general picture holds true, although
strategy of the first move thus consists of a one can distinguish different stages of EU envi-
rather promising way to effectively influence the ronmental policy, with certain approaches and
output of international and European coopera- instruments playing a more or less dominant
tion (Kern, Jörgens and Jänicke 2000, 512; role. Initial attempts of the Community to
Andersen and Liefferink 1997). What are the control environmental pollution were largely
effects of this type of regulatory competition for based on a reactive approach relying on quality-
influencing European cooperation on policy based instruments and were directed at single
design and levels of environmental standards? environmental media. During the 1980s, the
First, diverging interests of the member Commission – strongly inspired by the first
states play an important role in promoting the mover role of Germany – abandoned this
regulatory variety of EU environmental policy. strategy in favor of emission-based instru-
As regulatory and institutional arrangements ments and placed greater emphasis on the pre-
differ from member state to member state, the ventative character of EU environmental
regulatory contest favors the emergence of EU policy. From the late 1980s onwards, a further
environmental policy as a regulatory patch- stage of EU environmental regulation can be
work. Not only within a single policy area but observed. Strongly influenced by the pioneer-
sometimes within one and the same directive, ing role of Britain, the focus is on so-called
divergent approaches can be on parallel offer if ‘new instruments’, emphasizing regulatory
it has proved impossible to negotiate a consen- transparency, participation, self-regulation
sus (Héritier, Knill and Mingers 1996, 21–22). and voluntary agreements, as opposed to clas-
On the other hand, the regulatory contest sical forms of hierarchical top-down regula-
favors an equal distribution of adaptational tion (Knill and Lenschow 2000a). However, the
requirements between different member emergence of new regulatory stages character-
states, given that regulatory competition is ized by the dominance of certain approaches
unlikely to always favor the same ‘winner’ increased rather than reduced the regulatory
(Knill 1997). As a consequence, member states variety of EU environmental policy, as new
are generally more prepared to accept adapta- regulatory concepts were generally added to
tions in one area as long as their interests are existing approaches rather than replacing the
sufficiently taken into account when other latter. Moreover, the dominance of certain
measures are decided. approaches only implied the reduction rather
The patchwork character of EU environmental than exclusion of concepts that were still associ-
policy can be illustrated with respect to several ated with approaches dominant during earlier
dimensions. First, there is a variance in regulatory regulatory stages.
principles, including both reactive and preventa- Second, the regulatory contest for influence
tive approaches to environmental protection. in international cooperation activities gener-
Second, we find a multitude of regulatory instru- ally helps to avoid the often expected scenario
ments, including (1) numerical standards con- of an environmental dumping or race to the
cerning the state of the environment (quality bottom. At the European level, this can be
standards), the maximum emission of pollutants traced to the fact that member states striving to
Peter-3383-Chapter-14.qxd 6/8/2006 8:06 PM Page 259
set the agenda in line with their interests have for sustainable development, as well as regulatory
to win the support of the Commission. Since measures opening up the access to environ-
the Commission has a strong institutional self- mental information for the general public
interest in increasing its regulatory competen- (Tews, Busch and Jörgens 2003, 571). For all of
cies towards the member states, it will promote these measures, a still increasing number of
only such proposals which fit with these objec- OECD countries has progressively adopted
tives. From this constellation, we thus can similar legislation and created similar adminis-
expect the gradual expansion and upgrading of trative structures, although neither market
EU environmental policy over time (Eichener forces nor international agreements would
1997). Although this general tendency does have suggested such a broad pattern of cross-
not exclude potential failures of negotiations national convergence.
on individual policy issues, or the possibility To account for this development, the lit-
that in some cases European standards will erature emphasizes varying convergence mech-
remain below the level of the highest-regulating anisms which are based on transnational
member state, the long-term dynamics of the communication networks. The first scenario of
regulatory contest increase the scope and level policy emulation implies the simple copying of
of environmental protection within the EU policy decisions taken elsewhere. This pattern is
(Knill 2003, 135). Similar dynamics have also generally explained by a broad variety of
been reported for environmental cooperation factors, including the number of countries
at the international level (Jänicke 2003; Kern, which have already adopted a certain policy
Jörgens and Jänicke 2000). (Meyer and Rowan 1977), the striving for legit-
imacy in constellations of high uncertainty
(DiMaggio and Powell 1991, 70), the desire of
actors ‘not to be left behind’ (Meyer et al. 1997),
TRANSNATIONAL COMMUNICATION
the existence of time-pressures (Bennett 1991,
223), or the striving to avoid high costs of
The interplay of economic pressures emerging information which are probably much less with
from global market integration and the spe- simple imitation than with more demanding
cific dynamics underlying international coop- forms of learning (Tews 2002, 180).
eration can thus be seen as important factors Second, policy convergence can be the result
which explain not only the broad convergence of learning processes, understood as the rational
of national environmental policies, but also utilization of available experience elsewhere
the fact that convergence does not automati- (Rose 1991). Learning through transnational
cally coincide with a downgrading of levels communication, however, is not restricted to
of environmental protection. As empirical bilateral policy transfer (Dolowitz and Marsh
evidence shows, in many instances, quite the 2000), but can also result from the development
opposite pattern is the case. Notwithstanding of common problem perceptions and corre-
the relevance of the explanatory factors inves- sponding solutions within transnational elite
tigated so far, however, we are still left with networks or epistemic communities (Haas
some unresolved questions. This holds true 1992). Convergence in this sense results from
in particular for those constellations in which the development of shared ideas and beliefs
national environmental policies converge even amongst a relatively coherent and enduring net-
in the absence of either economic pressures or work of elites engaging in regular interaction at
corresponding legal obligations emerging from the transnational level (Bennett 1991, 224).
international agreements. Empirical studies A third convergence factor linked to
indeed reveal many cases for which this transnational communication is the promo-
scenario applies, including, for instance, the tion of policy models by international institu-
establishment of environmental ministries and tions. They often play a highly active role,
agencies, the introduction of energy or carbon promoting the spread of distinctive policy
taxes, national environmental plans or strategies approaches by learning through performance
Peter-3383-Chapter-14.qxd 6/8/2006 8:06 PM Page 260
comparisons (Tews 2002, 174). For instance, represented in the network, but also that they
with regard to sustainable development, the are at the same time ‘observed’ by their coun-
United Nations Commission on Sustainable terparts; i.e., they have to demonstrate the
Development regularly publishes status reports quality and legitimacy of their concepts in the
on the national implementation of Agenda 21 – transnational discourse. National representa-
the latest of which was on the occasion of the tives have to legitimize and justify national
2002 World Summit of Johannesburg (Tews, developments in light of ‘transnational scru-
Busch and Jörgens 2003). In a similar vein, tiny’. In this context, difficulties to legitimize
think tanks, consultancy firms or foundations distinctive national approaches will increase,
regularly disseminate information on best the more other countries have already moved
practice and new policies at the national and to another model which is perceived as the
international levels. most promising approach within the transna-
Cross-national policy transfer is stimulated tional setting (Meyer and Rowan 1977; Meyer
by international agreements on broad goals et al., 1997).
and standards that national policies should The competition of ideas, as it emerges in
aim to achieve, institutionalized peer review particular from the dissemination and evalua-
and identification of best practice (bench- tion of best practice, can generally be expected
marking) as well as the construction of league to result in an overall strengthening of regula-
tables ranking national policies in terms of tory concepts; hence, inducing an upward shift
performance to previously agreed criteria of environmental standards, as international
(Bomberg and Peterson 2000; Humphreys organization will generally promote the most
2002, 54). In constantly searching for new innovative national approach (Holzinger and
policy ideas, disseminating best practice and Knill 2004). For instance Jörgens (2003) shows
evaluating domestic policy performance, inter- that the worldwide proliferation of green plans
national institutions function as mediators of and sustainable development strategies was to
cross-national policy transfer, driving national a large extent based on the effects of transna-
governments to adopt successful policy models tional communication and benchmarking
(Kern, Jörgens and Jänicke 2000, 10). activities by different public and private insti-
In general, the impact of transnational com- tutions and networks.
munication on the development of national
environmental policies is expected to increase
with the density of information exchange within
CONCLUSION
transnational networks (Simmons and Elkins
2003). This includes not only the frequency of
interaction, but also the breadth of interaction; This article started from two general empirical
i.e., the functional differentiation of transna- observations on the development of national
tional networks. It is well-acknowledged in the environmental policies. First, we observe an
literature that interaction density between states increasing similarity of domestic policies and
increases with their membership in interna- regulatory frameworks in many issue areas.
tional institutions which strongly facilitate and Second, this pattern of convergence does not
intensify transnational information exchange coincide with an often feared scenario of a race
(Kern 2000, 267). Moreover, geographical to the bottom in levels of environmental protec-
proximity and cultural ties between states were tion. Several explanatory factors were presented
identified as important factors increasing cross- in order to explain these patterns, including reg-
national interaction density (Rose 1991). ulatory competition, international cooperation
The embeddedness of national bureaucrats and transnational communication.
and policy-makers in transnational expert On the one hand, the analysis of these
networks implies not only that these actors factors indicates that the observed patterns
can observe developments in other countries and developments of national environmental
Peter-3383-Chapter-14.qxd 6/8/2006 8:06 PM Page 261
policies, in many instances, can hardly be fully to the narrow meaning of international cooperation. The
understood by focusing on the isolated effects meaning of ‘cooperation’ is thus equivalent to its meaning
in cooperative game theory.
of individual factors. Rather, we need to apply
a broader perspective, taking into account that
several causal factors might operate in parallel. REFERENCES
In this context it is well conceivable that the
different factors interact. It is thus an impor-
tant area of future research to develop Andersen, M. S. and Liefferink, D. (1997). European
hypotheses on the interaction effects of all Environmental Policy. The Pioneers. MUP:
Manchester.
causal factors of convergence.
Beers, C. and van der Bergh, J. C. J. M. (1999). An
On the other hand, the consideration of the Empirical Multi-Country Analysis of the Impact
three convergence mechanisms reveals that of Environmental Regulations on Foreign Trade
fundamental shifts in the level of intervention, Flows. In: Kyklos 50, 29–46.
in the sense of races to the bottom, can only be Bennett, C. (1991). What Is Policy Convergence and
expected in rather specific constellations. More What Causes It? In: British Journal of Political
specifically, such scenarios are restricted to Science, 21, 215–233.
situations in which regulatory competition is Bomberg, E. and Peterson, J. (2000). Policy Transfer
the only or dominant mechanism to be effec- and Europeanization. Passing the Heinecken
tive. In all other constellations, upward shifts Test? Queens Papers on Europeanization 2, at
are a more likely scenario. http://www.qub.ac.uk/ies/ accessed in May 2001.
Brickman, R. and Jasanoff, S. and Ilgen, T. (1985).
These findings, together with the fact that dif-
Controlling Chemicals. The Politics of Regulation
ferent causal factors are often operating in par- in Europe and the United States. Ithaca, New York:
allel, indicate that an environmental dumping is Cornell University Press.
indeed a highly unlikely scenario. This does not Bungarten, H. H. (1978). Umweltpolitik in
automatically imply that we should be overopti- Westeuropa. EG, internationale Organisationen und
mistic about environmental problem-solving at nationale Umweltpolitiken. Bonn: Europa Union.
both the national and international level. But we Cowles, M. and Carporaso, J. and Risse, T. (2001).
should be aware that global market integration, Transforming Europe. Europeanization and
international cooperation and transnational Domestic Change. Ithaca, London: Cornell
communication, offer promising ways to effec- University Press.
tively address the problem of environmental DiMaggio, P. J. and Powell, W. W. (1991). The Iron
Cage Revisited. Institutionalised Isomorphism and
pollution.
Collective Rationality in Organizational Fields.
In: Powell, W. W. and DiMaggio, P. J. (eds.), The
New Institutionalism in Organizational Analysis.
NOTES Chicago: Chicago University Press, 63–82.
Dolowitz, D. P. and Marsh, D. (1996). Who Learns
1. Notwithstanding this general discussion over new What From Whom. A Review of the Policy Transfer
instruments, interventionist patterns still constitute the Literature. In: Political Studies, 44, 343–357.
dominant form of environmental regulation both at the Dolowitz, D. P. and Marsh, D. (2000). Learning from
domestic and international level (cf. Héritier 2002; Abroad: The Role of Policy Transfer in Contem-
Holzinger, Knill and Schäfer 2003). porary Policy Making. In: Governance, 13, 5–24.
2. I use the term international cooperation here in its Drezner, D. W. (2001). Globalization and Policy
strictest meaning, as being equivalent with international Convergence. In: The International Studies
legal obligation. In a wider understanding, international Review, 3, 53–78.
cooperation might also include non-binding agreements
Eichener, V. (1997). Effective European problem-
and all kinds of voluntary activities of international orga-
nizations. However, as the effects of non-binding policies
solving. Lessons from the regulation of occupa-
are even more difficult to predict than the effects of legal tional safety and environmental protection. In:
obligation and of regulatory competition, the inclusion of Journal of European Public Policy 4(4), 591–608.
non-binding international policies would strongly Goodman, J. and Pauly, L. (1993). The Obsolescence
increase the complexity of the argument. I therefore stick of Capital Control. In: World Politics, 46, 50–82.
Peter-3383-Chapter-14.qxd 6/8/2006 8:06 PM Page 262
15
Cultural Policy
K E V I N V. M U L C A H Y
This essay has four major goals. First, it the cultivation and refinement of the mind;
attempts to operationalize the elusive notion “the artistic and intellectual side of civiliza-
of culture and the broad outlines of what is tion;” finally, “the distinctive customs, achieve-
entailed in a cultural policy. Second, there is an ments, production, outlooks, etc., of a society
overview of what is entailed in public culture or group; the way of life of a society or group.”
as a public policy. Third, the objectives and The latter can be characterized as the “anthro-
justifications of public culture are surveyed. pological” sense of culture.
Fourth, the relationship between political The American Heritage Dictionary first defines
cultures and the particular expressions of culture as: ‘the totality of socially transmitted
public cultures is specified. Finally, the future behavior patterns, arts, beliefs, institutions, and
trends in cultural policy are briefly evaluated. all other products of work and thought; the pre-
dominant attitudes and behavior that character-
ize the functioning of a group or organization.”
Secondly, culture is “intellectual and artistic
WHAT IS CULTURAL POLICY?
activity and the works produced by it;” a high
degree of taste and refinement formed by
Culture, according to literary critic Raymond esthetic and intellectual training.
Williams, is one of the two or three most com- Typically, culture is used in political dis-
plicated words in the English language course as the “arts.” A Ministry of Culture is an
(Williams, 1977: 76). administrative agency responsible for public
It is worth noting that the root of the word support of these artistic activities and often the
is from the Latin colere, to till. There is the cul- humanities and scholarship as well (Williams,
tivation of a field as there is the cultivation of 1977: 80).
intellectual and esthetic sensibilities; the In its broadest sense, “public policy is what-
process of becoming educated, polished, ever governments choose to do or not to do”
refined; that is, cultured: the state of being (Dye, 2005: 1). Similarly, public policy can be
civilized. In sum, culture suggests a process for seen as a purposive decision “followed by an
the deliberate and systematic acquisition of an action or set of actions in dealing with a
intellectual sensibility. problem or matter of concern” (Anderson,
The Oxford English Dictionary first defines 1975: 3). “Stated most simply, public policy is
culture with reference to tillage. Secondly, it is the sum of government activities, whether
Peter-3383-Chapter-15.qxd 6/8/2006 7:38 PM Page 266
(including the for-profit cultural industries), major vehicles for delivery of the arts”
the humanities, and the heritage” (Schuster, (Cummings and Katz, 1987: 359). With the
2003: 1). Cultural policy, then, involves gov- prominent exception of the United States,
ernmental strategies and activities that pro- where the Broadcasting Act of 1920 essentially
mote “the production, dissemination, marketing, licensed the airwaves to commercial networks,
and consumption of the arts” (Rentschler, broadcasting was from its earliest days a public
2002: 17). Yet, if policy is the “intentionality” of responsibility. Official control, however, is
programs that seek to achieve certain out- often delimited by the creation of some sort of
comes in a field of activity, there are also goals autonomous governing board. Governments
that are “implicit” in a policy and its programs often saw broadcasting as a means of fostering
(Schuster, 2003: 1). national bonds (for example, the Canadian
Using the metaphor of “mapping,” Mark Broadcasting Corporation with both French
Schuster argues that understanding a policy and English programming) and sponsoring
requires viewing programmatic activities as a shared national rituals (such as the sovereign’s
“terrain of intentions.” The actions that a state Christmas address on the British Broadcasting
and its many operational entities take that Corporation).
affect the cultural life of its citizens, whether It should be noted, however, that “public”
directly or indirectly, whether intentionally or broadcasting in the Unites States is provided
unintentionally, collectively constitute the programmatically by the Public Broadcasting
cultural policy of the state. In sum, a state’s System (PBS) and National Public Radio
cultural policy “can best be understood once one (NPR). These are both 501(C)(3)s, that is,
has an atlas of such maps”(Schuster, 2003: 3). private, not-for-profit organizations. The
A cultural policy encompasses a much Corporation for Public Broadcasting (CPB) is
broader array of activities than what was tradi- a quasi-autonomous, government corporation
tionally associated with an arts policy. The that provides limited funding to local stations
latter typically involved public support for for technical assistance and program acquisi-
museums, the visual arts (painting, sculpture, tions; however, CPB is forbidden by law from
and pottery), the performing arts (symphonic, producing programming. Public radio and
chamber and choral music, jazz, modern television in the U.S. is essentially a confedera-
dance, opera and musical theatre, “serious” tion of independent entities, which are free to
theatre), historic preservation, and humanities schedule such programming as they deem
programs (such as creative writing and appropriate. Moreover, each station is respon-
poetry). A cultural policy would involve sup- sible for its own financial support. In essence,
port for all the aforementioned activities, but American public broadcasting rests on a
also other publicly-supported institutions such “bedrock of localism” in both administration
as libraries and archives; battlefield sites, zoos, and funding (Katz, 1982; Mulcahy and Widoff,
botanical gardens, arboretums, aquariums, 1986; Mulcahy and Widoff, 1988).
parks; as well as community celebrations, fairs Another important example of the broader
and festival; folklore activities such as quilting, net cast by the concept of cultural policy is the
“country” music, folk dancing, crafts; perhaps role of the education community. There is a
certain varieties of circus performances, natural affinity between education and
rodeos and marching bands. culture. In countries with well-established and
Television and radio, although considered widely recognized cultural traditions, cultural
separately as broadcasting policy, have long offerings are core components of the educa-
functioned as “major supporters of the arts tional curricula. The U.S. is an exception again
by purchasing the work of performing artists with arts and cultural courses being the most
on a massive scale, by developing audiences frequently cut offerings. Moreover, there are
for live performances, and sometimes even by decided benefits from an alliance between
making direct grants to artistic organizations. the cultural and educational communities
Moreover, television and radio have become (Cummings and Katz, 1987: 358). First, it is an
Peter-3383-Chapter-15.qxd 6/8/2006 7:38 PM Page 268
Versailles symbolized the king’s personal rule theatre to the remotest regions of the country
and was widely imitated by European monarchs and to culturally underserved areas within
and princelings. cities (Bakke, 1994: 115). Under Charles
For the great mercantile princes of the nine- DeGaulle, the first modern Minister of
teenth and early twentieth century, cultural Culture, Andre Malraux established a network
patronage also represented a form of glorifica- of maisons de la culture throughout the French
tion. Admittedly, the question of motivation is provinces. As “beacons of hope” in the provin-
complex, but great palaces of culture, in the cial darkness, these cultural agencies would
form of museums open to the public, signified serve as venues for Parisian and international
a grandness of philanthropic spirit and created offerings, as well as showcases for high-quality
an edifice that visually bespoke the donor’s local productions (Lebovics, 1999).
personal grandeur. Of course, support for The objective of cultural democratization is
the arts could also serve to legitimize these the esthetic enlightenment, enhanced dignity,
“robber-barons” and confirm the social status and educational development of the general
of the nouveaux riches. As with royal patron- citizenry. “Dissemination was the key concept
age, the art forms subsidized were a matter of with the aim of establishing equal opportunity
personal taste preferences. for all citizens to participate in publicly orga-
By contrast, a democratic state cannot be seen nized and financed cultural activities”
as simply indulging the esthetic preferences of a (Dueland, 2001: 41). To further this goal, per-
few, however enlightened. Consequently, a formances and exhibitions are low cost; public
democratic cultural policy must articulate its art education promotes equality of esthetic
purposes in ways that demonstrate how the opportunity; national institutions tour and
public interest is being served. perform in work places, retirement homes and
housing complexes.
Democratization of Culture
Cultural Democracy
Since culture is a “good,” and one that is “good
for you,” governments have pursued programs As indicated earlier, the “democratization of
to promote greater accessibility. In this concep- culture” is a top-down approach that essentially
tualization, significant esthetic works should privileges certain forms of cultural program-
be made broadly available to the public. In ming that are deemed to be a public good.
other words, “high culture” should not be the Clearly, such an objective is open to criticism
exclusive preserve of a particular social class or for what is termed cultural elitism; that is, the
of a metropolitan location. Rather, the benefits assumption that some esthetic expressions are
of the highest reaches of cultural excellence inherently superior – at least as determined by a
should be made available broadly and widely. cognoscenti concerned with the acquisition of
In sum, national cultural treasures should be cultural capital (Bourdieu, 1984). “The problem
accessible without regard to the impediments with this policy was that, fundamentally, it
of class circumstances, educational attainment intended to create larger audiences for perfor-
or place of habitation. mances whose content was based on the experi-
Typically, the cultural programs following ence of society’s privileged groups. In sum, it has
this policy formulation have been vertical in been taken for granted that the cultural needs of
nature; that is, top-down, center-periphery. all society’s members were alike” (Langsted,
For example, Norway is a large, sparsely popu- 1990: 17). The objective of cultural democracy,
lated country, with its cultural institutions on the other hand, is to provide for a more par-
concentrated in Oslo, the capital and largest ticipatory (or populist) approach in the defini-
city. With public subsidies, these national insti- tion and provision of cultural opportunities.
tutions have extensive touring programs to In essence, there is a shift from a top-down to
bring symphonic music, opera, ballet, and a bottom-up policy; that is, the government’s
Peter-3383-Chapter-15.qxd 6/8/2006 7:38 PM Page 270
these positions as mutually exclusive, rather argued that there is a parity between the state’s
than complementary. “Elitists” are denounced responsibility for its citizens’ social-economic-
as “high brow snobs” advocating an esoteric physical needs and their access to culture and
culture; populists are dismissed as “pandering opportunities for artistic self-expression.
philistines” promoting a trivialized and com- However, the esthetic dimension of public
mercialized culture. However, these mutual policy has never been widely perceived as intu-
stereotypes belie complementariness between itively obvious or politically imperative.
two bookends of an artistically autonomous Accordingly, the cultural sector has often
and politically accountable cultural policy. argued its case from the secondary, ancillary
There is a synthesis that can be termed a “lati- benefits that result from public support for
tudinarian approach” to public culture; that is, programs that are seemingly only esthetic in
one that is esthetically inclusive and broadly nature. Cultural policy is not justified on the
accessible (Mulcahy, 1995b: 180–181; Mulcahy, grounds that it is a good-in-itself, but rather
1995c: 223–224). that it yields other good results; culture is good
Such a public-cultural policy would remain because of its utilitarian value, not because of
faithful to the highest standards of excellence its inherent value.
from a broad range of esthetic expressions, The most commonly invoked argument
while providing the widest possible access to from utility is the “economic impact of the
people from different geographic locales, arts.” As a staple of political advocacy, the data
socio-economic strata, and educational back- from a veritable cottage industry of commis-
ground (Mulcahy, 1991: 22–24). In conceiving sioned studies document the contributions of
of public policy as an opportunity to provide arts organizations to the local economy and
alternatives not readily available in the market- serve to dispel any notions that cultural subsi-
place, public cultural agencies would be better dies are a “handout.” A quantitative justifica-
positioned to complement the efforts of the tion is provided, demonstrating that every
private sector rather than duplicate their expenditure on arts activities produces a mul-
activities. Similarly, cultural agencies can pro- tiplier that ripples through the local economy,
mote community development by supporting with increased spending on hotels, restaurants,
artistic heritages that are at a competitive dis- taxis/car parks; also, arts organizations buy
advantage in a cultural world that is increas- supplies from local vendors and employ people
ingly profit-driven. In sum, excellence should who pay taxes and consume goods and services
be viewed as the achievements of greatness (Cohen et al., 2003; Myerscough, 1988) There
from a horizontal, rather than a vertical, per- is no doubt that there is an important eco-
spective and a cultural policy as supporting the nomic contribution that culture can make. On
totality of these varieties of excellence. the other hand, the methodology of these
economic-impact studies, as well as the self-
serving, exaggerated and uncritical nature of
Cultural Utilitarianism their findings, has come under strong attack by
professional economists (Sterngold, 2004;
Given the demands in a democratic system Bianchini and Parkinson, 1993).
that public policies show a return to the tax- However, the methodological fallacies that
payer, cultural policy has frequently argued for flaw the validity of economic-impact studies
support on the basis of utility. Governments may constitute a secondary objection (Caust,
have certainly supported the arts and culture 2003). Arguably, the limits of utilitarianism have
“for their intrinsic value to the fulfillment of been reached. Arguments from economic utility
the human potential of their citizens. Art and may tell us how valuable the arts are as goods,
culture are, from this perspective, essential but not why they are good things. Economic-
elements to a life that is worth living…” impact studies are understandably valued given
(Cummings and Katz, 1987: 351). It can be the controversies over the arts and culture as
Peter-3383-Chapter-15.qxd 6/8/2006 7:38 PM Page 272
merit goods (Netzer, 1978; Cwi, 1982). provocative, if controversial, recent works in
Consequently, the politically expedient justifica- urban studies and cultural economics, Richard
tion for a cultural policy is an appeal to the Florida’s The Rise of the Creative Class, offers
numbers, not to its values. Yet, questions of conclusions that are remarkably similar.
value “remain at the heart of cultural policy Following are some of the major findings con-
even when they have been strangely silenced by cerning the attractiveness of localities for the
the relativizing language of economics and mar- admittedly amorphous, and relatively privi-
kets” (McGuigan, 1996: 71). What an ideology leged, “creative class:” (1) these are places that
of cultural utilitarianism does not articulate is offer stimulating cultural environments;
an understanding of the role that a cultural (2) there is a “street-level culture – a teeming
policy can play in preserving, transmitting, and blend of cafes, sidewalk musicians, and small
expanding a community’s cultural heritage. An galleries and bistros;” (3) various “hybrid
alternative ideology of merit good would argue spaces” and “small venues” exist like coffee
that culture, “like parks, libraries and schools, shops, restaurants and bars, art galleries, book-
provide benefits all out of proportion to the stores, alternative theaters for film and live per-
amount of their subsidies and merit support formances; (4) there is an authenticity that
because of their contributions to the general comes from “historic buildings, established
welfare” (Mulcahy, 1986: 46). neighborhoods, a unique culture scene”
(Florida, 2002: 95, 166, 183, 226).
Understandably, Florida’s findings have
Culture and Creative Cities been celebrated in the cultural sector in the
support that they offer for investments in a
Much research has argued that the true signif- variety of “lifestyle amenities” that include
icance of the arts may not be in their direct or comprehensively developed recreational, edu-
indirect economic effects as much as in their cational, entertainment, and artistic infra-
more induced and qualitative effects (Mulcahy, structures. It would appear that the “creative
1982b). In the case of cities, culture adds a class” is attracted to cities with a cultural (not
dimension of attractiveness that, while difficult simply an arts) policy that emphasizes spheres
to quantify, is very real. for public interaction rather than institutional
First, cultural organizations are important edifices.
to a city’s self-worth – enhancing its attractive-
ness to residents, visitors and businesses. Cities
are particularly desirable for firms that have
PUBLIC CULTURE AND
highly educated and well-paid personnel.
POLITICAL CULTURE
Other things being equal, a culturally vibrant
city is likely to improve its economic health
because businesses are attracted to locations At the risk of stating a self-evident truth, a
with strong cultural amenities. Second, certain nation’s public policies reflect the historical
industries such as publishing, advertising, experience and value systems that have charac-
broadcasting and fashion depend on the con- terized its societal development. This orientation
centration of cultural workers found in many toward politics involving “general attitudes
urban centers. Third, cities are also venues for about the system and specific attitudes about the
fashion and cuisine. Given their education, role of the self in the system” is termed a nation’s
income and lifestyle, cultural consumers are political culture (Almond and Verba, 1965: 13).
attracted to the goods and services provided in “To understand the cultural politics of a country,
urban markets. Fourth, culturally vibrant cities one must first understand its political culture.
often attract “urban homesteaders” who are State policies toward the arts are shaped by wider
committed to community improvement. beliefs about how government ought to be con-
These research findings were established ducted and what it should try to do” (Ridley,
in the 1980s. Interestingly, one of the most 1987: 225). Depending on their political
Peter-3383-Chapter-15.qxd 6/8/2006 7:38 PM Page 273
cultures, governments vary in the way that their Lang challenged Malraux, substantially rewriting
cultural policies are conceptualized. “This vari- the department’s initial mission statement in
ety reflects not only differing national traditions order to place creativity and creation above
in the organization of public functions and the democratisation, though in practice there was a
delivery of public services, but differing philoso- good deal of continuity between the two pio-
phies and objectives regarding the whole area of neers” (Looseley, 2003: 228; Looseley, 1995).
culture and the arts” (Cummings and Katz, French intellectuals frequently position them-
1987: 4). Cultural policies, then, need to be selves as the last exponents of high culture and
understood not simply as administrative mat- esthetic discernment in the face of the onslaught
ters, but as reflecting a societal Weltanschauung; of mass-entertainment culture appealing to the
that is, a world view that defines the character of lowest common denominator of taste. In partic-
a society and how its citizenry define themselves. ular, it is American popular culture that is iden-
With regard to the variety of institutions and tified as the enemy of esthetic excellence and
programs that have been created to implement a French cultural traditions (Ahearne, 2002). For
cultural policy, their esthetic values reflect many French intellectuals, Euro-Disney was not
deeper popular cognitions. In this sense, cultural just a theme park, but a “cultural Chernobyl.”
policies represent a microcosm of social and However, it should be noted that resistance to
political world views. At the risk of oversimplifi- Americanization and the more excessive asser-
cation, certain Weberian “ideal types” of cultural tions of France’s status as an alternative cultural
patronage rooted in different socio-historical hegemon has not been without criticism
traditions can be identified (Mulcahy, 2000b; (Fumaroli, 1999).
Zimmer and Toepler, 1996). These ideal types Despite much rhetorical hyperbole, what is
are useful for understanding why nations attach important to note is that the French see culture
an importance to supporting cultural activities as an essential part of the sens civique; that is, a
through public intervention (or choose not to sense of civic solidarity that has traditionally
do so). However, it is important to remember been regarded as a distinguishing characteris-
that any ideal type is a generalized construction tic of French society, especially when com-
that may not reflect particularized exceptions. pared to the Anglo-Saxon alternative (Konig,
With these caveats in mind, four cultural 1995: 95–106). Although there is constant
value-systems are posited: patrimony (cultural debate about the content of French cultural
preservation); identity (cultural sovereignty); policy, “at least there are cultural policies, at
social welfare (cultural entitlements); libertar- least there is public patronage of the arts, both
ian (cultural laissez-faire). Frequent reference national and local, at least the French remain
is made to four representative nations: France, self-conscious about their creative genius”
Canada, Norway and the United States, (Gildea, 1996; 232).
respectively. France may be the preeminent patron state
in the preservation and promotion of its cul-
tural heritage; in sum, France is l’Etat Culturel.
Patrimony However, Austria could be described as a
“Kulturstaat” and Vienna as a “Kulturstadt.”
The hegemonic status of French culture; that is, Also, Italy is endowed with so rich a cultural
the claim of its language, literature, philosophy heritage that its preservation absorbs almost all
and fine arts as universal accomplishments of the resources available. Each of these
worthy of preservation and emulation, is a widely nations, characterized in many ways by their
accepted principle of French political dis- cultural heritages, pursues a cultural policy in
course. Whether Andre Malraux (1959–1969) in which its patrimony is a central concern. This
DeGaulle’s government, or the Socialist Jack is realized through a highly developed system
Lang (1981–1986 and 1988–1993), ministers of of subsidies for the arts throughout the
culture have often employed aggressive policies country and direct management of national
to promote these hegemonic claims. “Ostensibly, cultural institutions.
Peter-3383-Chapter-15.qxd 6/8/2006 7:38 PM Page 274
What is most notable about Canada’s cultural The four nations surrounding the Baltic Sea, as
policy is the importance in political discourse well as Iceland, represent a distinctive eco-
of the relationship between cultural identity nomic and political unity in their shared com-
and political sovereignty. For the Royal mitment to social-democratic principles and
Commission on National Development in the the welfare state. Each is a small nation; each
Arts, Letters and Sciences, chaired by Vincent has a huge measure of ethnic and religious
Massey from 1949 to 1951, it was an article of homogeneity; their militaries are small; their
faith that there was an identifiable Canadian foreign relations are pacific and are distin-
identity. In particular, opposition to American guished by high per-capita levels of humanitar-
mass culture was the basis of its cultural iden- ian assistance to the international community.
tity (Meisel, 1989: 22–23). Indeed, many Most notably, there is a common commitment
Canadians mobilized to fight the North to equality, egalitarianism, and equity that is
American Free Trade Association as a threat to realized through long-standing public policies.
national cultural values. Cultural policy is part of a much broader array
This cultural “crisis-mentality” is under- of governmental efforts to provide a high-level
standable in a nation of 27 million adjacent to quality of life that is accessible, sustainable, and
one of 270 million whose popular culture representative. In sum, the Nordic cultural
dominates the world’s entertainment venues. model reflects this ideological superstructure;
The fear of “cultural annexation” can best be cultural democracy is an analog of social
understood when one realizes that 95 percent democracy (Dueland, 2003).
of their movies, 75 percent of their prime-time For example, Norway is a social-democratic
television, 70 percent of radio airtime (despite state with a well-articulated policy of cultural
the latter two media having Canadian-content democratization and a strong emphasis on pro-
quotas), 80 percent of the magazines, 70 percent moting maximum feasible accessibility to its
of the books are American products (Acheson national cultural heritage. Funding for culture
and Maule, 1999: 16). in Norway is rooted in a social-democratic
The issues of Canadian cultural identity, ideology that views government as the primary
with a vertical cleavage of asymmetry with the actor for providing social goods. “The welfare
U.S., is compounded by a horizontal cleavage state’s task is to make sure that the good are pre-
with Quebec and its 7 million francophones. It sent, meaning that they are created or made,
may be that an unanticipated consequence of and that the goods are distributed equally
the Massey Commission’s efforts to create a among the population” (Bakke, 1994: 124).
distinct Canadian culture was to encourage Norwegian governments in the post-World
artists and intellectuals in Quebec to achieve a War II era have accepted responsibility for public
“societé distincte;” that is, to assert the distinc- culture as a logical extension of the welfare state.
tiveness of their francophone culture and “The welfare ideology implied that ‘cultural
separate identity. Since the “quiet revolution” goods’ should be fairly distributed throughout
of the 1960s, the Quebecois developed an the country, and that the population should have
outward-looking cultural awareness along extended influence upon decisions affecting the
with strong cultural institutions. Quebec’s cultural life of its own community” (Mangset,
Ministere des Affaires Culturelles has actively 1995: 68). The welfare principle also applies to
supported the epanouissement (blossoming) the artists’ right to economic security and recog-
of its arts and literature (Mulcahy, 1995d; nizes that cultural activities – the crafts as well as
Mulcahy, 1995e). Canada is effectively a the fine arts – are “a national resource for social
“consociational society” that must adjust its and economic development” (Kangas and
national policies to accommodate the special Onsér-Franzén, 1996: 19).
status of a large, historically recognized cultural Overall, the social-democratic model views
region (Lijphart, 1977). culture as one of those rights to which all
Peter-3383-Chapter-15.qxd 6/8/2006 7:38 PM Page 275
citizens are “entitled;” that is, having a defined and 1,899 symphony orchestras among other
right, in the same sense that they are to other components of its cultural infrastructure
benefits of the welfare state. As a cultural- (President’s Committee on the Arts and the
policy commitment, the state intervenes to Humanities, 1997).
correct free-market inequalities in the distrib- The U.S. government promotes culture
ution of cultural products and opportunities through philanthropy; that is support for non-
through subsidies to national cultural institu- profit arts organizations through special pref-
tions, through sinecures to guarantee the erences in its tax code (Heilbrun and Gray,
status of artists, and through support for local 1993). For example, like all nonprofit
cultural heritage as well as for opportunities 501(c)(3)s, cultural organizations benefit from
for individual self-expression. provisions allowing corporations, foundations,
and individuals to deduct the full amount of
their charitable contributions when filing
Libertarian taxes. Also, nonprofit arts organizations gener-
ally do not pay income tax, local property
There is no “ministry of culture” in the United taxes, or local sales taxes on revenue that is
States; that is, a Cabinet-level department related to their mission. Nonprofits also
responsible for comprehensive cultural policy- receive substantial subsidies through preferen-
making and for administering a wide range of tial postal rates (Cummings, 1991: 39–41).
artistic activities. The cultural programs of the What makes American public policy unique
federal government are highly fragmented, is the extent to which the indirect mechanisms
located in a variety of administrative agencies, of tax exemptions are a means by which the
overseen by different congressional committees, government empowers private institutions and
supported by and responsive to a variety of individuals to address public purposes, includ-
interests and articulating the policy perspectives ing culture. As indicated earlier, a public policy
of discrete segments of the cultural constituency can be whatever a government chooses to do, or
(Cherbo, 1992). This institutional fragmentation chooses not to do. In this sense, the American
reflects both the diffuse nature of artistic activity government takes a laissez-faire approach,
in the United States and a fear of the effects that where culture is indirectly supported by allow-
a unified cultural bureaucracy might have on ing individuals through tax incentives to shape
artistic expression (Shattuck, 2005). the nature of cultural activities. Whether such a
The symbolic importance of the NEA in the privatized culture best serves the public’s inter-
cultural world is decidedly greater than its mon- est is a matter of political values.
etary resources (Mulcahy, 1995b). Overall, gov-
ernment is a minority stockholder in the
business of culture (Mulcahy, 1992). Public sub-
THE FUTURE OF CULTURAL POLICY
sidy from all levels of government accounts for
about 6 percent of the resources of performing
arts organizations and 30 percent of museums. The conventional wisdom of cultural policy
The American cultural organization is typically a has traditionally compared the reputedly
private, not-for-profit entity, supported by deplorable condition of public support for cul-
earned income, individual philanthropy, corpo- tural activities in the United States with an ide-
rate sponsorships and (mostly local) govern- alized conception of European public culture
ment grants; it is neither a public agency nor one (Schuster, 1989). Like most observations about
that is largely supported by public funds. The comparative public policies, however, broad
indirect public support provided by tax-exempt generalizations often disguise substantial
charitable deductions is the crucial element in exceptions. Four examples will suffice.
sustaining America’s 8,000 museums, 2,000 local First, while it is certainly true that European govern-
arts councils, 351 public television stations, 548 ments substantially support a broad array of cultural
public radio stations, 7,300 community theaters, endeavors, these subsidies have declined in recent years.
Peter-3383-Chapter-15.qxd 6/8/2006 7:38 PM Page 276
Moreover, if the esthetic dimension of public policy- In a system of mixed funding, public culture
making is more apparent in many European nations, can nurture the arts groups and cultural activ-
this aspect of public policy is not totally absent in the
United States.
ities that contribute to individual self-worth
Second, the universe of funded culture is very different and community definition, even if counting
outside of the United States, as this includes support for for less in the economic bottom-line. At root, a
what is primarily commercial in the U.S. such as film, cultural policy is about creating public spheres
books and audio-visual products. On the other hand, that are not dependent upon profit motives
many European nations are considering the reputed
virtues of privatization and searching for alternative
nor validated by commercial values. As politi-
sources of support for cultural activities. cal democracy is dependent upon the existence
Third, the role of the not-for-profit sector distin- of civil society and socio-economic pluralism,
guishes the American case from that of other nations. To cultural policy stands as an essential public
an extent unknown elsewhere, the American govern- commitment in realizing these fundamental
ment through its tax code has delegated broad policy-
making powers to private institutions in the pursuit of
preconditions.
various eleemosynary goals. The American system of
cultural patronage is, in effect, much broader and
stronger than would appear from considering just REFERENCES
the NEA’s budget or even that of all cultural agencies
collectively.
Fourth, generally public culture is associated posi- Acheson, Keith and Christopher Maule. 1999. Much
tively with identity and heritage, with how people define Ado about Culture: North American Trade
their communities and see themselves in the world and Disputes. Ann Arbor: University of Michigan.
in history. A cultural policy that would appear to offer Ahearne, Jeremy. 2002. French Cultural Policy
the best opportunity for public support is one that posi- Debates: A Reader. New York: Routledge.
tions itself as part of broad public culture that includes
Almond, Gabriel A. and Sidney Verba. 1965. The
the humanities, historic preservation, public broadcast-
Civic Culture: Political Attitudes and Democracy in
ing and arts education. This ideal cultural policy would
promote a sense of communal continuity and definition Five Nations. Boston: Little, Brown.
while continuing to support esthetic discourse and Anderson, James E. 1975. Public Policy-Making. New
artistic creativity. York: Praeger.
Bakke, Marit. 1994. “Centralized Decentralization in
The future of cultural policy would seem to Norwegian Cultural Policy.” The Journal of Arts
predict an increasingly inexorable demand that Management, Law, and Society, 24: 111–127.
the arts “carry their own weight” rather than Bennett, Oliver. 2004. “The Torn Halves of Cultural
rely on a public subsidy to pursue “art for art’s Policy Research.” The International Journal of
Cultural Policy, 10: 237–248.
sake” (Wyszomirski, 1995a). This “cultural
Bianchini, Franco and Michael Parkinson, eds. 1993.
Darwinism” is most pronounced in the United
Cultural Policy and Urban Regeneration: The West
States, where public subsidy is limited and European Experience. Manchester: Manchester
publicly-supported esthetic activities are University Press.
expected to demonstrate a direct public benefit Bourdieu, Pierre. 1984. Distinction: A Social Critique
(Mulcahy, 2003). Non-American cultural of the Judgment of Taste. Cambridge: Harvard
institutions are less constrained by the need to University Press.
maintain diversified revenue streams that Caust, Jo. 2003. “Putting the “Art” back into Arts
demand high levels of earned income and indi- Policy Making: How Arts Policy has been
vidual and corporate donations to compensate “Captured” by the Economists and the
for limited government appropriations. On the Marketers.” The International Journal of Cultural
Policy, 9: 51–59.
other hand, cultural institutions everywhere are
Craik, Jennifer, Libby McAllister and Glyn Davis.
increasingly market-driven in their need for
2003. “Paradoxes and Contradictions in
supplementary funds and as a justification for Government Approaches to Contemporary
continued public support. The American model Cultural Policy: An Australian Perspective.” The
of an essentially privatized culture is increas- International Journal of Cultural Policy, 9: 17–34.
ingly attractive to governments seeking to cur- President’s Committee on the Arts and the Humanities.
tail their cultural subsidies. 1997. “Creative America.” Washington, D.C.
Peter-3383-Chapter-15.qxd 6/8/2006 7:38 PM Page 277
Chartrand, Harry Hillman and Claire McCaughey. Florida, Richard. 2002. The Rise of the Creative Class.
1989. “The Arm’s Length Principle and the Arts: An New York: Basic Books.
International Perspective – Past, Present, and Foucault, Michel. 1991. “Governmentality.” The
Future.” Who’s to Pay for the Arts?: The International Foucault Effect: Studies in Governmentality.
Search for Models of Arts Support, eds. Milton C. Graham Burchell, Colin Gordon, and Peter
Cummings Jr. and J. Mark Davidson Schuster. Miller. London: Harvester Wheatsheaf.
Cherbo, Joni M. and Margaret J. Wyszomirski. 2000. Fumaroli, Marc. 1999. L’Etat Culturel: Une Religion
The Public Life of the Arts in America. New Jersey: Moderne. Paris: Fallois.
Rutgers University Press. Galligan, Ann. 1993. “The Politicization of Peer
Cherbo, Joni M. 1992. “A Department of Cultural Review Panels”, in Judith H. Balfe, ed., Paying the
Resources: A Perspective on the Arts.” Journal of Piper: Causes and Consequences of Arts Patronage.
Arts Management, Law, and Society, 22: 44–62. Chicago: University of Illinois Press.
Cohen, Randy, William Schaffer and Benjamin Gans, Herbert J. 1999. Popular Culture and High
Davidson. 2003. “Arts and Economic Prosperity: Culture: An Analysis and Evaluation of Taste, Revised
The Economic Impact of Nonprofit Arts and Updated Edition. New York: Basic Books.
Organizations and theirAudiences.” Journal of Gildea, Robert. 1996. France Since 1945. New York:
Arts Management, Law and Society, 33(1): 17–31. University of Oxford Press.
Cummings, Jr., Milton C. 1991. “Government Heilbrun, James and Charles Gray. 1993. The
and the Arts: An Overview”, in Public Money Economics of Art and Culture: An American
and the Muse, ed. Stephen Benedict. New York: Perspective. New York: Cambridge University Press.
W.W. Norton. Kangas, Anita and Jill Onsér-Franzén. 1996. “Is
Cummings, Jr., Milton C. and Richard S. Katz. 1987. There a Need for a New Cultural Policy Strategy
The Patron State. New York: Oxford University in the Nordic Welfare State?” The International
Press. Journal of Cultural Policy, 3: 15–26.
Cummings, Jr., Milton C. 1982. “To Change a Katz, Richard S. 1982. “Public Broadcasting and the
Nation’s Cultural Policy,” in Public Policy and the Arts in Britain and the United States”, in Kevin V.
Arts, eds. Kevin V. Mulcahy and C. Richard Mulcahy and C. Richard Swaim, Public Policy and
Swaim. Boulder: Westview Press. the Arts. Boulder: Westview Press.
Cwi, David. 1982. “Merit Good or Market Failure: Konig, Hans. 1995. “A French Mirror.” Atlantic
Justifying and Analyzing Public Support for the Monthly, December.
Arts,” in Kevin V. Mulcahy and C. Richard Swaim, Langsted, Jorn, ed. 1990. Strategies. Studies in
Public Policy and the Arts. Boulder: Westview modern cultural policy. Aarhus: Aarhus University
Press. Press.
DiMaggio, Paul J., and Francie Ostrower. 1992. Race, Lebovics, Herman. 1999. Mona Lisa’s Escort: André
Ethnicity, and Participation in the Arts. Washington, Malraux and the Reinvention of French Culture.
D.C.: National Endowment for the Arts and Ithaca: Cornell University Press.
Seven Locks Press. Lijphart, Arend. 1977. Democracy in Plural Societies.
Dorian, Frederick. 1964. Commitment to Culture. New Haven: Yale University Press.
Pittsburgh: University of Pittsburgh Press. Litt, Paul. 1992. The Muses, the Masses, and the Massey
Dorn, Charles M. 1995. “Privatization of the Arts Commission. Toronto: University of Toronto Press.
and the Public Interest: An Issue for Local Arts Looseley, David. 2003. “Back to the Future: Rethinking
for Local Arts Agencies.” The Journal of Arts French Cultural Policy, 1997–2002.” The Inter-
Management, Law and Society, 25: 182–91. national Journal of Cultural Policy, 9: 227–234.
Dueland, Peter. 2003. Nordic Cultural Model. Nordic Looseley, David. 1995. The Politics of Fun: Cultural
Cultural Institute, Copenhagen. Policy and Debate in Contemporary France.
Dueland, Peter. 2001. “Cultural Policy in Denmark.” Washington D.C.: Berg.
The Journal of Arts Management, Law and Society, Mangset, Per. 1995. “Risks and Benefits of Decen-
31: 34–57. tralisation: The Development of Local Cultural
Dworkin, Ronald. 1985. “Can a Liberal State Support Administration in Norway.” International Journal
Art?” in Ronald Dworkin, A Matter of Principle. of Cultural Policy, 2: 67–86.
Cambridge: Harvard University Press, 221–233. McGuigan, Jim. 1996. Culture and the Public Sphere.
Dye, Thomas R. 2005. Understanding Public Policy, New York: Routledge.
Eleventh Edition. New Jersey: Pearson Prentice Meisel, John. 1989. “Government and the Arts in
Hall. Canada”, in Who’s to Pay For the Arts, eds. Milton
Peter-3383-Chapter-15.qxd 6/8/2006 7:38 PM Page 278
C. Cummimgs, Jr. and J. Mark Davidson Schuster. Judith H. Balfe and Margaret Jane Wyszomirski.
New York: American Council for the Arts. New York:Praeger.
Miller, Toby and George Yudice. 2002. Cultural Mulcahy, Kevin V. and C. Richard Swaim. 1982a.
Policy. Thousand Oaks: SAGE Publications. Public Policy and the Arts. Boulder: Westview
Mulcahy, Kevin V. “Entrepreneurship or Cultural Press.
Darwinism: Perspectives on the American System Mulcahy, Kevin V. 1982b. “Culture and the Cities,” in
of Cultural Patronage,” Journal of Arts Management, Kevin V. Mulcahy and C. Richard Swaim, Public
Law and Society, (Fall 2003): 165–184. Policy and the Arts. Boulder: Westview Press.
Mulcahy, Kevin V. 2002. “The State Arts Agency: An Myerscough, John. 1988. The Economic Impact of the
Overview of Cultural Federalism in the United Arts in Britain. London: Policy Studies Institute.
States.” Journal of Arts Management, Law and Netzer, Dick. 1978. The Subsidized Muse: Public
Society, 32: 67–80. Support for the Arts in the United States. Cambridge
Mulcahy, Kevin V. 2000a. “Cultural Imperialism and University Press, Cambridge and New York.
Cultural Sovereignty: U.S.–Canadian Cultural President’s Committee on the Arts and Humanities.
Relations.” The American Review of Canadian 1997. Creative America: A Report to the President.
Studies, 30: 181–206. Washington, D.C.
Mulcahy, Kevin V. 2000b. “The Government and Perret, Jacques and Guy Saez. 1996. Institutions et vie
Cultural Patronage: A Comparative Analysis of culturelles. Paris: La Documentation Francaise.
Cultural Patronage in the United States, France, Peters, B. Guy. 1996. American Public Policy: Promise
Norway, and Canada,” in Joni M. Cherbo and and Performance, Fourth Edition. New Jersey:
Margaret J. Wyszomirski, The Public Life of the Arts Chatham House.
in America. New Jersey: Rutgers University Press. Pratt, Andy C. 2005. “Cultural Industries and Public
Mulcahy, Kevin V. and Margaret Jane Wyszomirski. Policy: An Oxymoron?” The International Journal
1995a. America’s Commitment to Culture. of Cultural Policy, 11: 31–44.
Boulder: Westview Press. Public Law 89–209. 1965. Congressional Declaration
Mulcahy, Kevin V. 1995b. “The NEA and the of Purpose Establishing the National Endowment
Reauthorization Process: Congress and Arts Policy for the Humanities and the National Endowment
Issues,” in Mulcahy and Wyszomirski, America’s for the Arts; National Foundation on the Arts and
Commitment to Culture. Boulder: Westview Press. Humanities Act of 1965. Washington, D.C.: U.S.
Mulcahy, Kevin V. 1995c. “The Public Interest and Government Printing Office.
Arts Policy,” in Mulcahy and Wyszomirski, Rafool, Mandy and Laura Loyacono. 1995 Creative
America’s Commitment to Culture. Boulder: Solutions for Funding the Arts. National Conference
Westview Press. of State Legislatures.
Mulcahy, Kevin V. 1995d. “Public Culture and Rentschler, Ruth. 2002. The Entrepreneurial Arts
Political Culture,” in Quebec Under Free Trade: Leader: Cultural Policy, Change and Reinvention.
Making Public Policy in America, ed. Guy Brisbane: University of Queensland Press.
Lachapelle. Quebec: Presses de l’Universite du Ridley, F.F. 1987. “Tradition, Change, and Crisis in
Quebec. Great Britain.” in Milton C. Cummings, Jr. and
Mulchay, Kevin V. 1995e. “Public Culture and Richard S. Katz, The Patron State. Oxford: Oxford
Political Culture: La Politique Culturelle du University Press.
Quebec.” Journal of Arts Management, Law and Robinson, P. John. 1993. Arts Participation in
Society, 25: 25–49. America, 1982–1992: A Report of the National
Mulcahy, Kevin V. 1992. “Government and the Arts Endowment for the Arts. NEA Report no. 27.
in the United States,” in Public Policy and the Prepared by Jack Faucett Associates.
Esthetic Interest, eds. Ralph A. Smith and Ronald Rouet, Francois and Xavier Dupin. 1991. Le Soutien
Berman. Aux Industries Culturelles. Paris: Department des
Mulcahy, Kevin V. 1991. “The Public Interest in Etudes et de la Prospective.
Public Culture.” Journal of Arts Management, Law Saez, Guy. 1996. “Les politiques culturelles des
and Society, 21: 5–25. villes,” in Institutions et vie culturelles, eds. Jacques
Mulcahy, Kevin V. 1986. “The Arts and Their Perret and Guy Saez. Paris: La documentation
Economic Impact: The Values of Utility.” Journal Francaise.
of Arts Management, Law and Society, 16: 33–39. Schuster, J. Mark. 2003. Mapping State Cultural
Mulcahy, Kevin V. 1985. “The NEA as Public Patron Policy: The State of Washington. Chicago: The
of the Arts,” in Art, Ideology, and Politics, eds. University of Chicago, Cultural Policy Center.
Peter-3383-Chapter-15.qxd 6/8/2006 7:38 PM Page 279
Schuster, J. Mark Davidson. 1989. “The Search for Wyszomirski, Margaret J. 2002. “Arts and Culture,”
International Models: Results from Recent in The State of Nonprofit America, ed. Lester M.
Comparative Research in Arts Policy,” in Who’s to Salamon. Washington D.C.: Brookings University
Pay for the Arts?, eds. Milton C. Cummings, Jr. Press.
and J. Mark Davidson Schuster. New York: Wyszomirski, Margaret J. 1995a. “Federal Cultural
American Council for the Arts. Support: Toward a New Paradigm?” Journal of
Shattuck, Roger. 2005. “In the Thick of Things.” Arts Management, Law, and Society, 25: 69–83.
New York Review of Books. May 26: 19–22. Wyszomirski, Margaret J. 1995b. “From Accord to
Stanbridge, Alan. 2002. “Detour or Dead-End? Discord: Arts Policy During and After the Culture
Contemporary Cultural Theory and the Search for Wars,” in America’s Commitment to Culture, eds.
New Cultural Policy Models.” The International Kevin V. Mulcahy and Margaret Jane Wyszomirski.
Journal of Cultural Policy, 8: 121–134. Boulder: Westview Press.
Sterngold, Arthur R. 2004. “Do Economic Impact Wyszomirski, Margaret J. 1982. “Controversies in
Studies Misrepresent the Benefits of Arts and Arts Policymaking,” in Public Policy and the Arts,
Cultural Organizations?” Journal of Arts eds. Kevin V. Mulcahy and C. Richard Swaim.
Management, Law and Society, 34: 166–188. Boulder: Westview Press.
Williams, Raymond. 1977. Keywords: A Vocabulary Zimmer, Annette and Stefan Toepler. 1996.
of Culture and Society. New York: Oxford “Cultural Policies and the Welfare State: The
University Press. Cases of Sweden, Germany, and the United
Wyszomirski, Margaret J. 2004. “From Public States,” Journal of Arts Management, Law, and
Support for the Arts to Cultural Policy.” Review of Society, 26: 167–193.
Policy Research, 21: 469–484.
Peter-3383-Chapter-15.qxd 6/8/2006 7:38 PM Page 280
Peter-3383-Chapter-16.qxd 6/8/2006 7:38 PM Page 281
16
Tax Policy
B. GUY PETERS
Taxation is one of the oldest and most some products and actions by taxing them
important activities of government. Govern- heavily, or attempt to reduce consumption by
ments have always needed money in order to high taxes (tobacco, for example). Taxation can
pay for the services they provide, and that has also be an important tool for managing the
meant that they needed to tax. Despite its economy, with lower taxes presumed (espe-
necessity, taxation has never been popular with cially by analysts and political parties on the
the public and therefore governments have had right) to stimulate economic growth, and
to find ways of raising the money with the higher taxes being useful for controlling infla-
minimum political opposition.1 As govern- tion. In short, taxation is a powerful instrument
ments have undertaken more functions, grow- for the public sector that can be used in a num-
ing from the Nightwatchman State to the ber of ways to move resources, both from the
contemporary Mixed Economy Welfare State, private sector to government and from various
the need for tax revenue has grown, and with segments of the private sector to others.
that the need to find politically acceptable Politically, tax policy is often a more conve-
means of taxing has also become more impor- nient mechanism for coping with complex
tant for governments. Citizens have become issues than are public expeditures, or many
somewhat more resistant to taxation, even in other forms of implementing public policy,
high-tax, high service political systems such as such as regulation (see Peters, 2001). Citizens
the Nordic countries, so government also con- talk a great deal about taxes but the process
sider more carefully other forms of revenue through which taxes are made is often more
collection, such as fees for services that do tend technical, and less visible to the public, than are
to be more palatable to the public. expenditure politics (see Steinmo and Tolbert,
Although the primary purpose of taxation is 1998). Benefits created for individuals or busi-
to raise revenue for government, tax policy also nesses through taxes tend not to involve creat-
can be used to achieve a number of substantive ing service programs, hiring employees or all
policy goals. Governments can advantage some the other trappings of “big government”, but
products and activities in the private sector by rather can be created through complex changes
taxing them more lightly than other types of in the rules used to collect taxes. These com-
income (capital gains), or by allowing the tax- plexities also create difficulties in the interna-
payer to deduct expenditures from taxable tional environment as organizations such as the
income (mortgage interest payments in many World Trade Organization attempt to maintain
countries). Governments may effectively ban fair competition in the world economy.
Peter-3383-Chapter-16.qxd 6/8/2006 7:38 PM Page 282
the same opportunities for evasion, having a For the less developed countries of the
good deal of their income in cash. Some coun- world consumption taxation tends to be
tries have established themselves as tax havens the principal form of revenue collection. If the
for corporations, with very low taxes and strict economy is highly decentralized and has a
laws preventing corporate or individual number of segments that are not conducted in
income information to tax authorities as a cash, then counting money incomes may be
means of assisting individuals who want to difficult. Further, to the extent that incomes
evade taxes to do so. are in money they may well be in cash that is
The second major alternative source for tax difficult to trace, as compared to salaries paid
revenue is consumption. The Value Added Tax, by larger firms. Therefore, these countries may
for example, is paid by consumers in European find it easier to collect excises and customs
countries whenever they purchase an item in rather than attempt to tax incomes, or even
the stores, although they may not notice attempt to impose consumption taxes on per-
because the tax is included in the price. sonal consumption, such as sales taxes.
Consumers in North America and Australia Finally, governments can also tax accumu-
pay sales taxes and goods and services taxes lated wealth. One of the most common forms of
that are added to the costs of the product on tax using this “handle” is the property tax, a
the shelf, and hence are more visible to the mainstay for local taxation in Anglo-American
consumers or taxpayers. Tariffs levied on countries. This tax also is justified on the ability
imports are also consumption taxes because, in to pay, with ownership of a house, or other
the end, they will be passed on by the importer assets, being evidence of that ability, although it
and paid by the final consumer of the prod- is also assumed that homeowners benefit more
ucts. Excise taxes – the famous sin taxes on from services such as fire protection and public
products such as alcohol, tobacco and gam- utilities than do other citizens. Inheritance taxes
bling, and simple necessity taxes on products also use wealth as a basis for collecting revenues
such as gasoline and phone service – also come as that wealth is passed between generations.
out of the pockets of the consumers of these While these taxes do appear fair, given the pos-
products. Most citizens find that they are help- session of the assets, some people (especially the
ing to fund government every time they go to elderly) find that their incomes do not match
the shops. their assets, and have to sell their homes simply
In most instances4 consumption taxes fall because of the taxes imposed. Further, heavy
more heavily on the less affluent, but may be taxes on accumulated wealth may discourage
considered fair politically because everyone in investment and savings, and hence have nega-
society should be obliged to pay at least some- tive long-term economic consequences.
thing for the maintenance of the society. The Again, governments can fine tune a tax
actual incidence of consumption taxes can be instrument in order to produce the type of
altered somewhat by careful design of the outcomes desired. For example, if there is a
taxes, e.g. not charging VAT on food or medi- desire to keep the elderly in their homes, gov-
cines, but the impact still tends to fall more ernments can provide a “homestead exemp-
heavily on those less capable of paying. tion” up to a modest amount to protect those
Consumption taxes also are generally more homeowners from tax. Likewise, property may
difficult to evade than are income taxes, and be passed on to spouses without tax, and to
have the additional advantage that government children with limited tax, but other forms of
can use stores (or casinos, or wherever) that transfer may be taxed more heavily. These tax
sell the products to collect the taxes. Given that instruments may appear to be rather blunt
the incidence of consumption taxes is general when first considered, but governments can
rather than on industry, many countries have make them into very sensitive tools for altering
been shifting their tax burdens somewhat in the distribution of income and providing
this direction in the face of pressures from incentives for certain types of economic and
globalization and tax competition. social behavior.
Peter-3383-Chapter-16.qxd 6/8/2006 7:38 PM Page 284
To this point I have been speaking about present those somewhat more systematically.
taxes per se, but governments have other As noted, tax policy tends to involve very large-
important revenue sources as well. In particu- trade-offs of these criteria. Economists have
lar, governments provide a number of goods tended to dominate policy-making for taxation,
and services for which they can charge their and hence the economic criteria discussed
“customers”. Governments have been raising below tend to be the dominant values. That
an increased amount of revenue from fees and having been said, in the public sector politics is
charges over the past several decades. This shift usually trumps, so, even though there may be
in revenue patterns has come about for several good economic arguments for a tax policy if
reasons. First, this form of revenue collection that policy produces political difficulties, it
tends to be more acceptable politically than may well be rejected for a program that econo-
does taxation, given that there is a direct mists might consider inferior. Economists
exchange of money for the service rather than do vote, but there are fewer of them than of
the perception that tax money goes into some many other groups in society and politicians
vague general fund. Secondly, governments in democratic systems have to appeal to
also use fees as a means of rationing consump- majorities.
tion of certain types of services that might other- Many of these criteria for taxes are distribu-
wise be overused by the public. Finally, the tive, and are concerned with the economic,
ideology of the New Public Management and political and ethical consequences of who pays
of the political right has emphasized the capa- taxes. That assessment is, however, somewhat
city to use more business-like mechanisms incomplete without a simultaneous assessment
within the public sector. of where the money is being spent. A tax, such
The choices among these tax handles, as well as the Social Security “contribution” in the
as among individual forms of taxation within United States, that is negatively redistributive
those broader categories of taxes, involve eco- because it is not charged on earned incomes
nomic, political, administrative and ethical above $90,000 (2005) and not charged on
criteria. One of the most important strategies incomes from investments, is used to fund a
that governments have adopted in attempting program that is slightly progressive. The net
to be able to raise adequate revenue without effect, therefore, may be roughly proportional
creating large-scale resistance among taxpayers among economic classes. In general tax sys-
is to spread their tax net as widely as possible. tems have become less progressive in most of
Most governments take at least some revenues the industrialized democracies, but spending
from as many sources as possible. This broad- continues to go to the programs of the Welfare
based approach to taxation may help reduce State, and perhaps in larger proportions,
political opposition, but it may also make because of the aging of the populations, so that
achieving other policy goals through taxation the net effects may not be altered as much as
more difficult. If all aspects of the economy are expected.
being taxed, or being given tax relief, then there
is little that any one type of economic action is
likely to gain from government. Thus, taxation, Economic Concerns
perhaps more than policy areas, involves a set
of complex trade-offs among a number of Economic values tend to be considered first
competing values and critieria for assessment. when assessing taxes, and the political criteria
discussed below are often seen by economists
as introducing distortions into the more effi-
cient solutions that might be produced by
CRITERIA FOR ASSESSING TAXES
focussing more single-mindedly on the eco-
nomics of taxes. Indeed, to some analysts
The above discussion has implied the applica- any tax may be considered a distortion in
tion of criteria for assessing taxes, but I should what might otherwise be a more efficient
Peter-3383-Chapter-16.qxd 6/8/2006 7:38 PM Page 285
economy, but, given the needs for contemporary Economic analysis of taxation focuses some
governments to raise money in large amounts, of its attention on the incidence of taxes.
these distortions must be minimized rather Levying a tax on a business or on an individual
than eliminated. The tax distortions of mar- is one thing, but the question of incidence is
kets may be especially apparent when tax sys- who actually bears the burden on that tax. For
tems favor one use of money over others. For example, the corporation tax is apparently
example, tax expenditures5 for housing favor a tax on businesses, but in reality that tax is
personal homes over other forms of invest- paid by employees, by customers, and/or by
ment, even though housing may be less pro- investors. The actual incidence of the tax will
ductive than many other potential uses of depend upon factors such as the elasticity
capital. Politically, however, houses are widely of demand for the product and therefore the
held in many economies so that a tax advan- ability to pass along the tax in higher prices.
tage for this use of capital is very good politics Determining the true incidence of a tax is
indeed. important because it allows determining just
The desire to eliminate distortions arising where the burden of the tax falls within the
from taxation does not, however, necessarily economy. For example, if corporations can
imply removing all special treatments of income pass along their taxes they may be less inter-
and expenditures. Indeed, creating a tax system ested in opposing those taxes politically, and
that is free of distortions might involve intro- indeed may be more than willing to publicize
ducing a large number of special treatments and the amount that they pay as “good corporate
the creation of a very complex tax code. For citizens”.
example, although depletion allowances for oil
and gas producers have been written into tax
codes in part because of the political influence Political Criteria
of those industries, they also reflect a real eco-
nomic consideration about the short lifetime of The economic criteria above may appear scien-
returns from what can be significant invest- tific and rationalist, while the political criteria
ments. Given this and numerous other needs to discussed here are more difficult to quantify or
fine tune taxes, the resultant tax code may have link to specific analytic justifications. That
to be extremely complex if the goal is eliminat- having been said, however, these criteria are
ing distortions in the tax system. often crucial to the success of the policy. The
Economists are also concerned with the most important political criterion for taxation
incidence of taxation. That is, who finally pays is fairness, and most citizens are willing to pay
the tax? While for the personal income tax it is their taxes if they believe they are fair. People
clear that the individual is the payer, for the do not like paying taxes, and, unfortunately for
corporate tax there continues to be a debate governments, a significant propotion of the
about who pays, and under what conditions. populations in many countries do not believe
Do consumers pay corporate taxes through that the tax system to which they are a party is
higher prices, or do stockholders pay through fair. Most people believe that they and people
lower dividends, or do employees through lower like them pay too much, and other people
wages, or is it a combination? The answer is (whether the rich or the poor, or whoever) do
that it is probably a combination, but that the not pay enough. Likewise, many citizens
mixture will differ under different conditions, believe that other people in the society are
e.g. the elasticity of demand for the products evading taxes, although the respondents to
of the firm.6 The incidence of taxes is impor- surveys often lump together illegal evastion
tant for more than technical reasons. If taxes with legal avoidance of taxes through tax
are being used to attempt to affect the behav- expenditures (Seldon, 1979).
ior of firms, or to alter the distribution of Perhaps the most fundamental political cri-
income, then it is important to understand terion for taxation is that an old tax is a good
who really pays. tax. This statement reflects not only public
Peter-3383-Chapter-16.qxd 6/8/2006 7:38 PM Page 286
resistance to new taxes but also indicates the Earmarking is another of the political
strong path dependence in tax policy (Kato, strategies that can be utilized to enhance polit-
2003). For example, the property tax, especially ical support for taxes, or at least minimize
as a means of funding local governments, is a adverse political reactions. A good deal of
traditional form of revenue collection in survey evidence indicates that citizens are
Anglo-American countries and, despite a more likely to accept a tax if they can be sure
number of serious problems in administration that it is being used in a particular way. For
and equity, remains in place. The political tra- example, gasoline taxes in a number of coun-
ditions of countries in taxation persist and tries have been dedicated to highways and
make some taxes appear quite normal and other transportation purposes. In the United
acceptable to citizens, while introducing others States and Switzerland, even in periods in
might be politically infeasible. This cultural which there is strong public resistance to taxa-
impact on taxes may even affect the level at tion in general, citizens have voted for taxes so
which a tax can be used. Citizens in the United long as they could identify the actual use of the
States are accustomed to consumption taxes revenues. So-called “sin taxes” may be a partic-
levied at the state and local level, but the idea ular form of earmarked taxes. Although gov-
of a national consumption tax is considered ernments do derive some general revenues
innovative, and perhaps too innovative. from taxes on alcohol, tobacco and gambling,
As already noted, one of the most important increasingly some or all of that revenue is ded-
political criteria is minimizing the visibility of icated to particular purposes, and often to deal-
the tax to the public. Taxes are not popular, so ing with the problems created by the substance
if the money can be raised without the public or activity being taxed. Revenues from taxes are
understanding the exact impact of the tax on often used to pay for advertising against smok-
their own resources, then political resistance ing, and some revenue from alcohol is used to
can be reduced. Therefore, although econo- fund rehabiliation and advertising.
mists might prefer direct taxation, e.g. taxes on To this point I have been talking about taxa-
income, politicians might prefer indirect taxes tion, but it is important to remember that this
that are included in the price of a product is but one of the possible revenue sources for
rather than charged directly to the citizen. government. Governments have a number of
Another means of reducing visibility is to use other ways to raise revenues, notably through
the full range of taxes available, charging each charging fees for services. For example, when
at a small rate, but have the total add up to the taxation may be politically impossible, govern-
needed revenue. Citizens may see that their dis- ments can charge for services such as garbage
posable income is lessened, but not have a clear collection, recreation facilities, passports, and a
target as they might with a few large taxes. host of other activities. This source of revenue
Although most politicians in government has the obvious advantage of being directly
may like the idea of less visible taxes, some related to services, and, further, are paid only
politicians on the political right have been by the people who actually avail themselves of
opposed to reducing tax visibility. They argue the services.
that if citizens really understood how much Public lotteries are a special case of “user
they were paying in taxes they would mobilize fees” to finance public services. These gam-
politically (presumably on behalf of those con- bling programs have been successful in provid-
servative politicians) and oppose the expan- ing a great deal of money for the public sector
sion of the public sector. Constitutional rules, where they have been used. For example, in
such as requiring referenda for tax increases in the American states in which they are used, lot-
some American states, and in Switzerland, have teries provide an percent of public revenues.
been devised to ensure that citizens are indeed Likewise, in the United Kingdom the lottery
aware of the level of taxation, and have the has been used to fund a number of major
opportunity to oppose any new taxes or public works. In most instances lotteries are in
increases in rates. essence an earmarked tax, paying for services
Peter-3383-Chapter-16.qxd 6/8/2006 7:38 PM Page 287
such as education and the support of the deal of the economic activity is not traded in
elderly. Although lotteries are often critisized the market for money, attempting to collect
as being negatively redistributive, and as gov- taxes from most of the population may be
ernment sponsored gambling (see below), they rather pointless.7
are too dependable a source of income for gov- One common strategy for coping with the
ernments that have the option to use them to problem of collecting taxes is for government
ignore. to have private sector actors bear a major part
As well as minimizing the resistance to taxa- of the burden. This practice goes back to “tax
tion, redistribution is a major political concern farming” in Roman times, in which tax collec-
when making tax policy. While economists tors were private individuals given the right to
may want to ensure the neutrality of taxes, collect taxes in return for the option of keep-
politicians generally want to use the tax system ing a part of the take. In more recent times
to move money around among social groups, governments have come to depend upon
especicially social classes (for a somewhat firms to do a major part of the record-keeping
polemical view of this debate, see Shugart, 1997). and administration. Most employers have to
Thus, the income tax – the principal source of withdraw a portion of their employees’
income for industrialized democracies – tends incomes as tax and turn that over to govern-
to be at least moderately progressive. Taxpayers ment. Likewise, firms and individual propri-
pay a higher marginal rate after certain levels etors are responsible for collecting the
of income, although the strength of conserva- Value-Added Tax (VAT) and other consump-
tive politics in recent decades has tended to tion taxes for government. In the case of the
reduce the level of progression. The Bush tax VAT, this can impose a substantial adminis-
cuts in the United States, for example, have trative burden on the private sector actor,
gone primarily to the most affluent members given the often complex questions involved in
of the society, with the economic justification this tax.
that their investments will produce economic Ordinary citizens also do a good deal of
growth. work for government in collecting taxes. Many
countries now use self-assessment of taxes,
requiring citiziens to retain records and then
file annual tax forms detailing income and
ADMINISTRATIVE CRITERIA
deductions. Government will still have to
review these returns and audit a percentage to
A good tax (although some people might argue prevent fraud, but much of the work is done by
there is no such thing) must be fair, it must not the citizens. The Internal Revenue Service
distort economic activity more than is in the United States estimates that the average
absolutely necessary, and it must also be col- filler of the principal tax form (1040) spends
lectible without imposing excessive adminis- nn hours per year keeping records and another
trative costs. There is little use in having a tax mm hours filling out the form. More complex
in the law books if it cannot be collected, or if tax issues require the citizen to engage lawyers
it imposes so many costs on government to and accountants, again imposing a good deal
collect that it is not worth the effort. While the of the cost of administration on the private
fortunes of government in collecting taxes are sector.8
to some extent a function of the laws and poli- The necessity of actually administering the
cies themselves, they are also to some extent a tax code and collecting money from the public
function of the social and economic environ- may conflict with several of the economic and
ment in which the taxation is being conducted. political criteria already advanced. For exam-
Some countries have a long tradition of ignor- ple, if governments attempt to fine tune taxes
ing tax laws, and reducing evasion in these in order to prevent significant distortions in
cases may present a major challenge for gov- the economy, they increase the administrative
ernment. Likewise, in societies in which a good problems of government (and citizens) in
Peter-3383-Chapter-16.qxd 6/8/2006 7:38 PM Page 288
would like and, therefore, the overall quality of The pressures for reducing taxation are
government programs have been declining.9 obviously related to the issue of tax competi-
It is important to note that taxes are but one tion mentioned just above. The competition
of many factors involved in making decisions over taxation is often discussed in terms of
about industrial location, and perhaps not the corporate taxation, or perhaps individual
most important. Some research indicates that income taxation, but some of the important
taxes, in fact, fall rather far down the list of the contemporary issues concerning cutting are
factors that are crucial for location decisions, about the social costs of employment. When
although firms may still act as if receiving tax firms are deciding whether to add new
breaks will be the central factor in the decision. employees, in addition to the actual wage
Indeed, for some types of industries requiring being paid to the individual, the firm must
highly skilled workforces, low levels of public consider the tax costs – especially for social
revenue, usually associated with a poor educa- security taxes – that are involved with the job.
tion system, may in reality be an impediment Especially in countries where collecting direct
to persuading a firm to locate in an area. taxes is difficult,11 there has been a tendency to
Further, the range of possibilities available to impose high rates of social insurance contribu-
governments in raising revenues means that tions on both individuals and firms, so that the
sacrifices in some types of taxation may be marginal cost of the additional employee can
made up in other forms of taxation, or in fees be very high.
levied for services. The logical question then becomes whether
tax cuts are indeed an important tool for
economic management, or simply an ideologi-
Tax Cuts cally driven device to aid the more affluent to
reduce their contributions to the public purse.
