Professional Documents
Culture Documents
Changes for
Democracy
LEONARDO MORLINO
..................
' '
Accountability Politics:
Power and Voice in Rural Mexico
Jonathan A. Fox
Regimes and Democracy in Latin America:
Theories and Methods
Edited by Gerardo L. Munck
Democracy and Diversity:
Political Engineering in the Asia-Pacifi.c
Benjamin Reilly
Democracy and the State in the New Southern Europe
Edited by Richard Gunther, P. Nikiforos Diamandouros,
and Dimitri A. Sotiropoulos
International Democracy and the West:
The Role of Governments, Civil Society, and Multinational Business
Richard Youngs
Democrarie Accountability in Latin America:
Edited by Scott Mainwaring and Christopher Weina
Democratization: Theory and Experience
Laurence Whitehead
OXFORD
UNIVERSITY PRESS
Contents
List of Figures
List of Tables
lX
Introduction
Part I The Basics
1. What Theory in Democratization Studies?
1.1 Theories of institutional change
1.2 A Retreat from theory?
1.3 What aie the possible theoretical developments?
2. Definitional Conundrums
2.1 Core definitions
2.2 Minimalist empirical definition
2.3 Main normative definitions
3. Are There Hybrid Regimef.
3.1 A widespread phenomenon?
3.2 Definition and analytic dimensions
3.3 Empirical cases of hybrid regimes, or sheer fantasy?
3.4 What kind of classification?
3.5 Where to go from here?
7
7
11
18
25
26
31
34
48
49
so
57
60
67
77
77
91
97
106
5. Domestic Anchoring
5.1 Do consolidation and crisis exist?
109
109
79
83
85
Contents
Vlll
5.2
5.3
5.4
5.5
6. External Anchoring
6.1 Methods of influence: what to consider?
6.2 Layers of irnpact and cycles of change: the framework
6.3 Testing the framework empirically
6.4 Key final remarks
112
119
128
132
144
144
150
160
179
191
191
194
195
196
197
205
208
211
215
216
8. TODEM Applied
8.1 Accountabilities and connections
8.2 From procedures to results and contents
8.3 Where are the qualities?
Appendix
240
249
258
9. Concluding Remarks
264
223
223
List of Figures
2.1 Most salient definitions of democracy
44
63
80
84
90
96
120
140
157
214
225
226
243
250
251
254
Bibliography
269
Index
255
293
256
257
257
Introduction
St Anthony's. I would like to thank Jan Zielonka for this possibility and all
consequent opportunities I had to work on the revisions.
I would also like to acknowledge very gratefully the comments on the entire
manuscript or on some chapter of it of a few colleagues and friends, including
Nancy Bermeo, Ian Carter, Donatella Della Porta, Michael Freeden, Timothy
Garton Ash, Morgan Uewellyn, Liborio Mattina, Angelo Panebianco, Riccardo
Pelizzo, Daniela Piana, Francesco Raniolo, Eugenio Somaini, Claudius
Wagemann, Laurence Whitehead, and Jan Zielonka.
Thus, looking back over these years of research, if I recall the famous
metaphor by the Greek poet Archilochus, quoted by Isaiah Berlin (1953: 1)
at the beginning of his essay on Tolstoy: 'the fox knows many things, but the
hedgehog knows one big thing', this work should show that my model is the
hedgehog. Whether the goal of knowing 'one big thing' has been even
partially achieved is eventually up to the reader to decide.
PART I
THE BASICS
1
What Theory in Democratization Studies?
For many years a key issue in comparative studies has been: how far can
democratic theory be developed while avoiding the double trap of the banality of over-general assertions, on the one hand, and the short-sighted view
of an important phenomenon, on the other hand, because of the limits and
constraints imposed by a specific analysis of either one case or few cases
within a limited time span? To deal with this topic we can start by asking: does
a theory of dem9cratization or of regime change exist? This is an issue that has
been around for decades in dassie political science literature, but it has been
recast here in the new context of the spreading of democracies since the early
1970s. A reasonable response to this question should consider two other
streams of Iiterature in addition to the one focusing on the most recent
democratization phenomenon alone. They consist of the attempts to find
theoretical propositions and ~podels at a high level of abstraction to explain
the institutional changes, ancf the quantitative analyses of a large number of
cases that basically develop the path-breaking work by Lipset in the 1950s
(1959). As we will see in the following sections, what immediately distinguishes these two groups of scholars from the democratization scholars is how
theoretical goals take priority in the two groups whereas there is a sort of
retreat from theory in the third group. To appreciate better these differences
the theoretical Iiterature and the empirical studies will be discussed first
before examining the theoretical contributions of democratization scholars.
The Basics
the analysis of critical junctures by Capoccia and Kelemen (2007) (just to cite
the best-developed ones), cannot be considered useful pieces of a theory of
democrafuation or of regime change for a number of reasons. One is their
very high level of abstraction. But the main reason is simply that they have
been devised with different research goals to those of transition to democracy
or crisis or other aspects of democratization. As their principal focus is
change, gradual or otherwise, in well-established democratic contex:ts they
1
cannot cope with key questions about democratization or regime change.
Although we will come back to some of these theoretical proposals in
Chapter 8, here we might immediately add that a useful and crucial notion
for understanding political change might have been that of 'critical juncture'.
But as Capoccia and Kelemen wisely state, critical junctures, even in the
moderate way they define them, i.e. as 'relatively short periods of time during
which there is a substantially heightened probability that agents' choices will
affect the outcome of interest' (2007: 348), are 'rare events' as 'the normal
state of an institution is either one of stability or one of constrained adaptive
change' (2007: 368) of the kind suggested by Mahoney and Thelen (see
above). In actual fact, as anyone who has clone field research in countries
where there has been a transition to democracy will be tempted to say: they
are very rare events, if they exist at all. This view is backed up by the literature,
which stresses that 'no politipl elites operate in a vacuum ... politicians ...
risk being displaced by a nef.r set of representatives if they do not meet the
interests and demands of their supporters' (Boix 2003: 9). Thus, at the very
least the suggestions ernerging from this stream of literatme should be
'translated' and mademorerelevant for the theoretical analysis of democratization (for more on this, see Chapter 8).
The second group of quantitative scholars who have developed relevant
and highly ambitious theoretical propositions are those who adopted a
quantitative strategy with a large number of cases. Vanhanen (1990, 1997),
in his search for 'the ultimate underlying factor of democratization' or 'the
common factor of democratization' (Vanhanen 1997: 25-6), first in 147
countries and then in 172 countries, developed an 'evolutionary theory of
democratization' where this factor is identified as the distribution of power
resources2 and democracy is measured by aggregating the degree of competition and that of participation. 3 lt is not relevant here to emphasize that this
seems to involve a quantitative reshaping of a dassie theory in the analysis of
democracy. In fact, the distribution of power resources is actually a different
way oflabelling econornic and social pluralism. Moreover, pluralism, competition, and participation have been the dassie loci of democratic theory at least
since Rousseau, Tocqueville, and several more recent authors such as Lipset
(1960), Dahl (1970), or Sartori (1987). Notwithstanding a few reasonable
The Basics
criticisms that can be made of the method used to measure pluralism and
democracy (for example, Norris 2008: 63-5), the study by Vanhanen, who is
well aware of the limits ofhis explanation (see 1997: 155-66), is a significant
preliminary step towards a more precise explanation of democratization.
Vanhanen's work can be complemented by the analysis of Przeworki et al.
(2000) on the connections between economic development and democracy,
though some of their most interesting observations concern the performance
of democracy and dictatorships in different domains. 4 Their key point shows
how 'strong and tight' the relationship is between economic development,
measured by per capita income, and democracy. More controversial is the
relative importance they attribute to this factor compared to others, such as
politicallegacy, social structure, cultural traditions, institutional framework,
and international context (Przeworki et al. 2000: Chapter 2, esp. p. 79). They
also clarify very effectively that there is no linear relationship between economic development and transitions to democracy that is compatible with a
wide range of economic Ievels and that there is a definite strong relationship
with the survival, if not consolidation, of democracy (Przeworki et al.:
98, 103).
Although a relevant early chapter ofhis book concerns 'a theory of political
transition', the issue Boix (2003) chiefly tackles is the conditions of stable
democracies, in addition to other less directly relevant aspects. He does so by
developing a model for the preferences and resources of social actors and
individuals, explained by three factors: the degree of economic equality; the
nature of economic assets, determined by their mobility; and the distribution
of political resources to repress or Outmanoeuvre opponents. There is no
doubt that Boix's model, which also recalls aspects explored by Vanhanen, is
one of the most successful attempts since Lipset to provide a socio-economic
explanation of democratic stability, with data going back about two hundred
years to the early nineteenth century.
The work of Acemoglu and Robinson (2006) can also be mentioned in this
review. These two authors propose a simplified theory of conflicts where the
role of socio-economic forces in the creation and consolidation of democracy
is analysed. A set of game theoretical hypotheses is developed with reference
to the strength of a well-organized civil society, the impact of economic crises,
the sources of elite income, the Ievel of inter-group inequality, the role of the
middle dass, and the positive irnpact of economic globalization complemented by non-majoritarian democratic institutions. The weakness of this analysis is that the key hypotheses are not really tested. In fact, besides a general
quantitative picture of existing democracy, basically in the 1990s, there is a
cursory analysis of four cases in different moments (Great Britain, Argentina,
Singapore, South Africa) and with different paths to democracy. In short,
there is an unacceptable disparity between the theoretical apparatus, a mixture of common sense and clever observations, and the empirical support. It
seems ~ excellent example ofhow not to do good empirical theory. 5
The 'revised theory of modernization' by Inglehart and Welzel (2005)
complements the previous works by adding a better knowledge and understanding of economic aspects in connection with the cultural changes that are
so important in democratization. More precisely, with the essential contribution of the results of several analytic surveys, 6 the authors single out the
cultural changes brought about by socio-economic development, which
include higher individual autonomy, the emergence of self-expression values,
and other additional consequences for human development. The findings of
these and other works with sirnilar characteristics mark an irnportant step
towards explaining and developing a theory of regime change. However, these
contributions are more pertinent when the focus is on democratic survival,
consolidation, _and stable democracies rather than when it comes to dealing
with the transition to democracy or democratic deepening. 7 Moreover,
because of the approach and the research question, attention is actually
devoted to the phenomenon as a whole rather than to di:fferences between
cases and above all-and this is very irnportant for the perspective being
pursued here-to the different, fundamental, and most specific recurring
aspects within the macro-p:t;ocesses of democratization. Finally, the authors
of these works are themselves lucidly aware of the limits of an exclusively
statistical analysis. On the whole, their work can be regarded as a first,
necessary step, which requires a subsequent, deeper and more qualitative
analysis of the key aspects involved in the transition to democracy, consolidation, crisis, and an issue far removed from the perspective of these works,
namely, the deepening of democracy, as well as of the evolution and di:fferences between cases or groups of cases. In short, they are theoretical proposals
and significant research accomplishments. But they still do not answer the
questions that are irnportant for a deeper understanding of the whole phenomenon of democratization.
10
11
The Basics
adopted concepts, the quantitative and qualitative analyses carried out, as well
as the main empirical findings in the field. The main findings of the very full
account of democratization provided by Berg-Schlosser are as follows:
research conceming social classes as 'prime agents' of democratization is
inconclusive; the analysis of consolidation should be substituted by an analysis of stability for more focused and clearly formulated questions; the explanations of democratic transitions are extremely varied; the old proposition
about the association between the level of economic development and democratic stability is still very solid; the influence of international factors can be
strong; and, above all, there is a lack of a theory of democratization. 9 On
. international factors, specifically, they became a core feature of other analyses
of transitions, especially Eastern European transitions, through authors such
as Whitehead (1996), who points to the three mechanisms of 'contagion',
'control', and 'consent', or Linz and Stepan (1996: 72-81), who discuss the
salience of the foreign policies of other countries-the USA for one-together
with 'zeitgeist' "and 'diffusion', or those who have been doing research on
the enlargement of the European Union (EU) (see Pridham et al. 1994; but
also Pevehouse 2002; Schimmelfennig and Secleimeier 2005; Magen and
Morlino 2008; and others).
