Tax Expenditure Management A Critical Assessment

TA X E X PE N DI T U R E M A NAGE M E N T
A tax expenditure is a ‘tax break’ allowed to a taxpayer or group of

taxpayers, for example, by way of concession, deduction, deferral or
exemption. he tax expenditure concept, as it was i rst identiied, was
designed to demonstrate the similarity between direct government
spending on the one hand and spending through the tax system on the
other. he identiication of beneits provided through the tax system as
tax expenditures allows analysts to consider the iscal signiicance of
those parts of the tax system which do not contribute to the primary pur-
pose of raising revenue. Although a seemingly simple concept, it has gen-
erated a range of complex deinitional and practical issues, and this book
identiies and critically assesses the controversial aspects of tax expend-
iture and tax expenditure management.
m a r k b u r t o n is an Associate Professor in Law at the University

of Melbourne and a Fellow of the Taxation Law and Policy Research
Institute, Monash University.
k e r r i e s a d i q is a Professor at the Queensland University of

Technology and a Fellow of the Taxation Law and Policy Research
Institute, Monash University. She is also a Barrister, Supreme Court of
Queensland, and a Chartered Tax Adviser.
Cambridge Tax Law Series
Tax law is a growing area of interest, as it is included as a subdivision in many

areas of study and is a key consideration in business needs throughout the world.
Books in this series will expose the theoretical underpinning behind the law
to shed light on the taxation systems, so that the questions to be asked when
addressing an issue become clear. hese academic books, written by leading
scholars, will be a central port of call for information on tax law. he content will
be illustrated by case law and legislation.
he books will be of interest for those studying law, business, economics,
accounting and inance courses.
Series Editor
Dr Peter Harris, Law Faculty, University of Cambridge,
Director of the Centre for Tax Law.
Dr Harris brings a wealth of experience to the Series. He has taught and pre-
sented tax courses at a dozen diferent universities in nearly as many countries
and has acted as an external tax consultant for the International Monetary Fund
for more than a decade.
TA X E X PEN DI T U R E
M A NAGE M E N T
A Critical Assessment
M A R K BU RTON
and
K E R R I E SA DIQ
C A M BR I D G E U N I V E R S I T Y P R E S S
Cambridge, New York, Melbourne, Madrid, Cape Town,
Singapore, São Paulo, Delhi, Mexico City
Cambridge University Press
he Edinburgh Building, Cambridge CB2 8RU, UK
Published in the United States of America by Cambridge University Press, New York
www.cambridge.org
Information on this title: www.cambridge.org/9781107007369
© Mark Burton and Kerrie Sadiq 2013
h is publication is in copyright. Subject to statutory exception

and to the provisions of relevant collective licensing agreements,
no reproduction of any part may take place without the written
permission of Cambridge University Press.
First published 2013
Printed and bound in the United Kingdom by the MPG Books Group
A catalogue record for this publication is available from the British Library
Library of Congress Cataloguing in Publication data

Burton, Mark.
Tax expenditure management : a critical assessment / Mark Burton and Kerrie Sadiq.
p. cm. – (Cambridge tax law series)
Includes bibliographical references and index.
ISBN 978-1-107-00736-9 (hardback)
1. Tax expenditures. 2. Tax expenditures – Law
and legislation. I. Sadiq, Kerrie. II. Title.
HJ4642.B87 2013
352.4′4–dc23
2012029786
ISBN 978-1-107-00736-9 Hardback
Cambridge University Press has no responsibility for the persistence or

accuracy of URLs for external or third-party internet websites referred to in
this publication, and does not guarantee that any content on such websites is,
or will remain, accurate or appropriate.
C ON T E N T S
Acknowledgements page viii
1 Introduction 1
1.1 he subject of this book 1
1.2 What are tax expenditures? 1
1.3 Why single out tax expenditures? 3
1.4 Explaining the focus upon tax expenditures – the
signiicance of tax expenditures 4
1.5 What else is controversial about tax
expenditures? 8
1.6 he irresolvable nature of these tax expenditure
controversies 10
1.7 Tax expenditure controversies and liberal-democratic
political theory 11
1.8 he thesis of the book 13
1.9 Outline of the book 14
2 he tax expenditures concept 15
2.1 Introduction 15
2.2 he concept of tax expenditures 17
2.3 he purpose of categorisation 23
2.4 he function of identifying tax expenditures 28
2.5 he identiication of tax expenditures 35
2.6 he measurement of tax expenditures 55
2.7 Contemporary expansions of the traditional
concept of tax expenditures 59
2.8 Conclusion 63
v
vi Contents
3 Reporting on tax expenditures 65

3.1 Introduction 65
3.2 International aspects of tax expenditures 68
3.3 he impact on reporting of the purpose and function of
recognising tax expenditures 76
3.4 Tax expenditures so deined 81
3.5 he benchmark tax law 83
3.6 Identiication of all types of tax expenditures 91
3.7 Measurement of tax expenditures 93
3.8 Structure and information contained in tax expenditures
reports 97
3.9 Conclusion 101
4 he practical signiicance of tax expenditures 103
4.1 Introduction 103
4.2 Where to start? Holism, atomism and tax expenditure
management 104
4.3 he economic signiicance of tax expenditures 107
4.4 Public administration 128
4.5 Legal signiicance 130
4.6 Administration of taxation law 134
4.7 he legitimacy of government 144
4.8 Tax expenditures – the weak link in the iscal
chain? 145
5 he politics of tax expenditure management 147
5.1 Introduction 147
5.2 he connection between moral objectivity and the politics
of tax expenditure management 149
5.3 Deontological norms and their relevance to tax
expenditure management 154
5.4 Democracy as political procedure – accountability and
transparency 169
5.5 Ontological accounts of tax expenditure
management 182
5.6 Deliberative concepts of accountability 191
Contents vii
6 Managing tax expenditure controversies 203

6.1 Introduction – epistemic and political tensions 203
6.2 What is to be done? 208
6.3 Towards a democratic tax expenditure management
framework 217
6.4 Into the political future 230
Bibliography 232
Index 249
ACK NOW L E D GE M E N T S
For Marie-Louise and Will – tax might not be the centre of the moral uni-
verse, but you are both the centre of mine: Mark.
For Mohsan, Lily and Erin – you are my constant rays of sunshine:
Kerrie.
From both of us, we would like to express thanks to all of those who
have advised, encouraged and helped us over this journey. Any fault of the
text is, however, down to us.
viii
1
Introduction
1.1 he subject of this book

his book is about how tax expenditures might be dei ned and managed
in a modern democracy. Others before us have ventured onto this ter-
rain with a view to identifying an ideal approach to these matters. Unlike
those commentators, we are not conident that there is an uncontroversial
tax expenditure management framework that can be prescribed for any
community at any time. he question that the book deals with is how
this absence of an ideal approach should be dealt with in a community
with ‘democratic’ aspirations. Some argue that such a controversial sub-
ject should be ignored by abandoning it altogether. However, we maintain
that both tax expenditures and the process for dealing with the contro-
versial aspects of the subject are too signiicant to ignore in a contempor-
ary democratic setting.
1.2 What are tax expenditures?

At irst blush, the concept of a tax expenditure is beguilingly simple. For
present purposes1 it is suicient to note that a tax expenditure is generally
understood to be a ‘tax break’ allowed to a taxpayer or a limited group
of taxpayers, for example, by way of concession, deduction, deferral or
exemption. Alternative phrases are also used to describe the same concept.
In Europe, these ‘tax breaks’ are oten referred to as ‘tax subsidies’,2 whilst
many OECD countries refer to tax expenditures as ‘tax reliefs’ or ‘tax aids’.3
hey have previously been known in the United States as ‘tax preferences’.4
A layman may refer to them as ‘loopholes’ or ‘escapes from tax’.5
1
In Chapter 2 we consider the tax expenditure concept in more detail.
2
See EuroSAI, Report on the Coordinated Audit of Tax Subsidies (2008).
3
OECD, Tax Expenditures in OECD Countries (2010), 12.
4
Surrey and McDaniel, Tax Expenditures (1985), 25.
5
Surrey and McDaniel, Tax Expenditures (1985), 1; for discussion of the political colour of
such terms see Bittker, ‘Income Tax “Loopholes” and Political Rhetoric’ (1973).
1
2 Introduction
Where a taxpayer is entitled to a tax expenditure, the outcome is com-

parable to that taxpayer having paid the tax due under the normal tax
structure and then having received a direct cash grant from the govern-
ment that is equal to the amount of the tax expenditure. In other words,
a tax expenditure is essentially a shortcut to a direct payment by a gov-
ernment as it saves the government from collecting tax and then paying
money back to the same person.6
In Chapter 2 we note that the nature of the ‘tax expenditure’ concept
will vary depending upon the purpose for which tax expenditures are
being identiied. First coined by Stanley Surrey in the late 1960s,7 the ‘tax
expenditure’ oxymoron was intended to highlight the incorporation of
‘spending’ rules within the ‘taxation’ law. Surrey hoped to shock legislators
into recognising that tax spending rules should be subjected to the same
merits review as direct spending measures. Some suggest that Surrey’s
motive was to abolish all tax expenditures, although this appears to reach
beyond Surrey’s overt ambitions for tax expenditure ‘management’. More
likely, Surrey hoped that budgetary committees would cast a cold eye
upon all government spending with a view to weeding out unmeritori-
ous spending no matter what form it took.8 To this end, Surrey adopted
a limited tax expenditure concept that focused upon expressly legislated
income tax rules that were substitutable for direct spending measures and
that departed from what Surrey identiied as a ‘normal’ income tax.9
One consequence of the limited deinition of the tax expenditure con-
cept adopted in the United States is that substantial tax breaks, more
broadly understood, are ignored. For example, excise tax breaks are
ignored because they do not arise under the income tax.10 Even if Surrey’s
concentration upon ‘spending reform’ were maintained,11 there is a good
case for relaxing the concentration upon tax expenditures within the
income tax alone. In Chapter 2 we note that a broader approach to iden-
tifying tax expenditure reaches beyond the narrowly targeted ‘spending
reform’ envisaged by Surrey and that such broader approaches entail
6
Surrey, Pathways to Tax Reform: he Concept of Tax Expenditures (1973), 6–7.
7
Although note the earlier German recognition of the ‘tax spending’ phenom-
enon: Shannon, ‘he Tax Expenditure Concept in the United States and Germany: A
Comparison’ (1986), 203.
8
9
Surrey, Pathways to Tax Reform: he Concept of Tax Expenditures (1973), 6f.
10
OECD, Inventory of Estimated Budgetary Support and Tax Expenditures for Fossil Fuels
(2011), 30.
11
By contrast to reform of the legislative rules underpinned by tax policy principles: Surrey,
Pathways to Tax Reform: he Concept of Tax Expenditures (1973), 30–1.
Why single out tax expenditures? 3
relaxation of some of Surrey’s tax expenditure criteria.12 Reaching beyond

identifying tax expenditures as a vehicle for spending reform, we argue
that identifying tax expenditures could be directed towards broader pub-
lic policy objectives such as assessing the performance of a taxation system
overall against a deined ideal tax system. We also consider whether non-
legislative departures from a speciied ‘normal’ tax should be recognised
as tax expenditures. If the broader objective of reporting upon the per-
formance of a tax system (or of a tax within a tax system) is adopted, there
is no reason to restrict the tax expenditure concept to expressly legislated,
income tax rules that are substitutable for direct spending measures.
1.3 Why single out tax expenditures?

In the preceding section we noted that the purpose for identifying tax
expenditures will shape the tax expenditure concept. Specifying the pur-
pose of identifying tax expenditures is therefore one controversial aspect
of the tax expenditure concept. Indeed, some take this purposive inquiry
to the point of questioning whether a community should focus upon tax
expenditures as a separate subject at all. hese commentators agree that
critical appraisal of government policy needs to be enhanced, but argue
that focusing upon ‘tax expenditures’ is at best a distraction from more fun-
damental matters of public policy. If a purpose of enhancing public admin-
istration is adopted, there is a view that a ‘whole of government’ approach
ought to be adopted when assessing government action. From this broader
standpoint, some commentators argue that there is no reason to segregate
‘special’ tax treatment in the form of tax expenditures from other forms of
preferential treatment. hese commentators observe that special govern-
ment treatment can be found in other government mechanisms as well –
such as the payment of a government subsidy, the disposal of public assets
for less than their ‘real’ value, diferential recognition of private property
rights and/or diferential regulatory treatment. From ‘whole of govern-
ment’ viewpoints, tax expenditure management should not be undertaken
in isolation because all government actions should be subjected to critical
assessment against the same organising principle.13 Weisbach and Nussim,
for example, argue that all government interventions should be assessed
against an organisational standard that requires identiication of measures
12
Surrey only briely considered the application of the tax expenditure concept to other
taxes: Surrey, Pathways to Tax Reform: he Concept of Tax Expenditures (1973), 26–9.
13
Roin, ‘Truth in Government: Beyond the Tax Expenditure Budget’ (2003).
4 Introduction
which are most efective in achieving a desired outcome.14 Similarly, a

broader approach suggests that critical assessment of the ‘tax/transfer’ sys-
tem should displace the appraisal of discrete ‘tax’ or ‘spending’ rules.15
Implementing such broad analyses of public policy would require con-
siderable resources to gather and interpret all relevant information. If
gathering, interpreting and analysing information were costless, it might
readily be accepted that calling for more information is banal.16 However,
as Downs noted,17 one of the key challenges confronting democratic gov-
ernment is the management of imperfect information in an environment
where information is not costless. ‘Whole of government’ perspectives
may not incorporate adequate consideration of the considerable resources
necessary for such comprehensive assessments of government action.
Further, the opportunity cost associated with devoting public resources
to such inquiries should be considered. Finally, the pragmatic aspect of
‘whole of government’ analysis must be addressed – how could a govern-
ment gather the requisite information in the face of opposition to what
can be portrayed as intrusive government imposition of tax compliance
costs upon the entrepreneurs in a market economy?
Every community, then, must make diicult choices about the proce-
dures for making substantive policy choices with imperfect information
and imperfect means of analysing and interpreting that information.18 he
signiicance of this issue is perhaps most stark for a developing economy,
where the beneit of comprehensive assessment of all government action
may be outweighed by beneits derived from targeting limited resources
upon speciic aspects of tax administration. In such contexts, it is possible
that a limited form of tax expenditure management delivers the greatest
public beneit by reining the taxation system which underpins meritori-
ous direct spending programmes.
1.4 Explaining the focus upon tax expenditures – the

signiicance of tax expenditures
Recognising these costs of appraising public policy complicates the con-
sideration of whether or not tax expenditures should be examined as a
14
Weisbach and Nussim, ‘he Integration of Tax and Spending Programs’ (2004).
15
Kaplow, ‘Taxation and Redistribution: Some Clariications’ (2007). Hence greater policy
emphasis upon the tax/transfer system. See, for example, Australia, Australia’s Future
Tax System (2009), 15.
16
Shaviro, ‘Rethinking Tax Expenditures and Fiscal Language’ (2004), 219.
17
Downs, An Economic heory of Democracy (1957), 215.
18
McGarity, Reinventing Rationality (1991).
Explaining the focus upon tax expenditures 5
discrete subject. If there are insuicient resources to critically assess all

government action, what should be done? We argue that tax expendi-
tures require separate consideration because they are a special case of
government action and also because they are so signiicant from several
standpoints.
Tax expenditures are signiicant by virtue of their number, distribu-
tion,19 impact upon the iscal position of states, constitutional signiicance,
impact upon public administration in general and tax administration in
particular and also because of their relevance to the legitimacy of demo-
cratic governments. hese and other impacts are considered in Chapter 4.
However, at this point it is appropriate to outline some of the impacts of
tax expenditures for two purposes. he irst purpose is to reject the ‘whole
of government’ proposition that tax expenditures should not be exam-
ined in isolation. he second purpose is to justify the book upon the basis
that tax expenditures warrant further consideration to which, we hope,
this book makes a useful contribution.
1.4.1 he pragmatic politics of tax expenditures in

contemporary democracies
Tax expenditures are a special case of government action because of the
unique and politically powerful combination of ‘tax’ and ‘spending’. As
many commentators have noted, tax expenditures are politically signii-
cant in contemporary democracies because they aford politicians the
opportunity to combine the politically powerful message of providing
a ‘tax cut’ while simultaneously delivering government ‘spending’ upon
politically signiicant issues such as welfare, small business, the environ-
ment and other substantive policy areas.20
1.4.2 Economic signiicance

From a iscal perspective, the most signiicant impact of a government’s
tax expenditures programme is the scale of the programme both in terms
of gross revenue foregone21 and also relative to the scale of direct spend-
ing. By virtue of their scale, tax expenditures are a major feature of gov-
ernment policy in many countries.22
19
20
Howard, he Hidden Welfare State (1997).
21
he diiculties of measuring the cost of tax expenditures are discussed in section 2.6.
22
he signiicance of tax expenditures is relected in the work that the OECD has under-
taken in this area. In 1984 and 1986 the OECD published studies speciically dealing
6 Introduction
Assuming that the iscal needs of a government remain constant, the

efect of a tax expenditures programme is to shit part of the tax con-
tribution from the beneiciaries of that programme to the broader tax
community. By relieving some taxpayers from paying at least some tax
that they would otherwise have paid, governments can use tax expendi-
tures to favour particular activities over others, and hence induce pri-
vate investors to invest in the favoured sector. hus, tax expenditures
are a way that governments afect the allocation of a community’s lim-
ited resources to particular activities. ‘Neoclassical’ economic theory
holds that the ‘invisible hand’23 of self-interest ensures that perfectly
functioning free markets are the best means for a society to extract the
greatest beneit from its limited resources. Some tax expenditures can
be defended upon the basis that they minimise the social cost of imper-
fect markets. However, there are empirical diiculties in demonstrat-
ing that the tax expenditure maximises social welfare. h is is because
of the diiculty of establishing that the social cost of a tax expend-
iture is less than the social beneit generated by introducing the tax
expenditure (i.e. the reduction of the social cost caused by the market
imperfection).24
By favouring some taxpayers over others, tax expenditures also afect
the distribution of income and wealth within a community. While
many tax expenditures are justiied upon the basis that they redistrib-
ute resources to those with lower incomes, many other tax expenditures
favour those with higher incomes. For example, the rationale of taxing
income rather than other tax bases is that progressive taxation of income
constitutes a fairer means of distributing a community’s tax burden
than other tax bases would allow25 and are also one means of pursuing
with tax expenditures. In 2003 it issued a special feature about tax expenditures in
Revenue Statistics (Special Feature for the 2003 Edition of Revenue Statistics: Note by the
Secretariat). he OECD document Best Practices for Budget Transparency (2002) con-
tains basic guidelines for the treatment of tax expenditures. In 2004, it elaborated on
these guidelines in Best Practice Guidelines – Of Budget and Tax Expenditures. Canada,
the Czech Republic, Germany, France, Greece, Hungary, Italy, Mexico, the Netherlands,
New Zealand, Norway, Poland, Spain, Sweden, the United Kingdom and the United
States were represented at the expert meeting to discuss the 2004 document. Australia
and Japan provided written information.
23
Smith, he Wealth of Nations (1993), para IV.2.
24
Burton, ‘he Australian Small Business Tax Concessions – Public Choice, Public Interest
or Public Folly?’ (2006).
25
Stanley, Dimensions of Law in the Service of Order (1993); Fleming and Peroni,
‘Reinvigorating Tax Expenditure Analysis and its International Dimension’ (2008).
Explaining the focus upon tax expenditures 7
a community’s aspirations to economic democracy.26 To the extent that

income tax expenditures redistribute a community’s resources to those
with higher incomes/wealth, tax expenditures undercut the legitimacy of
the income tax.27
1.4.3 Public administration