Although they have been most evident in the The answer to that question may depend in
United States, many of the industrialized large part upon the political persuasion of the
democracies have been reducing some of their individual answering. The answer may also
taxes, especially taxes on personal and corpo- depend upon the nature of the economy at the
rate income. The assumption is that this time, and the need for stimulus from more cash
reduction in government revenue will result in in the hands of citizens – even the more affluent.
more spending by the public, which will, in What is clear, however, is that many, if not most
turn, increase economic growth. This strategy governments, now face increasing pressures to
is a variation on the familiar Keynsian logic of justify any taxes they may impose, or even those
enhancing effective demand as the means of they choose to retain. At the same time that
coping with a slow-down in the economy. The governments may want to reduce levels of taxa-
difference is, in part, that many of these tax tion, many are constrained by their involvement
reductions (again, especially those in the in institutions such as the European Union that
United States (Pollack, 2003)) have benefitted constrain their economic actions. Even in those
the more affluent rather than the less affluent, constrained situations, however, there are
who are most likely to spend any additional increasing pressures in some countries to reduce
income they may receive. On the other hand, taxes to address continuing unemployment,
the advocates of the contemporary tax cutting even in countries, such as Germany, with a
strategies for the top of the economy argue significant social welfare system.
that these are the people who are most likely to
invest in the economy and produce long-term
economic growth.10 For some countries the Tax Expenditures
important issues are, however, cutting corpo-
rate income taxes and other business taxes as a At the same time that there are some political
means of promoting investment and economic pressures to reduce taxation there are other
growth. changes in tax systems that are having the effect
Peter-3383-Chapter-16.qxd 6/8/2006 7:38 PM Page 291
of increasing taxes. In particular, many have used the tax system as a means of
countries (see Brixi, Valenduc and Swift, 2004) providing something approaching a guaran-
have been eliminating tax expenditures, or teed income to the working poor. Again, these
“loopholes” that had advantaged certain types benefits might not have survived the more
of activity. Even some of the more popular tax visible politics involved in adopting expendi-
expenditures, such as supports for home own- ture programs.
ership, have been reduced or eliminated to sim-
plify tax systems. The logic of tax expenditures
has been that the tax system can be as efficient,
SUMMARY AND CONCLUSIONS
or even more efficient, a mechanism for mov-
ing resources about in the economy as is public
expenditure. Allowing citizens to escape taxa- Taxation is one of the oldest areas of govern-
tion on certain types of activity, or to receive ment policy, and many of the questions raised
tax credits for others, have had something of the by the first attempts to raise revenue continue
same effects as spending, but did not require to be important today. Although primarily
governments to go through the political pain of designed to raise the money governments need
taxation to fund the program, nor the economic to fund their spending activities, taxes can
costs of managing a program.12 also be used for a variety of other purposes,
While there is certainly some evidence of advantaging some activities and disadvantag-
abuse of tax expenditures, providing tax relief ing others. Even the simple choice of what type
for some activities can also constitute an impor- of tax to impose in order to raise the revenue
tant policy tool for governments. In addition to will have consequences other than simply col-
the political advantage of lessened visibility of lecting money. Therefore, taxation involves a
the programs that are developed, using tax number of complex decisions that involve
expenditures also displaces much of the admin- political, economic, administrative and ethical
istrative burden for programs on to the individ- concerns. Different political forces will stress
ual taxpayers and can save the public sector one or another of these criteria as the most
itself a good deal of money. If we use the famil- important for any decision, but there will have
iar example of deductibility of mortgage inter- to be some trading-off of values.
est as a tax expenditure, then a major housing In the end, however, governments have to
program is being administered through the raise revenue, and their first priority must be
individual actions of taxpayers, who retain their to find the money they need. Even when faced
own information about their costs, file the with an international environment that poses
forms, and do all the work that would normally substantial economic competition for indus-
have to be done in the public sector. trial location, governments have to find a way
Tax expenditures do, however, tend to have of getting enough money to fund their expen-
negative redistributive consequences. Tax relief diture commitments. This ever increasing
from these exemptions and deductions is more need for public revenue now means that a
significant the higher the individual’s tax rates greater proportion of the burden of govern-
are. Therefore, most tax expenditures tend to ment finance is falling on average citizens and
benefit the more affluent differentially (Howard, perhaps even the less affluent in society. On
1997; Greve, 1994). These benefits available the one hand, that change in revenue collec-
through the tax system, therefore, tend to tion patterns makes funding government very
reflect the political power of the middle and democratic, and everyone is contributing
upper classes. That having been said, however, to maintaining the public sector. On the other
there are notable examples of tax expendi- hand, however, the process may now be
tures being used to produce benefits for the less fair than in the past, given that the more
poor, notably the Earned Income Credit in the affluent are in many societies now bearing
United States, and the Integrated Child Credit a smaller proportion of the total costs of
in the United Kingdom. These mechanisms governing.
Peter-3383-Chapter-16.qxd 6/8/2006 7:38 PM Page 292
NOTES REFERENCES
1. As Richard Rose has written, the golden goose must Brannland, R. and I.-M.Gren (1999) Green Taxes:
be relieved of the maximum feathers with the minimum of Economic Theory and Empirical Evidence From
squawking. Scandinavia (Cheltenham: Edward Elgar).
2. This logic is contested by the advocates of the “flat Brixi, P., C. M. A. Valenduc and Z. L. Swift (2004)
tax” , who argue that it is in fact fairer for everyone to
Tax Expenditures: Shedding Light on Government
pay the same rate (usually with few if any deductions
permitted).
Spending Through the Tax System (Washington,
3. This may appear more distributive but employees DC: World Bank).
may actually pay the tax through lower wages or the Greve, B. (1994) The Hidden Welfare State, Tax
unemployed may pay it because they will not be hired Expenditures and Welfare Policy, Scandinavian
when the marginal costs of hiring an additional worker Journal of Social Welfare, 3, 203–11.
are high. Howard, C. (1997) The Hidden Welfare State: Tax
4. A major exception would be higher taxes on luxury Expenditures and Social Policy in the United States
items such as jewelry. (Princeton, NJ: Princeton University Press).
5. A tax expenditure (see Howard, 1997; Also see Hughes, J. F. and P. Moizer (2005) Taxation and
Wildavsky, 1985) is government assistance given to a par-
Ethics, in M. Lamb, A. Lymer, J. Freedman, and
ticular industry or use of resources through the tax system
rather than through an expenditure. Politically, these may
S. James, eds, Taxation: An Interdisciplinary Approach
be more palatable given that it keeps the public budget to Research (Oxford: Oxford University Press).
smaller than pursuing the same ends through a subsidy. James, S. R. and C. Nobes (1996) The Economics of
Further, providing the support through the tax system Taxation (New York: Prentice-Hall).
tends to be less visible than doing it through the expendi- Kato, J. (2003) Regressive Taxation and the Welfare
ture budget. State: Path Dependence and Policy Diffusion
6. The ability to pass taxes along is one explanation for (Cambridge: Cambridge University Press).
why corporation income taxes have been relatively high in Lymer, A. (2005) Taxation in an Electronic World, in
countries such as Switzerland and the United States that M. Lamb, A. Lymer, J. Freedman, and S. James,
might have been expected to have their tax systems influ-
eds, Taxation: An Interdisciplinary Approach to
enced heavily by business interests.
7. In a good deal of sub-Saharan Africa and South Asia
Research (Oxford: Oxford University Press).
a significant part of the population are in subsistence agri- Peters, B. G. (2001) The Politics of Policy Instru-
culture and have little money income to be subject to an ments, in L. M. Salamon, ed., Handbook of Policy
income tax. Instruments (New York: Oxford University Press).
8. Although a cost home by the taxpayer, the principal Pollack, S. D. (2003) Refinancing America: The
reason for taxpayers engaging these professionals is Republican Anti-Tax Agenda (Albany: State
to reduce tax, hence producing a net reduction of tax University of New York Press).
costs. Radian, A. (1980) Resource Mobilization in Poor
9. Reduced tax revenue has contributed to the reduced Countries: Implementing Tax Policy (New
quality of some programs, but there have been more con-
Brunswick, NJ: Transaction Books).
scious choices to reduce levels of regulation so that again
firms may find it more attractive to locate.
Salanie, B. (2003) The Economics of Taxation
10. This is a variant of the “supply side economics” of (Cambridge, MA: MIT Press).
Ronald Reagan, based on the assumption that, if, money is Seldon, A. (1979) Tax Avoision (London: Institute of
placed in the hands of the most affluent members of Economic Affairs).
society, they will invest that money and that will, in turn, Shugart, W. F. (1997) Taxing Choice: The Predatory
create economic growth. Politics of Fiscal Discrimination (New Brunswick,
11. In particular, countries that have high levels of tax NJ: Transaction Publishers).
evasion tend to rely more heavily on indirect forms of tax- Steinmo, S. and C. J. Tolbert (1998) Do Institutions
ation. Individual employees have an incentive to see that Really Matter?: Taxation in Industrialized Demo-
the social taxes are collected, since their pensions and
cracies, Comparative Political Studies, 31, 165–87.
health insurance coverage are dependent directly upon
the taxes.
Tobin, J. (1978) Proposal for International Monetary
12. Some loopholes, such as the deductibility of keeping Reform, Eastern Economic Journal 4, 153–9.
racehorses (usually by the very affluent) might not have Wildavsky, A. (1985) Keeping Kosher: The
stood scrutiny if they had been considered as subsidies for Epistemology of Tax Expenditures, Journal of Public
the same activities as a part of the expenditure process. Policy 5, 413–31.
Peter-3383-Chapter-17.qxd 6/8/2006 7:39 PM Page 293
17
Industrial Policy in Developed Nations
RICHARD D. BINGHAM
Industrial policy is a generally misunderstood A nation’s industrial policy is the sum total of
concept. When most people think of indus- its sector specific industry policies.
trial policy, if they think of it at all, they The above definition is a modification of
think of development banks to aid manufac- one formulated by historian Otis Graham, Jr.
turing, or government policies to stimulate in his book Losing Time: The Industrial Policy
exports. They think of the central bank Debate (1992, p. 6). Graham defined industrial
lowering interest rates to stimulate economic policy as a nation’s declared, official, total effort
growth and productivity, or extending the to influence sectoral development and thus,
North American Free Trade Agreement national industrial portfolio (1992, p.6). But the
(NAFTA) to South America. Yet none of these United States (and other developed nations
actions would be part of a nation’s industrial as well) can never have a “declared, official”
policy; because industrial policy is sector industrial policy because to do so would
specific. Sector specific means that the policy require industrial planning, something the
is applied to a specific industry, like steel modern American government is unlikely to
production. do – because the American political economy
In 2002, U.S. President George W. Bush is a Broker State (term coined by Chamberlain,
placed tariffs on a variety of steel products 1941). In a Broker State:
being imported into the United States. This is a
government did not look ahead, aiming at coherent strate-
sector specific policy – applying only to the
gic objectives. Instead, the Broker State orchestrated the
steel industry. Thus, steel tariffs became a part conflict of organized interest groups. Washington was not
of America’s national industrial policy. the site of a government with its own purposes, but of a
sort of modified marketplace, a “parallelogram of pres-
sures,” or place of political exchange where groups within
INDUSTRIAL POLICY DEFINED the economy and society brought their special problems
and bargained for state-conferred benefits. The economy
aimed itself; the government’s role was to ensure that it did
Industrial policy is a nation’s official total effort not slow down, too much, for too long. The style of poli-
to influence sectoral development and thus, the cymaking was incremental and piecemeal, the govern-
national industrial portfolio (Bingham, 1998: ment’s time horizon close in, timed to the electoral cycle.
It displayed a weak sense of an overriding public interest
p. 6). But a policy or a set of policies for a
and a deferential reliance upon the electorate’s and in par-
specific industry (e.g., steel) is not, in and of ticular large corporations’ private agendas. It was a reactive
itself, industrial policy – it is an industry policy system, flexible and responsive to its admirers, aimless and
because it pertains to one specific industry. without vision to its detractors (Graham, 1992: p. 280).
Peter-3383-Chapter-17.qxd 6/8/2006 7:39 PM Page 294
A broker state can never have a “declared, New Generation of Vehicles in 1993 (to be
official” industrial policy – it can only have a de discussed later). It was in the nation’s best
facto one. And that is what the United States interest to develop a fuel-efficient, nonpollut-
has – “an unacknowledged melange of all federal, ing automobile, but this was not in the best
state, and local government policies affecting interests of the big-three auto makers.
goods producing sectors” (ibid., p. 3). So, to Also, there is a belief among some that
identify industrial policy in a Broker State it is technology-intensive industries make special
not possible to focus on what government says contributions to the long-term health of
about industrial policy, for it says nothing. national economies. Laura Tyson, President
Instead, it is necessary to focus on what govern- Clinton’s Chair of the Council of Economic
ment does in order to be able to identify indus- Advisors, believes that success in high-technology
trial policy. This is what will be done here. industries bestows national benefits in produc-
tivity, technology development, and job crea-
tion and that nations must devise industry
policies that promote high-technology indus-
WHY INDUSTRIAL POLICIES?
tries (Tyson, 1993: pp. 2–3):
A dollar’s worth of shoes may have the same effect on
Nations adopt industrial policies for a variety the trade balance as a dollar’s worth of computers.
of reasons. For example, industrial policies But … the two do not have the same effect on employ-
are sometimes developed to provide the home ment, wages, labor skills, productivity, and research – all
nation with the presence of an industry which major determinants of our long-term economic health
it deems vital to its economic future. Back in (Tyson, 1993: p. 12)
the late 1950s this was the case with Japan’s But, while the above suggest that industrial
successful development of the semiconductor policies are adopted by nations for rational, and
industry (Prestowitz, 1988: pp. 33–36; Lodge, even laudable purposes, many are not. Most
1990: pp. 76–78). It was also the reason for the common is old-fashioned protectionism. Many
United States’ attempt to develop a flat panel countries go to great lengths to protect inefficient
display industry during the 1990s (Bingham, and even corrupt industries – often for political
1998: pp. 120–126). purposes. This has historically been the case with
Market failure is an oft-cited reason for Japan, where the government goes to great
national industrial policies. In many nations, lengths to protect the inefficient agriculture, dis-
most utilities are publicly licensed monopolies. tribution, and construction industries. Diet
Thus there is little incentive for these utilities (Japanese legislature) members have close ties to
to be particularly innovative, or for that matter all three industries (Nester, 1991: pp. 44–76).
efficient. Industrial policies might thus be nec-
essary to move one of these protected indus-
tries in a direction it might have taken itself in
THE DIMENSIONS OF A NATIONAL
a more competitive environment. In industri-
INDUSTRIAL POLICY
alized nations this would be one reason for
government support for the development and
commercialization of high-temperature super- Observation shows that there are regularities in
conductivity initiatives. There is every reason the ways developed nations support specific
for governments to want to move in this direc- industries. These regularities, or dimensions, are
tion but there are few incentives for public util- a useful way to view the various policies toward
ities to be directly involved in these efforts. industries that advanced nations have adopted.
Industrial policies are often adopted to One of the editors of this volume, Professor Jon
correct negative externalities and/or to pro- Pierre, suggested that one potentially useful way
mote the national interest. These were certainly of viewing or classifying industrial policies was
reasons why the Clinton administration (in the to make the distinction between “offensive,
United States) initiated the Partnership for a strategic elements of industrial policy, on the
Peter-3383-Chapter-17.qxd 6/8/2006 7:39 PM Page 295
one hand, and defensive, ‘bail-out’ types of case may be. The most commonly used tool in
policy on the other.”An excellent suggestion and response to Too Big to Fail is the bailout.
one which will be followed here. Bailout can be in the form of some favored
Both offensive and defensive policies are treatment (e.g., special government contract),
common to most industrialized nations. By loan guarantee, or direct cash transfer. Two
offensive policies we mean policies which are examples from the United States come imme-
designed to develop or enhance an industry’s diately to mind: First is America’s bailout of
position in the international economy. These the airline industry immediately following the
industries are often, but not always, in the tragedy of 9–11. The second is George H.W.
high-technology arena. Defensive policies are Bush’s bailout of the American savings and
less elegant and less justifiable in today’s era of loan industry.
free(er) trade. They are policies designed to
protect or revive existing industries and even
Subsidies
specific enterprises. They are often the result of
intense political pressure to “do something” to
This dimension of industrial policy was diffi-
preserve jobs.
cult to isolate because subsidies are often tools
It is suggested here that there are seven
to achieve other defensive or offensive indus-
dimensions to the industrial policies of devel-
trial policy dimensions, and thus are not
oped nations – three which can be classified as
dimensions themselves (a government subsidy
defensive policies and four as offensive.
simply directs resources to preferred indus-
tries). For example, subsidies were used to sup-
Defensive policies Offensive policies port and develop Airbus Industries. But the
Protectionism Urban development/ development of Airbus Industries and its series
Too Big to Fail redevelopment of Airbus aircraft was an offensive industrial
Subsidies Aggressive unilateralism policy designed to bring a new technology
National defense (aviation) to the nations supporting Airbus,
High technology with the eventual goal of competing with
Boeing and the United States.
In the United States, subsidies were origi-
DEFENSIVE POLICIES nally used as bailout strategies for American
agriculture – first in 1933 in response to the
Protectionism Great Depression and again in 2002. But these
bailouts became permanent industry policies.
Protectionism broadly defined is simply the And this is the distinction. Traditional subsi-
erection of some form of barrier to free trade. dies are a tool (with a discrete time limit) while
These barriers come in many forms, including subsidies, in the industrial policy sense, are
protective tariffs, import restrictions, volun- essentially permanent.
tary export restrictions, and other similar
obstacles. One clear example of old-fashioned
protectionism is George W. Bush’s tariffs on
OFFENSIVE POLICIES
imported steel, designed to prop-up America’s
steel industry.
Urban Development/Redevelopment
Too Big to Fail Policies designed to support and aid the real
estate and construction industries are common
Too Big to Fail is a policy dimension in response components of the industrial policies of
to significant industries, or even firms, whose nations throughout the world. But what is
failure is unacceptable because it would create unique about this dimension is that in many
havoc in the country or in the industry, as the cases real estate and construction industries
Peter-3383-Chapter-17.qxd 6/8/2006 7:39 PM Page 296
are the (industrial policy) beneficiaries of less than in the past, it is still significant. Dual
policies implemented for other purposes. use technologies funded by the Defense
In other words, they are unintended conse- Advanced Research Projects Agency (DARPA)
quences. This has been the case in the United provide such benefits.
States with President Jimmy Carter’s urban
policy, Urban Development Action Grants
Promoting High Technology Industries
(UDAG), and with Ronald Reagan’s economic
recovery package, the Economic Recovery and
Industrial policy through the development of
Tax Act of 1981 (ERTA). It may also be the case
high technology is the final dimension of indus-
in Korea (dating back to the 1970s) with the
trial policy. The rationale was stated earlier.
various measures to limit the growth of Seoul
Technology-intensive industries make special
and promote balanced regional development.
contributions to the long-term health of
national economies. In the United States, devel-
Aggressive Unilateralism oping a new generation of automotive vehicles
met this criterion. The same is true of the
Aggressive unilateralism represents an assertive impending development of the mag-lev train.
trade stance for a home nation in demanding
that trading partners reduce barriers to the
home country’s exports and foreign invest- DEFENSIVE INDUSTRIAL POLICIES
ments. A classic example of aggressive unilat-
eralism is U.S. President Bill Clinton’s 1995
Protectionism as a Dimension
attempt to open the Japanese market to U.S.
of Industrial Policy
autos and automotive products. A more recent
example is the European Union’s efforts to force
Developed nations adopt protectionism as a
the United States to remove its steel tariffs and
dimension of their industrial policies when the
repeal long-standing export subsidies.
impacted industries are able to apply sufficient
pressure to the political system to potentially
Industrial Policy through affect future electoral outcomes.
National Defense
Old-fashioned protectionism appears to be
In the United States particularly (although in creeping back into American and European
other nations as well), industrial policy has industrial policies after years of liberal trade
been a product of defense research and devel- actions (multi-lateral free trade agreements).
opment for decades. Historically, spin-offs into Protectionism here is called old-fashioned
the civilian market from military products has because it brings to memory the impacts of the
produced these impacts. The aircraft industry protectionist Tariff Act of 1930 (commonly
is an obvious example. Jet engines were devel- known as “Smoot-Hawley”) in the United
oped by the military during World War II States, establishing the highest general tariff
(WWII) and were significantly improved over rate structure the country has ever experi-
the next 20 years. The C5A transport project enced. By one estimate, the tariff on dutiable
provided the technological breakthrough for imports, on average, reached 60 percent.
the high-bypass-ratio engine used in commer- Country after country retaliated, driving
cial wide-bodied jets. And, with the KC-135 American exports down from $5.2 billion in
tanker aircraft, military R&D and procurement 1929 to $1.7 billion in 1933. The Great
reduced the cost of commercial development Depression deepened and became global.
and tooling for the Boeing 707 (Pascall and Smoot-Hawley is generally acknowledged to
Lamson, 1991: pp. 65–73). be one of the major causes of the Depression
While evidence today suggests that the rate (Bingham, 1998: p. 70). The Smoot-Hawley
of civilian spillover from military spending is attitude is back with the Democratic (party)
Peter-3383-Chapter-17.qxd 6/8/2006 7:39 PM Page 297
candidate for president making serious These steel consuming industries saw their
protectionist noises during the 2004 campaign. earnings decline by $601 million, while the
Example. In late 2002, the price of a ton of earnings of steel producers, mining companies,
hot-rolled steel fell below $200; low steel etc., rose by $67.4 million, Thus, the overall
prices, in part, causing massive lay-offs in the economic impact of the tariffs was trivial –
American steel industry and culminating in costing the economy 0.0003 percent of output
the bankruptcy of 15 steel makers. Steel com- (Weisman, “Tariffs Help Lift,” 2003).
pany executives and union officials claimed However, the conclusions were subject to
they needed trade protection to upgrade their some debate. A study backed by steel-using
mills and become competitive again. companies concluded that, through the end of
On Tuesday March 5, 2002, President George 2000, higher steel prices had cost the country
W. Bush responded to this pressure and imposed about 200,000 manufacturing jobs.
“temporary safeguards” (the administration’s But the steel producers had their own
term for Tariffs), effective on March 20th, on key numbers. A report by University of Maryland
steel products to provide relief to those parts of business professor, Peter Morici, commis-
the U.S. steel industry that had been most sioned by the steel industry, claimed that
damaged by import surges. These tariffs ranged 16,000 steel jobs were resurrected, more than
from 8 to 30 percent on 10 separate steel prod- 30,000 when steel suppliers were included
ucts for a period of three years – presumably (Allen and Weisman, 2003).
enough time for the U.S. steel industry to get But, in the scheme of things, the studies were
back on its feet (Rust never sleeps, 2002). While virtually irrelevant. In November 2003 the
the move was expected, it was somewhat unusual WTO issued its final ruling against the steep
in that, according to a White House news release steel tariffs President Bush had imposed (ruling
(www.whitehouse.gov), “Free trade is a corner- them illegal). European Union (EU) Trade
stone of President George W. Bush’s agenda to Commissioner Pascal Lamy threatened to
help generate jobs for American workers, open impose $4 billion in retaliatory tariffs on 1,866
markets to American products and services, and U.S. products if the tariffs were not ended, with
spur economic growth” (p. 1). $2.2 billion in tariffs to be imposed immedi-
The tariffs were particularly beneficial to steel ately. The U.S. exports were selected for their
companies and workers in three states – Ohio, own political resonance and included products
Pennsylvania, and West Virginia – all of which such as Harley-Davidson motorcycles, South-
were closely contested in both the 2000 and eastern textiles and Florida citrus fruit
2004 presidential elections. While steel has only (Weisman, EU Trade Negotiator, 2003).
about 160,000 workers, its extraordinary politi- President Bush gave in. On December 4,
cal clout comes from its concentration in these 2003, Bush revoked the tariffs he had imposed
three key political states. on imported steel, averting a potential trade war
The move infuriated the Europeans, as more with Europe. In a prepared statement, read by
than one-third of their $4 billion of steel exports his press secretary, the president said “These
to America were hit by the tariffs, prompting a safeguard measures have now achieved their
swift complaint to the World Trade Organization purpose, and, as a result of changed economic
(WTO) (the WTO eventually ruled against the circumstances, it is time to lift them … U.S. steel
tariffs). companies are now once again well-positioned
In late September 2003, about half way to compete both at home and globally.” Minutes
through the tariffs’ planned three-year dura- after the statement, the European Union can-
tion, the U.S. International Trade Commission celled plans to launch the retaliatory $2.2 billion
(ITC) issued an assessment of the tariffs’ in sanctions (Entous and Palmer, 2003).
impact. The report concluded that the tariffs But, of course, the steel tariffs were never an
had cost the U.S. economy about $30.4 million economic issue – they were a political issue.
a year, with steel users facing sharply higher From the beginning, Bush’s economic advisors
steel prices in the early months of the tariffs. had advised against the tariffs, but his political
Peter-3383-Chapter-17.qxd 6/8/2006 7:39 PM Page 298
advisors (recognizing the centrality of Ohio, life is so important that failure is unthinkable.
Pennsylvania, and West Virginia to the 2004 This is exactly what happened with the U.S.
presidential election) convinced him to impose savings and loan industry.
the tariffs. By the fall of 2003, however, there was Example. Examples of Too Big to Fail
near consensus in the White House that the tar- abound. The case selected here is far from the
iffs should go. The revocation marked a rare latest incidence of Too Big to Fail, but was
about-face for the Bush Administration and the selected because of its enormous proportion.
action, although expected, brought angry reac- The savings and loan bailout in the United
tions from labor unions, steel executives, and States cost the American taxpayers over $200
politicians in the steel-producing states. On the billion – an amount equal to about 3 percent of
other hand, executives from steel-consuming gross domestic product. The savings and loan
industries, who had lobbied hard for the tariffs’ fiasco forced the closing of about one-third of
lifting, praised the action (Weisman, “Bush the S&Ls in the United States.
Rescinds Tariffs,” 2003). From the 1930s through the 1960s, savings
and loans in the United States grew like weeds,
although their functions were limited to lend-
Too Big to Fail ing for home mortgage loans and liabilities
restricted almost entirely to savings deposits.
The doctrine known as “Too Big to Fail” is an Passbook savings accounts were a stable and
industrial policy dimension adopted by industri- low-cost source of funds. During this lengthy
alized nations that protects the largest corpora- period it was profitable to “borrow short and
tions and industries from failure – typically with lend long.” That is, to use short-term savings
bailouts. accounts to fund long-term, fixed interest rate
mortgage loans. But, as it turned out, “borrow
Too Big to Fail is a policy in response to firms or short and lend long” was a formula for disaster.
industries whose dominant position is so large There were a number of reasons for this (see
that their failure would create havoc in the indus- Bingham, 1998: pp. 55–56), but chief among
try or the country, as the case may be. This was them was the period of high and volatile inter-
the case with the 1983 failure of the Continental est rates during the late 1970s and early 1980s.
Illinois Bank. It was such an important part of This was a period of superinflation. To combat
the banking industry, particularly in the this inflation, the Federal Reserve forced inter-
Midwest, that failure was never a possibility. est rates to unprecedented levels, severely
Too Big to Fail also concerns matters of affecting the financial condition of the savings
public convenience. Penn Central Railroad, and loan industry. When interest rates sky-
with all of its commuter trains, was Too Big to rocketed, S&L net operating income plum-
Fail because of the millions of commuter pas- meted. It was here that “borrow short and lend
sengers on the East Coast. Events like the fail- long” came unglued.
ure of Penn Central would have created not The federal government tried to cope with
only economic havoc but political havoc as the situation by giving the S&Ls new powers.
well. The same can be said of the airline bailout The Carter administration began, and the
in the United States following 9–11. Reagan administration accelerated, reductions
Similarly, Chrysler was Too Big to Fail for in regulations in the industry patterned after
political reasons. It could not have been allowed reductions in government regulations of the
to fail because Chrysler parts suppliers, banks, airlines, trucking, railroads, and communica-
and dealerships were spread throughout the tions. The policy was a disaster.
country – virtually in every House and Senate Throughout most of the 1980s, S&L prob-
district in America. The pressure on Congress lems were like the proverbial hot potato – no
was simply too great to allow failure to occur. one wanted to touch it; not the Federal Home
Finally, Too Big to Fail is also a policy Loan Bank Board, not the Reagan administra-
response to failing industries whose position in tion, not Congress. They all hoped that by
Peter-3383-Chapter-17.qxd 6/8/2006 7:39 PM Page 299
protected industry in Japan. And it is the of the nation’s population. The average farm
Japanese consumers who take the hit. In many size was 155 acres and farming income was
cases, food goods of the store shelves in Japan often the only source of household wealth.
cost five or six times the world market prices. In contrast, today there are only 2.2 million
Permanent agricultural subsidies are a sig- American farmers, making up just two percent
nificant component of Japanese farm industry of the population. Average farm size is now 435
policy. Agricultural price supports account for acres, but most commodity production comes
about one-fourth of MAFF’s annual budget, from large, mass-producing firms. Farm house-
with about 60 or 70 percent of all agricultural hold income is now about 115 percent of non-
products receiving price supports. For the farm income (Allen, 2002).
individual farm families these price supports Yet American farm policy has remained rel-
have been a blessing indeed. Largely through atively unchanged, and very expensive, over the
income received from supports, farm income years in spite of a drastically reduced farm
has risen from about 70 percent of nonfarm population. It is easy to understand farm influ-
income in 1950, to 90 percent in 1970, 100 per- ence in the 1930s when farmers had such a
cent in 1975, and is about 115 percent today. dominant role in society, but it is not as easy to
A second major component of Japan’s agricul- understand their influence today.
tural industry policy is protectionism. Nowhere It is because of the peculiarity of America’s
is this more obvious than with rice. The rice bicameral legislature that farm influence has
market is monopolized by the powerful nokyo remained strong. The 435 members of the
(Central Union of Agricultural Cooperatives), U.S. House of Representatives are appor-
which sells about 95 percent of the crop at fixed tioned based on the states’ population. Thus,
price – a price about five times the international populous California has 53 representatives
market level. Rice is heavily protected by both while rural North Dakota has 1; New York has
tariffs and import quotas. In fact, Japan justifies 29 while Vermont has 1. But in the upper
its protectionist policies toward rice by claiming chamber, the Senate, each state has equal
it is a central aspect of Japanese national culture representation – two senators. Thus the most
(Bingham, 1998: pp. 182–183). Political scientist rural state has the same numerical influence as
William Nester concludes that: the most populous. So numbers do not always
count. In addition, many of the largest states
By neoclassical economic measures, Tokyo’s farm policies
have been an immense failure since protection annually
have significant farm areas and thus sympathetic
adds $65 billion to Japan’s food bill … By neomercantilist senators – California, Texas, Florida, Illinois,
measures they have been highly successful, the political and Ohio, to name a few. Thus farm influence
trade-off of protectionism for votes has allowed the LDP has remained strong.
to remain in office and continue the industrial policies America’s contemporary farm policy tool –
that helped transform Japan from a developing country
into the world’s manufacturing, financial, and technolog-
price supports – is rooted in the events associated
ical powerhouse (Nester, 1991: p. 207). with the end of World War I, but actually began a
few years later with the Great Depression. During
The United States In the United States, the the decade 1910–1920, American farms pros-
national government’s policy toward the agri- pered, with an increased demand for food
cultural industry is a product of the industry’s from a growing American urban population
historical dominance in society and its present and from the world demand for food created
disproportionate influence in the American by the war. But, after the war, food prices
political system. plummeted as European producers recovered.
Modern American agricultural policy was a American farm leaders then called for a
product of the Depression and the recovery national program to support farm prices.
measures of President Franklin Roosevelt’s The Depression gave the farmers what they
New Deal. At the time, farmers were a political wanted with the Agricultural Adjustment
force to be reckoned with. There were over 6 Act of 1933. The solution to rapidly falling
million farmers in the United States, one-quarter farm incomes was primarily price supports,
Peter-3383-Chapter-17.qxd 6/8/2006 7:39 PM Page 301
achieved through dramatic reductions in action against perceived trade barriers in order
supply. Supply controls for staple commodities to enhance their export markets and slow the
included payments for reduced planting and movement of jobs offshore.
government storage of market-depressing sur-
pluses when prices fell below a predetermined Aggressive unilateralism is a power play by
level. This combination of price supports and developed nations to force trading partners to
supply management, with several modifica- make trade concessions that they do not wish
tions, was the essence of federal farm policy to make. The trade stance is aggressive because
through 1996 (Effland, 2000: 23–24). it is backed by threats of retaliation. It is uni-
The Federal Agricultural Improvement and lateral because (1) the home nation unilater-
Reform Act (FAIR) of 1996 legislated a dramatic ally decides that the actions of a trading
shift in federal assistance to farmers. The Act was partner are unfair, and (2) it requires that
in response to both budget-cutting measures in its partner unilaterally liberalize without any
Congress and a desire by farmers for more pro- reciprocal concessions from the home nation
duction flexibility. The hope was to move farm- (Bayard and Elliott, 1994: p. 1).
ers into a free market system. FAIR removed the Example. One of the most highly publicized
link between money income support payments attempts (more discussion on this later) at
and farm prices by providing for seven annual aggressive unilateralism over the past several
fixed, but declining “production flexibility con- decades was U.S. President Bill Clinton’s 1995
tract payments”, whereby participating produc- actions designed to open Japanese markets to
ers would receive government payments largely U.S. autos and auto parts, and to convince the
independent of farm prices. Farmers had much Japanese to purchase more U.S. parts for their
greater flexibility to make planting decisions, transplant assembly operations in the United
with the elimination of acreage idling programs States. The June 1995 agreement was the cul-
and given the freedom to plant any crop on con- mination of almost two years of on-again-off
tract acres. As a result, producers were forced to again negotiations in which, ultimately, the
rely more heavily on the market as a guide for threat of retaliation by the United States seems
production decisions. Farmers also bore a to have played a role in the final reaching of an
greater risk because payments to them were fixed agreement.
and unrelated to market prices. The conflict began on May 6, 1995, when
But the economic situation for farmers President Clinton’s economic advisor recom-
changed in the late 1900s and it became clear mended that he impose significant sanctions
that American farmers could not succeed in a against Japanese imports in retaliation for
free market. Thus, in 2002, Congress passed Tokyo’s refusal to open up its auto markets to
the Farm Security and Rural Investment Act of the United States. The recommendation came
2002 which moved farm policy back to where the day after negotiations between U.S. Trade
it had historically been. The key features of the Representative (USTR) Mickey Kantor and the
Act were a return of direct payment supports MITI (Ministry of International Trade and
for farm products and a series of counter- Industry) Minister Ryutaro Hashimoto broke
cyclical payments to covered commodities down. Several days later President Clinton
whenever the effective price was less than the ordered an unfair trade complaint to be filed
target price (The 2002 Farm Bill). with the World Trade Organization (WTO).
The United States then formally notified the
WTO that it would file its case within the next
OFFENSIVE INDUSTRIAL POLICIES 45 days.
In late May the United States announced
Aggressive Unilateralism that its sanctions for failing to reach an agree-
ment by the end of June would be a 100 per-
Developed nations, particularly those with sub- cent punitive tariff on 13 Japanese luxury car
stantial trade deficits, are taking more aggressive models. The tariffs, worth about $5.9 billion,
Peter-3383-Chapter-17.qxd 6/8/2006 7:39 PM Page 302
would be retroactive to May 20. The U.S. foreign cars should increase by about 200
imposed deadline was June 28, 1995. during the first year of the agreement and
As has been typical with so many of the should reach 1,000 by the year 2000 (Bingham,
U.S.–Japanese negotiations in the past, agree- 1998: pp. 92–83).
ment was not reached until the very last But, of course, very little of this ever
minute – the day before sanctions were to go occurred, as is often the case with negotiations
into effect. The agreement came after three between the U.S. and Japan. The Japanese
days of intensive talks between Kantor and market is difficult to crack. Reaching agree-
Hashimoto. With the new agreement in hand, ments like this U.S.–Japanese Auto Accord of
the United States did not impose the threat- 1995 literally takes years. Then it takes several
ened sanctions. more years before measurable results can
In the agreement, the major Japanese car occur. In this there were no significant results,
makers, with MITI guidance, announced plans and when the Clinton administration pushed
to open up their dealerships to American cars. for an explanation, the Japanese reply was that
All five major automakers said that they would the severe economic downturn in Japan was
reassure their dealers that they are totally free responsible and not Japanese policy. President
to deal in products of other companies, both at Clinton was not willing to push the issue fur-
home and abroad. This is a far cry from past ther in the final years of his administration.
practices, where only seven percent of Japanese Nor was George W. Bush – especially after his
dealers handle foreign cars. unhappy experience with the steel industry.
With regard to autos and parts, Toyota
agreed to increase North American production
by 200,000 units from 1994 to 1998. Nissan Urban Development/Redevelopment
agreed to consider producing transmissions in
Tennessee at a future date and will establish a Industrial policies which ultimately aid in
forklift engine plant in the United States with renewal and development of cities are common
an annual production of 20,000 units. Honda to industrialized nations. Unlike most other
will start up a parts development division for industry policies, however, many of the policies
the purchase of after-market parts. Mazda which promote cities, or spur renewal, are the
will expand the number of vehicles produced unintended consequences of other policies.
by Ford plants in the United States, and
Mitsubishi will invest an additional $300 million The development of world-class cities has
in U.S. production facilities (the Japanese com- been a crowning achievement of industrialized
panies, in announcing these plans, were quick societies during the 20th century. One only has
to claim that these moves were already in to visit London, New York, Tokyo, Paris, Berlin,
the works and were not in response to U.S. or Seoul to recognize this. Much of the vitality
pressures). of these and other urban areas is due to gov-
What were the outcomes that the Clinton ernment policies promoting cities. This is true
Administration expected? Washington esti- not only for world-class cities, but for smaller
mated that by 1998 Japanese transplants would urban places as well.
have increased their auto production in the Back in 1991, Washington Post journalist Joe
United States by 550,000 units, increased pur- Garreau examined the development of new
chase of U.S. parts by $6.75 billion, and raised urban places in the United States – places he
the local content of transplant-assembled vehi- called “Edge Cities.” He reported that there
cles to 56 percent. Washington also expected were more than 200 new (less than 30 years
Tokyo to import $6 billion worth of parts for old) Edge Cities in the United States. But the
assembly operations in Japan. USTR Kantor incredible statistic is: “two thirds of all
was also optimistic about Japan opening its American office facilities are in Edge Cities,
markets to Big Three autos. He said that and 80 percent of them have materialized in
the number of Japanese dealerships handling only the last two decades” (Garreau, 1991:
Peter-3383-Chapter-17.qxd 6/8/2006 7:39 PM Page 303
p. 5). And much of this development has been consequently, the real estate sector of the
fostered by the federal government, both economy boomed. The result was an astonishing
through direct subsides for real estate con- boom in urban development. Unfortunately, it
struction and through tax benefits stimulating also resulted in significant overbuilding in the
this development. real estate sector, resulting in high vacancy
It is easy enough to see why governments rates, foreclosures, and the savings and loan
single out real estate and its related industries crisis discussed earlier. For example, in terms
for favored treatment. In the United States, real of commercial space, from 1983 to 1986 there
estate employs about 7.6 million people if one were 937 million square feet in new construc-
includes all of the related activities like finance, tion starts but the market could absorb only
construction, building supplies, brick, lumber 575 million square feet. In the residential
and the like. And these workers are well repre- market the apartment vacancy rate went from
sented in all congressional districts. 5.5 percent in the second quarter of 1984 to 6.9
Example. The example discussed here is dif- percent in the first quarter of 1986. Five years
ferent from the other examples in this chapter after ERTA, Congress would conclude that the
in that the policy outcome – urban develop- ERTA tax incentives had contributed to the
ment – was an unintended consequence. overbuilding of office space, high vacancy
In 1981, the United States was suffering rates, and distorted business decisions at all
from both high inflation and high interest levels (Veres, 1986: 62–63).
rates and Ronald Reagan had just been elected What ERTA did was give several million
president. The Economic Recovery and Tax Act wealthy people the opportunity to invest in
of 1981 (ERTA) was one of the new adminis- real estate that actually produced a large cash
tration’s first legislative initiatives and an return, but, once depreciation was taken, pro-
example of its supply-side policies. It was duced a paper loss that individuals then used
designed to reduce tax rates and thus stimulate to reduce their reported ordinary income and
the economy to recover from a serious reces- thus their taxes. These tax shelters led to
sion. Both inflation and interest rates were very unfairness and inefficiencies in the tax struc-
high and the unemployment rate had been ture. Although these tax shelters could be
rising. Productivity and savings were stagnant. employed through a number of different legal
The objective of ERTA was to upgrade the entities, limited partnerships became the main
nation’s industrial base, stimulate productivity vehicle. They were attractive mainly because
and new business, lower personal taxes, and of the limited liability of the partners (the
restrain the growth of the federal government. number of limited partners reporting losses on
As a result of ERTA’s focus on investment, the their tax returns went from 2.55 million in
real estate sector of the economy was a major 1981 to 5.11 million in 1986 (with $35.5 billion
beneficiary of the law. ERTA included provi- in losses). One thing about wealthy Americans –
sions that provided incentives for real estate they are quick learners).
investors by reducing their tax liabilities. The But limited partnerships and high-income
Act also reduced the marginal rate on long- individuals were not alone in skimming the
term capital gains. Treasury; corporations were also able to take
It turned real estate into a true “tax shelter.” advantage of these tax preferences to reduce
A tax shelter in real estate is property with the their tax liabilities. Major U.S. corporations
following characteristics: (1) it can be depreci- were significantly reducing their tax liabilities,
ated over a shorter period of time than its useful paying little or no federal income taxes, while
life, (2) when sold, the profit will be taxed as a their financial reports to stockholders showed
capital gain, and (3) the tax loss deducted (prob- strong earnings.
ably over several years) is greater than the As time progressed, Congress, and even the
amount of cash that the investor has in the deal.1 Reagan administration, began to find politi-
These changes significantly enhanced the cally unacceptable inequities in ERTA. The
tax incentives for investing in real estate and, generally accepted view was that ERTA had
Peter-3383-Chapter-17.qxd 6/8/2006 7:39 PM Page 304
dioxide emissions. To have this substantial engines, using diesel fuel. All three incorporated
impact the vehicles must be sold in high volume. significant technical advances developed in the
This required that: PNGV program to reduce the energy used to
they meet all emission and safety requirements and propel the vehicle (NRC, 2001: pp. 1–19).
include all of the characteristics that result in strong But, by 2001, in spite of its successes, the
customer appeal. In addition to product redesign the program was in trouble. It became clear that all
program embraced research aimed at improving the of the elements of Goal 3 (the heart of the
effectiveness of the broad manufacturing enterprise,
program) would not be met in production/
including everything from the development of new ana-
lytical tools to the use of new materials and manufac- prototype vehicles in 2004. First, were prob-
turing processes (NRC, 2001: p. 13). lems with the engine. From the beginning, the
The specific objectives of the program were:
power train with the highest probability of
achieving the 80 mpg fuel target was the CIDI
Goal 1. Significantly improve national competitiveness
engine. But, in 1999, about midway through
in manufacturing for future generations of vehicles.
Improve the productivity of the U.S. manufacturing the program, the Environmental Protection
base by significantly upgrading U.S. manufacturing Agency (EPA) promulgated new admission
technology, including the adoption of agile and flexible standards substantially more stringent than
manufacturing and reduction of costs and lead times, those at the start of the program, requiring a
while reducing the environmental impact and improv-
radically better emission control technology.
ing quality.
The required after-treatment devices would
Goal 2. Implement commercially viable innovations significantly degrade the efficiency of the CIDI
from ongoing research on conventional vehicles.
Pursue technology advances that can lead to improve-
engine and increase its cost.
ment s in fuel efficiency and reductions in the emissions Second, the automotive market changed
of standard vehicle designs, while pursuing advances to substantially during the program. Light trucks
maintain safety performance. Research will focus on and SUVs now account for about 50 percent of
technologies that reduce the demand for energy from the automotive market and these vehicles are
the engine and drivetrain. Throughout the research pro-
gram the industry has pledged to apply those commer-
heavier and bigger than automobiles and have
cially viable technologies resulting from this research lower fuel economy. Why produce a fuel effi-
that would be expected to increase significantly vehicle cient sedan if no one wants to buy it?
fuel efficiency and improve emissions. Finally, the National Research Council com-
Goal 3. Develop vehicles to achieve up to three times mittee charged with evaluating the program
the fuel efficiency of comparable 1994 family sedans. reports that the cost premium of a PNGV-type
Increase vehicle fuel efficiency up to three times that of vehicle with a fuel economy close to 80 mpg
the average 1994 Concorde/Taurus/Lumina automo-
would be several thousand dollars more than
biles, with equivalent cost of ownership adjusted for
economics (NRC, 2001: p. 14). a competing conventional vehicle. The new
vehicles would simply not be cost-competitive
At program inception, three milestones were (ibid.).
established: (1) a technology selection in 1997, But, on the bright side, Toyota and Honda
(2) concept vehicles in 2000, and (3) produc- have already introduced hybrid-electric vehi-
tion/prototype vehicles in 2004. At the end of cles; the Toyota Prius and the Honda Insight.
1997, PNGV made its technology selections. And Ford, General Motors, and Chrysler all
The year 2000 concept vehicle goal was met planned to introduce some form of hybrid-
when the three manufacturers each introduced electric vehicle in 2004 (although it had not
concept cars – the Daimler/Chrysler ESX3, the occurred as of this writing).
Ford Prodigy, and the General Motors Precept. In late 2002, following critical interpreta-
Each company adopted a different approach to tions of the NRC reports, President George W.
their vehicles, although there were some signifi- Bush announced that his FY 2003 budget
cant PNGV-related commonalities. All three proposal directed that PNGV be terminated
concept vehicles utilized hybrid-electric power and that research be transferred from the
trains designed around small, turbo charged, Department of Commerce to the Department
compression-ignition direct-injection (CIDI) of Energy in a new Freedom Cooperative
Peter-3383-Chapter-17.qxd 6/8/2006 7:39 PM Page 307
Automotive Research Program (Freedom policy dimensions. There are probably others
CAR) with a focus on fuel cells (U.S. Depart- which the author has not discovered.
ment of Energy, December 2002) (fuel cells But there are some conclusions which can
were already a significant part of PNGV). This be reached through the dimensions and case
meant a move from high mileage automotive studies presented here. First, industrial policies
technologies with a short-term payoff to a are pragmatic, not ideological. Conservative
program which uses hydrogen as a fuel and American presidents were equally at home
will take years, or even decades, to bring to using the policy dimensions as liberals. While
market. they spoke one way and acted in another, the
actions they took had very practical hoped-for
outcomes.
CONCLUSIONS Advanced industrial societies will adopt
defensive industrial policies to protect their
In an earlier work on industrial policy in the globally inefficient/ineffective industries for
United States, the author wrote: political and not economic reasons. This was
very clear in the discussion of President
George [HW] Bush announced to the world that he did
not have an industrial policy. Presumably Jimmy Carter
George W. Bush’s decision to protect the U.S.
and Ronald Reagan would have said the same thing of steel industry in 2002. The three large steel-
themselves. Even Bill Clinton, who is consistently more producing states of Pennsylvania, Ohio, and
active in promoting specific industries than any of his West Virginia were critical to his election in
three predecessors, would presumably also deny that he 2000 and were also critical in 2004 (Bush won
has an industrial policy. And all four would be correct to
some degree. As presidents, they never articulated a
Ohio and West Virginial). His political advi-
coherent and comprehensive industrial policy. sors recommend for the tariffs while his eco-
And yet, while these presidents may not have had an nomic advisors recommended against. And
industrial policy, the country they have governed cer- when he removed the tariffs two years later it
tainly does, and it is actually coherent and reasonable. was because of the political pressure from the
But the presidents would not be alone in thinking
that the United States lacked an industrial policy. Much
steel-using states (e.g., Ohio), also critical to
of the policy community has been vocal in its criticism Bush’s 2004 reelection effort.
of any government efforts at industrial development. Finally, it appears that economic logic, and
Mainstream economists are critical because these less so political considerations, are the driving
proposals usually violate some of their theories. forces behind the adoption of offensive indus-
Conservatives are critical because they see industrial
policy as nothing more than meddling with the free
trial policy dimensions. Offensive policies
enterprise system – the only “perfect” system known to really seem to be good for the host nation.
humankind. Liberals are not much happier. They see There was compelling economic logic behind
industrial policy efforts as disconnected and uncoordi- Bill Clinton’s PNGV program. A fuel efficient
nated, sometimes working at cross purposes (Bingham, car is good for the United States, although it did
1998: p. 151).
nothing for the auto makers. There was also
And, while the author has limited knowledge compelling logic in George Bush’s replacement
of industrial policy systems beyond the United of PNGV with the FreedomCAR program – at
States, the above is undoubtedly true in many least at the time it was adopted.
industrialized societies. Yet one must be careful not to read too much
This chapter has sought to identify the into industrial policies. They do not impact
dimensions of industrial policies in developed on all industries in a nation – just a few. And
nations. Three defensive dimensions and four those few industries, while important, are hardly
offensive dimensions have been identified. at the heart of economic growth or decline.
While most of these have been discovered Industrial policies may be economically or polit-
while examining United States industry poli- ically necessary, but they are used sparingly as
cies, they undoubtedly hold elsewhere in the components of a nation’s total economic policy.
developed world. This is not to suggest that the As such, they have a small but vital place in
seven dimensions are the totality of industrial developed nations’ economic arsenals.
Peter-3383-Chapter-17.qxd 6/8/2006 7:39 PM Page 308
18
Agriculture and Food
WYN GRANT
Agriculture and food policy is characterized by contestation about the nature and extent of the
its simultaneous exceptionality and centrality. changes taking place (Coleman, Josling and
Exceptionality is to be found in the persistence of Grant, 2004).
high levels of subsidy of agricultural production
in many countries of the world, ranging, from
Japan to the United States. It is also to be found
THE POLITICS OF PRODUCTION
in the high sensitivity of consumers to any
perceived risk associated with food, even if the
actual level of risk is much lower than that Agriculture and food was for a long time the
encountered in many everyday activities. ‘You prototypical example of the politics of produc-
are what you eat’ is a frequently repeated tion and it should be emphasized that many of
mantra and citizens expect wholesome and these features of the policy-making process
affordable food to be readily available. ‘Unlike have not disappeared, even if they have been
other commodities … food isn’t viewed, read, increasingly eroded and challenged. A politics
played or worn. It enters the body and of production is concerned with the organiza-
becomes part of the consumer’ (Schlosser, tion of the production process and the battle
2002: 10). As new issues such as obesity, origi- about who should benefit from the surplus
nating in health policy, enter into discussions of value that is generated. In many sectors, it was
food policy, the centrality of the policy domain a battle between employers and employees, but
is enhanced. The controversy about genetically agricultural workers have generally been a
modified (GM) crops has also enhanced the weakly organized group, a few exceptions
visibility of discussions of agriculture and food aside. Moreover, production in farming has
policy. However, although agriculture and often been dominated by the ‘family farm’,
food policy has a number of special features, it although this as much an ideological and
is not so distinctive that it cannot be used to rhetorical label as it is a clearly defined analyt-
illustrate more general points about the policy- ical construct. Many family farms, particularly
making process. It is evident that many of the in countries where large-scale farming is com-
processes of internationalization and global- mon, such as Australia and the United States,
ization observable in many other policy sectors have developed into family controlled corpo-
can also be found in agriculture and food, even rations that have moved upwards into various
if there is considerable uncertainty and much forms of food processing. ‘Traditional’ family
Peter-3383-Chapter-18.qxd 6/8/2006 7:39 PM Page 310
farms still exist, especially in Europe, but they Table 18.1 Producer support estimates as a
are often able to survive because family labor is percentage of farm income by selected countries
(2002 provisional)
not fully costed or because the farmer has
diversified into on or off farm businesses, or Australia 5%
Canada 20%
the farmer’s partner (still most usually a European Union 36%
woman) works off the farm. Japan 59%
The politics of production in agriculture New Zealand 1%
was usually characterized by tightly knit policy Norway 71%
communities in which farming interests were Switzerland 75%
United States 19%
the policy makers and consumers and other OECD average 31%
actors were the policy takers. They were typically
Source: Organization for Economic Cooperation and
organized around an agriculture department in Development
government which had a close and clientilistic
relationship with a farmers’ organization or
organizations, usually dominated by larger-scale although these were ameliorated in the medium
farmers. The pattern in the United States was term by the phasing in of subsidies in the
somewhat different because of the importance accession states. Budgetary pressures, however,
of commodity rather than general farmers’ orga- rarely provide a sustained form of pressure for
nizations and the significance of committees and change in farm policies. Other actors soon lose
sub-committees in the legislature in the policy- interest in the complexities of farm policy and
making process. However, the key point was it is often possible to come up with a policy
that the complexity and distinctiveness of the ‘fudge’ that delivers less than it promises and is
policy-making process made it very difficult for misunderstood by those outside the agricultural
outsiders to have an effective impact on the deci- policy community.
sions that were taken. This was exemplified by Trade policy brings a new set of actors into
the Common Agricultural Policy (CAP) of the consideration of agricultural policy issues.
European Union (EU), where decisions are Industrial and other business interests are ulti-
processed through a Special Committee on mately unwilling to see trade policy negotia-
Agriculture and a tightly knit club-like Council tions wrecked by agricultural policy issues
of Agriculture Ministers, with the details of and urge national heads of government to
the policy being handled by Management become involved in bringing about a solution.
Committees accessible to special interests. Before the Uruguay Round of trade negotia-
This politics of production has by no means tions, agriculture was effectively excluded from
disappeared. Evidence of its continuation is to the General Agreement on Tariffs and Trade
be found in the high level of producer subsi- (GATT), largely at the behest of the two lead-
dies still provided to agriculture in most devel- ing agricultural powers, the United States and
oped countries (see Table 1) and the way in the EU. The negotiations in the Uruguay
which the United States backed away from a Round systematically addressed agricultural
more market oriented approach to agriculture issues for the first time. Although some changes
embodied in the 1996 Federal Agriculture were made in tariffs and agricultural subsidies,
Improvement and Reform Act when the sector what was important was the conclusion of an
encountered a period of crisis. Nevertheless, Agreement on Agriculture which embedded
the traditional politics of production has agricultural trade issues as an agenda item for
been challenged on a number of fronts. The the new World Trade Organization (WTO).
budgetary burden of the CAP became so great The agricultural negotiations in the Uruguay
in the 1980s that it threatened to bankrupt the Round were effectively concluded by a series of
EU and a number of significant adjustments private deals between the US and the EU with
were made to policy. The accession of ten new France playing a key role. This power duopoly
member states to the EU in 2004 promised a was viewed less tolerantly by other WTO
new set of budgetary challenges to the CAP, members in the Doha Round negotiations,
Peter-3383-Chapter-18.qxd 6/8/2006 7:39 PM Page 311
which finally got under way after the collapse countries and that the generous provision of
of the 1999 Seattle ministerials. The Cairns subsidies continues, might seem to be as influ-
Group of agricultural exporting countries led ential as ever. However, there is some evidence
by Australia continued to seek to shape policy. of a process of ‘hollowing out’ of these organi-
They were joined by a new group of emerging zations. As the number of farmers decline, it is
countries led by Brazil, India and China, sometimes difficult to maintain the levels of
known as G-20 or G-21 (the membership var- resources necessary to sustain an effective role
ied). The long run outcome is likely to be con- in the policy-making process. Staff numbers
tinued pressure from the Global South on the may not appear to decline that much, but often
EU, Japan, the US and Canada to lower the experienced staff members may be replaced
protectionist barriers and generous subsidies by inexperienced recent graduates hired on
they provide for their domestic agricultures. low salaries. One survival strategy might be
This in turn would further undermine the tra- to transform farm organizations into more
ditional politics of production. broadly based bodies looking after rural inter-
Evidence of the erosion of the traditional ests, recruiting the growing numbers of ‘hobby
politics of production is to be found in the farmers’ and other countryside dwellers.
weakening of the traditional domestic politics However, that can set up new tensions within
duopoly of an agriculture department in the organization over policy that then hampers
government closely linked to a national farm its effectiveness.
organization. Most countries still retain an More moderate farm organizations that
agriculture department (and the EU DG-Agri), based their strategy on a dialog with govern-
a situation that gives agriculture a privileged ment often find themselves increasingly chal-
position in relation to other sectors of industry. lenged by more militant movements, led by
It means that agriculture has its own voice farmers who have particularly suffered from
within government and the agriculture depart- processes of change, often smaller scale farm-
ment usually also has a key role in the farm ers. This has long been a feature of farm poli-
related aspects of international trade negotia- tics in France, but the more militant bodies
tions. However, there is an emergent trend for seemed to have gained even more ground,
traditional agriculture departments to be bolstered by the use of popular anti-globalization
replaced by ministries with a much broader discourses. In Britain, the National Farmers’
mandate. Thus, in Britain the Ministry of Union (the NFU) offered the classic example
Agriculture, Fisheries and Food was replaced by of an ‘insider’ group locked in a close relation-
a Department of Environment, Food and Rural ship with government. However, the emer-
Affairs (Defra), sometimes referred to by farmers gence of Defra, and the hard financial times
as the Department for the Extinction of experienced by some farmers, led to an
Farming and Rural Activity. A similar develop- increase in questioning of this strategy and the
ment has taken place in Germany, with the emergence of a direct action group based on
formation of a new ministry giving a greater the French model, Farmers for Action. The
emphasis to food quality and safety and NFU decided to move its headquarters out of
consumer concerns. There has also been a trend London and signalled that it would move away
in Europe, in part influenced by the model of from its traditional ‘insider’ strategy.
the US Food and Drug Administration, to cre- The traditional production oriented policy
ate new food safety agencies separate from the community has increasingly been challenged by
agriculture ministry, such as the Food Standards groups emerging from a politics of collective
Agency in Britain. These are matched by a new consumption. Their agendas come from outside
EU level food safety agency, the European Food farming: the environment; health; global social
Safety Authority, although its role is confined to justice. Because they do not share the assump-
risk assessment. tions that have underpinned agriculture and
The traditional farm organizations persist food policy for so long, they are able to chal-
and, given that farm ministries survive in most lenge the way in which issues have been framed
Peter-3383-Chapter-18.qxd 6/8/2006 7:39 PM Page 312
and to penetrate and disrupt issue networks modification technologies to create new crops,
based on production based paradigms of policy. new drugs and potentially new breeds of
Later in the chapter, three sets of issues that have animal. There have also been important trans-
been raised by these new actors are considered: formations in the technology of food process-
food safety; obesity; and environmental issues. ing, starting with canning and bottling and
It is first necessary to place public policy in the moving on through the development of frozen
area of agriculture and food in the context of foods that met the growing need of consumers
the other influences on policy outcomes and the for convenience foods. More recent technolo-
efficacy of policy itself. gies such as irradiation have proved too con-
troversial to deploy. In many ways, however,
food processing represents a relatively mature
set of technologies and the technological fron-
THE DRIVERS OF POLICY
tier is to be found in seeds, crops and livestock.
Under the politics of production, technology
What happens in the area of agriculture and was seen as a benefactor from a public policy
food is influenced by four main drivers of policy: perspective. Considerable investments were
made by national governments in demonstra-
• Technology
tion farms and teams of field agents to ensure
• Market structure
that new innovations were disseminated to
• Consumer preferences
farmers as quickly and as effectively as possible.
• Public policy
The development of a more technologically
intensive agriculture also had a transformative
Technology effect on its politics. Commentators are often
puzzled why a sector as economically small
Let us consider each of these drivers in turn. as agriculture in advanced countries should
Technology has had a major impact on the nevertheless exert considerable political influ-
ways in which food is grown, processed and ence. There are a number of reasons why this is
delivered to the consumer. The mechanization the case, but one that is often overlooked is the
of agriculture, involving the replacement of extent to which the suppliers of agriculture are
human, horse and oxen power by tractors and linked to its economic fate. These include the
combine harvesters, a process not complete in multinational firms who manufacture various
all parts of the world, had a major effect on farms of farm machinery, agrochemicals and
levels of agricultural productivity. It can also veterinary drugs, as well as the increasingly
have a profound effect on the nature of the concentrated group of firms who provide
product that the consumer is offered, as the seeds. At a national and international level,
shape, texture and content of a natural product these firms and their representative organiza-
may have to be changed to make it amenable to tions can supplement the lobbying activities of
mechanical harvesting and transportation to farmers’ associations. At a local level, their rep-
distant markets. Technology, however, extends resentatives, along with other service providers
far beyond the use of mechanical power to sow to agriculture such as accountants, auctioneers
and harvest crops. It embraces the shift from and lawyers, may well identify with the agricul-
hand to automated forms of milking; the ture community to the extent that they effec-
development of new seeds and new varieties of tively form part of the ‘farm vote’.
livestock; and the deployment of new veteri-
nary drugs to combat animal diseases. There
was also an important ‘chemical revolution’ in Market Structure
agriculture involving the widespread use of
fertilizers and various types of agrochemical to Changes in market structure have seen power
deal with plant diseases, insect pests, etc. Most flow down the food chain from farmers to
controversially, it involves the use of genetic processors and retailers. Although there are
Peter-3383-Chapter-18.qxd 6/8/2006 7:39 PM Page 313
some very large scale farmers who are able to company in the world but ranked only 59th in
engage in vertical integration, the typical terms of sales in the year 2000. Of course, any
farmer is not large enough to shape the market account of market structure would have to also
and is essentially a price taker rather than a take account of the role of major trading com-
price maker. Even for a large scale farmer, very panies such as Cargill that play a key role in
small fluctuations in price can have a big effect grain markets. Processors themselves have to be
on profitability. For most farmers, farming is a to some extent eclipsed by retailers, most
high volume, low margin business. Despite notably in Britain which has a highly concen-
technological advances, farming is very suscepti- trated food retailing sector. Retailers are then
ble to variations in the weather, and if global able to drive down profit margins, impose con-
warming leads to more volatile weather patterns, ditions for access to shelf space and keep deliv-
these effects could be enhanced. Weather vari- eries waiting until they are ready for them. There
ations can lead to production cycles. In one are also incipient trends towards globalization in
year, favorable weather conditions increase food retailing, although the opportunities are
output and the price falls. In the following constrained by differences in consumer prefer-
year, farmers plant less of the crop in response ences and the significance of local land use and
and the price rises, and so the cycle continues. other laws, some of which are designed to pro-
It was the need to dampen down these cycles tect smaller-scale, domestic retailers.
that was one of the factors initially prompting An important development has been the
government intervention in agriculture, increasing importance of global supply chains
although the policies then created went far which can tie, say, apple producers in South
beyond simple ‘safety nets’ which, in any case, Africa to supermarkets in Britain where
could in principle be provided by insurance consumers are reassured that products are air
markets. freighted for freshness. These supply chains
Another reason for government interven- ensure that consumer demands for the avail-
tion in agriculture was to protect potentially ability of fresh products out of season is met.
vulnerable farmers against oligopolistic The tomato is a good lens for these develop-
processors, particularly in dairying. The dereg- ments, given its ubiquity both as a fresh product
ulation of dairy markets, after an initial com- used extensively in salads and increasingly seen
petition for contracts, has often left dairy as having health giving properties. It is also used
farmers in a much less strong market position. as the basis for tomato soup, juice and ketchup
More generally, the trend towards multina- converted into tomato paste, a highly tradeable
tional organization of food processing has product, which is used extensively in a number
continued to exploit economies of scale and of popular food products such as pizza.
the exploitation of brands, a key investment in Nevertheless, the analysis of the industry by
food processing and retailing. These Fordist Pritchard and Burch (2003: 260) leads them
forms of standardized production are, how- to the conclusion that ‘Whereas the dimensions
ever, complemented by post-Fordist produc- and importance of global restructuring in this
tion of high quality, high value added products industry are profound, changes are occurring
occupying market niches and often with a within the contexts of production–consumption
strong regional identification. ‘In terms of complexes that remain organized at national
global market concentration, food and agricul- or regional scales’. They argue (2003: xi) that
tural industries are less dominated by interna- ‘Global agri-food restructuring needs to be
tionally coordinated production systems understood as an intricate set of processes oper-
linked through mechanisms such as common ating at many scales, and on many levels, rather
ownership or strategic alliances, compared than as a unilateral shift towards a single global
with industries such as electronics, automo- marketplace’. Globalization in agriculture and
biles, energy production and chemicals’ food is an emergent and contested paradigm
(Pritchard and Burch, 2003: 254). As Pritchard with a number of outcomes possible (Coleman,
and Burch point out, Nestlé is the largest food Grant and Josling, 2004).
Peter-3383-Chapter-18.qxd 6/8/2006 7:39 PM Page 314
new forms, e.g., about food safety. Yet other production oriented agricultural policies that
debates introduce issues that are new to the emphasized security of supply, secured if
policy arena, e.g., about obesity. Environmental necessary by generous subsidies to farmers.
narratives have slowly inserted themselves into Nevertheless, food policy in the Second World
debates about agricultural policy and have War and the immediate post-war period, at
influenced the development of a new para- least in those countries that experienced food
digm, that of multifunctionality. rationing, was able to use the insufficiency of
food to push policies that emphasized a nutri-
tious and balanced diet. Thus, while less food
Food Safety was available to consumers, their consumption
of it may have been healthier.
Food safety has been a long standing issue in Food safety policy is a complex area, with
debates about agriculture and food policy and a many dimensions to it, but the central story to
justification for government intervention. In be told is one of politicization and institutional
large part this is because it involves narratives reform, particularly in Europe, at the end of the
that extend beyond agriculture and food policy, twentieth century. Indeed, it has almost received
e.g., about public and environmental health an attention that is out of proportion to its
and fair treatment for consumers. Early food importance relative to other illnesses. Barling
safety legislation was particularly concerned and Lang (2003: 5) point out that ‘heart disease,
with the issue of adulteration of food and drink which fells 1.5 million Europeans prematurely
products, e.g., the watering down of milk or the each year, receives next to no attention while
introduction of foreign substances to bulk out food safety, which kills barely a thousand,
bread. Adulteration activities were common as dominates legislatures’. Nevertheless, there is a
early as the eighteenth century ‘when pepper certain compelling force in statements such as
was adulterated with glove dust, or mustard, that ‘Every day in the United States, roughly
butter and coffee were mixed up with flour, 200,000 people are sickened by a foodborne dis-
grass, radish seeds or lard’ (de Vroom, 1987: ease, 900 are hospitalized, and fourteen die’
181). It was relatively easy for campaigners to (Schlosser, 2002: 195). The United States has a
portray consumers as the innocent victims of long record of effort on food safety issues.
scams that might, at worst, be dangerous to Nevertheless, ‘The long-established US Food
them, but at least involved a form of cheating and Drug Administration and Center for
and to induce state intervention. However, one Disease Control, after all, have not been able to
consequence was that food safety regulation prevent 5,000 deaths from food poisoning every
became treated as a relatively technical and year’ (Lang and Rayner, 2003: 68).
depoliticized matter. ‘The law appeared An underlying problem is that regulatory
comparatively early in response to food adul- arrangements in many countries have had a
teration scandals, but implementation was dis- strong local focus, for understandable reasons.
persed, was low key, and involved the familiar The food industry is necessarily a dispersed
construction of regulatory offences as technical one at the point of supply because individual
and economic, rather than criminal, in charac- consumers are being served, notably in the
ter’ (Moran, 2003: 149). Although the British catering sector, where some of the most serious
system had its own special features, much of food safety incidents have occurred. There is
Moran’s comment would apply elsewhere for therefore a logic in having localized systems of
much of the twentieth century. Not just in enforcement, even if they produce inequities
Britain, mid-twentieth century ‘health thinking that result from discrepancies in resources and
was based on the assumption that the main different regulatory styles. However, modern
health problems of the UK population were food supply chains are often highly elongated,
and would continue to be caused by the insuf- so the problem may not be simply one of a
ficiency and unaffordability of food’ (Lang and caterer with poor hygiene standards. Flynn,
Rayner, 2003: 67). This in turn underpinned Marsden and Smith’s comments (2003: 40–41)
Peter-3383-Chapter-18.qxd 6/8/2006 7:39 PM Page 317
apply to Europe, but could apply equally to junior minister was forced to resign and the
North America or China: Government was forced to abandon its prefer-
ence for secrecy:
Within Europe the speed and distance at which foods
move along supply chains can make it very difficult for It is clear that with the politicization of food and new
regulators to confine a food safety problem geographi- entrants into the policy network, food policy-making is
cally. By the time regulators are aware of a problem in not what it was. The salmonella affair reveals the extent
one place, the product may have moved to another loca- to which food has become a key political issue and how
tion or locations elsewhere. The result is that regulators the consensual food policy community has been more
firefight along a supply chain of whose start and end divided and more conflictual and much more open to
points they may not even be aware. new interests (Smith, 1991: 251).