Several other subsequent contributions basically confirm the key point I
wish to make in this chapter and also clarify and specify the positions and
rationale of different authorl. Thus, for example, Whitehead (2002) establishes the background for an analysis of democratization as a long-term, nonlinear, open-ended process, but by setting up an original 'interpretavist'
approach, in which he also states that 'it avoids spurious rigour and untenable
claims of causal necessity' (Whitehead 2002: 34) and displays scepticism
about tne possible merits of such a theory. 10 O'Donnell et al. (1986) seem
to have been aware of this since the beginning, as already said above, and
McFaul (2002) echoed them by developing a contrasting hypothesis to the
one on pacts. Linz and Stepan (1996) demonstrate a similar awareness.
I might add here that the analyses proposed by these and other authors who
have worked on several cases in Southern Europe and Latin America focus in
particular on the following: the main characteristics of the previous regime;
the important role performed by 'pacts' or elite agreement on the institutions
to build (see also above); the 'resurrection' of civil society; the limited role of
political parties; the salience of contingent consensus on the institutions to be
set up; the enormaus uncertainties of the entire process of transition; and the
importance of the first, fodnding elections. Put differently, what all these
authors actually propose is a theoretical framework that points to key factors
to look at when conducting an analysis of one or a small nurober of cases. In
such a framework, actors, institutions, timing, and the very notion ofprocess play
12
13
14
The Basics
a central role in the analysis of countries in two geopolitical areas, i.e. Southern
Europe and Latin America.
The research led by Gunther et al. (1995), who arealso the chairmen of the
Committee on Southern Europe of the American Social Science Research
Council (SSRC), takes the same basic position with a definite nurober of
cases (Southern Europe only) and the proposal of a theoretical framework.
The main focus of the research they conducted was not to draft a nurober
of empirical theoretical propositions on the phenomenon in Italy, Spain,
Portugal, and Greece, but to elaborate a theoretical framework to scrutinize
the consolidation in those and supposedly other countries. In their view, the
historicallegacy, the differentiating elements of the previous non-democratic
regimes, the characteristics of democratic transitions, and the emergence of
partial regimes are the main domains that need to be analysed to ex:plain
consolidation.
With reference to democratization, Pridham (2000: esp. eh. 1) effectively
pursues a similar design by suggesting an appropriate framework for analysing mainly Southern and Eastern European changes as an overall phe:riomenon comprising historical determinants, modes of authoritarian breakdown,
formal regime transition, the role of elite actors, economic transformations,
the mutual influence of elite and civil society, the possible role of statehood
and national identity in the transition, and the impact of international
factors. All of these are aspects that should be considered together in the
'dynamics of transition' when analysing specific cases.
Apparently, Huntington (1991: 30) follows a different path by setting a
temporally and spatially defined explanatory goal ('explain why, how and
with what consequences a group of roughly contemporaneous transitions to
democracy occurred in the 1970s and the 1980s') for transitions to democracy
only. In his work, which can be paralleled to the quantitative analysis in terms
of the search for ex:planation if not for theory, he ex:plicitly mentions five
changes as the main causes of transitions in about thirty countries in those
decades: the legitimacy problems of previous authoritarian regimes, closely
connected to poor domestic performance; global economic growth during the
1960s; basic changes in the doctrine and activity of the Catholic Church;
the new policies of external actors (the EU, the USA, and the breakdown of
the USSR); and 'snowballing' or demonstration effects (Huntington 1991:
45-6). That is, within a multi-causal ex:planation, 11 a few cultural, economic,
and international aspects are considered the key factors and there is no theory
built on the results.
When a systematic explanation of Southern European and Latin American
cases is attempted (see Morlino 2003), the political traditions of the country
stand out as a key factor. More precisely, the key variables are: the
15
16
The Basics
proposal on other cases such as the dissolution of the USSR and the transition
to democracy in Poland and other Eastem European countries. Although
limited to the phenomenon of transition only, and in this case to a specific
kind of transition, that by pact or agreement, the suggested theoretical
perspective is worthy of consideration.
It should also be added that Spain is a case that has attracted considerable
attention from a nurober of scholars who have developed other theoretical
proposals on the basis of just that case. Share (1987), Fishman (1990), and
Gunther (see especially Gunther et al. 2004) are just some ofthe authors who
have focused on Spain in particular. Moreover, some of the theoretical frarneworks or propositions formulated for Spain have also influenced the analysis
of Latin American and even Eastem European cases investigated by other
authors, including Colomer (2000). The recurring references to pacts, the
moderation of elites, the resurrection of civil society, the salience of memory
of the past, and even some of the regional propositions mentioned by Bunce
emerge from the closer attention paid to Spain. But this is really a countryspecific analysis that cannot be extended far beyond that case, as a cursory
overview of other, different situations in the sarne and other areas would
immediately suggest.
In addition to theoretical frarneworks, able to cover more than one area
and time or applicable to country-specific mechanisms, there has been a sort
of intermediate set of proposals involving the development of models or
pattems of transition for a definite, usually small, nurnber of cases. Over
the years Stepan (1986), Karl and Schmitter (1991), Higley and Gunther
(1992), Munck and Skalnik Leff (1997), and Berins Collier (1999) are some
of the main authors who have proposed such models. Some of the differences
between them can be explained simply in terms of the different cases considered. For example, Stepan and Berins Collier also include the dassie Western
European cases of the past in addition to Southem and Eastem European
ones; Karl and Schmitter and also Munck and Skalnik Leff encompass the
Latin American transitions as well as the Eastem European ones of the early
1990s. The similarities between these attempts lie in the fact that all the
authors quoted above mainly focus on two macro-variables: the actors of
transition, whether authoritarian incurnbent elites or those of the opposition,
and the strategies they pursued, either accommodating or conflictive.
Thus, Higley and Gunther (1992) singleout two main pattems of democratic consolidation. Emphasizing the key role of elites, their structural
integration, and consensus on a set of values, the two pattems are: (1) there
is consolidation from the beginning if the elites consciously choose accommodation and cooperation in the democratic installation (settlement); or (2)
there is consolidation when, following democratic installation, divided elites
17
gradually converge by accepting the electoral rules and fair competition and
by takingmore ideologically mod~rate ~ositions (convergen.ce). In this pr~
posal the salience of ~e ~ul~ural di~~ns10~ must be emphasiZ~d? althoug~ 1t
has to be considered m 1ts mtertwmmg With .the effect of polic1es and With
other structural aspects. By contrast, Munck and Skalnik Leff (1997) come up
with four models: 'revolution from above', if the actors of transition are the
authoritarian elites who pursued a conflictive strategy of confrontation;
'conservative reform', if those elites chose agreements and compromises;
'social revolution', if counter-elites were at the heart of transition and pursued
a conflictive strategy; and 'reform from below', if counter-elites at the heart of
transition adopted an accommodating strategy. The advantages and limits of
such models are fairly evident and connected. One of the main points is that
the most immediate understanding of a country is counterbalanced by a
strong sirnplification of many relevant aspects. In addition, the adoption of
mixed models is very common. Consequently, strong simplification is accompanied by a loss of theoretical efficacy that would have been to some extent
rescued with the 'pure' models. On the whole, despite a nurnber of attempts,
the 'impossibility' of a general or regional theory for dealing with the transition to democracy or to some other regime is confirmed.
It seems, then, that a general theory of regime change or democratization,
even one that is temporally bound, will not suffice if we really want to analyse
and explain the cases on which we are working. When focusing more specifically on the theoretical choices made by a large number of authors, a retreat
from theory or the fear of developing a theory seems the dominant mood.
This is not only a retreat from orthodox, predictive kinds of theories, reminiscent of the 1960s (e.g. Kaplan 1964), but also from explanatory, comparative theory, and the actual explanation comes with a reconstruction of every
case as a whole with its uniqueness. Within this choice, of course, there are
various more specific solutions and positions, from utter scepticism about the
possibility or even the desirability of theory, to a more self-conscious choice of
a theoretical framework by a few distinguished senior scholars in the field, so
as to reconstruct pattems that encompass one or very few cases, through to
proposals of country-specific models. Overall, this retreat from theory appears to stem from two different factors, which ultimately move in the sarne
direction.
The first is the result of a conscious reflection on the failure of general
functionalist theories, systems analysis, formal rational choice, and other
general theories, which were in 'fashion' in the 1950s, 1960s, and early
1970s, but, when submitted to empirical tests, displayed all their analytic
and explanatory flaws and were practically abandoned or subjected to a major
overhaul, with much better results, as happened for rational choice theory. 12
19
The Basics
This has led to an evident search for different, maybe less ambitious, but more
empirically solid, theoretical choices and the consequent results. The consciousness of this failme and the new direction has been thoroughly charted
by several scholars, especially by Ostrom (1982: 13, 26) when she stresses the
need for 'the development of theory' in political science and at the same the
fact that what 'we do achieve will be lirnited in scope to specific types of
theoretically defined situations rather than sweeping theories of society as a
whole'.
The second factor derives from an awareness of the complexities and major
differences between various cases around the world. Such differences ex:tend
to the temporal plane as well, even in the arc of a lirnited forty-year period
(1970-2010) during which several profound social and economic changes
have taken place internationally. 13 Such awareness is additionally strengtherred by the evident fact that in most comparative political research conducted
in recent decades, democratization has been the dominant leitmotiv, following in this respect the spectacular developments of reality: when the world's
five main geographic areas are considered, what irnmediately stands out is
that hundreds of articles and books have been written on the topic in English
alone, not to mention the ones in Spanish and various other languages.
Therefore, reflection on past failmes as well as an awareness of empirical
complexities and !arge differences among the cases, emphasized by a very
extensive literature, suggest great caution in articulating theoretical ambitions
and choices. Furthermore, the irnpossibility of theory seems to lie at the core
of the very phenomena to be ex:plored, and as such is effectively captmed by
March and Olsen (1989: 65-6) when they write: 'institutional change rarely
satisfies the prior intentions of those who initiate it ... Change cannot be
controlled precisely ... there are frequently multiple, not necessarily consistent, intentions ... intentions are often ambiguous ... initial intent can be
lost'. If the 'carnation revolution' triggered by the golpe of Portuguese
captains and a nurober of other examples are considered, the words of March
and Olsen become very telling, and additionally strengthen the point just
made above.
18
at least three more precise, complex phenomena (transition to and installation of a democracy, consolidation, and deepening), and their contrary
(transition from democracy to authoritarianism or other regime, crisis, and
worsening), each one requiring its own empirical analysis with different
possible theoretical results. Thus, transition to democracy requires a different
analysis than consolidation, as in this second macro-phenomenon the range
of variation is unavoidably smaller because of the already existing institutional constraints; research into qualities' deepening or worsening, where the
implementation of democratic values-yet another aspect-is on the spot
entails a different analysis vis-a-vis transition and consolidation. This enables
us to better understand why tt.e authors working on democratic transitions
were more sceptical about thl possibility for theory development than those
studying consolidation, even when the same areas and the same or similar
spans of time are considered. The key consequence of these considerations is
that the different and more specific phenomena requiring analysis may have
different theoretical goals. Such theoretical pluralism emerges simultaneously
from the literatme and from direct research ex:perience. 15 Although the
problern of how to achieve theoretical advances may still be considered
unsolved, at least it becomes more empirically accessible.