From a public administration perspective, tax expenditures can reach
into any domain across the policy spectrum – from social welfare to
industry assistance to national security. Typically tax expenditures are
administered by the tax authority because of the formal categorisation of
these ‘spending’ rules to the tax law. his is the case notwithstanding the
fact that any particular tax expenditure might be managed more efect-
ively within the relevant substantive government portfolio where there is
greater policy expertise and understanding of the interaction of the tax
spending with other government programmes.
Further, the integration of a plethora of public policy measures within
the tax framework engenders challenges surrounding the administration
of tax expenditures,28 particularly in the context of a system where tax-
payers self-assess their taxation liability. If the tax authority concentrates
upon revenue collection and is perceived not to have devoted suicient
resources to the spending aspects of the tax law, the legitimacy of the tax-
ation system can be threatened with consequent negative impacts upon
taxpayer voluntary compliance.29
1.4.4 he uneasy case for examining tax expenditures in isolation

In a perfect world communities would be able to subsume critical
appraisal of tax expenditures into a ‘whole of government’ selection of the
best means of achieving a desired public policy. However, the direct costs
and political challenges associated with this course of action mean that
a comprehensive whole of government analysis is not feasible. Tradeofs
must be made between competing objectives and this means that a com-
prehensive, whole of government approach must be compromised. At
26
For an illustration of this theme in the popular media see: Stiglitz, ‘Of the 1%, for the 1%,
by the 1%’ (2011).
27
Fox, If Americans Really Understood the Income Tax (2001), ch. 5.
28
United States of America, A Reconsideration of Tax Expenditure Analysis (2008), 41;
Brooks, ‘Comment’ (1988).
29
See section 4.6.
8 Introduction
the other end of the ‘information requirement continuum’ are those who
suggest that information diiculties mean that public government should
be kept to a minimum.30 If government policy cannot rest upon a irm
empirical basis, they say, government should not interfere. We disagree
with this libertarian view. Leaving a justiication for an interventionist
state for another day, suice to say that contemporary democratic govern-
ments are some considerable distance from libertarian minimalism.31 In
the present era it is diicult to imagine a truly non-interventionist state
where ‘whole of government’ analysis could be achieved at little social
cost because the scope of government was so small.
Once both comprehensive, whole of government analysis and libertar-
ian minimalism have been excluded upon pragmatic grounds, a point on
the ‘uneasy’ middle ground between these two extremes must be adopted.
A more selective approach must be adopted, notwithstanding the limita-
tions of such selectivity. We suggest that the practical signiicance of tax
expenditures means that they should be selected for critical scrutiny. In
itself, this selection of tax expenditures is controversial.
1.5 What else is controversial about tax expenditures?
1.5.1 he tax expenditure benchmark

Even if it is accepted that a community should focus its limited resources
upon examining the ‘proper’ functioning of its taxation system, it could
be argued that the community should critically assess the operation of its
entire taxation system as a routine part of its budgetary process. However,
for the same reasons as were noted in section 1.4, in most if not all com-
munities limited public resources are insuicient to allow such a compre-
hensive analysis of an entire taxation system on a routine basis. Further, a
judgement must be made regarding the likely beneits relative to the costs
of such ongoing systemic review. In this context, focusing a community’s
attention upon particular aspects of its tax framework is a pragmatic
response.
30
See the discussion of libertarian political theory in section 5.4.
31
And perhaps even if government is restricted to libertarian minimalism, as comprehen-
sive, whole of government analyses of alternate policy options with respect to the minim-
alist domains of property and national defence could, in themselves, absorb considerable
resources. In this regard, note Epstein’s more nuanced appreciation of the limitations of
simplistic libertarianism: Epstein, ‘Taxation with Representation: Or, he Libertarian’s
Dilemma (2005).
What else is controversial about tax expenditures? 9
One means by which eforts might be focused is by separating less

contentious tax policy decisions from those which a community consid-
ers to be more contentious. Identifying tax expenditures by separating
them from ‘normal’ taxation rules fuli ls this function. In Chapter 2 we
review the literature regarding various ways that this crucial aspect of tax
expenditure management has been undertaken. Suice to say that iden-
tiication of the ‘normal’ rules is undertaken by identifying an ideal tax-
ation system against which the present taxation rules can be compared.
Surrey’s ideal tax system was framed upon the basis of what he proclaimed
to be tax policy principles that were grounded upon a consensus amongst
tax policy experts.32 Surrey therefore constructed a position from which
he could criticise the taxation law enacted by a ‘democratic’ legislature.
hose rules which do not correspond with the ideal taxation rules would
be tax expenditures and would be subjected to tax expenditure analysis –
critical assessment of their merits as spending measures. Classiication of
a rule as a ‘tax expenditure’ does not mean that the rule should be abol-
ished – it simply means that the rule should be subjected to closer scru-
tiny under a routine process of tax expenditure management.33 hat is, it
should be critically assessed against speciied public policy principles.
Surrey’s approach relects the conidence of his time in a science of
public policy.34 Like the scientiic study of material phenomena, there
was a general portrayal of, and faith in, good public policy constructed
upon rational grounds. Like all scientists, Surrey hoped that his rational
approach to tax expenditure management would dispel subjective judge-
ments from the realm of tax policy. He constructed his ‘view from nowhere’
upon the ‘general consensus’ among tax policy experts, purportedly free
of any subjective element, from which he could identify tax expenditures.
However, the foundations of Surrey’s concept were quickly subjected to
critical assessment pointing to the insecure foundations upon which his
tax expenditure analysis was constructed.35 he scorn with which Bittker
described the use of ‘tax expenditure’ in the pejorative terms of ‘political’
and ‘rhetoric’ sounds through the ages.36 More recently, others similarly
32
Surrey, Pathways to Tax Reform: he Concept of Tax Expenditures (1973), 17.
33
Surrey and McDaniel, Tax Expenditures (1985), 5–6.
34
For consideration of an early instance of this presentation of the science of tax policy,
see Murnane, ‘Selling Scientiic Taxation: he Treasury Department’s Campaign for Tax
Reform in the 1920s’ (2004).
35
Bittker, ‘A “Comprehensive Tax Base” as a Goal of Tax Reform’ (1967), 985.
36
Bittker, ‘Income Tax “Loopholes” and Political Rhetoric’ (1973). For discussion of the
concept of rhetoric which conveys the scorn with which positivist social scientists use the
term see McCloskey, he Rhetoric of Economics (1998), ch. 1.
10 Introduction
have attempted to construct an apolitical ‘view from nowhere’,37 but their

eforts are subject to the same criticism that they are just another form of
partisan pleading dressed in the ‘objective’ robes of ‘policy science’.38
In chapters 2 and 3 we describe and explore these controversies in
more detail and, we should confess now, throw some fuel upon the ires
of controversy by suggesting that the concept of a tax expenditure could
be expanded to include what we call non-express tax expenditures. Non-
express tax expenditures include a legislative rule which accords with the
benchmark but that is not applied in practice. A non-express tax expend-
iture also includes the case of a government simply not enacting an entire
taxation law notwithstanding that the benchmark suggests that such a
law should be enacted.
1.5.2 Tax expenditure measurement

here are also considerable empirical controversies which we explore in
chapters 3 and 4. Even if a community could agree upon a tax expenditure
concept and upon the appropriate benchmark taxation system, the meas-
urement of tax expenditures is challenging for all sorts of reasons.39 While
it is diicult to measure the iscal cost of tax expenditures as addressed
in Chapter 3, measurement issues extend beyond the mere dollar report-
ing as highlighted in Chapter 4. Identifying the true beneiciary of a tax
expenditure, as opposed to the formal legal beneiciary, involves specula-
tion as to the interaction of market and non-market transactions in rela-
tion to the recipient’s beneit. Moreover, the impact of a tax expenditure
might reach beyond immediate cash transactions – as where an envir-
onmentally damaging tax expenditure creates negative impacts for the
health of a person living in a future time and in another country to that
where the tax expenditure was provided.
1.6 he irresolvable nature of these tax expenditure controversies

he seemingly simple concept of a tax expenditure, created upon a ration-
alist foundation for the apparently noble objects of objectively improv-
ing budgetary process and outcomes, has been mired in controversy.
37
United States of America, A Reconsideration of Tax Expenditure Analysis (2008).
38
Fleming and Peroni, ‘Can Tax Expenditure Analysis Be Divorced from a Normative Tax
Base?: A Critique of the “New Paradigm” and Its Denouement’ (2010).
39
Base: A Critique of the “New Paradigm” and Its Denouement’ (2010), 178–9.
Tax expenditure controversies and political theory 11
he deinition of tax expenditures, the appropriate benchmark and the

measurement of tax expenditures are subjects upon which consider-
able ink has been spilt. Notwithstanding this collective endeavour, there
seems little prospect that the controversies can be resolved. Authors of
comparative studies seem resigned to the fact that diferent jurisdictions
have adopted vastly diferent approaches to tax expenditure manage-
ment, including the very deinition of the concept.40 his variety makes
international comparisons of tax expenditures of limited value – as there
is no uniform basis upon which tax expenditure data is prepared.
In our view, these controversies are irresolvable because they entail
disagreements on fundamental empirical questions and also fundamen-
tal moral disagreements. he empirical disagreements revolve around the
nature of human nature – do we really it the description of homo eco-
nomicus or are we constituted as social beings? he moral disagreements
revolve around the world that we should create for ourselves (and perhaps
other beings as well). here is, we argue, little prospect of discovering the
universal path to resolving those disagreements by recourse to a consti-
tutive rule of the kind liberal philosophers have sought in their quest for a
universal liberal theory of justice that can accommodate democratic pub-
lic institutions.
hus, we agree with the critics of the tax expenditure project – the con-
cept of tax expenditures and the processes by which tax expenditures
might be managed are deeply controversial in the sense that there are no
‘right answers’ as Surrey and others ater him proposed. Where we difer
from the critics is in our response to the existence of these controversies.
he critics would have communities abandon tax expenditure manage-
ment as conceived by Surrey and others. We disagree with this response.
his brings us to the place of tax expenditure management within con-
temporary liberal democracy.
1.7 Tax expenditure controversies and liberal-democratic

political theory
Liberal democratic political theory is concerned with the challenge of
identifying the constitutional rules that underpin legitimate law making.
here are many aspects of liberal political theory, but for present purposes
suice to say that liberalism proceeds upon the basis that an individual
has fundamental rights which no community has authority to take away.
40
OECD, Tax Expenditures in OECD Countries (2010).
12 Introduction
Liberals are uncomfortable with democracy because an unrestrained

democratic legislature might make laws which override what liberals con-
sider to be the fundamental rights of the individual. herefore, liberal
democratic political theory is concerned with specifying the limits of a
democratic legislature to impinge upon what is taken to be the individ-
ual’s domain of freedom and property. If a tax expenditure imposes a bur-
den upon some members of a community, why should those members of
the community be obligated to obey the law imposing that burden?
Within democratic political theory there is a tension between populism
and elitism – should the will of the community be pursued notwithstand-
ing the fact that technical experts consider that the general community’s
will is misconceived? As we have already noted, the literature regarding
tax expenditures typically proceeds upon the basis that tax expenditures
have been expressly legislated. If democratically elected representatives
have created a tax expenditure and/or allow a tax expenditure to con-
tinue, upon what basis can it be said that any tax expenditure is wrong
for a particular community? As we see in Chapter 5, one view is that the
public gets what it wants in the form of legislated tax expenditures and/
or that dissenters are obligated to accept the will of the majority trans-
parently expressed in the democratic process by which tax expenditures
are made. Further, critics of ‘tax expenditure management’ question the
ability of technical experts to identify those tax rules that are harmful
to a community. hese critics argue that the tax expenditure concept is
so imprecise that it cannot provide the ‘view from nowhere’ which can
underpin an objective, critical analysis of legislation passed by a demo-
cratic legislature.41 here is a hint of populist and/or libertarian distrust
of a policy elite purporting to second guess what a community should
achieve with its taxation system, albeit through the community’s elected
representatives.
In this context, it may seem counterintuitive for supporters of the tax
expenditure concept, as we are, to add to the controversy surrounding an
already controversial subject by questioning whether Surrey’s approach
is inappropriate having regard to liberal democratic political theory.
However, we argue that the controversial nature of the subject of tax
expenditures is inevitable for both deontological and empirical reasons.
41
Relected in the judgement of the administration of President George Bush Jr that the
tax expenditure concept was ‘of questionable analytic value’: United States of America,
Analytical Perspectives, Fiscal Year 2002, Budget of the United States Government
(2001), 61.
The thesis of the book 13
In Chapter 5 we argue that diferent approaches to the subject of tax

expenditures are appropriate if one adopts diferent moral imperatives.
his means that if a community is determined to maximise its wealth in
spite of the weak sufering some casualties ‘for the greater good’, a very
diferent approach to tax expenditure management would be appropriate
to that which would be adopted if a community were to set out to ensure
an equal capacity of all to become the person they want to be even if this
meant that communal wealth would be diminished.
1.8 he thesis of the book

As previously stated, the book argues that tax expenditure management is
so important to a government’s implementation of public policy that invi-
tations to abandon tax expenditure management ought to be rejected.
We see tax expenditure management as controversial, but suggest that
this subject is no more controversial than the many other substantive
policy matters with which governments routinely grapple. We note that
the irresolvable controversies surrounding the tax expenditure concept
call for consideration of how such deep moral controversies should be
resolved in a modern democratic setting.
However, the controversial nature of the tax expenditure concept is
compounded by the controversy regarding the constitution of democratic
government. Held42 identiies nine democratic ‘models’ and within any
one of those there is room for argument with respect to key elements of
the model. Meanwhile, liberal democrats grapple with the morally con-
troversial question of the relative weight to be accorded individual rights
within a democracy, no matter which set of procedural rules is taken to be
‘democratic’. hese matters are deeply controversial in the sense that they
are not capable of resolution by recourse to a fundamental constitutive
rule by which disputes might be resolved.
his pluralist approach necessarily limits the prescriptive component
of the book. We cannot construct a ‘view from nowhere’ as Surrey and
others have done, and deign to tell a community that its extant tax rules
are clearly wrong by reference to a universal standard of tax system per-
fection. And nor do we suggest, as the tax expenditure critics seem to
accept by their implicit advocacy for the status quo, that all is for the best
in the best of possible worlds.
42
Held, Models of Democracy (2006).
14 Introduction
In Chapter 6 we set out what we consider to be core elements of a tax

expenditure management system which would meet the demands of
transparency and accountability in a modern representative democracy.
Our views as to what tax expenditure management ought to entail are
grounded upon particular moral stances as to the nature of democracy
and the nature of the society that we think a democratic community
ought to aspire to. hese views are subjective and aspirational. We suggest
that these elements should be subjected to scrutiny in the course of fram-
ing a tax expenditure management system in a particular context. he
outcome of any such deliberation may relect diferent moral stances as
to the nature of democracy and the nature of a community’s aspirations
for what it wants to become. And that, we say, is as it has always been even
though many commentators have not described the practice of govern-
ment in such terms.
1.9 Outline of the book

To sustain this thesis the book is broadly divided into two parts. he pur-
pose of the irst part is to introduce the tax expenditure concept, critically
review the existing literature and to consider the range of intractable con-
troversies with respect to these matters. Chapter 2 examines the diferent
approaches to deining the tax expenditure concept and also the diferent
approaches to identifying tax expenditures. Chapter 3 surveys contem-
porary tax expenditure management practices.
Chapter 4 considers the signiicance of tax expenditures from various
perspectives while acknowledging the empirical diiculties with quanti-
fying the impacts of tax expenditures. Chapter 5 looks back over the the-
oretical controversies described in the preceding chapters and considers
the political challenges that these controversies present in the context of
contemporary understandings of liberal democratic government. he key
purpose of this chapter is to demonstrate that the manner of dealing with
the tax expenditure controversies itself is controversial. Chapter 6 reviews
the preceding chapters before setting out what we think should be consid-
ered during the democratic deliberation upon a community’s approach to
tax expenditure management.
2
he tax expenditures concept
2.1 Introduction
Despite the general acceptance of the existence of tax expenditures,
together with the increasing awareness of their importance and need for
recognition amongst nations around the world, the concept itself, along
with its various elements, is by no means settled. Debate on all aspects
of tax expenditures, whether the pure concept, its purpose, its identii-
cation or the mere measure of tax expenditures, continues unfettered.
his debate simply supports the importance of tax expenditures globally,
along with the need for ongoing contributions.1 Valuable insight into
the dynamic and ongoing nature of the tax expenditures debate can be
gleaned from the words of Surrey and McDaniel in 1985:
he fact that there is debate about the classiication and/or presentation of
a particular item does not mean that the tax expenditure concept is fun-
damentally lawed. he tax expenditure concept requires a dynamic and
continuing analysis of the provisions in a tax system. As the tax expend-
iture concept compels closer consideration of the role of a speciic tax
1
For example, we can contrast the statement contained in the 2002 United States Budget
documents (with the 2010 bipartisan Report of the National Commission on Fiscal
Responsibility and Reform). he 2002 Budget states:
Underlying the ‘tax expenditure’ concept is the notion that the Federal Government
would otherwise collect additional revenues but for these provisions. It assumes an
arbitrary tax base is available to the Government in its entirety as a resource to be
spent. Because of the breadth of this arbitrary tax base, the Administration believes
that the concept of ‘tax expenditure’ is of questionable analytic value.
United States of America, Analytical Perspectives, Fiscal Year 2002, Budget of the
United States Government (2001), 61
In contrast, the 2010 report, which called for the abolition of all tax expenditures, states:
America’s tax code is broken and must be reformed. In the quarter century since the
last comprehensive tax reform, Washington has riddled the system with countless
tax expenditures, which are simply spending by another name. hese tax earmarks –
amounting to $1.1 trillion a year of spending in the tax code – not only increase the
deicit, but cause tax rates to be too high. Instead of promoting economic growth
15
16 The tax expenditures concept
provision (or non provision) in the overall tax system, new studies are
undertaken, new data are developed, and continual rethinking of posi-
tions is required. Especially with respect to borderline issues, unanimity
of opinion on classiication issues is rare. But the debates and analyses are
themselves important contributions to the continuing improvement of a
country’s tax and spending structures. Moreover, the number of classii-
cation or presentation issues that have inspired debate is insigniicant in
relation to the number of items on the tax expenditure list about which
there is no disagreement.2
We aim, as Surrey suggests, to contribute to the necessary rethinking of