The politicization of food safety issues, particu- The impact of the salmonella affair, or of E. Coli
larly in Europe, at the end of the twentieth outbreaks, in both Europe and the US, was
century may be presented in terms of a series of relatively limited compared with the impact of
‘food scares’ fuelled by intense media interest. the emergence of BSE (Bovine Spongiform
A stylized account could, however, probably Encephalopathy), first recognized as a novel
focus on two scares, salmonella and BSE (or brain disease of cattle in 1986. The UK had expe-
‘mad cow disease’ as it popularly became rienced 180,000 confirmed BSE cases by the end
known), although an American account would of 2003, although as many as 750,000 animals
have to give a great emphasis to E. Coli. These may have entered the British food chain before
food scares were not accidents or simply the the disease was recognized and proper precau-
result of inadequate enforcement, but were tions put in place. Diseases that cross species
seen as having a systematic relationship with are relatively rare, but BSE became linked with
more intensive and industrialized forms of a new variant of Creutzfeld-Jakob Disease
agriculture. Animals kept at high densities, (CJD), an invariably fatal disease in which the
often among their own feces, mean that bacte- brain effectively deteriorates into a sponge-like
ria could spread more widely. The extensive use substance. The pathology of the disease is still
of antibiotics to keep animals healthy can lead imperfectly understood, involving as it does
to resistant strains of bacteria entering the prions, which are essentially distorted proteins.
human food chain. Schlosser argues (2002: It would appear that not everyone who con-
195), on the basis of a review of E. Coli. sumes offal from an affected animal is necessar-
episodes in the US, that ‘the nation’s industrial- ily infected, although younger people seem to
ized and centralized system of food processing be particularly vulnerable, whereas classical CJD
has created a whole new sort of outbreak, one is characteristically a disease of older people.
that can potentially sicken millions of people’. Probably the quantity of meat consumed and
The salmonella in eggs crisis in Britain in genetic considerations are intervening factors.
1988–9 prompted a new Food Safety Act and However, the key point as far as public
changes to institutional arrangements for food policy is concerned is that a number of people
safety. Salmonella in chickens was not a new (138 in the UK by the end of 2003) have
problem, but it became evident that a new strain suffered an avoidable and particularly unpleas-
had emerged that did not make birds ill and ant death as a result of consuming unsafe food
therefore did not have economic implications which was initially said by those in positions of
for a highly concentrated and industrialized responsibility, worried about the implications
segment of the food industry. ‘The recycling of for the industry, to be safe to eat. The financial
infected birds in poultry feed was thought to costs to the public purse were also consider-
have created the conditions in which a new able, estimated at £4 billion in the UK. What
dominant strand of salmonella emerged’ was undoubtedly a major policy failure was
(Orbach, 2001: 70). An attempt was made to not the result of negligence or poor judgement
shift the blame for the outbreak on to poor by individuals, but was more systemic in char-
hygiene standards lower down the food chain. acter. ‘It failed, first, because it subordinated
Food safety became a prominent media issue, a consumer protection and public health to an
Peter-3383-Chapter-18.qxd 6/8/2006 7:39 PM Page 318
economic and political agenda’. The state’s task crops, their traceability and their labelling, are
was seen as being ‘actively to support the compatible with international trade rules.
commercial and industrial objectives of farm-
ers and the food industries, and to give that
goal a very high priority within its overall Obesity
portfolio’ (Van Zwannenberg and Milstone,
2003: 28). The issue of obesity is a relatively new one in
The problem was initially seen as a purely debates about food policy, but it has profound
British one, but later developed in continental implications for the future of the sector. It is a
Europe. In 2003, an infected animal was found powerful narrative because it has close links with
in the US, albeit one that originated in Canada. health policy and is an increasing focus of activ-
BSE called into question intensive forms of ity for the World Health Organization (WHO).
agriculture that had moved away from tradi- ‘Obesity has reached epidemic proportions glob-
tional methods of raising cattle on grass and ally, with more than 1 billion adults overweight –
had used feeding methods that had made at least 300 million of them clinically obese – and
carnivores out of ruminants. is a major contributor to the global burden of
The BSE episode, reinforced by the experi- chronic disease and disability’ (World Health
ence of government mishandling of an animal Organization, 2003: 1). Obesity poses a major
health crisis in the form of the outbreak of foot- risk for serious diet-related chronic diseases,
and-mouth disease, undermined public confi- among them type 2 diabetes, cardiovascular dis-
dence in the handling of food safety policy, ease and certain forms of cancer, especially the
initially in Britain and then elsewhere in hormonally-related and large bowel cancers.
Europe. It led to the dissolution of the agricul- Chronic obesity can also contribute to
ture ministry in the UK, the creation of an inde- osteoarthritis, a major cause of disability in
pendent Food Standards Agency, the creation of adults. Obesity rates increased three times or
similar agencies in other EU states and institu- more since 1980 in some parts of North America
tional changes at the EU level. Above all, the and Europe and notably in the Pacific Islands,
food safety arena, formerly a low key and with an obesity rate of over 75 per cent in urban
highly technical area of policy, became highly Samoa. The attention given to a health issue is
politicized, a development that in turn affected enhanced if a link can be demonstrated with the
public attitudes towards GM crops and foods. health of children. ‘Childhood obesity is already
Given the media’s intense interest in any hint epidemic in some areas and on the rise in others.
of a food scare, and the way in which the new An estimated 17.6 million children under five
institutional arrangements have involved a are estimated to be overweight worldwide’
much wider range of actors, there is unlikely to (World Heath Organization, 2003: 2). Some
be a reversion to old style food safety politics in popular representations of the issue in the media
Europe. The position in the US is somewhat also seem sometimes to draw on narratives
different because of longer established regula- critical of the US, given the high proportions of
tory arrangements that seem to have retained overweight and obese individuals in the
public trust and different attitudes towards the American population.
use of technology in food production, even Genetic factors can help to explain the sus-
different cultural attitudes towards the ‘coun- ceptibility of a particular individual to obesity,
tryside’ and to food, although one has to be but they cannot explain away the overall upward
cautious about speculative claims about differ- trend. This has been caused by a combination
ence that are not backed by sufficient evidence. of reduced physical activity combined with
What is clear is that the divergent politics of increased consumption of more energy-dense,
food safety and consumption in the EU and nutrient-poor foods with high levels of sugar
the US has spilled into the arena of interna- and saturated fats. The chair of the House of
tional trade politics, with a dispute about Commons health committee in Britain has
whether EU policies on the regulation of GM claimed that a McDonald’s cheeseburger with
Peter-3383-Chapter-18.qxd 6/8/2006 7:39 PM Page 319
fries and milkshake would take a nine mile walk and other populations, have been allowed to
to burn off (Financial Times, 28 November, dry out or have been deliberately drained.
2003). Using an analogy with tobacco, political Coastal marshes have been reclaimed, destroy-
pressures are increasing to use taxes to discour- ing distinctive natural habitats. Agriculture is a
age people from eating too much sugar, salt and significant source of greenhouse gases, includ-
saturated fat. It has also been argued that food ing carbon dioxide and methane.
advertising aimed at children should be banned The accession of central and east European
and that restaurant menus should list calorie countries to the EU poses a number of threats to
levels in the items listed. The food industry has valuable natural habitats. Many of these coun-
responded by arguing that the problem is not tries are rich in biodiversity, ranging from plants
junk food, but junk diets. There are certainly to mammals. However, much of this biodi-
some interesting issues here about the relative versity has depended on the maintenance of
part to be played by individual responsibility relatively unproductive forms of farming. The
and government intervention. What foods an additional funds made available under the CAP
individual eats is very much a personal choice, may encourage the ploughing up of grassland
but it can also have wider social consequences, that has sustained particular forms of biodiver-
with obesity estimated to account for at least sity. More marginal farmers may feel that the
seven per cent of health care costs in developed funds they receive are insufficient to encourage
countries. This new food policy agenda item is them to continue farming, particularly given
likely to gain momentum. that the adoption and implementation rate of
agri-environmental schemes in these countries
is often slow. Land abandonment is often harm-
Environmental Policy and ful to biodiversity as the scrubland that develops
Multifunctionality is not supportive of a number of species. In
southern European countries, traditionally
A range of environmental issues arise from cultivated olive groves can support rich insect
modern forms of farming. Productivity oriented and bird populations, but these are not sus-
agricultural policies encouraged forms of tained if the land reverts to scrub.
farming that were environmentally damaging. Inserting environmental considerations into
Hedgerows were ripped up to create larger agricultural policy is not easy. Production
fields that were easier for modern equipment oriented approaches to agriculture have not
to plough and harvest. This in turn had an been concerned with the negative externalities
impact on biodiversity, particularly in terms of generated. Traditional policy communities have
bird populations. Chemical intensive farming not been greatly concerned with environmental
was encouraged and this led to problems of issues and when they have responded to them
water pollution in streams and rivers from the have often relied on arguments that ‘the coun-
run off from fertilizers. Modern forms of pig tryside’ is best looked after by its traditional
and dairy farming also produced serious prob- ‘stewards’, the landowners and farmers. As far as
lems of water pollution. Soil erosion is a seri- the environment is concerned, governments are
ous problem in drier climates and increased faced with a wide variety of problems and often
salinity in the soil is a significant problem in a wide range of approaches to dealing with
many countries. Agriculture is a major user of them. The ‘environmental’ lobby is fragmented
increasingly scarce water resources, with modern between a wide range of interests and perspec-
forms of agriculture often relying heavily on tives. Some groups are concerned with the
irrigation. In Australia, irrigation has permit- preservation of traditional landscapes and the
ted intensive agriculture in areas that would aesthetic value of the countryside’s appearance.
otherwise be barren but may ultimately Others are concerned with the way that farming
threaten the fragile ecology of the Murray- impacts on the marine environment through
Darling river basin. Elsewhere, wetlands, which water pollution. Yet others are concerned with
often support distinctive and endangered bird the well being of farm animals. The group that
Peter-3383-Chapter-18.qxd 6/8/2006 7:39 PM Page 320
can ensure that issues are framed in terms of its public goods that are seen to be provided
own particular approach can have a substantial include the rural landscape, the protection of
impact on the way in which policy is developed. the environment, the maintenance of viable
Groups concerned with birds are particularly rural communities as part of a cultural heritage
strong in northern Europe, with the Royal and high standards of plant, animal, occupa-
Society for the Protection of Birds in Britain tional and public health. This is a diverse list and
having more members than all the political it is not easy to devise appropriate policy instru-
parties combined. It has been successful in get- ments. Agri-environmental initiatives face the
ting farmland bird populations widely adopted problem that they may end up paying farmers
as a measure of environmental stress, although for doing something they would have done any-
it may not necessarily be the best indicator. way so that there is no value added from the
The EU has tried to respond to these diverse expenditure of public funds. The EU has placed
demands and issues by developing a new para- some faith in cross-compliance, which means
digm of agricultural policy, that of multifunc- that farmers would not receive their other sub-
tionality, sometimes known as ‘the European sidies if they did not meet certain environmen-
Model of Agriculture’. The development of this tal and other standards. However, such an
new paradigm is in part a response to the approach can require intrusive and expensive
exhaustion of the old production oriented monitoring at the farm level, replicating the
paradigm. It forms part of a long run effort by concern about the CAP that it involved high
the European Commission to convert the CAP transaction costs both for government agencies
into a more broadly based rural development and for farmers. Cross-compliance is ‘arguably
policy, with a greater emphasis on funding an inappropriate tool with which to address
through the so-called ‘Second Pillar’ that covers some of the priorities for the reform of agri-
such policies as agri-environmental schemes. enviromental measures highlighted by the
The EU is still at the beginning of its move Commission … cross-compliance measures
down the multifunctionality path and Second apply across whole market sectors within the
Pillar funding is likely to still be in the low teens CAP in an unfocused and non-targeted manner’
as a share of total agricultural spending by the (Rodgers, 2003: 287). Agri-environmental
end of the first decade of the twenty-first policy faces the problem often encountered in
century. The policy is also very contentious agriculture and food policy of deciding what the
internationally. Within the WTO it is supported policy goals and their priorities are, seeking to
as a policy approach by a loosely knit group of resolve any contradictions and then devising
‘Friends of Multifunctionality’ that includes appropriate policy instruments.
Japan, Norway and Switzerland. However, the
United States and the Cairns Group see it as a
means of perpetuating agricultural subsidies in
CONCLUSIONS
the EU and elsewhere by qualifying them for
placement in the ‘green box’ that covers subsi-
dies that are not judged to be trade distorting. It One theme that has emerged from this chapter
also sees multifunctionality as a means of erect- is the increasing intersection of agriculture and
ing new trade barriers around so-called ‘Non food policy with other areas of policy, notably
Trade Concerns’ such as animal welfare. health policy, environmental policy and trade
‘While precise definition is difficult, there policy. These phenomena are both a cause and
would none the less seem to be a general con- effect of the breakdown of traditional agricul-
sensus that key elements of multifunctional tural policy community. More space is created
agriculture are the joint production of com- for the entry of new actors, which in turn
modity and non-commodity outputs and the accelerate the breakdown of traditional policy
fact that some of these non-commodity outputs assumptions and modes of policy-making.
may be characterized as externalities or public However, it is also important to be aware of
goods’ (Rodgers and Cardwell, 2003: 11). The the limits of this process. Agriculture remains
Peter-3383-Chapter-18.qxd 6/8/2006 7:39 PM Page 321
19
Transportation and Infrastructure
KENNETH BUTTON
rail network was constructed largely on this network planning and decision-making have
premise, and the Trans-European Networks been made possible to do this as computer
initiative of the European Union (EU) has technology has developed; this has led to better
motivations of political cohesion underpin- informed decisions although not always better
ning it. Indeed, the very concept of the ‘King’s decisions.
highway’ that in various languages can be Underlying all this is the fact that trans-
found throughout Europe from the Middle portation, and its associated infrastructure, is
Ages, harks back to this idea of the role of not sought as an end in itself but rather there
roads as a tool in governance. From the 11th to is a derived demand for the services that it
the 13th century the road networks in Europe offers to fulfill some other end. It is an input
were largely within political boundaries and into industrial and social processes rather than
designed to secure local political integrity. being an output – it facilitates trade and allows
Economic analysis of transportation infra- individuals to access final goods and services.
structure as we understand it is relatively new This makes policy formulation, and the public
and can be traced back to the sea changes in administration of transportation, particularly
economic thinking of the late 18th century. challenging. In policy terms it has often meant
Adam Smith (1776) talks of its importance, that transportation has been treated institu-
and indeed one function that he concedes gov- tionally alongside other intermediate inputs
ernment should perform is to invest in major (such as energy policy or land use planning) to
infrastructure, because private concerns are exploit synergies, or within a larger setting
unlikely to be able to muster the necessary (such as part of an environmental ministry or
resources. Techniques for appraising the social department) to minimize institutional distor-
benefits of devoting resources to transportation tions. It has also led to wide variations in the
infrastructure investment, and also for defin- way responsibility for transport infrastructure
ing appropriate pricing for the use of such is allocated, with more or less devolution to
facilities, provoked much of the early develop- lower levels of government.
ment of modern microeconomics policy. This contribution is largely angled from
French engineers seeking to help move the the perspective of the now often unfashionable
national economy forward into the industrial political economy. It seeks to look at the ways
age were much concerned with the efficiency in which transportation infrastructure is viewed,
of the transportation system as well as its tech- both as a part of transport policy and in its
nical features. Dupuit (1844), and his innova- larger role in public policy, but in doing so the
tive work on cost-benefit analysis, is often cited focus is on the economic issues. It is not about
in this context, but there was an entire body of the more technical engineering debates or the
analysis conducted by these engineers during strict political trade-offs that are made.
the 1840s and 1850s.
As the railroads came on-line in the mid-
19th century new commercial appraisal tech-
WHAT CONSTITUTES TRANSPORTATION
niques (e.g., involving discounting costs over
INFRASTRUCTURE?
time) were developed, with the subsequent
move to mechanized road transport using
public infrastructure being accompanied by If is often, although not always, useful to have a
innovative social and economic assessment good idea of what one is talking about. There
tools. Technically, however, these variants on is, however, no agreed definition of what con-
cost-benefits analysis are a partial equilibrium stitutes transportation infrastructure. Indeed,
methodology, whereas the large infrastructure infrastructure in the general sense is often
projects of the later 20th century (such as the highly contextual. Whilst a politician or planner
Channel Tunnel or ‘Big Dig’ in Boston) can may think in terms of ports or roads as infra-
only truly be analyzed within a larger general structure, a logistics company may view its infra-
equilibrium framework. Methods of strategic structure to embrace its vehicles and buildings.
Peter-3383-Chapter-19.qxd 6/8/2006 7:40 PM Page 325
income and jobs from depressed areas to has been facilitated by the development of
relocate them in the prosperous regions – the containerization (unitization) since the late
‘Appalachians Effect’). Nevertheless, it often 1960s and this has led to technical standardiza-
seems to have a hold in the political psyche. tion of infrastructure in many areas – tunnel
Much of the immediate post-Second World heights, port facilities, loading/unloading bays,
War policy debate, especially in Europe and etc. The European Union has pushed measures
Asia, was understandably focused on infrastruc- to increase interoperability, intermodality,
ture reconstruction. This led to a heavy engi- and interconnectivity in terms of the creation
neering input into policy-making and, as time of Trans European networks (European
passed and reconstruction moved into develop- Commission (1989) and in financing individual
ment and renewal, this engineering focus investments. The World Trade Organization,
remained; the ethos was very much one of pro- by reducing tariffs and non-tariffs barriers,
viding infrastructure capacity to meet demand. acts to facilitate trade, with implications for
The increasing difficulty in many cases was that the overall, global demands on transportation
users of the infrastructure did not pay, or did infrastructure.
not pay directly, for its use. While resources are
abundant, and the true opportunity costs of
these actions are effectively hidden as part of a
TRANSPORT INFRASTRUCTURE AND
larger growth trend, this poses few political
ECONOMIC DEVELOPMENT
problems, although potentially economically
wasteful. As time passed, however, projects
became more costly for what they achieved One of the main, if not the dominant, on-
(simply following the law of diminishing effi- going public policy concerns with transporta-
ciency of capital). The result was a shift towards tion infrastructure involves the ties that it
management of infrastructure – which is dealt seems to have with economic development.
with in more detail later. Over the years the creation of international
But there was also another important trend at agencies such as the World Bank, and regional
work. As incomes in Western Europe, North bodies such as the European Union, have
America and parts of Asia rose, so did environ- sought to expand, up-grade, and maintain
mental awareness. First, local issues came to the transportation infrastructure as a driver or
fore (e.g., lead in petrol and traffic noise), fol- facilitator, dependent on the details of the pre-
lowed by wider geographical impacts (acid rain) vailing philosophy of the day, of economic
and then the whole question of environmental growth. The planned economies of Central
sustainability as evidence on global warming and Eastern Europe expended considerable
began to emerge. This has led to policy debates, resources on transport infrastructure, but per-
not only about the extent to which transporta- haps should ultimately be seen as prime exam-
tion as a whole should grow, but also, if it does, ples of quantity not being a good substitute for
whether specific ‘environmental friendly’ modes quality.
should be given preference. This, for example, This attitude is not a new one. In practice it
is the underlying ethos of the European goes back to the earliest states and empires, but
Commission’s (2001) medium-term approach its intellectual rationale is perhaps more mod-
to infrastructure policy and its emphasis on try- ern. Adam Smith (1776), for example, devoted
ing to move more freight to rail and urban pas- considerable attention to the role of infrastruc-
senger traffic to transit. ture and its pricing. In many ways his views fit
The increased globalization of activities has with the thinking of institutions such as the
added new dimensions to policy. In almost a World Bank, the European Bank for Economic
return to the days of the Roman Empire there is Reconstruction, and the European Union.
a focus on linking international infrastructure For instance, ‘Good roads, canals and naviga-
networks and allowing easier global move- ble rivers, by diminishing the expense of
ments of goods and people. Freight movement carriage, put the remote parts of the country
Peter-3383-Chapter-19.qxd 6/8/2006 7:40 PM Page 327
more nearly upon a level with those in the clearly manifest in the encouragement of the
neighbourhood of the town. They are upon private sector to invest in traditional transport
that account the greatest of all improvements. infrastructure, and the privatization of some
They encourage the cultivation of the remote, existing facilities. But it is also evident in the
which must always be most extensive circle of willingness of governments to allow the private
the country. They are advantageous to the development of new forms of infrastructure,
town, by breaking down the monopoly of the and especially the communications infrastruc-
country in its neighbourhood.’ ture that transports information.
But the issues involved are not perhaps as
simple as Smith and others believed, and the
empirical evidence that has emerged on the Theories of Transport Infrastructure
success of pump-priming economic develop- and Macro-economic Growth
ment with transportation infrastructure hard-
ware has not always been compelling. The exact importance of infrastructure as an
element in the economic development process
has long been disputed. Much seems to depend
Over-riding Philosophies upon the degree to which supply considera-
tions are thought important. The Keynesian
There are two broad approaches to economic economic approach, epitomized by the Harrod–
policy-making. These apply as much to trans- Domar framework, indicates that causality
port as anything else. First, there is the Anglo- runs from economic exploitation to income
Saxon approach, whereby the onus is on and infrastructure generation. In contrast,
leaving things to the market unless it is clear neo-classical economics is essentially supply-
that interventions would reduce the distor- driven and transportation and other infra-
tions of market failure – more is said on this structure are generally seen as important
in the transportation infrastructure context elements in the production function.
latter. Second, there is the continental approach, Much of the recent work follows the neo-
where government plans, in a general sense, classical mode in looking at the links between
supply to meet its objectives and only uses infrastructure provision and economic devel-
markets when these result in outcomes coinci- opment through some form of aggregate pro-
dental with these objectives. In fact, the reality duction function analysis. It has sought to see
is somewhere between these extremes and de how well it, and its individual elements, explains
facto debates focus on the degree of market economic performance. The usual approach has
orientation rather than any absolute. been to take a production function of the stan-
In terms of transportation infrastructure dard form, with public capital (which is largely
policy, the tradition has largely been one more transportation infrastructure) being one of the
akin to the continental philosophy. Government variables used to explain variation in output or
has exercised a heavy hand. Even in countries productivity.2 These types of model were used
such as the US, while private companies have to estimate the output elasticities of public
built facilities such as railroads, there has been infrastructure investments but the results gen-
immense political involvement and, of course, erated have not been highly robust, as can be
airports, seaports, and roads have largely been seen in the summary of some of the main
public sector activities. The activities of the studies set out in Table 19.1.
authorities in the former Soviet economies and Many of the early studies, especially from
of the socialist governments that remain in the US, at the national and state levels, pro-
many developing countries have been ones of vided statistically significant and apparently
direction rather than facilitation. robust evidence that well designed and oper-
Over time there have been shifts in emphasis ated infrastructure can expand the economic
with a more recent movement toward approaches productivity of an area. Aschauer (1989), for
closer to the Anglo-Saxon approach. This is most example, looking at data covering the period
Peter-3383-Chapter-19.qxd 6/8/2006 7:40 PM Page 328
1949 to 1985, concluded that a 1% increase in levels of utilization, maintenance and so on,
the public capital stock could raise total factor in addition to the stock of, and investment
productivity by 0.39%, while Munnell (1990) in, infrastructure.
tabulates output elasticities of public capital • Fourth, even within the very vague notion
derived from US studies in the 0.03 to 0.39 we have regarding what constitutes infra-
range, although with a preponderance of structure, there are numerous sectors and
results toward the upper end of the range. elements. From a policy perspective it is,
Biehl (1991) regarding the European Union therefore, important to isolate the roles of,
(EU) records similar positive findings. say, transport, energy and softer infrastruc-
More recently, these studies have been sub- ture such as law, education, business services
jected to a variety of criticisms. Gramlich and defense in influencing macroeconomic
(1994) and Button (1998) offer summaries performance.
but, very briefly, the key points include: • Fifth, there are similar issues on the output
side and Nadiri and Manuenias (1996) have
• First, while econometric studies may throw pointed to closer links between infrastructure
up positive correlation between economic investment and particular industrial sectors.
performance and the state of infrastruc- • Sixth, as more studies emerge they are pro-
ture, the direction of causation is not ducing much wider ranges of results; as
immediately clear. Wealthier areas may Morrison (1993) puts it, ‘A clear consensus
simply have more resources for infrastruc- about the impacts of infrastructure invest-
ture provision. The efforts at testing for ment has as yet been elusive, at least partly
causality are, as yet, minimal. because different methodologies generate
• Second, as seen above, the term infrastruc- varying results and implications’. A study
ture is a flexible one with no agreed by Sturm and de Haan (1995), deploying
definition, and simply taking official US and Netherlands data, for instance,
accountancy data may disguise important points to the fact that the data series in
measurement, qualitative and definition most studies looking at the economic
factors. In US work, for instance, while effects of public capital are neither station-
Aschauer and Munnell used a 4-category ary nor co-integrated and, thus, conclu-
‘core’ sub-set of the nine US Bureau of sions that public capital has a positive
Labor Statistics infrastructure categories, effect on private sector productivity are not
others have employed the full set. well founded. Equally, Jorgensen (1991)
• Third, the way in which infrastructure is has questioned the basic premise underly-
managed and priced may be as important as ing the use of a production function
the provision of infrastructure per se approach. This sensitivity may, of course,
(Winston, 1991). In terms of policy, there- go beyond simple matters of technique if
fore, account must be taken of the short-term there is, in fact, no underlying relationship.
Peter-3383-Chapter-19.qxd 6/8/2006 7:40 PM Page 329
• The studies that have been completed longer than the construction itself, and
frequently indicate rates of returns on public should be seen to act more as Keynesian
capital investments – Aschauer’s (1989) cal- stimuli than as development factors.
culations give returns of 38% to 56% – in • Secondary effects. These are on-going local
excess either of levels outside the range of economic benefits of running and operat-
any a priori expectation or of those found in ing the transportation infrastructure –
individual micro-project appraisals. employment in maintaining and repairing
the facility, in handling traffic, in policing and
so on. These secondary effects can be extre-
Local Development Effects of
mely important to some local economies in
Transportation Infrastructure
terms of employment, income and, for local
Investments
government, they can add to taxation
revenue. While important, and accepting that
While the debates over the macro significance
they can generate important multiplier
of transportation infrastructure still need final
effects, they are not the main development
resolution, the local geographical implications
rationale for infrastructure investments.
for locations adjacent to new facilities is less
• Tertiary effects. These stem from the stimu-
murky. This is important because many policy
lus to a local economy resulting from firms
questions ride at least as much on who benefits
and individuals having improved trans-
from an action as on the overall net effect of
portation services at their disposal. A large
that action. Distributional considerations are
part of modern industry relies on good
seldom insignificant.
quality transportation to support just-in-
One of the most immediate ways of gaining
time supply chain production. The impor-
an insight into the local importance of
tance of the mode varies according to
adequate transportation infrastructure is to
location and the type of production
consider economic impact studies. These are
involved, but access to high-quality and
often mandatory when investments are made
appropriate transportation infrastructure
in transportation infrastructure and essentially
can act as catalysts to attract mobile indus-
seek to assess the income and employment of
try and to nurture the growth of local
the action. Techniques vary, and do influence
firms. Technically, it can push output up
the results, but, quantitatively, the economic
the local production function and, through
impacts of an investment in transportation
resultant economies of scale and scope,
infrastructure depend primarily both on the
create centers of employment.
time frame examined and on the geographical
• Perpetuity effects. These reflect the fact that
space under review. In broad terms, infrastruc-
new forms of economic growth, once
ture enhancements have four potential eco-
started in a region, becomes self-sustaining
nomic impacts of varying duration and spatial
and may accelerate. Availability of good
coverage.
transportation links can change the entire
• Primary effects. These are the benefits economic structure of a region – technically
to a region derived from the construc- it can shift its aggregate production func-
tion or expansion of the transportation tion rather than just bring about a move-
infrastructure – the design of the facility, its ment along it as with tertiary effects. The
construction, the installation of supplemen- change from Florida being primarily an
tary hard- and software and so on. These agricultural based economy to one with
effects represent once-for-all injections a large and profitable tourist sector is an
of expenditure into the local economy, example, as are some of the mega high-
with associated employment in industries technology clusters that emerged. These
involved in infrastructure planning, con- changes would have been very unlikely
struction and development. These types of without good road and air transportation
effect are, however, transient, lasting no infrastructure.
Peter-3383-Chapter-19.qxd 6/8/2006 7:40 PM Page 330
Whilst it is interesting to isolate these beneficial have of standard impacts is not practically
effects, there remains the practical questions possible, but some general impressions can be
of how large they are, and whether they justify gleaned on the importance of air transport
major market interventions by policy makers. infrastructure from the selection of results
It is also often challenging to isolate the indi- seen in Table 19.2. The studies cited are in no
vidual components that may be important way atypical. The details of their findings dif-
when looking at specific sectors of the econ- fer, as might be expected given the geographi-
omy. There are basically three ways of getting a cal coverage they embrace, the variety of
handle on these issues: techniques used, and differences in the forms
of infrastructure investments being consid-
• It can be done through questioning those
ered. They all, however, offer fairly strong sup-
involved in the provision of the transpor-
port for the notion that infrastructure
tation services in the area or who are
investment can generate significant local eco-
involved in local industry – so-called expert
nomic benefits.
opinion analysis, or stated quantitative
analysis. This method does, however, pose
the problem of ensuring appropriate views
are elicited and that there is objectivity in INTERNATIONAL, NATIONAL, REGION
the responses obtained. It can become a AND LOCAL INTERESTS
basis for lobbying rather than for analysis.
• It can be done using local Keynesian multi-
The Interested Parties
pliers or input-output analysis. Basically
tracing through either in aggregate (multi-
Transportation infrastructure, because of the
pliers) or by sectors (input-output analy-
diversity and intensity of its implications,
sis) the implications of local expenditure
inevitably results in divergent views, and the
on transportation. It is a technique often
creation of various forms of coalitions. Some of
deployed in transportation impact assess-
these can exercise considerable lobbying power
ments. But there are issues of the geo-
outside of normal democratic processes, and
graphical coverage of the analysis and the
have what is often called ‘voice’ in debates. There
time frame to consider (e.g., which type of
has traditionally been a strong road construction
effect is actually being captured).
coalition in many developing countries, made
• Finally, there are econometric methods that
up of not only those who plan, build, and main-
make use of statistical techniques. These
tain the infrastructure, but also those that make
can be highly complex and can be struc-
use of it – the automobile and trucking associa-
tured to determine the specific implica-
tions – and those that provide the vehicles – the
tions of, say, an airport for an economy
car manufacturers. There are also well orga-
when a variety of other developments are
nized, and frequently very vocal, labor interests
taking place in parallel. But, while offering
that seek to further develop the transportation
the ability to isolate transportation effects
activities that they engage in.
on local development, they do pose prob-
These traditional types of structure, that
lems of ensuring an appropriate model
have really focused on the type of infrastruc-
specification and in using correct estima-
ture rather than its amount, have been con-
tion procedures. They can also be heavily
fronted in more recent times with groups
data intensive.
that have been more concerned about exces-
But, irrespective of the method used, the body sive investment in transportation hardware.
of evidence on the potential scale of local Environmentalists have been the most obvious
secondary and tertiary effects of most trans- of these new lobby groups, as they have sought
portation provision on economic growth is to highlight the local and global ecological
compelling. Pulling all the information that we damage that motorized transportation and its
Peter-3383-Chapter-19.qxd 6/8/2006 7:40 PM Page 331
• The Atlanta Chamber of Commerce found, from a survey 264 foreign-based firms, that availability of direct
international services was the third most important factor in location decisions. Subsequent study showed that the
number of foreign firms locating in the region from a particular country grew significantly after the introduction
of a non-stop service.
• Ernst and Young, looking at location decisions of 57 companies in Europe making decisions regarding the location
of a manufacturing plant, found that the air transportation network was the third most important factor in the
decision process. Air services were much more important for service sector companies.
• The Amsterdam Chamber of Commerce found that the availability of an airport was one of five key factors
considered in company relocation decisions.
• A survey of firms around Zurich found that 34% considered the airport as ‘very important’ and 38% as important as
a location factor.
• Loudoun Chamber of Commerce (Virginia) found that airport/freeway access was important to 68% of firms.
• A study of small business firms in the Washington area that were engaged in export activities found the availability
of easy access to international air transportation one of the six most important factors in their success.
Multipliers
• An academic study by Rietveld estimated that Schiphol Airport (Netherlands) generates about 85,000 jobs for the
country.
• A study of Vienna International Airport by Industriewissenschaftliches Institute in 1998 indicated that, on a turnover
of ATS25 billion in 1996, there was an impact of ATS11.2 billion on the local economy.
• The Institute of Social and Economic Research found that the total annual economic importance of the air trans-
portation services offered from Anchorage International Airport on local payrolls was $130 million above the $316
million for on-site activities.
Econometric
• Analysis by George Mason University, taking variations in high-technology employment across all US Metropolitan
Standard Areas, found that an airline hub in a region increases that region’s new economy employment by over
12,000.
• Brueckner, looking at the possible expansion of Chicago O’Hare airport, found that an increase of airline passenger
traffic of 50% would increase service related employment in the region by 185,000 jobs.
• An econometric study by Science Applications International of the implications of the Open Skies agreement between
Germany and the US on the regional economy around Hamburg Airport found annual gains for the regional
economy of $783,318 in 1994 prices.
• Button and Taylor examined the impact on US cities of having European services and found that employment
was systematically positively related to both the scale of services offered by the supplying airlines and the range of
destinations served.
Source: Button (2004)
associated infrastructure causes. Combined they have organized and more scientific
with these groups have been others with evidence that emerges about the wider impacts
a stronger social orientation that have ques- of transportation. Whist many of the resultant
tioned the equity of many infrastructure technical changes have been to vehicles (cars,
investment decisions that they see as largely planes, ships, etc.) there have also been new
benefiting higher income car owning groups in approaches to infrastructure design and man-
society. These types of groups question tradi- agement (e.g., traffic calming on street, sound
tional views about links between welfare, in its barriers, and tunneling) as well as more exten-
broadest sense, and the amount of mechanized sion consultation procedures.
transportation that is invested in. Over the past While many of these new pressure groups
thirty years the ‘voice’ of these environmental started at the local level, there are now often
and socially concerned groups have grown as parallel national and international groups that
Peter-3383-Chapter-19.qxd 6/8/2006 7:40 PM Page 332
located. Transportation is effective industry on perceived, has long been the subject of
wings or wheels and, as such, while most people regulations and frequently of state ownership.
like to have access to it, there are often vocal Whether regulation or ownership has dominated
objections to having the infrastructure in one’s public policy depends to some extent on the
‘back-yard’. To this end most countries have philosophical view taken, Anglo-Saxon or
devised various forms of local participation in Continental. But infrastructure also serves a
the decision-making process that seek to facili- variety of non-economic purposes. Over the
tate at least a degree of conflict resolution. To centuries roads and ports have often been seen
help structure such debates techniques such as primarily as logistic tools to wage war, with eco-
cost-benefit analysis and planning balance sheet nomic considerations coming a poor second. The
analysis have been developed and are now Roman road network had a commercial value
widely used in many countries. but was primarily seen as a way of maintaining
There are also larger interface problems at the an empire. In some cases transportation infra-
macro level. In the federal system, whilst roads structure that would have been of economic ben-
of national importance are often the domains of efit has been neglected to serve another end. It
the central government, there is often little was no accident, for example, that the French
coordination of links between states. Indeed, it road network along the Belgium border was poor
is not uncommon for one state to favor expand- in 1914; it could not sustain the advance of the
ing capacity at the same time as adjacent states German Imperial Army and directly led to the
seek to limit traffic growth, with the result that French victory at the Battle of the Marne.
infrastructure links vary in quality and capacity Combining economic, military, and social
along their length. Each country has its own considerations in assessing transportation infra-
approach to these types of problem, generally structure investment and control is complex.
closely related to other policies covering Normally a nation or super-national entity has a
such things as agriculture, land-use planning, strategic plan for the network as a whole. This
regional development, environmental protec- emerges through consultation and discussion of
tion, and industrial policy. relevant parties, and the outcome depends upon
Internationally, this problem is com- the strength of the various coalitions of interest
pounded and it is not uncommon for different that exist. In the case of the current effort to
technical standards (e.g., road pavement thick- develop a trans-European transport network,
ness, railway track standards, and air traffic concerns about economic, political, and social
control equipment) to exist on different sides integration dominate. But the US Federal
of national borders. Sometimes this is a histor- Highway System was the product of a military
ical legacy (the UK driving on the left and the need to move the US military rapidly across the
French on the right) but it has also been used country, with any consideration of macro eco-
as part of non-tariff trade barrier policy. Some nomic benefits being a poor second.
regional national groupings, notably the EU, This still leaves the question of who provides
seek to reduce them, as do some global organi- the facility. There are wide variations across
zations such as the UNs International Civil countries even for a single mode. In the US, for
Aviation Organisation and the International example, the private railway companies own
Maritime Organisation, but, other than allow- both the track and the rolling stock, in the UK
ing access on the terms of the host country, the track is state owned (after a short period of
these arrangements are the exceptions. privatization) but private railway companies
supply the rolling stock. In most of Continental
Europe both track and operating capital is state
owned.4
THE ROLE OF THE PUBLIC SECTOR
Setting aside any military or similar rea-
sons, the issue of who owns and operates
Transportation infrastructure, because of its transport infrastructure often rests on politi-
strategic economic importance, either real or cal and economic arguments. In some cases it
Peter-3383-Chapter-19.qxd 6/8/2006 7:40 PM Page 334
involves practical considerations; the private The changing ownership (or management
sector, for example, may not be sufficiently structure in the case of Crown Corporations)
coordinated to raise the funds and construct does not necessarily mean less government
the hardware. These are perhaps most pro- control over transportation infrastructure, but
nounced in less industrial countries that do rather the use of other regulatory controls. In
not have sophisticated governance structures the case of monopolies this can entail price
to allow private finance markets to operate capping, as with BAA that owns the three
with confidence. In other cases, most notably London airports and other provincial UK air-
when transportation is seen as a way of ports. This allows flexibility in pricing of indi-
attaining larger national objectives, it is some- vidual services but under a fixed, inflation
times thought that government control offers adjusted annual increase for average charges.
the most likely course to success. The idea that To limit environmental damage there are plan-
prevailed post-Second World War UK of the ning regulations for most forms of transporta-
government needing to control the ‘towering tion infrastructure that involve some form of
heights of industry’ to steer the national econ- cost-benefit assessment, embracing a wide
omy forward is a case in point. of external considerations as well as narrow
In other cases the rationale has been more financial effects (Jansson, 2000). Engineering,
towards state provision being the most effective design, and operating standards are set to meet
means of circumventing market failures that safety requirements.
were feared if the private sector were the suppli- This move away from public ownership and
ers. These failures include a potential lack of the greater participation of private investors
coordination of network based infrastructure, and contractors stems from changes, both in
exploitation of market power where the infra- public attitudes, fostered by intellectual shifts,
structure involves a degree of monopoly, and the regarding the effectiveness of state industries
potential for excessive external costs (embracing in meeting social and economic goals gen-
such things as environmental damage and inad- erally, and in the financial underpinning of
equate safety concerns). Of course, this begs the public investments. The traditional public
question, if indeed market failures are likely to be interest concepts of state ownership, whereby
excessive, whether ownership or regulation is the the controlling body seeks only to fulfill social
best way of combating them. objectives began to be questioned intellectually
In recent years there has been a shift, as in in the 1960s by economists of the Public
many other sectors, away from public owner- Choice persuasion, who argued that state
ship to more private sector involvement. The managers themselves have personal motiva-
first private tolled road of recent times, for tions. The tendency, which is seen as a normal
example, has opened in the UK, but before part of human nature, is for bureaucrats to have
then many countries such as New Zealand had their own agendas that include their career
privatized air traffic control, airports and advancement, job security, and other factors
seaports in many countries have been taken out that can conflict with their seeking to meet
of state control, and outsourcing of many activ- social criteria. The result is inefficiency that
ities such as maintenance had become com- manifested itself most clearly in the state
monplace. In some cases, as with some build owned transportation systems in Central and
and operate schemes, the structure has been one Eastern Europe before 1989.
more akin to leasing than freehold ownership, Even where state owned entities were more
with the infrastructure reverting to the state efficiently managed, and social objectives still
after a predetermined period. In other cases, came near the forefront of activities, develop-
state ownership has remained but the control of ments outside of the transportation have forged
decision-making has been depoliticized. The change. In particular, the traditional view that
Crown Corporation owning air traffic control state ownership is needed to finance large infra-
in Canada (NAV Canada) is an example of this. structure investments, a position even Adam
Peter-3383-Chapter-19.qxd 6/8/2006 7:40 PM Page 335
Smith took, has been dented as private sector Access is not an issue if this were the case
financial markets have become more sophisti- because, by definition, public goods are non-
cated and international. Macroeconomic policy excludable. While things like national defense
has also moved away from the idea that the state may be seen as a public good, essentially it is
can steer the economy through its public sector impossible for the nation’s army not to protect
investments. Added to this has been the shift in me if it is protecting everyone else, this is
thinking towards economic efficiency per se. manifestly not the case with transportation
This has raised questions about access to trans- infrastructure. A simple barrier can thwart
portation infrastructure and making sure that assesses to a road. Equally, there is no problem
priority is with those who actually gain most with efficient use of a public good because, by
from it. User charges have become common, definition it is non-rival; there is no conges-
and technological advances in areas of tion. But this is hardly a characteristic that per-
Intelligent Transport Systems (ITS) have help in tains to transportation infrastructure. To put it
the facilitation of this. simply, generically transportation infrastruc-
But issues of ensuring that the private sector ture has all the features of a private good.
meet larger social goals as well as narrower The general lack of appreciation of the intrin-
financial ones still remains a concern of policy sic economic nature of transportation over the
makers. Essentially, the approach has shifted years has resulted in a continued display of
from always owning the infrastructure to one policy gymnastics to try to make the best use of
where the authorities set objectives or character- it. Some of these have entailed simple physical
istics (that may be in design standards, quality of command-and-control rationing devices; e.g.,
service, capacity, etc.) but then devising mecha- the allocation of airport runway take-off slots.
nisms whereby facilitating provision by the pri- Others have relied, rather perversely, when
vate sector at minimum cost. The result has been infrastructure is taken as public good, on con-
the creation of a variety of bidding and tender- gestion to deter use by deteriorating the quality
ing systems that seek to produce competition for of service offered; generally slower and less reli-
the provision of these predefined packages of able movement. In some cases where privately
infrastructure characteristics. Overlaying this are provided infrastructure is regulated, or pro-
regulatory structures governing things like prices vided publicly without subsidies, and costs
(often involving price-caps) to prevent exploita- require to be recovered, revenues are collected
tion as markets change over time. from users. The normal basis for this, however,
is aimed at cost recovery without distorting
excessively the pattern of use from what it
would be with no charges; technically, Ramsey
PUBLIC POLICY ON TRANSPORTATION
pricing is advocated for this. What this does not
INFRASTRUCTURE ACCESS
do, however, is ration the overall use of the
transportation infrastructure to the economi-
Whilst many of the traditional debates on cally efficient level.
infrastructure policy, or at least the economic What is noticeable is that where transporta-
debates, have centered on ways of assessing of tion suppliers, either of services or infrastruc-
alternatives (e.g., the development of various ture, are left to their own devices is that they
forms of cost-benefit analysis techniques) and seek to levy fares or charges that both give pri-
on means of finance, there has been a recent ority to those most willing to pay and act to
upsurge of interest in access issues; basically make the overall optimal use of the facilities.
making better use of existing transportation Most roads, for example, can either be used
facilities. One contributing factor to the long- without any direct payment6 or by paying a
standing focus on investment and capacity was fixed toll designed to cover the capital and
that much transportation infrastructure was maintenance costs of roads. Excessive traffic
fallaciously viewed as being a public good.5 congestions of roads, in particular in cities but
Peter-3383-Chapter-19.qxd 6/8/2006 7:40 PM Page 336
of pieces of transportation infrastructure, some of dubious European Commission (1989) Council Resolution in
merit, bearing the name of some defunct politician or Trans-European Networks, COM(89)643/FIN,
dignitary, hints at the legacy effects of getting involved in Brussels.
supporting investments. European Commission (2001) European Transport
6. Where there are dedicated ‘road funds’ financed by
Policy for 2010: Time to Decide, White Paper,
fuel or vehicle taxation, these sources of payment are
remote from the actual use of a road for any particular
Brussels.
trip. As such they have very little impact on peoples’ deci- Eberts, R. W. and Fogarty, M. S. (1987) Estimating
sions whether to drive or not. the relationship between local public and private
investment, Federal Reserve Bank of Cleveland
Working Paper 8703.
Frey, B. S. (1985) Democratic Economic Policy,
REFERENCES
Martin Roberson, Oxford.
Gramlich, E. M. (1994) Infrastructure investment:
Andersson, Å. E., Anderstig, C. and Hårsman, B. a review essay, Journal of Economic Literature,
(1990) Knowledge and communications infra- 32: 1176–96.
structure and regional economic change, Regional Holtz-Eakin, D. (1992) Public sector capital and
Science and Urban Economics, 20: 359–376. productivity puzzle, National Bureau of Economic
Aschauer, D. A. (1989) Is public expenditure produc- Research Working Paper 4122.
tive?, Journal of Monetary Economics, 23: 177–200. Hirschman, O. A. (1958) The Strategy of Economic
Biehl, D. (1991) The role of infrastructure in Development, New Haven: Yale University Press.
regional development, in R.W. Vickerman (ed), Jansson, J. O. (2000) Transport infrastructure:
Infrastructure and Regional Development, London, the investment problem, in J. B. Polak and
Pion. A. Heertje (eds), Analytical Transport Economics:
Button, K. J. (1998) Infrastructure investment, An International Perspective, Edward Elgar,
endogenous growth and economic convergence, Cheltenham.
Annals of Regional Science, 32: 145–162. Jorgensen, D. (1991) Fragile statistical foundations:
Button, K. J. (2002) High-technology employment the macroeconomics of public infrastructure
and hub airports: infrastructure’s contribution to investment, comment on Hulten and Schwab,
regional development, in Y. Higano, P. Nijkamp, presented at the American Enterprise Institute
J. Poot and J. J. van Wijk (eds), The Region in the New Conference on Infrastructure Needs and Public
Economy: An International Perspective on Spatial policy Options for the 1990s, Washington.
Dynamics in the 21st Century, Ashgate, Aldershot. Lakshmanan, T. R. (1989) Infrastructure and eco-
Button, K. J. (2004) Wings Across Europe: Towards an nomic transformation, in A. E. Andersson,
Efficient European Air Transport System, Ashgate, D. F. Batten, B. Johansson and P. Nijkamp (eds),
Aldershot. Advances in Spatial Theory and Dynamics,
Button, K. J. and Rietveld, P. (2000) A meta-analysis Amsterdam: North-Holland.
of the impact of infrastructure policy on regional Leitham, S. (1993) Predicting the economic devel-
development, in H. Kohno, P. Nijkamp and opment effects of transportation projects,
J. Poot (eds), Regional Cohesion and Competition Proceedings of the 25th Annual Conference of
in the Process of Globalisation, Edward Elgar, the Universities Transport Research Group,
Cheltenham. Southampton.
Costa, J., da Silva Ellson, R. W. and Martin, R. C. Mera, K. (1973) Regional production functions
(1987) Public capital, regional output and devel- and social overhead capital: an analysis of the
opment: some empirical evidence, Journal of Japanese case, Regional and Urban Economics,
Regional Science, 27: 419–37. 3: 157–86.
Dupuit, J. A. (1844) De la mesure de l’ utilité des Morrison, C. J. (1993) Macroeconomic relationships
travaux publics, Annuales des Ponts et Chaussés, 8. between public spending on infrastructure and
Duffy-Deno, K. and Eberts, R. W. (1991) Public private sector productivity in the United States, in
infrastructure and regional economic develop- J. M. Mintz and R. S. Preston (eds), Infrastructure
ment: a simultaneous approach, Journal of Urban and Competitiveness, John Deutsch Institute for
Economics, 30: 329–43. the Study of Economic Policy, Ottawa.
Eisner, R. W. (1991) Infrastructure and regional eco- Munnell, A. H. (1990) Is there a shortfall in public
nomic performance, New England Economic capital investment? New England Economic
Review, September/October: 47–58. Review, September/October, 11–32.
Peter-3383-Chapter-19.qxd 6/8/2006 7:40 PM Page 338
Nadiri, I. and Manuenias, T. (1996) Contribution of Sturm, J. E. and de Haan, J. (1995) Is public
Highway Capital to Industry and National expenditure really productive? New evidence
Productivity Growth, Federal Highway Adminis- from the USA and the Netherlands, Economic
tration Office of Policy Development, Washington. Modelling, 12: 60–72.
Nurske, R. (1953) Problems of Capital Formation in Winston, C. (1991) Efficient transportation infra-
Developing Countries, Oxford: Basil Blackwell. structure policy, Journal of Economic Literature,
Smith, A. (1776) An Enquiry into the Nature and 5: 113–27.
Causes of the Wealth of Nations. 2000 edition pub- Youngson, A. J. (1967) Overhead Capital. A Study in
lished by The Modern Library, New Work. Development Economics, Edinburgh: Edinburgh
Stough, R., Vickerman, R., Button, K. J. and University Press.
Nijkamp, P. (eds) (2002) Transport Infrastructure,
Edward Elgar, Cheltenham.
Peter-3383-Chapter-20.qxd 6/8/2006 7:41 PM Page 339
20
Foreign Policy
W A LT E R C A R L S N A E S
Taking the broad historical perspective is policy analysts today can and do agree on
often a suggestive strategy for gauging the cur- (and there is not much else), it is that they
rent state of a field of study. The chapter on belong squarely to the scholarly domain of
‘Foreign Policy’ in the magisterial eight-volume International Relations (IR) rather than to any
Handbook of Political Science, published in of the policy sciences.
1975, is in this regard insightful for at least two This is not to say, however, that the study of
reasons. The first is its tone, which is guardedly foreign policy currently enjoys an undisputed
optimistic about the future of foreign policy professional domicile within IR. This uneasy
analysis, despite deep-rooted disagreements state of affairs is due, at least in part, to the fail-
within the field regarding both its conceptual ure of foreign policy researchers, during the
boundaries and the most appropriate manner past twenty-five years, to consolidate their field
to analyze its substance. There is, the two in the manner once envisioned. Instead, their
authors write, a ‘sense of movement at last, practice has to a considerable degree become
akin to one’s first responses as a traffic jam one of eclecticism and defensiveness within a
unlocks and cars begin, hesitantly and tenta- larger scholarly milieu which, on the whole, is
tively, to pick up forward speed’ (Cohen and not especially engaged with the issues at the
Harris, 1975: 381). The second reason is the head of the agenda of foreign policy analysis.
unquestioned assumption that the subject A quick perusal of the table of contents of the
matter of foreign policy belongs naturally to major IR journals published during the past
the empirical domain of public policy rather decade or so is quite clear on this score: very
than of international relations – so much so few contain titles in which the concept of
that Cohen and Harris’ chapter was published ‘foreign policy analysis’ plays a prominent role.
in the volume on ‘Policies and Policymaking’ At the same time interest in the development
rather than that on ‘International Politics’. To of IR theory itself has grown exponentially, but
most readers today, a quarter of a century later, for the most part with little or no reference to
both of these characterizations will undoubt- ‘foreign policy’, either as an integral part of
edly raise more than a few puzzled eyebrows. such theory or as a separate but important
The first due to its misplaced (if admittedly approach in its own right. On the contrary,
guarded) optimism about the future discipli- most of the time it is simply ignored in these
nary development of the field; and the second debates and discussions, or politely dismissed,
because, if there is anything which all foreign with reference to the distinction between
Peter-3383-Chapter-20.qxd 6/8/2006 7:41 PM Page 340
system level and unit level theories, the former perhaps to a considerable extent as a result of
pertaining to international politics proper, the it – this dominant perspective has increasingly
latter ‘merely’ to the behaviour of individual had to provide space for a view of the sub-
states. ‘Theory development at this level’, a stance of interstate interactions which is more
recent review of theories of foreign policy thus in tune with some of the basic premises of
states laconically, ‘has received comparatively foreign policy approaches. In other words, a
little attention’ (Rose, 1998: 145). Alexander case can be made for why a focus on foreign
Wendt’s declaration of (a lack of) interest is policy is once again regaining ground within
equally symptomatic: ‘Theories of interna- IR, and why it should indeed do so.
tional politics are distinguished from those The way I intend to proceed is as follows. In
that have as their object explaining the behav- the next section an intellectual history of
iour of individual states, or “theories of foreign foreign policy analysis will be presented, pri-
policy” … Like Waltz, I am interested in inter- marily covering developments during the past
national politics, not foreign policy’ (Wendt, half-century. After that a conceptual and
1999: 11). Perhaps of equal significance, foreign analytical overview of the field itself will be pro-
policy analysts themselves seem to have lost vided, in which I will first very briefly discuss
heart. Hence, as a British scholar noted in 1999, fundamental definitional issues and present
‘These are testing times for foreign policy ana- four rock-bottom types of explanatory frame-
lysts. At issue is whether their area of study works defined, not in terms of ‘schools’, ‘grand
remains a major sub-field of International debates’ or ‘contending approaches’, but with
Relations or whether it has become anachro- reference to two fundamental meta-theoretical
nistic, either subsumed or replaced by other dimensions within the philosophy of social
approaches to understanding and explaining science. On the basis of these four generic per-
state behaviour’ (White, 1999: 37). Similarly, a spectives, my intention in the subsequent and
German colleague has noted that, despite a core part of the chapter is to highlight and
plethora of publications on the topic in his briefly discuss some of the more prominent
home country, the study of foreign policy itself contemporary attempts to structure and to
is currently in the throes of a conceptual crisis pursue analysis within the field. After this the
and theoretically at a standstill (Schneider, question will be raised – and a brief answer
1997). However, let me already at this point suggested – whether a synthetic or integrated
signal that, although there is some justification approach to foreign policy analysis is at all fea-
for the bleak picture of the sub-field of foreign sible. The concluding section will pinpoint a
policy analysis adumbrated above, it by no few current and contentious issues straddling
means represents the whole picture. It reflects the various approaches discussed, indicating
a disciplinary development during the past two some areas of potential development within
to three decades which has put a strong struc- the field.
turalist-systemic stamp on IR, and hence also
an effective damper on approaches – such as
foreign policy analysis – premised not primar-
INTERNATIONAL RELATIONS AND
ily on the international system as the generator
FOREIGN POLICY ANALYSIS:
of behaviour but on the importance of unit-
A SHORT INTELLECTUAL HISTORY
level factors and actors for understanding and
explaining state behaviour. But this structuralist-
systemic perspective has never been totally As is the case with IR itself, most historical
hegemonic, even in North America, and in accounts of foreign policy analysis – and there
Europe it has failed to achieve the same grip on are not many available – tend to suffer from a
the scholarly imaginations of its mostly small, Whig interpretation of this history, or from
eclectic and not equally ‘scientistic’ or ‘ratio- what Brian C. Schmidt has called the problem
nalistic’ IR communities. More importantly, of ‘presentism’: ‘the practice of writing a
since at least the end of the Cold War – and history of the field for the purpose of making a
Peter-3383-Chapter-20.qxd 6/8/2006 7:41 PM Page 341
point about its present character’ (Schmidt, beliefs (Cohen, 1968). The first was to focus on
2002: 5). These accounts are also to a consider- how the governmental institutions responsible
able degree infused with parochialisms of for the formulation and implementation of
various shades, both of a geographic, scholarly foreign policy could be made more efficient in
and subdisciplinary nature. The combination the pursuit of their tasks. The second had a
of these two characteristics makes for interest- more ideological thrust, essentially involving a
ing reading but hardly for a fully illuminating plea for the democratization of foreign policy –
overview of this historical development. In of why and how public values and interests
other words, they have on the whole con- should be introduced to every stage in the for-
tributed to conventional images of the pro- mulation and execution of such policy.
gression, and hence identity of the field, that However, concomitant with this institution-
need to be challenged and corrected. ally focused and policy-oriented tradition in
As suggested above, the conception of the academic study of foreign policy, which
foreign policy as an academic subject matter enjoyed its American heyday during the two
has had strong roots in the broader domain of decades following the Second World War, we
public policy, especially in the United States. also find a second major tradition, and one
However, this is not where the field originated which has left a much stronger and seemingly
but is, rather, a reaction to the earlier tradition – indelible imprint on the subsequent develop-
primarily of a European provenance, with ment of the field. I here have in mind the
origins in the seventeenth century and the rise induction into American thinking of a power-
of the modern state thereafter – of viewing ful European influence, and one that stands in
foreign policy as a distinct domain differing in marked contrast to the indigenous strands of
fundamental respects from all other spheres of the liberal Wilsonian project. Realism is its
public policy. ‘The leading assumption’, name, and Hans Morgenthau was for decades
Bernard C. Cohen thus noted some years ago, its undisputed high priest (Morgenthau, 1948).
‘is that foreign policy is “more important” than As argued by Stefano Guzzini in his compre-
other policy areas because it concerns national hensive sociological analysis of the history of
interests, rather than special interests, and realism, Morgenthau’s main concern, as that of
more fundamental values’ (Cohen, 1968: 530). realists more generally, was to resuscitate an
A further consequence of this doctrine of the older tradition by translating ‘the maxims of
‘primacy of foreign policy’ was, of course that, [the] nineteenth century’s European diplo-
being distinct in this manner, political elites matic practice into more general laws of an
demanded that it be treated differently from all American social science’ (Guzzini, 1998: 1; see
other areas of public policy, that is, beyond also Dunne, 1998; Kahler, 1997). To summarize
democratic control and public scrutiny. a complex argument, he did this by claiming
However, the experiences leading up to, and ‘that the inherent and immutable self-interested
the consequences of, the First World War con- nature of human beings, when faced with a
vinced some influential statesmen – in partic- structure of international anarchy, results in
ular Woodrow Wilson – that an end should be states maximizing one thing, power’ (Smith,
put to the traditional secretive practices of 1986: 15). By linking this view of power to the
statecraft and diplomacy. concept of the national interest, he believed
Despite the subsequent failure of the that he could provide a universal explanation
Wilsonian project, the study of foreign policy for the behaviour of particular states.
was deeply affected – especially in the United The behaviouralist turn in American social
States – by this liberal and democratic ideol- science in the 1950s and 1960s had a decisive
ogy, with the result that much of its activities effect on both of these approaches to the study
subsequent to the Second World War, when of foreign policy. Its impact on the institution-
foreign policy analysis first came to be firmly ally oriented research tradition was perhaps
established academically, was concerned with the more deep-going, in the sense that it
the study of two major implications of these changed its character altogether from being
Peter-3383-Chapter-20.qxd 6/8/2006 7:41 PM Page 342
an essentially idiographic and normative which all along had focused mainly on the
enterprise – analyzing particular forms of processes involved in foreign policy decision-
policy or prescribing better means for its for- making, or on contextual or sociopsychologi-
mulation and implementation – to one which cal factors influencing such behaviour
now aspired to generate and to test hypotheses (Hudson and Vore, 1995: 216–19). The former,
in order to develop a cumulative body of with roots going back the pioneering work on
empirical generalizations. The main out- decision-making by Snyder, Bruck and Sapin
growth of this fundamental theoretical and (1954), developed into extensive research
methodological reorientation was a move- exemplified by, for example, studies focusing
ment, starting in the late 1960s, which became on small group dynamics (C. Hermann, 1978;
known as the comparative study of foreign Janis, 1982; Tetlock, 1979), the ‘bureaucratic
policy, or CFP for short. Its strong behav- politics’ approach made famous by the publi-
iouralist character is manifested in its focus on cation in 1971 of Graham Allison’s study of the
explaining foreign policy in terms of discrete Cuban crisis, as well as Steinbruner’s attempt
acts of ‘behaviour’ rather than in the form of to present foreign policy-making as analogous
‘purposive’ state actions in the realist mode; to cybernetic processes (Steinbruner, 1974).
and taking its cue from how American behav- The latter type of research focus, concentrating
ioural political science focused on the ‘vote’ as on more particular aspects of the decision-
its fundamental unit of analysis, it posited the making process, produced a number of dis-
‘event’ as its dependent variable. In this view tinguished studies, ranging from Michael
foreign policy is seen as the exercise of influ- Brecher’s (1972) work on Israel, Robert Jervis’s
ence in international relations, with ‘events’ (1976) book on perceptions and mispercep-
specifying ‘who does what to whom, and how’ tions, and a long series of studies – continuing
(Hudson and Vore, 1995: 215). As a conse- to the present time, as we shall see below – on
quence, the task of collecting data on and ana- the role of cognitive and psychological factors
lyzing such events, with the aim of generating in the explanation of foreign policy actions
and accumulating empirical generalizations (Axelrod, 1976; Cottam, 1977; M. Hermann,
about foreign policy behaviour, became a 1974, 1980a, 1980b; Holsti et al., 1968).
major industry within CFP (Brecher, 1972; What can be said generally about this broad
East, 1978; McGowan and Shapiro, 1973; tradition is that, whereas there was perhaps a
Rummel, 1972; Wilkenfeld et al., 1980). It was brief moment in time when it could be
also an activity generously funded by a federal asserted that foreign policy analysis was self-
government fully in tune with these ambitions consciously in the process of achieving an
(Andriole and Hopple, 1984). identity of its own (‘all the evidence’, James N.
However, it is generally acknowledged by Rosenau thus proclaimed in 1974, in a state-
friend and foe alike that this programme of ment that was soon and forever after to cause
establishing a truly ‘scientific’ approach to the him considerable chagrin, ‘points to the con-
analysis of foreign policy was, on the whole, a clusion that the comparative study of foreign
significant if commendable failure. The empiri- policy has now emerged as a normal science’),
cal results of the major research programmes this is certainly not the case at the beginning of
which had been launched during these years the new millennium (quoted in Smith, 1986:
turned out to be disappointing (Hudson and 20). Instead, if anything is typical of its practi-
Vore, 1995: 215–16), and it became increasingly tioners at present, it is the almost total lack of
evident that the aim of a unified theory and a such a sub-disciplinary identity. In the words
methodology based on aggregate analysis had to of one of its contemporary chroniclers, the
be rejected as empirically impracticable and ana- attitude today is instead one of allowing ‘a
lytically unfruitful (Caporaso et al., 1987; East, hundred flowers to bloom’ (Hudson and Vore,
1978; Kegley, 1980; Munton, 1976; Smith, 1987). 1995: 22); or, as another reviewer has put it (in
The CFP programme did not, however, a slightly more upbeat locution), of opening ‘con-
eclipse the type of foreign policy analysis versational space’ to the multiple perspectives
Peter-3383-Chapter-20.qxd 6/8/2006 7:41 PM Page 343
and ‘new vistas’ of foreign policy analysis domestic independent variable … they all share
(Neack et al., 1995: 12). a common assumption – that foreign policy is
Turning to the development of realism in the best understood as the product of a country’s
face of the behaviouralist challenge, we are pre- internal dynamics’ (Rose, 1998: 148). Juxtaposed
sented with an intriguing paradox in the history against its explanatory logic we find realism
of foreign policy analysis. On the one hand, it broadly conceived, and for the sake of simplicity
was believed by many that, given the centrality (and linguistic consistency) we can refer to this
in Morgenthau’s approach of power defined in tradition as that of Realpolitik. Although not
terms of the innate, unobservable but crucial averse to allowing for the play of domestic
notion of a fixed human nature, it would not be factors in the pursuit of foreign policy, the major
able to withstand this confrontation. Yet, this is explanatory weight is here given to material sys-
precisely what it did, insofar as the behavioural- temic-level factors in one form or another.
ists never really challenged the underlying However, although this characterization in
assumptions of realism, only its methodology terms of the classical divide between domestic
(Vasquez, 1983). Nevertheless, while continuing and international politics has a long historical
to be the major intellectual force defining IR pedigree, it does have at least one major draw-
itself (Guzzini, 1998; Hollis and Smith, 1990), back as a criterion for classifying contemporary
realism became methodologically divided as a foreign policy analysis. For, while many scholars
consequence of the debate on its scientific continue to think of this analytical boundary as
status, and suffered a setback – by no means the major line of division within the field, and
permanent – with the publication of Allison’s one that continues to be conceptually fruitful in
in-depth penetration of the Cuba crisis in terms analysis, it is nevertheless based on an assump-
primarily of an analysis of unit-level rather than tion which is highly questionable as both an
systemic factors (Allison, 1971). Since the cele- empirical and a theoretical proposition: that it is
brated appearance of Kenneth Waltz’s Theory of indeed feasible to determine the nature and
International Politics (1979), an even clearer function of such a boundary, and to do so with-
bifurcation within realism has occurred, partic- out begging a fundamental question in the
ularly in response to the strong stand against all study of international relations. Thus, while it
forms of reductionist approaches – typified by can be argued that this characterization of the
most theories of foreign policy – which lies at field in terms of these two broad traditions con-
the core of his structuralist reformulation of tinues to reflect a sub-disciplinary selfunder-
realism. standing of its development, it will not be used
In summation of half a century of foreign below when discussing the current state of
policy analysis one can thus say that two broad affairs in foreign policy analysis. Instead of a
traditions have played a major role in it, and that criterion based specifically on the substantive
they continue to do so. The first is the more nature of foreign policy (and one of dubious
difficult to label, insofar as it contains a host of value), the discussion will proceed from two
different and disparate approaches, including meta-theoretical dimensions – one ontological,
work on cognitive and psychological factors, the other epistemological – which are entirely
bureaucratic and neoinstitutional politics, neutral with regard to the substance of foreign
crisis behaviour, policy implementation, group policy itself.
decision-making processes, and transnational
relations, to name some of the most important
(see Hudson and Vore, 1995: 222–8). If only for
CONCEPTUALIZING THE DOMAIN
lack of a better term, we can refer to this tradi-
tion in terms of the primacy allocated within it
to the role of Innenpolitik – of domestic factors – ‘There is a certain discomfort in writing about
in the explanation of foreign policy. As recently foreign policy,’ we are forewarned in the first
noted, although there ‘are many variants of this lines of the Handbook of Political Science
approach, each favouring a different specific chapter on foreign policy, ‘for no two people
Peter-3383-Chapter-20.qxd 6/8/2006 7:41 PM Page 344
seem to define it in the same way, disagreements As a starting point for discussing the types
in approach often seem to be deep-seated, and of explanatory factors that have characterized
we do not yet know enough about it to be able foreign policy analysis, it is necessary to con-
to say with confidence whether it may be sider two fundamental issues that have domi-
differentiated from all other areas of public nated current meta-theoretical debate within
policy’ (Cohen and Harris, 1975: 318). What its social theory (and IR). The first concerns the
two authors point to here is a twin problema- ontological foundation of social systems: the
tique which has occupied a central place in the type of issue exemplified by the claim, reput-
history of foreign policy analysis, and which edly made by Margaret Thatcher, that there is
needs to be addressed as much today as in ‘no such thing as a society’, but ‘only individu-
the past. The first of these concerns the cru- als’. Essentially, it revolves around the question
cial issue of what constitutes the particular of where the dynamic foundations of social
explanandum of the study of foreign policy: systems are located. This dynamism either has
what it is that is to be explained. For while this its origin in ‘the effects, intended or not, of
definitional issue may on first sight seem triv- individual action; or from the slowly evolving
ial, it in fact goes to the core of what distin- rules of the self-reproducing structure’
guishes this field of study from that of both (Guzzini, 1998: 197). This classic distinction in
domestic and international politics, and hence social theory is usually expressed in terms
lies at the heart of the long-standing issue of of the dichotomy between ‘individualism’
where and how to draw the analytical bound- and ‘holism’, the former holding ‘that social
ary between a sub-field that willy-nilly strad- scientific explanations should be reducible
dles these two major disciplinary foci of to the properties or interactions of indepen-
political science. Secondly, this issue is also dently existing individuals’, while holism
crucial to the choice of theoretically feasible stands for the view ‘that the effects of social
instruments of analysis, since the nature of a structures cannot be reduced to independently
given explanandum has obvious and funda- existing agents and their interactions’ (Wendt,
mental implications for the types of explanans, 1999: 26).
that is, explanatory factors, which in principle This ontological polarity between individu-
are appropriate and in practice fruitful. alism and holism should be clearly distin-
Although there is today (in contrast to a gener- guished from the epistemological issue of
ation ago) a relatively stable consensus with whether social agency is to be viewed through
regard to the explanandum, which therefore an ‘objectivistic’ or an ‘interpretative’ lens.
need not detain us for long, this is not the case Using a different metaphor, two choices are
with respect to the considerably more con- available here: to focus on human agents
tentious meta-theoretical issue. and their actions, either from the ‘outside’ or
This consensus boils down to a specification from the ‘inside’, corresponding to the classi-
of the unit of analysis that emphasizes the pur- cal Weberian distinction between Erklären
posive nature of foreign policy actions, a focus (explaining) and Verstehen (understanding).
on policy and the crucial role of state bound- As argued by Martin Hollis and Steve Smith,
aries. The following stipulation is intended to these two approaches tell two different types of
capture these definitional aspects: foreign poli- ‘stories’ about international relations, each
cies consist of those actions which, expressed with its own view of human nature and a con-
in the form of explicitly stated goals, commit- comitant range of ‘appropriate’ theories
ments and/or directives, and pursued by gov- (Hollis and Smith, 1990). The choice is thus
ernmental representatives acting on behalf of between an approach that models itself on the
their sovereign communities, are directed natural sciences, and one premised on the
toward objectives, conditions and actors – independent existence of a social realm con-
both governmental and non-governmental – stituted by social rules and intersubjective
which they want to affect and which lie beyond meanings. Whereas the former is based on a
their territorial legitimacy. ‘naturalistic’ epistemology self-consciously
Peter-3383-Chapter-20.qxd 6/8/2006 7:41 PM Page 345
between ‘aggressive’ and ‘defensive’ types Lynn-Jones and Miller, 1995: xi; Snyder, 1991;
(Snyder, 1991: 11–12; see also Lynn-Jones and Van Evera, 1990/91: 11–17; Walt, 1995; Zakaria,
Miller, 1995: xi–xii; Rose, 1998). During the 1995: 475–81).
past decade aggressive neorealism has been pre- Neoclassical realists should be distinguished
eminently represented by John Mearsheimer, from both offensive and defensive neorealists.
who has argued that, whereas the Cold War, They share with neorealists the view that a
based on bipolarity, military balance and country’s foreign policy is primarily formed by
nuclear weapons, produced peace in Europe for its place in the international system and in par-
45 years, its demise will – contrary to the con- ticular by its relative material power capabilities.
ventional wisdom – perforce have deleterious However, and here the classical roots of this
effects in the long run. This pessimistic scenario approach come to the fore, they also argue that
follows from a strict application of neorealist the impact of systemic factors on a given
tenets, especially of the view that, insofar as the country’s foreign policy will be indirect and
international system invariably fosters conflict more complex than neorealists have assumed,
and aggression, rational states are compelled to since such factors can effect policy only through
pursue offensive strategies in their search for intervening variables at the unit level (Rose,
security (Mearsheimer, 1995: 79–129; see also 1998: 146). This view is clearly contrary to the
Layne, 1995: 130–76). It also emphazises the whole tenor of offensive neorealism, but neo-
role of the polarity of the international system – classical realists also fault defensive neorealists,
bipolarity being more conducive to peace than mainly because it is claimed that their systemic
multipolarity – as well as the effects of changes argument fails to explain much of actual foreign
in the relative power of states. policy behaviour and hence needs to be aug-
Defensive neorealists, on the other hand, do mented by the ad hoc introduction of unit-level
not share this pessimistic and essentially variables (see, e.g., Schweller, 1996: 114–15;
Hobbesian view of the international system, Zakaria, 1995). As a consequence of the stress
instead arguing that, although systemic factors on the role of both independent (systemic) and
do have causal effects on state behaviour, they intervening (domestic) variables, research
cannot account for all state actions. Instead of within neoclassical realism is generally con-
emphasizing the role played by the distribution ducted in the form of theoretically informed
of power in the international system, scholars narratives – ideally supplemented by counter-
such as Stephen Walt and Charles L. Glaser thus factual analysis – that trace how different factors
instead pointed to the importance of the source, combine to forge the particular foreign policies
level and direction of threats, defined primarily of states (Rose, 1998: 153). More specifically,
in terms of technological factors, geographic this has yielded extensive narrative case studies
proximity, offensive capabilities and perceived of how twentieth-century great powers –
intentions (Glaser, 1995; Walt, 1995; see also the especially the United States, the Soviet Union
references in Rose, 1998: 146, fn. 4). The picture and China – have reacted to the material rise or
presented here is that states pursuing security in decline of their relative power in the interna-
a rational manner can on the whole afford to be tional system (Christensen, 1996; Schweller,
relatively relaxed, except in rare instances; and 1998; Wohlforth, 1993; Zakaria, 1998).
that security can generally be achieved by bal-
ancing against threats in a timely way, a policy Neoliberal institutionialism Although not
that will effectively hinder most forms of actual generally touted as an approach to the analysis
conflict. ‘Foreign policy activity’, Rose thus of foreign policy, it is obvious that the type of
explains, ‘is the record of rational states reacting focus that usually goes under the name of
properly to clear systemic incentives, coming neoliberal institutionalism is as relevant to the
into conflict only in those circumstances when study of foreign policy as are realism and neo-
the security dilemma is heightened to fever realism in their various configurations. Indeed,
pitch’ (Rose, 1998: 150; see also Glaser, 1995; insofar as this school of thought is posited
Peter-3383-Chapter-20.qxd 6/8/2006 7:41 PM Page 347
as an alternative to realism (and, the view of latter conform to the demands of the
some, as the only one), it also pari passu entails former – or between agencies and individuals;
an alternative approach to foreign policy on this level a structural view would imply that
analysis (see Baldwin, 1993). individual decision-makers do not act inde-
Neoliberal institutionalism is a structural, pendently but generally in conformity with the
systemic and ‘top-down’ view for some of the dictates of the agencies employing them
same reasons that realism constitutes such an (Hollis and Smith, 1990: 8–9, 196–202).
approach. It assumes that states are the pri- The latter type of claim has become known
mary actors in the international system; that as the organizational process approach ever
they behave like egoistic value maximizers; and since the celebrated publication of Allison’s
that the international system is essentially Essence of Decision in 1971. With roots in orga-
anarchic (Baldwin, 1993: 8–14; Grieco, 1993). nizational theory, it focuses on decisions not in
It is also for this reason that Andrew Moravcsik terms of instrumental rationality but as out-
has claimed that ‘neoliberal institutionalism’ is puts of large organizations functioning accord-
a misnomer insofar as it essentially constitutes ing to regular patterns of behaviour, usually
a variant of realism (Moravcsik, 1997: 537). referred to as standard operating procedures.