The problern was addressed in a different way by rational choice scholars,
who considered theoretical achievements as a priority and solved it by
focusing on 'ex:planatory mechanisms'. Elster (1989: esp. 9-10), in particular,
ex:plicitly states that the key theoretical goal should be singling out ex:planatory mechanisms of 'human action and interaction' as recurring 'ways in
which things happen'. In addition to rational choice analyses of transitions
only (see esp. Colomer 2000, but also Przeworski 1986) and other analyses
focusing on Spain (see Cololl},er 1995), this is also a path undertaken by other
comparativists working in areas other than democratization, such as Pearson
(2004), who ex:plicitly shares Elster's point, or Tsebelis, with his analyses of
nested games (1990) and veto players (2002). The appropriateness of such a
proposal is that it may allow theoretical advances while at the same time not
20
The Basics
21
being at so high a level of generality that our Statements become platitudes or
Moreover, such a starting point helps to clarify how the oft-proposed
pompous affirmations of the obvious, a trap which-as is well known-a
distinction between 'structure' driven and 'process' driven explanations (e.g.
number of rational choice contributions were unable to avoid.
Kitschelt i992), which are usually focused on transition to democracy only,
If the 'solution' suggested by rational choice is accepted with its stress on
can be overcome: different interactions among actors and structures, to be
theoretical priorities, looking for causal or explanatory 'mechanisms' still
considered as salient contextual variables, are recurring elements of analysis
leaves open some important issues. First, despite the broader formulation
within transition or consolidation processes. 16 There is also no doubt that not
by Elster, the core meaning of 'mechanism' always brings to mind some
only is there a random component in the actual unfoldirig of those macrocombination of cams, gears, belts, and chains or at least a set of links or
processes, but they can be open-ended, often convoluted, and never teleologconnections designed to achieve a certain outcome. That is, a sort of deterical, as Whitehead (2002: 238ff.) rightly stresses. To better understand this
minism comes with the term and this is unacceptable in o_ur topic for
point it suffices to recall that an individual or collective action or set of
everything empirical research has shown in these years, if not for another
actions-for example, an implemented political strategy-can be teleologimethodological reason, namely that attaching some sort of determinism to
cally driven, but a process by itself cannot be such because it unfolds through
this term mutes a basic feature of democratization phenomena: they are
several, often unexpected or unwanted interactions, even among different
'open-ended' changes. Second, in all phenomena of democratization time,
strategies, within a given or changing context with again sometimes unextiming, sequences, and identification of time-bound windows of opportunity are
pected, unwanted results. 17
key aspects to analyse (see esp. Linz 1998; Schedler and Santiso 1998; .SchmitOn the whole, these considerations suggest that advances in theory are
ter and Santiso 1998) and, although possibly present within the notion of
indeed necessary. But to achieve such results, first, the whole, too-broad topic
mechanism, time does not lie at the core of this notion. Despite what Pearson
of democratization has to be broken down into more palatable, albeit still
affirms about mechanisms that are or should be 'temporally oriented' (2004:
thorny, phenomena or processes. Those processes include at least: transition
7, but also 1-16 and 54-78), the same author adopts the term 'process' when
towards democracy, consolidation or crisis, and democratic qualities' deepthe time to be considered is a long one (Pearson 2004: 79-102).
ening or worsening. Second, in
connection with the reality that is analysed
l
Third, however, when conducting empirical field research it is not always
and the data we are actually able to detect and collect we have to be ready to
possible and is often difficult to gather consistent (fairly) complete data
accept different theoretical achievements. Thus, within the macro-processes
that are time bound. Thus, all considered, singling out empirical mechanisms
mentioned above, we can come out with key, salient, and recurring more
is a first, theoretical necessary step. But in addition we should embed the
circumscribed sub-processes, with simpler theoretical frameworks, with
mechanism/s we are able to find into a meaningful 'process', where time,
recurrent patterns or other theoretical results when the former are not
timing, and sequencing, when singled out, are essential components.
achieved just because that comes out as an impossible task. The consequent
To be more precise, at the core of our theoretical research there is the
empirical question is: what sort of more specific theoretical results are possisingling out of a 'process' as a 'set of recurring interactions among individual
ble when analysing transitions to democracy, consolidations and crisis, and
and collective actors within changing structures, which is spread out over
democratic qualities' deepening or worsening?
time, may or may not unfold in an expected result, is on occasion unilinear,
In order to reply to this question and to discuss those theoretical results, a
but is always open ended'. Inside this definition of process there is room for
step-by-step strategy seems_necessary. Accordingly,
mechanisms minimally defined as 'recurrent links or connections'. These
- first, an empirical definition of democracy that is suitable for conducting
definitions help to overcome a possible objection by Vanhanen (1997: 26)
'
empirical research into the different democratization macro-processes,
and other scholars, who stress how 'process-oriented analysis resorting to
including
the pertinent normative aspects, has to be proposed;
various proximate factors cannot lead to any general theoretical explanations,
second,
the
issue of how to define and empirically grasp ambiguous,
although they may produce useful descriptions of democratization': as a ,
intermediate
Situations has to be addressed;
r
general theoretical explanation is actually impossible, as shown by empirical
third,
these
two
steps
should
pave
the
way
for
delineating
more
precise
research in these years, singling out key processes and related mechanisms,
indications to detect when there is a change of regime or not, with a full
conceived as above, is the best theoretical achievement we can actually obtain.
The Basics
awareness of all the lirnitations and constraints this involves (see also
Whitehead 2002: eh. 1);
- fourth, when necessary, empirical definitions of the macro-processes to
analyse should be provided (transition towards democracy and installation,
consolidation and crisis, qualities' deepening and worsening);
- fifth, where feasible and with all necessary distinctions between the different macro-processes, the more specific theoretical results should fie singled
out with possible limits and constraints in terms of time and area 'travel
capabilities' as weil as with regards to mutual exchange or influence
between factors that are mainly domestic and others that are mainly of
international origin;
- sixth, and this is especially salient for democratic qualities' deepening or
worsening, to single out the different aspects of and connections between
the dimensions of deepening.
freedom, justice, a better life, and a fairer society. If democracies do not work
better to contain crime and corruption, generate economic growth, relieve
econornic inequality, and secure justice and freedoms, sooner or later, people
willlose faith and embrace (or tolerate) other-non-democratic-alternatives.' From our perspective this means that it is especially important to have
the necessary analytic tools to check whether, how, and to what extent the
people's needs have been achieved, or not. And the analytic tool mentioned
above could be useful for that purpose.
With the goals that have just been set up, in the next chapter it will be
necessary to take a stand on, if not to solve, the contested issue of the empirical
definitions of democracy from the perspective of democratization. In the same
chapter the need to cope with the empirical 'translation' of normative conceptions of democracy is spelled out, although it will be necessary to come
back to this in the third part ofthe book (esp. in Chapter 7).
22
Moreover, the analysis of the qualities of democracy should at least cope with
two additional issues. On the one hand, it should be able to detect and
empirically relate citizens' or people's demands or needs and politicians'
decisions and implementations with actual impacts. This has been one of
the key flaws of old general theories, such as systernic analysis or functionalism, which rightly addressed this issue, but failed to empirically detect those
relationships in adequate ways. On the other hand, research on that topic can
turn into a salient tool for subsequent analysis of the qualities of democracy
and the aspects capable of deepening them. If adequate, the development of
such a tool would be a more important result in terms of its possible future
applications than the substantive empirical results presented and discussed in
the following chapters. In fact, the implementation of such a tool could enable
us to better understand the directions in which contemporary democracies
are heading.
As known, in the three Southem European countries the transitionwas the
result of the 'natural' exhaustion of long-term established authoritarian
regimes (Spain and Portugal) or of an unconsolidated military regime
(Greece) in the face of new problems and the transformation of societies in
that part of Europe (see Morlino 1998). In Latin America the end of military
regimes came about due to economic failure or even to econornic success, but
with unacceptable, unbearable costs-if protracted-in terms of suppression.
In Eastem Europe the end of the communist regimes happenecl because of
econornic bankruptcy and the end of hegemony in the region of the USSR.
The indirect consequence of this is that unlike authoritarian solutions, as
Diamond (2008: 294) eloquently affirms, 'democracies .... must demoostrate
that they can solve governance problems and meet citizens' expectations for
23
Notes
1. See below to see what these are more precisely.
2. Such a distribution of economic, intellectual, and organizational power resources
is measured by urban population, non-agricultural population, number of university students, literacy, arb of family farms, and decentralization of nonagricultural economic resources (see Vanhanen 1997: 42-60). Most of the data
cover the early 1990s, but for some variables an earlier period, 1850-1980, is
considered.
3. Competition and participation respectively measured by the smaller parties'
share of the votes cast in parliamentary and/or presidential elections, calculated
by subtracting the percentage of votes won by the largest party from 100; and by
the percentage of total population who voted in the related election (see also
Vanhanen 1990: 17-24).
4. They especially include economic growth aspects, such as consumption, investment, labour force, distribution of income, and demographic issues such as birth
and death rates, fertility, and life expectancy. The data cover the 1950-90 period
and the whole 141 countries.
5. See a similar opinion expressed by Houle (2009) and the consequences for other
aspects of their analysis.
6. The empirical source is a database built on the World Value Surveys and European
Values Surveys, that is, a number of different surveys conducted in several
COuntries between 1980 and 2001, but some of them arealso earlier (see Inglehart
and Welzel 2005).
7. Although there is some reference to crisis in Przeworki et al. (2000).
24
The Basics
8. The first four chapters of the book by Gill (2000: 1-123) are also a good, balanced
2
Definitional Conundrums
Definitions are necessary compasses for empirical research. As Collier and
Adcock (1999: 562) remind us, 'concepts, definitions, and operationalization
may evolve with changes in the goals and context of research', and goals and
context of research may change because key events and related processes have
been taking place. In addition, there are terms, such as 'democracy', which
have at the same time an empirical reference and a normative, ideal connotation. Consequently, for decades 'democracy' has been an essentially contested
concept. 1 Thus, we have a widespread phenomenon of democratization with
about ninety political regirnes labelled in 2008-9 as free and democratic, and
an additional sixty cases of partially free regimes (see Freedom House 2008),
and thus the need to revise the definition of democracy to adapt it to the
renewed research goals. As the main goal is the analysis of a few macroprocesses of change, such as bjp.nsition towards democracy, consolidation, or
crisis, and democratic qualiti~s' deepening or worsening, the definition/s of
democracy we need should help to lay the foundations for that analysis.
Hence; first, it is necessary to spell out the core definition of democracy.