the positions required for tax expenditure management. he diiculties
surrounding the tax expenditure concept and tax expenditure analysis
have been discussed in both academic literature and government reports
since the genesis of the concept by Surrey in 1967 – over 40 years ago. hese
diiculties began with its theoretical constructs but have now extended
to the practical implementation of the concept by governments around
the world. It is not easy to unravel the diiculties associated with the tax
expenditure concept as most, if not all, are inextricably tied to each other.
Yet, ultimately, the diiculties stem from the basic requirement to identify
a provision of the tax system as a tax expenditure so deined. his iden-
tiication is generally determined according to a deined normative tax
base or, more precisely, as a deviation from a deined tax benchmark. he
essential nature of this question, along with the inevitable disagreement
both domestically and internationally as to what is a true representation
of the tax base, has at times distracted from the practical signiicance of
the identiication of tax expenditures and the beneits of tax expenditure
analysis, reporting and management which we discuss in Chapter 4. his
controversy in terms of tax expenditure analysis, which is at the end of
the day an academic one, has ultimately led to calls by some to abandon
the tax expenditure analysis and management process or to consider tax
expenditures not as a tax policy tool or budgetary tool, but rather solely,
and more broadly, as a matter of institutional design. he purpose of this
chapter is to outline those controversies which we do not dispute are argu-
ably signiicant. In Chapter 5 we argue that the way forward with respect
to the controversial, but socially signiicant tax expenditure concept is to
and competitiveness, our current code drives up health care costs and provides spe-
cial treatment to special interests. he code presents individuals and businesses with
perverse economic incentives instead of a level playing ield.
United States of America, he Moment of Truth (2010)
2
The concept of tax expenditures 17
explicitly acknowledge that the nature and purpose of tax expenditure

analysis and management is governed by one’s moral and political philoso-
phy. In essence, if we accept that there will always be divergence in the aca-
demic debate surrounding tax expenditures and focus on this acceptance,
a pragmatic perspective can prevail; one which requires a jurisdiction to
identify its own purpose of tax expenditure reporting and analysis.
h is chapter critically analyses, through an historical lens, the dif-
iculties associated with the theoretical constructs of tax expenditure
analysis. he object of this analysis, rather than being to resolve these
controversies, is to demonstrate the moral pluralism that abounds, with
the diverse approaches relecting the subjective preferences of their
respective author(s). By introducing the tax expenditure concept and
critically reviewing the existing literature, we are able to identify the
range of intractable controversies regarding the concept itself and, in
Chapter 3, its application in speciic jurisdictional contexts. A critical
review of the controversies surrounding the tax expenditure concept,
its application and the problematic measurement of tax expenditures
is essential to the understanding of the concept of tax expenditures. In
particular, the existing theoretical literature demonstrates the breadth
and depth of the controversies surrounding the tax expenditure concept.
Moreover, the controversies regarding the concept, whilst grounded
in the dei nition of the tax benchmark, extend to the debate on rival
measures of the quantum of tax expenditures, the speciication of a tax
expenditure reporting standard and the process by which tax expendi-
tures ought to be managed.
his discussion on the controversies surrounding tax expenditures
does not attempt to provide a contemporary, potentially elitist, prescrip-
tion for the reporting, management and analysis of tax expenditures.
Rather, it aims to highlight the choices one must make when designing a
robust tax expenditure management regime which accepts the inluences
of a jurisdiction’s political philosophy. To achieve this goal, this chapter
speciically considers the concept of tax expenditures, the purposes of the
categorisation of tax expenditures such as accountability and transpar-
ency, the diferent identiication of tax expenditures, the measurement of
tax expenditures and contemporary expansions of the concept.
2.2 he concept of tax expenditures

he concept of tax expenditures is elusive. Yet, the central tenet to the
tax expenditure debate is the deinition of the concept itself. As such, a
discussion on tax expenditures of the type taken in this book warrants

an initial analysis of how the term is deined because, despite the concept
being around for half a century, tax expenditures are neither well known
outside those directly involved with their use, analysis and reporting,
easily deined even by those who are directly involved, nor easily under-
stood. Ask the average citizen what a tax system is designed to do and
they will tell you that it raises money from the general population to con-
tribute to the public cofer. Any tax system is generally, and primarily,
regarded as a mechanism by which governments raise revenue. Much of
this revenue raised is used to support public spending programmes, par-
ticularly the provision of goods and services which may not be adequately
provided by the market. here is, however, a secondary purpose for a tax
system: it is a useful instrument though which a government can imple-
ment social and economic policy.3 he use of the income tax regime for
social and economic purposes allows a government an indirect (as con-
trasted with direct spending) means of potentially achieving its policy
objectives. While outside the traditional revenue raising objective of
any income tax regime this is a common phenomenon, with the rele-
vant provisions known as ‘tax expenditures’. When a tax system is used
for this purpose, tax expenditures have the potential to steer taxpayers
towards or away from certain behaviour by either imposing costs on, or
providing beneits to them. However, both the use and the efect of tax
expenditures within a country’s tax regime is generally not well under-
stood by citizens. Yet, the use is oten signiicant both in terms of the
sheer number of expenditure provisions within tax legislation and the
efect of tax expenditures from a iscal perspective, as they oten result in
a signiicant reduction in the amount of revenue raised compared with a
benchmark tax system. he benchmark tax system, one of the most con-
troversial aspects of tax expenditure reporting and analysis, is discussed
later in this chapter.
2.2.1 he phrase ‘tax expenditures’

he phrase ‘tax expenditures’, as i rst coined by Stanley Surrey in
1967, has been deined in various ways and, to date, despite decades of
3
For an analysis of speciic examples see Stead, ‘Implementing Disaster Relief hrough Tax
Expenditures: An Assessment of the Katrina Emergency Tax Relief Measures’ (2006) who
argues that aid for Hurricane Katrina victims in the form of tax relief was distorted in a
manner typical of similar tax relief. See also Cavanaugh, ‘On the Road to Incoherence:
Congress, Economics, and Taxes’ (2002).
controversy and an increasing number of governments adopting compre-

hensive tax expenditure reporting, there is no universally agreed dein-
ition. Deinitions adopted by various nations and international bodies are
discussed in Chapter 3. However, a few of those deinitions which have
developed include:
• ‘those revenue losses attributable to provisions of the Federal tax laws
which allow a special exclusion, exemption, or deduction from gross
income or which provide a special credit, a preferential rate of tax, or a
deferral of tax liability; and the term “tax expenditure budget” means
an enumeration of such tax expenditures’;4
• ‘the estimated costs to the tax revenue of preferential treatment for spe-
ciic activities’;5
• a provision that is a ‘substitutable tax provision’. hat is, a tax provision
‘that can be replaced with a non-tax-based federal program that fuli ls
the current tax provision’s purposes at least as efectively as does the
current provision itself’;6
• a provision that performs a mainly allocative task and therefore can be
identiied as exceptional within the primarily distributive income tax
framework;7
• a provision comprising a ‘tax subsidy’8 and/or a ‘structural income dis-
tortion’. A tax subsidy is a speciic legislative rule that departs from a
general taxing rule that might be discerned from the legislation over-
all. A structural income distortion is, it seems, a speciic legislative rule
that ‘materially afect[s] economic decisions in a manner that imposes
substantial eiciency costs’;9 and
• a tax expenditure as a provision in the tax code that is expressly identi-
ied as promoting a ‘spending’ purpose.10
he various diferent ways that the same concept is described means
that the phrase ‘tax expenditures’ is simply an idea, an expression, or the
4
2 United States Code § 622(3).
5
OECD, Best Practices for Budget Transparency (2002), 7 (para 2.2).
6
huronyi, ‘Tax Expenditures: A Reassessment’ (1988), 1186. Similarly, see Steuerle,
Contemporary US Tax Policy (2004) (dei ning ‘tax expenditures’ as ‘spending programs
channelled through the tax system … hese tax provisions generally grant special tax
relief to encourage certain kinds of behaviour by taxpayers or to aid taxpayers in special
circumstances.’).
7
Shaviro, ‘Rethinking Tax Expenditures and Fiscal Language’ (2004).
8
Witte, he Politics and Development of the Federal Income Tax (1985).
9
10
McIntyre, ‘A Solution to the Problem of Dei ning a Tax Expenditure’ (1980–81).
language used to describe a group of laws within a tax regime. Within

the context of a government’s iscal policy, it is no diferent to terms such
as ines, fees, levies, deicits, surpluses and pensions, to name a few.11
Each is an integral part of the language used by governments around
the world. Each is grounded in theoretical or normative underpinnings.
Each is attached to a jurisdiction’s moral and political philosophy. Yet no
other term has attracted the international debate that has attached to tax
expenditures and tax expenditure analysis. We propose that the reason
for this is the very reason why tax expenditure management cannot be
abandoned. Tax expenditures are simply too signiicant in terms of their
impacts on the international economy, the iscal position of the states, the
constitutional framework of countries, public administration in general
and tax administration in particular, and the legitimacy of governments.
We justify this claim later. For now, we use the more pragmatic justiica-
tion that tax expenditures are part of the tax systems of every developed
country around the world.12
2.2.2 he original intent of the ‘tax expenditure’ concept

he ‘tax expenditure’ concept, as it was irst identiied, was designed to
demonstrate the similarity between direct government spending, on the
one hand, and spending through the tax system on the other. In essence,
the identiication of beneits provided through the tax system as ‘tax
expenditures’ allows analysts to consider the iscal signiicance of those
parts of the tax system which do not contribute to the primary purpose
of revenue raising. Where a taxpayer is entitled to a tax expenditure that
taxpayer has paid the tax due under the normal tax structure and then
received a reduction in tax liability due to the tax expenditure entitle-
ment. It is essentially a short cut to a direct payment by a government. he
reason for this is that government programmes can generally be imple-
mented in one of two ways. his can be demonstrated by considering two
alternate scenarios:
11
12
OECD, Tax Expenditures in OECD Countries (2010), 14. See also Polackova Brixi et al.
(eds), Tax Expenditures – Shedding Light on Government Spending hrough the Tax
System: Lessons from Developed and Transition Economies (2004); Nordic Working
Group, Tax Expenditures in the Nordic Countries (2010); Villela et al., Tax Expenditure
Budgets: Concepts and Challenges for Implementation (2010); and Burton and Stewart,
Promoting Budget Transparency hrough Tax Expenditure Management: A Report on
Country Experience for Civil Society Advocates (2011).
Scenario 1:
A government wants to contribute to half the costs associated with a child’s educa-
tion up to a maximum amount of $500. he government can collect $500 in tax from
a taxpayer (a parent of the child) and pay a direct subsidy of $500 to that taxpayer
(parent) on proof that $1,000 has been spent on a child’s education. his would be
considered a direct government expenditure.
Scenario 2:
A government decides not to tax the taxpayer (parent) to the extent of $500 in
circumstances where tax policy indicates the government ought to collect that tax
(a tax expenditure). h is is achieved by the taxpayer (parent) being given a $500
credit of their tax liability upon proof that the taxpayer has spent $1,000 on a child’s
education. hat is, a 50 per cent credit is given up to a maximum of $500.
In both of the scenarios outlined above, the taxpayer is in the same pos-
ition economically, with the process of collecting the tax and then mak-
ing a payment avoided in Scenario 2, arguably resulting in administrative
eiciency. However, in almost all respects the delivery of these govern-
ment beneits is treated diferently. At the broadest level, for budget pur-
poses in most, if not all jurisdictions, the direct subsidy of $500 outlined
in Scenario 1 will be exposed to greater political and/or critical scrutiny
than the tax expenditure outlined in Scenario 2.
he example provided is one which uses a tax credit as the form of deliv-
ery of the tax expenditure. However, there are many other ways to achieve
a reduction in tax payable to speciic classes of taxpayers either because
of certain taxpayer characteristics or because of certain behaviour. his
may be achieved via a deduction for expenses not normally allowed, an
accelerated or increased deduction for normal expenses, the deferral of
the recognition of certain income or an outright exemption from tax-
ation for income that would otherwise be assessable. What is common to
the delivery of a tax expenditure to taxpayers, however, is the necessity of
the taxpayer or the activity to meet the qualifying factor/s. For example,
a tax deduction is oten available for philanthropic activities. he quali-
fying behaviour may be a donation to speciied charities. Research and
development undertaken by certain entities may attract a concession in
the form of accelerated deduction or credit. he qualifying behaviour is
the need to undertake eligible research and development. he very need to
meet criteria of itself raises issues such as gaming by taxpayers, which we

address later.
2.2.3 Negative tax expenditures

he tax expenditures discussed so far are known as positive tax expen-
ditures and are the main focus of any tax expenditure reporting and
analysis undertaken by jurisdictions around the world. However, there
is a second category of tax expenditures: negative tax expenditures.
Jurisdictions oten also report on negative tax expenditures, or what are
regarded as provisions of the tax regime which treat certain taxpayers or
activities disadvantageously. Where a negative tax expenditure applies,
those taxpayers or activities are taxed at a higher rate compared to the
benchmark tax system so deined. For example, taxes imposed on ciga-
rettes and alcohol may be regarded as negative tax expenditures. hese
types of tax expenditures, despite being revenue raising in nature, have
the same purpose as positive tax expenditures, that is, the achievement of
a social or economic policy goal.
2.2.4 A useful starting point

Dei ning tax expenditures in terms as we have just done is a useful start-
ing point in any discussion on tax expenditures. Understanding what
constitutes a tax expenditure in simple terms means that it is possible
to consider the purpose of classifying parts of a tax regime as belong-
ing to that category of provisions which are not about revenue raising
but rather are part of the broader social and economic policy of a gov-
ernment. It allows the type of analysis undertaken in a book like this,
and it allows an understanding of the importance of the reporting and
management of tax expenditures as part of any democratic process.
However, we do not purport to claim that this discussion on the dei n-
ition of tax expenditures is comprehensive and complete – in fact, far
from it. To this end, we also argue that the dei nition of tax expendi-
tures should go beyond those traditional dei nitions discussed above to
remove the prerequisite of being expressly legislated. We argue that this
prerequisite provides an unnecessarily narrow view of what constitutes
a tax expenditure and is a restriction on the types of activities captured
within the dei nition. Later, we outline our reasoning for expanding
the dei nition to include implicit and operational tax expenditures. For
The purpose of categorisation 23
now, we turn to the purpose of categorising a provision of the tax regime

as a tax expenditure.
2.3 he purpose of categorisation

By now it will be obvious to the reader that inherent in tax expenditure
analysis is the need to categorise the various provisions of an income tax
regime as tax expenditures. Before undertaking that categorisation, how-
ever, a consideration of ‘why’ this categorisation is undertaken is required
as it clearly inluences and ultimately shapes the identiication process.
his is also a problematical question, as the purpose of categorisation is
no less decisive than the concept itself. Some see tax expenditure analysis
as an important tool in tax reform.13 Others see it as a tool in spending
reform.14 Yet, these diferent purposes will shape a nation’s fundamen-
tal choice as to how it identiies and treats its tax expenditures. A nation
which adopts the view that the purpose of tax expenditure analysis is
designed to promote eiciency and/or equity within its tax regime is
likely to undertake a diferent approach to tax expenditure analysis than
a nation which adopts the view that tax expenditure identiication and
reporting is designed to promote political engagement. he purpose of
tax expenditure analysis and reporting adopted by a nation will also have
13
For further explanation of the two purposes see: Shannon, ‘he Tax Expenditure Concept
in the United States and Germany: A Comparison’ (1986). Shannon, at p. 202, states:
he tax expenditure concept and analysis have important implications for tax reform.
he concept divides tax law into those provisions that constitute tax expenditures
and those provisions that do not. According to tax expenditure analysis, tax expen-
ditures should be evaluated as spending, rather than as taxing provisions. hus tax
expenditure analysis helps focus tax reform. With respect to tax expenditures, tax
reform becomes the process of purging from existing law and proposed legislation
those tax provisions identiied as tax expenditures that would not be justiiable if cast
as direct expenditures.
He then goes on to state:
In addition to providing a focus for tax reform the tax expenditure concept also has
important implications for the budgetary process. If tax expenditures are essentially
equivalent to direct expenditures channelled through the tax system, it would be
irresponsible not to take them into consideration in the budgetary process. If tax
expenditures represent money the government spends, any government budget that
does not account for tax expenditures will be incomplete. he problem of calculat-
ing accurately the amount of a tax expenditure, therefore, is much more important
within the context of the budgetary process, than within the context of tax reform.
14
McDaniel, ‘Identiication of the “Tax” in “Efective Tax Rates”, “Tax Reform” and “Tax
Equity” ’ (1985), 277.
a fundamental efect on the benchmark it adopts to identify tax expendi-

tures within its tax regime.
his diference in ‘purpose’ has stemmed from Surrey’s initial adoption
of a tax expenditures list where the purpose was narrowly deined. His
aim was to make a limited list of provisions which did not it the usual
purpose of taxation law and for this reason he adopted a narrowly deined
tax expenditure concept framed upon the criteria of the provision being
(1) an expressly legislated taxation rule, (2) within the income tax, (3)
that departs from the speciied income tax benchmark and (4) that is dir-
ectly substitutable for direct spending.15 Surrey’s purpose was to identify
spending rules which were ‘hidden’ in the United States’ federal income
tax code with a view to applying spending criteria to assessing the merits
of these rules. Surrey therefore envisaged tax expenditure reporting as a
key step in fuli lling his relatively narrow purpose of treating all govern-
ment spending uniformly.16
An alternate purpose of tax expenditure reporting can be critical assess-
ment of the operation of a taxation system assessed against deined moral
norms. For example, if a community seeks to maximise economic ei-
ciency and/or social justice, it might express such aspirations by deining
how its taxation system might promote such goals. To assess the impact of
the tax system against such goals a broader approach to tax expenditure
reporting would relax Surrey’s restrictive speciication of the tax expend-
iture concept. In particular, the irst, second and fourth criteria might be
excised so that a tax expenditure would arise whenever a tax rule departed
from the speciied benchmark.
2.3.1 Framing the question of what is a tax expenditure

Tax expenditure analysis, or the identiication of certain provisions of
a tax regime for the purposes of comparison between provisions of the
tax law which are special deductions, credits and other allowances, with
government subsidies, was irst recognised in Germany as early as 1954.17
However, it wasn’t until 1967 that the phrase ‘tax expenditures’ entered
the sphere of mainstream tax policy and tax expenditure analysis became
15
For a discussion by Surrey on what he considers the dei nitional aspects see Surrey,
Pathways to Tax Reform: he Concept of Tax Expenditures (1973), 15.
16
For a recent consideration of tax expenditures in the context of the United States’ tax
regime, see United States of America, he Moment of Truth (2010).
17
Shannon, ‘he Tax Expenditure Concept in the United States and Germany: A
increasingly popular amongst developed jurisdictions. he phrase ‘tax

expenditures’ was initially coined for the purpose of highlighting provi-
sions of the tax regime which were, in efect, direct spending provisions.
his purpose remains a central tenet to the reporting of tax expenditures
by an increasing number of jurisdictions. However, as previously stated,
despite almost universal acceptance of the concept of tax expenditures,
there is no one universally accepted deinition of the term. We argue
that the lack of consensus stems from the divergence of opinion on the
underlying and central concept that tax expenditures are attempting to
identify.
Until this point in the book, we have couched tax expenditures in sim-
ple terms as being the equivalent of direct spending programmes; how-
ever, the issue of identifying tax expenditures is arguably much more
complex. Is a tax expenditure a provision contained in the tax legislation
that is directly substitutable for a spending programme? Is a tax expend-
iture a provision which provides preferential tax treatment to one group
of taxpayers over another group of taxpayers? Or, is a tax expenditure
a much wider concept that captures all departures from a normative or
benchmark tax system regardless of whether they can be substitutes for a
direct spending programme or provide beneits to a speciic class of tax-
payer? Put simply, the answer will vary depending on who is asked this
question and, as such, despite the longevity of the concept its deinition
is by no means settled. We do not purport to provide a deinitive answer
to the question of what a tax expenditure is, but rather argue that there is
no one answer. It is precisely a jurisdiction’s underlying purpose for cat-
egorising provisions of a tax regime as tax expenditures which frames the
answer as to how that jurisdiction deines a tax expenditure for its own
purposes. However, identiication of tax expenditures is at the core of any
analysis as before tax expenditure analysis and management can occur,
in whatever form deemed appropriate by a jurisdiction, it is necessary to
determine which provisions of a tax regime are considered tax expendi-
tures. hroughout the literature, various criteria have been recognised as
possible approaches to identifying tax expenditures, each of which have
an inherent purpose attached. hese include the following: the pursuit
of a non-iscal policy goal, convertibility of the provision into a direct
expenditure, the beneit of a limited group of taxpayers, the reduction of
revenue and the deviation from a benchmark tax structure.18
18
OECD, Tax Expenditures in OECD Countries (2010), 15–16, citing the criteria identiied
by Ende et al., ‘Tax Expenditures in the Netherlands’ (2004).
2.3.2 he elusive underlying concept