What then is distinctive about the neoliberal The most prominent recent research in which
institutionalist approach to foreign policy organizational theory has been used in foreign
analysis? Very briefly, the following: whereas policy analysis has focused on decision-
both realists and neoliberals view foreign making in general, and on the role of decision-
policy-making as a process of constrained making units – particularly small groups – in
choice by purposive states, the latter under- this process. This has been the case, for exam-
stand this constraint not primarily in terms of ple, in recent work reconsidering and going
the configurations of power capabilities facing beyond Irving Janis’s notion of ‘groupthink’,
policy-makers, but in terms of an anarchic focusing on the interplay between group
system which, while it fosters uncertainty and dynamics and the role of broader organiza-
hence security concerns, can nevertheless be tional cultures and socialization in foreign
positively affected by the institutional provi- policy decision-making (Beasley, 1998; ‘t Hart
sion of information and common rules in the et al., 1997; Ripley, 1995). This type of research
form of functional regimes. The result is that points to the applicability of recent organiza-
international cooperation under anarchy is tional theory (see, e.g., March and Olsen,
possible in the pursuit of given state prefer- 1998), in particular the celebrated (if not
ences (Oye, 1985); and hence certain specific entirely transparent) distinction between the
features of an international setting can explain logic of ‘consequences’, defining the type of
state outcomes in the form of cooperative action appropriate within both realist and
foreign policies (Axelrod and Keohane, 1993; neoliberal thinking, and the logic of ‘appropri-
Keohane, 1993). ateness’, which – as Allison and Zelikow have
claimed in their recent and substantial updat-
Organizational process approaches While ing of the organizational model – is very much
both realism and neoliberal institutionalism at the heart of the organizational process
are structural approaches of a systemic kind, approach to decision-making (Allison and
foreign policy analysis can be pursued ‘struc- Zelikow, 1999: 146).
turally’ on a lower level of analysis as well, in
which the structural factor driving foreign
policy behaviour is not external but internal to Approaches from an Agency-Based
the state. As argued by Hollis and Smith, a ‘top- Perspective
down’ approach on the sub-systemic level
either focuses on the causal relationship Cognitive and psychological approaches
between the state and its agencies – how the Although research on the cognitive and
Peter-3383-Chapter-20.qxd 6/8/2006 7:41 PM Page 348
here; he has already swept much of it for us policy, and, although earlier claimed to be
(Levy, 1994). excessively US-centred in its empirical appli-
cability, it is slowly finding its way to Europe
Bureaucratic politics approach Although as well (see, e.g., the contributions in Stern
the so-called bureaucratic politics – or govern- and Verbeek, 1998). Allison (with his coau-
mental – approach to the analysis of foreign thor) has also upgraded the chapter on
policy, first popularized by Allison in his study governmental politics in the second edition of
of the Cuban crisis, is often assumed to be his study, including in it a host of empirical
closely similar to the organizational process examples postdating the Cuban crisis (Allison
model discussed above (and sometimes con- and Zelikow, 1999: 255–378; see also Karbo,
flated with it), it is premised on an agency- 1998).
oriented rather than a structural view of the
field (Allison, 1971). Insofar as it focuses on Liberal approach Although it has roots
interaction among organizational players going back to the early Rosenau (Rosenau,
involved in bargaining games and competing 1969) and prominent European scholars of
preferences, it does not aim to explain in terms foreign policy (Czempiel, 1981; Hanrieder,
of organizational outputs but on the basis of 1967), as well as to research on the role of
the actual ‘pulling and hauling that is politics’ domestic structures in foreign policy analysis
(Allison and Zelikow, 1999: 255). At the same pioneered by Peter Katzenstein (Katzenstein,
time, although in a certain sense akin to ratio- 1976, 1978) and subsequently developed by
nal choice thinking insofar as its main ratio- Matthew Evangelista, Thomas Risse-Kappen
nale is to explain why decisions often take the and others (Evangelista, 1988, 1995; Risse-
form of ‘resultants’, as distinct from what any Kappen, 1991; Snyder, 1991), Andrew
person or group intended, it does this, not in Moravcsik must nevertheless be given primary
terms of given preferences and strategic moves, credit for having put the liberal approach
but ‘according to the power and performance squarely on the contemporary IR agenda
of proponents and opponents of the action in (Moravcsik, 1997; but see also Doyle, 1997). In
question’ (Allison and Zelikow, 1999: 256). The his view, three core assumptions underlie this
power in question is not in the first hand per- challenge to neorealism and neoliberalism: the
sonal but bureaucratic, insofar as the actors primacy of societal actors over political insti-
involved in these bargaining games represent tutions, the implication of which is that being
sectional or factional rather than individual based on a ‘bottom-up’ view of the political
interests. Hence the famous apothegm (reput- system, individual and social groups are
edly minted by Don Price, but also known as treated as prior to politics, insofar as they
Miles’s law) which encapsulates this bureau- define their interests independently of politics
cratic link between individual actors and their and then pursue these through political
organizational anchorage: where you stand exchange and collective action; state prefer-
depends on where you sit (Hollis, 1994; Stern ences represent the interests of a subset of
and Verbeek, 1998: 206). society, in the sense that state officials define
Although explicitly theorized on the basis of state preferences and act purposively in world
the empirical realities of how governments politics in terms of these interests; and state
actually work (at least in the United States), behaviour in the international system is deter-
this view of foreign policy decision-making mined by the configuration within it of inter-
has over the years received considerable criti- dependent state preferences, that is, by the
cism, both with reference to conceptual confu- constraints imposed on a given state by the
sion and poor empirical performance (see, e.g. preferences of other states (Moravcsik, 1997:
Bendor and Hammond, 1992; Bernstein, 2000; 520). Each of these core assumptions, Moravcsik
Rhodes, 1994; Welch, 1998). Nevertheless, it argues, supports a specific variant of liberal
continues to stimulate research on foreign theory, that is, ideational, commercial and
Peter-3383-Chapter-20.qxd 6/8/2006 7:41 PM Page 350
republican liberalism, respectively. The first Michael Barnett (1998), Jeff Checkel (1999),
pertains to the generation of domestic social John Ruggie (1998), Peter Katzenstein (1996),
demands, the second to the causal mechanisms Thomas Risse-Kappen (1995b) and Martha
by means of which these are transformed into Finnemore (1996b). Followers of ‘thicker’ ver-
state preferences, and the third to the resulting sions range from what Adler terms ‘modernist
patterns of national preferences in interna- linguistic’ (or ‘rule-oriented’) constructivists
tional settings (Moravcsik, 1997: 524–33). such as Friedrich Kratochwil (1989) and
Nicholas Onuf (1989), the ‘discursive’ group to
be discussed below, to the ‘postmodernists’
Approaches Based on a such as Richard Ashley (1984) and Rob Walker
Social-Institutional Perspective (1993), in addition to a number of feminist
scholars, particularly Spike Peterson (1992),
Social constructivism Although ‘social con- J. Ann Tickner (1993; 2002) and Christine
structivism’ (or simply ‘constructivism’), like Sylvester (1994). Since Wendt’s type of con-
‘rational choice’, is essentially a meta-theoretical structivism is explicitly not designed for the
standpoint in the study of social phenomena, analysis of foreign policy (Wendt, 1999: 11),
and hence is foundational to political analysis I will not discuss it further here. Similarly,
rather than being a specific analytical or insofar as postmodernist versions are difficult
‘theoretical’ approach within IR, it will here – to incorporate within a foreign policy frame-
following most constructivist scholars (Adler, work as defined here, these too will be left
1997; Dunne, 1995; Guzzini, 2000; Hopf, 1998; aside. The specifically discursive approach will,
Ruggie, 1998; Wendt, 1999: 31) – be used to des- however, be discussed below.
ignate a more or less coherent and emerging This leaves us here with contributions to the
body of thought in IR, including foreign policy study of foreign policy from within the ‘mod-
analysis. Although it has roots going back to ernist’ type of constructivism. This stream can
Grotius, Kant and Hegel, and was embedded be said to consist, first of all, of a normative
already in some of the classic contributions by and ideational strand, which emphasizes that
Karl Deutsch, Ernst Haas and in particular the the world of international relations does not
English School (Bull, 1977; Deutsch, 1954; exist independently of human action and cog-
Haas, 1964, 1990; see also Dunne, 1995), it is nition but, rather, that it is an intersubjective
nevertheless regarded by most IR scholars today and meaningful world whose rules and prac-
as a relative newcomer to the sub-discipline; the tices are made and reproduced by human
term itself was first introduced to IR by interactions. A second strand, often inter-
Nicholas Onuf as recently as 1989 (Onuf, 1989). twined with the first, emphasizes the role of
At the same time, however, it has quickly estab- identities in international relations, and does
lished itself as perhaps the main contender to a this by pointing to the ‘constitutive’ role that
mainstream perspective in IR usually desig- norms and ideas play in defining identities and
nated as ‘rationalist’ (see Katzenstein, 1996; hence prescribing proper behaviour on the
Fearon and Wendt, 2002). part of given types of actors.
Although this is not the place to go into the Both these strands are exemplified in the
details of social constructivism (for these, various chapters of the influential volume on
see, e.g., Adler, 2002; Guzzini, 2000), it is The Culture of National Security (1996), edited
fruitful to distinguish between essentially by Peter J. Katzenstein. Although it by no means
‘thinner’ and ‘thicker’ versions, since construc- cuts its roots to mainstream social science (see
tivism incorporates – rather uneasily – an Ruggie, 1998: 38), it takes issue with the ratio-
increasingly broad spectrum of views. The for- nalism of both neorealism and neoliberalism
mer is quintessentially represented by Wendt with regard to the role of both norms and iden-
in his recent treatise on international politics tities in world politics. In particular, it ‘makes
(Wendt, 1999), but also by other ‘modernist’ problematic the state interests that predomi-
constructivists, including Emanuel Adler and nant explanations of national security often
Peter-3383-Chapter-20.qxd 6/8/2006 7:41 PM Page 351
take for granted’, as Katzenstein writes in his et al., 1998) – has as its starting point a critique
introduction (1996: 1). In this volume two of the use of psychological and cognitive factors
studies in particular exemplify a constructivist in the explanation of the role of belief systems in
analysis of foreign policy. The first, by Richard foreign policy, in particular a tendency to focus
Price and Nina Tannenwald, shows that, while a exclusively on individual decision-makers,
rationalist analysis of the non-use of both viewing and analysing beliefs in positivists
nuclear and chemical weapons cannot account terms, and an assumption that language is a
for such policies, a constructivist view, empha- transparent medium without an inner dynamic
sizing the socially constructed nature of deter- of its own (Larsen, 1997: 1–10). Instead of
rence and deterrence weapons, shows that the analysing the belief systems of individual deci-
non-use of these weapons can only be under- sion-makers in this conventional manner, the
stood if one takes ‘into account the develop- emphasis is here put on viewing the discourse
ment of prohibitionary norms that shaped characterizing the foreign policy domain as a
these weapons as unacceptable “weapons of powerful structural constraint, on a high level of
mass destruction”’ (Price and Tannenwald, generality, shaping the foreign policy of the state
1996: 115). Similarly, Martha Finnemore has in question. More specifically, drawing on social
focused on another form of foreign policy constructivist premises, Henrik Larsen has
behaviour which cannot be adequately argued that ‘the framework of meaning within
explained by either realist or liberal theories: which foreign policy takes place is seen as the
humanitarian interventions which have no basis of the way in which interests and goals are
geostrategic and/or economic importance to constructed’ (Larsen, 1999: 453). However,
the interveners in question (Finnemore, contrary to ‘thinner’ constructivists, the
1996a; see also Finnemore, 1996b). Instead, she assumption in this type of discursive approach
argues, this type of behaviour, and the manner is that intersubjective meaning cannot be
in which it has changed and developed since apprehended in or by itself but, rather, that it is
the nineteenth century, cannot be understood constituted by language. As a consequence, dis-
apart from the changing normative context in courses ‘provide the basis on which policy pref-
which it occurs, insofar as ‘international nor- erences, interests and goals are constructed’
mative context shapes the interests of interna- (Larsen, 1999: 453; Waever, 1998). Along simi-
tional actors and does so in both systematic lar lines, Ole Waever has argued for a concep-
and systemic ways’ (Finnemore, 1996a: 154). A tualization of security – as ‘securitization’ –
third study which also exemplifies a construc- based not on the ‘objective’ measures of tradi-
tivist analysis of foreign policy along these tional security studies but on speech act theory
lines is Audie Klotz’s analysis of the role of and its emphasis on language as a privileged
international norms in the international vehicle for gaining and exercising social power.
embargo against the apartheid regime in South In this view, he writes, ‘security is not of inter-
Africa (Klotz, 1995). She argues that the emer- est as a sign that refers to something more real;
gence of an international norm of racial equal- the utterance itself is the act’, and hence ‘some-
ity led states – such as the United States – to thing is a security problem when elites declare
redefine their foreign policy interests despite a it to be so’ (Waever, 1995: 55, 54).
lack of material incentives for so doing. A second, different and broader strand has
recently been presented and discussed by
Discursive approaches Following the Jennifer Milliken (Milliken, 1999: 225, 228–30;
so-called linguistic turn in philosophy and see also 2001). She characterizes discourse the-
social theory, a second holistic-interpretative orists as crossing over and mixing ‘divisions
approach, focusing on the role of language in between poststructuralists, postmodernists and
social inquiry, is slowly but determinedly some feminists and social constructivists’, shar-
making inroads into foreign policy analysis. ing at least the following three commitments:
One strand of this movement – belonging to the viewing discourses as systems of signification
so-called Copenhagen School (see, e.g., Buzan that construct social realities (see, e.g., Milliken,
Peter-3383-Chapter-20.qxd 6/8/2006 7:41 PM Page 352
1996; Mutimer, 1999; Weldes and Saco, 1996); The historical antecedents of this approach
the claim that discourses are productive of the go back to the pioneering work of Richard C.
things defined by the discourse, such as com- Snyder and his associates, focusing on a sys-
mon sense and policy practices (see, e.g., tematic empirical analysis of the actual delib-
Campbell, 1993; Doty, 1996; Huysmans, 1998; erations of foreign policy decision-makers
Waever, 1995; Weber, 1995; Weldes, 1999; (Snyder et al., 1962; see also Paige, 1968).
Weldes and Saco, 1996); and ‘studying domi- Insofar as the focal point in studies of this kind
nating or hegemonic discourses, and their are the reasoned – rather than rational –
structuring of meaning as connected to imple- choices made by decision-makers, certain
menting practices and ways of making these aspects of role theory also exemplify this
intelligible and legitimate’ (see, e.g., Bartelson, approach, at least insofar as the analysis of par-
1995; Fierke, 1998; Neumann, 1998; Sylvester, ticular role conceptions puts the focus on the
1994). Discourse analysts thus focus on signi- reasoning of individual national foreign
ficative practices and the knowledge systems policy-makers and their understanding of the
underlying them, and are as such not only con- international system and the perceived role of
cerned with meta-theoretical critique but also their own states within this larger system (see,
with critical theorizing about the knowledge/ e.g., Holsti, 1987; Hyde-Price, 2000: 42–7; and
power nexus (on the latter, see also Guzzini, the discussion of ‘role-players’ in Hollis and
2000; Neufeld, 1993). Smith, 1990: 155–9, 214–16). The same goes
for more classical understandings of the role of
the ‘national interest’ in foreign policy deci-
Approaches Based on an Interpretative sion-making, based on individual interpreta-
Actor Perspective tions of this much maligned but exceedingly
flexible concept, as well as to the study of the
In their book-length discussion of core metathe- role of crucial decision-makers during crises
oretical issues in IR, Martin Hollis and Steve (see, e.g., Bernstein, 2000: 161–4).
Smith have described individualist interpretative However, a more illustrative and contempo-
approaches to foreign policy as follows: rary exemplar of this type of analysis is Philip
Understanding proceeds by reconstructing at an indi- Zelikow and Condoleezza Rice’s detailed study
vidual level. This Weberian line has been much used in of German reunification (Zelikow and Rice,
International Relations, especially in the sub-field 1995). It offers an insider’s view of the inner-
known as Foreign Policy Analysis. Here the concern is
to understand decisions from the standpoint of the
most workings of the top elites of the United
decision-makers by reconstructing their reasons. The States, the Soviet Union, West Germany, East
foreign policy behaviour of states depends on how indi- Germany, Britain and France in the creation of
viduals with power perceive and analyse situations. a united Germany. The logic of explanation is to
Collective action is a sum or combination of individual determine the thinking of these elites – the rea-
actions. (Hollis and Smith, 1990: 74)
soning behind their choices – and then to prof-
In addition, they make the distinction within fer it in explanation of the immense changes
hermeneutics – which the above approach that occurred during the year following the col-
exemplifies – between understanding individ- lapse of the Berlin Wall. This is ‘thick descrip-
ual actions through social rules and collective tion’ at its best; and although they have been
meanings (a top-down procedure), and chided for eschewing theory altogether in fol-
understanding collective policy through their lowing this strategy (see, e.g., Risse, 1997), it
individual elements (bottom-up). Inasmuch should at the same time be emphasized that,
as the top-down view is quintessentially although no causal analysis (or theorizing) in
the one discussed above in terms of social- the conventional sense is provided, the focus is
institutional approaches, we are here left with most certainly not simply on ‘what’ occurred,
the latter type of focus, which also happens to but also on the ‘why’ and ‘how’ aspects of this
be the least utilized today in the study of process. The assumption underlying this type of
foreign policy. analysis is the counter-factual argument that,
Peter-3383-Chapter-20.qxd 6/8/2006 7:41 PM Page 353
had not the main actors in this historical process integration, usually combining this with
reasoned and made choices the way they actu- empirical research that has lent strong support
ally did, the history of this period would have for such an integrative view of foreign policy
been different. In any case, insofar as ‘why’ issues analysis.
can have both a ‘because of’ and an ‘in order to’ Perhaps the most notable recent example of
implication, and since there are strong philo- such an ambition is provided by a number of
sophical arguments in favour of imputing some studies that have focused on the link between
form of causality also to purposive behaviour domestic structures and foreign policy actions.
(see Carlsnaes, 1986: 32–8), there is no justifica- Peter Katzenstein’s early work (1976, 1978) has
tion for off-hand denigrating this type of an played a pioneering role in paving the way for
approach for being ‘descriptive’ rather than studies of this kind, which have often had the
‘explanatory’. In this connection it should also added advantage of being comparative and
be noted that, despite a deep concern with its hence reaching back – albeit without the same
lack of theoretical anchorage, Risse has been ‘scientistic’ ambitions – to earlier work within
able to utilize this descriptive-analytic study to CFP. Significant research stimulated by this
illustrate the role of ‘communicative action’ and approach has included studies by Matthew
‘friendly persuasion’ in international relations Evangelista (1988, 1995), Risse-Kappen (1991,
(Risse, 2000). Indeed, insofar as the ‘logic of 1994, 1995a) and Jack Snyder (1991; see also
arguing’ – as distinct from the logics of ‘conse- the discussion in Evangelista, 1997). Some of
quentialism’ and ‘appropriateness’ – aims at this work has also taken its cue from Peter
achieving a reasoned consensus on the part of Gourevitch’s notion of the ‘second-image
real life decision-makers (such as Kohl and reversed’, focusing on how international insti-
Gorbachev), this approach seems to be ideally tutions affect foreign policy change via its
suited for analysis from within the interpretative effects on domestic publics and hence on state
actor perspective. actions (Gourevitch, 1978; Checkel, 1999; Keck
and Sikkink, 1998; Risse et al., 1999).
However, the main problem with ‘domestic
structure’ as an integrative bridge is that it
IS A SYNTHETIC APPROACH TO FOREIGN
assumes and hence reinforces the divide
POLICY ANALYSIS FEASIBLE?
between domestic and international politics,
which, as I have suggested above, is highly ques-
This rich flora – indeed, surfeit – of alternative tionable as a feasible foundational baseline for a
approaches to foreign policy analysis raises the sub-discipline that needs to problematize this
question whether it is possible to synthesize or boundary rather than positing it by assumption.
integrate at least some of these, or if we are Furthermore, this argument has various strands
willy-nilly obliged to choose between them. that are not necessarily mutually compatible as
Hollis and Smith, for example, have claimed explanations. Thus, it can refer to an essentially
that there are always two stories to tell – that of holistic structural view, as in Katzenstein’s work
‘explanation’ versus ‘understanding’, corre- on the role of weak versus strong societies
sponding to the distinction above between (Katzenstein, 1976), or in the ‘democratic peace’
‘objectivism’ and ‘interpretativism’ – and that argument (Russett, 1993); to an agency-based
they cannot be combined into one type of nar- view in terms of which domestic structures act
rative (Hollis and Smith, 1990). Similarly, as intervening factors between societal actors
‘holism’ and ‘individualism’ have most often and state action (Checkel, 1997; Risse-Kappen,
been assumed to be in principle mutually 1991); or to more recent constructivist
exclusive categories, forcing us into either a approaches emphasizing the impact of ideas
‘top-down’ or ‘bottom-up’ mode of analysis. and norms – either domestic or international –
However, other scholars – often with a less as sources of foreign policy change (Checkel,
pronounced meta-theoretical bent – have 1997; Finnemore and Sikkink, 1998; Reus-Smit,
argued for the feasibility of such analytical 1999). Given these contending uses to which the
Peter-3383-Chapter-20.qxd 6/8/2006 7:41 PM Page 354
of an enabling kind, causally affecting policy revealed the embarrassing fact that, not only
actions via its effects on the dispositional char- were many of these changes unanticipated, but
acteristics of the agents of policy. (Although also that the events in question were difficult to
not indicated in the figure, foreign policy explain even ex post facto in terms of existing
actions can in turn affect – either by intention theories, models or analytic approaches.
or unintentionally in the form of outcomes – Although this theoretical dearth led to a
both the structural and dispositional dimen- renewed interest in the analysis of foreign
sions, providing for the dynamic interaction policy change (Carlsnaes, 1993; Gustavsson,
over time between agential and structural 1998, 1999; Hermann, 1990; Rosati et al., 1994;
factors, thus invoking the agency–structure see also Koslowski and Kratochwil, 1995),
issue, to be discussed briefly below.) there is little consensus on the best way of
Although this type of an integrative doing so. Given the eclectic nature of the field
framework eschews the dichotomization of as such, as well as the fundamental differences
approaches discussed above, it does not as such between the types of perspectives presented
negate the applicability of any of these – as above, some of which are inherently more
long as they are used when and if analytically amenable to the study of change than others,
appropriate. Indeed, approaches from all the this should of course not come as a surprise.
four types of rock-bottom perspectives dis- At the same time, this issue is seminal to
cussed above can be fully utilized: the ‘struc- the future of the field as a whole, given the
tural’ and ‘social-institutional’ when analysing increased globalization of international
causal links between the structural and dispo- relations – a process arguably undermining the
sitional dimensions; ‘agency-based’ perspec- relative autonomy of the state qua foreign
tives when tracing causal patterns between the policy actor – as well as the emergence of new
dispositional and the intentional dimension; types of foreign policy actors, such as the EU,
and the ‘interpretative actor’ perspective when claiming not only foreign policy competencies
the purpose is to penetrate the teleological of their own but also as representing their
links between intentions and foreign policy member states, hence eroding the sovereignty
actions. of the latter (see, e.g., White, 2001).
A second contentious and topical concern
within the field pertains to the role of ideas in
the explanation of foreign policy. For long
CONCLUDING REMARKS
banished from mainstream social science
explanations, the ideational factor finally gained
To round up this overview I would like to con- full admission in the early 1990s with the pub-
clude by briefly pointing to three theoretical lication of the edited volume on Ideas and
issues that cut across the perspectives dis- Foreign Policy (Goldstein and Keohane, 1993).
cussed above and which, in my opinion, will Underlying this introduction lay the realization
continue to remain topical and controversial in that explanations based solely on rational actors
the study of foreign policy (as in the social maximizing a utility function rooted in material
sciences in general). interests were often inadequate to account fully
The first of these pertains to explaining the for the foreign policy behaviour of states.
dynamics of foreign policy change, both in Instead, it was suggested, ideas too can have an
terms of actor and policy characteristics. independent causal effect on foreign policy
Except for a short burst of interest in the early ‘even when human beings behave rationally to
1980s (Buzan and Jones, 1981; Gilpin, 1981; achieve their ends’ (Goldstein and Keohane,
Goldmann, 1982; Holsti, 1982; Holsti et al., 1993: 5; see also Checkel, 1997; Jacobsen, 1995;
1980; see also Goldmann, 1988), this was not a Risse-Kappen, 1994; Yee, 1996). Although wel-
topic that attracted much attention until the comed by scholars on the interpretative side of
profound transformations occurring at the the epistemological fence, this admission of the
end of that decade. These developments causal efficacy of ideas has nevertheless led to
Peter-3383-Chapter-20.qxd 6/8/2006 7:41 PM Page 356
considerable controversy within the field (see, this overview of the vicissitudes and current
e.g., Koslowski and Kratochwil, 1995; Laffey and condition of foreign policy analysis, since it
Weldes, 1997). The basic criticism is that ‘ratio- touches on the central core of the field itself:
nalists’ continue to think in terms of ‘naturalis- the fact that foreign policy actions are located
tic’ factors, even when conceptualizing ideas at the very centre of the international relations
(viewing them as cognitively and individually of states, incorporating a multitude of influ-
held ‘beliefs’ with causal effects, as well as being ences – structural and agential, as well as inter-
distinct from ‘interests’), whereas the social national, societal and individual – that
constructivist view is that ‘ideational factors continually impinge on them and on their
relate to social action in the form of constitu- decision-makers. To capture these complex
tive rules’, as Ruggie notes (Ruggie, 1998: 38). In and reciprocal processes, and to do so well, is
the light of this view, to speak of ‘ideational the challenge that will persist in energizing this
variables’ is tantamount to perpetrating an oxy- field of study as long as states continue to
moron – a begging of the whole question of remain viable actors within the international
what ideas are and are not, and hence how they system.
are affected by and affect social interaction.
Clearly, this debate is only at its beginning and
will continue to be a focal point for critical NOTE
discussion.
Finally, a third issue, and one that has This is an emended version of a chapter first published in
received considerable theoretical attention Carlsnaes et al. (2002). The author would like to thank the
during the past decade, and continues to do so following colleagues (as well as an anonymous reviewer)
today, is the agency-structure problematique in for commenting on earlier versions of the chapter: Stefano
foreign policy analysis. For all practical pur- Guzzini, Valerie Hudson, Jennifer Milliken, Thomas Risse,
Jerel Rosati, Beth Simmons and Colin Wight. The author
poses Wendt put it on the agenda in a much- also thanks participants at various seminars in Uppsala,
quoted article published in 1987, and since Oslo and Gothenburg.
then it has been hotly debated but hardly
resolved to the satisfaction of all concerned
(Bieler and Morton, 2001; Carlsnaes, 1992, REFERENCES
1994; Dessler, 1989; Doty, 1997; Friedman and
Starr, 1997; Guzzini, 1993; Hollis and Smith,
Adler, Emanuel (1997) ‘Seizing the Middle Ground:
1991, 1992, 1994; Patomäki, 1996; Suganami,
Constructivism in World Politics’, European
1999; Wight, 1999). At the heart of this Journal of International Relations, 3: 319–63.
problem lies the increasingly widespread recog- Adler, Emanuel (2002) ‘Constructivism and
nition that, instead of being antagonistic part- International Relations’ in Walter Carlsnaes,
ners in a zerosum relationship, human agents Thomas Risse and Beth A. Simmons (eds),
and social structures are in a fundamental sense Handbook of International Relations. London:
dynamically interrelated entities, and hence Sage.
that we cannot account fully for the one with- Adler, Imanuel and Barnett, Michael (eds) (1998)
out invoking the other. The ‘problem’ is that, Security Communities. Cambridge: Cambridge
although such views of reciprocal implication University Press.
suggest that the properties of both agents and Allison, Graham T. (1971) Essence of Decision:
Explaining the Cuban Missile Crisis. Boston: Little,
social structures are relevant to a proper under-
Brown.
standing of social behaviour (including the Allison, Graham and Zelikow, Philip (1999) Essence
study of change), we nevertheless (as Wendt of Decision: Explaining the Cuban Missile Crisis.
noted in his original article) ‘lack a self-evident New York: Longman.
way to conceptualize these entities and their Andriole, Stephen J. and Hopple, Gerald W. (1984)
relationship’ (Wendt, 1987: 338). ‘The Rise and Fall of Events Data: From Basic
This is also, perhaps, an appropriate issue Research to Applied Use in the US Department of
and tone of voice with which to put an end to Defense’, International Interactions, 11: 293–309.
Peter-3383-Chapter-20.qxd 6/8/2006 7:41 PM Page 357
Ashley, Richard (1984) ‘The Poverty of Neorealism’, Campbell, David (1993) Politics without Principle:
International Organization, 38: 225–86. Sovereignty, Ethics, and the Narratives of the Gulf
Axelrod, Robert (ed.) (1976) Structure of Decision: War. Boulder: Lynne Rienner.
The Cognitive Maps of Political Elites. Princeton: Caporaso, James A., Hermann, Charles F. and
Princeton University Press. Kegley, Charles W., Jr (1987) ‘The Comparative
Axelrod, Robert and Keohane, Robert O. (1993) Study of Foreign Policy: Perspectives on the
‘Achieving Cooperation Under Anarchy: Future’, International Studies Notes, 13: 32–46.
Strategies and Institutions’, in David A. Baldwin Carlsnaes, Walter (1986) Ideology and Foreign Policy:
(ed.), Neorealism and Neoliberalism: The Problems of Comparative Conceptualization.
Contemporary Debate. New York: Columbia Oxford: Blackwell.
University Press. Carlsnaes, Walter (1992) ‘The Agency–Structure
Baldwin, David A. (1993) ‘Neoliberalism, Problem in Foreign Policy Analysis’, International
Neorealism, and World Politics’, in David A. Studies Quarterly, 36: 245–70.
Baldwin (ed.), Neorealism and Neoliberalism: The Carlsnaes, Walter (1993) ‘On Analysing the
Contemporary Debate. New York: Columbia Dynamics of Foreign Policy Change: A Critique
University Press. and Reconceptualization’, Cooperation and
Bartelson, Jens (1995) A Genealogy of Sovereignty. Conflict, 28: 5–30.
Cambridge: Cambridge University Press. Carlsnaes, Walter (1994) ‘In Lieu of a Conclusion:
Baumann, Rainer, Rittberger, Volker and Wagner, Compatibility and the Agency–Structure Issue in
Wolfgang (2001) ‘Neorealist Foreign Policy Foreign Policy Analysis’, in Walter Carlsnaes and
Theory’, in Volker Rittberger (ed.), German Foreign Steve Smith (eds), European Foreign Policy: The EC
Policy Since Unification: Theories and Case Studies. and Changing Perspectives in Europe. London: Sage.
Manchester: Manchester University Press. Carlsnaes, Walter, Risse, Thomas and Simmons,
Beasley, Ryan (1998) ‘Collective Interpretations: Beth A. (eds), (2002) ‘Handbook of International
How Problem Representations Aggregate in Relations’. London: Sage.
Foreign Policy Groups’, in Donald A. Sylvan and Checkel, Jeffrey T. (1997) Ideas and International
James F. Voss (eds), Problem Representation in Political Change: Soviet/Russian Behavior at the
Foreign Policy Decision Making. Cambridge: End of the Cold War. New Haven: Yale University
Cambridge University Press. Press.
Bendor, Jonathan and Hammond, Thomas H. Checkel, Jeffrey T. (1999) ‘Norms, Institutions, and
(1992) ‘Rethinking Allison’s Models’, American National Identity in Contemporary Europe’,
Political Science Review, 86: 301–22. International Studies Quarterly, 43: 83–114.
Bernstein, Barton J. (2000) ‘Understanding Decision Christensen, Thomas J. (1996) Useful Adversaries:
making, U.S. Foreign Policy, and the Cuban Grand Strategy, Domestic Mobilization, and Sino-
Missile Crisis’, International Security, 25: 134–64. American Conflict, 1947–1958. Princeton:
Bieler, Andreas and Morton, Adam David (2001) Princeton University Press.
‘The Gordian Knot of Agency–Structure in Cohen, Bernard C. (1968) ‘Foreign Policy’, in David L.
International Relations’, European Journal of Sills (ed.), International Encyclopedia of the Social
International Relations, 7: 5–35. Sciences. New York: Macmillan and The Free Press.
Bonham, G. Matthew, Sergeev, Victor M. and Parshin, Cohen, Bernard C. and Harris, Scott A. (1975)
Pavel (1997) ‘The Limited Test-Ban Agreement: ‘Foreign Policy’, in Fred I. Greenstein and Nelson
Emergence of New Knowledge Structures in W. Polsby (eds), Handbook of Political Science,
International Negotiation’, International Studies vol. 6: Policies and Policymaking. Reading, MA:
Quarterly, 41: 215–40. Addison–Wesley.
Brecher, Michael (1972) The Foreign Policy System of Cottam, Martha (1994) Images and Intervention: US
Israel: Setting, Images, Process. London: Oxford Policies in Latin America. Pittsburgh: University
University Press. of Pittsburgh Press.
Bull, Hedley (1977) The Anarchic Society. New York: Cottam, Richard (1977) Foreign Policy Motivation: A
Columbia University Press. General Theory and a Case Study. Pittsburgh:
Buzan, Barry and Jones, J. Barry (1981) Change and University of Pittsburgh Press.
the Study of International Relations. London: Czempiel, Ernst-Otto (1981) Internationale
Pinter. Beziehungen: Ein Konfliktmodell. Munich: UTB.
Buzan, Barry, Waever, Ole and Wilde, Jaap de (1998) Dessler, David (1989) ‘What’s at Stake in the
Security: A New Framework for Analysis. Boulder: Agency–Structure Debate?’, International Organiza-
Lynne Rienner. tion, 43: 441–73.
Peter-3383-Chapter-20.qxd 6/8/2006 7:41 PM Page 358
Deutsch, Karl (1954) Political Community at the Finnemore, Martha (1996b) National Interests in
International Level. Garden City, NY: Doubleday. International Society. Ithaca: Cornell University
Doty, Roxanne Lynn (1996) Imperial Encounters. Press.
Minneapolis: University of Minnesota Press. Finnemore, Martha and Sikkink, Kathryn (1998)
Doty, Roxanne Lynn (1997) ‘Aporia: A Critical ‘International Norm Dynamics and Political
Examination of the Agent–Structure Problem- Change’, International Organization, 52: 887–917.
atique in International Relations Theory’, European Friedman, Gil and Starr, Harvey (1997) Agency,
Journal of International Relations, 3: 365–92. Structure, and International Politics: From Ontology
Doyle, Michael (1997) Ways of War and Peace: to Empirical Inquiry. London: Routledge.
Realism, Liberalism, and Socialism. New York: George, Alexander L. (1979) ‘The Causal Nexus
W.W. Norton. Between Cognitive Beliefs and Decision Making
Dunne, Timothy (1995) ‘The Social Construction of Behavior: The “Operational Code” Belief System’,
International Society’, European Journal of in L.S. Falkowski (ed.), Psychological Models in
International Relations, 1: 367–89. International Politics. Boulder: Westview.
Dunne, Timothy (1998) ‘International Theory and Gilpin, Robert (1981) War and Change in World
the Mirror of History’, European Journal of Politics. Cambridge: Cambridge University Press.
International Relations, 4: 347–62. Glaser, Charles L. (1995) ‘Realists as Optimists: Co-
East, Maurice A. (1978) ‘National Attributes and operation as Self-Help’, in Michael E. Brown, Sean
Foreign Policy’, in Maurice A. East, Stephen A. M. Lynn-Jones and Steven E. Miller (eds), The
Salmore and Charles F. Hermann (eds), Why Perils of Anarchy: Contemporary Realism and
Nations Act. Beverly Hills: Sage. International Security. Cambridge, MA: MIT Press.
Elman, Colin (1996a) ‘Horses for Courses: Why No Goldmann, Kjell (1982) ‘Change and Stability in
Realist Theories of Foreign Policy?’, Security Foreign Policy: Détente as a Problem of
Studies, 6: 7–53. Stabilisation’, World Politics, 34: 230–66.
Elman, Colin (1996b) ‘Cause, Effect, and Con- Goldmann, Kjell (1988) Change and Stability in
sistency: A Response to Kenneth Waltz’, Security Foreign Policy: The Problems and Possibilities of
Studies, 6: 58–61. Détente. New York: Harvester Wheatsheaf.
Evangelista, Matthew (1988) Innovation and the Goldstein, Judith and Keohane, Robert O. (eds)
Arms Race: How the United States and the Soviet (1993) Ideas and Foreign Policy: Beliefs,
Union Develop New Military Technologies. Ithaca: Institutions, and Political Change. Ithaca: Cornell
Cornell University Press. University Press.
Evangelista, Matthew (1995) ‘The Paradox of State Gourevitch, Peter (1978) ‘The Second Image
Strength: Transnational Relations, Domestic Reversed: The International Sources of Domestic
Structures, and Security Policy in Russia and the Politics’, International Organization, 32: 881–912.
Soviet Union’, International Organization, 49: 1–38. Grieco, Joseph M. (1993) ‘Anarchy and the Limits of
Evangelista, Matthew (1997) ‘Domestic Structure and Cooperation: A Realist Critique of the Newest
International Change’, in Michael W. Doyle and Liberal Institutionalism’, in David A. Baldwin
G. John Ikenberry (eds), The New Thinking in (ed.), Neorealism and Neoliberalism: The
International Relations Theory. Boulder: Westview. Contemporary Debate. New York: Columbia
Farkas, Andrew (1996) ‘Evolutionary Models in University Press.
Foreign Policy Analysis’, International Studies Gustavsson, Jakob (1998) The Politics of Foreign
Quarterly, 40: 343–61. Policy Change. Lund: Lund University Press.
Fearon, James and Wendt, Alexander (2002) Gustavsson, Jakob (1999) ‘How Should We Study
‘Rationalism v. Constructivism: A Skeptical View’, Foreign Policy Change?’, Cooperation and
in Walter Carlsnaes, Thomas Risse and Beth A. Conflict, 34: 73–95.
Simmons (eds), Handbook of International Guzzini, Stefano (1993) ‘Structural Power: The
Relations. London: Sage. Limits of Neorealist Power Analysis’, International
Fierke, Karin (1998) Changing Games, Changing Organization, 47: 443–78.
Strategies: Critical Investigations in Security. Guzzini, Stefano (1998) Realism in International
Manchester: Manchester University Press. Relations and International Political Economy.
Finnemore, Martha (1996a) ‘Constructing Norms of London: Routledge.
Humanitarian Intervention’, in Peter J. Katzenstein Guzzini, Stefano (2000) ‘A Reconstruction
(ed.), The Culture of National Security: Norms and of Constructivism in International Relations’,
Identity in World Politics. New York: Columbia European Journal of International Relations,
University Press. 6: 147–82.
Peter-3383-Chapter-20.qxd 6/8/2006 7:41 PM Page 359
Haas, Ernst (1964) Beyond the Nation-State. Hollis, Martin (1994) The Philosophy of Social
Stanford: Stanford University Press. Science. Cambridge: Cambridge University Press.
Haas, Ernst (1990) Knowledge is Power. Berkeley: Hollis, Martin and Smith, Steve (1990) Explaining and
University of California Press. Understanding International Relations. Oxford:
Hanrieder, Wolfram F. (1967) ‘Compatibility and Clarendon Press.
Consensus: A Proposal for the Conceptual Hollis, Martin and Smith, Steve (1991) ‘Beware of
Linkage of External and Internal Dimensions of Gurus: Structure and Action in International
Foreign Policy’, American Political Science Review, Relations’, Review of International Studies,
61: 971–82. 14: 393–410.
Hart, Paul ‘t, Stern, Eric K. and Sundelius, Bengt Hollis, Martin and Smith, Steve (1992) ‘Structure
(eds) (1997) Beyond Groupthink: Political and Action: Further Comments’, Review of
Dynamics and Foreign Policy-Making. Ann Arbor: International Studies, 18: 187–8.
University of Michigan Press. Hollis, Martin and Smith, Steve (1994) ‘Two Stories
Heradstveit, Daniel and Bonham, G. Matthew about Structure and Agency’, Review of
(1986) ‘Decision-Making in the Face of International Studies, 20: 241–51.
Uncertainty: Attributions of Norwegian and Holsti, K.J. (1982) Why Nations Realign: Foreign
American Officials’, Journal of Peace Research, 23: Policy Restructuring in the Postwar World.
339–56. London: Allen & Unwin.
Hermann, Charles F. (1978) ‘Decision Structure and Holsti, K.J. (1987) ‘National Role Conceptions in
Process Influences on Foreign Policy’, in Maurice A. the Study of Foreign Policy’, in Stephen Walker
East, Stephen A. Salmore and Charles F. Hermann (ed.), Role Theory and Foreign Policy Analysis.
(eds), Why Nations Act. Beverly Hills: Sage. Durham, NC: Duke University Press.
Hermann, Charles F. (1990) ‘Changing Course: Holsti, Ole R., North, Robert and Brody, Richard
When Governments Choose to Redirect Foreign (1968) ‘Perception and Action in the 1914 Crisis’,
Policy’, International Studies Quarterly, 34: 3–21. in J. David Singer (ed.), Quantitative Inter-
Hermann, Margaret G. (1974) ‘Leader Personality national Politics: Insights and Evidence. New York:
and Foreign Policy Behavior’, in James N. The Free Press.
Rosenau (ed.), Comparing Foreign Policies: Holsti, Ole R., Siverson, Randolph M. and George,
Theories, Findings, and Methods. New York: Alexander L. (1980) Change in the International
Sage–Halsted. System. Boulder: Westview.
Hermann, Margaret G. (1980a) ‘Assessing the Hopf, Ted (1998) ‘The Promise of Constructivism in
Personalities of Members of the Soviet Politburo’, International Relations Theory’, International
Personality and Social Psychology Bulletin, Security, 23: 171–200.
6: 332–52. Hudson, Valerie M. (ed.) (1997) Culture and Foreign
Hermann, Margaret G. (1980b) ‘Explaining Foreign Policy. Boulder: Lynne Rienner.
Policy Behavior Using Personal Characteristics of Hudson, Valerie M. and Vore, Christopher S. (1995)
Political Leaders’, International Studies Quarterly, ‘Foreign Policy Analysis Yesterday, Today, and
24: 7–46. Tomorrow’, Mershon International Studies Review,
Hermann, Margaret G. (1993) ‘Leaders and Foreign 39: 209–38.
Policy Decision Making’, in Dan Caldwell and Huysmans, Jef (1998) ‘Security! What Do You
Timothy McKeown (eds), Diplomacy, Force, and Mean? From Concept to Thick Signifier’,
Leadership: Essays in Honor of Alexander George. European Journal of International Relations,
Boulder: Westview. 4: 226–55.
Hermann, Margaret G. and Preston, Thomas (1998) Hyde-Price, Adrian (2000) Germany and European
‘Presidents, Leadership Style, and the Advisory Order: Enlarging NATO and the EU. Manchester:
Process’, in Eugene R. Wittkopf and James M. Manchester University Press.
McCormick (eds), Domestic Sources of American Jacobsen, John Kurt (1995) ‘Much Ado about Ideas:
Foreign Policy. New York: Rowman and Littlefield. The Cognitive Factor in Economic Policy’, World
Herrmann, Richard K. (1985) Perceptions and Politics, 47: 283–310.
Behavior in Soviet Foreign Policy. Pittsburgh: Janis, Irving J. (1982) Groupthink: Psychological
University of Pittsburgh Press. Studies of Policy Decisions and Fiascoes. Boston:
Herrmann, Richard K. and Fischerkeller, Michael Houghton Mifflin.
(1995) ‘Beyond the Enemy Image Spiral Model: Jervis, Robert (1976) Perception and Misperception
Cognitive- Strategic Research After the Cold War’, in International Politics. Princeton: Princeton
International Organization, 49: 415–50. University Press.
Peter-3383-Chapter-20.qxd 6/8/2006 7:41 PM Page 360
Kahler, Miles (1997) ‘Inventing International Laffey, Mark and Weldes, Jutta (1997) ‘Beyond
Relations: International Relations Theory After Belief: Ideas and Symbolic Technologies in the
1945’, in Michael W. Doyle and G. John Ikenberry Study of International Relations’, European
(eds), The New Thinking in International Journal of International Relations, 3: 193–237.
Relations Theory. Boulder: Westview. Larsen, Henrik (1997) Foreign Policy and Discourse
Kahler, Miles (1998) ‘Rationality in International Analysis: France, Britain and Europe. London:
Relations’, International Organization, 52: 919–41. Routledge.
Kahneman, D. and Tversky, A. (1979) ‘Prospect Larsen, Henrik (1999) ‘British and Danish European
Theory: An Analysis of Decision Under Risk’, Policies in the 1990s: A Discourse Approach’,
Econometrica, 47: 263–91. European Journal of International Relations,
Karbo, Juliet (1998) ‘Power Politics in Foreign Policy: 5: 451–83.
The Influence of Bureaucratic Minorities’, European Larson, Deborah W. (1985) Origins of Containment:
Journal of International Relations, 4: 67–97. A Psychological Explanation. Princeton: Princeton
Katzenstein, Peter J. (1976) ‘International Relations University Press.
and Domestic Structures: Foreign Economic Larson, Deborah W. (1997) Anatomy of Mistrust:
Policies of Advanced Industrial States’, U.S.–Soviet Relations during the Cold War. Ithaca:
International Organization, 30: 1–45. Cornell University Press.
Katzenstein, Peter J. (1978) Between Power and Plenty: Layne, Christopher (1995) ‘Kant or Cant: The Myth of
Foreign Economic Policies of Advanced Industrial the Democratic Peace’, in Michael E. Brown, Sean
States. Madison: University of Wisconsin Press. M. Lynn-Jones and Steven E. Miller (eds), The Perils
Katzenstein, Peter J. (1996) ‘Introduction: of Anarchy: Contemporary Realism and Inter-
Alternative Perspectives on National Security’, in national Security. Cambridge, MA: MIT Press.
Peter J. Katzenstein (ed.), The Culture of National Levy, Jack S. (1994) ‘Learning and Foreign Policy:
Security: Norms and Identity in World Politics. Sweeping a Conceptual Minefield’, International
New York: Columbia University Press. Organization, 48: 279–312.
Keck, Margaret E. and Sikkink, Kathryn (1998) Levy, Jack S. (1997) ‘Prospect Theory, Rational
Activists Beyond Borders: Advocacy Networks in Choice, and International Relations’, Inter-
International Politics. Ithaca: Cornell University national Studies Quarterly, 41: 87–1121.
Press. Lynn–Jones, Sean M. and Miller, Steven E. (1995)
Kegley, Charles W., Jr (1980) ‘The Comparative ‘Preface’, in Michael E. Brown, Sean M. Lynn-
Study of Foreign Policy: Paradigm Lost?’, pub- Jones and Steven E. Miller (eds), The Perils of
lished by Institute of International Studies, Anarchy: Contemporary Realism and International
University of South Carolina, Essay Series No. 10. Security. Cambridge, MA: MIT Press.
Keohane, Robert O. (1993) ‘Institutional Theory March, James G. and Olsen, Johan P. (1998) ‘The
and the Realist Challenge After the Cold War’, in Institutional Dynamics of International Political
David A. Baldwin (ed.), Neorealism and Orders’, International Organization, 52: 943–69.
Neoliberalism: The Contemporary Debate. New McDermott, Ross (1998) Risk-Taking in International
York: Columbia University Press. Politics: Prospect Theory in American Foreign Policy.
Khong, Yuen Foong (1992) Analogies at War: Korea, Ann Arbor: University of Michigan Press.
Munich, Dien Bien Phu, and the Vietnam McGowan, Patrick J. and Shapiro, Howard B. (1973)
Decisions of 1965. Princeton: Princeton University The Comparative Study of Foreign Policy: A Survey
Press. of Scientific Findings. Beverly Hills: Sage.
Klotz, Audie J. (1995) Protesting Prejudice: Apartheid Mearsheimer, John J. (1995) ‘Back to the Future:
and the Politics of Norms in International Instability in Europe After the Cold War’, in
Relations. Ithaca: Cornell University Press. Michael E. Brown, Sean M. Lynn-Jones and
Koslowski, Rey and Kratochwil, Friedrich V. (1995) Steven E. Miller (eds), The Perils of Anarchy:
‘Understanding Change in International Politics: Contemporary Realism and International Security.
The Soviet Empire’s Demise and the Cambridge, MA: MIT Press.
International System’, in Richard Ned Lebow and Milliken, Jennifer (1996) ‘Prestige and Reputation in
Thomas Risse-Kappen (eds), International American Foreign Policy and American Realism’,
Relations Theory and the End of the Cold War. in Francis Beer and Robert Hariman (eds), Post-
New York: Columbia University Press. Realism: The Rhetorical Turn in International
Kratochwil, Friedrich (1989) Rules, Norms, and Relations. East Lansing: Michigan State University
Decisions. Cambridge: Cambridge University Press. Press.
Peter-3383-Chapter-20.qxd 6/8/2006 7:41 PM Page 361
Milliken, Jennifer (1999) ‘The Study of Discourse in Rationality in International Relations Theory.
International Relations: A Critique of Research Princeton: Princeton University Press.
and Methods’, European Journal of International Rhodes, Edward (1994) ‘Do Bureaucratic Politics
Relations, 5: 225–54. Matter? Some Disconfirming Findings from the
Milliken, Jennifer (2001) Conflict Possibilities: The Case of the US Navy’, World Politics, 47: 1–41.
Social Construction of the Korean War. Manchester: Ripley, Brian (1995) ‘Cognition, Culture, and
Manchester University Press. Bureaucratic Politics’, in Laura Neack, Jeanne
Moravcsik, Andrew (1997) ‘Taking Preferences A.K. Hey and Patrick J. Haney (eds), Foreign
Seriously: A Liberal Theory of International Policy Analysis: Continuity and Change in Its
Politics’, International Organization, 51: 513–53. Second Generation. Englewood Cliffs: Prentice
Morgenthau, Hans J. (1948) Politics Among Nations: Hall.
The Struggle for Power and Peace. New York: Risse, Thomas (1997) ‘The Cold War’s Endgame
Alfred A. Knopf. and German Unification’, International Security,
Munton, Don (1976) ‘Comparative Foreign Policy: 21: 159–85.
Fads, Fantasies, Orthodoxies, and Perversities’, in Risse, Thomas (2000) ‘“Let’s Argue!”: Communicative
James N. Rosenau (ed.), In Search of Global Action in World Politics’, International Organization,
Patterns. New York: The Free Press. 54: 1–39.
Mutimer, David (1999) The Weapons State: Risse, Thomas, Ropp, Steven and Sikkink, Kathryn
Proliferation and the Imagination of Security. (eds) (1999) The Power of Principles: International
Boulder: Lynne Rienner. Human Rights Norms and Domestic Change.
Neack, Laura, Hey, Jeanne A.K. and Haney, Patrick J. Cambridge: Cambridge University Press.
(1995) ‘Generational Change in Foreign Policy Risse-Kappen, Thomas (1991) ‘Public Opinion,
Analysis’, in Laura Neack, Jeanne A.K. Hey and Domestic Structures, and Foreign Policy in
Patrick J. Haney (eds), Foreign Policy Analysis: Liberal Democracies’, World Politics, 43: 479–512.
Continuity and Change in Its Second Generation. Risse-Kappen, Thomas (1994) ‘Ideas Do Not Float
Englewood Cliffs: Prentice Hall. Freely: Transnational Coalitions, Domestic
Neufeld, Mark (1993) ‘Reflexivity and International Structures, and the End of the Cold War’,
Theory’, Millennium: Journal of International International Organization, 48: 185–214.
Studies, 22: 53–76. Risse-Kappen, Thomas (ed.) (1995a) Bringing
Neumann, Iver (1998) Uses of the Other: The ‘East’ Transnational Relations Back In: Non-State Actors,
in European Identity Formation. Minneapolis: Domestic Structures, and International Institutions.
University of Minnesota Press. Cambridge: Cambridge University Press.
Onuf, Nicholas (1989) World of Our Making. Risse-Kappen, Thomas (1995b) Cooperation among
Columbia: University of South Carolina Press. Democracies: The European Influence on US
Oye, Kenneth (ed.) (1985) Cooperation Under Foreign Policy. Princeton: Princeton University
Anarchy. Princeton: Princeton University Press. Press.
Paige, Glenn D. (1968) The Korean Decision. Rosati, Jerel (1987) The Carter Administration’s
New York: The Free Press. Quest for Global Community: Beliefs and Their
Patomäki, Heikki (1996) ‘How to Tell Better Stories Impact on Behavior. Columbia: University of
about World Politics’, European Journal of South Carolina Press.
International Relations, 2: 105–33. Rosati, Jerel A. (1995) ‘A Cognitive Approach to the
Peterson, V. Spike (ed.) (1992) Gendered States. Study of Foreign Policy’, in Laura Neack, Jeanne
Boulder: Lynne Rienner. A.K. Hey and Patrick J. Haney (eds), Foreign
Price, Richard and Tannenwald, Nina (1996) ‘Norms Policy Analysis: Continuity and Change in Its
and Deterrence: The Nuclear and Chemical Second Generation. Englewood Cliffs: Prentice
Weapons Taboo’, in Peter J. Katzenstein (ed.), The Hall.
Culture of National Security: Norms and Identity in Rosati, Jerel A. (2000) ‘The Power of Human
World Politics. New York: Columbia University Cognition in the Study of World Politics’,
Press. International Studies Review, 2: 45–75.
Renshon, S.A. and Larson, Deborah W. (eds) (2001) Rosati, Jerel, Hagan, Joe D. and Sampson, Martin W.
Good Judgement in Foreign Policy. Lanham, MD: (eds) (1994) Foreign Policy Restructuring: How
Rowman and Littlefield. Governments Respond to Global Change.
Reus-Smit, Christian (1999) The Moral Purpose of Columbia, SC: University of South Carolina
the State: Culture, Social Identity, and Institutional Press.
Peter-3383-Chapter-20.qxd 6/8/2006 7:41 PM Page 362
Rose, Gideon (1998) ‘Neoclassical Realism and Policymaking? A Symposium’, Mershon Inter-
Theories of Foreign Policy’, World Politics, 51: national Studies Review, 42: 205–55.
144–72. Suganami, Hidemi (1999) ‘Agents, Structures,
Rosenau, James N. (ed.) (1969) Linkage Politics: Narratives’, European Journal of International
Essays on the Convergence of National and Relations, 5: 365–86.
International Politics. New York: Free Press. Sylvan, Donald A. and Voss, James F. (eds) (1998)
Ruggie, John Gerard (1998) Constructing the World Problem Representation in Foreign Policy Decision
Polity. London: Routledge. Making. Cambridge: Cambridge University Press.
Rummel, Rudolph J. (1972) The Dimensions of Sylvester, Christine (1994) Feminist Theory and
Nations. Beverly Hills: Sage. International Relations in a Postmodern Era.
Russett, Bruce (1993) Grasping the Democratic Cambridge: Cambridge University Press.
Peace: Principles for a Post-Post-Cold War World. Tetlock, Philip E. (1979) ‘Identifying Victims of
Princeton: Princeton University Press. Groupthink from Public Statements of Decision
Scharpf, Fritz W. (1997) Games Real Actors Play: Makers’, Journal of Personality and Social
Actor-Centered Institutionalism in Policy Research. Psychology, 37: 1314–24.
Boulder: Westview. Tickner, J. Ann (1993) Gender in International
Schmidt, Brian C. (2002) ‘On the History and Relations. New York: Columbia University Press.
Historiography of International Relations’, in Tickner, J. Ann (2002) ‘Feminist Perspectives on
Walter Carlsnaes, Thomas Risse and Beth A. International Relations’, in Walter Carlsnaes,
Simmons (eds), Handbook of International Thomas Risse and Beth A. Simmons (eds), Hand-
Relations. London: Sage. book of International Relations. London: Sage.
Schneider, Gerald (1997) ‘Die bürokratische Politik der Van Evera, Stephen (1990/91) ‘Primed for Peace:
Aussenpolitikanalyse: Das Erbe Allisons im Licht Europe After the Cold War’, International
der gegenwärtigen Forschungspraxis’, Zeitschrift für Security, 15: 7–57.
Internationale Beziehungen, 4: 107–23. Vasquez, John (1983) The Power of Power Politics: A
Schweller, Randall L. (1996) ‘Neorealism’s Status Critique. New Brunswick: Rutgers University
Quo Bias: What Security Dilemma?’, Security Press.
Studies, 5: 90–121. Vertzberger, Yaacov Y.I. (1990) The World in Their
Schweller, Randall L. (1998) Deadly Imbalance: Minds: Information Processing, Cognition, and
Tripolarity and Hitler’s Strategy of World Perception in Foreign Policy Decisionmaking.
Conquest. New York: Columbia University Press. Stanford: Stanford University Press.
Singer, Eric G. and Hudson, Valerie M. (eds) (1992) Waever, Ole (1995) ‘Securitization and Desecuritiza-
Political Psychology and Foreign Policy. Boulder: tion’, in Ronnie Lipschutz (ed.), On Security.
Westview Press. New York: Columbia University Press.
Smith, Steve (1986) ‘Theories of Foreign Policy: An Waever, Ole (1998) ‘Explaining Europe by Decoding
Historical Overview’, Review of International Discourses’, in Anders Wivel (ed.), Explaining
Studies, 12: 13–29. European Integration. Copenhagen: Copenhagen
Smith, Steve (1987) ‘CFP: A Theoretical Critique’, Political Studies Press.
International Studies Notes, 13: 47–8. Walker, R.B.J. (1993) Inside/Outside: International
Snyder, Jack (1991) Myths of Empire: Domestic Relations as Political Theory. Cambridge:
Politics and International Ambition. Ithaca: Cambridge University Press.
Cornell University Press. Walker, Stephen G. (1990) ‘The Evolution of
Snyder, Richard C., Bruck, H.W. and Sapin, B. Operational Code Analysis’, Political Psychology,
(1954) ‘Decision-Making as an Approach to the 11: 403–17.
Study of International Politics’, Foreign Policy Walker, Stephen G. (1995) ‘Psychodynamic Processes
Analysis Project Series No. 3. Princeton: Princeton and Framing Effects in Foreign Policy Decision-
University Press. Making: Woodrow Wilson’s Operational Code’,
Snyder, Richard C., Bruck, H.W. and Sapin, B. (eds) Political Psychology, 16: 697–717.
(1962) Foreign Policy Decision Making: An Walker, Stephen G., Schafer, Mark and Young,
Approach to the Study of International Politics. Michael D. (1998) ‘Operational Codes and Role
New York: Free Press. Identities: Measuring and Modeling Jimmy
Steinbruner, John D. (1974) The Cybernetic Theory Carter’s Operational Code’, International Studies
of Decision. Princeton: Princeton University Press. Quarterly, 42: 173–88.
Stern, Eric and Verbeek, Bertjan (1998) ‘Whither Walt, Stephen M. (1995) ‘Alliance Formation and
the Study of Governmental Politics in Foreign the Balance of World Power’, in Michael E.
Peter-3383-Chapter-20.qxd 6/8/2006 7:41 PM Page 363
Brown, Sean M. Lynn-Jones and Steven E. Miller Wight, Colin (1999) ‘They Shoot Horses Don’t
(eds), The Perils of Anarchy: Contemporary They? Locating Agency in the Agent–Structure
Realism and International Security. Problematique’, European Journal of International
Waltz, Kenneth (1979) Theory of International Relations, 5: 109–42.
Politics. Reading, MA: Addison–Wesley. Waltz, Wilkenfeld, Jonathan, Hopple, Gerald W., Rossa,
Kenneth N. (1996) ‘International Politics is not Paul J. and Andriole, Stephen J. (1980) Foreign
Foreign Policy’, Security Studies, 6: 54–7. Policy Behavior: The Interstate Behavior Analysis
Weber, Cynthia (1995) Simulating Sovereignty: Model. Beverly Hills: Sage.
Intervention, the State and Symbolic Exchange. Wohlforth, William (1993) The Elusive Balance:
Cambridge: Cambridge University Press. Power and Perceptions during the Cold War.
Welch, David A. (1998) ‘A Positive Science of Ithaca: Cornell University Press.
Bureaucratic Politics?’, Mershon International Yee, Albert S. (1996) ‘The Causal Effect of Ideas on
Studies Review, 42: 210–16. Politics’, International Organization, 50: 69–108.
Weldes, Jutta (1999) Constructing International Young, Michael D. (1996) ‘Cognitive Mapping
Interests. Minneapolis: University of Minnesota Meets Semantic Networks’, Journal of Conflict
Press. Resolution, 40: 395–414.
Weldes, Jutta and Saco, Diana (1996) ‘Making State Young, Michael D. and Shafer, Mark (1998) ‘Is There
Action Possible: The United States and the Method in Our Madness? Ways of Assessing
Discursive Construction of “The Cuban Cognition in International Relations’, Mershon
Problem”, 1960–1994’, Millennium, 25: 361–98. International Studies Review, 42: 63–96.
Wendt, Alexander (1987) ‘The Agency–Structure Zakaria, Fareed (1995) ‘Realism and Domestic
Problem in International Relations’, International Politics: A Review Essay’, in Michael E. Brown, Sean
Organization, 41: 335–70. M. Lynn- Jones and Steven E. Miller (eds), The
Wendt, Alexander (1999) Social Theory of Perils of Anarchy: Contemporary Realism and
International Politics. Cambridge: Cambridge International Security. Cambridge, MA: MIT Press.
University Press. Zakaria, Faheed (1998) From Wealth to Power: The
White, Brian (1999) ‘The European Challenge to Unusual Origins of America’s World Role.
Foreign Policy Analysis’, European Journal of Princeton: Princeton University Press.
International Relations, 5: 37–66. Zelikow, Philip and Rice, Condoleezza (1995)
White, Brian (2001) Understanding European Germany Unified and Europe Transformed.
Foreign Policy. London: Palgrave. Cambridge, MA: Harvard University Press.
Peter-3383-Chapter-20.qxd 6/8/2006 7:41 PM Page 364
Peter-3383-Chapter-21.qxd 6/8/2006 7:41 PM Page 365
21
Criminal Justice Policy
TIM NEWBURN1
It is clear enough that criminal conduct does not policing and punishment there have developed
determine the kind of penal action a society adopts … huge state-managed apparatuses, and vast
[I]t is not ‘crime’ or even criminological knowledge
about crime which most affects policy decisions, but
bodies of laws, rules and regulations, aimed at
rather the ways in which ‘the crime problem’ is officially controlling crime. What I want to do in this short
perceived and the political positions to which these overview is examine this shifting landscape,
perceptions give rise. (Garland, 1990: 20) looking at the developing functions, purposes
(avowed and implied) and techniques that have
Intriguingly, in the main, scholars of social and developed in response to changing perceptions
public policy have tended to ignore the area of of the problem of crime and social order. My
criminal justice. Compared with, say, health, focus will be primarily upon the USA and the
education, welfare, and culture, criminal justice UK, partly because these are the jurisdictions
has been relatively invisible. And, yet, the cre- I am most familiar with, and whose histories of
ation and maintenance of systems for protect- crime and control have been most extensively
ing against the breakdown of internal social documented, but also because it is argued that
order is generally thought to be among the key they are, in different ways perhaps, at the fore-
characteristics and functions that define the front of the changes sweeping the penal systems
modern nation state (Weber, 1978). Interestingly, of a great number of late modern societies
just as it appears that the ability of modern (Wacquant, 1999).
states to lay claim to ‘sovereignty’ in the area I begin by briefly looking at historical trends
of security and order is in decline, so criminal in crime before moving on to look at the emer-
justice and penal policy are drawing greater gence of the modern criminal justice state. As
attention from scholars beyond the immediate the penal system has developed and changed
confines of criminology. Indeed, it is arguably so the aims of criminal justice policy have
the case that those self-same social transforma- altered and the third section of the chapter
tions that are reconfiguring our responses to looks at these changing rationales and philoso-
criminality also help to account for the degree phies and how they are related to broader
of attention such matters now enjoy. structural processes. In recent times, in partic-
The last two centuries or so have seen the ular, it appears those broad social, cultural,
progressive rationalisation and bureaucratisa- political and economic processes that are often
tion of criminal justice and penal processes. encapsulated within the term ‘late modernity’
From localised, community-based systems of have been accompanied by a set of often
Peter-3383-Chapter-21.qxd 6/8/2006 7:41 PM Page 366
quite radical shifts in the organisation and and, indeed, there are now regular comparative
deployment of punishment and other strate- surveys, such as the International Crime Victims
gies for the control of crime. In this connection, Survey (Nieuwbeerta, 2002).
the final two sections of the chapter explore the It is widely believed that we live in times of
emergence of what has been characterised as a unprecedented levels of crime. Whether such
new and harshly punitive ‘culture of control’ beliefs are accurate rather depends on the
(Garland, 2001a) in America and Britain, as timeframe being utilised. It is certainly the case
well as briefly looking at some of the slightly compared with, say, the 1940s, 1950s or 1960s
divergent trends evident in other jurisdictions. that current levels of crime are very high.
However, if we take a longer historical perspec-
tive then there is rather reliable evidence to
suggest that previous eras were characterised
TRENDS IN CRIME
by very high levels of crime and disorder, even
by contemporary standards (Gurr, 1989). In
Before looking at what has happened to levels public policy terms, of course, political time-
of crime it is important to pause momentarily frames tend to be very much shorter. It is here
to consider the basis upon which we ‘know’ that we confront an interesting irony. It is that,
and judge levels of crime. Broadly speaking although the public tend to believe that crime
there are two main methods used for measur- is rising, it is, in fact, falling and has been doing
ing and tracking trends in crime. One is taken so for some time. Figures 21.1 and 22.2, based
from data collected routinely by law enforce- on NCVS data, show trends in the two main
ment agencies concerning crimes reported by categories of crime in the past 30 years in
the public or otherwise coming to the atten- America. The former indicates that property
tion of the authorities. In the United States crime – which accounts of the bulk of crime –
these are referred to as the Uniform Crime has been in fairly consistent decline since
Reports (UCR) and are collected and collated around 1975. Although the trend in violent
by the FBI. In England such data are collected crime is somewhat different, again it hardly
by the police and are generally referred to as indicates a particular ‘crisis’ in most recent
recorded crime statistics. Similar methods of times. Thus, although there were some rises
tracking crime via police records are used in and falls, violent crime remained relatively sta-
most jurisdictions. The second main method ble between the mid-1970s and the mid-1990s.
uses survey methods to elicit information from Since 1995 violent crime has been declining
a representative sample of the population significantly in the USA. In neither case does
about their experiences of crime – primarily as this sit easily with popular representations.
victims wherever this is the case – usually over Similarly, if we turn to crime trends in the
the previous 12 months. The US National UK we see a similar pattern. Figs 21.3 and 21.4
Crime Victimization Survey (NCVS) was are based on official crime statistics and the
established in 1972 and has been undertaken BCS respectively. The former appears to show
twice a year since then. In the UK, a similar something of a rise in crime over the past
approach underpins the British Crime Survey decade. However, the bulk of this change is
(BCS). This was first undertaken in 1981 and accounted for by changes in the rules by which
has run intermittently since then – though it crime is counted. By contrast, crime trends, as
too is now undertaken annually. For a number measured by the BCS – generally considered to
of reasons it is now generally assumed by most be more accurate – indicate that crime overall
commentators that victimisation surveys are a is likely to have diminished since the mid-
more accurate measure of crime levels and 1990s.
trends than data collected by law enforcement As the quote from Garland at the top of this
agencies (Farrington and Langan, 2004). As a chapter indicates, in recent times at least it has
consequence such surveys are becoming not been trends in crime that have been the
increasingly common in other jurisdictions dominant feature in determining the shape
Peter-3383-Chapter-21.qxd 6/8/2006 7:41 PM Page 367
600.0
500.0
per 1000 households
400.0
300.0
200.0
100.0
0.0
73 75 77 79 81 83 85 87 89 91 93 95 97 99 01 03
19 19 19 19 19 19 19 19 19 19 19 19 19 19 20 20
60.0
50.0
per 1000 persons
40.0
30.0
20.0
10.0
0.0
73
75
77
79
81
83
85
87
89
91
93
95
97
99
01
03
19
19
19
19
19
19
19
19
19
19
19
19
19
19
20
20
and nature of criminal justice policy in many greater detail below, the dislocation between
developed societies (the USA and UK in par- crime levels and crime policy is evident else-
ticular perhaps). As we have seen for much of where. In the UK, for example, the period since
the past 30 years crime in the USA has either the early 1990s has seen a rapidly expanding
been stable or in decline. During the same prison population – at almost precisely the
period the prison population in the USA has point at which crime started to reduce. Given
been expanding at an unparalleled rate. In that it appears that criminal justice policy
some resects it is possible to argue that appears not to be some direct response to
America is something of an exception in this immediate trends in crime, we are required to
regard; it is certainly the case, for example, look elsewhere in order to understand how
that its incarceration rate is incomparably penal policies come to take the form they do.
higher than that in any other developed lib- We begin by looking at the origins and devel-
eral democracy. Nevertheless, as we will see in opment of the modern criminal justice system.