This is especially important given that this kind of regime has become 'the
only game in town' and is no Ionger challenged as such. The procedural
definitions so irnportant in a different historical period as the most empirically solid point of reference within the liberal tradition have to be complemented by other definitions referring to substantive aspects relevant to
empirical research. Second, a minimalist definition is essential for understanding when in a transitional process a regime turns into a democracy, or is close
to doing so. Such a definition should also be able to capture the complexities
of a transitional period when a regime may already be democratic in some
respects but continues to remain authoritarian in others. Third, if there are so
many democracies, an empirical analysis of the actual irnplementation of the
main democratic values or tenets, or better an analysis of the quality/ qualities
of democracy, seems obvious.,2 This, however, implies some standard or ra sort
of maximum definition of democracy as a frame of reference in the development of quality. These three sets of definitions stillleave open the definition of
all striking cases of uncertainty and ambiguity; those labelled by Freedom
26
The Basics
The first conundrum is characterized by the long and heated debate in past
decades about democracy as 'form' and as 'substance'. It started with the
critique of nineteenth-century liberal doctrine by socialist thinkers and politicians, and acquired fresh momentum during and after World War Two,
when 'democracies' were considered by people and scholars to be the Western
political regimes, but the same label was also applied to the so-called real
socialisms or people's democracies of the USSR and the other Bastern
European countries. The prevailing conclusion ofthat debate in democratic
theorywas to regard the definition of democracy as form or, better, procedure,
as preferable to democracy as substance (see below). The notion of procedural
democracy was the most important attempt to give a solid theoretical ground
to the definition. Such a conclusion was based on the so-called non-reversible
relationships between freedom, understood as civil and political rights, and
equality. In the light of the experiences of Western Europe vis-a-vis Bastern
Europe the key point to emerge was that no kind of equality can be achieved if
there is no actual guarantee of civil and political rights as a prerequisite (see
Sartori 1987). The consequence of this has been not only a growing shift
towards substantive aspects in research, which in our field has been mirrored
by enormous developments in policy studies, including welfare rights, but also
to political equality as suggested by Dahl (2006) and to political justice as
developed by astring of authors, from Rawls (1974) toSen (2009), and to the
ways of complementing equality and freedom by giving to democracy a
substantive, important content, as in Ringen (2007). 3
The procedural definition, however, has remained salient and deserves a
closer look. It was initially proposed by Schumpeter (1942: 269, 1964: 257), to
which several other authors began referring: 'The democratic method is that
institutional arrangement for arriving at political decisions in which individuals acquire the power to decide by means of a competitive struggle for the
people's vote.' A few years earlier and again basically in the same period,
Kelsen (see Kelsen 1981: 190-1, and 1988) 4 had stated: 'as method or procedure, democracy is a form. In fact, the procedure through which a social order
is devised and implemented is considered formal, to distinguish it from the
Definitional Gonundrums
27
28
29
The Basics
Definitional Gonundrums
30
The Basics
parliament or are the result of direct election by the electorate; a set of intermediary structures represented by political parties and interest groups; apart from
all the bureaucratic structures (public administration, magistracy, armed forces,
police, etc.) whlch continually interact with the directly elected democratic
structures. These institutions and norms presuppose, albeit to different degrees,
a genuine guarantee of political rights and liberty, such as freedom of e:xpression,
union, and association and alternative sources of information and therefore also
the e:xistence of other norms and of a bureaucratic apparatus that guarantees
such rights. According to Schmitter and Karl (1993: 45-6), it is also important
that these institutions (and, one might add, these rights) arenot subject to or
conditioned by 'non-elected parties' or exponents of other external regimes. In
the first case, one example might be the desire of the armed forces to influence
democratic decision-making processes or the overall functioning of a democracy. In the second, the conditioning might come from an external power that
undermines the independence and sovereignty of the democracy in question.
The former scenario would lie outside the bounds of the democratic genus in a
non-democratic or ambiguously democratic institutional arrangement. In the
second scenario, there may be democracy within the country, but not independence and sovereignty, namely sovereignty may be consciously surrendered to
achleve other important goals, as is happening in the case of the EU. The two
forms of conditioning may therefore have very different outcomes, and the
second one does not necessarily move in an anti-democratic direction.
The same phenomenon of democratization stresses the importance of
another additional perspective in the discussion about definitions. In fact, it
is relevant in the subsequent analysis of macro-processes of change to point
out a genetic definition of democracy as that set of norms and procedures that
stem from a compromise agreement for the peaceful resolution of conflicts
between politically significant social actors and other institutional actors
present in the political arena (see Przeworski and Wallertein 1982; Przeworski
1991: 26-34; see also Weingast 1997: 245-63). Moreover, this definition
differs from the ones mentioned earlier not only because it lays more stress
on the genetic aspect-how a democratic regime comes into being and
forms-but because it indirectly makes reference to certain basic socioeconomic conditions.
The genetic definition also helps us to gain a better grasp of a fundamental,
but often neglected or underestimated, feature of any democratic regime: a
democracy is a regime concretely characterized by rules and institutions that
moderate or offset different aspects. As the procedural definition makes clear
(see earlier), democracy requires the existence of some basic consensus on the
rules and at the sametime must accept dissent and conflict about the content.
It must accept the uncertainty of decisional results, but requires certainty
Definitional Gonundrums
31
about the rules so. that the uncertainty is relative and limited. It cannot,
for example, abolish the right to private property. It must apply the principle
of majority rule, the main decision-making rule to ensure the e:ffective
functioning of its decisional organs, but must protect the rights of minorities.
In some cases, therefore, it needs to obtain broader majorities or even
unanimity. There must be the broadest possible representation of interests
and identities in the appropriate government bodies, such as parliament
and regional or local assemblies, but at the same time a democracy relies on
some degree of decisional efficacy and functionality, which are much harder
if not impossible to obtain when seeking to achleve a representative presence
of a11 the components of fragmented and complex societies. Finally,
and parado:xically, the stronger the consensus about democracy, the more
conflict it can sustain; the smaller and more integrated the rninorities,
the more government institutions can resort to the majority principle;
the Iess complex the political structures, the greater the decision-making
efficacy.
r
This discussion helps to establish the bases for developing the definitions
that are more appropriate for our research. In fact, we need a narrow
empirical definition, indicated by several authors as 'minimalist', in order to
better analyse cases of transition as weil as of consolidation and crisis, but we
also need a 'maximalist' or rather a normative definition, which is necessary
in the analysis of the qualiti!s of democracy, that is in the processes of
deepening and worsening of qualities.
After the procedural and genetic definitions, it is now possible to retum to the
first of the two issues raised by the first definition. In order to conduct an
empirical analysis of the establishment of and transitions within democracies,
it is important to arrive at a minimalistdefinition specifying a limited nurober
of features that are more immediately controllable and essential at an empiricallevel, thereby permitring the establishment of a threshold beneath which a
regime cannot be considered democratic.
The three reasons why such a minimalist definition is empirically helpful
can be more explicitly stated:
r
The Basics
Definitional Gonundrums
32
The best minimalist definition is still the one inspired by Dahl several years
ago (1971: esp. eh. 1): a regime should be considered democratic if it has at
least the following: (a) universal male and female suffrage; (b) free, competitive, periodic, and fair elections; (c) more than one political party; (d)
different and alternative sources of information. Such a minimalist empirical
definition has been considered as a procedural one (see e.g. Whitehead 2002:
10).8 However, it must be noted that ultimately it considers genuine respect of
at least civil and political rights to be essential, assuming that these rights exist
if there is effective universal suffrage, the quintessential expression of political
rights, that is, if the right to vote extends to the whole adult demos; if, as a
consequence, there are free, fair, and regular elections, an expression of the
effective existence of freedom of speech and thought; if there is more than one
genuinely competing political party, a manifestation of the existence of a real
and practised right of association; and if there are different sources of media
information with different proprietors, proof of the existence of the abovementioned freedoms. 9 In other words, in this definition strictly procedural
aspects such as competition and participation are actually complemented by
an effective guarantee of freedoms where the substantive elements are also
present (see also Part III on this).
One important aspect of such a definition is that if just one of these features
is absent or ceases to exist, the regime concerned would no Ionger be a
democracy but some other political-institutional arrangement, possibly an
intermediate one characterized by differing degrees of uncertainty and ambiguity. Finally, it is again useful to stress that the minimalist definition must
focus on the characteristic institutions of democracy-elections, competing
political parties (at least potentially), media pluralism-thereby tying in with
the definitions given above, such as those of Dahl and Schumpeter, and
shifting the level of abstraction of those two definitions on to the more
immediately empirical plane of the institutions that are indispensable for a
democratic regime.
An important addition was made to this definition by Schmitter and Karl
(1993: 45-6; also see above), who stress that democratic institutions, existing
rights, and also the decision-making process should not be constrained by either
non-elected elites or external powers. When considered jointly, all six requested
characteristics are clearly demanding, and the four adjectives attached to elections
33
34
The Basics
Definitional Conundrums
35
36
The Basics
Definitional Conundrums
37
institutions fundamental for a society, the relations of Subordination stem16
ming from them, arid democracy. The bonds are based on dyadic relationships. Strong arid concrete inequalities are created within these institutions.
In their research Bachrach and Baratz affirm that citizen participation is a
form of political action that influences the allocation of values. Participatory
processes tend to redistribute power, authority, and influence in favour of
participants. In anti-poverty programmes, the involvement of disadvantaged
citizens is incentivized in order to arouse their political interest and, in this
way, to strengthen their ability to pursue their own interests. In the view of the
two authors, well-structured and guided participatory initiatives offer a
means of anticipating and dealing with conflicts. In this sense, it has been
observed that participatory practices can be viewed as technologies instrumental to the process of government. Such a perspective is effective ifbuilding
and maintaining consensus is considered a priority for the stability of democratic systems. There is no doubt that empirically participatory phenomena
change considerably according to the level focused on by research, be it statenational or local government. In the former, specific representative institutions and political participation in the circuits of representative democracy
remain absolutely central. Moreover, the solid tradition of research into
participation does not create unsurmountable problems in the empirical
assessment of this normative notitn, be it in one of the different conceptions
that are present in the literature, and which we have briefly commented on
here.
38
The Basics
through interaction, and may change quite significantly. The path taken to
reach collective decisions and build consensus and agreement about a choice
is very different to the procedures and mechanisms of representative democracy. On the one hand, deliberative democracy is based on a set of principles
that tend to be expressed in universalistic terms, especially in Habermas's
interpretation of public democracy as a rediscovery of 'govemment through
discussion' and in the work of other North European, Anglo-Saxon, and
American scholars. 17 On the other hand, in cantrast to the wave of participation that took place in the 1960s and 1970s, deliberative democracy grants
little or no space to spontaneaus impulses, to the forms of protest and the
claims of movements, as the dimension of 'bottom up' participation. Even in
the model proposed by Habermas, the organizational networks of civil society
,converge in the public sphere in a structured way, institutionalizing a kind of
ideal communicative community through practices of public deliberation.
These are the presuppositions of the current of thought that adopts a radical
interpretation of deliberative thought, defined as a 'process based on public
18
discussion between free and equal individuals'.
Moreover, from Elster's (1993, 1998) different perspective, another theoreticalline of deliberative democracy emerges, which stresses the 'strategic'
interweaving of argumentation and negotiation in deliberative processes and
presupposes a strategic result-oriented rationality. Butthisvision of deliberative democracy weakens the role of dialogue as a 'neutral' discursive practice
animated by universalistic ideals and offers space for differences in the
expression of vested interests and reaching agreements based on an instrumental adjustment of preferences rather than a profound change in points of
view where values would be affected. Comparatively, the Habermasian perspective is more radical in affirming the desire of participants to reach
agreement and the power of the best argument, and in the capacity to go
beyond the instrumental agreement, developing a genuine sharing of views
between the parties involved, and hence consensus, against a background of
communication inspired by a faith in dialogic rationality. The empirical
translation of such a notion, already attempted by a few scholars (see e.g.
Bchtiger et al. 2005), partially overlaps with the previous normative notion,
and its main difference is in the more specific aspects of participation,
especially in the decision-making process, which should be analysed to detect
how much and with what characteristics some deliberative features have been
implemented in the reality of certain democracies. In fact, Bchtiger, Steiner,
and associates who translated the notion ofHabermas pointed to the working
of parliaments. 19
In the associative democracy, as theorized by Hirst (1997) and others,
accountability, participation, and freedom are key elements, that is, there is
Definitional Gonundrums
39
40
The Basics
Definitional Gonundrums
41
Rule oflaw and freedom are the key empirical elements of good governance.
To better clarify this point, it should immediately be added that, within a very
lively and rich debate, the notion of governance has usually been grounded on
empirical analyses (see e.g. Pierre 2000). It refers to a complex set of structures
and processes, both public and private, which are not synonymous with
'government' (see Weiss 2000) and can be effectively C<?nceptualized as the
'capacity of public sector of steering, alone or in cooperation with other
actors, economy and society' (Peters, 2008: 445). 22 An excellent reconstruction is offered by Rhodes (1996), who stresses how with this notion there is
not a 'crisis' or the 'disappearance' of government and the state, but a
redefinition and repositioning of its role and function in a pluralized context
of actors taking part in public decision-making processes. If anything, it does
mark the end of the paradigm of government as a centralized, coercive, and
hierarchical form of power through a redefinition of the form of political
regulation in the interaction with social and economic regulations. Hooghe
and Marks (2001) classify forms of governance on the basis of the prevalent
definitions found in studies of the EU, international relations, federalism, '
local government, and public policies. From our perspective, the key feature
they draw attention to is the need to analyse different levels of government
even when considering normative conceptions of democracy. The discussion
could continue and doubts rlmain about what is the most effective definition
of this notion and the level niost appropriate for close analysis. But we are still
within an empirical debate.