It has been stated that there is no universally accepted theoretical or prac-
tical deinition of a tax expenditure. Further, those who have attempted to
deine the term invariably link it to another, just as elusive, concept. hese
secondary concepts are generally associated with some form of preferential
tax treatment, being a substitutable (for direct spending) tax provision, or
promoting a purpose other than direct iscal revenue raising. Each of the
diferent deinitions gleans insight into, in the case of academic work, the
author’s or, in the case of country reporting, the reporting body’s particu-
lar views on the purpose of identifying tax expenditures. Consequently,
the concept is irst and foremost dependent on the purpose for which the
term is being used. he literature on tax expenditures reveals the pur-
poses of identifying tax expenditures as falling into two broad categories:
irst, promoting eiciency within the tax regime speciically and budget-
ary process generally; and second, promoting equity along with political
engagement via a discussion and debate on the relative merits of speciic
tax expenditures and their associated iscal costs.19 Jurisdictions do not
have to explicitly state one purpose but rather may adopt one or both of
these broad purposes for identifying tax expenditures. Ideally, the pur-
pose adopted is explicit; however, oten it is merely implicit in the use for
which a jurisdiction analyses tax expenditures and must be extrapolated
from various documents. Whether explicit or implicit, once the purpose
of identifying tax expenditures is determined, it is then possible to ascer-
tain the means by which they are identiied or, more speciically, the cri-
teria which provide the benchmark which underpins tax expenditure
identiication, which is the most controversial of all discussions in the tax
expenditures debate. It is also possible to consider other issues such as
measurement which arise as a consequence of the diferences in the expli-
citly deined or inherently adopted purpose of tax expenditure reporting
and analysis by a jurisdiction and subsequent variances in the recognised
benchmark.
We argue that lack of agreement in relation to the tax expenditure con-
cept can be explained if we accept that the controversial nature of the sub-
ject of tax expenditures is inevitable for both deontological and empirical
19
Within the context of the United States, see, for example, Woellner, ‘Spending on an
Empty Wallet: A Critique of Tax Expenditures and the Current Fiscal Policy’ (2006), who
analyses several tax expenditures which he claims are counterintuitive to accepted tax
policy norms.
reasons. History has proven that an elitist model is not possible. As such,
it is the criteria underlying the benchmark which we argue low from the
diferent moral assessments of what a community values which shapes a
nation’s tax expenditure analysis and management. A community must
make choices about how its tax system is to operate and must weigh the
competing imperatives against each other, with these choices becoming
embedded in the particular view of the tax expenditure concept adopted.
he reason why the identiication of tax expenditures which requires the
identiication of a benchmark is by no means settled is arguably because
there are two broad schools of thought as to the imperative which tax
expenditure analysis is attempting to achieve. Fleming and Peroni make
the point that for tax expenditure analysis to be more than rhetoric, it
must possess a principled basis for it to have any normative force,20 a point
with which we agree. However, that basis will be inluenced by a nation’s
views on fundamental tax policy principles or, more speciically, ei-
ciency and equity.21
2.3.3 ‘Good’ tax criteria

Traditionally, tax expenditure analysis has been grounded in equity
principles on the basis of the normative benchmark being the Schanz–
Haig–Simons deinition of income, which we discuss later. As Fleming
and Peroni point out, however, the Schanz–Haig–Simons dei nition of
income derives from the ‘ability to pay’ principle.22 As such, where it is
accepted that a nation models its tax system on ability to pay, the bench-
mark will relect this. Equity, or fairness, nevertheless will not be the only
criteria upon which a tax system will be based, and oten a competing
concept is one of eiciency or neutrality. As such, there is the argument
that the tax expenditure benchmark should be designed on the basis of
eiciency, where tax expenditure analysis will reveal those provisions of
the tax regime which distort the efect of tax. We do not purport to argue
that equity is more important than eiciency, or vice versa. Rather, we
argue that this is for a nation to decide. Many jurisdictions will ultimately
20
Fleming and Peroni, ‘Reinvigorating Tax Expenditure Analysis and its International
Dimension’ (2008), 450.
21
See Fleming and Peroni, ‘Reinvigorating Tax Expenditure Analysis and its International
Dimension’ (2008), for a comprehensive discussion on the fundamental tax policy prin-
ciples relating to tax expenditure analysis.
22
he ‘ability to pay’ principle is the proposition that taxes should be levied according to a
taxpayer’s ability to pay tax. herefore, the more income, the higher the tax.
adopt a hybrid approach to the benchmark used for deining tax expen-
ditures, starting with the Schanz–Haig–Simons deinition of income and
then adapting it for other criterion of good tax policy, such as economic
eiciency, but also factors such as enforceability, administrability, simpli-
ication and efectiveness.23
An analysis of the approaches by diferent countries, which is more
comprehensively discussed in Chapter 3, reveals that jurisdictions have
historically considered that tax expenditure analysis and reporting serves
diferent purposes, with their own assessment of the conlicting impera-
tives of good tax policy. Inextricably tied to this assessment is the func-
tion for which tax expenditures are then identiied by these nations. he
impetus for tax expenditures reporting in the United States, by exposing
tax expenditures as the equivalent in outlays of direct spending, was tax
reform. Whilst in Germany, the emphasis on the tax expenditure con-
cept has centred around budgetary policy rather than reform24 with the
budgetary process in Germany taking into account the cost of tax expen-
ditures.25 We argue that no matter the fundamental purpose adopted by
a nation and no matter the tax policy imperatives adopted, a robust tax
expenditures management framework is arguably an essential element to
both of these purposes. his leads us to a discussion on the function of
identifying tax expenditures.
2.4 he function of identifying tax expenditures

Tax expenditures are now a major feature of not only the tax system, but
also government policy and the overall economy and, consequently, the
tax expenditure concept is central to a government’s iscal function.26 In
essence, the identiication of beneits provided through the tax system as
tax expenditures allows analysts the obvious function of considering the
iscal signiicance of those parts of the tax system which do not contrib-
ute to the primary purpose of revenue raising.27 However, we argue that
23
Dimension’ (2008), 458.
24
25
26
See the material cited in Chapter 1, note 22.
27
For a discussion of the Bush Administration’s criticisms of the concept of tax expend-
iture and its questioning of the analytic value, see Roin, ‘Truth in Government: Beyond
Tax Expenditure Budget’ (2003).
The function of identifying tax expenditures 29
there is more to tax expenditures than simply determining the iscal value
of each identiied expenditure and evaluating that identiied expend-
iture within a direct expenditure framework. If a nation so chooses, it
also creates the opportunity for enhancing the legitimacy of tax systems
speciically and government more generally, that is, the promotion of pol-
itical engagement. As such, tax expenditure analysis has two distinct but
equally important functions: it has important implications for a nation’s
budgetary process, and it is a tool used in the tax reform process. A third
function that is also becoming increasingly relevant, and which may be
described as a by-product of the two primary functions, relates to the
international signiicance of tax expenditures, particularly in a compara-
tive context.
2.4.1 Reporting on tax expenditures

he genesis of modern tax expenditure analysis was the simple process
of the reporting of those provisions regarded as tax expenditures for the
purposes of revealing them as the equivalent of direct spending outlays.
Put simply, traditionally, the central tenet of tax expenditure analysis is
the notion that tax expenditures are directly comparable to direct gov-
ernment expenditures. he argument profered is that if equivalent direct
government expenditure is not justiiable because it is either ineicient
or inequitable, the tax expenditure is also ineicient or inequitable and
should not be part of any tax regime. he comparison of tax expendi-
tures with direct expenditures adopts the view that tax expenditures are
a transfer of public resources, achieved by reducing tax obligations with
respect to a benchmark tax.28 Determining which tax expenditures are
equivalent to direct spending programmes, along with the iscal costs
associated with the tax expenditure, has traditionally been the primary
purpose for reporting on tax expenditures. In turn, this has led to a sig-
niicant number of governments producing tax expenditures statements.
Consequently, this makes tax expenditures an integral part of a coun-
try’s budgetary process. Once it is accepted that tax expenditures are part
of the budgetary process, the substantive elements of the reporting are
revealed. hat is, there is a need for the reporting of tax expenditures to
include not only the measurement of those expenditures as is tradition-
ally done, but also contain a rigorous review and oversight process. We
refer to this as the management of tax expenditures.
28
here is no doubt that the most signiicant advances in tax expenditure

analysis by revenue authorities around the world has been in the report-
ing of tax expenditures, with a traditional emphasis on iscal costs. his
increased level and sophistication of reporting is evident in the discussion
in Chapter 3 when we consider the various approaches adopted by dif-
ferent jurisdictions. From an economic perspective, the most signiicant
impact of a government’s tax expenditures programme is the scale of the
programme. he scale of the programme relates not only to gross rev-
enue foregone as measured in the reporting process but also as a measure
relative to the scale of direct spending.29 If it is accepted that tax expendi-
tures are directly comparable to direct expenditures as is suggested here,
it necessarily follows that the cost should also be included in any govern-
ment’s budget. As such, the reporting of tax expenditures from a iscal
cost perspective has traditionally been and continues to be the domin-
ant purpose for the categorisation. A consequence of the reporting of tax
expenditures is the ability of a nation to expose many of the perceived
advantages as well as failings of these expenditures as contrasted with dir-
ect expenditures. We discuss the importance of revealing these failings in
Chapter 4.
he decision by a nation to report tax expenditures is a signiicant part
of the reason for categorising provisions of a tax regime as tax expen-
ditures as contrasted with other provisions of the tax regime. hat is,
reporting is a means in itself through the public disclosure of tax expen-
ditures. However, arguably, this traditionally adopted reason for report-
ing tax expenditures has its limitations, and jurisdictions which limit the
purpose of identifying tax expenditures to mere reporting (disclosure)
are not fully embracing the broader purpose of the tax expenditure con-
cept. Generally, the reporting of tax expenditures is simply designed to
facilitate a comparison between direct spending decisions and revenue
foregone through tax expenditures. Consequently, a limiting factor is the
fact that tax expenditure statements generally only consider existing tax
expenditures, and then, as previously emphasised, only consider those
existing tax expenditures in terms of reporting rather than any substan-
tive management. As such, the traditional approach to tax expenditures
is arguably deicient in that it is both limited to an ex post analysis and it
is a mere analysis rather than a genuine management framework. If the
29
Many countries do not report aggregate tax expenditure data. However, in Australia
aggregate measured tax expenditures for the 2007/2008 i nancial year amounted to 7.1
per cent of GDP: Australia, 2007/2008 Tax Expenditures Statement (2008), 14.
reporting of tax expenditures serves the purpose of being an important

tool in spending reform it is imperative that the type of analysis under-
taken is extended to an ex ante review of all proposed tax expenditures.
his is simply because budget control is diicult where new tax expendi-
tures are introduced without being subject to rigorous review and future
oversight.
2.4.2 Accountability and transparency

Whilst the traditional rationale for categorising a provision of the tax
regime as a tax expenditure is the reporting of that provision as a devi-
ation from the benchmark so deined and an ascertaining of the is-
cal signiicance of both the individual and aggregated tax expenditures,
jurisdictions have also recognised the signiicance of the identiication
of tax expenditures from a public policy and administrative perspective.
A democratic government should aspire to openness, accountability and
transparency in relation to all of its policy measures and this will include
tax expenditures. Tax expenditures are a signiicant social and economic
policy instrument applied across the full policy spectrum.30 By their very
nature, tax expenditures are designed to compensate or reward particular
taxpayer behaviour, or, in the case of negative tax expenditures, designed
to punish particular behaviour and impose a cost on negative external-
ities. Given that tax expenditures are a signiicant policy instrument, ques-
tions will always arise as to accountability and transparency within the tax
regime when tax expenditures are used. he tax expenditure concept and
analysis, by dividing provisions of tax law into those that are part of the
normative base and those which are not, plays an important role in both
the reform and integrity of any tax regime for jurisdictions which adopt
the view that the tax expenditure concept is designed to aid reform.
he usefulness of tax expenditure analysis as a policy tool leads us back
to the earlier discussion on the purpose of categorisation by individual
nations and the competing tax policy imperatives. Previously, authors
have argued that the usefulness of tax expenditure analysis depends on
one’s view as to the objective or subjective nature of the identiication of
tax expenditures. Many view tax expenditure analysis as highly subjective
but used under the guise of being an objective standard, while there are
others who do not perceive any subjectivity in the process and argue that
30
Howe and Landau, ‘Do Investment Attraction Incentives Create Decent Jobs? A Study of
Labour Conditions in Industry Assistance Contracts’ (2008).
‘tax expenditure analysis is attractive as a tool for tax reform because it is

objective’.31 We argue that it is neither objective in the traditional sense
of the word, nor is it without merit. Surrey originally proposed an elitist
tax expenditure model, in which he attempted to build in objectivity, with
the purpose of designing tax expenditure reporting and analysis to reveal
tax expenditures as the equivalent of direct spending programmes with
the aim of tax reform, or essentially a reduction in the number of spend-
ing programmes implemented through the tax regime. To this end, tax
expenditure reporting has been a dismal failure as we are at a point in time
where the use of tax expenditures is pervasive and growing.32 his growth
is not only in the iscal value of tax expenditures but the number of tax
expenditures in existence. However, this does not automatically lead to the
conclusion that a lack of objectivity results in a call for the abandonment of
tax expenditure reporting and analysis. Rather, it leads to the conclusion
that tax expenditure reporting and analysis is grounded in a nation’s own
moral and political philosophy. here will in this case still be a principled
basis for using tax expenditure analysis as a policy tool.
While we argue that the tax expenditure concept is not objective,
the view that tax expenditure analysis is an objective analytical tool is
not without merit, particularly when we consider the number of items
which are considered tax expenditures, and to which no one would doubt
this categorisation. To this end, the underlying purpose for tax expend-
iture analysis may not be objective, meaning that the benchmark across
nations will vary, but the low on processes, that is, the actual analysis
and management, may be objective. he advantage of realising the ben-
eits of accepting tax expenditure analysis and management as objective
‘at a time when many government budgets are threatened by population
ageing and adverse cyclical developments’33 is that tax expenditure ana-
lysis aids governments in avoiding ineicient spending programmes,
some of which may utilise tax expenditures.34 Further, if we accept that
tax expenditure reporting and analysis is unlikely to reduce the num-
ber of tax expenditures within a regime and the iscal cost of those tax
expenditures, there is even greater importance in the recognition of tax
expenditures for accountability and transparency purposes. he OECD
seems to adopt this stance when it states that ‘there is a perceived need
31
32
33
34
for greater understanding of the issue, of the trend in tax expenditures,

and of successful practices with respect to their enactment, budgetary
reporting, and review’.35 he OECD is not focusing on a comparison of
tax expenditures with direct expenditures but rather the accountability
and transparency which results from the reporting of tax expenditures
per se. As such, it becomes increasingly important to distinguish between
the separate parts of any income tax regime to ensure that all government
programmes, whether via direct spending or tax expenditures, are sub-
ject to the same degree of scrutiny and accountability.
Opponents to the tax expenditure concept would disagree with us
as they are likely to continue to argue that the notion that tax expend-
iture analysis is an objective policy tool can be dismissed on the basis
of the controversies surrounding the identiication of tax expenditures.
However, where it is agreed within a nation as to the fundamental under-
pinnings of the tax expenditure concept, reporting on tax expenditures
promotes analysis of tax policy principles and potential improvements
to the tax system for that particular community. Put simply, the discus-
sion that surrounds tax expenditure reporting, such as the deinition of
tax expenditures and the requisite tax benchmark, allows a consideration
both by a country’s citizens as a whole as well as its government as to
what that society wants for its tax system. Citizens are then able to make
a more informed choice about what it considers to be a fair, eicient and
simple tax system. By reporting on tax expenditures, not just in terms of
iscal value, but also for the purpose of accountability and transparency,
many more of their negative characteristics may be revealed leading to
the conclusion that tax expenditures are oten a ‘second best’ policy ater
direct spending.36 Surrey and McDaniel, in 1985, provided a list of the
characteristics they believed were exposed by tax expenditure analysis,
cataloguing the diferent negative features of tax expenditures.37 hese
negative features have been highlighted subsequently in voluminous lit-
erature. We return to these in Chapter 4.
2.4.3 International aspects of tax expenditures

Within domestic jurisdictions, the identiication of tax expenditures has
been for the purpose of budgetary considerations and/or for the purpose
35
36
Lienert, Manual on the Role of the Legislature in the Budget Process (2010), 13.
37
of driving tax reform. However, over the decades the reporting of tax
expenditures has resulted in a third use: as a tool to be used for com-
parative analysis purposes at an international level. his use is not a new
phenomena; however, the number of bodies and parties undertaking the
studies, along with its level of sophistication, has increased over the dec-
ades. he practical work of international organisations is considered in
Chapter 3, but there are also academic reports which have undertaken a
comparative study. Historically, McDaniel and Surrey, in 1985, undertook
the irst analytic and data collection process needed to facilitate a study of
the international aspects of tax expenditures. In doing so, they considered
six industrialised countries and developed lists for these countries using
uniform criteria. he purpose for undertaking such a study was to allow
for international comparative analyses as well as for the purposes of inter-
national economic relationships.38 he lack of international consensus on
the benchmark tax system makes international comparability diicult;
however, commonalities do exist between jurisdictions.39 For example,
OECD countries consistently use what is known as the comprehensive
income tax benchmark as contrasted with the expenditure tax bench-
mark,40 resulting in agreement as to a key element of any tax regime.
Further, over time the work undertaken by the OECD, which we discuss
in Chapter 3, is becoming more sophisticated in its ability to adjust for
diferences. Limitations, however, should not be dismissed, as it is cur-
rently not possible to ensure that the data is fully comparable,41 nor would
we suggest that this is ever likely to be the case in the future.42
McDaniel and Surrey in 1985 recognised that the most important appli-
cations of tax expenditures reporting and analysis are and will remain in
the context of the budget and tax policies of the individual jurisdictions,43
a point with which, over a quarter of a century later, we agree. However, it
38
McDaniel and Surrey, International Aspects of Tax Expenditures: A Comparative Study
(1985), 5.
39
For example, all OECD countries regard a progressive income tax rate structure as part
of the tax base.
40
he diference being that under a comprehensive tax benchmark savings are included
whereas with an expenditure tax benchmark savings are excluded.
41
Heady, ‘Tax Expenditures: Deinitional and Policy Issues’ (2011).
42
he OECD explicitly recognises that there are diferences between various countries’ tax
expenditure methodologies in general and their benchmark tax systems in particular but
argues that where the goal of international studies is better policy, comparative studies
are useful: OECD, Tax Expenditures in OECD Countries (2010), 18.
43
(1985), 4.
The identification of tax expenditures 35
is remiss not to note the production and use of international comparative

studies by organisations such as the OECD with the aim of its most recent
publication, Tax Expenditures in OECD Countries, designed to shed light
on the use of tax expenditures through a study of ten OECD countries.
he stated purpose of the study was to help government oicials and the
public better understand some of the technical and policy issues behind
the use of tax expenditures by highlighting ‘key trends and successful
practices’ and addressing a ‘broad range of government inance issues,
including tax policy making, tax and budget eiciency, iscal responsi-
bility and rule making’.44 When used for these purposes, international
comparative studies have signiicant beneits.
We have argued that the identiication of the purpose and function
of tax expenditure analysis and management is what allows a nation to
determine what elements of its tax regime it considers to be tax expen-
ditures. his is because these decisions form the basis for determining
the benchmark against which tax expenditures are identiied. As such, we
argue that the purpose and function drives the identiication, and at the
heart of tax expenditure analysis lies the identiication of speciic elem-
ents of a tax regime as tax expenditures. Herein lies the greatest of tax
expenditure analysis controversy.
2.5 he identiication of tax expenditures