Peter-3383-Chapter-21.qxd 6/8/2006 7:41 PM Page 368
7000
20000
No. of incidents in 000s
15000
10000
5000
0
1981 1983 1987 1991 1993 1995 1997 1999 2001/2 2002/3 2003/4
THE EMERGENCE OF THE MODERN system. Though the criminal justice systems of
CRIMINAL JUSTICE SYSTEM the US and the UK have been shaped by differ-
ent historical pressures and circumstances,
Walker (1998) divides his history of American there is nevertheless considerable commonal-
criminal justice into three eras: the colonial ity between the two jurisdictions, especially
era, which lasted roughly 150 years from the over the past 200 years.
late 1600s until the 1820s prior to the estab- In the colonial era in the USA, and prior to
lishment of a formal criminal justice system; the industrial revolution in the UK, crime con-
an era from the 1820s to the 1920s in which all trol was relatively informal and community-
the main institutions of criminal justice were based. Communities were small and enforcing
established; and the modern era, in which dual moral codes was relatively straightforward. As
processes of consolidation and reform have one legal historian puts it, they ‘had both the
shaped and reshaped the criminal justice will and the ability to enforce laws against
Peter-3383-Chapter-21.qxd 6/8/2006 7:41 PM Page 369
fornication, sins of the flesh, minor vices, and From this point on, however, transportation
bad behaviour. They punished these offences declined and the use of imprisonment began
the way autocratic fathers or mothers punish to grow.
children; they made heavy use of shame and The system of punishment in Victorian
shaming’ (Friedman, 1993: 37). Confession Britain differed quite significantly from that of
and repentance were key parts of the system, the late eighteenth century. The use of the
and public identification and humiliation death penalty declined markedly throughout
through branding and letter-wearing (‘A’ for the 1800s, public ceremonies of execution
adulterer for example) were not uncommon. ceased in 1868 and corporal punishment of
The stocks, the ducking stool and, for more adults was rare by the second half of the
serious offences, bodily mutilation or banish- century. Similarly, whereas in the Colonial era in
ment, were also invoked. Then, finally, there America punishment was conducted publicly,
was the death penalty, though, according to the Revolution had a major impact on criminal
Friedman (1993: 41), ‘on the whole, English justice, bringing formal criminal codes and,
law was more liberal with capital punishment from the late eighteenth century, the creation
than colonial law. In England, men and women of the modern prison. As Garland (1985) and
swung from the gallows for theft, robbery, bur- others have noted, the steady decline in the
glary; in the colonies, this was exceptionally infliction of pain upon the body was paralleled
rare’. Where it was used, the death penalty was by increasing use of incarceration as the main
much more common in the southern United form of punishment of adult offenders.
States than in the north, and was used most According to Ignatieff (1978: 90) this shift
frequently on slaves. indicated ‘a loss of confidence in the morality
In Britain Radzinowicz (1948: 231) notes and efficacy of ritual punishments, a growing
that according to ‘one line of contemporary resistance to the idea that the state should
thought on penal matters eighteenth century share the infliction of the punishment with the
criminal law was insufficiently severe to afford community’.
adequate protection against crime. Death, Put crudely, imprisonment moved from
which … was then the appointed penalty for a being merely a repository for those awaiting
large – and growing – number of offences, was trial, sentence or death in the sixteenth and
considered too mild a punishment for a great seventeenth centuries, to a site where punish-
many of them’. Though the death penalty was ment was inflicted on an increasingly wide
the focus of the penal system in medieval range of offenders during the course of the
times, and levels of capital punishment were eighteenth and nineteenth centuries. Although
high, it underwent something of a brief boom in Victorian Britain and late nineteenth century
from about the mid-seventeenth century America the formal priority of the prison
through to the early eighteenth. The so-called system remained ‘the repression of crime’
‘bloody code’, under which a large number of through deterrence (Garland, 1985) neverthe-
new capital offences were placed on the statute less a growing emphasis on the reform and
book from 1688, came under increasing attack rehabilitation of offenders was emerging.
and much of the century was characterised by From the mid-twentieth century, for approxi-
a search for viable secondary punishments mately three decades, the notion of ‘treatment
(Sharpe, 1990). Transportation was the other and training’ became the explicit guiding prin-
major form of judicial punishment in Britain ciple of the prison system.
in the latter half of the seventeenth century, Like the prison, so a body of uniformed police
but by the turn of the century a number of agents organised bureaucratically, functioning to
British colonies were beginning to refuse to maintain order and detect and investigate crime,
receive any further convicts. Nevertheless, by is similarly a product of the modern industrial
the 1760s transportation to the American nation state. In the pre-industrial era, ‘policing’
colonies accounted for at least 70% of all sen- was a community-based, less formal set of activ-
tences at the central criminal court in London. ities. In eighteenth century America, for example,
Peter-3383-Chapter-21.qxd 6/8/2006 7:41 PM Page 370
a system of ‘brotherly watch’ existed in which legitimacy. Declining police legitimacy coincided
community members were expected to monitor broadly with, and in part signalled, the ending
each others’ behaviour and to report drunken- of the ‘modern’ period in criminal justice and
ness, sexual promiscuity, heretical views and fail- penal policy. This was the period in which the
ure to attend Church, among other activities. In ‘solidarity project’ (Garland, 1996) – the post-
the UK the establishment of formal policing was Enlightenment aim in which the state was the
preceded by community-based systems, such as guarantor of full citizenship and security for
the ‘hue and cry’, in which local citizens took all – was increasingly eclipsed by market
responsibility for raising the alarm and for chas- forces. Recent decades have seen the emer-
ing down the offender. On both sides of the gence of a rapidly expanding mixed economy
Atlantic at this time the maintenance of order in policing and in other areas of criminal jus-
was essentially a civic duty shared, at least in tice. Crucially, this period has also seen what to
principle, by all. many appears to be a decisive shift in what are
Eighteenth-century England was charac- believed to be the purposes and ambitions of
terised by increasing concerns about crime. By our criminal justice and penal policies, and it is
the mid- to late eighteenth century, crime and to these that we turn next.
disorder was perceived to pose a threat to
social stability. It was around this time that
what we now understand as ‘the police’
THE CHANGING GOALS OF CRIMINAL
emerged on both sides of the Atlantic – in 1829
JUSTICE AND PENAL POLICY
in London, in Boston in 1838 and New York
City in 1845 (Miller, 1975). The nineteenth
and early twentieth centuries saw the creation In pre-industrial/colonial times much punish-
of all the fundamental institutions of the ment was public in character and, as such, was
modern criminal justice system: the prison, the designed to shame, to bring forth expressions
police, the courts and related systems of crim- of guilt, remorse and repentance. Loss of free-
inal prosecution, probation and, in due course, dom – through imprisonment – was far from a
an increasingly complex array of non-custodial common response to criminal infractions and
penalties. Toward the end of the nineteenth was not assumed, as yet, to be an effective
century separate systems for dealing with method for stimulating reform. By the nine-
juvenile offenders also emerged in both the teenth century this had all changed.
US and the UK. The first half of the twentieth The post-colonial era in America saw the
century, dominated of course by two world operation of two distinct approaches to
wars, saw the consolidation and reform of imprisonment. One, most notably at the
modern criminal justice systems. Policing Eastern State penitentiary, opened in 1829,
provides a good example. Kelling and Moore subjected inmates to total solitary confinement
(1988) argue that the period between the with only one hour of exercise allowed and no
introduction of formal police departments in contact permitted with other prisoners. By
the early to mid-1800s and the turn of the contrast, the Auburn system in New York per-
twentieth century can be characterised as a mitted communal work and eating, though in
‘political era’, characterised by the struggles silence, otherwise confining inmates to indi-
between various groups to control the police. vidual cells. Both systems reflected the view
The ‘reform era’ emerged in the early twentieth that crime was the product of corrupting
century, had its high point in the 1950s and external influences (Walker, 1998) and
1960s and eroded through the 1970s, eventu- reflected a shift from the earlier era when
ally to be replaced by a community problem- crime was viewed as an individual failing or
solving era. simply as the product of sin. In 1779 in the UK
The reform era involved the progressive the Penitentiary Act was passed with the inten-
professionalisation of policing and saw, toward tion of building two establishments to house
the end of the period, the high point of police convicts who otherwise would have been
Peter-3383-Chapter-21.qxd 6/8/2006 7:41 PM Page 371
regard was an article written by Robert than all nations other than China, Iran and
Martinson (1974) exploring existing academic Vietnam (Amnesty International, 2005).
evidence on the impact of correctional interven- Criminal justice policy has been caught up in
tions. Though Martinson’s article was entitled the battle between two competing versions of
‘What Works?’, and his conclusions were mixed – the role of the state. The first emphasises wel-
limited impact was found in something under fare and civil rights, and the reduction of social
half of all interventions – the interpretation gen- inequalities. The other rails against ‘big govern-
erally drawn from his work was that ‘nothing ment’ and seeks to limit state intervention in
works’. Though Martinson quickly distanced citizens’ lives in most areas, with the exception
himself from such a conclusion (Martinson, of criminal justice. In this second model the
1979) penal pessimism was deeply embedded by state has a much diminished role in managing
this time and was anyway being overtaken by and protecting social welfare, something
broader social processes associated with neolib- increasingly left to the market, but has an
eralism that were undermining penal-welfarism. increasingly enhanced role in the management
The last 30 years have seen the emergence of a of social order. Indeed, for commentators such
qualitatively different model of penal trends, in as Charles Murray, rising crime and disorder
which the systems of punishment and control are precisely a product of welfare dependency.
have been reconfigured and which are punitive In this clash of ideologies, the vision espoused
and populist in intent (Sparks, 2003). by the moral conservatives won out despite the
The most obvious signal of this change was fact that such calls for tighter control should
the rapidly increasing use of imprisonment in have clashed with the deregulatory trends set in
many jurisdictions, most spectacularly the train by neoliberalism in the economic sphere
USA, where its ‘imprisonment orgy’ (Walker, (Garland, 2001a). That they did not illustrates
1998) resulted in a fivefold increase in the the extent to which the neo-conservative argu-
prison population between 1970 and 2003 ment successfully portrayed the problems of
(Tonry, 2004). The outcome of the ‘crisis of crime, disorder – and immorality – as specifi-
penal modernism’ (Garland, 1995) has seen cally problems of particular social groups: the
the discretion of judges and other sentencers unemployed, those on welfare, offenders and
considerably limited, indeterminate sentenc- drug-users. As a consequence effective criminal
ing progressively replaced by determinate justice and penal policy ‘came to be viewed as a
alternatives as a result of legislative restriction matter of imposing more controls, increasing
and sentencing guidelines. Indeed, in many disincentives, and, if necessary, segregating the
jurisdictions mandatory forms of sentencing – dangerous sector of the population’ (Garland,
the most infamous being California’s ‘Three 2001a: 102). Over the course of the past 20
strikes and You’re Out’ laws – became increas- years such punitiveness has become the stan-
ingly popular, particularly as a result of the dard political position on crime and order and,
‘war on drugs’. Perhaps the other most obvious indeed, ‘is a stance that no serious politician
indicator that the tenor of the times has can safely disavow’ (Sparks, 2003: 170).
changed markedly, within the US at least, has
been the return of the death penalty. By 1967,
when arguably the modern rehabilitative ideal
A NEW POLITICS OF CRIME?
was at its height, an effective moratorium on
executions was established in America. It lasted
10 years and ended, in Utah, with the execu- The starting gun in penal policy’s equivalent
tion of Gary Gilmore. Though the rate of of the arms race was fired in 1964 by US
executions rose only relatively slowly, and even Republican presidential candidate Barry
at its current (2005) level of 60, remains con- Goldwater. In his acceptance speech he argued,
siderably below the levels reached earlier in the ‘History shows us that nothing prepares the
century, it is now the case that the United way for tyranny more than the failure of public
States uses the death penalty more frequently officials to keep the streets safe from bullies
Peter-3383-Chapter-21.qxd 6/8/2006 7:41 PM Page 373
and marauders. We Republicans seek a phenomena in criminal justice in the last half
government that attends to its fiscal climate, century. The bare facts are startling. There were
encouraging a free and a competitive economy approximately 200,000 people in state and fed-
and enforcing law and order’ (quoted in eral prisons in 1972. Twenty five years later the
Chambliss, 1999). Though his position was number was well over 1.1 million. Similarly, in
controversial, and ultimately unsuccessful, the 1972 there were 130,000 in local jails; by 1997 it
tough on crime message he sought to use has was 567,000. In two and half decades – a
become a staple of national politics in recent period, let us remind ourselves, in which crime
times. The ‘war on poverty’ fought through the was falling – the number of Americans incar-
Great Society programs, which Lyndon cerated increased fivefold. The overall numbers
Johnson also believed to be anti-crime pro- incarcerated exceeded two million in 2002. The
grams, were eventually to be replaced by the sheer scale of this imprisonment orgy now
conservative view, expressed by Richard Nixon, means that commentators routinely refer to
that the ‘solution to the crime problem is this aspect of penal policy in the US as ‘mass
not the quadrupling of funds for any govern- incarceration’ (see Garland, 2001b).
mental war on poverty but more convictions’ Of course, we focus on imprisonment
(quoted in Beckett, 1997: 38). As Beckett because it somehow represents the symbolic
astutely observes, however, at that time the heart of criminal justice policy, the apex of our
President had relatively little leverage over contemporary, hierarchical system of punish-
street crime outside of the confines of ments. However, prison numbers only convey
Washington DC. The initial reaction was to part of what has happened in the US. The reach
take refuge in tough sounding rhetoric and of the criminal justice system is much broader
largely symbolic legislation. Successive admin- than just the incarcerated population. If one
istrations have sought both to increase federal includes in the calculations those also under
aid to, and federal leverage over, local and state criminal justice supervision via probation or
law enforcement in order to mitigate this parole then the numbers are even more star-
problem, not least through successive wars on tling. As Fig. 21.5 illustrates, the numbers sub-
drugs (Zimring and Hawkins, 1992). ject to parole and probation have also increased
Although criminal justice policy took some- and the total under correctional supervision
thing of a backseat under both Presidents Ford reached almost seven million by 2003.
and Carter, it re-emerged with a vengeance Again, however, staggering as they are, the
from 1980 onward under Reagan and not just bare facts of mass imprisonment and mass
his Republican but also his Democrat succes- correctional supervision actually hide some
sors. According to Reagan, ‘the war on crime important features of what has occurred in the
will only be won when an attitude of mind and US in the last 30 years. As Garland (2001b)
a change of heart takes place in America – notes, mass imprisonment has an important
when certain truths take hold again … truths feature beyond the fact of the sheer numbers
like: right and wrong matters; individuals incarcerated. Its second important feature is
are responsible for their actions; retribution that it involves the systematic imprisonment of
should be swift and sure for those who prey on whole groups of the population. In the USA this
the innocent’ (quoted in Beckett, 1997: 47). is young black urban males. As Mauer (1999:
The wars on crime and drugs adopted by the 118–119) observes, ‘A walk through nearly any
Reagan and Bush administrations involved courtroom or prison in the United States reveals
tougher sentencing laws, a profound shift back a sea of black and brown faces at the defendant’s
toward determinate sentencing, increased use table and in the prison yard. Half of all prison
of preventive detention, vastly upgraded inter- inmates are now African-American, and
diction arrangements, and increased street- another 17% are Hispanic’. The lifetime likeli-
level enforcement. hood of imprisonment for African-Americans is
The rising prison population in the United 32% compared with 17% for Hispanics and 6%
States is quite simply one of the most staggering for Whites. Approximately one in eight black
Peter-3383-Chapter-21.qxd 6/8/2006 7:41 PM Page 374
8,000,000
7,000,000
6,000,000
5,000,000
4,000,000
3,000,000
2,000,000
1,000,000
0
80
82
84
86
88
90
92
94
96
98
00
02
19
19
19
19
19
19
19
19
19
19
20
20
Total Probation Prison Parole Jail
males aged 25–29 was in prison at midyear 2003 American politics, making crime a salient issue
compared with 1 in 27 Hispanic males and 1 in in national politics. ‘Conservatives blamed the
63 white males of the same age (Sentencing increase in crime on liberal social policies,
Project, 2004). arguing that the courts had “handcuffed” the
What makes for such unprecedented growth police, that bail reform and rehabilitation
in the prison population? In part the war programs turned predatory criminals loose
on crime and, more particularly, the war on on society, and that the moratorium on
drugs have had a dramatic on the supply side – executions removed the deterrent effect of the
providing ‘an almost limitless supply of death penalty’ (Walker, 1998: 201). Similarly,
arrestable and imprisonable offenders’ (Simon, it was not until the mid-1970s in the UK that
2001). Second, there are features of the ‘macho ‘law and order’ became an important feature
penal economy’ itself (Downes, 2001), not of national, electoral politics (Downes and
least the expanding private corrections sector, Morgan, 1997).
which tend toward expansionism (Christie, In the criminal justice arena, politicians’
1993). The third, and possibly most important, concern with how they are likely to be per-
factor is the general transformation of politi- ceived has had a profound effect on policy-
cal culture that has occurred in America and making in recent times. Crucially, as numerous
elsewhere. commentators have noted, by the 1990s the old
Crime is now a staple of political discourse divisions between ‘conservative’ and ‘liberal’
and of electoral politics. Whilst this may not political positions on crime had disappeared,
feel surprising it is, in fact, a relatively new and had been replaced by what appeared to be
political phenomenon. Until the early 1960s in a straightforward ‘tough on crime’ message.
the USA and the early 1970s in the UK, for The past two decades have seen a progressively
example, criminal justice policy barely fea- intensifying battle by the major political par-
tured in major elections and certainly was far ties to be seen as the party of law and order.
from the ‘wedge issue’ it has often been since. Initially in the USA, and subsequently in the
As we have seen, the 1964 presidential election UK and other jurisdictions, a ‘tough on crime’
campaign signalled an important shift in stance has come to be associated with electoral
Peter-3383-Chapter-21.qxd 6/8/2006 7:41 PM Page 375
success and its opposite, being ‘soft on crime’, Clinton administration (Newburn and Jones,
with electoral failure. 2005) and the promise of similarly punitive
Crucial in this regard was Michael Dukakis’ policies. The amalgam has never been more
defeat in the 1988 presidential election in the successfully captured than in Tony Blair’s 1993
US. In the aftermath of the defeat the soundbite, ‘tough on crime and tough on the
Democratic Party sought to reinvent and repo- causes of crime’. Here, in only 10 words, the
sition itself on a number of core public policy relegation of traditional liberal concerns with
issues, not least crime. The ‘New Democrats’ poverty and social inequality as the genera-
sought to break free from what they perceived tors of criminality was successfully conveyed
as the failed ‘liberalism’ of the Party, regaining whilst also using the new symbolic keywords
the ‘mainstream’ voters that had been alienated (Fairclough, 2000) ‘tough’ and ‘crime’ twice
over the course of the 1980s. In the aftermath each. Though the rise is by no means as spec-
of the Dukakis defeat, Bill Clinton, then an tacular as in America, nevertheless the conse-
active member of the reformist Democratic quence of the emergence in the early to
Leadership Council observed that: ‘No matter mid-1990s in Britain of a punitive bipartisan
how popular your programs may be, you consensus on crime is a rapid and substantial
must be considered in the mainstream on the increase in the prison population. In 1994 the
shared values of the American people, the abil- prison population in England and Wales stood
ity to defend the nation and the strength to at a little over 44,000. A decade later it had
enforce its laws’ (quoted in Baer, 2000: 126). reached 75,000, with the Labour government
Subsequently, he became a ‘tough on crime’ looking to institute measures to cap the popu-
Democratic president, supportive of the death lation at 80,000. Again, the bulk of this prison
penalty, introducing federal ‘three strikes’ sen- expansion took place during a period of
tencing and, though supportive of drug treat- declining crime. With the main political par-
ment programs and community policing, by ties seemingly wedded to tough on crime
1994 was overseeing ‘a repressive criminal jus- rhetoric, on the surface there seems little
tice climate rivalling that of any time during prospect of breaking out of the ‘new iron cage’
the preceding twenty years’ (Mauer, 1999: 71). (Garland, 2001a) of punitive justice.
The parallels between the politics of crimi- The ‘culture of control’ in America and
nal justice in America and Britain at this time Britain that Garland so vividly describes is by
are strong. The lengthy political dominance of no means confined to those nations. Elements
Conservatism during the 1980s in the UK led of this new configuration are visible elsewhere.
to vociferous debates within the British Labour Prison populations are rising sharply in many
Party over the possible sources of electoral suc- jurisdictions, including hitherto quintessen-
cess in what were clearly changed times. tially liberal societies such as the Netherlands
Paralleling the remodelling of the Democratic and punitive rhetoric, often originating in the
Party in the aftermath of the 1988 presidential US, has spread through much of northern
election, the Labour Party sought to dump its and southern Europe (Wacquant, 1999) and
various ‘hostages to fortune’ (Downes and beyond. Just as there are global pressures trans-
Morgan, 1997) not least of which was its previ- forming and seemingly homogenising many
ously more liberal policies on crime control. aspects of national political cultures, the struc-
‘New Labour’ in the UK embraced so-called tural changes accompanying the shift to late
‘third way’ politics. In the criminal justice modernity appear to be creating the condi-
arena this meant attempting to modify the old- tions across the developed economies in which
fashioned liberal penal-welfarism that the party a newly punitive culture of control can flour-
had largely clung to throughout the 1980s and ish. These developments are so pervasive that
into the 1990s by adding into the mix what was they frequently appear irresistible. And, yet, it
by now considered the sine qua non of success- would be a mistake not to acknowledge the
ful electoral politics: a healthy dose of punitive continued existence of considerable variations
rhetoric – much of which was drawn from the in the localised cultures of control that are
Peter-3383-Chapter-21.qxd 6/8/2006 7:41 PM Page 376
visible internationally, for, not only are they policy in Scandinavia? Bondeson suggests that
important illustrations of some of the limits of there is some evidence, though far from
current trends, but potentially also help identify unequivocal, that attitudes to punishment are
the sources of resistance to the new iron cage. less punitive in Scandinavia and that fear of
crime is also lower. Moreover, it seems possible
the media are also more constrained, and
thoughtful, in their treatment of crime and
INTERNATIONAL VARIATIONS
justice issues. A number of broader socio-
cultural explanations also seem plausible. On a
As an increasing body of scholars have noted general level it appears to be the case that there
(inter alia, Tonry, 2004; Newburn and Sparks, remains greater support for a broadly welfarist
2004; Melossi, 2004) that beneath what appears approach to public policy, including penal
a general trend in the direction of increased policy, in Scandinavia than in some other parts
punitiveness and exclusionary practices in of Europe and certainly than in the US. The
crime control, there remain some considerable Nordic countries have more homogenous pop-
and apparently quite deeply embedded national ulations than, say, the UK and the US, and
differences in criminal justice and penal poli- higher levels of functional democratisation,
cies. Indeed, there exist considerable variations both of which, arguably, have served to protect
within some jurisdictions. Thus, the USA is against the emergence of those forms of pop-
made up of many criminal justice systems and ulist punitiveness that are increasingly visible
that there exist considerable variations between elsewhere.
these. As one example, taking incarceration The number of people in prison in Italy has
rates in the 30 years 1970–2000, Hinds (2005) been rising slowly, but the incarceration rate
has shown that custody rates were consistently remains close to the European average of 100
higher in this period in the Southern States and per 100,000 population (Solivetti, 2004). Italy
lower in the Northeast and Midwest compared does differ somewhat from much of the rest of
with other regions. In 2000 the custody rate in Europe, however, in the way it reaches its aver-
the Southern States, at 8.34, was twice that of age. The prison population in Italy has both a
the Northeastern States (at 4.09). Similarly, higher proportion of very long prison sen-
by no means all the states in the Union still tences compared with, say, the Scandinavian
have the death penalty, let alone still use it. countries, and also a significant proportion of
Mandatory sentencing varies hugely by state those in prison are detained awaiting final
(Jones and Newburn, forthcoming) as does the sentence (Nelken, 2005). In explaining the rel-
nature of policing. Whilst identifying general atively lenient criminal justice responses found
trends continues to be important, it is equally in Italy, particularly in relation to dealing with
vital not to lose sight of the nature and sources youthful criminality, Nelken (2005) points to a
of continuing variation. number of important socio-cultural factors,
Thus, in Scandinavia for example, despite including the relative absence of ghetto hous-
generally rising crime rates over the past 40–50 ing estates, the strength and surveillance
years, prison populations have remained rela- capacities of the family, and the lack of a
tively stable (Denmark, Norway and Sweden) youthful drinking culture (at least when com-
or reduced from a previously enormously high pared with Northern Europe).
level to something akin to other Scandinavian One of the most interesting and intriguing
countries (Finland). It is important to note, contrasts with the USA is Canada. Mulroney’s
however, that there appear to have been some Progressive Conservative government of the
increases in incarceration rates in recent years late 1980s largely continued the approaches
(von Hofer, 2004), possibly related to extended established by its Liberal predecessor (Hatt
sentence lengths, particularly for drug-related et al., 1992) and during the 1990s the new Liberal
crimes (Bondeson, 2005). How might one federal government set about establishing a
explain the existence of generally milder penal distinct ‘Canadian way’ for dealing with social
Peter-3383-Chapter-21.qxd 6/8/2006 7:41 PM Page 377
problems, including crime (Meyer and O’Malley, and cultural frames through which crime is
2005). This distinctive way was explicitly understood. That the whole area of penal
designed to differ from the example being set policy has become intensely and minutely
by the nation’s nearest neighbour: [W]e have politicised is one of the central changes of
two choices – go the way of the US and build recent times. Crime and criminal justice is now
more prisons, or develop meaningful, lasting accepted as being a major political issue – if
alternatives to incarceration for low-risk not yet a major area of scholarly study among
offenders who can be better managed in the political scientists. Not only do we expect
community. (Canada, 2001, quoted in Meyer politicians to spend much of their time talking
and O’Malley, 2005). In sum, such an approach about crime and criminal justice, but we
to criminal justice policy has led to a federal expect them to disagree. It was not always thus.
prison system with a comprehensive pro- As many commentators have noted (Beckett,
gramme of health, educational and training 1997; Chambliss, 1999; Downes and Morgan,
programmes for inmates, incarceration rates 1997) for much of the twentieth century there
that are both low and declining, together with existed something close to a bipartisan consen-
the removal of the death penalty. That said, as sus about how best to respond to crime, both
Moore and Hannah-Moffatt (2005: 97) con- in the US (certainly at a Federal level) and the
clude in their recent review, ‘[A]ppearances UK. Broadly speaking, the consensus view was
can be deceiving. The liberalism of Canadian that a set of strategies aimed simultaneously at
punishment is a veil underneath which remains punishing and reforming – characterised by
an extremely punitive system’ albeit of a dis- Garland as ‘penal-welfarism’ – was the most
tinctive form in which therapeutic initiatives appropriate and efficacious means by which
are central to the system of punishment. the modern state could maintain order and
The varied cultures of control in contempo- reduce crime.
rary advanced economies clearly have many Though there remain jurisdictions in which
commonalities. Moreover, such shared charac- an attachment to penal-welfarism persists, the
teristics are undoubtedly the product of broad general trend appears to be away from such
social, cultural, political and economic pres- approaches and toward a set of highly puni-
sures. However, it is also clear that particular tive and exclusionary strategies. Indeed, it is
socioeconomic, cultural and political contexts arguably the case that a new political consen-
frame and shape local cultures of control in sus has emerged around precisely such ‘tough
quite different ways. As Lacey (2003: 86) on crime’ rhetoric and practice. At the heart
observes, ‘it is crucial for us to recognise that of the emergent punitive strategies lies the
the salience and politicisation of criminal language of risk. The shift away from concerns
justice vary from country to country’ and that with causes and toward the management and
our understanding of this variation remains very administration of crime problems has led
far from developed. Understanding similarities some commentators to argue that we are
and differences in the pattern of contemporary witnessing the emergence of a distinctive form
systems of crime control is arguably, therefore, of criminal justice policy that they characterise
one of the most pressing tasks facing us. as a ‘new penology’ (Feeley and Simon, 1992).
Such strategies are risk-oriented and based on
actuarial calculations rather than individual
diagnoses, and future preventive-orientation
CONCLUSION
rather than the imposition of sanctions ex
poste ante. Such strategies are clearly compati-
I opened this chapter with a quote from David ble with the broader neo-liberal political pro-
Garland’s magisterial study of punishment and ject, which seeks to roll-back the welfare state,
modern society in which he observes that both resting on a view of those caught in poverty,
criminal justice and penal policy have less to offending and drug dependence as straightfor-
do with trends in crime than with the political wardly feckless and morally culpable. The
Peter-3383-Chapter-21.qxd 6/8/2006 7:41 PM Page 378
result in the crime control arena is policies of crime and drug use is also shown to be
dominated by rational choice models, the much more complex than the ‘democracy-
racheting up of penalties and hazards atten- at-work’ thesis would allow. This, of course, is
dant on behaviour deemed to fall short of not to deny any link between public opinion
required standards, together with vast and political initiative in this area – it is highly
increases in the technologies of surveillance likely that they are mutually reinforcing in
available to monitor populations of risk. many respects – merely to reiterate the point
In discussing the politics of crime control, that the ‘punitive turn’ in contemporary crim-
Simon (1997) has described the strategy of inal justice policy can neither be explained by
‘governing through crime’. Simon’s central reference to trends in crime nor as a simple
argument is that advanced industrial societies extension of public concerns and fears about
are experiencing a crisis of governance (rather crime. Rather, as in so much else in public
than a crisis of crime) and that the response to policy, it is the broader restructuring of social
this has been to prioritise crime and punish- relations and political cultures associated with
ment as the preferred contexts for governance. late modernity that is central to the explana-
Though the argument may be overstated in tion of contemporary developments. The par-
some respects, he is undoubtedly right in argu- ticular irony here is that, as the ‘limits of the
ing that crime currently casts a ‘disproportionate sovereign state’ become more visible, so part of
shadow over what we primarily identify with the response appears to have been to engage ‘in
governance, i.e. politicians and the electoral a more expressive and more intensive mode of
process of democracy’ (1997: 174). This has punishment that purports to convey public
been true for well over a decade in the USA, for sentiment and the full force of state authority’
a similar period in the UK, most evidently (Garland, 2000: 349). The result is a set of
under the New Labour administrations of the arrangements for dealing with criminality that
past eight years, and appears to be a growing pay precious little regard to root causes, and
force in other jurisdictions besides. have little interest in punishment beyond its
One possible explanation for the rise up the utility as a means of reducing threats and man-
political agenda of crime policy generally, and aging risks. Churchill’s observation that ‘The
punitive policies particularly, is what Beckett mood and temper of the public in regard to
(1997) calls the ‘democracy-at-work’ thesis. the treatment of crime and criminals is one of
This thesis has it that the significant increases the most unfailing tests of the civilisation of
in crime that have occurred have led to any country’ (quoted in Radzinowicz and
increases in public fear of crime. Heightened Hood, 1990: 774) is often cited – particularly
fears put crime on the political agenda, in turn now that the mood and temper so often seems
this leads to demands for harsher punishments so illiberal. Less often quoted are the sentences
for offenders and eventually to a significant that followed this most famous of observa-
growth in the number of people incarcerated. tions. I conclude with them as it is they per-
She quotes James Q. Wilson (1975), arguing haps that best capture the general philosophy
along these lines that ‘public opinion was well that underpinned criminal justice policy until
ahead of political opinion in calling attention as recently as 30 years ago and so clearly
to the rising problem of crime’. In fact, using demonstrate how far, in what direction, we
public opinion data during the US war on have travelled since:
crime period (1964–1974) and the war on A calm and dispassionate recognition of the rights of
drugs era, (1985–1992), she is able to show the accused against the State, and even of convicted
that, in practice, it was political initiatives in criminals against the State, a constant heart-searching
relation to crime and drugs that was a signifi- by all charged with the duty of punishment, a desire and
eagerness to rehabilitate in the world of industry all
cant factor in shaping public concerns rather
those who have paid their dues in the hard coinage of
than the other way around. Moreover, the punishment, tireless efforts towards the discovery of
assumption that there is a close link between curative and regenerating processes, and an unfaltering
levels of public concern and the reported rate faith that there is a treasure, if you can only find it, in the
Peter-3383-Chapter-21.qxd 6/8/2006 7:41 PM Page 379
heart of every man – these are the symbols which in the Farrington, D. P. and Langan, P. (2004) England and
treatment of crime and criminals mark and measure the Wales, in Farrington, D. P., Langan, P. and Tonry, M.
stored-up strength of a nation, and are the sign and (eds), Cross-National Studies in Crime and Justice,
proof of the living virtue in it. Washington DC: U.S. Department of Justice.
Feeley, M. and Simon, J. (1992) The new penology:
Notes on the emerging strategy of corrections
NOTE and its implications, Criminology, 30, 4, 449–474.
Friedman, L. (1993) Crime and Punishment in
American History, New York: Basic Books.
1. Professor of Criminology and Social Policy and Garland, D. (1985) Punishment and Welfare: A
Director, Mannheim Centre for Criminology, London history of penal strategies, Aldershot: Gower.
School of Economics. Contact details: Dept of Social
Garland, D. (1990) Punishment and Modern Society,
Policy, LSE, Houghton St, London WC2A 2AE, UK; t.new-
burn@lse.ac.uk
Oxford: Clarendon Press.
Garland, D. (1995) Penal modernism and postmod-
ernism, in Cohen, S. and Blomberg, T. G. (eds),
REFERENCES Punishment and Social Control: Essays in Honor of
Sheldon L. Messinger, New York: Aldine de Gruyter
Garland, D. (1996) The limits of the sovereign state,
Amnesty International (2005) The Death Penalty British Journal of Criminology, 36, 4, 445–471.
Worldwide: Developments in 2004, http://web. Garland, D. (2000) ‘The culture of high crime
amnesty.org/library/Index/ENGACT500012005. societies: Some precondition of recent ‘law and
Baer, K. (2000) Reinventing Democrats: The politics order’ policies’, British Journal of Criminology,
of liberalism from Reagan to Clinton, Kansas: 40, 347–375.
University Press of Kansas. Garland, D. (2001a) The Culture of Control: Crime
Beckett, K. (1997) Making Crime Pay: Law and Order and social order in contemporary society, Oxford:
in Contemporary American Politics, New York: Oxford University Press.
Oxford University Press. Garland, D. (2001b) ‘The meaning of mass impris-
Bondeson, U. V. (2005) Levels of punitiveness in onment’, in Garland, D. (ed.), Special Issue on
Scandinavia: description and explanations, in Mass Imprisonment in the USA, Punishment and
Pratt, J. Brown, D. Brown, M. Hallsworth S. Society, vol. 3, no. 1.
and Morrison W. (eds), The New Punitiveness: Gurr, T.R. (1989) Violence in America, Thousand
Trends, theories, perspectives, Cullompton: Willan Oaks, CA: Sage.
Publishing. Hatt, K., Caputo, T. and Perry, B. (1992) Criminal
Chambliss, W. J. (1999) Power, Politics and Crime, justice policy under Mulroney, 1984–90: Neo-
Boulder, Co: Westview Press. conservatism, eh?, Canadian Public Policy,
Christie, N. (1993) Crime Control as Industry, 18: 245–260.
London: Routledge. Hinds, L. (2005) Crime control in Western coun-
Dodd, T., Nicholas, S., Povey, D. and Walker, A. tries, 1970–2000, in Pratt, J. Brown, D. Brown, M.
(2004) Crime in England and Wales 2003/2004, Hallsworth S. and Morrison W. (eds), The New
London: Home Office. Punitiveness: Trends, theories, perspectives,
Downes, D. (2001) The macho penal economy, in Cullompton: Willan Publishing.
Garland, D. (ed.), Special Issue on Mass Hofer, H. von (2004) ‘Crime and the reaction to
Imprisonment in the USA, Punishment and crime in Scandinavia’, Journal of Scandinavian
Society, vol. 3, no. 1. Studies in Criminology and Crime Prevention, 5,
Downes, D. and Morgan, R. (1997) Dumping the 2, 148–166.
‘Hostages to Fortune’? The politics of law and Home Office (1979) Committee of Inquiry into the
order in post-war Britain, in Maguire, M., United Kingdom Prison Services (The May
Morgan, R. and Reiner, R. (eds), The Oxford Inquiry), Cmnd 7673, London: Home Office.
Handbook of Criminology, Second Edition, Ignatieff, M. (1978) A Just Measure of Pain: The pen-
Oxford: Clarendon Press. itentiary in the industrial revolution 1750–1850,
Emsley, C. (1991) The English Police: A Political and Harmondsworth: Penguin.
Social History, Hemel Hempstead: Harvester Jones, T. and Newburn, T. (forthcoming) Three
Wheatsheaf. Strikes and You’re Out: Exploring Symbol and
Fairclough, N. (2000) New Labour, New Language?, Substance in American and British Crime
London: Routledge. Control Politics. British Journal of Criminology.
Peter-3383-Chapter-21.qxd 6/8/2006 7:41 PM Page 380
Kelling, G. L. and Moore, M. H. (1988) The evolving Newburn, T. and Sparks, R. (eds) (2004) Criminal
strategy of policing, reprinted in Newburn, T. Justice and Political Cultures: National and
(ed.), Policing: Key Readings, Cullompton: Willan International Dimensions of Crime Control,
Publishing. Cullompton: Willan.
Lacey, N. (2003) Principles, politics and criminal Radzinowicz, L. (1948) A History of English Criminal
justice, in Zedner, L. and Ashworth, A. (eds), The Law and its Administration, Vol. 1, The Movement
Criminological Foundations of Penal Policy: Essays for Reform, London: Steven and Stevens.
in Honour of Roger Hood, Oxford: Clarendon Press. Radzinowicz, L. and Hood, R. (1990) The Emergence
Martinson, R. (1974) What Works?, The Public of Penal Policy in Victorian and Edwardian
Interest, March, 22–54. England, Oxford: Clarendon Press.
Martinson, R. (1979) New Findings, New Views: Sentencing Project (2004) Facts about prisons and pris-
A note of caution regarding sentencing reform, oners, Washington DC: The Sentencing Project.
Hofstra Law Review, 7, 243–258. Sharpe, J.A. (1990) Judicial Punishment in England,
Mauer, M. (1999) The Race to Incarcerate, New York: London: Faber and Faber.
The New Press. Simon, J. (1997) ‘Governing through crime’, in L.M.
Melossi, D. (2004) The Cultural Embeddedness of Friedman and G. Fisher (eds), The Crime
Social Control: Reflections on the Comparison of Conundrum: Essays on Criminal Justice,
Italian and North-American Cultures Concerning Boulder, CO: Westview Press.
Punishment, in Newburn, T. and Sparks, R. (eds), Simon, J. (2001) Fear and loathing in late moder-
Criminal Justice and Political Cultures, Cullompton: nity, in Garland, D. (ed.), Special Issue on Mass
Willan. Imprisonment in the USA, Punishment and
Meyer, J. and O’Malley, P. (2005) Missing the puni- Society, vol. 3, no. 1.
tive turn? Canadian criminal justice, ‘balance’ and Solivetti, L. (2004) Italian Prison Statistics,
penal modernism, in Pratt J., Brown D., Brown M., Department of Statistics, La Sapienza, Rome, 21.
Hallsworth S. and Morrison W. (eds), The New Sparks, R. (2003) States of insecurity: punishment,
Punitiveness: Trends, theories, perspectives, populism and contemporary political culture, in
Cullompton: Willan Publishing. McConville, S. (ed.), The Use of Punishment,
Miller, W. (1975) Cops and Bobbies 1830–1870, Cullompton: Willan Publishing.
Journal of Social History, vol. IX, 73–88. Tonry, M. (2004) Thinking About Crime: Sense and
Moore, D. and Hannah-Moffat, K. (2005) The liberal sensibility in American penal culture, New York:
veil: revisiting Canadian penality, in Pratt, J., Oxford University Press.
Brown, D., Brown M., Hallsworth S. and Morrison Wacquant, L. (1999) How Penal Common Sense
W. (eds), The New Punitiveness: Trends, theories, per- Comes to Europeans: Notes on the transatlantic
spectives, Cullompton: Willan Publishing. diffusion of the neoliberal doxa, European
Nelken, D. (2005) When is a society non-punitive? Societies, 1(3), 319–352.
The Italian case, in Pratt, J. Brown, D. Brown, M., Walker, S. (1998) Popular Justice: A history of
Hallsworth S. and Morrison W. (eds), The New American criminal justice, New York: Oxford
Punitiveness: Trends, theories, perspectives, University Press.
Cullompton: Willan Publishing. Weber, M. (1978) Economy and Society, Roth, G. and
Nieuwbeerta, P. (2002) Introduction, in Wittich, C. (eds), 2 vols, Berkeley, CA: University
Nieuwbeerta, P. (ed.), Crime Victimisation in of California Press.
Comparative Perspective: Results from the ICVS, Wilson, J. Q. (1975) Thinking About Crime,
1989–2000, Den Haag: Boom Juridische uitgevers. New York: Basic Books.
Newburn, T. and Jones, T. (2005) Symbolic politics Zimring, F. and Hawkins, G. (1992) The Search for a
and penal populism: the long shadow of Willie Rational Drug Control Policy, Camb, MA:
Horton, Crime, Media, Culture, 1(1), 72–87. Cambridge University Press.
Peter-3383-Chapter-22.qxd 6/8/2006 7:42 PM Page 381
22
Privatisation by Divestment
IAN THYNNE
In all cases, there was a distinct move away from SOME FACTS AND FIGURES
politics and administration to money and man-
agement. A central concern was to ensure a
Public enterprises around the world have been
results-oriented response to market forces and
constituted in a variety of forms (Thynne and
demands, while often retaining varying degrees
Wettenhall 2001). They have ranged from
of state involvement through joint ownership,
departments subject to direct government
regulatory and funding arrangements, and
control, through to statutory bodies, compa-
other mixes of public and private activity.
nies, trusts and so on, with varying degrees of
Privatising action in its different forms con-
legal and operational autonomy. They have
tinues to be taken throughout the world, but in
usually had a capacity to enter into contracts,
many instances it is now being complemented or
to buy and sell property, and to sue and be
tempered by important initiatives involving civil
sued in their own corporate names. But only
society organisations as well as market firms.
those in company form have been open to
Competition often gives way to collaboration,
immediate divestment. All of the others have
with synergies being sought through partner-
needed first to be reconstituted as companies,
ships, networks and other alliances which adopt
which are the only organisations capable of
a more cooperative, integrative approach to the
having a divisible share capital that can be
management of public affairs. The flow of
offered for sale and subsequently traded.
market forces and related exchange relationships
Beyond the form of organisation, govern-
based on contracts are being contained or soft-
ments have had considerable choice concern-
ened by associational arrangements which can
ing the objectives, targets and methods of
enhance the collaborative bases of collective
divestment, as well as the means and extent
endeavours in pursuit of public goals (Evans
of any ongoing control over the companies
1997; Flynn 2002, ch 7; Huxham 2000; Lowndes
involved. Not all choices and related decisions
and Skelcher 1998; Wettenhall 2003).
have been rational in the usual sense of the
The focus hereafter is on privatisation by
word, either at the time or with hindsight. But
divestment, which has enabled states to with-
most have been strategic, with a focus on aims
draw from the ownership of selected public/
and purposes of present and future signifi-
state/government enterprises (“public enter-
cance, including the development of political,
prises”) and associated assets, without
economic and social capital in various guises.
necessarily giving up all interest in and respon-
The choices made can be addressed here quite
sibility for the work involved, be it in the
briefly, with reference to objectives, targets, etc.
domestic or international arena (OECD 2003;
as identified in a few surveys and other perti-
Lane 2002; Thynne 1995; Wettenhall 1993;
nent discussions.
World Bank 1995). Often, the withdrawal has
only been partial, resulting in joint or mixed
public and private ownership. Often, too, the Objectives
ownership, production and provision roles of a
state have been replaced by those of regulation A questionnaire-based analysis of the divest-
and facilitation, such that state involvement in ment experience of numerous countries in
certain sectors has become less direct and less Europe, Africa, Asia, the Pacific and Latin
active, but remains significant. Meanwhile, America in the mid-1990s considered, among
many enterprises and assets have not actually other things, the rating of six likely objectives
become the targets of divestment, and some of divestment, namely to improve business
new ones have been established or acquired efficiency, to reduce state activity, to reduce
which could well become future targets. So, at state debt and/or cut budget deficits, to obtain
one and the same time, states have often been funds, to enlarge the capacity of a stock
both “rolled back” and “rolled over”, while also exchange, and to build a share-owning democ-
being “rolled on” (Ariff, Asher and Thynne racy (Gonzalo, Pina and Torres 2003). While
1995; Steel and Heald 1984). the findings are of a decade ago, they apply to
Peter-3383-Chapter-22.qxd 6/8/2006 7:42 PM Page 383
Table 22.1 Objectives of divestment (with percentage ratings as one of the top three)
Enlarge Build
Improve Reduce Reduce stock share-
business state state debt/ Obtain exchange owning
efficiency activity cut deficit funds capacity democracy
19 developed 79% 47% 63% 37% 16% 21%
countries
a time when there was considerable divestment share-owning democracy has had a particular
activity in many countries, with keen attention appeal or ring to it, its likely benefits have been
being given to aims, processes and achieve- less tangible than those attainable through the
ments. Accordingly, they are a valuable basis pursuit of more concrete objectives and thus
on which some relevant lessons and conclu- not necessarily a matter of central concern. It
sions can be drawn. has remained a good idea without frequently
The findings are summarised in Table 22.1 being reflected in specific divestment decisions.
They are clear enough from the table, but at Overall, the objectives of divestment have
least three are worth highlighting. The first is inevitably varied from one sector to another.
that more than 60 percent of the countries This has been the case both within countries
beyond Eastern Europe rated the reduction of and internationally, as affected by the particu-
state debt and budget deficits as one of their top lar production or provision roles being per-
three divestment objectives, whereas none of formed in each sector.
the countries in Eastern Europe did so. The
second is that 100 percent of the latter countries
rated the improvement of business efficiency as Targets
one of their top three objectives, and 90 percent
of them also rated the reduction of state activity Within and across sectors, the actual targets of
in the top three; while the percentages for divestment have also varied significantly.
the other countries were lower in both cases. The Numerous enterprises have been subject to
third is that only 40 percent or less of all of the some degree of divestment as a means of meet-
countries rated the building of a share-owning ing one or more of the objectives to which
democracy as one of their top three objectives. governments have been committed.
None of these findings is at all surprising. A survey of OECD countries covering
Over the last two decades, many governments the 11-year period 1991–2001 considered
have had to cope with substantial public debts five sectors (plus a catch all “other”) in terms of
and budget deficits, and the former communist the annual revenue generated by the divest-
states have needed to recast their public man- ment of companies and related assets in
agement systems and affairs substantially in each sector (Mahboobi 2002). The five were
response to the demands of a market economy. financial intermediation, manufacturing,
Also, while for some governments the idea of a public utilities (including electricity, gas and
Peter-3383-Chapter-22.qxd 6/8/2006 7:42 PM Page 384
Table 22.2 Methods of divestment (with number ratings as most commonly used)
Stock Direct
market sale MBO Gift Auction Combination Others
17 developed 4 5 2 1 1 3 1
countries
7 Eastern European 0 1 2 1 2 1 0
countries
15 developing 6 6 3 1 4 2 0
countries
The developed countries: Australia, Austria, Canada, Denmark, Finland, France, Germany, Greece, Iceland, Ireland,
Japan, Netherlands, New Zealand, Norway, Portugal, Sweden, United Kingdom
The Eastern European countries: Albania, Bielorusia, Czech Republic, Estonia, Hungary, Lithuania, Poland
The developing countries: Algeria, Bahrain, Barbados, Brazil, Chile, Costa Rica, Indonesia, Israel, Korea, Maldives,
Mexico, Morocco, Nigeria, St Kitts & Nevis, Turkey
sectors have been ready and willing to invest in and New Zealand, have been able to preserve
public enterprises. In countries of Eastern control over such matters as the appointment
Europe, stock markets have usually developed of company boards and the making of changes
after various divestment initiatives have been to company constitutions, including any pro-
taken rather than being established in advance. visions limiting foreign ownership (Graham
Also, private firms in other countries have and Prosser 1988; Taggart 1990). The latter
often been a little slow in buying companies in type of share has been criticised in the United
that part of the world, at least partly because Kingdom as giving the government an oppor-
such companies understandably have been tunity to determine relevant policy “by stealth”
quite difficult to value in terms of their likely (Graham and Prosser 1988). Its use, both in
productivity and profitability. the United Kingdom and in Spain, has also
The earlier-mentioned survey of OECD been condemned by the European Court of
countries found that, over the 12-year period Justice as unjustifiably restricting the flow of
1990–2001, public flotations were clearly the capital between or among countries in the
most favoured method of divestment (Mahboobi European Union (BBC News 2003).
2002). This was followed by the direct sale of The external means of government control
companies as trade sales, and then by other over company activities has comprised the estab-
methods such as management or employee buy- lishment of appropriate regulatory arrange-
outs and the sale of non-organisational assets. ments. In the United Kingdom, for example,
Over the period, more than US$670 billion various sector-based regulatory bodies have
was generated by the divestment of companies been created, such as the Offices of Electricity,
and related assets in the OECD countries. The Gas, Telecommunications and Water Services
high peak was 1997–99, during which some (Maloney and Richardson 1992; Vass 1992;
US$290 billion was raised. The big sales were Veljanovski 1987). These specialist bodies stand
in the telecommunications, financial and in contrast to the broadly-focused, generalist
public utility sectors. Commerce Commission which serves to regu-
late divested and other companies in New
Zealand (Taggart 1990).
Ongoing Control
The various regulatory arrangements
have also included anti-trust schemes which
The objectives pursued, targets selected, meth-
have long existed in the United States and have
ods chosen and proceeds expected as integral
more recently become a feature of initiatives in
components of divestment have all had a bear-
the European Union (Lane 2002). In addition,
ing on the means and extent, if any, by and to
a range of specialist, international regulatory
which governments have continued to exercise
bodies now exist in non-government form,
control over divested companies. Broadly,
with power which is often more extensive and
there have been two means, one internal and
more effective in protecting public interests
the other external, with both taking one or
than the power of government bodies (Scott
other of two forms and being open to varying
2003).
degrees of involvement (Thynne 1994, 1995).
The internal means of control have been
based on two forms of ownership. One has
SOME THEORIES AND MODELS
seen governments in many countries often
retaining a sufficiently significant ordinary
tradeable shareholding in divested companies The facts, figures and associated material just
to enable them to continue to control the addressed have been underpinned by various
whole range of company affairs (Thynne 1995, theories and models on why and how choices
1998). The other has been in the form of a spe- are or could be made within and beyond a
cial non-tradeable “golden share” by which state. Individually or in clusters, the latter have
governments, notably in the United Kingdom served to stimulate, justify and/or describe
Peter-3383-Chapter-22.qxd 6/8/2006 7:42 PM Page 386
what governments have done with regard to degree of self-interest in their support for
divestment and other reforms involving signif- divestment. They have seen it as a possible
icant structural and operational change. vote-winning initiative by generating resources
which could be used for electorally popular
purposes, including the off-setting of timely
The Why-type tax-cuts. This has often been buttressed by
a practice, if not a policy, of governments’
Why-type theories and models have been under-pricing the shares in companies to be
grounds-based, with much of the emphasis divested as a means of ensuring widespread
being on actual or assumed weaknesses of the investor interest and commitment. In the
state and strengths of the market. They have process, members of the public have essentially
combined state failure ideas with thoughts been hoodwinked into becoming individual
about likely market successes, along with some owners of enterprises of which they were
reflections on political manoeuvring. They already owners in their collective capacity as
have focused on the relationships between citizens and taxpayers. The result is that divest-
form and function and also function and space ment in such circumstances has reaffirmed the
in the state and market. Those of public choice, publicness of the enterprises involved, but in
agency and transaction costs have been partic- individual rather than collective terms.
ularly prominent (Boston, Martin, Pallot and The rational self-interest basis of public
Walsh 1996, ch 2; Hood 1994; Self 1993, 2000; choice theory has been given a specific focus by
Williamson 1985). agency theory. The latter concentrates directly
Public choice theory has cast individuals as on the relationship between principals and
rational utility maximisers who will use the agents and on means by which principals can
organisations in which they work as avenues to seek to have agents act in accordance with their
advance their self-interests. This idea has been expectations and objectives rather than the
translated to mean that in state organisations agents’ own interests. This too has been related
officials will pursue goals and objectives which to public enterprises in terms of the difficulties
are often at odds with what the organisations faced by politicians as principals in exercising
are expected to be achieving in terms of gov- meaningful, goal-oriented control over their
ernment policy. Thus, it has been thought that, agents as the top management in such enter-
where such organisations are public enter- prises. Again, supported by the same line of
prises which need not be state-owned, then thought as in public choice theory, divestment
divestment will serve to bring their top man- has been seen as a necessary step towards
agement to heel by subjecting them to the dis- containing the assumed self-interested propensi-
ciplines of the market and the active scrutiny ties of top managers, with any ongoing govern-
of shareholders. Thereafter, if need be, politi- ment policy and public interest concerns
cians can pursue relevant policy goals and pro- similarly being responded to by appropriate
tect the public interest by using external regulatory means.
regulatory means rather than by internal Ironically, this solution to the principal-
means based on ownership. In essence, for the agent problem was actually an arena in which
performance of commercial roles, companies the problem was first identified, namely in the
in private ownership and subject to market relationship between the owners as principals
forces, but possibly operating within a state and the managers as agents in large private
regulatory framework, have been the favoured firms, especially publicly listed ones with a
form of organisation over any continuation of wide shareholder base. Thus, in assuming the
direct state involvement through various forms problem was now one confronting state organ-
of public enterprise. isations and particularly public enterprises, the
Notwithstanding the perceived need to limit advocates of divestment have simply or conve-
or counter managerial self-interest, politicians niently overlooked its origins and proposed a
themselves have often been motivated by a solution that was already known to be flawed.
Peter-3383-Chapter-22.qxd 6/8/2006 7:42 PM Page 387
The basis of public choice theory and the because of considerable pressure from the
focus of agency theory have been broadened by World Bank, International Monetary Fund,
transaction costs theory, which addresses and other international or regional lending
factors affecting the relative costs of respond- agencies.
ing to market circumstances through different A second element has to do with market
governance arrangements, including means by globalisation involving international competi-
which principals can cope with agents. The tion and investment. Governments have fre-
factors include market size and certainty, the quently seen the need to divest enterprises,
availability of required skills, the extent of notably in the airline, banking and telecommu-
opportunistic behaviour on the part of key nications sectors, as a means of opening them
individuals, and the physical and cognitive up directly to international market forces and
capacity of individuals to collect and use infor- equity with the aim of expanding their scope of
mation. These factors have been relevant to influence, enhancing their performance, and
public enterprises, as they have to private increasing their financial resources.
firms, but their implications in terms of divest- A third element concerns both ideas and
ment have been mixed. For example, the market forces. Governments have often
divestment of a monopoly with no change in rethought what constitutes an asset of strategic
its monopoly status will not have altered its significance. In doing so, they have realised, for
basic mode of operation and associated trans- example, that utilities such as electricity, gas
actions in any appreciable way. This contrasts and water can be managed effectively by pri-
with the breaking up and divestment of a vate firms with either a domestic or an inter-
monopoly, say, in the electricity industry into national base, though subject to some form of
generation, transmission and supply. Such state regulation.
action will have served to decrease intra-com- A further line of argument addresses
pany transaction costs and also to foreclose on the cases where governments have only par-
any internal cross-subsidisation, while increas- tially divested selected enterprises, resulting
ing external costs because of there now being in mixed public and private ownership
more companies involved and possibly a (Thynne 1995, 1998; Thynne and Ariff 1989;
degree of competition, particularly with regard Wettenhall and Thynne 2002a). The impetus
to supply. Thus, overall, while transaction con- has often been for governments to seek to
siderations have been used to support divest- enjoy the best of both worlds. They have been
ment, resulting in companies in private able to acquire needed resources from the
ownership, specific features of the market from sales involved, while still retaining a degree of
one sector to another have obviously impacted internal control through the ongoing exercise
on company performance. of ownership rights. This has enabled them to
Another body of theory – or, perhaps, theory influence the operation and direction of the
in the making – concerns elements of global- enterprises at the same time as facilitating the
isation which have stimulated divestment injection of private funds and some market
(Farazmand 2002; Hood 1994; Ikenberry 1990; control.
Lane and Ersson 2002). At least three interre- The retention of state ownership, partially
lated elements are worth noting. All three have for some enterprises and fully for others,
been significant, but not necessarily applicable to has raised questions about the way in which
the experience in all countries. governments have actually perceived and
One element is the global spread of ideas approached their ownership and control
and practices. Thus, one country after another responsibilities. In this regard, there are two
has seemingly felt compelled to get on the broad models (Wettenhall and Thynne 2002b).
divestment “bandwagon”. This has often been One is “public” in the sense that it emphasises
by choice, with governments consciously seek- the centrality of the state and the public at
ing to reap the financial benefits of enterprise large. The other is essentially “private” to the
and other asset sales. But it has often also been government of the day.
Peter-3383-Chapter-22.qxd 6/8/2006 7:42 PM Page 388
The public model suggests that a government The theories, models and related ideas
can create and hold shares in a company only addressed here have all had some bearing on
where it has an explicit, legally prescribed right the development of capital, be it economic,
to do so. The company in question is best political or social (Fukuyama 2001; Hodge
described as being “publicly-owned” or “state- 2000; Schuurman 2003; Self 2000). Various
owned”. The government acts on behalf of the relationships have been assumed or actually
public as a delegate or proxy, with the public forged by the choices and decisions made on
being the real owners in their collective capacity divestment and other reforms. Some have been
as members of the state. The activities of the based on an identifiable cause and effect.
company are rightly subject to parliamentary Others have been less certain in this regard, but
and judicial review as a matter of public law. still important.
The private model, by contrast, implies that Capital can be an elusive thing which is hard
a government can create and hold shares in a to pin down and define, especially when
company largely by reason of its prerogative. applied not just to economic affairs but also to
The company concerned is best described as political and social life. In political, economic
being “government-owned”. It essentially exists and social terms, it broadly comprises two
and operates at the behest of the government, interrelated elements, respectively. First, there
within the bounds of private law. is the support, wealth and trust which are gen-
Companies functioning partially or fully in erated by some kind of promise or action,
state hands have tended to be in line with the exchange and association. Second, the support,
private model. This has meant that govern- wealth and trust so generated could well be
ments have often been able to interact and deal essential to the achievement of desired goals
with them, including organising their divest- within a state, market and society. Thus,
ment, almost as though they were personal in political, economic and social analyses, cap-
possessions. They have thus proved to be valu- ital is potentially both a dependent and an
able instruments of government action, with independent variable.
policy goals, party influences and individual Public choice, agency and transaction costs
political commitments all potentially coming theories, along with ideas on globalisation,
into play without necessarily impeding com- have had implications particularly for eco-
pany efficiency and profitability. Indeed, given nomic capital formation through increases in
their status as free-standing legal persons, they company productivity, business competition,
will sometimes have been used by govern- investor interest, and so on. These have all been
ments to achieve purposes that the govern- aims and possible consequences of divestment
ments themselves have actually lacked the in individual cases or sectors and more gener-
constitutional or statutory power to achieve ally within an economy. They can be assessed
(Seddon and Bottomley 1998). with reference to actual results, which in turn
The internationalisation of company affairs need to be causally linked to economic growth
involving mixed ownership arrangements and increased levels of economic development
across state borders has opened the possibility as wider policy goals. Thus, a two-stage analy-
of governments directly influencing specific sis is required for any convincing or com-
commercial developments beyond their con- pelling conclusions to be reached.
stitutional and political spheres of responsibil- Such an analysis is also necessary in relation
ity. This has created an incentive for them to to political and social capital. These are more
support company mergers, share-swapping subjective forms of capital than the economic
agreements and other kinds of business and, accordingly, are much less amenable to
alliances in the international arena, but unfor- quantification. But this is not to say anything
tunately with little thought being given to the about their relative importance to the opera-
implications for public control, accountability tional dynamics of a state, market and society.
and legitimacy (Lane 2002; Thynne 2003a; For the enhancement of political capital,
Wettenhall 1993). governments have needed to play-up the
Peter-3383-Chapter-22.qxd 6/8/2006 7:42 PM Page 389
expected advantages of divestment in terms of These are matters that governments could
debt and deficit reductions, caps on tax rises, well have played on in taking various divest-
and investment opportunities, all as a counter ment decisions. In doing so, however, they will
to public concerns about likely price hikes, have needed to recognise that the principal-
staff lay-offs, inflated salaries for top manage- agent relationships involved have inevitably
ment, and so on following divestment. Any become more complex. Employees will have
support that they have been able to garner by become principals through share-ownership,
stressing the possible benefits of divestment while still being agents through their employee
will have had little political significance unless positions. Customers, too, will have become
it was manifest in an increased vote for them, principals, as well as being recipients and ben-
coupled with an ability on their part to use the eficiaries of the goods and services being pro-
increase to meet important policy commit- duced or provided by the companies. Also,
ments and objectives. Thus, it is necessary to where the companies have remained partly
consider the relationship between expressed state-owned, both groups will have continued
support and voting percentages, followed by to be principals collectively as members of the
the impact of the percentages on policy capa- public within a state. In all cases, there will
city and action. have been significant implications for com-
Similarly, of relevance to social capital, it has pany control and accountability.
been expected that, where members of the
public, management and employees become
shareholders in divested companies, they will The How-type
develop a particular affinity with those compa-
nies and possibly also have some beneficial Why-type theories and models such as those
associational opportunities. But, in practice, to just considered have been complemented by
what extent, if any, have these factors had a various process-based, how-type theories and
positive affect on their willingness to trust one models about the scope, direction and style of
another, the companies, and other organisa- policy-making in government, including the
tions? In turn, has this led to any increase in existence of reform drivers and supporters.
social togetherness, cohesion and well-being These theories, etc. can be used to throw light
within the companies and the community? on how divestment and associated initiatives
These are empirical questions which ideally have been approached, with politicians,
need to be addressed, but no doubt with some administrators and others coming together in
difficulty. differing ways to bring about or limit reform.
Aspects of economic, political and social Notable are theories concerning rational-
capital potentially have been intertwined when comprehensiveness, mixed-scanning, incre-
employees of divested companies have invested mentalism, opportunities and rummaging in
in those companies and, as a result, have contri- the policy process (Cohen, March and Olsen
buted more actively to company productivity, 1972; Etzioni 1967; Howlett and Ramesh 1995,
profitability and growth, have become politi- chs 5, 7; Kingdon 1984; Lindblom 1959, 1979).
cally disposed towards the government which The comprehensiveness of the rational-
provided them with the opportunity to invest, comprehensive theory has had some descrip-
and have felt much more part of the social life tive value, especially where there has been a
of the companies and the immediate environ- fundamental re-assessment of the nature and
ment thereof. A broadly similar result could extent of state activity, leading to widespread
possibly also have flowed from customer divestment as, for example, in the United
investment, with positive contributions being Kingdom and New Zealand in the 1980s and
made to profitability through increased pur- early 1990s, as well as in countries in Eastern
chases, along with a change in political dispo- Europe. The detailed analyses of values, goals,
sitions and in feelings towards the companies problems, options and possible consequences
concerned. have resulted in considerable structural and
Peter-3383-Chapter-22.qxd 6/8/2006 7:42 PM Page 390
bonds between politicians and administrators, well understood in the case of many countries,
producing a quite tightly integrated elite, have the capital development and associated fea-
resulted in the two groups being joint drivers, tures have not been studied and highlighted to
with little need to manage the legislature. any appreciable extent. They clearly remain an
The first model has covered situations in important item of unfinished business in
which divestment and related reforms have been research and analysis. The theories and models
extensive, with the roles, organisations and man- considered here serve generally to identify
agement practices of the state being reshaped areas of future inquiry in which questions
and transformed quite substantially. The state about capital and reform impacts in their
has been “rolled back” as an active owner, pro- various forms need to be addressed. Such
ducer and provider, while also being equipped, inquiry ought ideally to refine and extend the
or re-equipped, with a facilitatory and regula- theories and models, with detailed reference to
tory capacity to determine national agendas and ongoing practical experience in a wide range of
to protect matters of public interest. countries and with a keen eye on the ways in
The second model has applied to reform that which choices and related developments are, or
has been relatively modest, with an emphasis can become, closely intertwined in the social,
more on changing organisations and manage- economic and political life of societies in
ment practices than on reshaping the state various parts of the world. Only then will it be
through significant divestment and associated possible to reach some very meaningful con-
initiatives. The state has largely being “rolled on” clusions about whether privatisation by divest-
in its existing roles, while having its organisa- ment is something largely to be heralded or to
tional and management capability upgraded. be scorned.
The third model has seen considerable
reform, but less than in the first model, with
some reshaping of the state involving strategic
divestment, along with other changes to organ- REFERENCES
isations and management practices. The
reshaping has led to the state’s being “rolled Ariff, M. Asher, M. and Thynne, I. (1995) “Singapore”
over” into some new areas of activity, while in I. Thynne (ed.), Corporatization, Divestment and
also being both “rolled on” and re-equipped, the Public-Private Mix: Selected Country Studies,
with a concomitant transformation of organi- Hong Kong: AJPA in collaboration with IASIA.
sations and management. BBC News (2003) “BAA ‘Golden Share’ Ruled Illegal”,
http://news.bbc.co.uk/2/hi/business/3022809.
Obviously, numerous other drivers and sup-
stm 8 August.
porters have also contributed to divestment Boston, J. Martin, J. Pallot, J. and Walsh, P. (1996)
and related reforms, including personal advisers Public Management: The New Zealand Model,
to politicians, hired consultants, merchant Auckland: Oxford University Press.
bankers and associated financial experts, the Cohen, M. March, J. and Olsen, J. (1972) “A Garbage
World Bank and IMF, the media, interest Can Model of Organisational Choice”, Adminis-
groups, and the voting public. All have had trative Science Quarterly, 17(1): 1–25.
some form of influence and impact, positively Coopers and Lybrand (1988) Privatisation: Its Place
or negatively, on the political, economic and in Public Enterprise Reform, London: Coopers and
social affairs of communities through their Lybrand.
reform activities. Curtis, D. (1991) Beyond Government: Organisations
for Common Benefit, London: Macmillan.
Etzioni, A. (1967) “Mixed Scanning: A ‘Third’
Approach to Decision-Making”, Public Adminis-
CONCLUDING COMMENT tration Review, 27(5): 385–392.
Evans, P. (1997) “Government Action, Social Capital,
and Development: Reviewing the Evidence on
While the strategic choice aspects of divest- Synergy”, in P. Evans (ed.), State-Society Synergy:
ment as a significant reform policy are now Government and Social Capital in Development,
Peter-3383-Chapter-22.qxd 6/8/2006 7:42 PM Page 392
in Comparative Perspective”, Public Management Wettenhall, R. and Thynne, I. (guest eds) (2002a)
Review, 5(3): 449–459. “Symposium on Turn-of-the-Century Trends and
Thynne, I. and Ariff, M. (eds) (1989) Privatisation: Future Prospects in Public Enterprise and
Singapore’s Experience in Perspective, Singapore: Privatization: Contexts, Structures and Dynamics”,
Longman. Public Finance and Management, 2(1): 1–153.
Thynne, I. and Wettenhall, R. (2001) “Public Wettenhall, R. and Thynne, I. (2002b) “Public
Enterprises: Many Faces, Much Questioning, New Enterprise and Privatization in a New Century:
Challenges”, International Review of Public Evolving Patterns of Governance and Public
Administration, 6(1): 1–9. Management”, Public Finance and Management,
Vass, P. (1992) “Regulated Public Service Industries”, 2(1): 1–24.
in F. Terry and P. Jackson (eds), The Public Williamson, O. (1985) The Economic Institutions of
Domain: The Public Services Year-Book, London: Capitalism: Firms, Markets, Relational Contracting,
Chapman and Hall. New York: Free Press.
Veljanovski, C. (1987) Selling the State: Privatisation Wiltshire, K. W. (1987) Privatisation: The British
in Britain, London: Weidenfeld and Nicolson. Experience – An Australian Perspective, Melbourne:
Wettenhall, R. (1993) “The Globalization of Public Committee for Economic Development of Australia
Enterprises”, International Review of Administrative and Longman Cheshire.
Sciences, 59(3): 387–408. World Bank (1995) Bureaucrats in Business: The
Wettenhall, R. (2003) “The Rhetoric and Reality of Economics and Politics of Government Ownership,
Public-Private Partnerships”, Public Organization New York: Oxford University Press.
Review, 3(1): 77–107.
Peter-3383-Chapter-22.qxd 6/8/2006 7:42 PM Page 394
Peter-3383-Chapter-23.qxd 6/8/2006 8:26 PM Page 395
Section Three
Evaluating Policy
Peter-3383-Chapter-23.qxd 6/8/2006 8:26 PM Page 396
Peter-3383-Chapter-23.qxd 6/8/2006 8:26 PM Page 397
23
Evaluation Research
EVERT VEDUNG
(R E S U L T S)
Figure 23.1 The general system model adapted to government intervention evaluation
Source: Adapted from Vedung 1997: 5.
three of them. The term implementation 5. The implementation problem: Are there
usually covers administration and output. The any obstacles or malfunctionings in the
reasoning is summarized in Figure 23.1. execution phase between the formal
instigation of the intervention through the
final outputs? How can such problems be
THE EIGHT PROBLEMS APPROACH mitigated?
TO EVALUATION 6. The outcomes problem: What are the pos-
sibly relevant outcomes – immediate,
intermediate and ultimate, intended,
The problems to be attacked in evaluation can
unintended – of the intervention?
be phrased as eight questions. I shall call this
7. The impact (effect) problem: To what extent
the Eight Problems Approach to Public Policy
is the outcome effects of the intervention?
Evaluation.
Besides the intervention, what other contin-
gencies (operating causal forces) or factors
1. The comprehensive purpose problem: For
(mechanisms) contributed to the outcome?
what overall aims is the evaluation
8. The utilization problem: How is the evalu-
launched?
ation to be utilized? How is it actually used?
2. The organization (evaluator) problem:
How can utilization be improved?
Who should exercise the evaluation and
how should it be organized? Source: Adapted from Vedung 1997: 93 f.
3. The value criteria problem: By what value
criteria can and should the merits of the
intervention be assessed? By what stan-
COMPREHENSIVE PURPOSES
dards of performance on the value criteria
OF EVALUATION
can and should success or failure or satis-
factory performance be judged? And what
are the actual merits and demerits of the The overall rationale of evaluation is to create
intervention? repositories of descriptive and judgmental
4. The intervention analysis problem: How insights for reasoned practical thought and
is the evaluand, that is, the government action. Under this practice-servicing umbrella,
intervention, e.g. the policy, the program, evaluation is performed for either accountabil-
the components of policies and programs, ity, improvement or basic insights. In addition,
the project, the provision of services and evaluation is laced with strategic aims to gain
goods, or the management strategy, to be time, or to show up a front of rationality to the
characterized and described? world.
Peter-3383-Chapter-23.qxd 6/8/2006 8:26 PM Page 399
The key rationale of accountability evaluation scheme to use evaluation to gain time, hide
is to find out how agents have exercised their shortcomings from their principals, display
delegated powers so that principals can judge attractive images of programs and projects, and
their work and take appropriate action. The in general provide appearances more flattering
first dimension of accountability is reporting than reality. Evaluations are launched to become
in which descriptions are provided. A second Potemkin villages.
function is a justifying analysis or explanation.
In instances where a justifying analysis recog-
nizes deficiencies, true accountability requires THE ORGANIZATION
answerability – that is those in charge must be (EVALUATOR) PROBLEM
held responsible.