The normative element intervenes when the adjective 'good' is added. In
this perspective one of the mostoriginal definitions (and analyses) of 'good
governance' is the one by Rothstein and Teorell (2008: 165-90), who consider
this notion as synonymous with 'quality of government' and strongly argue
that a key aspect of good governance is the 'impartiality in the exercise of
public authority' (p. 166) as a 'moral principle' for civil servants (p. 176) with
democracy as a necessary condition and impartiality implying rule of law as
well as enhancing efficiency/effectiveness (pp. 178-83). The World Bank
prefers a broader definition where 'good governance' is 'an efficient public
service, and independent judicial system and legal framework to enforce
contracts; the accountable administration of public funds; an independent
public auditor, responsible to a representative legislature; respect for the law
and human rights at all levels of government; a pluralistic institutional
structure; and a free press' (World Bank 1992). Or, in other words, once
again those of the World Baclc, there is a 'predictable, open and enlightened
policymaking (that is, transparent pro~sses); a bureaucracy imbued with a
professional ethos; an executive arm of government accountable for its
actions; and a strong civil society participating in public affairs; and all
The Basics
Definitional Conundrums
behaving under the rule oflaw' (World Bank 1994: vii). The notion has also
been adopted by other international organizations, and the most recurrent
empirical dimensions mentioned in different documents have been: rule of
law, government effectiveness, legitimacy and voice, political stability and
absence of violence, transparency, accountability, participation, responsiveness, and so on. 23 This means that ultimately we are not far away from other
normative notions when the empirical translation has been made.
A different normative path is to examine the main values that a contemporary democracy is supposed to implement with its policies. As already mentioned above, for several authors these are mainly freedom and equality/
solidarity. This can lead to a definition of a 'good' democracy as 'the set of
institutions that create the best opportunities to carry out freedom and
equality' or, in a more developed way, 'a stable institutional structure that
realizes the liberty and equality of citizens through the legitimate and correct
functioning of its institutions and mechanisms'. Thus, a good democracy is a
broadly legitimated regime that satisfies its citizens. 24 When institutions have
the full backing of civil society, they can pursue democratic values. If, by
contrast, institutions have to postpone their objectives and expend energy and
resources on consolidating and maintaining their legitimacy, crossing even
the minimum threshold for democracy becomes a remarkable feat. Second, in
a good democracy the citizens themselves have the power to check and
evaluate whether the government is pursuing the objectives of liberty and
equality according to the rule of law. They can monitor the efficiency of the
application of the laws in force, the efficacy of the decisions made by government, and the political responsibility and accountability of elected officials in
relation to the demands expressed by civil society. All this implies that the
different levels of government-local, regional, national, supranational (especially for European countries)-cannot be overlooked (see also Morlino
2004). The affirmation of the two principles of freedom and equality must
occur at the local (municipal or regional), national, and supranationallevels,
with all the implications and especially the limitations that this inevitably
entails (see Dahl 1994). It must also take into consideration the various
spheres of everyday life, from the workplace to the educational system and
associational activities. Consequently, any attempt to come up with an ideal
definition must adopt a perspective that starts with the individual and society
rather than with government institutions. To ensure a greater degree of
democracy, one cannot avoid looking first of all at the establishment and
extension of the rights of individuals or communities of varying sizes. If this is
the case, the next step is that the realization of freedom and equality, even
though these may be differently conceived, must necessarily be based on an
increasingly rich and extensive set of individual rights.
Dahl (1998), Beetharn (1999), and other authors help to explain the further
step required to arrive at a clarification and formulation of rights. In fact, they
suggest a n umber of more specific principles that are also means of affering
the best jnstitutional opportunities for ensuring freedom and equality. Above
all, these principles do not just include aspects such as recognition of the
political inclusion of all adult individuals, equal voting rights, the promotion
of the effective participation of all citizens, and the availability of clear and
correct information for all. Drawing once again on Dahl and Sartori, they also
encompass the promotion of the accountability of government and of its
capacity to respond to the demands of citizens and communities, and control
.of the decisional agenda and of the outcomes of such decisions on the part of
citizens themselves or in various other ways that may be devised.
If these aspects, which are essential for the achievement of freedom and
equality, are to be effectively pursued, contemporary democracies will also
have to attend to issues such as environmental conservation, the right to
healthcare, assistance for the elderly and disabled, the right to a job, provisions for the unemployed, the need to ensure everyone has a reasonable
staridard of living, the right to educational opportunities, and also the
promotion of equity in private disputes or between public and private interests. Not to include in an analysis of the ideal democracy the safeguarding of
the substantive elements outliped above would paradoxically mean ignoring
the steps already taken by many real democracies to promote equality. It
would also involve a failure to recognize that in today's world the protection
and promotion of those social rights have become just as indispensable for
democracy as the principles of inclusion, participation, equal voting rights,
accountability, and so on, and indeed are closely related to them. In short, it
would result in a definition of the ideal democracy that in some ways falls
short of what real ones have already achieved.
A concrete understanding of the affirmation and protection of the values
and rights mentioned above requires a third step, namely the pinpointing of
the institutional tools that best realize them. Once again, Dahl (1971: 3 and
1982, 10-ll) is useful here, with his emphasis on the need to have the
following eight institutional guarantees: (a) freedom of association and organization; (b) freedom ofthought (and expression); (c) the right to vote; (d)
the right of politicalleaders to compete for (electoral) support; (e) alternative
sources of information; (f) the possibility of being elected to public office
(passive electorate); (g) free and fair elections; and (h) the existence of
institutions that make government policies dependent on the vote ruid on
other expressions of preference.
The centrality of rights assumes the possibility that they may be extended
and deepened, and may be concretely exercised with greater satisfaction. One
42
43
44
The Basics
procedural
genetic
Definitional Gonundrums
45
Notes
minimalist
normative
- liberal, representative democracy
- responsive democracy
- participatory democracy
- deliberative democracy
- associative democracy
- egalitarian or social democracy
- good governance
- good democracy
Figure 2.1. Most salient definitions of dernocracy
example of this in relation to the right to vote might be the granting of the
right-which, indeed, has already become concrete reality in many
countries-to express a preference not only for a candidate or party but also
for a government. In this way, citizens would feel that they have more scope for
expressing their political will. Likewise, a more profound affirmation of
the principle of inclusion, and once again of voting rights, would involve
the granting of political citizenship to immigrant residents. Such a proposed
definition of good democracy is inevitably complex. It starts with the individual and develops in at least three steps: identification of the two most general
principles; preliminary delineation of more specific, intermediate principles;
and specification of the set of rights that stem from the first two steps. However,
the empirical translation should be simpler with the concrete emphasis on the
rights, the related content, and the ways of guaranteeing them.
To conclude this section and the chapter, it is worth stressing that empirical
notions of democracy, especially the minimalist one, are essential compasses
in the empirical research and the normative notions of democracy may be
empirically translated to assess their actual presence in the different countries
or political situations. But when we switch from the ideal discourse to
empirical analysis the enormous richness of different conceptualizations of
democracy can only be translated into a few recurrent, empirical features we
can explore. These include rule of law, electoral accountability, institutional
accountability, participation (the most common of the normative notions),
competition, freedom, equality/solidarity, and responsiveness. Chapters 7 and
8 will show such a translation and its consequences for the analysis of
democratic deepening. As a way of summing up the entire chapter, Figure
2.1 lists the most recurrent definitions of democracy that are useful for the
chapters that follow.
46
The Basics
Definitional Gonundrums
13. The term is used to refer to the 'practice or policy of making no more than a
minimal effort to offer minorities equal opportunities to those enjoyed by the
majority' (Webster Dictionary). According to Arnstein, the demands of minorities and weaker sections of society are managed through powerfully mediated
forms of listening designed to put a damper on conflicts in the interest of social
order, but which are not foilowed by any real measures to solve problems or
significant changes in the way decisions are made.
14. I would like to thank Francesca Gelli for calling my attention to this author.
15. The theoretical hypothesis in question had already been formulated by Bachrach
and ~aratz in two articles published many years earlier in the American Political
Science Review, entitled respectively 'Two Faces of Power' (1962) and 'Decisions
and Non Decisions: An Analytical Framework' (1963).
'
16. Pateman examines the institution of marriage in relation to the subordination of
women and in the ambit of work and employment, which creates bonds and
relations of Subordination particularly for women and the young. These institutions are viewed not so much in terms of relations between individuals, but as
social institutions--codes and conventions that take root in society and become
cognitive structures as weil as mechanisms of power.
17. The universalistic and normative tendencies are emphasized hereinordernot to
create misunderstanding, given that critics are prone to describe deliberative
democracy and its theoretical prernises as 'soft'.
18. The importance of dialogue lies in the comparing of views by the various parties
involved through discussion backed up by rational argumentation. This means
not merely asserting one's point of view, but publicly explaining the motivations,
in a situation where everyone enjoys equality and freedom of opinion.
19. On deliberative democracy especially there is a blossoming of mixed conceptions.
Among them those on 'liberal deliberative democracy' and 'participatory deliberative democracy', which Della Porta (2011: esp. sect. 3) singled out in her
analysis of the connections between these notions and social movements, can
be noted.
20. In the 'associative network' of civil society, people make decisions that may be
minor, but are capable nonetheless of reducing the distance that otherwise forms
between citizens and the state, and even the iniquitous outcomes of the market
can be lirnited in this way (Walzer 1992: 99). The conclusion reached by Walzer is
that the state cannot last long if it is alienated from civil society, which at the same
time cannot be viewed as an alternative to the state (governance does not exclude
government). There are at least two fundamental reasons for this. Firstly, because
the organizations belonging to 'civil society' rely on the help of the state in various
ways (specific prograrnmes and services, financial resources, tax reductions and
exemptions, protection, and, above all, the perforrning of a regulatory function
on the part of the state, which acts as a guarantoi:) in order to realize their
projects. Secondly, because mechanical procedures and rigid bureaucratic apparatuses can take over in 'civil society' as weil, leading to radically unequal power
relations and reproducing conditions of injustice (Walzer 1992: 104).
21. In the scheme proposed by Hirst (1997), associative democracy is a system for
organizing political and sociallife that encourages self-govemance through voluntary associations. This helps, in his view, to tackle three fundamental issues: (1)
to guarantee the responsibility ef representative democracies; (2) to deal with
increasingly pluralist societies with diverging values; (3) to enable those who are
successful to make use of services that have all the advantages of being collective
without, however, losing their flexibility, and maintaining attention for disadvan
taged individuals.
22. Other definitions of governance are: minimal state (Rhodes 1996); corporate
govemance (Rhodes 1996; World Bank); new public management (see Mathiasen
1996; Terry 1998; Peters and Pierre 1998; Pierre 2000; Lynn 2006); socio-cybernetic
system (see Kooirnan 1993; Rhodes 1996); networks (Rhodes 1996; Peters 2008);
multi-level governance (Hooghe and Marks 2001); democratic governance
(Franck 1992); and global govemance (Carin et al. 2007). Some ofthem (minimal
state, corporate govemance, new public management, good governance, sociocybernetic system, self-organizing networks) arealso discussed by Rhodes (1996).