A tax expenditure can only be identiied as such when it is contrasted with
or compared to another theoretical or practical base. To this extent, while
tax expenditure reporting and analysis has become increasingly popu-
lar, consensus as to what constitutes a tax expenditure, whether debated
by academics or pragmatically required by domestic jurisdictions to
enable reporting, has not been reached. Debate as to what constitutes a
tax expenditure has ensued for more than four decades, with that debate,
whilst undertaking an international focus, being the most ierce in the
United States. As such, in this part we critically analyse what constitutes
a tax expenditure by considering the historical developments of the def-
inition, most notably observing that by and large the debate has been
fuelled not by reporting bodies but by academics generally arguing about
the deinition.45 What is also revealed in this discussion is the need for an
44
45
See McDaniel and Surrey, International Aspects of Tax Expenditures: A Comparative
Study (1985) who argue that there is a great deal of consensus and that it is only at the
periphery that there is debate.
ongoing consideration of what constitutes a tax expenditure. As Surrey

and McDaniel prophetically stated in 1985, tax expenditure analysis needs
to undertake a dynamic and continuing analysis of the provisions in a tax
system.46 Further, any consideration of what constitutes an identiied tax
expenditure from a deinitional perspective needs to be placed in the con-
text of the fact that some of the concepts relating to tax expenditures are in
fact aforded a common consensus and this oten occurs no matter what
the purpose and function individual nations adopt. In essence, what must
not be lost in any discussion on the deinitional issues surrounding tax
expenditures is that there is a deal of consensus on what constitutes a tax
expenditure for reporting purposes regardless of how the benchmark is
deined by both academics and reporting bodies alike and, as such, some
of the arguments are quite correctly classiied as being at the margin. his,
however, does not detract from the signiicance of the debate.
he ensuing debate as to what constitutes a tax expenditure should also
lead the reader to the logical conclusion that it is not possible to suggest
that a universal deinition of tax expenditures is or can be ascertained
with certainty. We certainly do not intend to suggest this is the case.
Quite the contrary: we remind the reader that the overarching thesis of
this book is that a unanimous international, or even domestic, dein-
ition cannot be reached because neither international or domestic moral
objectivity is possible. It is the absence of moral objectivity that makes
the need to deine tax expenditures so as to aford identiication, one of
the most contentious questions surrounding any analysis of tax expen-
ditures. he underlying purpose makes it no less contentious, whether
theoretical and undertaken by academics or the purpose is much more
pragmatic with the deinition used by jurisdictions in their reporting of
tax expenditures.
Whatever the label applied to tax expenditures, dei nitional issues
arise out of the fundamental need to dei ne a benchmark or basic
tax structure from which tax expenditures are considered devia-
tions or exceptions. h is identiication of a country’s normative tax-
ation system is intrinsically contentious and can only be resolved,
albeit contingently, by a community’s adoption of a unique combin-
ation of normative tax principles that relects the community’s prior-
ities. Moreover, the process by which this normative tax framework is
adopted ought to relect the community’s normative tax policy process.
46
Surrey and McDaniel, International Aspects of Tax Expenditures: A Comparative Study
(1985), 196.
he choice of benchmark is signiicant not only in providing the def-

inition to determine what constitutes a tax expenditure but, where tax
expenditure analysis is used as a budgetary tool, it can have signiicant
bearing on the aggregate estimate of revenue foregone. For example,
where negative tax expenditures are included in the aggregate estimate,
the value of the revenue foregone from negative expenditures may be
signiicantly underestimated.
Below, we focus on non-legislative (academic) dei nitions of tax
expenditures as dei ned by reference to the alternative normative, or
benchmark, taxation systems adopted. Inherent in this discussion is
an acceptance that there are many diferent views of the normative or
benchmark tax base but, by and large, there is signiicant consensus
as to what is and is not a tax expenditure no matter how dei ned. In
Chapter 3, we consider legislative dei nitions of tax expenditures which
are generally adopted by jurisdictions for the purposes of tax expend-
iture reporting.
2.5.1 he history of the generally adopted ‘normative tax base’

We have already discussed the fact that the concept of tax expenditures
had its genesis in the works of Stanley Surrey. He irst used the expres-
sion on 15 November 1967 in a speech calling for ‘full accounting’ whilst
he was Assistant Secretary for Tax Policy in the United States Treasury
Department. In that speech, as reproduced in the Annual Report of the
Secretary of the Treasury on the State of the Finances for the Fiscal Year
Ended June 30, 1968, Surrey’s desire to have tax expenditures recognised
as the equivalent to direct spending outlays is obvious. Surrey was con-
cerned with the ‘hidden’ aspect of tax expenditures and the use of the tax
system for purposes other than raising revenue. He described this hidden
aspect of tax expenditures in the following terms:
A tax system presumably concerns itself with raising revenues rather than
spending funds. But a closer analysis of our present tax system would
reveal real substance to the phrase. hrough deliberate departures from
accepted concepts of net income and through various special exemptions,
deductions, and credits, our tax system does operate to afect the private
economy in ways that are usually accomplished by expenditures – in
efect to produce an expenditure system described in tax language.47
47
Surrey, Excerpts from Remarks Before the Money Marketeers on the US Income Tax
System – he Need for a Full Accounting in the Annual Report of the Secretary of the
Treasury on the State of the Finances for the Fiscal Year Ended June 30, 1968 (1968), 322.
Surrey, a supporter of comprehensive income taxation, undertook his ini-

tial work on tax expenditures driven by his desire to have identiied tax
expenditures recognised as equivalent to direct expenditures, with the
aim of consequently converting them into direct spending programmes
or having them repealed. he context in which the tax expenditure con-
cept was introduced by Surrey was one that involved tax reform in light of
the subsequent Tax Reform Act 1969. His theory of tax expenditures was
founded upon a model of government that called for rational and informed
members of government to procure institutional reform in conjunction
with taxation reform. In 1973, Surrey published his treatise Pathways to
Tax Reform: he Concept of Tax Expenditures48 providing further details
of his research into tax expenditures and defending his classiication and
analysis of tax expenditure provisions. he impetus for Surrey’s coining
of the phrase, or what he describes as an ‘illumination’,49 was a pragmatic
one; to improve the tax system by exposing tax expenditures as special tax
provisions which are similar to direct government expenditures such as,
for example, grants, loans and interest subsidies. Surrey, in that seminal
work, laid the foundation for what is generally understood to be the two-
part categorisation of a tax regime and the foundation for the deinition of
the term tax expenditure:
he federal income tax system consists really of two parts: one part com-
prises the structural provisions necessary to implement the income tax
on individual and corporate net income; the second part comprises the
system of tax expenditures under which Government i nancial assist-
ance programs are carried out through special tax provisions rather than
through direct government expenditures. h is second system is grated
on to the structure of the income tax proper; it has no basic relation to
that structure and is not necessary to its operation. Instead, the system of
tax expenditures provides a vast subsidy apparatus that uses the mechan-
ics of the income tax as the method of paying the subsidy. he special
provisions under which this subsidy apparatus functions take a variety of
forms, covering exclusions from income, exemptions, deductions, credits
against tax, preferential rates of tax and deferrals of tax.50
he work by Surrey laid the foundation for the assumption that provisions
of a tax system are neatly and easily divided into two distinct categories:
structural provisions, those provisions which serve the traditional tax
purpose of revenue raising; and tax expenditures, those provisions which
48
Surrey, Pathways to Tax Reform: he Concept of Tax Expenditures (1973).
49
50
are essentially inancial assistance programmes. he former, it is generally

argued, should be evaluated on the basis of traditional tax criteria, such
as equity, eiciency and simplicity, whilst the latter should be assessed
using budgetary criteria. he notion of dividing the tax regime into two
parts is not diicult to grasp. However, the issue then becomes one of how
to distinguish between the two separate parts of the tax system in order
to apply the diferent evaluative criteria, and this is diicult. At the root
of tax expenditure analysis is the basic deinitional question as to which
income rules are special provisions designed to achieve social and eco-
nomic objectives of a government and which income tax rules constitute
the basic structure of an income tax system and are thus integral to the
revenue raising function of income tax.51 As stated, the argument prof-
fered is then one of the criteria upon which the two separate parts of the
tax system are evaluated.
he process of determining the structural provisions, according to
Surrey, is not a process which can be applied to domestic jurisdictions uni-
formly. his is because, as he states, an income tax requires not only those
provisions which shape a normative income tax, but also those provisions
which are structural parts of an income tax. hose parts which shape the
normative income tax, Surrey argued, were those parts which receive
a general consensus and ‘would essentially be treated in the same way
by any group of tax experts building the structure of an income tax and
being governed in that task by all the requirements implicit in such a tax
because it is an income tax’.52 hose provisions which contain the struc-
tural parts of the income tax system, Surrey argued, ‘could, in the view of
such group of tax experts, conceivably be treated diferently from coun-
try to country depending on the views and the policies shaped by other
goals in the particular society, rather than by factors special to an income
tax’.53 According to Surrey, these provisions are not part of the normative
income tax but do become part of the structural elements of an income
tax and are therefore not considered tax expenditures.54 Surrey, along
with his future co-author McDaniel, maintained that the normative tax
base was that of the Schanz–Haig–Simons deinition of income and this
should be used for the purposes of determining whether a tax provision
is a structural provision or a tax expenditure. he Schanz–Haig–Simons
51
52
Surrey, Pathways to Tax Reform: he Concept of Tax Expenditures (1973), 17 (original
emphasis).
53
54
deinition of income is well known and understood as ‘the algebraic sum

of (1) the market value of the rights exercised in consumption and (2) the
change in the value of the store of property rights between the begin-
ning and the end of the period in question’.55 In other words, a person’s
income to be taxed is their annual consumption plus their annual sav-
ings. Although arguably the normative tax base can be better described as
the comprehensive tax base with amendments to take into account struc-
tural provisions.
Surrey regarded the determination of a normative model as requiring a
consideration of a variety of questions going to the fundamental income
tax structure. he questions that he highlighted were:
• What receipts should be included and what expenses allowed to obtain
the proper measure of net income for an income tax – ‘proper’ in the
sense that it is an income tax for which the measure is being sought;
• In what time periods should includable receipts be included and allow-
able expenditures be taken (e.g., cash and accrual accounting, expens-
ing of capitalisation of expenditures, and if the latter, how written
of – the method of depreciation, for example);
• Over what period of time should the measurement itself be made (e.g.,
averaging and net operating loss questions);
• What is the unit whose income is being measured (e.g., is the family to
be taxed as a unit or the members separately taxed);
• How should the income of organisations of individuals be treated (e.g.,
the relationship of corporate income and the corporate tax to the tax
treatment of the shareholders.56
Twelve years ater his initial work on tax expenditures was published,
Surrey, with his co-author Paul McDaniel, published a second book
on the subject simply entitled Tax Expenditures.57 Despite the ongoing
criticism, Surrey and McDaniel’s deinition of tax expenditures in 1985
remained the same as the initial formulation and, as such, they main-
tained their reliance on the same normative tax structure, with little vari-
ation to Surrey’s initial formulation of a tax expenditure. Ignoring the
substantive criticisms made of the concept and the politics of tax expend-
iture management, they restated Surrey’s initial broad deinition of tax
expenditures:
55
Simons, Personal Income Taxation (1938), 50.
56
57
Surrey and McDaniel, Tax Expenditures (1985).
he tax expenditure concept posits that an income tax is composed of
two distinct elements. he irst element consists of structural provisions
necessary to implement a normal income tax, such as the dei nition of
net income, the speciication of accounting rules, the determination of
the entities subject to tax, the determination of the rate schedule and
exemption levels, and the application of the tax to international transac-
tions. hese provisions compose the revenue-raising aspects of the tax.
he second element consists of the special preferences found in every
income tax. hese provisions, oten called tax incentives or tax subsid-
ies, are departures from the normal tax structure and are designed to
favour a particular industry, activity or class of persons. hey take many
forms, such as permanent exclusions from income, deductions, deferrals
of tax liabilities, credits against tax, or special rates. Whatever their form,
these departures from the normative tax structure represent government
spending for favoured activities or groups, efected through the tax sys-
tem rather than through direct grants, loans, or other forms of govern-
ment assistance.58
Surrey and McDaniel continued by stating that ‘tax expenditure analysis

is based on the concept of a normative tax of the type under consider-
ation’,59 expanding taxes for consideration beyond income tax for which
they continued to rely on the Schanz–Haig–Simons normative concept of
net income to distinguish between the two parts of an income tax system.
Any concession to the criticisms they faced was minimal; however, whilst
continuing to advocate for the Schanz–Haig–Simons base, Surrey and
McDaniel acknowledged that it was the mere starting point. hey stated:
Although the S–H–S dei nition is a useful starting point for identifying
the normative provisions in an income tax, it cannot be sued alone to dis-
tinguish normative provisions from tax expenditure provisions. One rea-
son for this is that Simons applied his deinition of income to only a few
items. He did not discuss the appropriate technical treatment of many
receipts (such as government transfer payments, personal damages and
scholarships) or of many expenditures or losses (such as medical expenses,
casualty losses, and charitable contributions). A second reason is that the
S-H-S dei nition does not address all the issues involved in framing a
normal income tax, such as the taxable period to be sued in applying the
dei nition. A third reason, more narrowly applicable to the dei nition of
income, is that the S–H–S dei nition, though theoretically correct, is too
rigid and demanding to be applied comprehensively in a national income
tax. It remains an ‘ideal’, a ‘theoretically pure treatment’.60
58
59
60
In the same year they published Tax Expenditures, McDaniel and Surrey61
also co-edited their study on the International Aspects of Tax Expenditures
which analysed the tax systems of six countries to identify and quantify
the tax expenditures within each of those systems. he purpose of this
study, discussed earlier in the context of the function of classiication,
chose the countries Canada, France, the Netherlands, Sweden, the United
Kingdom and the United States for the purposes of the comparative study.
Whilst this study did little in the way of contributing to the theoretical
underpinnings of the tax expenditure concept, it arguably laid the foun-
dation for future comparative studies and the recognition of the divergent
approaches to the normative tax base as deined by various developed
nations.
Ultimately, Surrey continued to argue that the benchmark for income
tax should be the Schanz–Haig–Simons dei nition of comprehensive
income. His co-author, McDaniel, continued to vehemently defend this
adoption up until his recent death in 2010.62 Further, this is a deinition
which, whilst facing a great deal of criticism, has endured and is gener-
ally adopted at a broad level by jurisdictions which engage in tax expend-
iture reporting and analysis. his deinition has not only received support
via revenue authority adoption, which is discussed in Chapter 3, but as is
noted later in this chapter, has received more recent support from authors
such as Fleming and Peroni. However, this deinition has not been with-
out its detractors.
2.5.2 Forty years of criticism of the ‘normative tax base’

Surrey’s work almost immediately came under criticism on the basis that
his deinition of tax expenditures was framed by reference to a norma-
tive taxation system grounded upon a consensus of tax experts where that
consensus simply does not exist. Essentially it was profered that Surrey’s
model was an elitist one that failed to take into account political and social
philosophy. he initial criticism which directly followed Surrey’s original
work, was maintained during the intervening period of this sole authored
work and endured through to his later work with McDaniel. Criticism
subsequent to the later work continues today. he most notable of critics
61
Surrey unfortunately passed away in 1984.
62
McDaniel, ‘he Staf of the Joint Committee on Taxation Revision of Tax Expenditure
Classiication Methodology: What Is to Be Made of a Change that Makes No Changes?’
(2011).
was Boris Bittker who, in 1969, criticised what he believed to be the sub-
jective element of the tax base so deined. In his view, ‘a systematic com-
pilation of revenue losses requires an agreed starting point, departures
from which can be identiied’.63 Bittker did not consider that Surrey’s
tax expenditures list was based on such an identiied starting point but
rather was of the view that it provided an ad hoc list of tax provisions.
He believed that what was needed was ‘a generally acceptable model, or
set of principles, enabling us to decide with reasonable assurance which
income tax provisions are departures from the model, whose costs are to
be reported as “tax expenditures” ’.64 Bittker went on to state:
the trouble is that, aside from the many ambiguities that become apparent
as soon as one attempts to apply the Haig-Simons dei nition to the pro-
tean stream of economic life, any system of income taxation is an aggre-
gation of decisions about a host of structural issues that the Haig-Simons
deinition does not even purport to settle. As to these, one could lock forty
tax experts in a room for forty days, and get no agreement – except as a
surrender to hunger or boredom – even if they could all recite the com-
plete works of Henry Simons by heart.65
Bittker, by suggesting that objectivity was not achievable, identiied a

key element to understanding the tax expenditure concept – the inher-
ently political nature of the concept encapsulating moral philosophy.
As Weisbach and Nussim explain, Bittker is suggesting that not only are
there inconsistencies in the deinition but even more of an issue is the fact
that implicit policy judgements are being made leading to a great deal of
subjectivity.66 Bittker’s criticism of the normative tax base so deined was
followed by further attacks not only on the works of Surrey but the Tax
Expenditure Budgets that were a product of his work. he strongest criti-
cisms continued in the vein of Bittker’s argument that the normative base
was subjective and ultimately political whilst at the same time, as Kahn
and Lehman suggest, adopting a tone of moral absolutism.67 In fact, Kahn
and Lehman took Bittker’s criticisms one step further arguing that tax
expenditure analysis creates ‘an illusion of value-free scientiic precisions
in a heavily politicized domain’.68 Again, it is the political nature that is
highlighted but not embraced.
63
Bittker, ‘Accounting for Federal Tax Subsidies in the National Budget’ (1969).
64
Bittker, ‘Accounting for Federal Tax Subsidies in the National Budget’ (1969), 247.
65
Bittker, ‘Accounting for Federal Tax Subsidies in the National Budget’ (1969), 260.
66
Weisbach and Nussim, ‘he Integration of Tax and Spending Programs’ (2004), 974.
67
Kahn and Lehman, ‘Tax Expenditure Budgets: A Critical View’ (1992), 1662.
68
Whilst some authors aimed criticism at Surrey’s model, others accepted