In the improvement perspective, evaluation
Who should exercise evaluation and how
aspires to guide amelioration of the interven-
should it be organized? Internal versus external
tion in its implementation. It is felt that the
evaluation is one option.
intervention will continue its operations in the
Internal evaluation is the use by an organiza-
foreseeable future and that it must function as
tion of evaluators who are employees of that
smoothly as possible. Sub-aims in this category
organization to assess the organization’s own
include effectiveness, cost-efficiency, sustain-
interventions. An internal evaluation is arranged
ability, equitability, and being adapted to client
and financed by the organization, carried out by
concerns and needs.
people from the organization, focused on the
The primary intended user of improvement
organization’s own activities and intended to be
evaluation is staff and management immedi-
used primarily by people in the organization.
ately responsible for or involved with the inter-
If activities of an organization are evaluated
vention. Accountability evaluation, on the
by employees of an outside body, the evaluation
other hand, is a tool for superiors to check
is external. External evaluations can be arranged
their subordinates and keep them and the
and used by insiders, but must by definition
intervention responsible for their actions.
be produced by outsiders. When a subordinate
Intervention improvement, or “learning”,
agency is commissioned by its principal to do an
“lesson-learning”, “formative evaluation”,
accountability report, and the subordinate con-
“promotion”, is a worthy aspiration of evalua-
tracts with some consulting firm to produce the
tion. Several experts, particularly in education,
study, the assessment will be external.
maintain that learning must be the major
External evaluations may be contracted out
comprehensive purpose of evaluation.
to for-profit consultants and non-profit insti-
In the third comprehensive purpose, basic
tutions. The latter include public investigatory
insights, evaluation is seen as a kind of funda-
commissions, government auditing agencies,
mental research that seeks to increase the gen-
universities and other institutions of higher
eral understanding of actions and events in the
learning, government-funded public policy
public sector. This is probably the major aim of
institutes, and non-profit think tanks and
evaluative investigations initiated, carried out,
foundations. Mass media, think tanks and
and funded by academics. In evaluations
universities may also carry out evaluations on
commissioned by public sector agents to be
their own initiative and financing.
swiftly reported, the basic insights purpose is
secondary to accountability and improvement,
and best regarded as a by-product consideration
CRITERIA OF MERIT
in the pursuance of the other two. In metaeval-
uation or synthesis analysis, the basic insights
ambition seems adamant. A key process in evaluation is to determine the
Since public intervention evaluation takes merit, worth or quality of the public intervention
place in action settings, it is always permeated under appraisal. The quandary is: what consti-
with strategic purposes, often hidden. Agents tutes a valuable public intervention and how
Peter-3383-Chapter-23.qxd 6/8/2006 8:26 PM Page 400
can it be appraised? This is captured by four Aside from the general orientations to be
exercises: adopted (ex ante – ex nunc – ex post, descriptive –
prescriptive) the particular values preferred in
• Identification of appropriate criteria of each orientation must be justified. The most
merit to be used in the assessment. commonly used substance criteria, economic cri-
• On the chosen criteria, selection of perfor- teria and process criteria are listed in Figure 23.2.
mance standards that constitute success or Substance-only criteria primarily address
failure. substantive intervention content, outputs and
• Ascertaining the actual performance of the effects (and, secondarily, the processes preced-
evaluand on each criterion and comparing ing and connecting them). Economic merit
it to each standard. criteria attend to either intervention costs only
• Decision whether or not to integrate the or to the relationship between intervention
judgments into a single, overall appraisal of costs and substantive output and impact.
worth of the intervention. Process criteria, finally, check for legality,
equity, publicity, representativeness, participa-
Criteria may be determined at the beginning of tory democracy, deliberative democracy, and
the evaluation (ex ante), during the process of other qualities of the procedures according to
doing the evaluation (ex nunc) or after the which the interventions are supposed to be
evaluation is finished (ex post). handled by ministries and agencies.
Furthermore, criteria can be descriptive or
prescriptive. In descriptive valuing, the evalua-
tor chooses the values of others as criteria and
GOALS
standards. In prescriptive valuing, the evaluator
herself advocates the primacy of particular
values, such as justice, equality and client needs, The use of the institutionalized goals and objec-
regardless of whether these values are adopted tives of the intervention is a classic in evalua-
by any decision-making body or held by some tion. Earlier literature took it for granted that
stakeholding constituency. evaluators should use prefixed intervention
Peter-3383-Chapter-23.qxd 6/8/2006 8:26 PM Page 401
goals as value criteria. Nowadays, many more Particular goals may be ambiguous and carry
criteria are actually applied. two or more simultaneous meanings. Yet ambi-
There are, however, several worthwhile guity in this sense of dual meanings is excep-
reasons in favor of goal-attainment assessment tional in political and bureaucratic language,
in the public sector. Fundamental is the argu- and barely bothers evaluators. More uncertainty
ment from representative democracy. is caused by vagueness. The outer border delim-
In a democracy, all power belongs to the iting the extension of a vague word is so fuzzy
people. Yet, the people cannot make every that within a certain range it is impossible to
decision concerning citizens and nature due know what is included in the extension and
to lack of competence and lack of time. The what is not.
people doesn’t have the necessary time and Yet, the real problem with haziness is produced
competence to participate in millions of deci- by goal catalog. Most large social reforms contain
sions on, for instance, placement of patients in impressive directories of diverse goals. While a
line for surgery, or day-to-day care for ailing single goal may be hailed as the major one, it
senior citizens in public-sector homes for the is often also maintained that this one must be
elderly. For these reasons the citizenry must balanced against all the others, maybe including
elect political representatives to make the deci- potentially conflicting ones. But the necessary
sions for them. But representatives in political trade-offs between the several goals are not indi-
assemblies don’t have time and competence to cated, which makes it impossible to elicit from
make all decisions. They must delegate their such lists of goals one distinct, transparent,
power to governments to make decisions for expected outcome. Thus, an array of program
them. But governments don’t have time and goals of this type are not lucid enough to be
crucial knowledge, so they in turn have to del- usable as value criteria against which to measure
egate to civil servants and professionals to take intervention successes and failures.
decisions, etc. etc. The public sector is made up The gist of the inflationary argument is that
of long chains of principal-agent relationships. goals are exaggerated. They are set not only to be
To safeguard that agents won’t cheat their achieved but to garner support from various
principals, public-sector goal-setting is cir- audiences. Therefore goals are not good value
cumscribed by formal rule systems. Goal- criteria for performance and achievement.
setters are obligated to honor the rules of the The most compelling rebuttal of the prede-
constitution and more specific rules of proce- termined goals approach emanates from its
dure. Once a decision on goals comes out of blindness to unintended side-effects. Public
the system, it has a special institutional status. sector interventions invariably lead to conse-
This is the case with all entities in the public quences which were not foreseen in the origi-
sector. If an agency adopts a program in order nal decision situation. Since nobody can state
to reach some goals, these stated goals get in advance any reasonable goals for something
their legitimacy from the fact that the agency’s unknown, the evaluator guided by premedi-
decision-making authority has been delegated tated stated goals only won’t be able to trace
to it by the government and that the govern- any unforeseen spillovers. And if she discovers
ment, in turn, has received its authority to do such things, there will be no prestated goals
so from parliament, and the parliament, in which can work as value criteria for judging
turn, from the people. It is a merit of the goal- them. The prestated goals approach is obvi-
attainment approach that it recognizes this ously deficient on this account.
democratic aspect of public-sector goals.
On the other hand, the goal-attainment
approach also suffers from weaknesses. The
CRITERIA FOR SIDE-EFFECTS
haziness argument maintains that predeter-
mined intervention goals are deficient as criteria
of merit due to their obscurity. Occasionally, It is a compelling duty of evaluation to map and
programs are based on indeterminate goals. assess serendipitous results and unanticipated
Peter-3383-Chapter-23.qxd 6/8/2006 8:26 PM Page 402
side-effects outside goal areas. The term “side- service process and service administration. Is
effects” refers to at least partial consequences the core service tailored to the demands of the
of the intervention occurring outside its target clients? Do the service employees encounter
areas. Taking this into consideration, evalua- the clients with respect and correctness? These
tors should always search for side-effects while are two questions that might be answered by
not ignoring main effects, of course. In these the service-users in their self-reports.
enterprises, preset intervention goals are The clients may also choose to raise the
retained as the fundamental value yardsticks effects issue, that is, estimate service impacts
for the main effects. But for the unanticipated on themselves or on the client community in
side-effects, value criteria must be developed general. In such effects-on-clients evaluation,
either in the evaluation process or ex post targets try to determine the relative change
when the evaluation is finished. in themselves or in the overall client body as a
result of their participation in specific treat-
ment modalities.
In data assembly, advocates of client-criteria
CLIENT-ORIENTED CRITERIA
prefer client self-observation and sustained
interviewing to questionnaires, documentary
Responsiveness to intervention client concerns methods, and evaluator observation. In-depth
has been suggested as an alternative or supple- interviewing of individual targets is one
ment to intervention goals and merit criteria favored technique, distribution of self-report
for side-effects. Cases of such concerns are the instruments that clients and their network can
desires, values, objectives, and expectations of easily complete themselves another. In some
the intervention addressees. At the heart of the cases, client-oriented evaluators endorse focus-
full-blown client-oriented approach is the group interviewing which allows for group
question whether the intervention, its imple- deliberations among the participants and
mentation, outputs, and effects satisfy client between the participants and the evaluator.
desires, values, and interests, or are in accord Such forums of debate will support the devel-
with client expectations, assumptions, and opment of new ideas, service concepts, solu-
needs. Client criteria can be weakly, moderately tions and technologies. It might also, as a
and strongly included in the evaluation. In the side-effect, educate participants to become
weakly-included case, participants respond to better citizens in the future.
data-gathering instruments and provide infor- Today, client-based value criteria are
mation but nothing more. Studies of client employed in numerous evaluative contexts, par-
satisfaction with public-sector services belong ticularly those concerning public service provi-
to this category. sion such as child care, nursing homes
Strongly-included means that the evalua- for the elderly, public housing, mental health,
tion is initiated, funded, designed, carried out urban transit, public utilities, parks and
and reported by the clients themselves. Let me recreation, and physical health services, where
reason from the moderately-included case clientele participation is crucial to program
where the evaluation (i) is commissioned by operation. Client-based criteria are used to eval-
the administrators but (ii) charted to involve uate library services, museums, national parks,
the service-users in important decisions swimming pools, soccer fields, tennis halls, trash
regarding its planning and execution. Once an hauling, street cleaning, snow removal, traffic
evaluation team is appointed and its terms of noise, traffic congestion, and urban transit. It is
reference agreed upon, intervention clients are a favorite with educators. At universities, stu-
located and sampled. Now clients are asked dents are routinely requested to share their
to select intervention dimensions to be judged opinions of courses, reading lists and lectures.
and value criteria to be used. For instance, They are asked to rate their teachers’ abilities to
clients are asked to judge program output, organize the course contents, to stimulate and
service availability, service quality, or even promote altercation and discussion, to stir
Peter-3383-Chapter-23.qxd 6/8/2006 8:26 PM Page 403
student interest, foster critical thinking, and to other things they are preoccupied with besides
show concern and enthusiasm for the students. providing good services.
In some contexts, these evaluations are occa- But evaluators must be aware of the ten-
sionally used to rank faculty and courses so that dency of the clientele to exaggerate complaints
prospective future students can make better- in order to get more service. Clients may also
informed choices. nurture fiscal illusions. Greater client involve-
The use of client concerns as value criteria ment in evaluation may surrender power to
will adjust public interventions to serve the groups with vested and narrow interests.
clients. It is based on the notion that public All in all, application of client criteria may
administration produces goods and services for supplement the previously presented merit
customers in the market place. In buying a com- criteria, since they pose other problems for con-
modity in the store, the consumer pays no sideration. They can make important contribu-
attention to producer goals. Her own assess- tions to evaluation, but must be balanced
ment of the value of the good is what counts. against other criteria like goal-attainment and
Public services should also be regarded in this professional norms for service excellence.
way and geared toward consumer tastes. The
main rationale of public services is actually to
serve the public. Therefore, the customer’s own
PROFESSIONAL MODELS:
assessment of the value of the good is what
SELF-EVALUATION PLUS PEER REVIEW
counts. In situations where recipients are mod-
erately to strongly involved in the evaluation,
an additional participatory-democracy and Another set of value criteria are professional
deliberative-democracy case is made for the use conceptions of merit. Members of a profession
of client-based criteria. According to this justifi- are entrusted to evaluate their own and their
cation, the customer parallel cannot be pushed colleagues’ performances with respect to the
too far, since the client notion includes a partic- profession’s own criteria of merit and stan-
ipatory as well as a deliberative aspect, which is dards of quality.
absent from the customer concept. The partici- In some areas of public life, quality criteria
patory feature suggests that clients are also citi- are so complex and the subject-matter so intri-
zens who may voice their complaints and cate that political officials have found it wise
desires to the evaluators and service providers, to leave the shaping and debating of them to
and to some extent influence and take responsi- well-educated professionals. Architects, judges,
bility for service content. The deliberative university professors, doctors, nurses, veterinari-
feature engenders a discursive, reasoning, dis- ans and engineers would be cases in point.
cussing, learning-through-dialog countenance, Hence, it is also considered natural to delegate ex
which may educate clients to become better nunc and ex post evaluation to the professions.
citizens in general: the customer as citizen Evaluation of research in an economics
rather than the customer as consumer. Still, the department of a European university may be
major idea is to get the services more geared used as a case. The professional evaluation
toward consumer desires. The better-citizens usually starts with self-evaluation by the evalua-
aspect is best regarded as a side-effect. tees. The professionals to be evaluated carry out
Adoption of client criteria may also increase an appraisal of their own performance. What
the legitimacy of the intervention. If clients are are the strengths and weaknesses of our depart-
allowed to participate and exert some influ- ment? What opportunities do we have?
ence on the intervention, their approval of the What threats are confronting us now or in the
intervention will probably increase. near future? In sum, they perform a SWOT
In addition, application of client criteria analysis.
may foster effectiveness and efficiency because Then comes a peer review. Highly renowned
concentration on clients will force service- scientists of the particular field, but from some
providers and managers to do away with many other universities, are assigned to appraise the
Peter-3383-Chapter-23.qxd 6/8/2006 8:26 PM Page 404
scholarly quality and relevance. Beside the She singles out the intervention’s primary
written self-evaluation, these peers base their target group, the clients, and the clients’ associ-
assessment on additional documentary evi- ations. She identifies relatives and relatives’
dence and site visits with face-to-face evaluator– associations. She may also include lay people.
evaluatee dialogs. The evaluators pass their And she searches for those who are unaware of
preliminary judgments in a draft report, often the stake they hold.
organized according to the SWOT-scheme. Advocates of stakeholder criteria nurture a
Then the evaluatees are given the opportunity strong penchant for qualitative, interactive,
to comment on the draft report before it is hands-on methodology. The evaluator must
finalized. The final peer review report is partic- talk to the stakeholders to elicit their narrative
ularly aimed at performing an overall quality histories and observational data. The evaluator
judgment of the evaluand. must be responsive and let the issues and con-
Usually, evaluation based on professional cerns of the affected people govern the next
merit criteria is interactive. The evaluators step in the investigatory enterprise. After a
meet and listen to the evaluatees and the eval- while, she might discover what issues various
uatees learn from the evaluators. Dialog, delib- stakeholders are occupied with and what con-
eration and interaction among the evaluatees cerns they nurture regarding these issues. The
themselves and between evaluatees and evalu- evaluation design will be gradually deter-
ators are important. mined. Stakeholder evaluation is responsive
Self-evaluations and peer reviews frequently evaluation.
produce questionable results. Studies with To gather final data on stakeholder issues and
matched panels show that peers use widely dif- concerns, stakeholder self-observation and sus-
ferent merit criteria and performance standards tained dialogical methods are preferred to sur-
and reach miscellaneous conclusions. However, veys, questionnaires and statistics. In-depth
in complex fields, interactive collegial evalua- interviewing of individual targets is one favored
tion is probably the finest method available to technique. Self-report instruments that clients,
judge the quality of what is produced. parents, relatives and other stakeholding net-
works can easily complete is used. In some
cases, stakeholder evaluators endorse focus-
group interviewing, which allows for group delib-
STAKEHOLDER CONCERNS AND ISSUES
erations among the participants and between
the participants and the evaluator. After data are
Finally, the concerns and issues of all actors amassed and processed, the reporting of find-
who have an interest in or are affected by the ings, which might vary from one stakeholder to
intervention can be employed as evaluative another, will commence. The key word seems to
yardsticks. be “portrayals”, that is, information-rich charac-
The stakeholders may constitute themselves terizations using pictures, anecdotes, thick
as the evaluation team and carry out the descriptions, and quotes.
evaluation. In the sequel I shall reason from The use of stakeholder criteria have several
another case where the evaluation is conducted advantages. The democratic arguments depart
by particular evaluators, who elicit the views of from participative and deliberative points of
the stakeholders. view. True, democracy means that citizens in
The stakeholder approach starts by evalua- general elections vote for competing elites
tor mapping of the major groups who are that are supposed to make decisions on their
involved or thought to have an interest in the behalf (representative democracy). Yet, the citi-
intervention. The evaluator identifies the zenry should also be able to partake in final
people who initiated, hammered out, funded, public decision-making between elections
and adopted the intervention. She identifies (participative democracy). Furthermore, dis-
those who are charged with its implementa- cussion, dialog and debate are also important
tion: managers, staff, and front-line operators. democratic values because they help people to
Peter-3383-Chapter-23.qxd 6/8/2006 8:26 PM Page 405
form and refine their beliefs and preferences to more than vaguely concerned groups and
(deliberative democracy). the majority of citizens.
Stakeholder models satisfy these participa- This raises a problem with stakeholder evalu-
tive and deliberative values somewhat more ation, or, more properly, collective decision-
than the goal-achievement and the side-effects making on the insights base brought forward by
approaches. It might also, as a positive spillover, stakeholder evaluations. Since representative
educate participants to become better citizens democracy is the dominant form of democratic
in the future. government in most developed societies, deci-
According to the knowledge argument, stake- sions on various sectoral policies cannot be left
holders nurture convictions about inadvertent to the stakeholders. The power belongs to the
side-effects, sophisticated implementation people, its representatives and their delegates.
barriers, and outright cheating which may fur- Stakeholder evaluation must work within the
nish the evaluator with ideas about topics frames fixed by representative democracy.
for further investigation. Since stakeholder- Another objection involves the risk that the
orientation will bring up more aspects of the stakeholder approach will embrace a pragmatic
subject-matter for discussion, the quality of theory of truth. Truth may turn into a matter of
the evaluation findings will increase. usefulness, utility or acceptability to stakehold-
There is also a utilization argument in sup- ers. Stakeholders often entertain politicized
port of stakeholder evaluation. The stakeholder views of program effects. Supporters ascribe
approach increases the chances that issues of everything positive occurring after the program
genuine interest to concerned parties will be as caused by the program and everything nega-
addressed. It brings to light information that tive as caused by something else. Detractors
meets the real requirements of the different hold the opposite view. Facts are essentially
stakeholders, thereby enhancing the probability contested. In these situations the various par-
that the findings actually will be put to use. ties will accept only those findings that fit into
A last rationale for stakeholder models is that their preformed opinions.
they might promote compromises, and forestall
bitter political struggle. Stakeholder assemblies
are vehicles for shaping agreement on the results
COST CRITERIA OF MERIT
of earlier efforts and, most importantly, propos-
als for future action. Consensus-building and
the rendition of legitimacy to fundamental Substance-only criteria pay no heed to costs. To
decisions are considered great advantages of remedy this omission, economists have devised
stakeholder evaluation. three cost criteria: economy, productivity and
There are also drawbacks with stakeholder efficiency. Economy is a cost-only criterion. Is the
criteria. Stakeholder evaluations are inordi- intervention carried out in a reasonably cheap
nately impractical and resource-demanding, fashion? Have cheaper options been preferred to
since every stakeholding constituency must be more expensive ones? Is implementation carried
contacted and nurtured. Stakeholder models out in the cheapest way possible?
are fuzzy. They provide no clear-cut answer Productivity is the ratio of outputs to costs, i.e.
to the question of who the stakeholders are. outputs:inputs. Productivity is operationalized
Furthermore, all the stakeholding audiences, differently due to type of output and type of
however selected, are treated as equals. But, in cost. A study of productivity in municipal
a representative democracy, elected politicians libraries can use cost productivity as a measure:
must carry more weight than administrators number of books borrowed: costs in euros (or
or experts on the substantive matters under dollars) to have these books borrowed. As an
consideration. The stakeholder model embod- alternative to cost productivity, work productiv-
ies no priorities among the stakeholders. There is ity can be used. It can be illustrated by the
a risk that small, well-organized and very com- expression: number of books borrowed: number
mitted stakeholders are consulted and listened of hours worked.
Peter-3383-Chapter-23.qxd 6/8/2006 8:26 PM Page 406
and the sixth problem in the Eight Problems intervention or are they brought about by
Approach. something else?
Qualified monitoring, commonly called for-
mative evaluation, checks the various stages in
the implementation chain through the outputs Types of Effects
and outcomes in search for malfunctions and
hindrances. The monitor searches for prob- To capture effects the evaluation community has
lems, discovers problems, verifies problems, developed an analytical language. Main effect is
disentangles causes of problems, extricates defined as the central substantive impact that the
mechanisms releasing problems, suggests solu- intervention instigators by intention wanted to
tions to problems. Is the intervention being achieve. Main effects are associated with the sub-
delivered to the clients in ways envisioned stantive objectives of the intervention-makers
in the formal intervention decision? Is the and with what the intervention was capable of
intervention as delivered reaching all the achieving. Aside from transpiring in the target
prospective participants? In addition, qualified area, main effects are by definition anticipated
monitoring also focuses on processes preced- as well as positively valued by the intervention
ing delivery. Qualified monitoring is actually a adopters. In a similar vein, a side-effect is defined
process evaluation, in the sense that the com- as a consequence outside the main target area(s)
plete implementation process from formal of the intervention. Side-effects can be antici-
intervention adoption to eventual addressee pated and unanticipated, negative or positive.
participation in delivery is scrutinized for pos- Perverse effects run exactly counter to the pur-
sible problems. The point is to examine poses stated in the intervention. These impacts
whether, for instance, an intervention decision may occur in the target area(s) of the public
is carried out according to plan at lower levels intervention. Yet, they may also arise outside the
of authority in order to correct mistakes and target areas and be side-effects. Perverse effects
omissions during execution and delivery. may ensue far down in the purported chain of
Simple monitoring, as opposed to qualified control, i.e. in the second, third, or even fourth
monitoring, engenders data assembly only on outcome stage. In addition, they often crop up
some variable without marrying this activity to after many years. Perverse effects are different
interventions, value criteria or problem search. from null effects. Null effects means that inter-
True, data gathered through simple monitor- ventions generate no impacts at all inside or out-
ing may be used for qualified monitoring and side their targeted areas.
impact assessment. But, in itself, simple moni- The effects tree in Figure 23.3 shows which
toring is not evaluation. aspects of effects that might be studied in eval-
uation research.
The importance of noticing main effects,
IMPACT ASSESSMENT perverse effects, and null effects should be obvi-
ous to every evaluator. But why is paying atten-
tion to side-effects vital? Because by-products,
Impact assessment probes the issue of interven- whether detrimental or beneficial, are crucial
tion effects upon outcomes, whether immediate, factors in every inclusive judgment of the oper-
intermediate, or ultimate. Impact assessment ation of an intervention.
attempts to determine to what extent outcomes
are produced by the intervention or by some-
thing else operating besides the intervention. Designs for Impact Assessment
Impact assessment addresses the seventh
problem in the Eight Problems Approach to There is no widely recognized solution to the
Evaluation. Are the outcomes, inside or outside impact problem in evaluation. A number of dif-
the targeted areas, at least indirectly and ferent designs are available. Sometimes, it
to some extent triggered and shaped by the is maintained that the approaches can be
Peter-3383-Chapter-23.qxd 6/8/2006 8:26 PM Page 408
Effects
(Intended − Unintended)
(Anticipated − Unanticipated)
Figure 23.3 Main effects, side-effects, perverse effects, and null effects
Source: Adapted from Vedung 1997: 54.
Experiments with randomized controls: In a provisional tryout before the permanent intervention is adopted,
targets are randomly divided into an experimental group, to whom the intervention is administered, and a
control group—randomized controls—from whom the intervention is withheld (classic experiments).
Experiments with matched controls: Targets to whom a provisional tryout is given or who have been
exposed to the permanent intervention are compared to a theoretically equivalent group, created nonran-
domly through matching—matched controls—from which the intervention is withheld or which has been
exposed to other intervention(s) (quasi-experiments).
Generic controls: Effects of a provisional or permanent intervention among targets are compared with
established norms about typical changes occurring in the larger population not covered by the intervention.
Reflexive controls: Targets who receive or have received the provisional or permanent intervention are com-
pared to themselves, as measured before the intervention.
Statistical controls: Participant and nonparticipant targets of the provisional or permanent intervention are
compared, statistically holding constant differences between participants and people not covered by the
intervention.
Shadow controls: Targets who receive or have received the provisional or permanent intervention are com-
pared to the judgments of experts, program managers and staff, and of the targets themselves on what
changes they believe would have happened should there have been no intervention.
Explanatory process evaluation: The development of the permanent intervention in its natural setting is
traced step by step from decision to outcome to find out facilitating and constraining factors.
Note. A permanent intervention is a “real” intervention by contrast with a provisional tryout intervention.
I. INTERVENTION FORMATION
1. Direction of proposed change
2. Political support
3. Symbolic politics
4. Participation of affected interests
V. ADDRESSEE RESPONSE
1. Comprehension, capability, willingness
2. Formative moments
3. Zealots
4. Capture
5. Free riders
perspective—can be formulated in the follow- in national agencies and among other actors in
ing fashion: (1) In what ways and to which the implementation stage.
extent does the outcome stem from the Symbolic politics may influence implementa-
intervention? tion and results. Symbolic politics means that
Seven broad factors and mechanisms which the intervention is inaugurated for other pur-
may have impact upon outcomes of public poses than to attain substantive results, i.e. to
interventions are arranged in Figure 23.5. give an impression of being concerned without
Intervention formation. The direction of actually being so. Policy-makers may want to
proposed change may affect outcomes through satisfy party opinion or strengthen their own
addressee considerations or indirectly via con- party leadership, to keep party membership
siderations among implementing actors. in line, to secure votes in general elections, or
Should the envisaged change point in a direc- to facilitate government coalitions. Naturally,
tion different from the one public officials and agency officials, low-level operators, and other
recipients are used to, it will become more official actors may perceive the symbolic con-
difficult to carry out. tent and devote less energy to implementation
The outcomes will be influenced by the politi- than they otherwise would have done.
cal support for the intervention in connection Participation of affected interests in the
with its instigation. Partisan and pressure group preparatory work leading up to intervention
disagreements create uncertainty about the adoption may well have an influence on inter-
future of a program, which will probably trigger vention results. If those to be affected are
reflections and reasonings and affect decisions allowed to participate in the formation, the
Peter-3383-Chapter-23.qxd 6/8/2006 8:26 PM Page 411
intervention will acquire a legitimacy, which in Still another subcomponent of the larger
turn will facilitate implementation. implementation factor suggests that the compre-
Intervention design. Intervention obscurity hension, capabilities, and willingness of formal
makes it difficult for implementing officials intermediaries may positively or negatively affect
and addressees to form a correct image the outcomes. Formal intermediaries enter the
of what policies are designed to achieve. stage as middlemen between national adminis-
Indeterminate words make intervention lan- trative agencies and target groups.
guage ambiguous and vague. Goal catalogs More and more non-governmental interme-
consist of an array of goals with no priorities diaries are used to carry out public inter-
among them. In both cases, it becomes impos- vention objectives. One category is private
sible to lay bare the overall purpose of the business contractors and their interest organi-
intervention as originally conceived. This pro- zations. Another is public interest organiza-
vides enforcement officials with a broad lati- tions or voluntary local associations like
tude of discretion. home-owners’ associations, sports clubs, or
Decision-maker support after instigation. tenants associations.
Decision-makers may back the intervention Non-governmental intermediaries may be
after its inauguration, for instance through negotiating partners when agencies work out
repeated endorsements of the intervention, may norms under a regulatory statute. They may be
influence the results. represented on agency boards. They may act
Implementation. Implementation might as real implementors. They can participate in
impinge on results. Three problem levels are evaluations.
discerned: national agencies, formal interme- The use of non-governmental intermedi-
diaries, and street-level bureaucrats. aries may increase cost-effectiveness and avoid
Deficient administrator comprehension of the build-up of large bureaucracies. Non-
intervention content may cause execution governmental intermediaries may enhance the
problems. The comprehension ingredient is program’s legitimacy in the eyes of the recipi-
important at all three implementation levels. ents and the general public. Yet, intermediaries
For simplicity reasons, I shall mention it only may also create drawbacks. The intervention
at the national agency level. may be diverted in directions not conforming
Administrative capability may modify results. with its stated original intentions.
If intervention execution requires appropria- Also street-level bureaucrats, “public service
tions, personnel, talent, time, energy, and workers who interact directly with citizens in
equipment unavailable to administrators, the the course of their jobs”, may influence out-
probability of achieving successful results will comes through the multitude of decisions they
be weakened. make in interacting with clients. They con-
Outcomes are affected by administrative sciously develop coping strategies like limiting
willingness. Public bureaucracies may have the information about their services, asking
agendas of their own, which may conflict with clients and inspectees to wait, making them-
the faithful implementation of the principal’s selves unavailable to contacts, making ample
directives and recommendations. A much use of referrals of clients to other authorities or
debated hypothesis says that bureaucrats are concentrating attention to a limited number of
budget-maximizers. To enhance their personal easy, well-defined cases.
reputation, salary, happiness, comfort, field of Addressee response. Addressee comprehen-
influence, and other self-serving resources, sion, capability, willingness, organizational
bureaucrats tend to expand their budgets. belongingness and general predicament may
A milder variation on the administrative- operate on outcomes. If clients neither know
willingness theme is that civil servants may nor understand, the intervention cannot work.
harbor doubts about the appropriateness of A grants program unknown to prospective
the intervention itself, which may hamper applicants will not contribute to the results it
their preparedness to act. was instituted to achieve.
Peter-3383-Chapter-23.qxd 6/8/2006 8:26 PM Page 412
pertinent and of higher quality through efforts Responsiveness to user worries is important.
directed at the evaluation process. The responsive evaluator should care for
Evaluators should conduct self-evaluation the questions of potential users, not the
of her own evaluation, preferably the final questions of academics. Preferably, evaluators
report, before it is left to the commissioners. and the likely recipients should frame the
Self-evaluation amounts to quality control. questions together before they are left to the
Another stratagem is to encourage critical evaluators for investigation. Alternatively,
commentaries from enlightened but dispas- recipients and evaluators should work
sionate scholars and evaluator colleagues. together throughout the entire assessment
Peter-3383-Chapter-23.qxd 6/8/2006 8:26 PM Page 415
MEANS Collection, (1999) Evaluation of Socio- Pawson, Ray and Nick Tilley (1997) Realistic
economic programmes. Luxembourg: Office Evaluation, London: Sage Publications.
for Official Publications of the European Commu- Rossi, Peter H., Mark W. Lipsey and Howard E.
nities. ISBN: 92-828-6626-2. 6 volumes. 1. Freeman (2004) Evaluation: A Systematic
Evaluation design and management, 2. Selection Approach, Thousand Oaks, CA: Sage, 7th ed.
and use of indicators for monitoring and evalua- Stern, Elliot (2005) ed., Evaluation Research
tion, 3. Principal evaluation techniques and tools, Methods, vol. I-IV, London: Sage Publications,
4. Technical solutions for evaluation with partner- Sage Benchmarks in Social Research Methods.
ship, 5. Transversal evaluation of impacts on the Vedung, Evert (2001) Public Policy and Program
environment, employment and other intervention Evaluation, New Brunswick, New Jersey and
priorities, 6. Glossary of 300 concepts and terms. London: Transaction Publishers.
Peter-3383-Chapter-24.qxd 6/8/2006 7:42 PM Page 417
24
Efficiency and Cost-Benefit Analysis
A I D A N R . V I N I N G A N D D AV I D L . W E I M E R
Which public policies promote the good Certainly, there is increasingly widespread
society? This central question of political phi- acceptance of CBA and related methods by
losophy must also be routinely confronted by national and sub-national governments in a
practicing policy analysts. Comprehensive broad range of policy areas, especially regula-
policy analysis requires the assertion and justi- tion. Hahn and Dudley (2004: 1) observe that:
fication of goals that provide a basis for “Countries and states throughout the world
systematically comparing policy alternatives. are requiring extensive use of cost-benefit
Almost always, efficiency is among the relevant analysis and related tools as a way of informing
policy goals. It is the primary concern of key regulatory decisions and reforming the
welfare (normative) economics. Indeed, some regulatory process.” Many countries with
economists even regard efficiency as the only primarily public health systems, such as
appropriate goal for comparing alternatives in Australia, the United Kingdom, and Canada, are
many substantive policy areas. attempting to use CBA and cost-effectiveness
Cost-benefit analysis (CBA) is the primary analysis to make important health resource
method for comparing alternatives in terms allocation decisions (Drummond, 2004).
of their efficiency impacts. CBA comprises the Many countries also require that transporta-
concepts, conventions, and techniques for tion infrastructure projects be subjected to
assessing efficiency in practice. This essay detailed CBA.
reviews the evolution of thinking about the Some observers object to CBA on the
meaning of efficiency and the use of CBA as a grounds that limited information and exper-
tool for selecting policies. As Joe Grice notes in tise make it impractical. Yet, even when per-
the preface to the Green Book, UK Treasury’s formed well, critics of CBA oppose its use on
guide to best practice in policy analysis (HM grounds that it is fundamentally flawed. Some
Treasury, 2003: v): critics view placing monetary values on goods,
such as lives saved, as inherently inappropriate
Appraisal, done properly, is not rocket science, but it is
(Kelman, 1981; Ackerman and Heinzerling,
crucially important and needs to be carried out care-
fully. Decisions taken at the appraisal stage affect the 2002). Other critics view its exclusive focus on
whole lifecycle of new policies, programmes and efficiency as rendering it inappropriate as a
projects. Similarly, the proper evaluation of previous decision rule in most policy contexts.
Peter-3383-Chapter-24.qxd 6/8/2006 7:42 PM Page 418
A definition of efficiency is a necessary starting CBA has steadily evolved since its first use in
point for understanding CBA. After briefly the evaluation of water projects during the last
describing the neoclassical perspective on effi- century. Two developments are particularly
ciency that underlies CBA, we consider three important. First, the increasing demand for
lines of criticism. First, the social choice litera- methods to value public goods, such as environ-
ture raises technical issues about the limitations mental quality, has resulted in the growing use
of various, indeed any, measures of efficiency. (and sophistication) of contingent valuation, a
Second, the transaction cost literature chal- method for eliciting stated preferences. Second,
lenges the conceptual basis for efficiency- the increasingly common imposition of require-
enhancing policies and the measurement of ments that certain classes of policies be subjected
opportunity costs. Third, the behavioral eco- to CBA raises a number of issues. These include
nomics literature raises questions about how questions about the circumstances in which the
well individual choice corresponds to the use of analytical resources for CBA is appropri-
axioms of rationality, and therefore about the ate. It also includes questions about the ability of
appropriateness of traditional neoclassical mea- government agencies to conduct sound CBA in
sures of efficiency, especially as they apply to the inherently political contexts.
treatment of decisions under risk and over time.
The discussion of efficiency sets the stage for
PERSPECTIVES ON EFFICIENCY
an overview of CBA. The distinction between
stated and revealed preferences in the measure-
ment of benefits and costs is fundamental to What does “efficiency” mean? It is important
understanding CBA in both theory and prac- to first distinguish between technical and allo-
tice. Measurement based on stated preferences cational efficiency. First, consider allocational
involves eliciting individuals’ willingness to pay efficiency. Cte in Figure 24.1 shows the total
for the states of the world that they believe cost of producing various quantities of a good
would be produced by alternative policies. It is given technical efficiency (see below). B shows
attractive because it avoids many of the concep- the total benefits accruing to society at each
tual difficulties posed by risk acceptance, time level of output. Allocational efficiency is
preference, and altruism; it is challenging straightforward: it results when the output
because it forces analysts to conduct surveys of level that maximizes the difference between
samples of respondents whose responses can total benefits and total costs, Qte, is selected.
reasonably be used to make inferences about the Second, consider technical efficiency. Cte
preferences of the population. Measurement shows the total cost of producing various quan-
of revealed preferences, in contrast, requires tities of a good assuming technical efficiency.
analysts to make many judgments about the Technical efficiency requires that the total cost
appropriate way to treat risk and time from of the resources employed in producing the
the social perspective as well as deal with all the good is the lowest feasible amount at every level
challenges of inferential social science. of output. If output is not produced using the
Yet, even when efficiency impacts can be least-cost combination of inputs, then a techni-
reliably monetized, the issue remains as to cally inefficient total cost curve, such as Cti,
the appropriateness of CBA as a decision rule. results. Maximizing the difference between total
Multiple goals and the problem of immoral benefits and total costs with this technical inef-
preferences pose challenges to the use of CBA ficiency would result in output level Qti. This
for selecting policies. Nonetheless, taking a output level is both allocationally and techni-
realistic view of the functioning of representa- cally inefficient. It is allocationally inefficient
tive democracy, and the role that CBA plays in because it results in too little output (Qti rather
that process, we argue that CBA can have, and than Qqe). It is technically inefficient because the
does have, a constructive role in informing the output that is produced costs more than the
policy-making process. minimal feasible amount (Ci rather than Ce).
Peter-3383-Chapter-24.qxd 6/8/2006 7:42 PM Page 419
Cti
Total Benefits
and
Total Costs
B
Cte
Ci
Ce
social choice perspective, centers around how Hicks (1940), it basically asks if a change
well the neoclassical measures of willingness to would generate sufficient net gains so that it
pay appropriately measure benefits (curve B in would be possible, with costless redistribution,
Figure 24.1). Another theoretical controversy, to achieve a Pareto improvement.
arising from the transaction cost perspective, The most common justification of the use of
concerns the interpretation of opportunity potential Pareto efficiency assumes that its con-
costs (curve Cte in Figure 24.1). A practical sistent application in the selection of policies will
controversy, arising from behavioral econom- not only produce the greatest total net benefits
ics, concerns the extent to which individuals’ but also make everyone better off in the long run
apparent deviations form utility maximization because different policies will tend to have dif-
affect the prediction and interpretation of their ferent winners and losers. Of course, this need
willingness to pay. not be the case in practice. Further, the theoreti-
cal reference of Pareto efficiency can itself be
challenged in that it depends on the initial
Neoclassical Perspective endowments of the members of society prior to
the consideration of policies. Different initial
Traditional welfare economics has its founda- endowments can result in different sets of Pareto
tions in the notion of the ideal competitive efficient moves from the status quo.
economy. Under certain assumptions about The dependence of Pareto efficiency on
individual preferences and technology, the initial endowments has led some to reject it in
model of the economy has an equilibrium favor of an interpretation of efficiency as max-
(Arrow and Debreu, 1954; for an overview, see imizing some social welfare function over the
Weintraub, 1983). Defining efficiency in the distribution of goods after the adoption of
sense of Pareto, specifically, that an allocation policies (Bergson, 1938). This approach, how-
is efficient if it is impossible to make someone ever, places a political and conceptual burden
better off without making anyone else worse on the analyst to choose a specific social wel-
off, the fundamental theorem of welfare eco- fare function that appropriately incorporates
nomics is that the equilibrium of the ideal distributional values. From a practical per-
competitive economy is efficient. Market fail- spective, social welfare functions are often
ures result from violations of the assumptions unworkable because they typically require
so that the resulting equilibrium is not Pareto extensive information, well beyond that pro-
efficient. For example, the market failure of vided by markets that often provide the basis
pure public goods occurs when the assump- for applying potential Pareto efficiency.
tions of the ideal competitive model that goods The guiding principle for measuring net
are rivalrous in consumption and exclusive in benefits within the framework of potential
ownership do not hold. Pareto efficiency is willingness to pay: how
The existence of a market failure indicates much money would someone be willing to
the possibility that a public intervention to pay to obtain the impacts of a policy? The
force a reallocation could offer sufficient gains resources required to implement a policy are
to permit compensations to be made so that a valued in terms of their opportunity costs, the
Pareto improvement, a change in allocation willingness to pay for what the resources
that makes at least someone better off without would have produced in their next best uses.
making anyone else worse off, could be Willingness to pay is a “money metric” for
achieved. In practice, however, it is almost persons’ utility changes from policies that
never feasible for government to discover and can be aggregated across persons to obtain the
execute the compensations required to make overall net benefits of a policy. If the net bene-
policies Pareto improving. Consequently, fits are positive, then the policy offers a poten-
potential Pareto efficiency is the actual guide tial Pareto improvement. Efficiency calls for
for applied welfare economics. Formalized in choosing the set of feasible policies offering the
slightly different ways by Kaldor (1939) and largest net benefits.
Peter-3383-Chapter-24.qxd 6/8/2006 7:42 PM Page 421
market failures, transaction cost critics argue resources required to implement and maintain
that such market failures do not necessarily the policies.
imply inefficiency because, to avoid the “Nirvana The problem for those advocating the trans-
Fallacy,” the institutional arrangements that action cost approach is how to treat the costs of
produce the market failure must be compared reaching political decisions to adopt policies.
to feasible alternative arrangements rather If one treats the costs of collective decision-
than to the ideal (Demsetz, 1969). making in comparing alternatives, then one
Some critics challenge the whole notion of risks sliding into the tautology that “what is,
market failures as indicating inefficiency, argu- must be efficient”—if gains from trade were
ing that transaction costs are ubiquitous larger than the collective action costs of
because property rights are never perfectly achieving them, then these gains would have
complete and effective (De Alessi, 1983; Zerbe been realized. In attempting to assess these
and McCurdy, 1999). This line of argument political costs, one would ultimately encounter
can become tautological: existing institutions what has been called the “determinacy para-
are efficient because any possible gains from dox”: political economy models that fully
trade would be realized in the absence of trans- endogenize politics would leave no room for
action costs. There is also a tendency not to giving normative advice about efficiency or
recognize the essential role of government in other values (O’Flaherty and Bhagwati, 1997).
shaping property rights. For example, the pri- Consequently, strong advocates of the transac-
vate provision of lighthouses in England, cited tion cost approach, who do not want to aban-
by Zerbe and McCurdy (1999) in their attack don analysis altogether, retreat to essentially
on the notion of market failures, required sub- the neoclassical approach of ignoring decision-
stantial government involvement in support- making costs (Zerbe, 2001: 17–18).
ing the collection of fees (Van Zandi, 1993).
Others argue that Pareto efficiency is too ambi-
tious a criterion in a world of transaction costs Behavioral Perspective
and that a simpler net profit test (which one
may interpret as simply a net benefits test for Individual rationality is an axiom of neoclassical
public policy purposes) is more realistic and economics. Individuals are assumed to have
therefore more appropriate (Furubotn, 1999). coherent preferences characterized by transitivity
The transaction costs approach seems most (if alternative A is preferred to alternative B, and
useful in comparing alternative property rights B is preferred to alternative C, then A is preferred
arrangements, or what Williamson (1985: 2) to C) and completeness (all alternatives can be
calls the “comparative costs of planning, ordered), and they act as if they are seeking to
adapting, and monitoring task completion obtain more over less preferred outcomes.
under alternative governance structures.” An Conceptually, preferences can be very broad,
empirical literature has had some success in including not only individuals’ assessment of
applying the theory to questions such as orga- impacts of alternatives on themselves (self-
nizational form (for a review, see Shelanski and regarding preferences), but also assessments of
Klein, 1995). To the extent that resources must the alternatives’ impacts on other persons (other-
be expended to establish and maintain prop- regarding preferences) or the ways in which
erty rights, they should be included in any effi- alternatives were selected (process-regarding
ciency assessment. The proper perspective is preferences). In practice, neoclassical economics
not just the comparison of the costs and bene- usually assumes that individuals are self-regarding
fits of alternative final states of the world, but for reasons of tractability. Nonetheless, it is con-
additionally the costs and benefits of moving ceptually possible to account for the more gen-
from the status quo to these alternatives. In eral preferences in CBA. A more conceptually
practical terms, this simply means that com- challenging problem arises when individuals
parison of policies should take account of all make choices that appear to be inconsistent with
their impacts, including any uses of scarce the rationality axiom.
Peter-3383-Chapter-24.qxd 6/8/2006 7:42 PM Page 423
Behavioral economics, which draws heavily apply them according to the expected utility
on the experimental findings and theoretical hypothesis. Prediction aside, whether or not
perspectives of psychology (Rabin, 1998), inte- these biases pose a problem for the assessment
grates research about how individuals actually of efficiency depends on the extent to which
make choices into models of economic behav- we believe people would make fully rational
ior. Among the important topics addressed by choices if they were fully informed.
behavioral economics are complexities in the
structure of preferences and judgment biases
arising in situations involving uncertainty.
COST-BENEFIT ANALYSIS
The various complexities in the structure of
preferences that have been identified do not
inherently violate the axiom of rationality, but CBA provides the science and craft for ranking
do so if simpler utility structures are assumed. alternatives in terms of neoclassical efficiency.
For example, there is considerable evidence When efficiency is the only goal relevant to the
people value goods more once they possess policy problem being addressed, CBA provides
them—in other words, they experience an a conceptually clear decision rule: choose the set
“endowment effect” (Kahneman, Knetsch, and of alternatives offering the largest present value
Thaler, 1990). Though it is possible to posit util- of net benefits. It also provides a methodology
ity functions that are consistent with the for assessing efficiency in contexts where effi-
endowment effect, they would differ substan- ciency is one of the relevant goals.
tially from those commonly assumed in welfare A full exposition of CBA requires book-
economics. In applied work, the endowment length treatment (see, e.g., Boardman et al.,
effect usually manifests itself as a large diver- 2001; Dinwiddy and Teal, 1996; Zerbe and
gence between the willingness of individuals to Dively, 1994; Gramlich, 1990). Briefly, the basic
pay for marginal expansions of a public good steps are as follows: First, identify the set of
and their willingness to accept marginal reduc- alternative policies, including the status quo
tions of the same magnitude. In experimental policy, to be assessed (for convenience, alterna-
settings, the divergence between willingness to tives are often assessed relative to the status
pay and willingness to accept declines with quo policy so that common effects can be
experience, but does not disappear when the netted out). Second, decide who has standing
goods being considered do not have close sub- (Whittington and MacRae, 1991). This involves
stitutes (Shogren et al., 1994). Another com- choosing a jurisdictional scope (usually
plexity of preference structures is behavior that national because of the shared economy and
exhibits time inconsistency. For example, indi- constitution), but it may also involve deciding
viduals may choose one option if payoffs are to who within the jurisdiction (citizens, legal res-
be immediate, but another if the payoffs occur idents, illegal aliens) has their costs and bene-
further in the future. Such choices are inconsis- fits counted or whether all preferences (such as
tent with the conventional assumption that those for illegal activities) are counted. Third,
individuals discount at a constant marginal rate comprehensively identify the impacts of alter-
of time preference, although they may be con- natives, select units of measure, and project the
sistent with a discount rate that declines over impacts into the future. In the case of invest-
time (Laibson, 1997). ment projects, the future is determined by the
Cognitive psychologists have identified a life of the longest alternative considered.
variety of biases that result from heuristics that Fourth, monetize all impacts. That is, put all
people commonly employ in situations involv- the impacts into a common metric. Doing so
ing uncertainty (Kahneman, Slovic, and will often require the use of shadow prices for
Tversky, 1982). These biases clearly pose a goods not traded in efficient markets, such as
problem for positive economics based on the statistical lives saved or commuting time
assumption that individuals rationally use avoided. Fifth, discount costs and benefits to
information to estimate probabilities and take account of the preferences of people
Peter-3383-Chapter-24.qxd 6/8/2006 7:42 PM Page 424
to consume sooner rather than later. Sixth, are other-regarding and process-regarding as
algebraically sum the discounted costs and well as private-regarding preferences for
benefits to obtain a present value of net bene- public goods that leave no behavioral traces to
fits, which provides the basis for ranking poli- support observational valuation. It would thus
cies in terms of their efficiency. Seventh, taking provide a much broader basis for assessing
account of uncertainties in projecting and efficiency than is possible in traditional
monetizing effects, select the policy with the revealed preference approaches.
largest present value of net benefits.
The discussion that follows highlights a few
of the important conceptual and practical Revealed Preferences: Expected
issues that arise in applying CBA. Surpluses
willingness of the poor to pay for such relationship between efficiency and equity as
reductions). In narrow applications, such as a trade-off. In the presence of market failures,
the comparison of policies with differential many redistributive policies may be efficiency-
effects on different age groups, there is consid- enhancing (Aghion, Caroli and Garcia-
erable controversy over whether group-specific Penalosa, 1999; Rogers, 2003). Thus, these
statistical values of life should be used (Aldy policies involve no trade-off. The trade-off is
and Viscusi, 2004). likely to arise largely from incentive effects:
A fundamental problem for the revealed pref- equity requires redistribution, which necessar-
erence approach arises when impacts do not ily transfers wealth from the more to the less
have behavioral traces that can be used to esti- wealthy. This reduces work incentives and
mate shadow prices. These impacts, called pas- aggregate welfare. However, the size, and there-
sive use benefits, are especially important in fore the policy significance, of this disincentive
environmental applications of CBA (Vining is unclear; it may be small in aggregate
and Weimer, 1998). One important category is (Slemrod, 2000). Additionally, the relationship
option value, the amount people would be will- between inequality and economic growth (at
ing to pay to have the opportunity to consume least a partial proxy for dynamic efficiency, the
some good in the future if specific contingencies improvement of productive capacity over
arise (Weisbrod, 1964). Assumptions about the time) is complex, suggesting that short run
shape of individual preferences can sometimes evidence on the existence of the trade-off may
be used to estimate option value (Larson and be deceptive. For example, “inequality can be
Flacco, 1992), which has the technical definition associated with crime, riots, disputes and alike,
of the difference between option price and which reduce productive activities and raise
expected surplus. Another important category uncertainty” (Rogers, 2003: 119).
of passive use benefit is existence value, the will- Weimer and Vining (1999) argue that if any
ingness of people to pay for public goods, such goal other than efficiency is relevant, then the
as environmental quality, beyond any direct policy problem always becomes one of multi-
consumption value the goods provide (Krutilla, ple goals. Take the case of equity. Even if one
1967). solely or mostly cared about redistributing
resources to poorer members of society, few
would argue that this should be done with
Challenges to CBA as a Decision Rule
Okun’s “leaky buckets” (that is, inefficiently), if
it can be avoided. Maximizing allocative effi-
Conceptual and practical arguments can be
ciency (subject to the redistributive constraint)
raised against the use of CBA as a decision rule:
maximizes the resources available for con-
goals beyond efficiency are relevant in the par-
sumption or redistribution; therefore, it
ticular policy context; not all efficiency impacts
potentially facilitates redistribution. The same
can be appropriately monetized; and CBA is
logic applies to any other goal.
inconsistent with democratic principles.
Moving to multi-goal analysis, however,
means that CBA is not an appropriate decision
Goals Other Than Efficiency rule. The argument that efficiency should also
generally remain a goal, however, means
CBA ranks alternatives in terms of efficiency. that analysts still face the task of assessing the
In many policy contexts, however, other goals relative efficiency of alternatives. The tech-
are clearly important (Okun, 1975; Myers, niques of CBA provide the science and craft for
2002). Nussbaum (2000), for example, argues doing so. In the special case when efficiency
that preservation of basic social entitlements and one other goal are relevant, and the
often should be legitimately viewed as a impacts of policy alternatives on the other goal
separate goal. Okun (1975) argues that the effi- can be quantified, analysts may resort to cost-
ciency-equity trade-off is the “big one.” Yet effectiveness analysis, ranking alternatives in
several caveats deserve note in describing the terms of how much of the non-efficiency
Peter-3383-Chapter-24.qxd 6/8/2006 7:42 PM Page 426
impact can be achieved per dollar of efficiency to be unwilling to monetize the value of
loss, or how many dollars of efficiency loss it statistical lives. Instead, change in either life-
takes to produce a unit of non-efficiency years or quality-adjusted life years (QALYs) is
impact (Gold et al., 1996). typically treated as a quantitative but non-
In some applications, it may be possible to monetized impact in cost-effectiveness analy-
disaggregate costs and benefits by groups so that sis. Considerable controversy still surrounds
it would be possible to apply distributional the measurement of quality of life (Blomqvist,
weights. That is, a dollar of costs or benefits 2002), the extent to which cost-effectiveness
accruing to members of some groups, such as rankings are consistent with welfare theory
those with lower levels of wealth or income, (Dolan and Edlin, 2002), and the extent to
would be counted as more than a dollar. The which quality of life measures should reflect
weights might be based on assumptions about the valuations of the public as well as those of
declining utility of income (Feldstein, 1972). As directly affected groups such as patients (Ubel
consensus on appropriate weights is unlikely, et al., 2000).
Harberger (1978) recommends that analysts
present standard CBA along with weighted-
CBA in Political Context
CBA to make clear the implications of the
assumed weights on policy rankings.
Is CBA a threat to democracy? Faced with the
choice of a world in which one was forced to
Non-Monetized Impacts choose either vibrant democracy or well-
applied CBA as a basis for collective choice, we
CBA requires that all efficiency impacts be doubt that even many economists would
monetized. Monetization can be stymied in choose CBA.
several ways: impacts cannot be reasonably A more realistic question is whether society
quantified, appropriate shadow prices cannot would be better off using CBA as a decision
be found for quantified impacts, or analysts rule in those areas of policy where efficiency is
may not be willing to apply available shadow legitimately the dominant goal and the full
prices. Any failure to monetize prevents the use range of impacts can reasonably be monetized.
of CBA as a decision rule. Additionally, would society be better off using
When an impact cannot be quantified or CBA to compare redistributive policies? The
monetized, analysts may still find it useful to first point to note in addressing these ques-
estimate the net benefits of all other impacts. tions is to recognize that, although CBA has
The sign of net benefits by itself does not indi- increased in visibility and influence, rarely
cate whether an alternative offers potential does it serve as a decision rule. Indeed, it is
Pareto improvement over the status quo. easier to find laments about its lack of ultimate
Instead, net benefits must be compared to the influence (Anis and Gagnon, 2000; Jan, 2003;
likely magnitude of the net benefits of the Drummond, 2004) than to document cases
excluded impact. For example, estimates of where it has been determinative.
the net benefits of public sports stadiums usu- We believe that the nature of representative
ally do not explicitly include the existence and democracy suggests that greater use of CBA
option values some residents might have for the would be beneficial. This argument rests on the
facilities. If the net benefits are negative, how- observation that a major weakness of represen-
ever, one could ask how large these benefits tative democracy is that it tends to favor con-
would have to be to make the overall net bene- centrated and organized interests over diffuse
fits positive. The answer for Camden Yards, for and unorganized interests. As a consequence,
instance, would be approximately $11 million many implemented policies are inefficient and
per year (Hamilton and Kahn, 1997). inequitable. Because it aggregates costs and ben-
It is quite common in the areas of health and efits over the entire population, the information
medicine for analysts and decision-makers provided by CBA often provides a voice for
Peter-3383-Chapter-24.qxd 6/8/2006 7:42 PM Page 427
diffuse interests (Vining and Weimer, 1992). In further development (Arrow et al., 1993).
this view, the information provided by CBA Almost every issue of the Journal of Environ-
helps compensate for an inherent structural mental Economics and Management and Land
weakness of representative democracy. Indeed, Economics, the two leading environmental eco-
one can conceptualize recent attempts by gov- nomics journals, carries articles dealing with
ernments to institutionalize CBA-like analysis methodological issues. CV studies are being
as an attempt to credibly commit themselves to widely applied to evaluate health policies
efficiency-enhancing policies (this has the (Diener et al., 1998); increasingly other applica-
added advantage of discouraging future rent- tions, such as the measurement of the existence
seeking by interest groups, as well as encourag- value of sports stadiums (Johnson and
ing efficient policies). All this is very worthy, but Whitehead, 2000) and the valuation of
no easier than other difficult self-commitment Aboriginal artifacts (Boxall et al., 2000).
exercises, such as quitting smoking. It is espe- Aside from all the general problems of sur-
cially difficult for Parliamentary systems that vey methodology, CV surveys must ensure that
have few mechanisms for credibly making long- respondents understand the good being valued
term commitments. and that they are asked questions that elicit
truthful (non-strategic) responses. If problems
of representative sampling can be overcome,
Technical Frontier: Contingent Valuation
internet-based surveys offer the potential
for providing more and different information
Economists continue to strive to find ways of
to respondents to expand the range of public
making inferences about willingness to pay
goods they can adequately understand
from data that can be observed, such as the
(Berrens et al., 2004). The theory of mecha-
travel costs people bear to visit recreational sites
nism design provides a basis for assessing the
(Clawson and Knetsch, 1966) or the capitaliza-
risks of strategic responses (Carson et al.,
tion of environmental amenities into property
1999). In particular, only the valuation of
values (Smith and Huang, 1995). Such efforts
public goods through referendum-type ques-
have moved the boundary line between use and
tions with dichotomous responses eliminate
passive use benefits. Simultaneously, econo-
incentives for strategic answers. At the same
mists have worked to improve methods for elic-
time, considerable progress has been made in
iting option prices for passive use benefits
linking this elicitation format to utility theory
through surveys that ask respondents to value
(Hanemann and Kanninen, 2000).
hypothetical, or contingent, goods (for a com-
prehensive review, see Bateman and Willis,
2000). Over the last twenty years elicitation of Policy Frontier: CBA as an
willingness to pay through contingent valua- Intra-Governmental Accountability
tion (CV) surveys has become more sophisti- Mechanism
cated, expanding the range of policies that can
plausibly be subjected to CBA. How useful are CBA and related economic eval-
An important factor in the methodological uation methods as an accountability mecha-
development of CV surveys is their use in envi- nism? Can these methods be used more
ronmental damage assessment. With U.S. courts effectively to achieve good policies? In the
giving CV surveys the status of “rebuttable pre- United States, the use of CBA for informing
sumption,” plaintiffs and defendants have had decisions about public investments can be
an incentive to invest in developing better and traced back to the River and Harbor Act of 1902
more defensible methods (Kopp et al., 1990). (Steiner, 1974). Since then, CBA has been widely
Although CV surveys remain controversial, adopted by national governments, state and
endorsement by a blue ribbon panel assembled provincial governments, as well as by interna-
by the National Oceanic and Atmospheric tional agencies as a preferred form of analysis.
Administration has encouraged their use and CBA in the United States is mandated within
Peter-3383-Chapter-24.qxd 6/8/2006 7:42 PM Page 428
the federal government across a wide spectrum and Waters (1993) provide models of such
of public policies. Both President Reagan’s behavior. There are two (not necessarily mutu-
Executive Order 12,291 and President Clinton’s ally exclusive) ways to produce biased analysis:
Executive Order 12,886 require federal agencies incorrect application of techniques or inaccu-
to prepare impact analyses for regulations that rate predictions.
are likely to result in significant economic Common incorrect applications of CBA
impacts. The Clinton Executive Order requires techniques include failure to discount or the
that analyses identify social costs and benefits use of inappropriate discount rates, selective
and attempt to determine if the predicted bene- inclusion or exclusion of cost or benefit cate-
fits of the proposed regulation justify the costs gories, double-counting of costs or benefits,
to society. Additionally, the Order “places and idiosyncratic definitions of standing,
greater emphasis on distributional concerns” among other problems. These kinds of errors,
than does previous Executive Orders (Hahn however, are generally observable provided the
et al., 2000; 860). Congress has included analysis is available in some form to external
requirements for CBA-like analysis in a variety review. Indeed, at least at the federal level,
of legislation, such as the Unfunded Mandates there is a growing trend for oversight agencies,
Reform Act of 1995. Even the federal courts in think tanks, and academics to identify and crit-
the United States are now utilizing a form of icize these analytic errors (GAO, 1998; Hahn
“cost-benefit balancing” where legislation autho- et al., 2000; Hahn and Dudley, 2004) and to
rizes them to do so (Hahn and Sunstein, 2002). publicize cases where agencies are attempting
Many states also require that regulatory initia- to proceed with projects or regulations with
tives include some form of cost-benefit analysis negative net benefits.
(Whisnant and Cherry, 1996; Zerbe, 1998). The problem of prediction bias is probably
Experience to date suggests that many federal more serious. Analysts have to predict ex ante
agencies have had difficulty performing CBA how many motorists will use a highway or how
well (and this is almost certainly generalizable to many illnesses will be avoided by a specific
other governments). Two explanations seem immunization program. It is impossible to show
important. First, in spite of the fact that it “ain’t that a given prediction is incorrect until the pro-
rocket science,” there are difficult technical ject is significantly under way or complete
issues, some of which we have already discussed. (Boardman, Mallery and Vining, 1994). Indeed,
These range from non-intuitive meanings of it is often even difficult for external reviewers to
costs and benefits (Boardman, Vining, and decide whether projections are plausible without
Waters, 1993) to the difficulty of appropriate access to detailed agency data. Nonetheless, there
discount rates for intergenerational projects is considerable ex post evidence that agencies do
(Moore et al., 2004) and the problems of mea- underestimate costs or overestimate benefits in
suring costs and benefits that have few behav- ways that inflate projected benefits of projects
ioral traces. As a consequence, good analysis (e.g., Kain, 1990; Flyvbjerg, Holm, and Buhl,
takes scarce analytic capacity and therefore can 2002; Seong and Mendeloff, 2003).
involve large opportunity costs. Such capacity, In view of all these barriers to effective use,
both general and specific to particular policy what is the prognosis for CBA? Clearly, the
areas, can take time to develop and often suffers efforts of scholars such as Robert Hahn and
during periods of budget cutting. others to induce agencies to engage in sound
Second, and probably of more importance and comprehensive economic analysis is useful.
for the future of CBA within government, There is also some systematic evidence,
politicians, senior civil servants, or those trying although not from North America, that good
to influence them often do not like the recom- analysis leads to good projects (Deininger
mendations that result from high quality CBA, et al., 1998; Vawda et al., 2003). In any event,
for self-interested reasons. Agencies, if they unless agencies are required to publish in easily
cannot avoid the requirement for analysis, may accessible format their complete analyses,
provide self-serving analysis. Boardman, Vining, including supporting data and assumptions,
Peter-3383-Chapter-24.qxd 6/8/2006 7:42 PM Page 429
there will remain much room for both unin- Arrow, Kenneth J. and Gerard Debreu (1954)
tentional error and strategic behavior. For “Existence of an Equilibrium for a Competitive
example, regulatory dockets, including agency Economy,” Econometrica, 22(3): 265–290.
CBAs and supporting materials, could be made Arrow, Kenneth J., Robert, Solow, Paul Portney,
Edward Leamer, Roy Radner and Howard
available on the Internet—for a detailed pro-
Schuman (1993) “Report of the NOAA Panel on
posal along these lines in the context of analy-
Contingent Valuation,” Federal Register, 58(10):
sis of infrastructure projects in Canada, see 4601–4614.
Vining and Weimer (2001). Bateman, Ian J. and Ken G. Willis (eds.) (2000)
Valuing Environmental Preferences: Theory and
Practice of Contingent Valuation Methods in the
CONCLUSION U.S., EC, and Developing Countries. Oxford, UK:
Oxford University Press.
Bergson [as Berk], Abram (1938) “A Reformulation
Economists generally argue that efficiency of Certain Aspects of Welfare Economics,”
should be considered an important goal of Quarterly Journal of Economics, 52(2): 310–334.
public policy. The proper measurement and Berrens, Robert P., Alok K. Bohara, Hank C. Jenkins-
interpretation of efficiency are not without Smith and Carol L. Silva (2004) “Information and
controversy. Nonetheless, economists continue Effort in Contingent Valuation Surveys: Applica-
to refine the science and improve the craft tion to Global Climate Change Using National
of CBA and related techniques for assessing Internet Samples,” Journal of Environmental
the relative efficiency of public policies. Economics and Management, 47(2): 331–363.
Blackorby, Charles and David Donaldson (1985)
Governments are increasingly demanding
“Consumers’ Surpluses and Consistent Cost-
CBA be conducted by their agencies for impor-
Benefit Tests,” Social Choice and Welfare, 1(4):
tant decisions. As CBA techniques continue to 251–262.
develop, the plausible range of application of Blackorby, Charles and David Donaldson (1990)
CBA will almost certainly increase. “A Review Article: The Case Against the Use of
the Sum of Compensating Variation in Cost-
Benefit Analysis,” Canadian Journal of Economics,
REFERENCES 23(3): 471–494.
Blomqvist, Ake (2002) “QALYs, Standard Gambles,
and the Expected Budget Constraint,” Journal of
Ackerman, Frank and Lisa Heinzerling (2002) Health Economics, 21(2): 181–195.
“Pricing the Priceless: Cost-Benefit Analysis Boardman, Anthony E., David H. Greenberg, Aidan
of Environmental Protection,” University of R. Vining and David L, Weimer (2001) Cost-
Pennsylvania Law Review, 150(5): 1553–84. Benefit Analysis: Concepts and Practice 2nd, Upper
Aghion, Philippe, Eve Caroli and Cecilia Garcia- Saddle River, New Jersey: Prentice Hall.
Penalosa (1999) “Inequality and Economic Boardman, Anthony E., Wendy L. Mallery and
Growth: The Perspective of the New Growth Aidan R. Vining, (1994) “Learning from Ex
Theories,” Journal of Economic Literature, 37(4): Ante/Ex Post Cost-Benefit Comparisons: The
1615–1660. Coquihalla Highway Example,” Socio-Economic
Aldy, Joseph E. and W. Kip Viscusi (2004) “Age Planning Sciences, 28(2): 69–84.
Variation in Workers’ Value of Statistical Life,” Boardman, Anthony E., Aidan R. Vining and William
National Bureau of Economic Research, W10199: Waters (1993) “Costs and Benefits Through
January. Bureaucratic Lenses: Example of a Highway
Allen, Douglas W. (1991) “What Are Transaction Project,” Journal of Policy Analysis and Management,
Costs?” Research in Law and Economics, 14: 1–18. 12(3): 532–555.
Anis, Aslin and Yves Gagnon (2000) “Using Economic Boxall, Peter C., Jeffrey Englin and Wiktor L.
Evaluation to Make Formulary Coverage Decisions: Adamowicz (2000) “Valuing Aboriginal Artifacts:
So Much for Guidelines,” PharmacoEconomics, A Combined Revealed-Stated Preference
18(1): 55–62. Approach,” Journal of Environmental Economics
Arrow, Kenneth J. (1963) Social Choice and and Management, 45(2): 213–230.
Individual Values, 2nd Edition, New Haven: Yale Carson, Richard T., Theodore Groves and Mark J.
University Press. Machina (1999) “Incentive and Informational
Peter-3383-Chapter-24.qxd 6/8/2006 7:42 PM Page 430
Hicks, John R. (1940) The Valuation of Social Nussbaum, Martha C. (2000) “The Costs of
Income, Economica, 7(26): 105–124. Tragedy: Some Moral Limits of Cost-Benefit
HM Treasury (2003) The Green Book: Appraisal and Analysis,” Journal of Legal Studies, 29(2): 1005–36.
Evaluation in Central Government, London: TSO. O’Flaherty, Brendan and Jagdish Bhagwati (1997)
Jan, Stephen (2003) “Why Does Economic Analysis in “Will Free Trade With Political Science Put
Health Care Not Get Implemented More? Towards Normative Economists Out of Work?” Economics
a Greater Understanding of the Rules of the Game and Politics, 9(3): 207–219.
and the Costs of Decision Making,” Applied Health Okun, Arthur M. (1975) Equality and Efficiency: The
Economics and Health Policy, 2(1): 17–24. Big Tradeoff, Washington, DC: The Brookings
Johnson, Bruce K and John C. Whitehead (2000) Institution.
“Value of Public Goods from Sports Stadiums: Rabin, Matthew (1998) “Psychology and Economics,”
The CVM Approach,” Contemporary Economic Journal of Economic Literature, 36(1): 11–46.
Policy, 18(1): 48–58. Rogers, Mark (2003) “A Survey of Economic
Kahneman, Daniel, Jack L. Knetsch and Richard H. Growth,” The Economic Record, 79(244): 112–135.
Thaler (1990) “Experimental Tests of the Endow- Scitovsky, Tibor (1941) “A Note on Welfare
ment Effect and the Coase Theorem,” Journal of Propositions in Economics,” Review of Economic
Political Economy, 98(6): 1325–1348. Studies, 41(1): 77–88.
Kahneman, Daniel, Paul Slovic and Amos Tversky, Seong, Si Kyung and John Mendeloff (2003)
editors (1982) Judgment Under Uncertainty: “Assessing the Accuracy of OSHA’s Projections of
Heuristics and Biases, New York: Cambridge the Benefits of New Safety Standards,” Washington,
University Press. D.C: AEI-Brookings Joint Center for Regulatory
Kain, John F. (1990) “Deception in Dallas: Strategic Studies, 03–8, July.
Misrepresentation in Rail Transit Promotion and Shelanski, Howard A. and Peter G. Klein (1995)
Evaluation,” Journal of the American Planning “Empirical Evidence in Transaction Cost
Association, 56(2): 184–96. Economics: A Review and Assessment,” Journal of
Kaldor, Nicholas (1939) “Welfare Propositions of Law, Economics, and Organization, 11(2): 335–361.
Economics and Interpersonal Comparisons of Shogren, Jason F., Seung Y. Shin, Dermot J. Hayes
Utility,” Economic Journal, 49(195): 549–552. and James B. Kliebenstein (1994) “Resolving the
Kelman, Steven (1981) “Cost-Benefit: An Ethical Differences between Willingness to Pay and
Critique,” Regulation, January/February: 33–40. Willingness to Accept,” American Economic
Kopp, Raymond J., Paul R. Portney and V. Kerry Review, 84(1): 255–270.
Smith (1990) “The Economics of Natural Resource Slater, Gary and David A. Spencer (2000) “The
Damages after Ohio v. U.S. Department of the Uncertain Foundations of Transaction Costs
Interior,” Environmental Law Reporter, 20(4): Economics,” Journal of Economic Issues, 34(1):
10127–10131. 61–87.
Krutilla, John V. (1967) “Conservation Reconsidered,” Slemrod, Joel B. (2000) Does Atlas Shrug? The
American Economic Review, 57(4): 777–786. Economic Consequences of Taxing the Rich,
Laibson, David (1997) “Golden Eggs and Cambridge, Mass: Harvard University Press.
Hyperbolic Discounting,” Quarterly Journal of Smith, V. Kerry and Ju-Chin Huang (1995) “Can
Economics, 112(2): 443–478. Markets Value Air Quality? A Meta-Analysis of
Larson, Douglas M. and Paul R. Flacco (1992) Hedonic Property Value Models,” Journal of
“Measuring Option Prices from Market Political Economy, 103(1): 209–227.
Behavior,” Journal of Environmental Economics Steiner, Peter O. (1974) “Public Expenditure
and Management, 22(2): 178–198. Budgeting,” in Alan S. Blinder et al. (eds.), The
McKenzie, George W. (1983) Measuring Economic Economics of Public Finance, Washington, D.C.:
Welfare: New Methods, New York: Cambridge Brookings Institution: 241–357.
University Press. Ubel, Peter A., Jeff Richardson and Paul Menzel
Moore, Mark, Anthony E. Boardman, Aidan R. (2000) “Societal Value, The Person Trade-Off, and
Vining, David Weimer and David Greenberg the Dilemma of Whose Values to Measure
(2004) “Just Give Me a Number!,” Journal of Policy for Cost-Effectiveness,” Health Economics, 9(2):
Analysis and Management, 23(4): 789–812. 127–136.
Myers, Samuel L. (2002) “Presidential Address— Van Zandi, David E. (1993) “The Lessons of the
Analysis of Race as Policy Analysis,” Journal of Lighthouse: ‘Government’ or ‘Private’ Provision
Policy Analysis and Management, 21(2): 169–190. of Goods,” Journal of Legal Studies, 22(1): 47–72.
Peter-3383-Chapter-24.qxd 6/8/2006 7:42 PM Page 432
Vawda, Ayesha Yaqub., Peter Moock, J., Price Gittinger Weintraub, E. Roy (1983) “On the Existence of a
and Harry Anthony Patrinos (2003) “Economic Competitive Equilibrium: 1930–1954,” Journal of
Analysis of World Bank Education Projects and Economic Literature, 21(1): 1–39.
Project Outcomes,” International Journal of Weisbrod, Burton A. (1964) “Collective Consumption
Educational Development, 23(6): 645–660. Services of Individual Consumption Goods,”
Vining, Aidan R. and David L. Weimer (2001) Quarterly Journal of Economics, 78(3): 71–77.
“Criteria for Infrastructure Investment: Normative, Whisnant, Richard and Diane DeWitt Cherry
Positive, and Prudential Perspectives,” 131–161 in (1996), “Economic Analysis of Rules: Devolution,
Aidan R. Vining and John Richards (eds.), Building Evolution and Realism,” Wake Forest Law Review,
the Future: Issues in Public Infrastructure in Canada. 31(3): 693–743.
Toronto: C.D. Howe Institute. Whittington, Dale and Duncan MacRae, Jr., (1991)
Vining, Aidan R. and David L. Weimer (1998) “The Issue of Standing in Cost-Benefit Analysis,”
“Passive Use Benefits: Existence, Option, and Journal of Policy Analysis and Management, 10(1):
Quasi-Option Value,” in Fred Thompson and 96–105.
Mark Green, eds, Handbook of Public Finance, Williamson, Oliver E. (1985) The Economic
New York: Marcel Dekker, 319–345. Institutions of Capitalism, New York: The Free Press.
Vining, Aidan R. and David L. Weimer (1992) “Welfare Zerbe, Richard O., Jr., (2001) Economic Efficiency in
Economics as the Foundation for Public Policy: Law and Economics, Northampton, Mass: Edward
Incomplete and Flawed but Nevertheless Desirable,” Elgar.