23. In addition to the World Bank the other international organizations that have
adopted the notion of 'good governance' include: United Nations Development
Programme (UNDP); United Nations Economic and Social Corninission for Asia
and the Pacific (UNESCAP); International Govemance Office (IGO); Canadian
International Development Agency (CIDA); and Austrian Development Cooperation and Cooperation with Eastem Europe (ADC). I am not going to discuss
why these organizations adop~ed the notion. I just wish to mention briefly that it
has been considered a way ofnot appearing to interfere in the domestic affairs of
the countries where they intervene in different ways. A similar notion has been
adopted by the EU, which defines 'good governance' in terms of five principles:
' [ ... J openness, participation, accountability, effectiveness and coherence [ ... ]
Each principle is important for establishing more democratic governance. They
unde;:pin democracy and the rule oflaw in the Member States, but they apply to
alllevels of government-global, European, national, regional and local' (European Commission 2001: 10). Later the notion of'democratic govemance' was also
proposed more explicitly: it 'spans a broad range of issues, such as respect and
promotion of human rights and fundamental freedoms, democratisation and
citizens' involvement in the political process, the rule oflaw and access to justice,
gender equality, human security, access to information, management of migration flows, access to basic public services, effective, transparent, responsive and
accountable State institutions, sustainable management of resources and promotion of sustainable economic growth and social cohesion' (European Commission
and Council General Secretariat 2009: 3-4).
24. Of course, such a definition overlooks a few dassie issues related to individual
satisfaction such as when individuals have incompatible desires. I thank Uaniela
Piana who called my attention to this, although I cannot develop the point here.
47
3
Are There Hybrid Regimes?
Finer (1970: 441-531) was one ofthe first authors to singleout the existence
of farade democracies or semi-democracies to indicate regimes that are no
Ionger authoritarian, but not yet minimally democratic and have institutions
that are recurrent in a democracy, such as a constitutional charter and
elections, but where the former is not actually implemented and the latter
are largely constrained. Other authors such as Rouquie (1975) and O'Donnell
and Schmitter (1986), who had a Spanish background and were working on
Latin American countries, Iabelied the phenomenon of ambiguous cases
dictablandas and democraduras. Thus, the notion of hybrid regimes has been
present in the dassie political science Iiterature for years. The basic change in
recent years is the sheer size and variety of the phenomenon.
As a consequence, the view of Croissant and Merkel (2004: 1) that 'the
conceptual issue of diminished sub-type of democracy ... has begun the new
predominant trend in democratic theory and democratization studies' comes as
no surprise. Nor does the assertion by Epstein et al. (2006: 556, 564-5) that
'partial democracies' 'account for an increasing portion of current regimes and
the lion's share of regime transitions', adding, however, that there is little available
information about 'what prevents full democracies from sliding back to partial
democracies or autocracies, and what prevents partial democracies from sliding
back to autocracy' and that 'the determinants of the behaviour of the partial
democracies elude our understanding'. Nor, finally, is it surprising that a variety of
Iabels have been coined for these regimes in addition to those mentioned above:
'ex:clusionary democracy' (Remmer 1985--6), 'serni-democracies' (Diamond et al.
1989), 'electoral democracies' (Diamond 1999; Freedom House), 'illiberal
democracies' (Zakaria 1997), 'competitive authoritarianisms' (Letvitsky and
Way 2002), 'serni-authoritarianisms' (Ottaway 2003), 'defective democracies'
(Merkel 2004), 'partial democracies' (Epstein et al. 2006), 'mi:x:ed regimes'
(Bunce and Wolchik 2008), to mention just some of the expressions and some
of the scholars who have investigated what is denoted here by the broader term of
hybrid regimes (see, in particular, Karl1995; Diamond 2002; Wigell2008). 1
In trying to gain a better understanding of the reasons for such attention it
is worth bearing in mind that complex phenomena such as democratization
49
are never linear, and cases of a return to more ambiguous situations have by
no means been exceptions to the rule in recent years. Moreover, cases of
democracies, even if minimal, going 'all' the way back to stable authoritarian
regimes have been much less frequent: 2 it is more difficult, though not
impossible, to recreate conditions of stable coercion once the majority of a
given society has been involved and become politically active in the course of
transition. If nothing else, as Dahl noted many years ago (1971), greater
coercive resources would be required. Furthermore, in periods of democratization, even if only as a result of an imitation effect, authoritarian crises and
the resulting initial phases of change should be more frequent. Consequently,
there are several reasons for the greater frequency of regimes characterized by
uncertainty and transition. Moreover, if the ultimate goal is to examine and
explain how regimes move towards democracy, it is fully justified, indeed
opportune, to focus on those phases of uncertainty and change. But here we
have a relevant aspect that-this time surprisingly-the Iiterature has failed to
address and solve: when considering those phases of uncertainty and ambiguity, are we dealing with an institutional arrangement with some, perhaps
minimal, degree of stabilization, namely a regime in a proper sense; or are we
actually analysing transitional phases from some kind of authoritarianism (or
traditional regime) to democracy, or vice versa?
Thus, starting with definitions of the terms 'regime', 'authoritarianism', and
t
'democracy' we have to discuss and clarify the reasons for the proposed
definition of hybrid regime; try to answer the key question posed in the
chapter title, which, as will become clear, is closely bound up with the
prospects for change in the nations that have such ambiguous forms of
political organization and, more generally, with the spread of democratization; propose a typology of hybrid regimes; and, in the last section, reach a
number of salient conclusions.
The simplest and most immediate way of understanding the nature of the
phenomenon under scrutiny is to refer to the principal sets of macro-political .
data that exist in the literature. Data have been gathered by international
bodies like the World Bank, the OECD, and the United Nations; by private
foundations, such as the IDEA and Bertelsman Stiftung; by individual scholars or research groups, like Polity IV, originally conceived by Ted Gurr, or the
project on human rights protection undertaken by Todd Landman (2005),
The Basics
relatively speaking, is the broadest notion, whereas most of the others (e.g.
'exclusionary democracy', 'partial democracies', 'electoral democracies', 'illiberal democracies', 'defective democracies', 'semi-authoritarianisms') seem to
refer to more specific models, mainly ilirninished forms of democracy. And it
seems appropriate to come up with a precise definition of the broadest notion
before going ahead and mapping out the diversified realities inside it. 9 In
short, we are looking here for the genus that comes before' the species.
Moreover, an adequate conceptualization of the 'hybrid regime' must start
with a definition ofboth the noun and the adjective 'trapped' between a nondemocratic (above all, traditional, authoritarian and post-totalitarian) and a
democratic set-up. As regards the term 'regime', consideration will be given
here to 'the set of government institutions and of norms that are either
formalized or are informally recognized as existing in a given territory and
with respect to a given population'. 10 Emphasis will be placed on the institutions, even if they are not formal, that exist in a given moment in a given
nation. While they no Ionger configure some form of non-democracy and do
not yet configure a complete democracy, such institutions still bear traces of
the previous political reality. In addition, in order to have so methingthat may
be labelled as a regime we need an at least minimal stabilization. Fishman
(1990: 428) recalls that regimes 'are more permanent forms of political
organization'.n Otherwise, we 'pick up fireflies for lanterns' by confusing a
temporary changing situatio~ with a more stabilized one, whatever the
reasons might be. Obviously, the consequences of making such a distinction
are very significant. In any case, we will totally misunderstand the entire
situation if we fail to assess whether there is or has been stabilization or
not. This will be a key element of our definition, one that differentiates it from
those existing in the literature, even the mostprominent ones (see above).
The second point that can be stressed, then, is that a regime may not fulfil
the minimalist requirements of a democracy, in other words it does not meet
all the more immediately controllable and empirically essential conditions
that make it possible to establish a threshold below which a regime cannot be
considered democratic. As recalled in the previous chapter, for a minimalist
definition of democracy, we need at the same time: (a) universal suffrage,
both male and female; (b) free, competitive, recurrent, and fair elections; (c)
more than one party; (d) different and alternative media sources. Tobetter
understand this definition, it is worth stressing that a regime of this kind must
provide real guarantees of civil and political rights that enable the actual
implementation of those four aspects. That is, such rights are assumed to exist
if there is authentic universal suffrage, the supreme expression of political
rights, that is the whole adult demos has the right to vote; if there are free, fair,
and recurrent elections as an expression of the effective existence of freedom
50
Before proceeding with our analysis, it is worth stating that we prefer the term
'hybrid regime' to all the others present in the literature (see above) as this,
51
The Basics
of speech and thought as weil; if there is more than one effectively competing
party, demonstrating the existence of genuine and practised rights of assembly and association; and if there are different media sources belanging to
different proprietors, proof of the existence of the liberties of expression and
thought. One important aspect of this definition is that in the absence of just
one of these requirements, or if at some point one of them is no Ionger met,
there is no Ionger a democratic regime, but some other political and institutional set-up, possibly an intermediate one marked by varying degrees of
uncertainty and ambiguity.
Finaily, it is worth stressing that this minimalist definition focuses on the
institutions that characterize democracy: elections, competing parties (at least
potentially so), and media pluralism and it can be added that it is also
important that these institutions and rights should not be subject to, or
conditioned by, 'non-elected actors' or exponents of other ex:temal regimes
(see previous chapter and Schmitter and Karl1993: 45-6). The former refers
to the armed forces, religious hierarchies, econornic oligarchies, a hegemonic
party, or even a monarch with pretensions to infl.uencing decision-making
processes or at any rate the overall functioning of a democracy; in the second
case, a regime might be conditioned by an ex:temal power that deprives the
democracy in question of its independence and sovereignty by pursuing nondemocratic policies.
To avoid terminological confusion it should be pointed out that the
'electoral democracies' defined by Diamond (1999: 10) solely with regard to
'constitutional systems in which parliament and executive are the result of
regular, competitive, multi-party elections with universal suffrage' are not
minimal liberal democracies, in which additionally there is no room for
'reserved domains' of actors who are not eleetorally responsible, directly or
indirectly, there is inter-institutional accountability, that is the responsibility
of one organ towards another as laid down by the constitution, and finally,
there are e:ffectively applied norms to sustain and preserve pluralism and
individual and group freedoms (Diamond 1999: 10-11) .12 The term 'electoral
democracies' is also used by Freedom House with a similar meaning: an
electoral democracy is understood as a multi-party, competitive system with
universal suffrage, fair and competitive elections with the guarantee of a secret
ballot and voter safety, access to the media on the part of the principal parties,
and open electoral campaigns. In the application of the term by Freedom
House, all democracies are 'electoral democracies' but not all are liberal.
Therefore, even those regimes that do not have a maximum score in the
indicators for elections continue to be considered electoral democracies.
More specifically, a score equal to or above 7, out of a maximum of 12, is
sufficient for partially free nations to be classified as electoral democracies. 13
52
53
54
The Basics
55
with a 'weak judiciary'. But again this definition only refers to authoritarianisms that partially lose some of their key characteristics, retain some authoritarian or traditional features, and at the same time acquire some of the
characteristic institutions and procedures of democracy, but not others. A
hybrid regime, on the other hand, may also have a set of institutions where,
going down the inverse path, some key elements of democracy have been lost
and authoritarian characteristics acquired. Thus, it has to be adequately
completed, for example, by induding some of the aspects mentioned by
Levitsky and Way (2002: 52-8) in their analysis of a specific model ofhybrid
regime (competitive authoritarianism), such as the e:xistence of 'incumbents
(who) routinely abuse state resources, deny the opposition adequate media
coverage, harass opposition candidates and their supporters, and in some case
manipulate electoral results'.