that tax expenditure analysis was a worthwhile task and a useful tool but
disputed the methodology involved in identifying provisions of the tax
regime as tax expenditures. To this end, several alternative proposals
have been suggested, although none has gained broad favour either from
academics or tax authorities alike.69
McIntyre, in his 1980 paper entitled ‘A Solution to the Problem of
Deining a Tax Expenditure’,70 proposed an entirely new methodology
for identifying tax expenditures. McIntyre relies on a linguistic approach
to ascertaining which provisions of the tax regime are considered tax
expenditures and postulates that the deinition of a tax expenditure may
be diferent for diferent purposes. His solution to what he sees as the def-
initional issues is to ‘apply tax expenditure analysis to any tax provision
defended on non-tax grounds’.71 hat is, the determination of whether a
tax provision is in fact a tax expenditure requires a consideration of the
policy goals of the provision. As McIntyre states, ‘assertion that the tax
rule under examination promotes a spending goal triggers a tax expend-
iture analysis’.72 In particular, he argues that his model is superior to the
required normative baseline of Surrey’s model as it bypasses the problem
of obtaining consensus on the features of the normal tax structure.73 his
argument is based on the premise that the ‘complex theoretical problem
of developing a general, all-purpose dei nition of a tax expenditure’ is
avoided because the identiication of tax expenditures is only for a limited
purpose.74
McIntyre’s suggested linguistics approach was followed in 1988 by
huronyi’s ‘substitutable tax provisions’ methodology. huronyi was also
of the view that Surrey’s normative income tax base, due to the signiicant
departures for economic income to accommodate political and admin-
istrative concerns, was ‘so inherently subjective that it deprives the tax
expenditure concept of its persuasive force’.75 His suggested alternative
was one which considered whether the tax provision was a tax law pro-
vision ‘whose purposes a non-tax-based federal program can achieve at
69
Apart from the short-lived adoption of an alternative baseline by the United States Joint
Committee of Taxation in 2008, which is discussed in this section.
70
McIntyre, ‘A Solution to the Problem of Dei ning a Tax Expenditure’ (1980–81).
71
McIntyre, ‘A Solution to the Problem of Dei ning a Tax Expenditure’ (1980–81), 101.
72
73
McIntyre, ‘A Solution to the Problem of Deining a Tax Expenditure’ (1980–81), 82.
74
75
huronyi, ‘Tax Expenditures: A Reassessment’ (1988), 1155.
least as efectively’.76 Maintaining that there is no single deinition of tax

expenditures that can serve all purposes, huronyi argued that it is ‘best
not to attempt to draw a bright line between tax and spending programs’
when considering provisions which may not be tax-based subsidies but
rather serve other purposes.77
In the early 1990s, criticism of the tax expenditure concept came from
Kahn and Lehman. heir criticism was not speciically aimed at Surrey’s
work, but rather at the United States tax expenditure budget resulting
from the initial proposals by Surrey. Speciically, Kahn and Lehman’s
criticism of the tax expenditure budget is its claim of distinguishing what
is ‘normal’ with what is ‘deviant’.78 heir argument is in line with the cau-
tion which must be given in relation to the stigma that can attach to a
provision labelled a tax expenditure. In the context of the language, Kahn
and Lehman hold the view that:
it suggests that provisions that it within the implicit baseline of the tax
expenditure budget are somehow pure, safe and good. hey should not be
changed because ‘neutral’ principles have blessed them. Conversely, the
language suggests that provisions that fall outside the implicit baseline of
the tax expenditure budget (tax expenditures) are somehow corrupt, dan-
gerous and evil. hey should be changed as soon as possible to conform
with the ‘neutral’ position.79
he language, they argue, suggests that to lirt with tax expenditures calls
one’s probity into question.80 Kahn and Lehman go further than simply
criticising the development of this stigma and believe that the provisions
of the tax regime should not be immune from political discussion81 and
tax expenditures analysis should not be obviated by ‘one particularized
vision of the “normal” or “ideal” tax base’.82 Again, we see the recognition
of the highly political nature of tax expenditure management entering the
debate.
Much more recently, we have continued to see an attempt to shit away
from a deined normative base towards an approach that continues to
argue for the adoption of a iscal language or descriptive method of iden-
tifying tax expenditures. Shaviro, in 2004, asked the question in relation
76
77
78
79
Kahn and Lehman, ‘Tax Expenditure Budgets: A Critical View’ (1992), 1662–3.
80
Kahn and Lehman, ‘Tax Expenditure Budgets: A Critical View’ (1992), 1662–1163.
81
82
to Surrey’s work that ‘even if one accepts his normative goals, one may
rightly ask to what extent a mere iscal language innovation either can or
should categorically shape political outcomes’.83 He went on to state that
‘perhaps more disappointing, therefore, is the repetitive and misdirected
conversation that the concept has generated in the tax policy community,
too narrowly focused on the defensibility of using a canonical “reference
tax base” in identifying tax expenditures’.84 Shaviro argues that rather
than classifying tax expenditures in terms of ‘spending’ programmes, a
more sound base from which to classify tax expenditures is according to
the Richard Musgrave classiication of any tax system into allocative and
distributional branches. hat is, tax expenditures should be considered as
rules which are part of the allocative branch of the tax system. In stating
this, however, Shaviro believes that:
the allocative or distributional character of a given rule is a matter both of
degree and, in some cases more than others, reasonably contested opin-
ion. Tax expenditure analysis ought to be more lexible and varied in its
groupings than it is in the Surrey tradition, where each rule is canonically
classiied as tax expenditure or not relative to a speciied reference tax
base that itself relects both contestable distributional judgments and a
set of administrative compromises.85
Shaviro, the most prominent of recent authors to contend that tax

expenditure analysis has gone ‘of the rails’, argues that ‘where tax
expenditure analysis went of the rails … was not in its aim of identify-
ing ‘special provisions … but in its means of doing so, through the iden-
tiication of a supposedly canonical, yet in practice under-theorized …
deinition of the “normative income tax base” ’.86 He goes on to suggest
that the ‘prevailing iscal language is too manipulable in some respects
and too rigid in others’.87 Like the earlier criticisms, Shaviro’s attack is on
the use of the normative tax base to identify tax expenditures. His spe-
ciic criticism goes to the claim that the normative tax base is objective
despite its fundamentally subjective nature. He provides that this ‘con-
notation is so pervasive that iscal language has a dual character. It is
83
Shaviro, ‘Rethinking Tax Expenditures and Fiscal Language’ (2004), 188. Shaviro, at
p. 187, likens tax expenditure analysis to a ‘hardy plant with shallow roots that spreads
widely, resisting the occasional efort to extirpate it, while having little if any efect on the
soils in which it sprouts’.
84
85
86
87
both a purportedly objective descriptive tool and a weapon of political

combat.’88 It is not so much the process of identiication that Shaviro sees
as subjective, but rather its subsequent use once provisions are identiied
as tax expenditure. He then sees the process as potentially being mean-
ingless. Shaviro states:
its use as a political weapon, however, is parasitic on its claim to ofer
objective description. For example, if calling a proposed rule a tax is rec-
ognized as merely a matter of convention, rather than relecting some-
thing signiicant about a rule’s substance, then any inference we are being
invited to draw from the label, such as that it is an example of ‘big govern-
ment’, is unlikely to persuade.89
Shaviro’s criticism should not lead the reader to believe that he argues
that tax expenditure analysis has no merit. Rather, he argues that there
is merit in such a process but it is the current classiication which falters
as ‘classiications must seem meaningful in order to associate broader
inferences with them’.90 His contention with the current state of tax
expenditure analysis is that the characteristics which underlie the iscal
language categories, that is the normative tax base, are oten purely for-
mal, rather than meaningful.91 Whilst criticising the current methods
associated with tax expenditure analysis, Shaviro claims its value as an
exercise lies in its ability to address the ‘confusion in public policy debate
that may occur when proponents of placing particular allocative rules
in the tax system exploit the common tendency to dei ne “taxes” and
“spending” entirely formally, and yet to treat the categories as genuinely
meaningful’.92
Shaviro suggests that a meaningful baseline would exist where the
alternative method of identifying tax expenditures using the iscal lan-
guage of allocative and distributional rules is adopted. He argues that
political warfare over that language would still exist, and baselines would
look normative; however, a meaningful baseline would be apparent.93 No
doubt, again, as Shaviro suggests, this would be a conceptually compli-
cated process and as such accepts that the use of the current ‘cruder and
more formal’ means of identifying tax expenditures is unsurprising. Even
if Shaviro’s alternative measure is not accepted, he believes that ‘more
88
89
90
91
92
93
ought to be done in developing the substantive measures … although the
ones used now could be improved as well as used more thoughtfully’’.94
In 2004, the same year that Shaviro suggested his new method of iden-
tifying tax expenditures, Weisbach and Nussim proposed a new theory of
tax expenditure analysis by arguing that the tax expenditure decision is
purely one of institutional design. In doing so, they rejected the leading
two theories underlying tax expenditure analysis (the comprehensive tax
base and functional equivalence), arguing that the focus should not be on
taxation policy but rather should focus on how the government chooses to
compartmentalise its functions.95 Weisbach and Nussim base their argu-
ment on the premise that, assuming the underlying policy remains the
same, there are no efects to putting a programme in or out of the tax
regime. As such, their focus is on the broader question of government
policy as contrasted with the more traditional approach of focusing on
the efects on the tax system in isolation. he real issue, they argue, is one
of how government spending should be organised.96 Central to Weisbach
and Nussim’s thesis is the underlying question of how to determine the
best way to implement a government programme. However, we would
argue that they do not address the preliminary question of whether the
programme should be implemented in the irst place. What Weisbach and
Nussim suggest is that ‘the attempt to identify tax expenditures should be
abandoned, and that all tax provisions should be compared with equiva-
lent expenditure programs in order to decide how best to achieve their
aims’.97 heir view is that there is no such thing as a normative tax base.98
Weisbach and Nussim refer to the 1969 work of Bittker to support
their argument that the broader organisation of a bureaucracy should
not depend on a dei nition of income. Rather, if the question is one of
how to implement a government programme they see the deinition of
income as entirely irrelevant. However, having expressed the view that
the underlying deinitional issues of the tax base are lawed, Weisbach
and Nussim recognise Surrey’s valuable contribution and focus on what
they describe as the substantive argument surrounding tax expenditures
being that of integration.99 Yet, despite their acknowledgement of Surrey’s
contribution, they argue that the direct comparison approach of Surrey is
94
95
96
97
Heady, ‘Tax Expenditures: Deinitional and Policy Issues’ (2011), 2:3.
98
99
fundamentally lawed on the basis that what Surrey did was not compar-
ing identical tax expenditures and direct expenditures but rather com-
paring diferent programmes. heir reasoning is that tax expenditures
and direct expenditures, implemented by diferent institutions with dif-
ferent policies, will have diferent features and as such cannot be treated
as identical for comparative purposes.100
he alternative to a direct comparative analysis, as suggested by
Weisbach and Nussim, is to set broad policy objectives and then decide
which institutional setting provides the best results.101 Essentially, their
thesis difers from Surrey’s design of tax expenditure analysis by arguing
that it is not the expenditure itself which is to be compared, but rather
the programmes which are best designed for the relevant institutional
settings.102 hey state:
When comparing methods of implementing policy, we should not com-
pare identical programs, but instead should compare how a policy is likely
to be implemented in any given institutional structure. he contribution
of this framework is signiicant, but the arguments of the tax expendi-
tures literature are unconvincing precisely because they fail to take full
advantage of the framework. Tax expenditure theory fails to account for
the inherent beneits of integration, and instead focuses on a tax-centric
consideration of complexity. In so doing, tax expenditure theorists blind
themselves to the diferences that matter most between tax and direct
expenditures: the simpliication that the tax system provides on the one
hand, and the tailoring and accuracy that direct spending programs pro-
vide on the other.
he criticisms levelled at tax expenditure analysis generally, and the nor-

mative tax base speciically, by prominent authors such as those discussed
above has recently been the subject of analysis by Fleming and Peroni,
leading advocates for the reinvigoration of tax expenditure analysis.103 In
particular, Fleming and Peroni are critics of the methodologies proposed
by McIntyre, as well as Weisbach and Nussim. Fleming and Peroni ’s lat-
est work on tax expenditure analysis can be placed in the context of the
recent rejection of the use of the Schanz–Haig–Simons deinition of the
income base by the Staf of the Joint Committee on Taxation of the United
States Congress. In 2008, the Joint Committee abandoned the use of the
100
101
102
103
Dimension’ (2008).
traditional income base and devised a system which provided for two
main categories of what it had previously considered tax expenditures:
‘tax subsidies’ and ‘tax-induced structural distortions’. It was argued that
this revised approach removed both the necessity to rely on a normative
tax base and the implicit criticism of the present law as to what was ‘nor-
mal’. his approach received support from some prominent academics
such as Kleinbard who argued that the redeining of tax expenditures by
the Joint Committee on Taxation in a more objective fashion ‘satisied a
critical precondition to bringing tax expenditures more efectively within
budget framework legislation’.104
he irst category of tax expenditures, known as tax subsidies, was
somewhat based on the work done by Seymour Fiekowsky in 1980 where
he argued that the Schanz–Haig–Simons baseline should be abandoned.
he Joint Committee on Taxation deined a ‘ “tax subsidy” as a speciic
tax provision that is deliberately inconsistent with an identiiable general
rule of the present tax law (not a hypothetical “normal” tax), and that
collects less revenue than does the general rule’.105 he second category
of tax expenditures, known as tax-induced structural distortions, were
supposedly introduced to compensate for the inappropriately narrow def-
inition of tax expenditures and were deined as ‘elements of the Internal
Revenue Code (not deviations from any clearly identiiable general tax
rule and thus not Tax Subsidies) that materially afect economic decisions
in a manner that imposes substantial economic eiciency costs’.106
he replacement by the Joint Committee on Taxation of the Schanz–
Haig–Simons baseline in 2008 was followed by, as Fleming and Peroni
describe, the ‘denouement of the “new paradigm” ’107 and a reversal of the
decision to abandon the normative approach. In their 2010 paper, Fleming
and Peroni correctly point out that ‘this event is a manifestation of the
practical and theoretical diiculties involved in the critically important,
longstanding TEA baseline controversy and it merits close analysis to
see what light it sheds on the correct resolution of that controversy’.108
Interestingly though, is that ater more than 40 years of controversy it is
the Surrey model of relying on a normative baseline as a deinition of tax
104
Kleinbard, Tax Expenditure Framework Legislation (2010), 3.
105
United States of America, A Reconsideration of Tax Expenditure Analysis (2008), 9.
106
United States of America, A Reconsideration of Tax Expenditure Analysis (2008), 10.
107
Base?: A Critique of the “New Paradigm” and its Denouement’ (2010), 180.
108
expenditures that the Joint Committee on Taxation returned to despite

those decades of controversy. Fleming and Peroni, whilst applauding the
return by the Joint Committee on Taxation to the Schanz–Haig–Simons
baseline, view the proper baseline as one which includes not only the con-
cept of Schanz–Haig–Simons, but is also grounded in the principles of
ability to pay and neutrality.109 We agree.
he consequence of the debate that has continued for over 40 years is
that, while criticisms of the benchmark as deined by Surrey abound, a
more appropriate model has not been devised. here is no doubting that
Surrey’s model is one which can be labelled ‘elitist’; however, when the
political philosophy of a jurisdiction is taken into account, it can also be
modiied to a baseline which takes account of a nation’s own policies.
2.5.3 Identifying the benchmark

It is clearly established that the deinition of tax expenditures is contro-
versial and that controversy is unlikely to be resolved. Despite this, we
propose that history has demonstrated that tax expenditure reporting
and analysis requires a benchmark or normative tax system so that tax
expenditures are deined as deviations from that base however deined. In
addition to deining the theoretical benchmark, identifying tax expendi-
tures also requires a consideration of the attributes speciic to a particular
jurisdiction which are considered to be part of the base itself, or what are
generally regarded as structural elements of the benchmark. his is oten
simply a matter for domestic jurisdictions and generally not controversial
on the international stage. Below, we discuss the more pragmatic elem-
ents to the identiication of the normative tax base, with the incorporated
structural elements.
As part of the process of tax expenditure management, where the
benchmark, so deined, uses the Schanz–Haig–Simons model of income
it will generally correspond to a position representing the standard tax
treatment that applies to similar taxpayers or similar types of activities.
In pragmatic terms, jurisdictions will consider the benchmark as repre-
senting this standard treatment, such as the comprehensive tax bench-
mark, with structural elements then incorporated to take into account
the diiculties in simply adopting the standard treatment as the bench-
mark. he structural elements take into account integral design features
109
of individual tax regimes and may vary signiicantly across both time
and jurisdictions. Consequently, it is only possible to consider the ques-
tions which a jurisdiction, in its consideration of what constitutes the
benchmark for its own purposes, may ask in order to ascertain those
structural elements. Each nation must answer these questions in view
of the inluences of its own political philosophy. Despite the criticisms
levelled at Surrey and McDaniel, in 1985 they clearly recognised that the
Schanz–Haig–Simons dei nition of income was only the starting point
for dei ning the benchmark and suggested that the identiication of the
normative tax structure could be ascertained by determining whether
a tax provision is responsive to one of a series of questions.110 Whilst
the Surrey and McDaniel list may arguably be too narrow to take into
account all of the factors needed to ascertain a jurisdiction’s bench-
mark, it does provide a sound platform from which to commence as
this allows many of the various unique elements of any tax regime to be
taken into account. he outcome being that if a particular provision was
not responsive, it was not part of the tax base. heir suggested questions
are:
(1) Is the provision necessary to determine the base of the tax, norma-
tively dei ned, in accordance with the fundamental nature of the
tax?
(2) Is the provision part of the generally applicable rate structure?
(3) Is the provision necessary to deine the taxable units liable for the
tax?
(4) Is the provision necessary to assure that the tax is determined within
the time period selected for imposition of the tax?
(5) Is the provision necessary to implement the tax in international
transactions?
(6) Is the provision necessary to administer tax?111
A consideration of these questions reveals that such elements as the selec-
tion of the tax base, the taxpayer/entity to which the tax base applies, the
time for the imposition of the tax liability, jurisdictional rules deining a
jurisdiction’s claim to tax, the tax rate structure along with any minimum
tax free threshold, and tax administration costs are all structural elem-
ents of the benchmark.
110
(1985), 9.
111
(1985), 9.
he irst step in identifying the benchmark is generally to determine the

tax base, that is, the activities or transactions subject to the tax. Countries
are likely to have many other taxes besides an income tax and it is neces-
sary to determine what is considered a tax for tax expenditure manage-
ment purposes as well as determining the base of those taxes. Diferent
taxes will have diferent tax bases. he discussion above focused on the
benchmark for income tax which will generally describe the benchmark
for taxes on such categories as personal income, business income, capital
gains, retirement savings and fringe beneits. On the other hand, a con-
sumption tax benchmark will not rely on the Schanz–Haig–Simons model
but rather will likely consider the standard tax arrangements in a par-
ticular jurisdiction for any direct or indirect consumption tax on goods
and services, whether that be an ad valorem (value) or volumetric (quan-
tity) approach. Some jurisdictions may also have an externalities taxation
benchmark to take into account the normal taxation arrangements on the
external costs of particular activities such as taxes on carbon emissions.
From a tax expenditure management perspective, the identiication of the
tax base and how a country deines that base plays a signiicant role in
what is considered part of the tax expenditure reporting process.
Once the tax base is determined, it is necessary to select the entity,
or tax unit, which is liable to pay the tax. he personal income tax sys-
tem may have as its benchmark unit the individual, joint or family unit.
Diferent jurisdictions adopt a diferent unit of taxation for personal
income with a majority of nations ofering some form of joint taxation
resulting in income averaging. Whether joint taxation is of itself regarded
as a tax expenditure is open to debate. It may be argued that the bench-
mark should be the individual tax position with joint taxation, which
provides for preferential treatment, a deviation from that benchmark.112
Others may argue that joint taxation merely captures ability to pay.113 he
benchmark unit for corporations can either be a single company or the
head entity of a consolidated group. Again, similar arguments may be
profered as to whether the ability to be assessed as a consolidated group,
like a joint return, is a deviation from the benchmark or simply founded
on ‘ability to pay’ principles. he interaction of the personal income tax
system with the corporate income tax system with the diferences in tax
rates also leads to questions regarding the benchmark. Countries, such as
Australia, which operate a dividend imputation system are likely to regard
112
113
corporate tax as equating to a withholding tax with ultimate liability fall-