Journal of Socio-Economics, 21(1): 25–37. Zerbe, Richard O. (1998) “Is Cost-Benefit Analysis
Viscusi, W. Kip (1993) “The Value of Risks to Life Legal? Three Rules,”Journal of Policy Analysis and
and Health,” Journal of Economic Literature, Management, 17(3): 419–456.
31(4): 1912–1946. Zerbe, Richard O., Jr., and Dwight D. Dively (1994)
Vitaliano, Donald F. (1997) “X-Inefficiency in the Benefit-Cost Analysis in Theory and Practice,
Public Sector: The Case of Libraries,” Public New York: Harper Collins.
Finance Review, 25(6): 629–643. Zerbe, Richard O., Jr., and Howard E. McCurdy
Weimer, David L. and Aidan R. Vining (1999) Policy (1999) “The Failure of Market Failure,” Journal of
Analysis: Concepts and Practice, 3rd Edition, Policy Analysis and Management, 18(4): 558–578.
Upper Saddle River: Prentice Hall.
Peter-3383-Chapter-25.qxd 6/8/2006 7:45 PM Page 433
25
Ethics and Public Policy
GARY BRYNER
Ethical issues permeate the design, implemen- expected to guide the conduct of policy analysis.
tation, and evaluation of many public policies, Policy analysts should conduct their profes-
particularly those policies that regulate some of sional activities with analytical integrity and
the most personal and intimate decisions indi- provide objective advice about policy options
viduals make, shape the distribution of essential and consequences and make predictions as
resources, determine national and personal accurately as possible, ensure the transparency
security, and influence the natural environment of their assumptions, and clarify value choices,
that makes life itself possible. We assume that but leave such choices and trade-offs to policy
such important decisions will be made rationally makers. However, policy analysts are also
rather than arbitrarily, with a careful weighing of responsible to their clients and expected to be
alternatives and selection of those that are most loyal to their clients and represent their inter-
beneficial to those affected. One critical chal- ests. Much like attorneys, policy analysts must
lenge for policy makers, policy analysts, and not misrepresent data, purposely omit impor-
others who care about how policy making deci- tant information, or otherwise provide mis-
sions are made, is how to integrate the analytic leading analyses, but also interpret their
tools and approaches used in analyzing policy analyses in ways that are most favorable to the
options with reasoning that is rooted in views interests of their clients. Analysts may also seek
about what actions are morally right and wrong. to use their analysis in ways that promote their
This chapter explores the intersection of policy conception of what kinds of public policies are
analysis and moral reasoning by examining how most desirable and advocate changes in which
cost-benefit analysis, the predominant method- they believe. Policy analysis is also expected to
ology of policy analysis, can help illuminate promote values such as consistency with con-
ethical issues underlying policy choices. stitutional standards and the rule of law. These
expectations may clash in a variety of ways as
the interests of clients may differ from those of
objective analysis or personal value commit-
THE PROBLEM: INTEGRATING COMPETING
ments. Since analysis almost always requires
METHODS OF POLICY ANALYSIS
judgment about what data to use, what models
and analytic methods to employ, what assump-
Public policy and ethics converge in a number tions to begin with, how to characterize uncer-
of ways. At one level, ethical considerations are tainties, and other choices, analysts cannot
Peter-3383-Chapter-25.qxd 6/8/2006 7:45 PM Page 434
escape the need to make ethical choices based sustainability, or some other distinction. They
on how they understand their role and what may review the results of a diverse set of analy-
expectations guide them (Weimer and Vining, ses, somehow weigh and balance the competing
1999, 47–53). conclusions, and then draw some conclusions.
At another level, the kind of reasoning that Or they may try to integrate these different
people use to assess ethical choices surrounding ways of assessing policy choices and designing
moral issues—choices about what they believe public policies into a more coherent, synthetic
to be right or wrong—may clash with analytic analysis. There are considerable challenges in
tools and methods used to evaluate policy seeking such an integration. Some decision
options. Cost-benefit analysis (CBA) is the makers may simply prefer to keep the analyses
most widely used analytic tool in policy analy- separate so that their strengths and advantages
sis. It appeals to common sense notions that the are preserved. An integration or synthesis may
benefits of actions should exceed their costs, make the analyses hopelessly complex. But
but its power lies in its claim that different analysis that does not incorporate some kind
factors can be translated into a common of assessment of benefit and costs is likely to be
metric, usually dollars, and then compared. significantly disadvantaged because of the
CBA is rooted in utilitarian, egalitarian values convenience, clarity, and preciseness (albeit
that policies should promote the greatest utility perhaps misleadingly) represented in CBA,
or happiness for the greatest number of people where values are quantified and comparable
possible. The value of a policy is the sum of the within one overall indicator. Quantification
utility it produces for each affected person. But and monetization of variables is often impre-
policies may also be expected to promote other cise and incomplete at best. Nevertheless, in an
values besides utilitarian ones. Policies may also economic system driven by markets, prices,
seek social justice and fairness, taking into and other economic indicators, a failure to
account the differing circumstances that people engage in the debate using the dominant terms
find themselves in, compensating for past of discourse risks having those values ignored.
actions that have been discriminatory, and cre-
ating opportunities for those whose choices and
chances have been limited in the past. Policies BENEFITS OF COST-BENEFIT ANALYSIS
may also be aimed at preserving inherent rights
such as personal autonomy and liberty that are
to be honored regardless of whether they con- In 1792, Benjamin Franklin, writing from
tribute to the aggregate happiness or utility. London, offered this explanation for how he
Protecting the rights of unpopular minorities determined what course of action to take when
may even diminish the overall utility of the confronted with difficult choices:
majority but nevertheless be compelled by the When those difficult cases occur, they are difficult,
chiefly, because while we have them under considera-
value of preserving individual rights. Policies
tion, all the reasons pro and con are not present to the
may seek to promote sustainability and ensure mind at the same time; but sometimes one set present
that future generations have at least the same themselves, and at other times another, the first being
options as the current one to meet their needs. out of sight. Hence the various purposes or inclinations
Policy analysis may assess options in terms of that alternatively prevail, and the uncertainty that per-
plexes us. To get over this, my way is to divide a half
the inherent rightness or wrongness of policy
sheet of paper by a line into two columns; writing over
choices or whether their consequences are the one Pro, and over the other Con. Then, during three
desirable or undesirable, however those terms or four days consideration, I put under the different
are defined. heads short hints of the different motives, that at time
What are decision makers to do when occur to me, for and against the measure. When I have
thus got them all together in one view, I endeavor to
confronted with the diversity of policy analy-
estimate their respective weights; and where I find two,
ses? They may simply choose to make deci- one on each side, that seem equal, I strike them both
sions based on one value—utilitarianism, out. If I find a reason pro equal to some two reasons
social justice, preservation of individual rights, cons, I strike out three. If I judge some two reasons con,
Peter-3383-Chapter-25.qxd 6/8/2006 7:45 PM Page 435
equal to some three reasons pro, I strike out five; and action and select those with the greatest ratio,
thus proceeding I find at length where the balance lies; or choose the option that produces the desired
and if, after a day or two of further consideration, noth-
ing new that is of importance occurs on either side, I
outcomes or benefits at the lowest cost.
come to a determination accordingly. And, though the Both Democratic and Republican policy
weight of reasons cannot be taken with the precision of makers in the United States, and a great many
algebraic quantities, yet when each is thus considered, public policy analysts and advocates, continue
separately and comparatively, and the whole lies before to have great faith in the power of cost benefit
me, I think I can judge better, and am less liable to take
a rash step, and in fact I have found great advantage
analysis to illuminate policy choices. Every pres-
from this kind of equation, in what may be called moral ident during the past 25 years has sought ways
or prudential algebra (Gramlich, 1998, 1). to require regulatory agencies to take greater
account of the economic impacts of their
More than two hundred years later decision actions through some form of cost-benefit
makers continue, like Franklin, to rely on a log- analysis. These efforts have been motivated by
ical delineation of the pros and cons of alter- a variety of concerns, such as reducing the cost
native courses of action. CBA allows decision of compliance with regulations in order to fos-
makers to rationally set priorities and allocate ter economic growth, and increasing the cost-
resources. Critics of government regulation, effectiveness of regulation so that benefits are
for example, argue that regulatory agencies fail maximized. Congress joined the effort in the
to address the most serious risks, and that mid-1990s, as members introduced a variety of
identifying the costs and benefits of alternative bills to mandate cost-benefit analysis in regula-
actions will help steer them towards the initia- tory agencies. In 1995, Congress required that
tives that promise to accomplish the most agencies identify the costs and benefits of major
good with the resources expended. CBA is also rules. It came close, but has not yet succeeded, in
championed as a way to avoid policies such as enacting the requirement that agencies only
overly stringent regulations that impose major issue regulations whose benefits exceed their
compliance costs but produce limited value costs. Federal Courts have joined the debate
and to ensure that only worthwhile projects are over how agencies are to make difficult deci-
undertaken. Since the costs and benefits of dif- sions. In one of the most far-reaching challenges
ferent policy choices vary significantly, CBA to regulatory agencies it has ever considered, the
can help ensure that risks are regulated and Supreme Court rejected an industry challenge
desirable actions encouraged in comparable, to the U.S. Environmental Protection Agency’s
fair ways. CBA is ultimately aimed at produc- national air quality standards for fine particles
ing efficient public policies—the kinds of deci- and ozone it issued in 1997, concluding that
sions that markets would produce if they were EPA use of cost-benefit analysis in formulating
fully functioning. regulations requires express congressional
Identifying costs and benefits is only the first approval.
step; cost benefit analysis involves quantifying The current debate in the United States over
and monetizing variables so they can be more energy policy provides examples of the latest in
easily compared. Concepts and methodologies a long line of disputes over how to think about
from economics are used to estimate the and assess public choices. Bush administration
market value of variables for which prices do officials have consistently argued that develop-
not exist, such as survey data that ask respon- ing new energy supplies and constructing new
dents to express their willingness to pay to power plants are the primary solutions to the
ensure the preservation of ecosystems and energy crisis, and that modern technologies
species. Risk assessment is used to estimate the allow expanded oil and gas production “with
probability and seriousness of risks that might minimal impact on our environment and
be avoided (and benefits produced) if preven- wildlife” (McAllister, 2001). Cost-benefit
tative action is taken. Cost-effectiveness analy- analysis is expected to provide the analysis to
sis permits decision makers to calculate the guide decisions about protecting natural
benefit-to-cost ratio of alternative courses of resources and promoting energy development.
Peter-3383-Chapter-25.qxd 6/8/2006 7:45 PM Page 436
President Bush, for example, said early in his and CBA is a critical tool in ensuring that
administration that at least some national policies promote efficiency.
monuments should be opened to drilling for • Cost-benefit’s bias against values that are
oil and gas: “there are parts of the monument not easily quantified need not be fatal. CBA
lands where we can explore without affecting can also lay out in qualitative terms the
the overall environment. It depends on the values to be compared; this systematic
cost-benefit ratio” (Soraghan, 2001). Critics identification of the costs and benefits of
reacted with dismay to the belief that preserva- alternatives can be a very useful decision
tion values could actually be reflected in a cost making tool and help illuminate the conse-
and benefit formula. Richard Fineberg, an quences of policy choices. However, the
environmental consultant in Alaska, argues inability of qualitative CBA to generate an
that the concept of wilderness “is immutable. It unambiguous bottom line—whether the
is like perfection—there are no degrees to it. benefits are greater than the costs—makes
Oil development in a wilderness, no matter it less useful, at least to some users.
how sensitive, changes the very nature of it. It
Cost-benefit analysis’ utilitarian roots make
means it’s no longer wilderness” (Friedman,
it attractive for a number of reasons. It calls
2001).
for a distribution of benefits, opportunities,
Proponents of CBA argue that it is an essen-
and burdens that generates the greatest wel-
tial tool analysts should use in assessing policy
fare for the greatest number. It is consistent
choices, for several reasons:
with economic efficiency; a failure to maxi-
• CBA is promoted as a solution to the mize benefits would be unjust, because of the
problem of lack of sensible priorities. Critics unrealized potential for generating wealth
of government regulation, for example, and the consequential well-being. It seeks to
argue that agencies fail to address the most maximize the economic value of goods and
serious risks, and that identifying the costs services or economic utility. It is also egalitar-
and benefits of alternative actions will help ian: every person’s utility or interests are given
steer them towards the initiatives that equal weight and reinforces democratic
promise to accomplish the most good with expectations of individual equality and
the resources expended. majority rule.
• CBA is championed as a way to avoid poli- Even ardent proponents of CBA may
cies such as overly stringent regulations acknowledge the difficulties in achieving the
that impose major compliance costs but expectations held for it. Many values cannot be
produce limited value. The analysis seeks to easily quantified. More precise economic costs,
ensure that only worthwhile projects are for example, may be given more weight than
undertaken and that imposing costs does imprecise estimates of the value of ecosystem
not waste resources that could be put to health, ecological services, public health, or
productive use. aesthetics. Others emphasize that CBA is not a
• The costs and benefits of different policy neutral analytic tool in practice, but has been
choices are believed to vary significantly, used to provide “objective” cover for decisions
and CBA can help ensure that comparable reached on questionable grounds. The Office of
factors are examined in comparable ways Special Counsel of the Department of Defense,
and options are selected that produce the for example, concluded that officials in the
greatest benefits. One variation of CBA, Army Corps of Engineers appeared to engage
cost-effectiveness analysis, seeks to ensure in a deliberate pattern of manipulating eco-
that the option with the greatest benefit to nomic cost analyses in order to justify expan-
cost ratio is identified. sion of Corps projects (Grunwald, 2000).
• CBA is aimed at identifying the kinds of But, from this perspective, these are largely
decisions markets would produce if they problems in implementation rather than a
were able to function. Efficient allocation reflection of fundamental shortcomings with
of resources is in the interests of everyone, the methodology.
Peter-3383-Chapter-25.qxd 6/8/2006 7:45 PM Page 437
situated are treated the same, and those who are quality, public health, personal freedom, and
in different circumstances are treated differ- equality of opportunity are most powerfully
ently. That is, there may be some factors that defended by arguing that they are priceless,
justify different treatment, although there and that their protection is compromised by
should be consistency within each category. For considerations of costs. Many values cannot be
example, those who benefit from a risky activity easily quantified in dollars; those values that
should bear the associated costs. Or equality can can be more precisely and unambiguously
be understood to require a minimum level of quantified will be given priority over those that
equality or access to benefits or imposition of are less certain. More precise economic costs,
burdens or risks. A commitment to equality for example, may be given more weight than
may conflict with simple utilitarian calculations imprecise estimates of the value of ecosystem
and suggest different distributions to different health, ecological services, public health, or
persons, according to their merit, conditions, or aesthetics. Cost-benefit analysis’s bias against
needs. A collectivist formulation links benefits values that are not easily quantified can be
and needs: everyone is to receive the level of overcome by resisting quantification and lay-
benefits consistent with their needs. In contrast, ing out in qualitative terms the values to be
a merit-based approach requires that benefits be compared. This systematic identification of the
distributed according to effort or contribution costs and benefits of alternatives can be a very
and burdens. Correspondingly, those who make useful decision-making tool and help illumi-
only partial contribution should receive partial nate the consequences of policy choices, but its
benefits (Paul, 1995, 27). inability to generate an unambiguous bottom
The nature of risk is a complicating but line—whether the benefits are greater than the
important factor here. Risks that are voluntarily costs—makes it less useful, at least to some
assumed and represent an informed choice pose users.
less of a problem than risks involuntarily More fundamentally, neoclassical economic
imposed on others. For example, the choice of theory underlies the calculation of costs and
individuals to live on the oceanfront and volun- benefits, but that theory is problematic, for
tarily accept the increased risk of damage or several reasons. At one level, the problem is
death from hurricanes and other storms does rooted in differences between private and
not pose a challenge to this notion of distribu- social costs. Markets work well when private
tive justice. On the other hand, a decision to live and social costs are equivalent, when prices
next to a toxic emitting factory or incinerator is include all the costs that matter to society and
problematic, because it may be in only a very to the environment. But prices often fail to
narrow sense “voluntary;” low-income persons reflect the true costs of production and exter-
may have few options besides choosing to live in nalize rather than internalize costs in the pur-
economically depressed areas where prices are suit of maximum profits. Producers and
low and environmental risks are high. consumers often make choices that reflect their
narrow economic interests at the expense of
broader societal and environmental interests.
There are often differences between the private
QUANTIFYING COSTS AND BENEFITS
and social benefits: the use of some resources
may produce economic gains for some indi-
Critics of CBA argue that it inevitably is biased viduals, while other activities could bring
in favor of easily quantified values, such as much greater gains. Planting a forest creates
industry compliance costs, and against more some level of benefits for a landowner who
intangible and hard to quantify ethical values eventually sells the timber, for example, but it
of biodiversity, environmental health, social also produces additional benefits of sequester-
equity and empowerment of politically dis- ing carbon, filtering air and water, and provid-
advantaged groups, and public well-being. ing habitat. Geoffrey Heal nicely summarizes
Others argue that values such as environmental this fundamental limitation of markets: “Just
Peter-3383-Chapter-25.qxd 6/8/2006 7:45 PM Page 439
as we will tend to do too much of the activities because scarcity “tends to induce competitive
shows social costs exceed the private costs, so bidding and preemptive buying, which lead to
will we also do too little of those for which price fluctuations, market disruption, and the
social benefits exceed private benefits” (Heal, inequitable or inappropriate distribution of
2000, p. 28). resources” (Ophuls and Boyan, 1992, 200). Even
The idea of the state of balance produced by if prices are carefully regulated to ensure they
the invisible hand of markets is rooted in the reflect true costs, the assumptions on which
work of Adam Smith, Thomas Malthus, and neoclassical economics rests are largely at odds
David Ricardo, who sought to discover a set of with ecological science. Assumptions such as the
natural laws of economics that mirrored the following are recipes for eventual environmen-
natural laws of Newtonian physics that came to tal collapse (Nadeau, 2003, p. 9):
dominate thinking in the 18th Century. Just as
• Markets are closed circular flow systems of
the universal force of gravity held the elements
production and consumption separate and
in balance, the natural law of economics
distinct from the external environment,
ensured that individual economic actions
• Inputs of raw materials from the environ-
would, without some overarching direction,
ment are “free” unless costs associated with
ensure order and stability. Smith argued that
their use are internalized,
the market’s invisible hand was similar to the
• Raw materials are inexhaustible or replace-
invisible force that caused a pendulum to oscil-
able by other resources or by technologies,
late in equilibrium or a liquid to flow through
• The external environment is a bottomless
connected containers to reach its own level, a
sink for waste materials and pollutants,
conclusion he drew not from empirical
• There are no biophysical limits to the
research but from metaphysical assumptions.
growth of closed market systems; markets
Subsequent theorists also sought to imbue eco-
will perpetually grow and expand.
nomics with the same kind of authority
enjoyed by physical scientists who substituted Market ideology fails to reflect the larger com-
economic concepts for physical variables devel- munity of land, resources, biomes, languages
oped in 19th-century physics (Nadeau, 2003). and cultures, and institutions in which eco-
Contrary to its theoretical underpinnings, nomic activity is embedded and on which it
however, markets are intimately intertwined depends (Daly and Cobb, 1989). The inherent
with the natural world. Natural resources are dynamic of markets has produced a global envi-
exhaustible and natural systems are, overall, ronmental deficit, the “collective and mostly
irreplaceable. In contrast, market ideology unanticipated impact of humankind’s alteration
demands that costs and benefits that extend of the earth’s atmosphere, water, soil, biota, eco-
very far into the future are discounted and logical systems, and landscapes;” this deficit has
devalued. As a result, as Ophuls and Boyan occurred because “the longer-term ecological,
(1992, 219) put it, “critical ecological resources social, and economic costs to human welfare are
that will be essential for our well-being even greater than the shorter-term benefits flowing
30 years from now not only have no value to from these alterations” (Borman and Kellert,
rational economic decision makers, but scarcely 1991, xii). Environmental deficits rob future
enter their calculations at all.” Markets typically generations as they permit profligate consump-
handle incremental change with relative ease tion by the current generation, who pursue their
but tend to break down when confronted with interests rather than ensuring that the needs of
absolute scarcity or even marked discrepancies current and future generations are preserved.
between supply and demand. In such situations Cost-benefit analysis has deep roots in natural
(e.g., in famines), market collapses or degener- resource policy making (Gramlich, 1990, 2).
ates into uncontrolled inflation, because the Much of the criticism of CBA has been devel-
increased price is incapable of producing an oped in response to its use in environmental
equivalent increase in supply. Markets also fail policy but similar criticisms can be made in
to respond to problems of ecological scarcity other areas of policy sectors. Can policy
Peter-3383-Chapter-25.qxd 6/8/2006 7:45 PM Page 440
Part of what is most central here is an “everything is connected to everything else,” that
endorsement of the idea of true costs—prices is at the heart of ecology. So too is it a commit-
should include the real or true costs of produc- ment to a land ethic. Aldo Leopold (1966,
tion, including the environmental and health 238–39) defined a land ethic, sounding much
consequences and impacts, rather than allowing like a proponent of sustainable communities:
sellers and purchasers to externalize those costs An ethic, ecologically, is a limitation on freedom of
to others. In a system committed to markets as a action in the struggle for existence… . All ethics so far
way of allocating scarce resources, determining evolved rest upon a single premise: that the individual is
value, and making choices from competing a member of a community of interdependent parts… .
The land ethic simply enlarges the boundaries of the
needs and wants, it is critical that prices provide community to include soils, water, plants, and animals,
accurate information about the true costs of or collectively: the land.
what is being sold. As real costs are better under-
stood, those affected by them can have more of Ecological sustainability can be understood in terms of
the information needed to make trade-offs and three core elements:
balance competing concerns such as environ- 1. Ecological integrity, ecosystem services, natural capital,
mental protection and economic development and true-cost prices:
(Prugh, Constanza, and Daly, 2000, 28–29).
Ensure economic activity is within ecological limits,
Cost-benefit analysis can be useful in assessing maintain ecological integrity, protect key ecosystem ser-
policy alternatives, but the analysis requires vices, ensure sustainable yield of renewable resources,
information about true costs. preserve and regenerate natural capital base, internalize
A central feature of ecological sustainability is environmental costs in market exchanges, reduce subsi-
dies that have harmful environmental consequences
its integration of ecological protection and eco-
nomic activity with social equity and political 2. Equity, wealth, and consumption:
empowerment. Sustainable development here
Ensure inter-and intra-generational equity, carefully
gives priority to reducing poverty and helping identify the equity consequences of costs and benefits,
the poor gain some measure of self-sufficiency address interaction of poverty and environmental degra-
through a more equitable distribution of dation, distribute resources in ways that increase oppor-
resources. Sustainability is largely a concept of tunity, transfer and disseminate widely cleaner
community, bound up with notions of strong technologies that benefit all persons, not just the wealthy
democracy, participation, community, and those 3. Democratic politics, community, and natural resource
social characteristics are fostered through a scale governance:
of personal interaction. Political participation is
Foster strong democracy and participation, provide eco-
a key ingredient in ensuring that decisions affect- nomic and technical resources to ensure effective partic-
ing economic and environmental conditions be ipation, strengthen community, civil society, and social
made more inclusive. Sustainability is not an capital, develop economic indicators and measures that
ecological concept alone, but also one of social reflect depletion of natural resources and pollution as
costs, devise broader measures of social and economic
justice, inclusion, fairness, community well-
factors.
being, and political engagement. Sustainability
requires new ways of thinking that integrate Debates over the depth of sustainability per-
ideas of individual rights, community responsi- meate efforts to determine guidelines for public
bility and accountability, material and spiritual policy and private behavior. Questions over the
well-being, and ecological health. It requires, scope of sustainability, and what behaviors
from this view, a radical set of changes in order to should come within its reach are just as con-
ensure fairness in the distribution of benefits and tested. If sustainability requires that the stock of
burdens, a perpetual resource base and ecologi- nonrenewable natural resources be preserved for
cal services, and a social system that secures the future generations, does that mean we have long
interests of all persons. Sustainability focuses on since violated its basic premise? Does sustain-
comprehensive solutions that reflect the inter- ability suggest a static goal of ensuring that
connections of ecology and respects the maxim, future generations have the same level of
Peter-3383-Chapter-25.qxd 6/8/2006 7:45 PM Page 442
resources and ecosystem services to rely on? assessments of discrete policy choices, now
How does sustainability make sense in light of dominated by cost-benefit analysis, with a richer
the dynamic nature of biological evolution? The form of analysis that incorporates a more diverse
value of natural resources, in part, lies in their set of values, perspectives, and objectives.
use, but human use has varied greatly. Uses of
forests have changed dramatically, for example,
from sources of strategic defense materials REFERENCES
(wood for battleships) to watershed functions of
cleansing water. What kinds of changes can and
Borman, F. Herbert and Stephen R. Kellert (1991)
should be limited in the pursuit of sustainability?
Ecology, Economics, Ethics: The Broken Circle. New
Perhaps most difficult is the question of whether Haven: Yale University Press.
sustainability is in the interest of the current gen- Daly, Herman and J. B. Cobb, Jr. (1989) For the
eration. Is it more morally compelling for them Common Good. Boston: Beacon Press.
to improve the quality of life of existing Friedman, Thomas (2001) “Drilling in the
members of that generation and reduce the Cathedral,” The New York Times (March 8), A25.
enormous disparities in wealth, or is it better to Gramlich, Edward (1990) A Guide to Benefit-Cost
set aside resources for those not yet born? Analysis. Prospect Heights, IL: Waveland Press,
Sustainability seems promising as an idea for 2nd ed.
governing a community. The problem is how to Grunwald, Michael (2000) “Shooting Pork in the
extend the idea beyond our community. We may Barrel,” The Washington Post (March 2): A17.
Heal, Geoffrey (2000) Nature and the Marketplace:
find that our own community is sustainable if
Capturing the Value of Ecosystem Series.
we export our wastes or import unsustainable Washington, DC: Island Press.
levels of resources. How does sustainability Leopold, Aldo (1966) A Sand County Almanac. New
guide us in these circumstances? York: Ballantine Books.
Developing good answers to these and other McAllister, Bill (2001) “Norton praises oil-drilling
questions is central to the viability of ecological efforts,” Denver Post (April 5) www.denverpost.
sustainability as a way to integrate CBA, moral com/business/biz0405c.htm.
and ethical frameworks, and other approaches to Merchant, Carolyn (1994) Ecology. Atlantic Highlands,
policy analysis. The idea of sustainable develop- NJ: Humanities Press.
ment has firmly taken root in global, national, Nadeau, Robert L. (2003) The Wealth of Nature.
and local political discourse in a remarkably New York: Columbia University Press.
Ophuls, William and A. Stephan Boyan, Jr. (1992)
short length of time. Government agencies at all
Equality and the Politics of Scarcity Revised.
levels, transnational corporations, multilateral New York: W.H. Freeman.
institutions, community-based collaborative Paul, Ellen Frankel (1995) “Set-Asides, Reparations,
groups, and many others have embraced the idea and Compensatory Justice,” reprinted in Kenneth
of sustainability. The impreciseness of sustain- A. Manaster, Environmental Protection and Justice.
ability makes it an attractive idea around which Cincinnati: Anderson Publishing Co., 27.
diverse expectations can congregate. But that Prugh, Thomas, Robert Costanza, and Herman Daly
impreciseness makes its use as a guide for (2000) The Local Politics of Global Sustainability.
decision-making problematic. The contested Washington, DC: Island Press.
nature of the idea of sustainable development is Rawls, John (1999) A Theory of Justice. Cambridge,
rooted in two primary dimensions: the depth of MA: Harvard University Press, revised ed.
Sandel, Michael (1996) Democracy’s Discontent:
sustainability and the level of change required,
America in Search of a Public Philosophy.
and the breadth of sustainability and the range Cambridge, MA: Harvard University Press.
of practices and behaviors it reaches. As these Soraghan, Mike (2001) “Bush: National monuments
provisions are debated and refined, we can have oil-drilling potential,” The Denver Post (March
develop a clearer picture of what an ecologically 15) www.denverpost.com/news/news0315d.htm.
sustainable technological society might look Weimer, David L. and Aidan R. Vining (1999) Policy
like. The development of sustainability-related Analysis: Concepts and Practices, 3rd ed., Upper
methods and tools can also serve to replace Saddle River, NJ: Prentice-Hall.
Peter-3383-Chapter-26.qxd 6/8/2006 7:46 PM Page 443
26
Performance and Performance
Management
Performance has become a defining feature of in isolation from other factors that make up
modern government. The age of performance public management and the more general
has produced commitment to, but not neces- public administration system. Our study is
sarily clarity about, the universal objective of grounded in public administration tradition
organisational and government improvement. that recognises rationality and trust as two
Performance and performance management fundamental dimensions of organisational life.
continue to be enigmatic concepts. There is a The performance focus not only has an impact
paradox too in that the adoption rate of per- on the key public management functions and
formance approaches continues to expand components (HRM, finance, strategy, etc.) but
despite the mounting critique. One limitation also changes the nature of policy and manage-
has been the failure to integrate performance ment in the public sector in itself.
and measurement within a broader system of
management and within a conception of
organisational life that invests practice with
WHY PERFORMANCE MANAGEMENT?
greater meaning.
Performance management lacks a coherent
treatment that explicates its significance, The need for the systematic study of perfor-
analyses its several dimensions as a working mance management is based on the increase in
system and challenges its shortcomings. The application and its current significance within
purpose of this chapter is to seek an under- OECD. The key debates have produced a bur-
standing of performance and performance geoning literature because they address core
management as concepts as they are applied in issues of public administration and policy, yet
practice and in relationship to public manage- some ambiguities, even confusion, remain
ment. Performance management has to be about the workings and standing of perfor-
located within a broader construction of mance management.
organisational life, which recognises that per- The most striking feature of performance
formance management cannot be considered management is its continuing expansion over
Peter-3383-Chapter-26.qxd 6/8/2006 7:46 PM Page 444
the last two decades, making this current forms the core of public management,
period its international apogee. As will be dis- especially NPM: is it possible to envisage man-
cussed later, the antecedents of performance agement in the public sector without due regard
management – various approaches to perfor- to the pursuit of results and the measurement of
mance and measurement – have a long lineage. performance? This correspondence can readily
But the high international commitment goes be established through stock specifications of
well beyond a resurgence of interest in mea- New Public Management features (e.g., Hood,
suring performance. Moreover, performance 1991). When Behn (1995) writes about the ‘Big
management is recognised to have evolved, to Questions of Public Management’ – microman-
be different and to involve more sophisticated agement, measurement and motivation – he is
measures (Schick, 2001: 40). essentially identifying questions about perfor-
International observers agree that some- mance management. When Hood and Peters
thing unusual was occurring internationally in (2004) reflect on ‘The Middle Aging of New
the 1990s, with ‘the rise and rise of “perfor- Public Management’ are they not writing for the
mance” as an issue in public sector theory most part about the side-effects of performance
and practice’ (Talbot, 1999). Similarly a US management?
observer reports that ‘if there is a single theme The field has become the subject of debates
that characterizes the public sector in the about the value of performance, measurement
1990s, it is the demand for performance. A and their management. Performance manage-
mantra has emerged in this decade, heard at all ment has become a small growth area within
levels of government, that calls for documenta- public administration and management, cen-
tion of performance and explicit outcomes of tred on critiquing aspects of performance and
government action’ (Radin, 2000: 168). The measurement (e.g., Forsythe, 2001). Three
penetration was significant by the mid-1990s: types of argument can be mentioned. The first
measurement was becoming ‘more intensive, is about the impact of rhetoric on poor imple-
more extensive and more external’ (Bouckaert, mentation (‘What makes performance man-
1996); and key performance management agement so attractive in theory, yet so difficult
questions were prominent in a ten-country in practice?’ Thomas, 2004: 1); and the
survey by the OECD but stronger in the Anglo- assumption of the performance management
American compared to the European (OECD, industry that an agency will be ‘transformed by
1997, Bouckaert, 1997). This trend continues measuring its performance. This is the logic of
in the 2000s, with no indication that it is GPRA’ (Schick, 2001: 40, 43). A second critique
abating with ‘performance measurement and explores the limits of rationality and unin-
reporting becoming even more important tended consequences. Managers are seen as
within most governments. “If you can’t mea- functioning within the narrow parameters of
sure it, you can’t manage it” has become a performance management and an imperfect
familiar refrain’ (Thomas, 2004: 1). model that is rationally defined and deficient.
Two big questions in public management A third type of argument points to apparent
during the last fifteen years – markets and per- confusion. There is a lack of agreement on how
formance – have attracted the most contro- to measure performance. Moreover, with per-
versy in an era of unprecedented public sector formance measurement there are an ‘array of
reform. Both have recent origins in New Public buzzwords… reinventing government, new
Management (NPM), although their lineage is public management, performance manage-
much longer, and reflect private sector ideas. ment, results-driven government, results-
Markets are examined in many other studies, based budgeting, performance contracting etc.’
but performance management remains some- (Thomas, 2004: 1); and authors move casually
thing of an enigma and is lacking coherent between one concept and another (e.g., a dis-
meta-analysis. cussion of performance management is really
One factor has been its centrality to public about performance measurement: Bovaird and
management. Performance management Gregory, 1996: 239). For all these reasons there
Peter-3383-Chapter-26.qxd 6/8/2006 7:46 PM Page 445
Performance Administration
Managements of Performances
Performance Management
Figure 26.1 Relationship between measuring, integrating, and using performance and the
evolving models
The relationship between the three 2003; Pollitt & Bouckaert, 2004): the United
performance models, and how each is succes- States, which pioneered the early experiments
sively nested within another, are illustrated in and significant developments up to managing
Figure 26.1. performances, and the United Kingdom,
One further question is the relationship which illustrates the evolution of a system
between performance and management. What of performance management over several
differentiates more developed systems is the decades.
application of management, but this is insuffi-
cient for fully-fledged Performance Manage-
Early Measurement of Government
ment as an ideal type. In Table 26.2 the types
Activities under ‘Performance
in the first column are narrowly conceived in
Administration’
terms of either limited conceptions of perfor-
mance or measurement or both. Whereas the
The early history of performance measure-
two types in the second column are more com-
ment is mainly centred on the United States,1
prehensive. The first line is more diffuse/less
and is complex because of variations in termi-
systematised; the types in the second line more
nology, differences of actors and the levels
integrated within their sphere (either measure-
of government involved. The initial decades
ment or management system).
of the twentieth century resulted in many
institutions focusing on mostly municipal effi-
ciency, but a Bureau of Efficiency (1912), and the
HISTORICAL APPLICATION
Institute for Government Research, the prede-
cessor of the Brookings Institution (1916) were
The framework provides a basis for analysing created at the national level.
the development of performance measure- First generation performance measurement
ment and management. It requires the identi- was concerned with measuring government
fication of the several elements and stages from activities that were eventually to be defined in
concerns with performance measurement in terms of service delivery. This was ‘the earliest
the early decades of the twentieth century to identifiable sustained effort to engage in per-
fully developed performance management at formance measurement and productivity
the end of the century. This is done with refer- improvement’ (Williams, 2003). In the first
ence to two closely-linked countries (Halligan, part of the 20th century the ‘Government by
Peter-3383-Chapter-26.qxd 6/8/2006 7:46 PM Page 449
the Good’, changed into the ‘Government by was on the budget agenda. The Bureau of the
the Efficient’: public administration and scien- Budget started a productivity project in 1962
tific management ‘proclaimed a new gospel to and the eventual report concluded that pro-
a new deity: efficiency’ (Mosher, 1968: 71, pas- ductivity could be measured and that there
sim). There was a focus on economy and effi- should be annual reporting to the President
ciency in a context of a division between and Congress (Kull, 1978: 5). Presidents Nixon
politics and administration. Politicians should and Carter established a succession of national
not be involved in administration since it is productivity commissions.
scientific and value-free, and aims at economy Although the interest in measuring produc-
and efficiency, a conviction strengthened by tivity from an administrative technical point of
the view that the activity was purely technical. view continued, savings and spending less
Although efficiency was generally defined tax money became the main issues in the
as obtaining a result with a minimum of 1980s. Under Reagan the President’s Private
resources, there was discussion from the begin- Sector Survey on Cost Control, the Grace
ning about definitions of efficiency, with ter- Commission, estimated yearly savings of $US
minology focusing on inputs, activities, 3 billion. The Commission was primarily con-
outputs, and efficiency, but not much on cerned with efficiency, expressed as cost
results and outcomes. There was also an savings and operated within an agenda that
assumption that more efficiency results in sought minimal government. This debate had
more effectiveness, allowing administration a technical and an increasingly dominant
to focus on the technical dimensions without ideological angle, which ultimately led to New
having to get involved in effectiveness. Public Management in the 1990s (Bouckaert,
1991).
For the United Kingdom, there were two
Broadening and Elaboration of
decades (1960 and 1970s) of ad hoc experi-
Performance, Measurement and
ments and dabbling in performance measure-
Productivity
ment during which programs were piloted
but were eventually discontinued. The earliest
A second generation of performance measure-
cases of performance measurement have been
ment activity emerged with post-World War 2
traced to the 1960s and associated with cost
experiments, led by the US central govern-
benefit analysis, management by objectives
ment’s interest in performance (in budgeting),
and output budgeting, but only two remained
measurement (more generally) and productiv-
significant in the 1970s. The two decades of
ity. This resulted in the development of tools
discussing and piloting performance measure-
and techniques like PPBS, MBO, or ZBB, all
ment yielded little that was convincing and
including performance information. The first
durable (Bovaird and Gregory, 1996: 239–40).
Hoover Commission (1949) recommended
performance-based budgets and the second
Hoover Commission (1955) on budgets, costs, The Advent of Managements of
and management reports. Economy and effi- Different Performances
ciency remained crucial, but a crucial driver
for the innovations was not the search for a Two decades of pursuing performance,
‘better government’ but an explicit desire to measurement and increasingly management
reduce expenses. For that reason productivity followed (1980s and 1990s). Performance
Peter-3383-Chapter-26.qxd 6/8/2006 7:46 PM Page 450
measurement became a growth industry in the agreeing that something exceptional was
UK following the launching of the Financial occurring (Talbot, 1999; Radin, 2000). For these
Management Initiative in 1982, which was developments to make a difference, it was nec-
designed to focus on objectives and to measure essary for performance to become embedded in
outputs and performance. A significant com- management framed in comprehensive terms.
ponent of the approach was the use of perfor- The Government Performance Results Act
mance indicators (PIs), Prime Minister 1993 provided the main US impetus, eventu-
Thatcher proclaiming in the ‘manifesto of the ally becoming mandatory for national
revolution – that a thousand PIs should flourish’ agencies, and revolutionised the systemic
(Carter, Klein and Day, 1992: 2; Cave, Kogan and implementation of outcome-related goals and
Smith, 1990). By 1987 departments had 1800 performance measurement, planning and
PIs (Pollitt, 1993: 54). reporting. For the UK, the transition from
UK’s FMI was only partly successful yet it Performance Administration to Managements
laid the foundation for the next stage. During of Performances occurred in the 1980s and was
the 1980s the fundamental shift occurred from arguably the more complete move (of the two
public administration to public management countries) to Performance Management in the
with the new philosophy and style being more 1990s.
receptive to and influenced by the private
sector. Public management had already been
identified in the US (Perry and Kraemer, APPLICATION OF TYPES TO COUNTRY
1983), although the more fundamental basics PERFORMANCE MANAGEMENT
of new public management did not register
until the 1990s. This management focus had
A significant element is contextual analysis of
major implications for the efficiency agenda,
different performance management systems to
and provided the basis for the active imple-
enable comparisons to be made across coun-
mentation of performance and measurement.
tries, while recognising the importance of their
The move to management of performances
cultural and administrative traditions. Three
was to emerge from this foundation with the
country approaches can be identified: those
management of individual and organisational
that have made a high commitment to perfor-
performance becoming increasingly common.
mance management, where the interest is in
There were aspirations in both countries to
how they frame their performance manage-
manage individuals in the respective senior
ment systems to handle the limitations to this
civil services through the use of performance-
approach; countries that have sought to bal-
related pay (Pollitt, 1993). Eventually it became
ance performance management with other fea-
possible for the management of performances
tures and how that plays out in practice; and
to assume a more integrated form as a con-
countries that have sought to selectively draw
struction of performance management. New
on performance management techniques
Public Management has an interesting place in
while operating within another type of system.
these developments and the progression
A purpose of the ideal types is to use them
towards performance management.
as a basis for examining the variations in
the orientation of the public administration
Public Management Trends Produce systems to performance management. Pre-
Fusing of Performance and Management liminary analysis indicates that Germany (at
the national level) continues to operate with-
If the 1980s saw the flowering of performance out commitment to, and application of, per-
measurement, the 1990s were years of perfor- formance, measurement and management
mance expansion and management consolida- principles, while France has made a major
tion. Observers in the UK and US record commitment to performance management but
the mushrooming of performance in 1990s, the relationship to management is tenuous
Peter-3383-Chapter-26.qxd 6/8/2006 7:46 PM Page 451
(Table 26.3). Three countries in different ways most of the attributes referred to above, but
approximate Managing Performances: Canada, there remain weaknesses in consistency of
the Netherlands and the United States. This approach, reporting of information externally
means that, while performance management and meaningful applications of principles in
principles and practice may be fairly well practice.
developed, they are inclined to be neither inte- A strength has been central and external
grated nor systemic. For the Performance oversight with a series of inquiries reviewing
Management type two countries, Australia and the principles and their application, but they
the United Kingdom, come closest to the strin- have revealed a persistent problem: the credi-
gent requirements, although there are weak- bility of performance management systems
nesses in both cases that make the fit less than (APSC, 2004). The performance aspects of
optimal. human resource management have attracted
The criteria for performance management sustained interest because of fundamental
include an identifiable framework that must be issues raised by the transformation of employ-
sustained over time and formally supported by ment relations under a highly devolved system.
key actors (e.g. central agencies and cabinet); A prominent issue has been the subjectivity of
and the presence of attributes identified with management assessment of performance, and
Model 4: performance management must be public servant motivation where remuneration
comprehensive (systemic), integrated, coherent is based on performance (e.g. O’Donnell and
and consistently applied across agencies. This O’Brien, 2000).
requires evidence of practice at the agency
level, of how the ‘system’ as a whole is operat-
ing and of a capacity to review and absorb
SOME KEY ELEMENTS OF
lessons. At the agency level, the prerequisites
PERFORMANCE MANAGEMENT ISSUES
include a performance focus, measurement
IN A POLICY CONTEXT
systems in place and a management approach
and a coverage that includes individual, organ-
isational, financial, etc. There also needs to be Performance is relevant for (policies of) single
evidence of the use of performance measure- organisations, for substantive policy areas (see
ment; taking responsibility for performance especially Part III in this volume) and for the
and being held to account; and relationships macro level of countries. The general purpose
between the elements (cascades, individual, is to describe the link of resources with activi-
organisational, etc.). ties and outputs, to link outputs to effects of
One of these cases illustrates the features of organisations and their policies, and some-
the Performance Management type. Australia times even to link this to satisfaction or confi-
has been more committed to performance dence in single organisations, policies or their
management than most OECD countries (see institutions.
OECD, 1997). Two management frameworks Essential information for policy makers is
have existed within which two generations of linking expenditure to effects and to confi-
performance management can be distinguished dence at all levels, if possible.
(Halligan, 2002; McKay, 2003). The perfor- Comparing the macro level between countries
mance management framework encompasses is obviously a challenge. There are differences
Peter-3383-Chapter-26.qxd 6/8/2006 7:46 PM Page 452
needs
policy objectives
operational objectives
indicator of effectiveness
measure of effectiveness
resources
resources activities
activities outputs
outputs effects
effects
indicator of efficiency
measure of efficiency
environment
economy cost-effectiveness
in institutions, definitions and registration But this systemic approach also causes some
procedures for data, and there are longitudinal significant problems affecting performance
inconsistencies. Also data are incorrect, prelimi- measurement and management of ‘substan-
nary, inconsistent, and incomplete. All these rea- tive’ policies, and therefore requires a ‘perfor-
sons are insufficient not to look for performance mance measurement and management’ policy
data and information, integrating and using this itself. It is clear that micro (single organisa-
in policy decision-making, results upgrading, tion), meso (substantive policy), and macro
and enhancing accountability. levels (government wide) are not necessarily
Figure 26.2 places inputs-throughputs- harmonious and compatible (Figure 26.3).
outputs-effects in a context of general and There are obvious conflicts between individual
operational needs and embedded policy objec- schools or hospitals (which may aim at max-
tives that are interacting with environments. imising numbers of students or patients) and
This scheme allows us to define the economy, educational and health policies that want to
efficiency/productivity, effectiveness, and cost- control or even limit (and definitely audit)
effectiveness of organisations, policies, and numbers of students and patients.
countrywide analysis. As a consequence the Also, well functioning and performing
available performance information allows us single organisations do not always guarantee a
to develop a policy focus on this economy, effi- well functioning and good performing policy.
ciency, effectiveness, cost-effectiveness, and Good performing micro levels are necessary
even confidence. It is obvious that defining the but insufficient for a well performing meso
responsibility for performance, however it is level. Performance measurement systems
defined, will immediately determine who is should make this visible and performance
accountable for that performance. management should take this into account by
Peter-3383-Chapter-26.qxd 6/8/2006 7:46 PM Page 453
MACRO
I O E
MESO
I O E
MICRO
I O E
Figure 26.3 Linking micro, meso and macro performance management. I: Input;
O: Output; E: Effect
looking at micro, meso, and macro perfor- This also requires a policy to integrate, to
mance management. add or to consolidate the micro into the meso,
Third, several and different single organisa- and the meso into the macro level.
tions may contribute to a general effect (see ‘Performance management’ also requires a
Figure 26.4). Environmental policies are imple- clear statement on how this performance
mented by a whole range of agencies (inspection, information will be integrated into the finan-
subsidies, direct intervention, communication, cial cycle (budgets, accounts, audits), the con-
etc.), which may be disconnected. Micro perfor- tract cycle, and the policy cycle. It is crucial
mance should be consolidated at the policy level. that performance based information is consis-
But defining the level of attribution of separate tently integrated in a coherent way in the three
well performing agencies to a general policy related financial, contract and policy cycles.
effect is impossible. Therefore, performance This is probably a requirement to guarantee
measurement systems should be developed at the evidence-based policies in all stages.
level of a substantive policy field (meso), and at a
single organisational level (micro). Obviously,
government wide indicators (macro) are also
ASSESSING PERFORMANCE
indispensable and are developed in several
MANAGEMENT IN A POLICY CONTEXT
OECD countries (US, Canada, Australia).
From a ‘performance management’ policy
point of view this implies that an integrated In the history of public management and
micro/meso/macro performance focus is public policy reform one of the common
needed. New Public Management has concen- denominators has been to increase the level
trated more on the need to guarantee an opti- of performance information in order to
mal micro level than a meso level, even up to improve the level of decision-making and the
ignoring the co-ordination of single organisa- level of effectiveness (OECD, 2000; Bouckaert,
tions in a substantive policy field. 1994).
Peter-3383-Chapter-26.qxd 6/8/2006 7:46 PM Page 454
Implementation is difficult and even subject to intensive, more extensive, and more external?
obstruction. And executive politicians lose There is still scope for a more extensive approach,
interest if the potential political yield is not i.e. an increased coverage rate in organisations
obvious. and policy fields. The absorption of performance
A major question is therefore what is going measures in financial cycles (budgets, accounts,
on in the practice of the reform agenda in and audits), in policy cycles (design, strategic
general, of which the performance issue is a plans, monitoring of implementation, and evalu-
crucial common denominator, and how can ation), in personnel functions, and in organisa-
we explain what is going on. tional intra- and inter-action still has potential
One key change has been the creation of for intensive expansion. Finally, a growing exter-
autonomous agencies. These agencies and their nal dimension is also correlated to a developing
guiding centres use a variety of tools and tech- accountability.
niques like, for example performance budgeting, Second, just like the accounting adagium
accrual accounting, performance evaluation and ‘different costs for different purposes’, a system
value for money auditing, personal and institu- needs different performance measurement sys-
tional performance contracts with rewards and tems for different purposes. Three main objec-
sanctions, target setting, benchmarking and tives of performance management include
performance monitoring systems. (re-)allocation of resources (of which savings
In any case, the general assumptions seem to is one specific variant), increase performance
be that more information is better than less, of the public sector as such, and enhance
that comparative information is better than accountability. Performance measurement sys-
non-comparative information, that timely tems will have to cover all these objectives.
information is better than untimely informa- How to balance costs and benefits of perfor-
tion, that output and effect information is mance measurement and systems of managing
better than single input information, and that performance? A performance measurement/
clear and quantitative objectives are better management Cost Benefit Analysis is always
than opaque ones. present, be it more implicitly than explicitly.
Listing costs and benefits is not easy.
Quantifying performance measurement/man-
agement costs and benefits is even more diffi-
CONCLUSION: PERFORMANCE
cult. One key problem is that in general costs
MANAGEMENT IN A CONTEXT OF
are more immediate, more tangible, and less
POLICY EVALUATION
conditional than benefits, which are less
immediate, less tangible, and more condi-
In practice the implementation of measuring tional. Costs precede benefits because one
performance attracts scepticism (Ammons, needs to collect data and process this into
2002), fallacies (Hatry, 2002), and even resis- information before one can collect the benefits
tance (Bouckaert and Peters, 2002). of it. Also costs seem to be more tangible than
Performance measures are Achilles’ heel, and the nature of the benefits, which are also and
measuring performance is Sisyphus’ job. perhaps predominantly intangible. A focus on
However, from the history of performance transparency, accountability, and trust are
measurement systems it is obvious that there is clear benefits but difficult to match to a mone-
a learning cycle and progress is present tary cost. Finally, costs are incurred almost as a
(Bouckaert, 1995, 1996). The study of the fact of an accounting nature, whereas benefits
history of measurement systems in the public depend on the functional use of this informa-
sector can be useful for future strategy. tion in a policy and management cycle.
First, learning from history, for example, the It is important to avoid measurement/man-
progress in measuring performance could result agement systems that are not used, that
in extrapolating the learning cycle. Could per- demonstrate failures (because of technical
formance measurement become even more problems, illegitimacy, or dysfunctionality), or
Peter-3383-Chapter-26.qxd 6/8/2006 7:46 PM Page 456
that clearly demonstrate a perverse effect education, values and commitment in order
(Bouckaert, 1995). Awareness of transitional to re-establish and rediscover democracy
problems like, for example, a temporarily dou- (Gawthrop, 1998). Education of the citizenry will
bling of red tape because of simultaneously empower citizens in a governance context and
reducing ex ante input controls and increasing equilibrate bureaucratic values. Values are crucial
ex post output and outcome controls, is also a and should not be replaced by facts and figures
key element. but balanced with these. Commitment and
Third, these performance measurement engagement imply that public servants know
strategies need to put in a broader context, i.e. what to do and go beyond what they have to do.
the 60-year-old Friedrich/Finer debate is rele- A third option is the dominance of perfor-
vant. The key question is whether there is a mance measurement/management systems that
need for a system based on subjective elements turn the whole system into performance bud-
(e.g., personal accountability based on gets, accounts, audits, contracts, evaluations, etc.
integrity and trust – Friedrich’s position) or for The cost of monitoring and compliance is con-
a system based on objective elements (e.g., rule- siderable. Principles and agents will try to create
directed accountability based on compliance disequilibrated information systems. The
and performance measurement systems – system would not use trust and integrity at all
Finer’s position). for its governing (Bouckaert, 1998).
According to Friedrich (1940), professionals A fourth option – balancing the two
themselves provide ‘fellowship’, which regu- positions – seems to be desirable. A fair mix of
lates the behaviour of fellow practitioners, and trust and measurement, of integrity and com-
their technical and situational knowledge pliance, of subjective and objective approaches
allows them to have a deeper knowledge of the is necessary. Upgrading the two positions and
public interest. Pride-related arguments of combining them seems to be a functional way
these professionals allow for a subjective to make things work. According to Balk ‘pro-
accountability mechanism derived from their fessionals as key actors in the network of stake-
values. According to Finer (1940), the track of holders in a democratic society, will develop
objective accountability should be followed, mutually supportive relationships at the work-
since those within agencies are obliged to place in order to maintain their integrity and
follow and implement the popular will as strengthen democratic political institutions.
expressed by political officials and agency They can take the lead by acknowledging that
executives. Control could then be linked to principles of tentative trust and contingent
performance-based elements. loyalty need not threaten the benefits of, nor
How can this dichotomy be resolved? One the necessity to respond to, appropriate hierar-
option is to upgrade both positions. Balk pro- chical direction’ (Balk, 1996: 190–1). This
poses two fundamental positions to orthodox implies that performance-based systems
practitioner perception. The first raises subjec- should be combined with trust based systems
tive agency accountability to a formal level of in a dynamic way.
visibility by providing an applied theoretical Fifth, there will remain tensions or contradic-
rationale. Then the realm of public administra- tions between responsibility and the account-
tion discourse is extended well beyond conven- ability on issues like output versus outcome
tional management philosophical domains so as and organisation versus policy and these often
‘to recognize employees as legitimate, proactive derive from the relationship between the civil
actors, citizens with a powerful commitment to service and the politicians. Performance
the values and actions essential to democratic measurement/management may clarify this but
political institutions’ (Balk, 1996: 13). will never solve these tensions.
The second option is to choose one position Sixth, the impact of introducing perfor-
as dominant. Gawthrop contends that there is a mance measurement/management in a
need to expand and implement the ethical Weberian context may lead to a neo-Weberian
imperatives for the 21st century through or a post-Weberian design of the public sector
Peter-3383-Chapter-26.qxd 6/8/2006 7:46 PM Page 457
Accomplished Reporting. Wesport, Conn: Quorum Hatry, Harry (2002) Performance Measurement:
Books. Fashions and Fallacies. Public Performance &
Carter, Neil, Rudolf Klein and Patricia Day (1992) Management Review, 25(4): 352–8.
How Organisations Measure Success: The use of Hood, Christopher (1991) ‘A Public Management
performance indicators in government. London for all Seasons?’, Public Administration, 69(1):
and New York: Routledge. 3–19.
Cave, Martin, Maurice Kogan and Robert Smith Hood, Christopher and Guy Peters (2004) ‘The
(1990) ‘Introduction’, in Martin Cave, Maurice Middle Aging of New Public Management: Into the
Kogan and Robert Smith (eds), Output and Age of Paradox?’, Journal of Public Administration
Performance Measurement in Government: The Research and Theory, 14(3): 267–82.
State of the Art. London: Jessica Kingsley Ingraham, Patricia W., Philip G. Joyce and Amy
Publishers. Kneedler Donahue (2003) Government Perfor-
Christensen, Tom and Per Laegreid (2001) (eds), mance: Why Management Matters. Baltimore and
New Public Management, The Transformation London: John Hopkins University Press.
of Ideas and Practice. Aldershot: Ashgate. Jones, L.R. and Donald Kettl (2003) Assessing Public
Cutler, Tony and Barbara Waine (1994) Managing Management Reform in an International
the Welfare State: The Politics of Public Sector Context. International Public Management Review
Management. Oxford/Providence: Berg. (electronic journal at http://www.ipmr.net)
de Bruijin, Hans (2001) Managing Performance in 4(1): 1–18.
the Public Sector. London: Routledge. König, Klaus (1996) On the critique of New Public
Finer, Herman (1940) ‘Administrative Responsi- Management, Speyer Forschungberichte 155.
bility in Democratic Government’, in C.J. Speyer: Institute fur Verwaltungs-wissenschaft.
Friedrich (ed.), Public Policy. Cambridge, MA: Kull, D. (1978) Productivity Programs in Federal
Harvard University Press. Government, Public Administration Review,
Flynn, Norman and Franz Strehl (1996) Public 38(1): 5–9.
Sector Management in Europe. New York Prentice- Liner, Blaine, Harry Hatry, Elisa Vinson, Ryan Allen,
Hall. Pat Dusenbury, Scott Bryant and Ron Snell (2001)
Forsythe, Dall W. (2001) Quicker Better Cheaper? Making Results-Based State Government Work.
Managing Performance in American Government, Washington, DC: Urban Institute Press.
Ithaca, NY: Rockefeller Institute Press. McKay, Keith (2003) ‘Two Generations of Performance
Friedrich, Carl J. (1940) ‘Public Policy and the Evaluation Management Systems in Australia’,
Nature of Administrative Responsibility’, in C.J. Canberra Bulletin of Public Administration, No. 110,
Friedrich and Edward S. Mason (eds), Public 9–20.
Policy. Cambridge, MA: Harvard University Press. Management Advisory Committee (2001)
Gawthrop, Louis C. (1998) Public Service and Performance Management in the Australian Public
Democracy: Ethical Perspectives of the 21st Service: A Strategic Framework. Canberra:
Century. New York: Chatham House. Commonwealth of Australia.
Halligan, John (2001) ‘Comparing Public Sector Miller, Gerald, Hildreth Bartley and Rabin Jack
Reform in the OECD’, in Brendan Nolan (ed.), (2001) Performance Based Budgeting. Boulder,
Public Sector Reform: An International Perspective. Co: Westview.
Hampshire: Macmillan. Morley, Elaine, Scott P. Bryant and Harry Hatry
Halligan, John (2002) ‘Public Sector Reform in (2001) Comparative Performance Measurement.
Australia’, in Byong-Man Ahn, John Halligan and Washington, DC: Urban Institute Press.
Stephen R.M. Wilks (eds), Reforming Public and Mosher, Frederick C. (1968) Democracy and the
Corporate Governance: Management and the Public Service. Oxford: Oxford University Press.
Market in Australia, Britain and Korea. O’Donnell, M and J. O’Brien (2000) ‘Performance–
Cheltenham: Edward Elgar. Based Pay in the Australian Public Service’,
Halligan, John (ed.) (2003) Civil Service Systems in Review of Public Personnel Administration, Spring,
Anglo-American Countries, Civil Service Systems 20(2): 20–34.
in Comparative Perspective Series. Cheltenham: OECD/Organisation for Economic Co-operation
Edward Elgar. and Development (1997) In Search of Results:
Hatry, Harry (1999) Performance Measurement: Performance Management Practices. Paris: OECD.
Getting Results. Washington, DC: Urban Institute OECD (1999) Integrating Financial Management
Press. and Performance Management. Paris: OECD.
Peter-3383-Chapter-26.qxd 6/8/2006 7:46 PM Page 459
OECD (2000) Government of the Future. Paris: Better Cheaper: Managing performance in
OECD. American Government. Ithaca, NY: The
Perry, J. and K. Kraemer (1983) Public Management: Rockefeller Institute Press.
Public and Private Perspectives. California: Talbot, Colin (1999) ‘Public Performance: Towards
Mayfield. a new model?’ Public Policy and Administration,
Pollitt, Christopher (1993) Managerialism and the 14(3).
Public Services, 2nd ed. Oxford: Blackwell. Talbot, Colin (2005 forthcoming) ‘Performance’, in
Pollitt, Christopher and Geert Bouckaert (2000) Christopher Pollitt and Lawerence Lynn Jr (ed.),
Public Management Reform, an International Oxford Handbook of Public Management. Oxford:
Comparison. Oxford: Oxford University Press. Oxford University Press.
Pollitt, Christopher and Geert Bouckaert (2001) Talbot, Colin, Lyn Daunton and Colin Morgan
Evaluating Public management Reforms: an (2001) Measuring Performance of Government
International Perspective. International Journal of Departments – International Developments,
Political Studies, Spring 2001, 167–192. A Report for the National Audit Office.
Pollitt, Christopher and Geert Bouckaert (2004) Abergavenny: Public Futures Ltd.
Public Management Reform: A Comparative Thomas, Paul (2004) Performance Measurement,
Analysis, 2nd ed. Oxford: Oxford University Press. Reporting and Accountability: Recent Trends and
Radin, Beryl (1998) The Government Performance Future Directions, Public Policy Paper 23. Regina:
and Results Act (GPRA): hydra-headed monster The Saskatchewan Institute of Public Policy.
or flexible management tool? Public Williams, Daniel W. (2003) ‘Measuring Government
Administration Review, 58(4): 307–16. in the Early Twentieth Century’, Public Admini-
Radin, Beryl (2000) Beyond Machiavelli: Policy Analysis stration Review, 63(6): 643–59.
Comes of Age. Washington, DC: Georgetown Williams, Daniel W. (2004) ‘Evolution of Perfor-
University Press. mance Measurement until 1930’, Administration
Rogers, Steve (1990) Performance Management in and Society, 36(2): 131–65.
Local Government. Harlow: Longman. Wollman, Hellmut (ed.) (2003) Evaluating Public
Schick, Alan (2001) ‘Getting Performance measures Sector Reforms. Aldershot: Edward Elgar.
to Measure Up’, in Dall W. Forsythe (ed.), Quicker
Peter-3383-Chapter-26.qxd 6/8/2006 7:46 PM Page 460
Peter-3383-Chapter-27.qxd 6/8/2006 7:46 PM Page 461
27
Argumentative Policy
Analysis
HERBERT GOTTWEIS
broad tradition of political and social theory, is not about revealing a truth but attempts to
or displayed a rather indifferent attitude convince (Breton and Gauthier, 2000: 3–5).
towards new social theoretical and philosophi- The Sophists emphasized the importance of
cal approaches. For many scholars in these rhetoric in politics and the idea that facts are
institutions, the idea that language is impor- what we are persuaded of (Danzinger, 1995).
tant in politics linked well with what most edu- Plato accused the Sophists of only dealing with
cated people associate with the Greek the appearances of truth, whereas philosophy’s
philosophy tradition, but this was hardly seen role was to deal with establishing the true and
as undermining any important epistemologi- the good (Meyer, 1994: 50–1). Aristotle, by
cal positions. It was in this intellectual climate contrast, attempted to accord a positive place
that from the late 1980s on important posi- to rhetoric by positioning rhetoric as part of
tions in policy analysis in countries such as in dialectic, along with poetics and the study of
the United Kingdom, Germany, Austria, the topics (Meyer, 1994: 119–23). As Michel Meyer
Netherlands, and Denmark were taken by points out, rhetoric appears forcefully in times
scholars working in the tradition of argumen- of crisis for the lack of directing principles in
tative policy analysis. the settling of questions that are being submit-
At the same time, the European Union not ted to controversial answers. In the absence of
only became a mounting challenge for political leading principles that could offer some defin-
scientists to comprehend, but it also matured itive, unequivocal answers problems are bound
into a major force of research funding. New to be disputed and solved ‘equivocally’. Just as
approaches to understanding Europe and the the Peloponnesian wars led in ancient Greece
unprecedented challenges for European gover- to a collapse of previous and well-established
nance were asked for, which opened up the values and modes of thought and the rise of
space for new theoretical avenues, such as con- rhetoric, the upheavals of our times have led to
structivist policy and communicative planning a new reconsideration of rhetoric, argumenta-
approaches. Most scholars in the argumenta- tion, persuasion and its relationship to logic
tive tradition emphasize their commitment to and communication (Meyer, 1994: 36–7).
offer useful and more practical policy analysis Rhetoric is a discourse in which one can hold
than normally done by the traditional policy opposite judgments on the same question.
approaches. This type of concrete and specific What is problematic remains so through the
analysis and advice was precisely what the displayed multiplicity of judgments (Meyer,
European Commission was looking for in its 1994: 52).
various Framework programs. Thus, a new In contemporary times, the theory of argu-
and important source of funding for argumen- mentation was taken up and further developed
tative policy analysis developed, and which by Stephen Toulmin (1958) and Chaim
helped further expansion. Perelman (1958) in the late 1950s, and the work
of both had a lasting influence in the field of
political science. Closely related to the develop-
ment of argumentation theory was the rise of
ARGUMENTATION, SOCIETY,
hermeneutics, phenomenology, structuralism
AND POLITICS
and post-structuralism, not only in philosophy
but also in the social sciences from the 1970s
Argumentation theory has a long history that on. The ascent of argumentative policy analysis
dates back to pre-Aristotelian philosophy. It must be seen in this complex intellectual envi-
is always connected to considerations and ronment as a result of a political constellation
reconsiderations of the notions of logic, com- of transformation and upheaval, when, during
munication, and persuasion. Mobilizing, posi- the 1980s, largely unanticipated by the interna-
tioning and transmitting arguments also tional political science community, the Soviet
requires appropriate sociopolitical conditions: Empire broke down, the ‘end of history’ was
argumentation is the antithesis to revelation; it proclaimed, the European Union finally rose to
Peter-3383-Chapter-27.qxd 6/8/2006 7:46 PM Page 463
there’, which for them is neither just there nor is traditions in argumentative policy analysis.
this reality the simple outcome of human activ- Whereas positivist social science was searching
ity (Yanow, 1996: 4–9). Many policy studies for universal laws or empirical regularities of
approaches follow the recommendation of human behavior, interpretative approaches
neo-positivist epistemology to engage in were more interested in the unique capacities
hypothesis testing, with the goal to find causal of the human capacity for creating, communi-
generalizations that are quantitatively testable. cating, and understanding meaning, which
In this perspective, language is viewed as a neu- is seen to be located in social practices and
tral medium to record observation. This con- texts (Yanow, 1996: 4–9). The hermeneutic
ceptualization of language and the idea to search approach to data collection provides access to
for universally generalizable findings is ques- facts via the understanding of meaning, which
tioned in the post-positivist, neo-pragmatic can be reconstructed on the basis of the inter-
perspective of argumentative policy analysis pretation of written or spoken texts (Verstehen).
(Fischer, 1998). With it, philosophical realism, Thus, the researcher may come to understand
the idea that objects, properties, and the rela- the actors’ motives to act (both the reasons
tions the world contains exist independently why and the intentions with which to act), and
from human thoughts or perceptions of them, their interpretation of their own actions and of
is put in doubt, or rejected. It is at this point those of others (Gadamer, 1977). A further key
that the relatively harmless assertion that lan- influence on argumentative policy analysis is
guage is important in the political process is in the Critical Theory of Jürgen Habermas,
epistemologically ‘radicalized’ and becomes a who, in his The Theory of Communicative
sharp dividing point within policy studies: Action (1985), had developed a universal prag-
although it is safe to assume that most policy matics through a grand synthesis of the major
scholars will attribute at least some importance traditions of social theory and philosophy. In
to the role of language in policy, many will be particular, Habermas’ normative concept of
hesitant to give up on philosophical realism, the communicative rationality, the process of
idea that ‘things out there’ exist independent of problem solving and conflict resolution
the human mind, a view which also tends to be through open discussion, should turn out to be
broadly shared within the more general political a key inspiration for a number of the authors
science community. working in argumentative policy analysis.
Although philosophical neo-pragmatism Within argumentative policy analysis, there is
and anti-representationalism (Rorty) seem to another strand of reasoning that explicitly wants
unify the various approaches in argumentative to go beyond an exclusive focus on how particu-
policy analysis, a great deal of variation exists, lar actors understand their social worlds and
however, due to the fact that a number of each other. Discourse theory in the tradition of
additional philosophical and sociological tra- Michel Foucault and others emphasizes that dis-
ditions have taken their impact on argumenta- course is not just a system of words, but an entity
tive policy analysis. The intellectual roots of that operates in contexts, in social, economic, or
the post-positivist movement in political in political contexts. Discourse is a historically
science trace as far back to German sociology specific system of signification in which human
and social theory of the late nineteenth cen- and nonhuman phenomena are given meaning
tury and early twentieth century, in particular and context. In this more structuralistic view,
to the work of Max Weber, Edmund Husserl, the idea that actors are ‘unified subject’ who
and Alfred Schutz. In sociology, their work encounter decision-making with an identity
constituted an important influence on the already formulated in terms of his or her prefer-
evolving sociological schools of symbolic ences are rejected. Instead, the notion of actor is
interactionism, critical phenomenology, and questioned in a radical way. Discourses are seen
ethnomethodology, as developed by Erving as constitutive of identities, subjects, and objects
Goffman (1974) and Harold Garfinkel (1967). and thus part of the structure that organizes life
Many scholars have taken up these sociological (Howarth and Stavrakakis, 2000: 6).
Peter-3383-Chapter-27.qxd 6/8/2006 7:46 PM Page 466
The Foucauldian influence in argumentative metaphysical thinking and a desire for control.
policy analysis is historically related to classical Structuralism searches for general ordering
structuralism, as paradigmatically developed principles and universal regularities which
in the work of Ferdinand de Saussure that, in make the world capable of technological and
general, constitutes a key departing point of scientific mastery, which give clear orientation
structuralistic and post-structuralism’s cri- in a world which otherwise would seem to be
tiques of language and meaning. Saussure out of control. According to Derrida, Saussure
developed his notion of structure through the has taken an important step in his critique of
analysis of language (Frank, 1989: 20–33; language by rejecting the idea that words are
Welsch, 1996: 245–255; Derrida, 1976). For somehow images of prior existing ideas; by
Saussure, structure refers to the ordering prin- pointing at the phenomenon that meaning is
ciple according to which the lexicon of a lan- constituted solely by differences between signi-
guage is articulated. This is done in such a way fiers within a system of interrelations; that we
that it can be recognized and mastered as the can only think about the redness of an apple in
lexicon of one and the same national language. relationship to yellow apples and green pears.
This process happens in acts of differentiation But Saussure had insisted on the existence of
and connection; that is all elements of expres- something like a structuring principle of
sion that make a sign audible and legible must language. This, argues Derrida, constitutes a
be clearly distinguishable from another, I must move in the language game of metaphysics.
be able to distinguish between signs. But, Metaphysics not only believes in a transsensual
according to Saussure, this is not immediately world, it also offers orientation in the form
possible merely based on the meaning of signs, of control and domination, for example by
but rather by means of their expression. searching for general ordering principles and
Meaning itself is amorphous, lacking a clear universal regularities. Derrida contests that
profile. Therefore, in order to be able to distin- structures can decompose, that we are entan-
guish among meanings, I can do this only by gled in structures and have no possibility of
differentiating between sound images or writ- getting beyond our ‘Being-inside-structures’.
ten images – the signifiers (acoustic or written He says, ‘Everything is structure’, but he does
image as opposed to the signified, the con- not mean ‘everything’ is taxonomy, but, rather,
cept). Hence, structure is a system of pairs every meaning, every signification, every view
meaning-expression, that is, signified/signifier – of the world is in flux; nothing escapes the play
such that one and only one signified is of the differences and thus nothing can be
assigned to every signifier. The individual signs tracked down and fixed in its meaning (1976;
are exact applications of an invariable law to Frank, 1989).
which they are related, they cannot proliferate Obviously, in this tradition of reflection the
out and become uncontrollable (Frank, 1989). relationship between actors and structures is
But it is exactly the concept of uncontrolla- interpreted in a different light than in the
bility that post-structuralism brings into the hermeneutic tradition. In argumentative
debate. Post-structuralism has taken influence policy analysis this has led to a strong interest
on a number of policy analysis scholars, such in phenomena beyond individual interpreta-
as Jacob Torfing (1999) and David Howarth tions of meaning, such as discourses, narra-
(2000) who, in the context of the ‘Essex School tives and other ‘structures of culture’ that
of Discourse Analysis’ have begun to use the were seen to play a key role in the shaping of
theoretical framework developed by Ernesto policies (mentioned below). There are, for
Lacalu and Chantal Mouffe (1985) for the example, striking differences how in different
purpose of policy studies. In this context, the countries embryo experimentation, cloning,
work of French philosopher Jacques Derrida and stem cell research are approaches on the
plays an especially important role for a new regulatory level. In argumentative policy analy-
reading of language and discourse (1976). For sis such differences are not simply explained
Derrida, the idea of a closed structure reflects by reference to differences in political culture,
Peter-3383-Chapter-27.qxd 6/8/2006 7:46 PM Page 467
such as between Germany and the United ‘contextualism’ (Flyvberg, 2001) as an important
Kingdom, but it is studied how historically feature of social science research. Context-
grown systems of meaning are interpreted and ualism is considered the solution to the seem-
also reinterpreted into new configurations of ing polarity between the idea of reproducible,
meaning and shape policy making (Gottweis, objective truth that compels universal validity,
2002). on the one hand, and sheer relativism (‘any
If language cannot be used like an instru- interpretation is as valid as the next’) on the
ment to describe reality and the goal of social other. Contextualism implies that the data col-
science is not to produce generalizable find- lected in a case study must be understood in
ings, what is the knowledge goal of argumenta- terms of the studied phenomenon’s own con-
tive policy analysis? Although a number of text, history, and sociality. It features a focus on
currents occur in argumentative policy analy- ‘text in context’, that is, on practice (Flyvberg,
sis, the different approaches share the position 2001: 115). The research does justice to the
that the ontology – epistemology distinction particulars of the situation under scrutiny as
that characterizes the conventional (neo-) well as to the meanings produced and con-
positivist inquiry paradigm is mistaken. This veyed by the actors in that situation. Such a
view implies that the social sciences deal with focus on practice and ‘contextualism’ is central
multiple realities that are constructed in con- in recent developments in ‘interpretive’ or
text; a view that holds obvious implications for ‘phronetic’ social science (Flyvberg, 2001).
the methodology of social inquiry. As a result, in Thus, argumentative policy analysis does
the act of inquiring, discovery and verification have a strong commitment to theory develop-
are two sides of the same coin (see for instance, ment, rigorous testing of prepositions, and the
Berger and Luckmann, 1967). This insight has study of the interplay between text and con-
implications for the organization of an inquiry text, the praxis of policy making embedded in
process, from the selection of methods to the its complex social, cultural and economic con-
presentation of the research findings. figurations. But it rejects abstractionism, what
Methodologically, in argumentative policy it regards as the pursuing of imaginary goals,
analysis, there tends to be a focus on entering such as finding insights of general validity or
in a ‘dialogue with the data’, researchers may identifying ‘laws of society’. Argumentative
identify particular instances of empirical real- policy analysis has an interest in variables
ity as exemplars of some conceptual category and their relationship but takes a cautious
of social phenomena, tying them together on approach towards questions of causality,
a higher level of abstraction: a ‘grounded’ dependence, and independence of factors, and
approach to generating theory (Glaser and it views policy as a complex configuration in
Strauss, 1967). Though there is no shared which the material (such as industrial produc-
methodology in argumentative policy research, tion or unemployed workers) is inextricably
the idea of ‘grounded theory’ explicitly or linked with the immaterial (such as statistical
implicitly plays an important role in much data and representational strategies) in a com-
research in the argumentative tradition. plex scenography (Maingueneau, 2002: 71–73)
‘Grounded theory’ means to develop theory located in space and time.
inductively from a corpus of data. Typically, a
textual corpus (such as field notes) are studied
to identify variables and their interrelation-
ARGUMENTATION: VARIATIONS
ships. In this approach, theoretical concepts
ON A THEME
are elaborated at the onset of a research pro-
ject, in the form of a conceptual lens and
further elaborated and refined in the course The phenomenon of argumentation is at the
of a research project in a close interplay with core of argumentative policy and gives it its
the progress made in empirical investigation. name. But, despite a certain shared philosophi-
In a similar way, Flyvberg has spoken of cal and epistemological heritage, there is hardly
Peter-3383-Chapter-27.qxd 6/8/2006 7:46 PM Page 468
one shared notion of argumentation dominant But what precisely is a ‘narrative’, one of these
in argumentative policy analysis (Fischer, 2003). much used and often little understood terms?