This discussion, however, prompts reflection about two elements. First, a
hybrid regime is always a set of ambiguous institutions that maintain aspects
of the past. In other words, and this is the second point, it is a 'corruption' of
the preceding regime, lacking as it does one or more essential characteristics
ofthat regime but also failing to acquire other characteristics that would make
it fully democratic or authoritarian (see definitions above). Consequently, to
define hybrid regimes more precisely it seems appropriate to take a different
line from the one suggested in the literatme and to explicitly include the past
of such regimes in the definillon itself. The term 'hybrid' can thus be applied
to all those regimes preceded by a period of authoritarian or traditional rule,
followed by the beginnings of greater tolerance, liberalization, and a partial
relaxation of the restrictions on pluralism; or, all those regimes which,
following a period of minimal democracy in the sense indicated above, are
subject to the intervention of non-elected bodies-the military, above allthat place restrictions on competitive pluralism without, however, creating a
more or less stable authoritarian regime. There are, then, three possible
hypotheses behind a definition taking account of the context of origin,
which can be better explicated as follows: the regime arises out of one of the
different types of authoritarianism that have existed in recent decades, or even
earlier; the regime arises out of a traditional regime, a monarchy or sultanism;
or the regime arises out of the crisis of a previous democracy. To these must be
added a fourth, which is an important specification of the second: the regime
is the result of decolonialization that has never been followed by either
authoritarian or democratic stabilization.
If, to gain a closer empirical understanding of a hybrid regime, onerdevelops at least the first and second of these hypotheses a little further-though
the majority of cases in recent decades would seem to fall into the first
category-it can be seen that alongside the old actors of the previous
57
The Basics
56
they stable or unstable, for at least a decade, have been preceded by authoritarianism, a traditional regime (possibly with colonial characteristics), or even a
minimal democracy, and are characterized by the break-up of limited pluralism
As a way of stressing the differences with the existing literature (see above)
and of making sense of the definition above, .it is useful to emphasize the
reasons that justify it: to better understand what hybrid regimes are, it is
necessary to disentangle cases of transitional phases from hybrid regimes
stricto sensu, where the extent of achieved stabilization has to be taken into
account. At the same time it is important to grasp the ambiguities and the
fuzziness of regimes in which features ofboth democracy and authoritarianism coexist, and in this vein, to consider the institutional past that is so
important to them. But if this is the case, two key questions need to be
addressed: ( 1) Are there actually cases of hybrid regimes, or does reality just
throw up transitional cases, as might sound more reasonable? (2) If there
actually are hybrid regimes and transitional phases as well, what characterizes
one and the other? In other words, is it possible to elaborate a good typology
of hybrid regimes and single out the recurrent characteristics of transitional
phases? The remairring sections of the chapter will be devoted to answering
these questions.
A key element that runs against the effective existence of hybrid regime, that
is, institutional set-ups that are not democracy, nor authoritarianism, nor
traditionalism, is the expected low probability of duration. In fact, once some
degree of freedom and competition exists and is implemented in various
ways, it seems inevitable that the process will continue, even though the
direction it will actually take is unknown. lt might lead to the establishment
of a democracy, but it could also move backwards, with the restoration of the
previous authoritarian or other type of regime, or the establishment of a
different authoritarian or non-democratic regime. Is this constitutive short
duration or high instability confirmed by the Freedom House data?
If we assuroe that we are facing a transitional period when the 'partially
free' assessment is assigned for more than two years but less than a decade and
there is a regime in the pro~er sense when the same/similar political set-up
has lasted for ten years or more, and we consider all countries that were
'partially free' between 1989 and 2010, we can immediately detect and differentiate the forty-six cases of transition from the hybrid regimes. Among the
59
The Basics
58
Transition to
authoritarianism
Uncertainty in a
democratic context
Uncertainty in an
authoritarian context
Brazil (9)
Croatia (9)
Domin. Rep. (5)
EI Salvador (8)
Ghana (8)
Guyana (4 +1)
India (7)
Indonesia (7)
Romania (5)
South Africa (5)
Taiwan (7)
Afghanistan (3)
Algeria (3)
Azerbaijan (6)
Bahrain ( 4+ 7)
Belarus (5)
Bhutan (3 +2)
Egypt (4)
Eritrea (4)
Kazakhstan (3)
Swaziland (4)
Thailand (7+ 1)
Togo (3) +3
Tunisia (4) +1
Bolivia (7)
Ecuador (10)
East Tirnor (11)
Honduras (11)
Malawi (11)
Papua N. Guinea (5+7)
Philippines ( 6+5)
Solomon Islands (10)
Venezuela (4+11)
Burundi (7)
Congo (Brazzav.) (6+6)
Cte d'Ivoire (3+3)
Djibouti ( 11)
Gambia (9)
Haiti (6+4)
Kenya (8)
Kyrgyzstan (9+4)
Lebanon (4+5)
Liberia (5+6)
Mauritania (3)
Niger (5+11)
Yemen (4+6)
Note: The number in parenthesis refers to the years the country was assessed as partially free (PF). When
there is a plus (+), it means an interruption in the continuity of assessment.
Source: Freedom House, Freedom in the World. Country Ratings 1972-2007 (www.freedomhouse.org) .
transitional cases (see Table 3.1), we can find four different situations. There
are: ( 1) countries that after years of uncertainty became democracies, such as
Brazil, Croatia, and El Salvador, which had long transitional periods, and
countries with shorter transitions, such as Guyana, Romania, and South
Africa; (2) countries where transition led to authoritarian regimes, such as
Algeria, Belarus, and Thailand; (3) countries where there is great uncertainty
because a decade has not yet elapsed, but which have a democratic legacy,
such as Bolivia, Ecuador, and Venezuela; (4) countries where there was also
non-stabilization, but which have an authoritarian past.
Table 3.2 surprisingly confirms that not only can hybrid regimes stabilize as
such, but they are half (45) of the entire group (91). Here, if we distinguish
between more stabilized hybrid regimes, that is regimes that have been 'partially free' for fifteen years or more, we have twenty-six cases where at least
there has been a continual stand-offbetween veto players and democratic elites
resulting in a stalemate, where all the main actors, especially the elites, might
even find satisfactory solutions for their concerns, perhaps not ideal but
nonetheless viewed pragmatically as the best ones currently available; or
because a dominant power or even a coalition keeps the regime in an intermediate limbo; or, finally, due to the lack of any central, governing institution.
Hybrid regirne +
transition to
democracy
Hybrid regirne +
transition to
authoritarianism
Albania (19)
Bosnia-Herzegovina (14)
Pakistan (10)
Armenia (19)
Bangladesh ( 17)
Burkina Faso (18)
Cent. Afr. Rep (12+5)
Colombia (21 )
Cornores (20)
Ethiopia (2+ 15)
Fiji (10+10)
Gabon (19)
Georgia ( 18)
Guatemala (21)
Guinea-Bissau (19)
Jordan (20)
Kuwait (18)
Macedonia (18)
Madagascar (21)
Malaysia (21)
Meldova (19)
Morocco (21)
Mozambique (16)
Nepal (12+4)
Nicaragua (21)
Nigeria (4+ 12)
Paraguay (21)
Seychelles ( 18)
Singapore (21)
Sri Lanka (21)
Tonga (21)
Turkey (21)
Uganda (16)
Zambia (2+ 17)
Senegal (13+2)
Sierra Leone ( 12)
Tanzania (15)
Antigua &
Barbuda (13)
Lesotho (11)
Mexico (ll)
Peru (12)
Suriname (11)
Ukraine (14)
Russia (13)
Zirnbabwe (12)
,.t
60
The Basics
Let's consider the possibility, however, that our assurnptions (less than ten
years is to be considered a transitional phase; a decade or more to be regarded
as a hybrid regime, and more than fifteen years as a more stabilized hybrid), as
reasonable or practical as they rnight sound, are not accepted: is one year's
difference (from 9 to 10) enough to call the first one a case oftransition and the
18
second one a hybrid regime? But, even ifwe do not make those assumptions,
two basic findings have to be accepted and are worth singling out and
emphasizing strongly: ( 1) hybrid regimes do exist, as the first colurnn of
Table 3.2 shows beyond any doubt; (2) the cases ofhybrid regimes (according
to our definition) that turn into democracies or authoritarianisms are very few
vis-a-vis the other ones: ten out of forty-five. In other words, at least the
traditional, recurrent hypothesis that once some seed of competition is installed then it is difficult to stop it and some sort of democratic arrangement
will emerge is heavily underrnined. On the contrary, uncertainty and constraints on competition (and rights) can last for decades: in the first column of
Table 3.2 there are fourteen cases out of twenty-six where the hybrid regime has
been in place for almost two decades. At this point, to try to understand the
reasons that might explain all this, we need to look more closely at those
regimes. The first important step in this direction is to develop a classification
of the thirty-five regimes present in the first two columns of Table 3.2.
61
62
The Basics
63
21
traditionallpersonal regime
roilitary authoritarianisrn
protected dernocracy
civil-roilitary authoritarianisrn
limited dernocracy
likely that the survival of authoritarian veto players points towards a single
solution, that of protected democracies. In any case, the elaboration of this
classification or typology leads us to propose three possible classes: the
protected democracy and the limited democracy, which have already been
described above; and a third, logically necessary one in which it is hypothesized that there are no relevantlegacies or powerful veto players, nor are there
any forms of state suppression or non-guarantee of rights, but simply a
situation of widespread illegality in which the state is incapable of performing
properly due to poorly functioning institutions. This third dass can be
defined as 'democracy without law' or rather, democracy without state, as
the state can be conceived as a government based on the primacy of the law
and organized as a hierarchic system of roles defined by laws.
The second perspective, i.e. the focus on processes of change undergone by
a country and the consequences for the institutional set-up that emerges, may
be complementary to the first one. If the contex:t is one of regime change and
there is a hybrid resulting from a process of transition during which the
characteristics of the previous regime have disappeared ( as mentioned above ),
starting with the break-up oflimited pluralism (the hypothesis here is obviously that of a transition from authoritarianism), it is necessary to see what
process of change has started and how, in order to assess and predict its future
course. The advantage of this classificatory perspective is that, given that the
regimes in question are undergoing transformation, the direction and possible outcomes of such change can be seen more clearly. So, if there is liberalization, without or with little resort to violence, that is, a process of granting
more political and civil rights from above, never very extensive or complete,
The Basics
but so as to enable the controlled organization of society at both the elite and
mass level, what one has is an institutional hybrid that should permit an
'opening up' of the authoritarian regime, extending the social support base
and at the same time saving the goveming groups or leaders already in power.
The most probable result, then, would be a hybrid that can be defined as a
protected democracy, capable, moreover, of lasting for a considerable or very
long time. In order to have some probability of persistence, such a political
hybrid must be able to rely not only on the support of the institutional elites,
both political and social, but also on the maintenance of a limited mass
parti~ipation (in other words, on the goveming elite's capacity to repress or
dissuade participation) and on the limited attraction of the democratic model
in the political culture of the country, especially amongst the elites. Another
possibility is the occurrence of a rupture of authoritarianism as a result of a
grassroots mobilization of groups in society or of the armed forces, or due to
foreign intervention. If it proves impossible to move towards a democratic
situation, even if only slowly, due to the presence of anti-democratic veto
players, then a more or less enduring situation characterized by a lack of the
guarantees regarding order and basic rights to be found in a limited democracy becomes a concrete outcome. In this dynamic perspective, the third
solution, that of the democracy without state, does not even entail liberalization or the break-up of limited pluralism as such, in that there is no
previously existing stable regime or working state institutions.
However, these modes of classification are a priori and do not include an
empirical survey of the countries defined as hybrid regimes. In this respect
they still fail to carry out one of the principal tasks of any good classification,
which is to examine all the empirical phenomena associated with hybrid
regimes and then arrive at some form of simplification that makes it possible
to grasp the phenomenon as a whole as weil as its intemal differences. So do
the three classificatory types outlined above hold up when empirical cases are
considered? Looking at the Freedom House data and assuming that the
regimes regarded as 'partially free' coincide with the notion of the hybrid
regime developed here, there seems to be a need for some revision and
integration. Above all, it is worth -examining in greater detail the ratings of
the countries belonging to the category of hybrid regimes (Table 3.2) on the
set of indicators relating to seven ambits that are important when analysing
any political regime, democratic or otherwise: rule of law, electoral process,
functioning of government, political pluralism and participation, freedom of
expression and beliefs, freedom of association and organization, personal
autonomy and individual freedom. 23 If those ambits are matched with the
elements that appear in the definition of the hybrid regime suggested above,
the connections can be immediately grasped: all aspects of minimal
64
65
Cornponents
2
Rule oflaw
Electoral process
GovernrnentfUnctioning
Pluralisrn and participation
Freedorn of expression & beliefs
Freedorn of association & organization
Personal autonorny and individual freedorn
Notes: Percentage of explained variance: 80.623%.