ing to the individual. As such, a lower corporate tax rate is simply part
of the benchmark. Alternatively, a jurisdiction which adopts the classical
approach of taxing corporations may regard a lower corporate tax rate as
providing preferential treatment to a particular category of taxpayers and
therefore the lower rate as a tax expenditure.114
he tax period, or time frame for tax liability, must also be established
as part of the integral design feature of the benchmark. he tax year is
considered a structural element of the benchmark, along with the conse-
quences of needing a tax period such as the timing of the recognition of
income and deductions. Generally, an assessment will be applied to nom-
inal rather than real income because expenses are deductible at historical
cost, and where beneits are realised across more than one tax period they
will potentially be pro-rated. Further, income may be recognised either
on a cash or accruals basis depending on the taxpayer. Other structural
provisions, which are likely to be considered integral design features, are
depreciation deductions spread over the life of an asset, and exemptions
for some personal gains. Again, each of these integral design features will
vary across individual nations.
he design features discussed above are focused purely on domestic
taxation principles. In addition, there are jurisdictional rules and inter-
national tax arrangements that may also be considered part of the bench-
mark design. In particular, provisions which prevent double taxation such
as the exemption or credit rules contained in domestic legislation, as well
as any treaty obligations under international tax agreements, will gen-
erally be considered structural elements of a benchmark. International
tax arrangements which are potentially incorporated into the benchmark
include controlled foreign company rules, transfer pricing rules and thin
capitalisation rules, along with variances in interest, dividend and roy-
alty withholding rates and any other variations from domestic legislation
due to treaty obligations. Finally, the tax rate, or rate of tax which applies
to the base, will be part of the benchmark. It is generally accepted that a
progressive rate regime, acknowledged as an integral and long-standing
feature of tax regimes around the world, will be part of the income tax
benchmark.
his discussion has focused on an income tax baseline, and generally,
debate has centred on this particular benchmark. However, it would be
remiss to fail to briely mention alternative baselines which have been
114
The measurement of tax expenditures 55
proposed. he most common alternative is the consumption tax baseline

with its advantage cited as being the removal of any bias between saving
and current consumption.115 Unlike the Schanz–Haig–Simons deinition
of income as the baseline, the consumption tax baseline would exempt
savings and investments.116 Such a baseline would reduce the number
of identiied tax expenditures for the simple reason that provisions that
resulted in reductions in the taxation of savings or investments would no
longer be considered tax expenditures.
2.6 he measurement of tax expenditures

he reporting of tax expenditures traditionally involves not only the list-
ing of those provisions considered deviations from the benchmark but
also their iscal measurement. By briely retracing our steps, it becomes
apparent that the signiicance of this measurement is dependent on the
purpose for which a nation categorises provisions of its tax regime as tax
expenditures. hat is, whether a nation is reporting tax expenditures for
the purposes of promoting eiciency within the tax regime and budgetary
processes or for the purposes of promoting political engagement and tax
reform. he identiication of tax expenditures rather than their measure-
ment has received the most criticism within the USA. he reason for this
is simple. he original purpose of tax expenditure analysis in the United
States was tax reform. And, tax reform depends on tax expenditures being
identiied rather than quantiied. hat is, the identiication may lead to the
conclusion that certain tax expenditures are inequitable and unjustiiable
because of the distortion in their distributional efects (the upside-down
efect) and, therefore, it is not necessary to get to the point where their
iscal value is relevant for the purpose of analysing them as spending pro-
grammes. his can be contrasted with the German approach to tax expen-
ditures where they are viewed as a legitimate means of implementing the
government’s economic and social policy.117 Consequently, in Germany
deinitional issues have received much less attention and the iscal value
115
Dimension’ (2008), from p. 508, for a discussion on consumption tax as the norm along
with a review of the literature supporting this baseline.
116
For a discussion on the implications of evaluating tax expenditures under a consump-
tion tax baseline, along with the conceptual diferences, see Carroll et al., ‘Income Versus
Consumption Tax Baselines for Tax Expenditures’ (2011).
117
becomes paramount to allow for an overview of the cost of this indirect

government spending.
We argue that whether tax expenditure analysis is undertaken for the
purpose of tax reform or as part of the budgetary process, once a pro-
vision of the tax regime is identiied as a tax expenditure, it is generally
accepted that a calculation of the iscal cost of that expenditure is required
to determine whether it is justiied. In practice, for the purposes of report-
ing tax expenditures, it is necessary to determine a basis for measuring
the iscal cost of each expenditure. Whilst not all tax expenditures can be
measured in iscal terms, where it is possible to do so, jurisdictions tend
to attempt such an exercise. Although, as indicated above, we argue that
measurement only becomes signiicant when tax expenditure analysis is
used for budgetary purposes. When used for tax reform purposes, mere
identiication is the crucial issue. he measurement of tax expenditures,
like other elements of tax expenditure analysis and reporting, has also
faced its share of critics. his criticism is directed to both the usefulness
of the measurement of tax expenditures in iscal terms and, where it is
accepted that there may be some usefulness, the methodology for deter-
mining the measurement.
At the most fundamental level, there are those who argue that the meas-
urement of tax expenditures is not a worthwhile exercise with detractors
believing iscal measurement is not a useful analytical tool. For example,
critics of the deinition of tax expenditures and its reliance on a normative
base, Weisbach and Nussim, also believe that the traditional tax expen-
ditures rationale, with what they describe as the ‘only one correct (and
strongly evaluative) method’ of presenting the information should be
replaced with a ‘variety of nonevaluative ways under an “information use-
fulness” rationale’, although they do not deny the normative and political
components of such an approach.118 Essentially, the argument becomes
one of what information is useful.119 If, however, it is accepted that there is
some analytical value to determining the estimated cost of individual tax
expenditures, the diiculty as to methodology arises.
Measurement of individual tax expenditures is oten diicult because,
whilst there is a theoretical justiication for arguing that tax expenditures
are the equivalent of direct expenditures, the iscal costs cannot be dir-
ectly transposed. he reason being is that stakeholders will modify their
behaviour to not only take into account the individual tax expenditure
118
Weisbach and Nussim, ‘he Integration of Tax and Spending Programs’ (2004), 976–977.
119
The measurement of tax expenditures 57
but also other related factors. For example, if we assume that a govern-
ment decides to abolish all tax expenditures for retirement saving by
individual taxpayers those taxpayers have several choices to make in
relation to the savings which have previously been concessionally taxed.
Taxpayers may choose instead to put some of their savings into another
concessionally taxed or exempt savings vehicle, such as tax deferred sav-
ings or, possibly, put more money into buying a family home, which is
frequently exempt from tax. As a result, the savings in tax revenue would
be lower than expected which is a major law in the quantitative aspect
of tax expenditure analysis. In fact, Fleming and Peroni, supporters of
the traditional benchmark, believe that the quantitative aspect of tax
expenditure analysis is the weakest part of the reporting process due to
its static nature.120 Where tax expenditures are considered the equivalent
of direct expenditures, the implication of any quantitative analysis of the
cost is that by repealing the tax expenditure a revenue gain of the equiva-
lent amount will occur. However, as indicated, this is an overly simplistic
interpretation of any measurements as it fails to take into account second
order efects. Fleming and Peroni, pointing out that it is not an accurate
measure of the gains to be made from repealing a provision and, as such,
is not an accurate measure of the cost of the tax expenditure, note that the
iscal estimate is far from precise,121 a point that few would dispute.
Where tax expenditure analysis is used as a tool for tax reform the is-
cal accuracy of a measured tax expenditure is warranted but may not
ultimately be signiicant in determining whether a provision is justiied
and equitable. Further, the characteristics of some tax expenditures argu-
ably immediately deem them unacceptable. A case in point is a deduction
classiied as a tax expenditure and available to all individual taxpayers
who are subject to progressive tax rates, as it is inherently biased towards
high income earners compared to low income earners. For example, a
taxpayer with a marginal tax rate of 20 per cent and a deduction of $100
receives a beneit of $20, while a taxpayer with a marginal rate of 40 per
cent and the same deduction of $100 receives a beneit of $40. Further, a
taxpayer with income below the tax free threshold will receive no beneit
from their $100 deduction. As previously discussed, this is known as the
‘upside-down efect’ or ‘inverted distributional efect’ of many personal
120
121
tax expenditures and no matter what a nation considers to be its good tax
policy imperatives, this result will be contrary to the most basic of the
principles making up the ability to pay concept.
Where tax expenditure analysis is used as part of the budgetary pro-
cess, and governments are taking into account the cost of tax expendi-
tures, the accuracy of the iscal cost of those expenditures becomes much
more signiicant. To this extent, because of the diiculties of measuring
the cost of tax expenditures accurately, tax expenditure reports state that
they provide estimates rather than precise budget allocations. As a result,
they cannot be directly compared with direct budget spending. hat is,
tax expenditures cannot be ‘summed’ or added up in the same way as
direct expenditures and concerns about tax expenditure estimates are
compounded when these estimates are added together to produce a total
estimate for government tax expenditures. A total tax expenditure esti-
mate may usefully indicate, very roughly, what proportion of tax expen-
ditures are comprised of government revenues or direct expenditures.
However, it must be remembered that this is only a rough estimate and
used cautiously in debates about public policy. Users of a tax expenditure
report should also be aware that the abolition of a tax expenditure will
not necessarily generate additional revenue to the extent implied by the
tax expenditure report. his is likely because, as previously indicated, the
report might not consider behavioural responses to the abolition of the
tax expenditure.
Traditionally there are three methods of calculating the iscal costs of
tax expenditures, whether calculated on a cash or accruals basis: the rev-
enue foregone approach, the revenue gain approach and the outlay equiva-
lence approach. he predominant method of calculating tax expenditures
is the ‘revenue foregone approach’ which calculates the tax which would
have been payable without the concession, and assumes that the economic
behaviour of the afected taxpayers does not change. However, as identi-
ied in numerous reports, there are several alternative approaches which
also measure tax expenditures with the ‘revenue gain approach’ and
‘outlay equivalence approach’ oten ofered as those alternative means.
he diferent approaches relect the diferent underlying assumptions
made about taxpayer behaviour and responses to potential amendment.
Pragmatic concerns relate to the measuring of tax expenditures as there is
the need for adequate data in tax returns, overall tax revenues and taxpay-
ers. hese methods are discussed in Chapter 3 where it is demonstrated
that these two further methods, whilst potentially more accurate because
of the factoring in of second line efects, require higher quality data.
Expansions of the concept of tax expenditures 59
2.7 Contemporary expansions of the traditional concept

of tax expenditures
By now, it should be obvious to the reader that tax expenditures take a
number of diferent forms and target a number of diferent taxpayers and
functions. Traditionally, tax expenditures are considered to fall within
one of ive categories: allowances, exemptions, rate relief, tax deferral and
credits.122 Each tax expenditure under this traditional categorisation is a
departure from the normative tax base which is substitutable for a direct
spending measure. he historical development of tax expenditure ana-
lysis along this line is understandable given Surrey’s object of having tax
expenditures identiied as the equivalent of direct outlay programmes.
His use of both the terminology and the categorisation of provisions of
the tax regime as ‘tax expenditures’ was designed to achieve this. Surrey’s
genius of the concept of a ‘tax expenditure’ is that it constitutes a rhet-
orical device intended to expose the framing efect arising from a reluc-
tance or failure to recognise that not gathering revenue from a particular
taxpayer is substantively the same as gathering revenue from the taxpayer
and then handing the same sum back to that taxpayer.123 He also believed
that this process would achieve a reduction in the number of tax expendi-
tures as they would fail to survive the same level of scrutiny faced by direct
spending programmes. However, Surrey was canny enough to know that
this rhetorical device was not enough in itself to produce the tax spend-
ing retrenchment that he desired. He knew the art of reformist rhetoric
enough to know that the success of his tax expenditure project depended
upon it being couched in such rhetorically watertight terms as to be a fait
accompli. he i xation upon expressly legislated departures from the nor-
mative tax framework contributed to the rhetorical force of his project.
It has already been seen that Surrey’s deinition of tax expenditures has
come under signiicant critical review, speciically in relation to two of its
elements, the normative tax base framed upon the consensus of experts
and the direct spending substitutability. However, time has demonstrated
that the deinition of tax expenditures provided by Surrey, along with
most subsequent deinitions, has also sufered from a third limiting fac-
tor; the exclusion of all but expressly legislated tax measures. As will be
122
123
For a discussion of framing efects in the context of tax expenditures, see Zelinsky,
‘Do Tax Expenditures Create Framing Efects? Volunteer Fireighters, Property Tax
Exemptions, and the Paradox of Tax Expenditure Analysis’ (2005).
evidenced by the discussion in Chapter 3, this third limitation, the con-
centration upon expressly legislated tax expenditures, has been adopted
without critical scrutiny in countries that have adopted the United States’
‘tax expenditure’ concept and also countries adopting the ‘tax subsidy’
nomenclature irst adopted in the Republic of Germany.124 Yet, it is clear
that Surrey’s rhetorical eforts have not borne the fruit of tax expenditure
retrenchment that he desired. In large part this is because Surrey’s con-
cept of a tax expenditure was built upon a false dichotomy – the distinc-
tion between policy neutral tax provisions ‘that are just tax provisions’125
and ‘special provisions representing Government expenditures made
through the income tax system to achieve various social and economic
objectives’.126 he subsequent literature regarding the tax expenditure
concept has demonstrated the falsity of this dichotomy by showing that
all tax rules embody policy tradeofs and therefore are policy-laden.127
Further, the literature suggests that the categorisation of rules into ‘tax
rules’, ‘regulation’ and ‘spending’ rules is problematic but nevertheless
consistent with constitutional and other institutional frameworks. With
the beneit of hindsight, it is not surprising that the ‘objective’ consensus
amongst tax experts regarding the normative framework has not materi-
alised. Further, it is not surprising to ind that the substitutability require-
ment has prompted a burgeoning literature demonstrating the absence
of platonic public policy forms such that the taxing/spending dichotomy
collapses.128 As such, we argue that, no matter what a nation’s adopted
purpose of the tax expenditure concept, this third limitation should be
abandoned.
Whether a tax expenditure is deined as a deviation from a deined
benchmark, considered to be convertible to a direct expenditure, or a pro-
vision of the tax legislation which has as its objective the pursuit of social
or economic policy, each of the deinitions seeks to posit tax expenditures
as a deliberate intent by the state. As explained, this is understandable
given tax expenditure analysis has been historically designed to result in
124
h is may be speciically stated as demonstrated by the deinitions previously extracted
in this book. Alternatively, work dealing with the tax expenditure concept impli-
citly may accept that tax expenditures can only arise by express legislative rule. For
example, see Zelinsky, ‘James Madison and Public Choice at Gucci Gulch: A Procedural
Defense of Tax Expenditures and Tax Institutions’ (1993); Laity, ‘he Corporation as
Administrative Agency: Tax Expenditures and Institutional Design’ (2008).
125
126
127
Eisenstein, he Ideologies of Taxation (1961).
128
Shaviro, ‘Rethinking Tax Expenditures and Fiscal Language’ (2004).
Expansions of the concept of tax expenditures 61
a pragmatic outcome of either budget control or tax reform. Countries
by necessity focus on their tax regime and seek to divide the parts of that
regime into those provisions which form part of the benchmark and those
which are considered deviations. But, the practical consequence is that
tax expenditures as deined by both academics and reporting bodies alike
tend not to include what may be labelled as implicit or operational tax
expenditures. By this, we mean those parts of the broader tax regime
which confer on certain taxpayers a ‘tax break’ which is not otherwise
aforded to all taxpayers. Yet, if a ‘tax break’ is conferred on a group of
taxpayers at the exclusion of others, whether or not deliberate, we argue
that they should also be considered as part of a broader deinition of tax
expenditures. In other words, arguably where taxpayers are aforded a tax
beneit via a means outside speciic legislative intent, those beneits should
be regarded as tax expenditures. Implicit or operational tax expenditures
fall within the scope of tax expenditures as deined simply because rev-
enue is not collected in circumstances where the benchmark indicates that
tax ought to be imposed. Essentially the deinition of tax expenditures is
then understood to include not only those which are posited but those
which are implied, generally through some form of omission, whether
administrative or judicial.
here are several signiicant non-legislative or implicit tax expendi-
tures of note, with the types of tax expenditures that we propose con-
tained within the extended deinition falling into two broad categories.
hese tax expenditures will either be the result of a deliberate admin-
istrative or judicial decision, or through an omission to undertake the
requisite administrative tasks.129 Express administrative rulings or
behaviour, for example, the exercise of administrative discretion by the
revenue collecting authority, such as the discretion to settle a tax dispute,
or the granting of dispensation from tax compliance obligations or tax
payment obligations, may lead to signiicant tax expenditures. hese are
all overt administrative actions that do not necessarily entail the adop-
tion of an administrative rule. By contrast, the interpretation of the law
by the same authorities in the granting of an administrative interpret-
ation that departs from the benchmark does entail the adoption of an
express legislative rule. It is also possible for an express legislative rule
129
For a discussion of the consequences of failing to recognise these implicit tax expen-
ditures see Krever, ‘Analysing Implicit Tax Expenditures’ (2011). Krever argues that
tax expenditure analysis will not yield better outcomes until it is extended to recognise
implicit tax expenditures as part of the tax expenditure analysis process. See also Krever,
‘Taming Complexity in Australian Income Tax’ (2003).
to be judicially interpreted in a manner which is inconsistent with the
benchmark.
In addition to express administrative and judicial rule or overt action,
there is implied legislative tax expenditure. hat is, tax expenditures
which arise by legislative omission. In other words, the legislature has
not enacted a rule that the normative framework indicated ought to be
enacted. hus, for example, there is no diference between a govern-
ment speciically exempting capital gains from income tax and a gov-
ernment maintaining an income tax that does not include any provision
regarding the taxation of capital gains.130 Assuming that the normative
tax system would impose income tax upon capital gains, in both cases
a tax expenditure should be reported. A consistent failure to enforce
an existing legislative provision or a departure from a stated tax norm
without express rule sanctioning that departure would also be impli-
cit tax expenditures. h is form of administrative tax expenditure can
be contrasted with an express administrative tax expenditure, which
arises where the law can potentially be applied in accordance with the
normative taxation system but the administrator exercises a discretion
not to apply that law. he most obvious example of an administrative
tax expenditure arising from administrative failure is the evasion of tax
through the cash economy. Estimates of the ‘tax gap’ are made in some
countries, while others believe that the estimates are too uncertain to
be credible.
Given the controversy surrounding the tax expenditure concept in its
more traditional form, it is easy to suggest that arguing for an expanded
deinition of tax expenditures is likely to contribute to, and support, the
critics’ call for a derailment of the concept itself due to the diiculties sur-
rounding the conservative and traditional deinition. Supporters of the
tax expenditure concept may argue that rigorous tax expenditure analysis
is undertaken for Surrey’s pragmatic reason of recognising tax expendi-
tures as being the equivalent of direct spending programmes and, as such,
the identiication should be limited to express expenditures. Similarly, an
eiciency argument may be mounted against the suggested expansion on
the grounds of increased costs to taxpayers in the form of compliance as
well as government authorities (both legislative and administrative) in
the form of identiication.
130
It may be argued that in countries where there is no or a limited capital gains tax regime,
the failure to legislate to capture capital gains is a departure from the normative tax
base.
Conclusion 63
However, arguing for an expanded deinition does not detract from the
concept, but rather enhances its legitimacy if we explicitly acknowledge
that the nature and purpose of tax expenditure management is governed
by one’s moral and political philosophy. his argument for expansion is
no diferent to arguments surrounding the traditional deinition of tax
expenditures and the justiications for tax expenditure management
remain the same, with arguably greater integrity to the process. In Chapter
5 we argue that the iscal and moral signiicance of tax expenditure man-
agement means that how a society deals with the irresolvable controversies
regarding tax expenditures provides instructive insights into the political
constitution of that society more generally. Such insights may then pro-
vide sound reasons for expanding the deinition. Each community should
be given the choice to take into account the tax expenditures which fall
within the scope of the expanded deinition of tax expenditures to allow
for critical scrutiny. Very little express support for the inclusion of impli-
cit tax expenditures can be found. However, recently in the United States
we have seen the recognition of implicit tax expenditures at a policy level,
with the Center on Budget and Policy calling for the inclusion of both
explicit and implicit tax expenditures in state tax expenditure reports.131
While implicit tax expenditures are deined narrowly in this document to
be ‘what is let out of the code, by a reference in the code, or by the code’s
departure from standard or historical practices’,132 it is signiicant that
they are recognised at all.
2.8 Conclusion
A fundamental problem associated with the categorisation of a provision
of a tax regime as a tax expenditure is the negative connotations which
accompany that classiication. Tax expenditure analysis over the years
has tended to stigmatise anything labelled as a tax expenditure as a pro-
vision worthy of suspicion and ultimately repeal. Weisbach and Nussim
succinctly and accurately describe this as the normative consequences of
the label133 and it is unlikely that many would argue such a stigma does
not exist. However, this should not be accepted as an absolute and doing
so defeats the purpose of tax expenditure analysis as it is then necessarily
131
Leachman et al., Promoting State Budget Accountability hrough Tax Expenditure
Reporting (2011), 14.
132
133
assumed that tax expenditures are ‘bad’. Ostensibly, many tax expendi-
tures do warrant an investigation as to their worth. However, the pur-
poses of tax expenditure analysis is not to simply label tax expenditures
as ‘bad’ and all other tax provisions as ‘good’. Rather, we suggest that this
stigma is a result of the fundamental necessity for the categorisation of
tax expenditures discussed here as well as a deinition of tax expendi-
tures. Such a view is consistent with mainstream tax expenditure analysis
and, to this end, Surrey himself stressed that the ‘classiication of an item
as a tax expenditure does not in itself make that item either a desirable
or an undesirable provision; nor does it indicate whether the inclusion
of the item in the tax system is good or bad iscal policy’.134 As such, an
open mind must be kept when considering the categorisation of provi-
sions of a tax regime as tax expenditures, as well as the purposes of tax
expenditure identiication and analysis. In essence, classifying provisions
of a tax regime as tax expenditures cannot and should not be viewed stat-
ically. McDaniel and Surrey were correct in 1985 when they stated that
the ‘classiication of provisions as tax expenditures or as parts of the nor-
mal tax structure is a dynamic and continuing process’.135 he concept
itself does not imply anything except a particular way to think about or
analyse a provision which is classiied as a tax expenditure. To this end,
the tax expenditure concept is exactly that, a ‘concept’ which allows us
to classify and categorise certain provisions of a tax system. As Brooks
explains, ‘just because there is some dispute about the precise dividing
line between tax expenditures and other tax provisions, and some provi-
sions that are sometimes labelled tax expenditures might also arguably
be justiied in terms of traditional tax criteria, does not mean the con-
cept is incoherent’.136 Tax expenditures remain a major part of tax systems
around the world and continue to grow in number. As such, the initial
aim of tax expenditure analysis is a failure. Yet, it is because the number of
tax expenditures is continuing to grow that tax expenditure analysis is so
important. To this end, we ask the reader also to keep an open mind as to
the tax expenditure concept.
134
135
(1985), 8.
136
Brooks, ‘he Under-Appreciated Implications of the Tax Expenditure Concept’ (2009),
234.
3
Reporting on tax expenditures
3.1 Introduction
Interest in and controversy surrounding tax expenditures has not only
been academic, as outlined in Chapter 2, but has also emerged as a major
democratic issue in mainstream government reporting, with an increas-
ing number of jurisdictions around the world providing detailed annual
tax expenditures statements. In addition, there has been a growing num-
ber of oversight agencies, such as National Audit Oices in various juris-
dictions, undertaking investigations into the completeness and reliability
of these annual statements and generally recommending greater transpar-
ency in the reporting of tax expenditures.1 Adding to this source of litera-
ture on the adoption and processes involved in tax expenditure reporting
are the international bodies such as the Organisation for Economic
Co-operation and Development (OECD), World Bank and International
Monetary Fund (IMF) which have contributed to ‘international best prac-
tice’, international comparative studies and the analysis of the process of
tax expenditure reporting itself.
As with the theoretical controversies, this history of tax expenditure
reporting dates back to its conceptual inception. We saw in Chapter 2
that the concept of tax expenditures was introduced by Stanley Surrey,
and one of the i rst tax expenditures reports, produced in 1968, was
a consequence of this conception when Surrey as the then Assistant
Secretary of the United States Treasury (tax policy) instructed his
staf to compile a list of ‘government spending for favoured activities
or groups, efected through the tax system rather than through direct
grants, loans, or other forms of government assistance’,2 or what we now
1
See for example, Australia, Preparation of the Tax Expenditures Statement (2008) and the
2007 Report by the National Audit Oice of Denmark as cited by the Nordic Working
Group, Tax Expenditures in the Nordic Countries (2010).
2
(1985).
65
66 Reporting on tax expenditures
know as tax expenditures. Although, across the Atlantic, Germany pub-