This is a strength, but also a weakness. The ‘The narratives of the world are numberless’,
many variations on the theme of argumentation Roland Barthes begins his famous essay on
display the plurality of approaches in argumen- narrative theory. Narrative is present in myth,
tative policy analysis, but the fact that different legend, novella, cinema, weather reports, gov-
scholars use different notions of argumentation ernment documents, political speeches, and
has also occasionally created some confusion articles about molecular biology. News of the
and might have hampered the diffusion of argu- world comes to us in the form of ‘stories’, our
mentative policy analysis. dreams narrate about the subconscious, and for
Majone, for example, defines argument as each of us there is a personal history, the narra-
‘the link that connects data and information tives of our lives, which enable us to construe
with the conclusions of an analytical study. who we are and where we are going (O’Neill,
The structure of the argument will typically be 1994: 11) However, it is only recently that the
a complex blend of factual statements and sub- study of narratives has emerged as a field of
jective evaluations’ (Majone, 1989: 10). For interest for a variety of fields in the social and
Majone, this definition of argument and argu- human sciences, from literary theory to philos-
mentation is crucial, because it determines his ophy, economics, ethnography, and, more cur-
project of policy analysis: the task of the policy rently, political science.
analyst is not the one of the nonpartisan, tech- This ‘narrativist turn’ must be contextual-
nical problem solver but as a producer of ized as part of the broader anti-representation-
policy arguments; the analyst needs to probe alist (Rorty) movement identified above as the
assumptions critically and produce and evalu- breakdown of the logico-deductive models of
ate evidence. reason and knowledge and of transcendental
But Majone, and this is a great element of truth claims. The anti-representationalist per-
the elegance in his approach, does not seem to spective asserted that there is no objective real-
attempt to embed his concept of argumenta- ity that can be universally represented through
tion in any other theoretical context than the a neutral medium of language. The methods
classical Greek tradition of rhetoric. Within of natural science, as interpreted by logical
argumentative policy analysis, an important positivism, were seen as inadequate for
strand of scholarship has begun to locate its an understanding of society and culture.
understanding of argumentation in the tradi- This development opened the way for new
tion of the theory of narration and discourse approaches, one of which is narrative analysis.
theory. Narrative and discourse are structural As Wallace Martin states, ‘Mimesis and narra-
phenomena that underscore that the bound- tion have returned from their marginal status
aries of politics, science, and technology are as aspects of “fiction” to inhabit the very center
always drawn within the larger semiotic con- of other disciplines as modes of explanation
text of the different stories that give a society necessary for an understanding of life’
its identity and hold it together. Theories of (Wallace, 1986: 7).
narrative and discourse have played an impor- The exact nature of narratives and their
tant role in the further development of argu- relationship to politics requires explication.
mentative policy analysis by providing a new, On the most general level, narrativity is the
sophisticated and elaborated theoretical representation of real events that arise out of
approach for linking argumentation with the desire to have real events display the coher-
social, literary and linguistic theory. They spec- ence, integrity, fullness, and closure of an
ify the Argumentative Turn’s doubt in the idea image of life that is and can only be imaginary
of the omnipotent, rational, utility maximizing (White, 1981: 23). Narratives bring elements of
human actor, and show how deeply embedded clarity, stability, and order into what usually
political actors are in complex structures of tends to be the complicated and contradictory
culture, language and symbolic orientation. world of politics. This power to create order is
Peter-3383-Chapter-27.qxd 6/8/2006 7:46 PM Page 469
other. The purpose of this exercise is to storylines as a generative sort of narrative that
determine how a metanarrative recast a allows actors to draw upon various discursive
problem in such a way as to make it possible categories to give meaning to specific phenom-
for opponents to act on an issue over which ena (Hajer, 1993: 56). The focus in his concept
they still disagree (Roe, 1994: 156). of narrative qua storyline is less on narratives
Policy narratives are more specific and as sources for attributing meaning and sense to
describe the frames or plots used in the social a complex policy setting, but as a semiotic tool
construction of the fields of action for policy for the creation of what he calls ‘discourse
making, for governmental activities from envi- coaltions’. Once a discourse is formulated, ‘ …
ronmental to technology policy. These frames it will produce story lines on specific problems,
or plots are principles of organization that employing the conceptual machinery of the
govern events and give orientation to actors in new discourse (e.g. Sustainable development).
a policy field (Gottweis, 1998). Narratives A discourse coalition is thus the ensemble of a
describe a structure of relationships that set of story lines, the actors that utter these
endows the events contained in the account story lines, and the practices that conform to
with a meaning by integrating these events these story lines, all organized around a dis-
into a narrative. Policy narratives or frames are course’ (47). In this respect, Hajer’s concept of
constructed and used to make sense out of the story line is more distant to narrative
events. Their ‘raw material’ are the dispersed theory, and closer to Laclau and Mouffe’s
events, such particularly exciting and econom- (1985) and the Essex School’s notion of the
ically promising research in a sub-field of ‘empty signifier’, that refers to the fact that
genetics that takes place in the laboratories of a important notions, in political struggle, such
number of important medical schools in a par- as environment or nation are not characterized
ticular country, the pharmaceutical industry of by special density of meaning, but, on the con-
this country, the world economic situation, trary, by a certain emptying of their content
and financial resources of the country’s which strengthens their role in the discursive
Ministry of Research. A policy narrative might struggle (Torfing, 1999: 98). Hajer exemplifies
begin to take shape when the Ministry of his concept of discourse coalition by showing
Research argues in an internal paper that spe- how in the United Kingdom the acid rain con-
cial support should be given to researchers cept and the story line to which it gave rise
working in the newly emerging genetics sub- brought a variety of actors with very different
field because they are working in such a background and beliefs together and thus cre-
promising area with high economic potential ated a new dynamics in environmental policy
that might eventually help the international making. Roe gives special attention to how
competitiveness of the country. The construc- narrative policy analysis can be constructed as
tion of such a narrative would also typically to examine competing narratives in policy
involve references to certain metanarratives, controversies (Roe, 1994). Rein and Schön
such as statements about the importance of speak, in vicinity to Goffman’s terminology, of
industrial policy for capitalist development, frames as ‘policy positions resting on underly-
the importance to do everything for the health ing structures of belief, perception and appre-
of the nation, or about the general character of ciation that define what counts as fact or
capitalist development. Political metanarra- argument and that are resistant to resolution
tives endow policy narratives with a higher by appeal to facts or reasoned argument’
authority and legitimacy as they tie policies to (Schön and Rein, 1994: 23). In their key study
the revered hold systems of values and identity ‘Frame Reflection,’ ‘they develop an elaborated
codes of a community. framework of analysis for policy analysis and
But, like with many concepts in argumenta- show how policy makers use frames for dealing
tive policy analysis, there is also not one with problems, manage solutions and find
broadly shared notion of narrative in argu- compromise. An explicit goal of frame analysis
mentative policy analysis. Hajer speaks of in the tradition of Rein and Schön seems to be
Peter-3383-Chapter-27.qxd 6/8/2006 7:46 PM Page 471
to somehow transcend frames via what they what may be spoken of in discourse, which
call frame-reflection, a process guided by statements survive or are repressed and what
Kuhn, Habermas, Foresters and others to individuals or groups have access to parti-
somehow arrive at viable solutions for policy cular kinds of discourses (Smart, 1985: 48).
problems through the process of “thinking However, it is important to see that the regu-
through” the constraining and often confusing lated and systematic character of a discourse
nature of frames in a struggle to avoid the does not imply a sense of closure. The rules
“relativist trap” ’ (38–49). delimiting the sayable or thinkable are not just
An obvious question arising from narrative those internal to discourse, but include rules of
policy analysis is where do political metanarra- combination with other discourses. In the
tives and policy narrative come from? European Community during the 1980s, for
Apparently, narratives could be construed as example, the rise of neo-liberalism to the
‘tools’ used by policy actors in policy contro- status of a political metanarrative was closely
versy. But are political narratives just vehicles connected to a discourse of ‘Europessimism’,
and instruments for skillful actors and inter- the creation of an image of Europe that was
ests in the game of power? There is a strong increasingly unable to compete with nations
tendency in argumentative policy analysis for a such as Japan and the United States. In fact, the
more ‘structuralist’ interpretation of narrative. systemic character of discourse implies its
Discourse, in particular historical and political articulation with other discourses.
discourses, are endowed with ‘narrative struc- In this way it could be stated that discourse
ture’. They are narrativizing discourses, as is always caught in a web of materiality that is
Hayden White puts it, discourses without a a historical product. Discourse is not just a
narrator. No one seems to speak and the events system of words, but an entity that operates in
seem to speak for themselves (White, 1987: contexts – in social, economic, or in political
202). Narrativizing discourse refers to the contexts. But discourses are not simply reflec-
numerous kinds of narratives that every tions of these contexts; rather, they are com-
culture disposes for those of its members (such plex mediations between various codes by
as administrators) who might wish to draw which reality is to be assigned possible mean-
upon them for the encodation and transmis- ing (White, 1987: 202). For instance, political
sion of messages (e.g., about pollution con- discourse on industrial development that dis-
trol). In other words, there are people and cusses different strategies for industry support
interests behind narratives who bring narra- cannot be separated from practices in compa-
tives into the world. But these individuals give nies to which it refers. At the same time, the
birth to narratives only within the confine- objects of this discourse do not coincide with
ments of the available discursive possibilities. the range of activities and phenomena they
Actors cannot freely choose the narratives they systematize as central to specific discourses,
deploy. The given discursive possibilities though they bear relations to them. For exam-
describe the large reservoir of narratives, ple, industrial development discourse is not
which can be mobilized for political purposes. equivalent to the investment and product
For example, whereas it had certainly been for development decisions of companies, though
decades within the realm of discursive possi- there is a specific relationship between these
bilities to craft neo-liberalism into a new polit- two domains of the theoretical and the mater-
ical metanarrative, it was not before the 1980s ial. We can say that, to some extent, context is
that neo-liberalism actually achieved the status already in discourse, that discourse bears the
of a new political metanarrative in many coun- traces of content, and that the study of dis-
tries. In this context, Foucault has emphasized course involves the question as to what extent
the importance of the archive, the general discourses are able to come to terms with con-
system of the formation and transformation of texts and vice versa. Discourse should be seen
statements existing at a given period within a as a practice of production that can be charac-
particular society. This archive determines terized as material, intertextual and complex,
Peter-3383-Chapter-27.qxd 6/8/2006 7:46 PM Page 472
always inscribed in relation to other discourses role of policy analysis and the policy
and thus part of a ‘discursive economy’ analyst in the policy process. As already shown,
(Henriques, Holway, Urwin, Venn and argumentative policy analysis sweepingly
Walkerdine, 1984: 105–106). rejects the idea of the ‘neutral’ ‘objective’ policy
Industrial production, for example, is gov- analyst qua social technician and, rather,
erned by a proliferation of discourses and it espouses the idea of the policy analyst as some-
cannot exist prior to or outside of its articula- thing like a lawyer (Majone), advocate, deeply
tion with a set of other social and political engaged in the policy process itself. Although
discourses in concrete historical conjunctures authors such as Majone and Stone have not
(Daly, 1991: 100). Discursive formations like gone further than rejecting the ‘objectivist’ idea
modern biotechnology are intertwined with of the policy analysist, in the wake of Fischer
other disparate factors, such as institutions or and Forester’s ‘Argumentative Turn’ and
economic forces, which establish flexible rela- Dryzek’s ‘Discursive Democracy’ the notion of
tionships and interactions that merge into a argumentative policy analysis as fostering and
single apparatus (Dryfus and Rabinow, 1983: encouraging political participation and delib-
121). The fact that one discursive formation eration has become very influential. With the
gains influence over another, that it becomes departure of the idea that the main task of the
hegemonic, is related to the degree of congru- policy analyst is to identify solutions for objec-
ence and complementarity that this discursive tive problems, the image of the professional
formation has within a given discursive con- expert is reconstructed as one of the facilitators
stellation and within an apparatus or historical of public learning and political empowerment
bloc. Rising unemployment rates, lack of (Fischer, 2003). Torgerson argues that ‘just as
success in high-technology fields such as in positivism underlies the dominant technical
information electronics, and the discourse of orientation in policy analysis, so the post-
Europessimism are examples of such a constel- positivist orientation now points to a participa-
lation in Europe, in which neo-liberal narra- tory project’ (Torgerson, 1986: 241). Forester
tives began to gain political currency. (1999), Healy (1997), and Innes (1999) have
By introducing the concepts of discourse advocated communicative policy analysis, the
and narrative in argumentative policy ana- idea that the main task of the policy analyst is
lysis (Fischer, 2003: 73–94; 161–180), the to facilitate process of deliberation and to help
‘Argumentative Turn’ has contributed signifi- planners to critically reflect on their own dis-
cantly to reshaping the agenda or rhetorical cursive practices. Underlying this ‘policy
analysis in argumentation theory. Discourse model’ is an approach towards communication
and narrative provide crucial and sophisticated and argumentation strongly influenced by
conceptual tools for argumentative analysis the late work of Jürgen Habermas. In his
and link micro- and macro-levels of analysis, ‘Theory of Communicative Action’, (1985, origi-
structure with action, culture with citizen, and nally in German, 1981), Habermas has devel-
institutions with everyday life practices. They oped a notion of ‘communicative rationality’,
bring the notion that ‘language matters’ to life which defines as rational what is communica-
by introducing an array of analytic techniques tively, intersubjectively justified or justifiable.
to deal with phenomena like language, sym- Rationality comes into existence via intersub-
bols, talk, and interaction and their role in the jectively grounded argumentation. This advo-
play of power. cacy for a ‘communicative policy analysis’ is in
most detail elaborated in Dryzek’s ‘Discursive
Democracy’, which discusses Habermas’s cri-
tique of instrumental rationality, the idea to
THE POLITICS OF ARGUMENTATION
devise, select, and effect good means to clarified
ends, and the alternative model of a communi-
One of the key characteristics of argumentative cated rationality, oriented towards the coordi-
policy analysis is its conceptualization of the nation of interactions via communication
Peter-3383-Chapter-27.qxd 6/8/2006 7:46 PM Page 473
(Dryzek, 1990). This idea of policy analysis as a structuralist relativization of the autonomy of
deliberative, participatory, communicative pro- human action for the purpose of discourse
ject can be followed from Torgesen, via Dryzek, analysis per se, without being prepared to con-
the ‘Argumentative Turn’ and, most recently, in sider the corresponding normative-political
‘Deliberative Policy Analysis’ by Hajer and consequences. The very optimistic approach
Wagenaar (2003), who bluntly state that policy towards ‘collaborative and participatory policy
analysis is deliberative (21). Clearly, this ‘poli- making’ of many proponents in argumentative
tics of argumentation’, the advocacy for a par- policy analysis seems to be especially problem-
ticular political engagement of policy analysis, atic in a situation, when, like at the beginning
is consistent with the dominant identifica- of the 21st century, many Western and non-
tion of the nature of policy analysis as being Western governments have enthusiastically
engaged in the clarification of argumentation, a embraced the idea of participatory democracy
process that by definition is open-ended and as a new material and symbolic method to
can hardly be defined as monopolized by any create political stability and, to paraphrase Ian
form of expertise. On the other hand, it cannot Hacking, ‘to make up’ a participatory citizenry
be ignored that an interesting tension exists as part of its governance agenda.
between the basically idealistic idea of commu- Via the Habermasian ‘Communicative
nicative rationality, the idea that, given the right Action’ model another important feature of
domination-free circumstances, individuals are argumentative policy analysis has been intro-
capable of making reasonable, informed, inde- duced in argumentative policy analysis: a cer-
pendent decisions benefiting the common tain tendency towards cognitivism. Discourse
good, and the embrace of the work of Foucault ethics in the Habermas tradition starts from the
and other post-Enlightment thinkers by many assumption that moral problems are capable of
authors in argumentative policy analysis. being solved in a rational and cognitive way.
Traditionally, the Derrida/Foucault tradition of However, students of rhetoric and persuasion
discourse analysis has been rather critical in politics can hardly subscribe to the notion
towards the last heroic efforts of the Frankfurt that policy disputes are always solved or settled
School personified by Jürgen Habermas to save by appeal to reason. Although argumentative
the Enlightment project and the idea that policy analysis clearly recognizes this phenom-
reason can be brought into the world through enon, much analysis in this tradition tends to
discussion and communication. Although the pay only scant attention to phenomena such as
comprehensive work of Jürgen Habermas passions and emotions in policy making, prob-
in the field of language, argumentation and ably because of an understanding of discourse
pragmatics has been to a considerable extent and argumentation that tends to reduce argu-
ignored in argumentative policy analysis, his mentation to the operation of logos than to
idea of communicative rationality and, in par- pathos and ethos, which, in the tradition of
ticular, the underlying human model and polit- Greek rhetoric, have received much attention.
ical philosophy, have been highly influential on In fact, there seems to be a tendency in argu-
it. This has led to the following paradox: on the mentative policy analysis to confine reasoning
one hand we see in argumentative policy analy- to deliberative and judicial reasoning as apart
sis a considerable influence of a Foucault- and separated from manipulative, negative
inspired understanding of discourse as far as rhetoric. Propaganda is clearly differentiated
epistemology goes, but great reluctance to deal from ‘genuine argumentation’, and in this
with the underlying rather more pessimistic respect argumentative policy analysis seems to
Foucaudian/structuralist critique of the idea of be closer to the Platonic ideal for science as a
the human subject and the potential of humans search for truth than to the Aristotelian/
for free, domination-free interaction when it Sophistic tradition (Turnbull, 2003). In its
comes to normative-political positions. It attempt to avoid what is seen as one of the
seems that quite a number of authors writing in main mistakes in neo-positivist policy analy-
the argumentative tradition can relate well to a sis, namely the confusion of reason with
Peter-3383-Chapter-27.qxd 6/8/2006 7:46 PM Page 474
as fear, anger, trust, and respect. We then have Berger, Peter L. and Luckmann, Thomas (1967). The
to pay close attention to the performative and Social Construction of Reality: A Treatise in the
situated dimensions of persuasion (Hajer, Sociology of Knowledge. London: Penguin Books.
2005), a key element in the decision-making Braun, Dietmar and Busch, Andreas (2000). Public
Policy and Political Ideas. London: Edward Elgar.
process, and try to better understand the inter-
Breton, Philippe and Gauthier, Gilles (2000).
mediation of pathos, ethos, and logos, or, in
Histoire des Théories de l’Argumentation. Paris : La
other words, the intersecting of argumenta- Decouverte.
tion, feelings, and status of speakers. What thus Daly, Glyn (1991). The Discursive Construction
comes into view is the complex scenography of of Economic Space: Logics of Organization
policy making, its argumentative performativ- and Disorganization. Economy and Society, Vol.
ity and location in time and space. Such an 20, 79–102.
approach is certainly compatible with the idea Danzinger, Marie (1995). “Policy Analysis Post-
of deliberative policy analysis but also opens modernized: Some Political and Pedagogical
up new avenues of analysis for policy making Ramifications.” Policy Studies Journals 23, No. 3,
beyond the local level and for the broad variety 435–50.
Derrida, Jacques (1976). Of Grammatology.
of strategies being used in policy making. The
Baltimore: Johns Hopkins University Press.
role of the policy analyst would still be the one
Dryfus, Hubert L. and Rabinow, Paul (1983). Michel
of the ‘clarifier’ of issues of public deliberation, Foucault: Beyond Structuralism and Hermeneutics.
as envisioned by Lasswell, without being Chicago: Chicago University Press.
limited to the (typical local) role of the ‘delib- Dryzek, John S. (1990). Discursive Democracy:
erative practitioner’ (Forester). This approach Politics, Policy, and Political Science. Cambridge:
would also open up the normative horizon for Cambridge University Press.
the consideration of political innovation Fischer, Frank (1980). Politics, Values, and Public
beyond participatory democracy. One of the Policy: The Problem of Methodology. Boulder, CO:
classical questions of liberal democracy, Westview Press.
namely how to cultivate appropriate compas- Fischer, Frank (1995). Evaluating Public Policy.
Chicago: Nelson Hall Publishers.
sion, and to link emotions and passions with
Fischer, Frank (1998). ‘Beyond Empiricism.’ Policy
the polity (Nussbaum, 2001: 422–423) might
Studies Journal 26, 129–146.
have to be asked anew by a policy analysis that Fischer, Frank (2000). Citizens, Experts, and the
has at its core a comprehensive understanding Environment: the Politics of Local Knowledge.
of argumentation. Durham: Duke University Press.
Fischer, Frank (2003). Reframing Public Policy:
Discursive Politics and Deliberative Practices.
REFERENCES Oxford: Oxford University Press.
Flyvberg, Bent (2001). Making Social Science Matter.
Why Social Inquiry Fails and How it can Succeed
Adam, Jean-Michel (1999). Linguistique Textuelle et Again. Cambridge University Press.
Analyse des pratiqueses Discursives, Nathan: Paris. Forester, John (1985) (ed.). Critical Theory and Public
Aristotle (1995). The Art of Rhetoric. London: Life. Cambridge, MA: MIT Press.
Penguin. Forester, John (1993). Critical Theory, Public Policy,
Abu-Lughod, Lily and Lutz, Catherine (1990). and Planning Practices. Albany: SUNY Press.
‘Introduction: Emotion, Discourse, and the Politics Forester, John (1999). The Deliberative Practioner:
of Everyday Life’, in Abu-Lughod and Lutz (eds), Encouraging Participatory Planning Processes.
Language and the Politics of Emotion. Cambridge: Cambridge, MA: Cambridge University Press.
Cambridge University Press, 1–23. Fortenbaugh, W.W. (1975). Aristotle on Emotion.
Austin, John, L. (1992). How to do Things with London: Duckworth.
Words. Cambridge, Mass. Foucault, Michel (1972). The Archaeology of
Baudrillard, Jean (1976). L’échange symbolique et la Knowledge and the Discourse on Language. New
mort. Paris: Gallimard. York: Pantheon.
Benveniste, Emile (1972). Problèmes de linguistique Frank, Manfred (1989). What is Neostructuralism?
générale, Volume II. Paris : Gallimard. Minneapolis: University of Minnesota Press.
Peter-3383-Chapter-27.qxd 6/8/2006 7:46 PM Page 478
Gadamer, Hans (1977). Philosophical Hermeneutics. Hume, David (1739). A Treatise on Human Nature.
Berkeley: University of California Press. Mossner, Ernst C. editor, Viking Press, 1986.
Garfinkel, Harold (1967). Studies in Ethno- Innes, J. E. (1996). ‘Planning through Consensus
methodology. Englewoods Cliff, NJ: Prentice Hall Building.’ Journal of the Americam Planning
George, Jim and Campell, David (1990). ‘Patterns of Association 62, 460–472.
Dissent and the Celebration of Difference: Critical Innes, J. E. and Booher, D. E. (1999). ‘Concensus
Social Theory and International Relations.’ Inter- Building as Role Playing and Bricolage: Toward a
national Studies Quarterly, Vol. 34, 269–293, 273. Theory of Collaborative Planning.’ Journal of the
Glaser, Barney G. and Strauss, Anselm L. (1976). American Planning Association 65, 9–26.
Discovery of Grounded Theory: Strategies for Keller, Douglas (1989). Jean Baudrillard: From
Qualitative Research. Chicago: Aldine. Marxism to Postmodernism and Beyond. Stanford:
Goffman, Erving (1974). Frame Analysis. Stanford University Press.
Cambridge, MA: Harvard University Press. Lacalu, Ernesto and Mouffe, Chantal (1985).
Gottweis, Herbert (1998). Governing Molecules: Hegemony and Socialist Strategy. Towards a
The Discursive Politics of Genetic Engineering in Radical Democratic Politics. London: Verso.
Europe and in the United States. Cambridge, MA: Lasswell, H. D. (1951). ‘The Policy Orientation.’ The
MIT Press. Policy Sciences. H. D. Lasswell and D. Lerner.
Gottweis, Herbert (2002). “Stem Cell Policies in the Stanford: Stanford University Press.
United States and in Germany: Between Bioethics Lasswell, H.D. (1971). A Pre-View of Policy Sciences.
and Regulation. Policy Studies Journal 30, New York: American Elsevier.
444–469. Lutz, Catherine and White, Geoffey M. (1986). ‘The
Gottweis, Herbert (2003). Theoretical Strategies of Anthropology of Emotions’. Annual Review of
Post-structuralist Policy Analysis: Towards an Anthropology, 15, 405–35.
Analytics of Politics, 247–265. In Hajer and Lyotard, Jean-Francois (1979). The Post-Modern
Wagenaar, 2003. Condition. Manchester: Manchester University
Habermas, Jürgen (1985). Theory of Communica- Press, 1984.
tive Action, Vols. 1 and 2. Boston: Beacon Press. Maingueneau, D. (1991) L’Analyse du Discourse.
Hajer, Maarten A. (1993). “Discourse Coalitions and Paris: Hachette.
the Institutionalization of Practice: the Case of Maingueneau, Dominique (2002). Analyser Les Textes
Acid Rain in Britain,” in Frank Fischer and Jon de Communication. Paris : Nathan.
Forester (eds.), The Argumentative Turn in Policy Majone G. (1989). Evidence, Argument and Persuasion
Analysis and Planning. Durham: Duke University in the Policy Process. New Haven, Conn: Yale
Press, 43–76. University Press.
Hajer, M. A. and Wagenaar, H. Eds. (2003). Deliber- Martin, Wallace (1986). Recent Theories of Narrative.
ative Policy Analysis: Understanding Governance in Ithaca: Cornell University Press.
the Network Society. Cambridge: Cambridge Meyer, Michel (1986). De la Problématologie.
University Press. Philososphie, Science et Langage. Bruxelles: Pierre
Hajer, M. A. (2005). Setting the Stage: A Dramaturgy Mardaga.
of Policy Deliberation Administration & Society, Meyer, Michel (1991). Le Philosophe et les Passions:
624–647. Esquisse d’une histoire de la nature humanine. Le
Hall, Peter (1986). Governing the Economy: The Livre de Poche, Paris.
Politics of State Intervention in Britain and in Meyer, Michel (1994). Rhetoric, Language, and
France. New York: Oxford University Press. Reason. University, Part: The Pennsylvania State
Healey, P. (1997). Collaborative Planning: Shaping University Press.).
Places in Fragmented Societies. London: MacMillan. Nussbaum, Martha C. (2001). Upheavals of Thought.
Henriques, Julian, Holway, Wendy, Urwin, Cathy, The Intelligence of Emotions. Cambridge: Cambridge
Venn, Couse and Walkerdine, Valerie (1984). University Press.
Changing the Subject: Psychology, Social O’Neill, Patrick (1994). Fictions of Discourse:
Regulation and Subjectivity. London: Methuen. Reading Narrative Theory. Toronto: University of
Howarth, David, and Stavrakakis, Yannis (2000). Toronto Press.
Discourse Theory and Political Analysis: Identities. Perelman, Chaim and Olbrechts-Tyteca, Lucie
Hegemonies and Social Change. Manchester: (1958). Traité de l’argumentaion, la nouvelle rhé-
Manchester University Press. torique. Paris : Presse Universitaire de France.
Peter-3383-Chapter-27.qxd 6/8/2006 7:46 PM Page 479
Roe, Emery (1994). Narrative Policy Analysis: Theory Toulmin, Stephen (1958). The Uses of Argument.
and Practice. Durham: Duke University Press. Cambridge: Cambridge University Press.
Rorty, Richard (1989). Contingency, Irony, and Turnbull, Nick (2003). ‘The Implications of the
Solidarity. Cambridge: Cambridge University Division of Logic and Argumentation for Policy
Press, 5. Theory’. Paper given at the 2nd ECPR Conference,
Sabatier, Paul A. (1988). ‘An Advocacy Coalition Marburg, 18–21 September 2003.
Framework of Policy Change and the Role of Marin, Wallace (1986). Recent Theories of
policy-Oriented Learning Therein’. Policy Sciences Narrative, Cornell University Press: Ithaca.
21, 129–88. Wells, Susan (1996). Sweet Reason. Rhetoric and the
Schön, Donald A. and Rein, Martin (1994). Frame Discourses of Modernity. Chicago: Chicago
Reflection: Towards the Resolution of Intractable University Press.
Policy Controversies. New York: Basic Books. Welsch, Wolfgang (1996). Vernunft. Die zeitgenoes-
Smart, Barry (1985). Michel Foucault. Chichester: sische Vernunftkritik und das Konzept der transver-
Ellis Horwood Limited. salen Vernunft. Frankfurt am Main: Suhrkamp
Stone, Deborah (1997). Policy Paradox: The Art of Verlag.
Political Decision Making. New York, London: White, Hayden (1981). ‘The Value of Narrativity in
W.W. Norton & Company. the Representation of Reality,’ in Mitchell, W.J.T.
Svasek, Maruska (2002). The Politics of Emotions. (ed.), On Narrative. Chicago: University of
Emotional Discourse and Displays in Post-Cold Chicago Press.
War Contexts, ‘Focaal – European Journal of White, Hayden (1987). The Content of the Form:
Anthropology’, No. 39, 9–27. Narrative Discourse and Historical Representation.
Torfing, Jacob (1999). New Theories of Discourse: Baltimore: Johns Hopkins University Press.
Laclau, Mouffe and Zizek. Oxford: Blackwell Yanow, Dvora (1996). How Does a Policy Mean?
Publishers. Interpreting Policy and Organizational Action.
Torgerson, D. (1986). Between Knowledge and Washington, D.C.: Georgetown University Press.
Politics: The Three Faces of Policy Analysis,
Policy Sciences, 19, 33–59.
Peter-3383-Chapter-27.qxd 6/8/2006 7:46 PM Page 480
Peter-3383-Chapter-28.qxd 6/8/2006 8:27 PM Page 481
28
Disciplinary Perspectives
JON PIERRE
Richard Rose once described the study of Applying a strict economic analysis when
public policy as “disciplined research but designing a policy towards homelessness is
undisciplined problems”.1 While academics probably not a recipe for appropriate public
think and act very much within disciplinary policy; nor would sociology probably have
boundaries, real-world problems are rarely very much to offer in terms of advice in the
understood by one academic discipline. foreign policy field.
Almost all policy problems have political, eco- To understand the contributions of different
nomic and social dimensions – we could easily academic disciplines to public policy analysis we
add several more – which means that our must also remind ourselves that these disci-
academic understanding of those problems plines are not only tools for research but are also
require some degree of inter-disciplinary present in the process which is analysed. Policy
work. As a result, the study of public policy has makers acquire advice from academics both
almost always been more or less multidiscipli- inside and outside the institutions of govern-
nary. In the Introduction to this volume we ment and civil servants usually have an
discuss the eclectic nature of policy research, academic degree in the social sciences or eco-
and it is safe to say that no single academic nomics. Moreover, think tanks and independent
discipline can claim monopoly on policy research institutions are omnipresent in the
research. Public policy is an umbrella concept political debate, presenting reports and policies
for government actions to address vastly dif- with which they try to influence elected officials.
ferent kinds of societal problems. Given that Here, too, we can see how public policy is eval-
government policy includes such different uated, and how alternative policies are propa-
policies as housing policy, economic policy gated, from an academic viewpoint.
and defense policy, it should not surprise any- This chapter will discuss the contributions
one to see that an academic understanding of of three academic disciplines – economic
public policy almost by design must incorpo- theory, political science and sociology – to the
rate theories and framework from several dif- analysis of public policy. Thus, we are leaving
ferent academic disciplines. However, different aside, for the most part, the role of these disci-
disciplines represent different discourses, draw plines within the policy process. It is not very
on different theories, focus on different eco- difficult to see, for instance, how the Zeitgeist
nomic, social and political phenomena and are of the 1960s and 1970s, with its emphasis
geared to solve different types of problems. on political solutions to most or all societal
Peter-3383-Chapter-28.qxd 6/8/2006 8:27 PM Page 482
problems, made sociologists and political to place more emphasis on the interests and
scientists attractive policy advisors, whereas goals of the actors than on the role of analysis
the more market-embracing policy style of the and causality in designing the policy. Also,
1980s and 1990s gave economists a privileged policy makers are sometimes guided almost
position in the policy process. Bearing these entirely by political or economic necessity.
caveats in mind, the overarching question During time periods of extensive cutbacks, poli-
which this chapter will address is what types of cies tend to be assessed in terms of their eco-
knowledge and information that the three aca- nomic potential, regardless of whether that was
demic fields can produce, given the core focus the theory which guided the initial design of the
of the disciplines and the typical research ques- policy.
tions which are raised within the disciplines. A final introductory observation is that
To what extent can these disciplines help the there has for a long time been some degree of
observer uncover the logic of public policy and rivalry between different academic disciplines
explain the policy process and its outcomes? with regard to their purported values, both in
Before we embark on that analysis, three understanding public policy and also the
general comments need to be emphasised. quality of the advice they can offer to policy
First, given the eclectic nature of policy makers. Economists from time to time tend to
research, scholars face a significant problem in think that social scientists rely too much on
applying the wrong discipline to a given policy. “soft data” and fuzzy theories and therefore fail
In the ideal world, there should be some degree to produce good scientific explanation.
of correspondence between the causal model Similarly, political scientists and sociologists
sustaining a policy on the one hand, and the sometimes accuse economists for using unreal-
academic theory used to analyse the policy on istic models of rationalistic, utility-maximising
the other. A policy which is aimed at increasing actors in their research; for instance, Peter Self
the number of new businesses, for instance, (1977) once referred to cost-benefit analysis –
departs from some model of what discourages and economics more in general – as “nonsense
potential entrepreneurs from launching new on stilts”. Given the growing interest among
business ventures and seeks to remove those political scientists in rational choice-based
obstacles. These problems could be assumed to models of political behaviour, some of the
be primarily economic in nature, hence, eco- antagonism between political science and eco-
nomics should be the preferred – although not nomics may have faded, but even so we need to
necessarily the only – analytical model to be aware of the tension that exists between
employ in the policy analysis. Sociologists and these different academic camps.
political scientists would probably be able to
offer some useful contribution to the analysis,
but if the problem is construed as primarily an APPROACHES TO PUBLIC POLICY
economic or financial problem, that is where
we would expect the most rewarding analysis
Economics, political science and sociology
to come from.
offer three different analytical vantage points
Secondly, we need to realise that the notion of
for policy analysis. Before we embark on a
a policy process guided by scientific logic and
more systematic comparison of the three disci-
causal models is a somewhat idealised represen-
plines we should first briefly go over the main
tation of policy making. In order to understand
feature of the three perspectives on public
policy design or policy change we need to not
policy.
only look at public policy in and of itself but
also at the politics of policy (Gourevitch, 1986).
Politicians and political parties have strategic Economic Theory
goals and long-term visions and manoeuvre
tactically to promote those objectives and to In order to forestall the impending attack from
cater to their constituencies. This pursuit tends economists, it must immediately be noted that
Peter-3383-Chapter-28.qxd 6/8/2006 8:27 PM Page 483
there is no single “economic theory”; there are has helped us better understand a number of
theories at the micro, meso and macro levels of complex policy-related issues such as tax
society just as there are economic theories that evasion and the consumption of collective
highlight different policy areas. In addition, goods. To be sure, given that political solutions
economic theory is beginning to pay greater to societal problems nearly always tend to be
attention to the role of institutions and the collective solutions, the nexus of collective
consequences of different institutional action and individual rationality offers a host
arrangements. Indeed, constitutionalism – a of important research questions.
research field that attracts both political scien- That having been said, we should also note
tists and economists – is an area of research that the agency-focus which is typical to much of
which seeks to assess the macro-economic economic analysis creates problems in account-
consequences of institutions for political rep- ing for the role of institutions. Furthermore, if
resentation and policy implementation. we allow institutions to incorporate not just
An important contribution of economics in structure but routines and practices as well, we
understanding public policy is the analysis of can see that economics, in the form of neo-
how markets can be employed to enhance classical micro-economic theory, has problems
allocative efficiency. Neo-classical market conceptualising or explaining behaviour which
theory is a powerful analytical tool in this is guided by non-rational consideration.
research. Applied more specifically on the Douglass North, one of the founding fathers of
public sector, public choice – the so-called the institutional school of economic theory,
Chicago school – helped understand some bemoaned these shortcomings and delivered a
problems in tax-financed service provision. It powerful attack on neo-classical economic
might be argued that economists do a better theory based on an institutional approach to
job at describing these problems than actually economic development. Development econom-
explaining them – largely because political ics, North argued, had failed to produce expla-
agency and institutional inertia are phenom- nations to cross-national variations in economic
ena that are difficult to accommodate in eco- growth and was simply “not up to the task”
nomic models – but even so it remains clear (North, 1990: 11).
that economic theory delivers an important Essentially, North’s criticism of not just
contribution to policy studies. development economics but much of the neo-
Even more importantly, economic theory is classical market theory as well takes aim at the
the keystone of cost-benefit analysis, an inability of these theories to account for coop-
approach which probably is the most signifi- eration and institutions. The argument is that
cant contribution of economic theory to theory does not take real-life factors into
public policy. Cost-benefit analysis, which has account and hence cannot provide a useful
developed into a key analytical model in policy framework for empirical analysis. This debate
analysis, accords values to different policy among economists has probably helped move
options and hence provides an elegant model the discipline towards greater policy relevance
for assessing alternative policies in addressing and understanding.
policy problems (see Vining and Weimer, this
volume).
Economics as a cluster of theories rests Political Science
firmly on an assumption of rationality. There
is a distinct focus on agency, and the notion of Public policy is to a very large extent a political
the rational actor has become a powerful pre- phenomenon and, as such, a field of expertise
dictive analytical model. While much of the for political scientists. In particular, political
focus is on individual market behaviour, the science can account for not only the substan-
model has also been immensely useful in studies tive side of policy but also the politics of public
on collective action and the “free-rider” policy, i.e. the compromises among political
problem (Olson, 1965). This research, in turn, actors and the trading of support for different
Peter-3383-Chapter-28.qxd 6/8/2006 8:27 PM Page 484
policy proposals as well as an understanding of political power and the institutions where it is
the institutional dimension of public policy. vested remain the key players in the policy
Thus, political science uncovers the process process. Certainly, actors are framed by institu-
through which policy evolves and what factors tions and previous policy decisions, but history
drove policy design and choice of policy is replete with examples of how the exercise
instruments. of political power has made an autonomous
Political science also has much to offer in and decisive impact on the course of events.
terms of understanding agency as well as insti- Politics, and public policy, is to a great extent
tutions, although the combination of the two about “who gets what, when and how”, to quote
seems to be somewhat of a weak spot; policy Lasswell’s (1936) classical book title, and polit-
studies typically focus either on institutions or ical science remains the discipline most unam-
actors. The stages model allows for analysis biguously devoted to study of those issues.
which takes into account both institutions and All of that having been said, political science
actors and the interplay between these two also has several shortcomings in understand-
variables (see Hill and Hupe, this volume). The ing public policy. One significant problem is
institutional dimension of public policy has related to the previously mentioned tension
been the dominant field of public policy between structure and agency and the com-
research over the past couple of decades. plexities of bringing the two together in an
Alongside that research there has also been a analytical framework. While most political sci-
rapid growth in the application of rationalistic entists probably agree that the interplay
models of political behaviour and decision- between structure and agency harbour much
making. Interestingly, however, so far rational of the explanation to public policy issues, very
choice has made rather few inroads into policy few have been able to provide ideas about how
research. to integrate them into a common framework
Futhermore, political science appears to be or theory.
geared to provide an understanding of how Another potential problem with the political
some types of public policy are shaped by science approach to public policy is its state-
intergovernmental relations. Decentralisation centric view on governance. Although the
has a profound influence on the state’s ability state-society literature has helped us under-
to impose policies on subnational institutions stand the often complex relationship between
and to ensure equal standard (or equalisation) policy makers and policy targets, political
across the nation. Intergovernmental relations science as a whole still seems to take comfort in
appear to a growing extent to be negotiated the notion that once a policy has been decided
rather than hierarchical, and the state finds by political institutions there is an immediate
itself increasingly often playing the role of a and appropriate societal response to that
mitigator between transnational and subna- policy. Implementation research has, from
tional institutional systems and ditto actors time to time, been successful in disturbing that
and interests. The resulting multi-level gover- comfort by showing the frequently rather mea-
nance, as shorthand for contextually defined gre outcome of policy (Winter, 2003). And,
and negotiated institutional relations, has more recently, governance scholars have con-
gained much attention in recent research, not vincingly argued that the state-society nexus is
least because it highlights significant limita- a two-way street and public policy is in no
tions to state power and centrality (see, for insignificant measure shaped by actors outside
instance, Bache and Flinders, 2004). the state (Pierre and Peters, 2000; Rhodes,
A final stronghold of political science in 1997). While there is much to suggest that the
public policy research is the discipline’s focus state remains the undisputed centre of gover-
on power and allocation of public goods. While nance, we need also be prepared to acknowl-
policy scholars frequently note the contextual edge the increasing societal complexities and
and ad hoc nature of policy processes (see how they shape much of contemporary gover-
Kingdon, 1995), it is difficult to ignore that nance (Pierre and Peters, 2005).
Peter-3383-Chapter-28.qxd 6/8/2006 8:27 PM Page 485
A third potential problem, finally, in the to a resolution through political means – but
political science approach to public policy is a rather as problems which are derived from
tendency to focus too much on interests and social dynamics or from the relationship
objectives while ignoring the significance of between social and economic organisation.
political ideas, ideology and vision. It is easy to If treating economics and political science as
argue that politicians are driven by strategic homogenous academic projects is problematic,
and tactical goals such as power, perks, pres- this becomes even more difficult with sociol-
tige, success, and re-election. The notion that ogy. This discipline displays the full range of
people sometimes choose to become involved different ontological and epistemological tra-
in politics because they have a vision about ditions. Some sociologists are anchored in
how to transform society is, while intriguing methodological positivism and conduct their
and far from unlikely, far more difficult to fit research using mass data. Others employ a
into an analytical model. Goal-oriented (teleo- more deductive, theoretical approach and use
logical) explanations are tricky – politician X empirical data basically to drive home a theo-
did what she did because she wanted to, where retical argument. Yet others are almost exclu-
goal and action become circular evidence – sively concerned with the development of
and are also not conducive to theory building. social theory. Sociological theory is equally
For successful and hugely influential politi- heterogeneous. The classical focus on struc-
cians like Margaret Thatcher, Francois ture-agency problems has spun off a wide vari-
Mitterand, John F. Kennedy and Olof Palme, ety of approaches such as neo-Marxism,
power and political office was probably more a institutional theory and theories focused on
means towards an end rather than a goal in sociological issues such as power and power-
itself, yet much of the political science analysis lessness, alienation and similar issues. Thus,
has remained focused more on strategy than sociology has much to offer to public policy,
on sheer political passion. The recent interest both in terms of presenting and describing
in the “ideational” approach which highlights social problems and also in measuring the
the significance of ideas and beliefs in shaping effects of policy implementation.
politics (see, for instance, Berman, 1998) holds Much of sociology has, with massive simpli-
some promise in ameliorating this problem. fication, a conceptualisation of the state, which
to some extent tends to obscure the potential
role which public policy may have in address-
Sociology ing social problems. Sociologists with a neo-
Marxist orientation argue that the state is a
Sociology was once argued by two prominent captive of the ruling class in society and/or that
sociologists to “not apply” to any areas of public there is little the state can do to mitigate the
policy (Scott and Shore, 1979). These two social ramification of the capitalist economy
scholars appear to have underestimated the (Therborn, 1978). A related perspective is that
value of their discipline for helping to solve the state caters to economic interests and
policy problems such as crime, poverty, and the therefore public policy is design to fulfill
disintegration of family structures. Sociology “reproductive” functions within that economic
has tended to address these issues as social order. Alternatively, the state is seen as a set of
problems, but attempts to understand the structures controlled by the political elite.
social causes of these conditions may be crucial Either way, policy design is divorced from
for designing policies to ameliorate them. social problems per se, hence the lower social
Thus, sociology’s perhaps main contribution to classes have little to expect from the state or
public policy lies in its ability to identify and public policy in terms of support or welfare.
analyse social problems. Again, however, there Further, the sociology of organisations is
is to some extent a trait in sociology not to important for understanding policy implemen-
identify social problems as political problems – tation. Implementation scholars tend to focus
that is to say, problems which are conducive just as much on the organisational context
Peter-3383-Chapter-28.qxd 6/8/2006 8:27 PM Page 486
within which implementation takes place as on Public policy is to a large extent a matter
policy objectives or design when accounting of choice. Policy making, for the most part, is
for the degree of success or failure in policy less about whether or not a policy should be
implementation (see, for instance, Pressman created towards some political objective but
and Wildavsky, 1984; Winter, 2003). To be sure, rather more about incremental adjustments in
sociology, particularly sociological institution- terms of spending and organisational design.
alism, thus seems to have much to offer in Sociology performs rather poorly in assessing
terms of understanding consequences of options or in understanding decision-making
organisational design and of changes in regula- process.
tions. Much of the institutional theory that has It almost goes without saying that this some-
become mainstream political science, and what rhapsodic review of three major academic
which more recently is making inroads into disciplines can hardly do justice to any of them.
economics, has its intellectual roots in sociology The purpose of this quick review has been to
and the structure-agency problematic as it is point at some strengths and weaknesses in the
treated by sociologists. three disciplines with regard to their contribu-
The most important contribution which tions in understanding public policy.
sociology has made to public policy research is
probably its analysis of framing. Sociologists
typically argue that, throughout the policy CONTENDING THEORIES AND
process, issues need to be framed in certain THE STAGES MODEL
ways in order to receive attention and to be
successfully resolved. The important point
We will now go into some more detail by dis-
here is that what matters is not so much issues
cussing the strengths and weaknesses of the
per se but rather how those issues are
three disciplines at different phases of the
described and politically packaged, which to a
policy process. We have already had some
large extent determines the future political
remarks on these issues; now is the time to
handling of them. More recently, framing has
look more closely at them.
found its way into public policy through the
constructivist approach to public policy (see
Gottweis, this volume). Agenda setting
There are also some significant shortcom-
ings with regard to the utility aspect of sociol- A number of observers have pointed to control
ogy in public policy analysis. First of all, over the political agenda as one of the most
sociology tends to portray and conceptualise powerful ways to control the political process,
power in ways which makes it difficult to (see e.g. Kingdon, 1995; Schattschneider, 1960).
employ in a meaningful way in public policy. The three disciplines discussed in this chapter
To most sociologists, power or control is perform rather differently with regard to their
derived from (nominally) non-political vari- utility in public policy analysis. Quite some
ables such as the economy or social status. time ago now, Bachrach and Baratz pointed
Thus, power is rarely seen as something com- out that “there are two faces of power, neither
ing out of the policy process. That said, there of which the sociologists see and only one of
may certainly be some value in the insight that which the political scientists see” (Bachrach
political power is not the only source of influ- and Baratz, 1962: 947). Their argument, in
ence and control in society but it might not be brief, was that political scientists tend to focus
useful when designing public policy. their analysis on power in the decision-making
Another aspect of sociological theory which context, forgetting that the power to keep
to some extent obstructs its utility in public issues off the agenda is an equally important
policy is that it is not a discipline directed expression of power and control. Sociologists
towards choice, options and decision-making. fail to see both of these manifestations of
Peter-3383-Chapter-28.qxd 6/8/2006 8:27 PM Page 487
power and control, according to Bachrach and predominant theory of agenda setting available
Baratz. While that was probably a correct today. At the same time, however, we can see
observation at the time of writing, sociology that politicians seem to be increasing attention
has since then made some significant contri- to controlling the political agenda and putting
butions to our understanding of the power a positive “spin” on issues on the agenda. Thus,
dimensions of the agenda-setting stage of the politicians are probably more aware of the
policy process. Steven Lukes’ (1974) analysis of significance of controlling this phase of the
power as an instrument of influencing the decision-making process than are most acade-
basic images and perceptions of personal effi- mic observers of the process.
cacy and social structure is a significant contri-
bution to the power literature at large and to
the study of agenda setting in particular.
ASSESSMENT OF POLICY OPTIONS
Bachrach and Baratz’ conclusion that
sociology is basically oblivious of the power
associated with agenda setting is somewhat The evaluation of decision alternatives is the
harsh. To be sure, one of the strengths of soci- next phase we will discuss. Here, it appears as if
ology is its research focus on social problems all three disciplines have some help to offer,
which do not make themselves easily visible to although each discipline can only produce a
the political and administrative elite. More somewhat limited and uni-dimensional assess-
importantly, sociology is highly useful for an ment of the policy options. One of the key
understanding the framing of issues, some- problems in assessing policy options is that
thing which can be decisive in whether or not such an assessment will almost by design have
an issue finds its way on to the political agenda to reflect the multi-dimensional nature of
(Schon and Rein, 1994). many policies. Thus, for instance, most regula-
Economics can make a contribution to tory reforms have some economic, social and
understanding agenda setting, but we should political or institutional ramifications, hence
acknowledge that this is not an area of expertise an assessment of such a policy option must be
for economists. For both sociologists and econ- conducted against the backdrop of the stated
omists their overall contribution to agenda set- objectives of reform. Some regulatory reforms
ting lies in the normative dimension of the may be costly but will be implemented regard-
actual state of affairs, to paraphrase the German less, since they serve other objectives than
philosopher Fichte; by drawing the attention of simply saving public resources.
politicians to social problems or misuse of Economics have much to offer in terms of
public resources, these disciplines can exert sub- assessing different policy options. Cost-benefit
stantive influence over the policy agenda. analysis is a powerful approach in policy
Given its overall focus on the political choice analysis (see Vining and Weimer, this
process, political science should be geared to volume). Again, however, economists have
understanding the agenda-setting phase of the problems accounting for those types of bene-
policy process. Interestingly, however, political fits which do not lend themselves to measure-
scientists tend to draw on the work of scholars ment which is compatible to financial costs.
from other disciplines, like, for instance, mass True, most economists seem to agree that you
communication, when approaching the can, indeed, measure the benefit of clean air
agenda-setting phase of the policy process (see or consumption of culture or personal safety,
McCombs, 2004). Alternatively, research on but not everyone would probably agree with
agenda setting tends to focus on contextual that.
richness more than theory building (cf. Political scientist E. E. Schattschneider once
Kingdon, 1995). suggested that “the definition of the alterna-
In sum, our knowledge on agenda setting tives is the supreme instrument of power”
is rather scattered and eclectic. There is no (Schattschneider, 1960: 68). While that is a
Peter-3383-Chapter-28.qxd 6/8/2006 8:27 PM Page 488
congenial observation, it has not left any critical phase of the policy process and the key
noticeable traces in the political science litera- focus of political science.
ture. To some extent, political scientists share That having been said, we should also note
some of the economists’ problems with quanti- that there is little agreement among political
tative assessment of decision alternatives. The scientists about how to approach decision-
typical criterion advanced by political scientists making. Influenced by economic theory,
would be goal attainment – i.e. the extent to organisation theory and decision theory, polit-
which the different options mean reaching the ical science today displays the full range of
goals and objectives of the policy – which cer- theories of decision-making, from rational
tainly is useful information, although it says choice theory to the “garbage can model of
rather little about policy substance. Sociologists, organizational choice” (Cohen et al., 1972). The
finally, can also assess policy options according fact that these contending views of these issues
to their consistency with the policy goals and are sit at the predominant line of skirmish in the
also well equipped to assess both the likely out- discipline today is a somewhat paradoxical
comes of policy options as well as assessing proof of the significance of decision-making
potential implementation problems. in political science. Thus, while decision-
To sum up, all three disciplines – but proba- making in principle should be perhaps the
bly most economics – can help us understand main stronghold of political science, there is
the phase of selecting policy options. A signifi- little agreement about which theory, or theories,
cant problem, however, is that this selection is best at conceptualising and explaining such
often is a multidimensional task, where vari- decision-making.
ables such as probability of successful imple- Understanding decision-making is a matter
mentation, costs and predicted outcomes, of understanding both the substantive process
must be weighed against each other, and the but also the strategy of the actors involved in
three disciplines tend to only be capable of that process. Decision theory provides an
assessing one or very few of these variables. excellent base for understanding the strategy of
the actors but says very little about the sub-
stantive aspects of the decision-making.
Decision-making One might argue that economics, at least
micro-economic theory, should be a theory of
Political science – if we stick with Bachrach decision-making, since this is a theory which
and Baratz’ portrayal of the social sciences and deals extensively with preferences and choice,
political power – does capture one aspect of broadly defined. However, this theory says sur-
power but it is not the type of power which prisingly little about decision-making strictu
manifests itself in control of the political sensu. Choice – the process of choosing – is not
agenda. Instead, they argue, political scientists given much attention, presumably because
have a strong expertise in understanding one choice is believed to stem automatically from
“face of power” – decision-making. It is diffi- preferences and objectives. Economic man
cult to take issue with that argument; political never makes irrational decisions, hence little
science is a discipline which directly and indi- need to focus on decision-making. Once the
rectly is geared to enhance our knowledge alternatives and preferences are known, choice
about how decisions are formed. David will follow suit.
Easton’s (1971) definition of politics as the Sociology too is rather quiet when it comes
“authoritative allocation of values” strongly to understanding decision-making. Its main
suggests that political science is first and fore- contribution probably comes from organisa-
most a discipline about decision making. tional sociology, which is focused on organisa-
Similarly, Harold Lasswell’s (1958) classical tional behavior, including decision-making.
book title “Politics: Who Gets What, When, Also, decision theory is influenced by organi-
How” also points at decision-making as the sational sociology; Herbert Simon’s seminal
Peter-3383-Chapter-28.qxd 6/8/2006 8:27 PM Page 489
analysis of bounded rationality grew out of political science literature on public policy and
work conducted by economists, sociologists implementation (Salamon, 2000).
and political scientists (Simon, 1967). Economic theory is not directly focused on
the complexities of policy implementation, at
least not as far as the organizational dimension
IMPLEMENTATION of implementation is concerned. However,
economics has proven to be useful in those
cases of policy implementation which involve
If sociology has fairly little to contribute to an economic incentives as a means of altering social
understanding of the decision-making phase or market behaviour. Significant segments of
of the policy process, it has substantively more environmental protection policies draw on
to offer when it comes to understanding incentives as a policy instrument. Similarly,
implementation. Policy implementation is financial incentives is a frequently used policy
normally an interorganisational process, instrument in industrial policy, and there are
frequently transcending the border between many features of the tax system of many coun-
the state and the surrounding society. Imple- tries which effectively serve as incentives to
mentation can also be conducted at the reward a certain behaviour.
regional and local levels of the public sector, To sum up, implementation is for the most
something which again highlights the need to part an interorganisational process, and disci-
ensure that different organisations interact plines geared to understanding that problem-
efficiently. atic will also be helpful in understanding
Political science too has – at least from time implementation. Moreover, by linking imple-
to time – had a significant interest in policy mentation to policy design, particularly the use
implementation (see Winter, 2003). Much of of economic incentives, economic theory has
the initial wave of implementation studies fol- much to contribute to the analysis of policy
lowed on the growth of policy programs in the implementation at the micro level.
United States during the late 1960s and early
1970s (Pressman and Wildavsky, 1984). When
the outcomes of those policies proved to be
EVALUATION
rather meagre, scholars turned their attention
to the implementation phase to see what types
of obstacles explained that pattern. Since then, To assess the utility of the three disciplines in
implementation research is an established sub- policy evaluation, it is useful to make a distinc-
field of political science. This research typically tion between process evaluation and outcome
argues that policy design is a key variable in evaluation. Although most evaluation seeks to
understanding the success or failure of policy measure different types of policy outcome –
implementation. Policy design, which does not the effects of the policy – process is sometimes
take into account the institutional context also evaluated in order to generate knowledge
within which the policy is to be implemented, (what organisational theory calls second-loop
runs a significant risk of being inefficient learning) about the efficiency of the process in
because it is not very likely to be properly devising appropriate policy. All three disci-
implemented. Closely related to matters of plines can make important contributions to
policy design is the cluster of issues on policy policy evaluation. Economics can assess the
instruments. The selection of policy instru- efficiency and effectiveness of the policy, e.g. to
ments is an important factor in understanding what extent the same outcomes could have
policy implementation. To a significant degree, been achieved at a lower cost. Political science
this argument relates to economic theory is good at process tracing and can link out-
(Kirschen et al., 1964), although lately this area comes to input and also generate explanation
of research has become more integrated in the to the degree of policy success or failure by
Peter-3383-Chapter-28.qxd 6/8/2006 8:27 PM Page 490
bringing political and institutional variables electoral input and political and administrative
into the analysis (see Vedung, this volume). accountability.
Sociology, finally, has a stronghold in analysing Turning to sociology, it is the unrivalled dis-
social problems and the extent to which they cipline in terms of contextualising; the key
have been ameliorated by a particular policy. sociological contribution to policy analysis is
Policy evaluation is sometimes part of the that it uncovers the significance of the institu-
legitimation process of public policy. By mak- tional and organisational within which policy
ing evaluation an integrated part of the policy is designed and implemented. Questions about
itself, policy makers provide critics with a why policy evolves in the fashion it does, or
channel to voice their discontent with either why the implementation of a policy was suc-
the policy process or with the policy outcomes. cess or a failure, cannot be fully answered with-
Evaluation in this perspective is a means of out taking these contexts into account. These
institutionalising a feedback loop in the policy contexts represent much of the backdrop
process. Sociology and political science have a against which economists and political scien-
great deal of input on these issues, which tists try to make sense of the assessment of
arguably are more salient to policy analysts policy options or the actual decision-making.
than they are to policy makers. To these three disciplines we should also add
decision theory as a means of understanding
strategising. This approach to public policy is
CONCLUDING DISCUSSION
entirely focused on the politics of policy mak-
ing and the behaviour of decision makers. It
Looking back at the discussion thus far, it says virtually nothing about the substantive
appears as if there are a couple of more overar- dimension of policy but it is certainly valuable
ching features of the three disciplines which to in terms of understanding agency in the policy
a large extent determine the utility of them in process. However, the analysis of policy sub-
understanding public policy. Economics, in stance cannot be divorced from the analysis of
this context, is first and foremost the art and policy strategy; since policy makers pursue
craft of optimising; economic theory is out- both substantive and strategic interests in the
standing in identifying allocative inefficiencies. policy process, decision theory can be of indi-
That having been said, optimising the use of rect use in understanding why the process gen-
scarce resources can be problematic in policy erated a particular outcome.
as it is not the only criterion which policy mak- We could certainly have added more disci-
ers take into consideration when policies are plines to the analysis. Law has probably
designed. become more and more important, along with
Political science, on the other hand, might the increasing significance of courts in policy
be able to tell us more about these trade-offs implementation and intergovernmental rela-
between conflicting policy objectives. If eco- tionships. Statistics crops up at every phase of
nomics is about optimising, political science the policy process as a prerequisite for good
deals to a large extent with issues of satisficing policy advice. Philosophy – particularly nor-
and democratising. Political science is weary of mative philosophy on state power versus
optimising because it draws on a rationalistic individual freedom – provides the point of
idealisation of the policy process. Satisficing – departure for much of policy theory and social
Herbert Simon’s model of bounded rationality theory, just to mention a few additional acade-
in organisations – seems to capture much of mic disciplines, which directly or indirectly
political decision-making under uncertainty speak to public policy. Again, this is proof of
and complex contingencies. Equally impor- the eclectic and inter-disciplinary nature of
tant, democratising is a process through which public policy.
policy, is perceived as a collective project, with Policy makers have looked at these disci-
all that entails in terms of transparency, plines with changing degrees of affection or
Peter-3383-Chapter-28.qxd 6/8/2006 8:27 PM Page 491
hostility during the post-war period. U.S. Cohen, M. D., J. G. March and J. P. Olsen (1972),
President Wilson is once said to have asked for ‘A Garbage Can Model of Organizational Choice’,
advice from a one-armed economist to escape Administrative Science Quarterly 17: 1–25.
the “on the one hand, but on the other hand” Beman, S. (1998), The Social Democratic Moment:
Ideas and Politics in the Making of Interwar
type of analyses that economists sometimes are
Europe (Cambridge, MA: Harvard University
accused of delivering. During the harsh eco-
Press).
nomic situation of the 1990s, economic con- Easton, D. (1971), The Political System: An Inquiry
siderations played a key role in the policy into the State of Political Science (Chicago: The
process in most western democracies. Given University of Chicago Press).
that policy objectives such as balancing the Gourevitch, P. (1986), Politics in Hard Times:
budget, curbing inflation and cutting taxes Comparative Responses to International Economic
were overriding political goals, economic crite- Crises (Ithaca: Cornell University Press).
ria became the main yardstick with which Kingdon, J. W. (1995), Agendas, Alternatives, and
policy proposals were assessed. Public Policies (2nd. edn) (New York: Harper
Seen in a longer time perspective, however, Collins).
Kirschen, E. S. et al. (1964), Economic Policy in Our
one cannot escape the notion that it is awk-
Time (Chicago: Rand McNally).
ward to believe that the answer to the question
Lasswell, H. D. (1958), Politics: Who gets What,
of what is good policy sits at the end of an eco- When, How? (New York: Meridian Books).
nomic formula. The “New Deal” and the Luke, S. (1974), Power: A Radical View (Basingstoke:
“strong society” in the United States and the Macmillan).
extensive welfare states in many Western McCombs, M. (2004), Setting the agenda: Mass
European countries were huge political pro- Media and Public Opinion (Oxford: Polity
jects that were launched, not because they were Press).
believed to be economically sound but because Meehan, E. J. (1982), Economics and Policymaking:
they were seen as politically desirable and The Tragic Illusion (Westport, CT: Greenwood
socially justified. Economics has little to offer Press).
North, D. A. (1990), Institutions, Institutional
in terms of a deeper understanding of why
Change and Economic Performance (Cambridge:
politicians behave the way they do. Instead, a
Cambridge University Press).
deeper understanding of policy rests to a large Olson, M. (1965), The Logic of Collective Action: Public
extent on an appropriate employment of all Good and the Theory of Groups (Cambridge, MA:
the disciplines discussed in this chapter. Public Harvard University Press).
policy, to reiterate a comment from the Ormerod, P. (1994), The Death of Economics
Introduction to this volume, is an eclectic and (London: Faber).
multi-disciplinary challenge. Pierre, J. and B. G. Peters (2000), Governance,
Politics and the State (Basingstoke: Palgrave).
Pierre, J. and B. G. Peters (2005), Governing Complex
NOTE Societies: Trajectories and Scenarios (Basingstoke:
Palgrave).
Pressman, J. L. and A. Wildavsky (1984),
1. I am, as ever, indebted to Guy Peters for valuable
Implementation (3rd edn) (Berkeley: University
comments on an earlier version of this chapter.
of California Press).
Rhodes, R. A. W. (1997), Understanding Governance:
Policy Networks, Governance, Reflexivity, and
REFERENCES
Accountability (Buckingham: Open University
Press).
Bache, I. and M. Flinders (eds) (2004), Multi-level Salamon, L. M. (2000), Handbook of Policy
Governance (Oxford: Oxford University Press). Instruments (New York: Oxford University
Bachrach, P. and M. S. Baratz (1962), “Two Faces Press).
of Power”, American Political Science Review, Schattschneider, E. E. (1960), The Semi-Sovereign
vol. 56: 947–52. People (New York: Holt, Rinehart & Winston).
Peter-3383-Chapter-28.qxd 6/8/2006 8:27 PM Page 492
Schon, D. and M. Rein (1994), Frame Reflection: Simon, H. A. (1967), Administrative Behavior: A
Toward the Solution of Intractable Policy Study of Decision Making in Administrative
Controversies (Cambridge, MA: MIT Press). Organization (New York: Free Press).
Scott, R. A. and A. R. Shore (1979), Why Sociology Therborn, G. (1978), What does the Ruling Class do
Does not Apply: A Study of the Use of Sociology in When it Rules?: State Apparatuses and State Power
Public Policy (New York: Elsevier). under Feudialism, Capitalism and Socialism
Self, P. (1977), Econocrats and the Policy Process: the (London: Verso).
Politics and Philosophy of Cost-benefit Analysis Winter, S. (2003), Handbook of Public Administration,
(Basingstoke: Macmillan). edited by B. G. Peters and J. Pierre (London:
Self, P. (1999), Rolling back the Market: Economic Dogma Sage).
and Political Choice (Basingstoke: Macmillan).
Peter-3383-Index.qxd 6/8/2006 7:45 PM Page 493
Index
A Agriculture cont.
EU, 310
Academic study expansion food, 309–321
of public policy, 1 food safety, 316
Acid rain government intervention, 313
United Kingdom, 470 and international affairs, 122
Across the board cuts, 142 market structure, 312–313
Action choices multifunctionality, 319–320
aiding identification, 26 narrative policy, 315–316
Action levels, 22 obesity, 318–319
governance, 13–28 policy drivers, 312–315
Action situations, 23 production politics, 309–312
Actors, 22 public policy, 315
Adequacy technology, 312
income security, 189 Agriculture Adjustment Act of 1933, 300–301
Ad hoc budgeting, 140–144, 147 Agriculture industrial policy
Administration Japan, 300
and policy, 118–119 Agri-food industries
Administrative coordination, 118 price relationship, 314
vs. policy coordination, 118 Air services
Administrative layers, 23–24 benefits, 331t
Administrative reform, 121–122 Western Europe, 323
Administrative system Allison, Graham, 342
internal fragmentation, 119 Allocation efficiency, 419
Adverse selection, 61 Alternative analytical frameworks, 20
Advisory committees Alternative general analytical frameworks
horizontal policy coordination, 130 policy process, 24t
variants, 130 America. See specific country
Advocacy Coalition Framework, 154 American Heritage Dictionary, 265
Advocacy coalition framework, 17 Amsterdam Treaty, 124
Agency Analytical frameworks, 14
autonomous or semiautonomous creation, 122 Analytic supply, 36
with portfolios Anglo-American
horizontal policy coordination, 130 OECD, 444
structure and process Anglo-Saxon approach, 333
effects on regulatory environment, 63 for transportation infrastructure, 327
Agency theory Anticipatory policy analysis
descriptive critiques, 63–64 policy design, 78–79
globalization, 388 Apples
and policy implementation, 60–62 South Africa, 313
Agenda setting, 55–56 Approach, 13. See also specific type
Aggressive neorealism, 346 Argumentative policy analysis, 461–477
Aggressive unilateralism anti-representationalism, 465
industrial policies, 301–302 causality questions, 467
offensive policies, 296 epistemology, 464–465
Agreement European philosophy, 461
accomplishing, 100 genuine argumentation, 473
Agriculture language, 464
consumer preferences, 314–315 Majone’s definition, 468
environmental policy, 319–320 methodology, 464–465
erosion, 311 Neo-positivist epistemology, 465
Peter-3383-Index.qxd 6/8/2006 7:45 PM Page 494
INDEX 495
INDEX 497
D Diffusion-oriented strategy
public policy evaluation, 413, 414f
DARPA. See Defense Advance Research Projects Dilulio, John, 371
Agency (DARPA) Direct democracy, 38
Data sources Direction setting
budget changes, 141 pension reform, 192
Decentralization Directive governance, 22
education policy, 236–237 Dirigiste central planning, 77
Decentralized centralism Discipline breakout
British Columbia, 237 policy design, 90
Norway, 237 Discursive Democracy, 472
Decision making, 55, 56 Discursive economy, 472
characteristics, 139 Divested companies
parsimonious model, 59 economic capital, 389
public policies, 488 Divestment
shift, 57 bandwagons, 387
Decision Making Through Mutual Adjustment, 100 company annual revenue generation, 383
Deck stacking thesis, 62, 63 how-type, 389–391
Defense Advance Research Projects Agency legislature, 384
(DARPA), 296, 304 media, 391
Defensive neorealists, 346 methods, 384–385, 384t
Defensive policies mixed-scanning theory, 390
protectionism, 295 models, 385–386
Deficits objectives, 383t
causes, 145 ongoing control, 385
reduction targets, 144–147 politically-driven reform, 390
Defra. See Department of Environment, Food, and proceeds, 384–385
Rural Affairs (Defra) public choice theory, 386
DeGaulle, Charles, 269, 273 public interest, 391
Delegation and control rational-comprehensive theory, 389
organizational decision making, 60 targets, 383–384
DeLeon, Peter, 17–18, 19 theories, 385–386
Deliberationists, 87 why-type, 386–389
focus, 88 Divide the dollar game, 58
vs. institutionalists, 88–89 Dror, Yehezkel, 16, 99
Deliberative-democracy case Dror optimal model, 99
goods, 403 Dukakis, Michael, 375
Deliberative policy analysis, 107–109 Dynamic, fluid, and loosely jointed
Democracy, 37–38 policy process, 14
education policy, 236–237
survival, 38 E
Democracy and Market System, 110
Democratization Earned Income Credit
elitism, 270–271 United States, 291
Demographic groups Eastern Europe
horizontal policy coordination, 131 OECD, 250
Department of Environment, Food, and Rural East Germany
Affairs (Defra), 311 social security spending, 215
Dependent variables Easton, David, 102, 488
budget changes, 142–143 EAZ. See Education Action Zones (EAZs)
Descriptive critiques Ecological sustainability, 441
agency theory, 63–64 cultural goals, 440
Deutsch, Karl, 102 economic goals, 440
Dewey, John, 232 social goals, 440
education policies, 233 Economic(s)
Deweyan pragmatist, 109 and deficits, 145
Dialogue of deaf, 14 equilibrium, 420
Dictator form goal-achievement, 406
divide the dollar game, 58 rationality, 483
Peter-3383-Index.qxd 6/8/2006 7:45 PM Page 498
INDEX 499
INDEX 501
INDEX 503
INDEX 505
INDEX 507
INDEX 509
INDEX 511