Method of extraction: Principal Component Analysis.
Method of rotation: Varirnax with Kaiser normalization.
Only factor loadings > 0.5 represented.
0.856
0.835
0.818
0.862
0.713
0.865
0.858
66
The Basics
67
Countries
Quasi-democracies
Albania
Bosnia-Herzegovilla
Colombia
Georgia
Macedonia
Madagascar
Malaysia
Moldova
Mozambique
Seychelles
Singapore
Sri Lanka
Tanzania
Turkey
Zambia
Limited democracies
Banglac:1esh
Cornoras
Fiji
Guatemala
Guinea-Bissau
Jordan
Nicaragua
Paraguay
Sierra Leone
Tonga
Armenia
Burkina Faso
Central African Rep.
Ethiopia
Gabon
Kuwait
Morocco
Nepal
Nigeria
Uganda
The Basics
the semantic field, that is, the notions of democracy, authoritarianism, and the
traditional regime. Third, once we went from an a priori classification, grounded
on the criterion of the constraints to an effective minimal democracy, to an
empirical one, a few changes and adaptations proved necessary.
On the whole, even if different trajectories are theoretically relevant (see
also Figure 3.1), most ofthe cases that might be considered as hybrid regimes
have either an authoritarian past or a traditional one. Moreover, because of
the data we had, the dass of protected democracies did not emerge. But, for
example, in cases such as Jordan and Morocco, which we would have unquestio~ably Iahelied as such because of the strong political role of a traditional
power like that of the monarchy, our empirical analysis suggests that the
inefficiency of the state (Morocco) or the limited guarantee of rights (Jordan)
is more relevant. As regards this, the data and a different theoretical emphasis
account for an unexpected result. In fact, the data clearly suggest that what
counts in hybrid regimes is not so much the existence of a legacy or of veto
players, but a lack of state or the effective guarantee of basic rights.
The quasi-democracies are a more relevant dass than expected, with fifteen
cases, and this is perhaps grounds for optimism in terms of the future
possibility of change towards democracy. In particular, in an international
context characterized by cooperation rather than high conflict, a stronger role
of international actors or of democratic governments is possible and seems to
have some chance of success in a democratic perspective. 30
Limited democracies and democracies without state are confirmed as empirically relevant classes for which different, contrasting elements need to be
stressed: on the one hand, the lack of an effective guarantee of rights despite
the presence of state institutions and, on the other hand, the lack of the rule of
law and of a functioning state, with laws that are not applied because the
judiciary has no effective independence, there is widespread corruption, and
the bureaucracy is tlawed and inefficient. In these cases too, governments and
international organizations have a potentially strong role to play in helping to
strengthen the respect of rights and building state institutions, prior even to
establishing democracies, in countriesthat have manifested a strong incapacity in this respect over the years. 31
There is a final important question that is still unanswered: What is the last
step? More explicitly, even if we accept that the threshold from hybrid regime
and minimalist democracy can be crossed in both directions, what should we
consider the additional change, i.e. the 'last step', for a regime or also for
transitional context to be considered a minimalist democracy? The minimalist definition of democracy, recalled at the beginning of this chapter-and
discussed in the previous one-complemented by the result of the empirical
dassification show the existence of three patterns with regards to the last step,
and a11 of them when not immediately taken, difficult to emerge in a situation
of stalemate. The first pattern, development of civil and political rights, is the
one requested in case oflimited democracy; the second one, building of rule of
law, is what the transition to a minimal democracy would entail in a democracy without state; the third mixed pattern, growth of right and rule of law,
clearly concerns the possibility of transitions for the quasi-democracies.
However, every pattern can be additionally and more ptecisely developed
only by following the specific directions the empirical analysis of each case
points to. Thus there is no possibility of finding a 'recipe' to be applied in
different cases: there is no set of recommendations that 'fits all'. Two final
considerations on this may be helpful. Although we are not going to develop
the point, it seems evident that this analysis can have policy recommendations. That is, at least in an abstract way the classification of each country in
the tables above, the consequent pattern, and the additional specific recommendations can show specific strategies of transitions. Second, although this
analysis is limited to hybrid regimes, patterns of such a kind can be traced in
the transitional processes, that is, also in the cases where a country is not stuck
into some dass of hybridism, but follows a path of change.
This last consideration brings us to the second and third parts of the book,
where the broader processes of change, i.e. transition and installation of
democracy, consolidation and crisis, qualities deepening or worsening, and
the related sub-processes willbe considered in terms of actor/structure led
change, multi-levellegitimation, domestic and external anchorings, and procedures, contents, and results in democratic deepening.
68
69
Notes
1. See also Collier and Levitsky (1997: 440), where there is an exhaustive Iist of
'dirninished sub-types', i.e. hybrid regirnes, that are present in the Iiterature on
the topic. I'm not taking into consideration here the issue of fragile and failed
states. This would open a largely different set of problems that are not the object of
the analysis here.
2. See below for more on the meaning of the terms as used here.
3. The main criticisms concerned the right-wing liberal bias of the Institute itself and
consequently the compiling of unfair ratings for the assessed countries. Although
basically appropriate at the beginning, these criticisms have been overcome and
neutralized by subsequent de},'elopments characterized by more reliable emJ?irical
results. This relevant conclusion is strongly supported by the high number of
quotations and attention that Freedom House data have received in recent years.
Even a quick look on the Internet will vividly show all this.
The Basics
4. Even some of the most interesting data, such as those of the Index ofFailed States
(see www.foreignpolicy.com/story/cms.php?story_id=3865) or of Polity N (see
www.systemicpeace.org/polity/polity4.htm) were discarded in favour ofFreedom
House data for the reasons mentioned in the text.
5. See below for a necessary and more specific definition.
6. It should be remernbered that Freedom House adopts a reverse points system: a
score of 1 corresponds to the greatest degree of democracy in terms of political
rights and civil liberties, while 7 corresponds to the most repressive forms of
authoritarianism as regards rights and freedom. The electoral democracies need
. to be distinguished amongst these. For more on the definition thereof, see below.
7. As is known, about half of this population lives in a single nation, China.
8. The inclusion ofTurkey in this group of countries has already prompted debate,
and other analysts, especially Turkish scholars, place it amongst the minimal
democracies, stressing the great and now long-standing fairness of the electoral
procedure, for which Freedom House does not award the maximum rating.
9. Other terms, such as 'mixed regimes', arealso very broad, but ultimately 'hybrid'
was preferred to 'mixed' as the former term gives more precisely the gist of the
phenomenon where new and old aspects change each other, that is, it is not just a
problern of'mixing' (see below for the definition and classification).
10. A more complex definition is offered by O'Donnell (2004: 15), who suggests
considering the patterns, explicit or otherwise, that determine the channels of
access to the main government positions, the characteristics of the actors who are
admitted or excluded from such access, and the resources or strategies that they
can use to gain access. An empirically simpler line is adopted here, which is based
on the old definition by Easton (1965). But seealso Fishman (1990).
11. He also adds that a state is a 'more permanent structure of domination and
coordination' than a regime (Fishman 1990: 428). But on this see below.
12. The other specific components ofliberal democracies are delineated by Diamond
(1999: 11-12).
13. The three indicators pertaining to the electoral process are: (1) head of government and principal posts elected with free and fair elections; (2) parliaments
elected with free and fair elections; (3) electorallaws and other significant norms,
applied correctly (see the website ofFreedom House).
14. These values include notions like homeland, nation, order, hierarchy, authority,
and such like, where both traditional and modernizing positions can, and sometimes have, found common ground. In any case, the regime is not supported by
any complex, articulated ideological elaboration. In other regimes, like the traditional ones, the only effective justification of the regime is personal in nature, that
is, to serve a certain leader, who may, in the case of a monarch who has acceded to
power on a hereditary basis, be backed by tradition.
15. For another analysis of non-democratic regimes, especially authoritarian ones,
see Brooker (2000).
16. Despite her empirical focus on Centtal America the analysis by Karl (1995) is also
useful to betterunderstand the conditions and perspectives ofhybrid regimes in
other ateas.
17. As mentioned above, many years ago Finer (1970: 441-531) seemed to have
detected the existence of hybrid regimes when he analysed 'fac;:ade democracies'
and 'quasi-democracies'. Looking more closely at these two models, however, it is
clear that the former can be tied in with the category of traditional regimes, while
the latter falls within the broader authoritarian genus. In fact; typical examples of
'quasi-democracies' are considered to be Mexico, obviously prior to 1976, and
certain African nations with a one-party system. A third notion, that of the
'pseudo-democracy', refers not to a hybrid regime, but to instances of authoritarian regimes with certain exterior forms of the democratic regime, such as constitutions claiming to guarantee rights and free elections but which do not reflect
an even partially democratic state of affairs. There is, then, no genuine respect for
civil and political rights, and consequently no form of political competition either.
18. On the whole, we stillthinkthat those hypotheses are reasonable and practical for
our purposes. Above all, they are obligatory when trying to make better sense of
the set of cases with which we are dealing.
19. For a more in-depth discussion of veto players in hybrid regimes and in the
democratization process, see Morlino and Magen (2008a, 2008b).
20. Such a regime combines 'democratic and authoritarian elements' (Diamond
2002: 23).
21. It should be noted that Diamond uses the term 'electoral democracy' with a
different meaning to that of ~{eedom House, as has already been clarified above.
For Diamond 'electoral democracy' and 'liberal democracy' are two different
categories, while for Freedom House all liberal democracies are also electoral,
but not vice versa. So, for example, according to Freedom House a nation like the
United Kingdom is a liberal democracy, but is also electoral, while for Diamond it
is not.
22. To build his typology Wigell (2008) develops the two well-known Dahlian criteria
(participation and competition/opposition) (Dahl 1971) into the notions of
'electoralism' and 'constitutionalism' seen at limited and at effective stages. If
there is a limited development then there are minimal electoral conditions, such
as free, fair, competitive, inclusive elections; and/or minimal constitutional conditions, such as freedom of organization, expression, alternative information, and
freedom from discrimination. If there is more effective development, then there
are additional electoral conditions, such as electoral empowerment, electoral
integrity, electoral sovereignty, electoral irreversibility, and/or additional constitutional conditions, such as executive accountability, legal accountability, bureaucratic integrity, local government accountability. The result is a four-cell matrix
with liberal democracy in the case of effective electoralism and effective copstitutionalism; limited democracy in the case of limited, minimal electoralism and
limited, minimal constitutionalism; electoral democracy in the case of minimal
70
71
72
The Basics
s.
73
For example, as regards the salience of the judicial arena, see Morlino and Magen
(2008b).
6. Piease riote that while the main purpose of such a method is explanatory, it is used
2
here to provide a better classification of my cases. It is worth recalling, however,
the existing connection between explanation and a good classification, although
this cannot be discussed here.
27. For details of the indicators see above, n. 23.
28. All the truth tables are available upon request.
29. Please note that because of the strong correlations the specific group of variables
(i.e. (1) electoral process + political pluralism and participation + freedom of
expression and beliefs + freedom of association and organization; (2) rule oflaw
and personal autonomy and individual freedoms) have been combined on the
basis of the rule of sufficiency, that is, the existence of one component is sufficient
to represent the dass.
30. On this point, see also the results of research on the role of international actors
in the democratization processes in a number of Europe's neighbour nations
(Morlino and Magen 2008b).
31. Please note that the role of the lOs can be double-faced, and sometimes vague and
slippery. They can recognize as interlocutor a corrupted government.
;'~
"