lished a report on tax subsidies and preferences one year earlier and, as
such, can lay claim to be the very irst nation to report on a tax expend-
iture budget.3 While the United States and Germany led the way in the
late 1960s, momentum in the 1970s saw the members of various inter-
national organisations such as the International Fiscal Association (IFA)
meet to examine both the concept of tax expenditures per se as well
as the reporting of tax expenditures, resulting in nations like Austria
(1978), Canada (1979), France (1979), Spain (1979) and Australia (1982)
all introducing tax expenditure reporting in some form. By the 1980s,
most OECD countries and a few developing countries were adopting the
practice of reporting on tax expenditures and this number has contin-
ued to grow over the last 30 years. While it is diicult to assess the exact
number of nations reporting on tax expenditures because of the difer-
ent ways and means of doing so, we estimate that today there are over 40
nations which do so.4
In Chapter 2, we discussed the theoretical constructs of the tax expen-
ditures concept along with the academic literature which demonstrates
the inevitable and irresolvable controversies surrounding the concept
itself as well as its various components. We concluded that these con-
troversies do not warrant the abandonment of tax expenditure analysis
and, in actuality, detractors to the concept have achieved very little in
the way of disrupting the practice of doing so. Legitimacy of the process,
evidenced by the increasing number of nations reporting on tax expen-
ditures as well as the increased sophistication of that reporting, indicates
that these controversies have neither derailed nor stemmed the report-
ing of tax expenditures by nations around the world. his is augmented
by calls for greater transparency of iscal policy from key drivers within
the international community in general, and speciic interested individ-
uals and groups in particular. here is a long list of nations which have
adopted a formal system of reporting on tax expenditures and each has
dealt with the theoretical controversies in practice. Countries such as
Australia, Canada, the United States, the United Kingdom, France, Korea,
Germany, Japan, Finland, Norway, Sweden, Denmark, Spain, India and
3
1er Subventionsbericht.
4
h is list includes Argentina, Australia, Brazil, Canada, Chile, Colombia, Denmark, El
Salvador, Finland, France, Germany, Ghana, Guatemala, India, Indonesia, Italy, Japan,
Kazakhstan, Kenya, Malaysia, Mexico, New Zealand, Norway, Papua New Guinea, Peru,
Philippines, Portugal, Russia, Serbia, Slovakia, South Africa, Spain, Sweden, Trinidad &
Tobago, Turkey, Uganda, United Kingdom, United States, Zambia.
Introduction 67
South Africa, to name a few, all report on tax expenditures in difering

forms and level of detail. Each has dealt with the theoretical controversies
relating to the purpose, identiication, measurement and management of
tax expenditures. In this chapter, we consider how various nations around
the world have dealt with these controversies and examine the pragmatic
approaches based on the diferent responses to the tension between elit-
ism and democracy, demonstrating, as we claimed in Chapter 1, that the
concept of tax expenditures and processes by which tax expenditures
might be managed are deeply controversial in the sense that there are no
‘right answers’. However, the pragmatic responses, demonstrating that
tax expenditures are a signiicant part of modern government because
they are an important part of government budgeting, and a government’s
budget is its core policy document, highlight the fact that tax expenditure
management ought to be grounded upon particular moral stances as to
the nature of democracy and the nature of the society that we think a
democratic community ought to aspire to.
In this chapter, we survey the current international approach of
attempting a ‘best practice’ for tax expenditure reporting by investigating
the diferent solutions that nations can and/or have adopted to overcome
the theoretical controversies. We argue that tax expenditure controver-
sies extend beyond conceptual matters to practical matters, such as those
highlighted in this chapter. We do not propose that it is possible to over-
come the contingent nature of a nation’s standpoint. We accept that the
controversies are irresolvable, as each democratic society will ground its
aspirations for its tax system in difering imperatives. We consider the
current ‘best practice’ approach to tax expenditure reporting not because
there is an international best practice to tax expenditure analysis. Rather,
we consider that a comparative survey of unique national approaches
to tax expenditure management provides suggestive examples for how
a particular nation’s process for managing tax expenditures might be
improved in furthering national public policy priorities. To this end, we
accept that the controversies are irresolvable, as each democratic soci-
ety will ground its aspirations for its tax system in difering imperatives.
However, the importance of tax expenditures warrants ongoing dis-
course with the aspirational goal of ‘good public policy’. Once a society
has determined its particular aspirational goals against which to measure
the use of tax expenditures, robust tax expenditure reporting and analysis
afords signiicant advantages. We address those advantages in Chapter 4.
In Chapter 5, we then deal with the politics of navigating the controver-
sies regarding tax expenditures.
3.2 International aspects of tax expenditures

Tax expenditure reports are produced at a domestic level and are, therefore,
able to be designed in a manner that allows nations to individually address,
in a modern democratic setting and without international conines, the
irresolvable theoretical controversies surrounding tax expenditures. his
does not mean that nations should operate in a silo and disregard what is,
albeit controversial, currently considered by some to be international ‘best
practice’. In addition to domestic tax expenditure reports, there are studies
which have been produced at an international level which may ofer higher
level guidance to the issues requiring resolution. hese reports, as noted
in Chapter 2, also provide an ancillary use for the reporting of tax expen-
ditures which has emerged over the decades; a comparative analysis at an
international level. his comparison allows for an assessment of both the
practices of various jurisdictions as well as an overall comparison across
nations of the iscal cost of individual tax expenditures, aggregated by cat-
egory and total tax expenditures relative to other government spending.
Despite the difering approach by nations as to how the controversies are
addressed, the type of comparative analysis undertaken in these studies
does provide signiicant, albeit limited, analytical value and we do not dis-
count the contribution this work has made. In this chapter, we attempt
to extend on this work by considering diferent nations’ tax expenditure
reporting elements to study the pragmatic adoption of the theoretical con-
troversies surrounding tax expenditure reporting and management. Before
we do this, however, the major and contemporary international reports and
comparative studies warrant a brief discussion for the two reasons already
mentioned. First, these international reports and studies ofer insight into
what is currently considered international best practice, and second, a cer-
tain limited amount of comparative information can be gained.
3.2.1 International reports

At an international level, both the OECD and the World Bank have under-
taken comprehensive comparative studies on tax expenditures with both
expressing concerns about, and releasing reports on, the treatment of tax
expenditures. he IMF has also commented on the role of tax expenditures
in a nation’s spending policy in its Manual on the Role of the Legislature
in the Budget Process5 and all three bodies have identiied aspects which
5
International aspects of tax expenditures 69
they consider to be best practice. Various other international bodies,

think tanks and academics have also produced either international ‘best
practice’ style reports or comparative studies.6 In fact, as discussed in
Chapter 2, complementing their theoretical work, McDaniel and Surrey,
in 1985, undertook a study of the international aspects of tax expendi-
tures where they considered six industrialised countries and developed
lists for these countries using uniform criteria. In this study they went so
far as to assert that broadly shared views as to the elements that constitute
a normative tax base exist7 and ‘despite the diferences among countries
with respect to forces that have shaped their tax systems – history, social
attitudes, culture, jurisprudence, political philosophies, etc. – tax experts
from diferent countries can and do maintain a fruitful dialogue about
their respective tax systems’. hey reason that this dialogue at an inter-
national level could only take place because of a shared understanding of
the fundamental elements of various taxes.8 In essence, while McDaniel
and Surrey maintained an elitist model for tax expenditure analysis and
suggested that it was only tax experts entering into dialogue, there was at
least limited recognition of broader social dialogue in shaping a country’s
tax regime.
3.2.1.1 he OECD
he OECD has led the way in the international reporting on tax expendi-
tures9 and adopts a pragmatic approach to its reports with the aspirational
goal of better policy.10 To this extent, its Working Party of Senior Budget
Oicials has the objective of improving the efectiveness and eiciency of
resource allocation and management in the public sector, with its work on
tax expenditures part of that agenda.
Despite the controversial nature of dei ning tax expenditures, the
OECD both recognises those controversies and deines tax expenditures
6
For examples of recent reports see: Nordic Working Group, Tax Expenditures in the
Nordic Countries (2010); Villela et al., Tax Expenditure Budgets: Concepts and Challenges
for Implementation (2010); and Burton and Stewart, Promoting Budget Transparency
hrough Tax Expenditure Management: A Report on Country Experience for Civil Society
Advocates (2011).
7
(1985), 5.
8
(1985), 5.
9
As previously noted, McDaniel and Surrey’s initial study on international aspects of tax
expenditures was published in 1985, around the same time as the OECD i rst reported on
tax expenditures.
10
as ‘provisions of tax law, regulation or practices that reduce or postpone

revenue for a comparatively narrow population of taxpayers relative to a
benchmark tax. For government, a tax expenditure is a loss in revenue;
for a taxpayer, it is a reduction in tax liability.’11 As noted, the OECD is
pragmatic in its approach and recognises that, in practice, the identi-
ication of tax expenditures may entail diicult judgements as to what
is part of the benchmark or an exception to it. In addition, as it states,
the identiication of a tax expenditure, which it considers may be allow-
ances, exemptions, rate relief, tax deferral or credits, requires more than
a broad and general dei nition.12 As will be seen below, this has trans-
lated into diferent countries adopting diferent dei nitions on what
constitutes the benchmark for their own reporting purposes, a propos-
ition which the OECD accepts as being the case and does not propose
a standard best practice dei nition. To this extent, the OECD has spe-
ciically stated in its most recent work on tax expenditures that ‘there
is no implication that the benchmark or normal tax system in any one
country should serve as a model for the benchmark for tax expenditure
analysis for all countries’13 – a point with which we agree. It also adopts
the view that the diferences between various countries’ tax expenditure
methodologies in general and their benchmark tax systems in particu-
lar does not prohibit analysis14 and nor should there be any implication
that the benchmark be devoid of politics.15
he earliest work on tax expenditures by the OECD Centre for Tax
Policy and Administration was its 1984 report entitled Tax Expenditures:
A Review of Issues and Country Practices.16 his was followed in 1996
by the report Tax Expenditures: Recent Developments17 and then the
2003 Special Feature for the 2003 Edition of Revenue Statistics: Note by
the Secretariat.18,19 Following these earlier works on tax expenditures,
it discussed at its 2004 meeting in Madrid the report by Kraan on
11
12
13
14
15
16
OECD, Tax Expenditures: A Review of Issues and Country Practices (1984).
17
OECD, Tax Expenditures: Recent Developments (1996).
18
OECD, Special Feature for the 2003 Edition of Revenue Statistics: Note by the Secretariat
(2003).
19
he OECD also provided a short section on tax expenditures in its Best Practices for
Budget Transparency (2002), stating:
‘Of-budget and Tax Expenditures’.20 Signiicantly, this meeting resulted

in the Working Party of Senior Budget Oicials producing Best Practice
Guidelines – Of Budget and Tax Expenditures.21 It is evident that these
best practice guidelines do not go to the fundamental controversies sur-
rounding tax expenditure analysis but, rather, focus on the reporting and
managements aspects. he guidelines relevant to tax expenditures are set
out below:
Guideline 10
Tax expenditures should be identiied by use of a benchmark tax. he benchmark
does not necessarily need to represent the normative tax base. he benchmark
should be comprehensive and unique.
Guideline 11
All tax expenditures should be estimated and integrated in the expenditure docu-
mentation that is presented to the budgetary authorities for all signiicant taxes.
Regular expenditures and tax expenditures should be shown in this documentation
side-by-side for the same number of years.
Guideline 12
Under nominal or structural deicit or operating/current balance rules tax expen-
ditures should either be included in the total expenditure cap that is set every year
during budget preparation or in a special tax expenditure cap. Under medium
term rules with multi-annual expenditure caps, tax expenditures should either be
included in the total expenditure cap of each year or in a special tax expenditure
cap of each year. Overspending on tax expenditures should be fully compensated,
at least in so far as it originates in policy change. If a special tax expenditure cap
is used, compensation can take place within that cap or through reduction of the
regular expenditure cap.
2.2 Tax expenditures

• Tax expenditures are the estimated costs to the tax revenue of preferential treat-
ment for speciic activities.
• he estimated cost of key tax expenditures should be disclosed as supplementary
information in the budget. To the extent practicable, a discussion of tax expendi-
tures for speciic functional areas should be incorporated into the discussion of
general expenditures for those areas in order to inform budgetary choices.
20
Kraan, ‘Of-budget and Tax Expenditures’ (2004).
21
OECD, Best Practice Guidelines – Of Budget and Tax Expenditures (2004).
Guideline 13
All tax expenditures should be reviewed in the same way as regular expenditures
in the annual budget process. hey should be reviewed by the i nancial staf of
spending ministers and the budget bureau in the same way as regular expenditures.
Special evaluation procedures, including program review, should be applied to tax
expenditures in the same way as to regular expenditures.
Guideline 14
Tax expenditures should be assigned to individual ministries. Objections of the
Minister of Finance against change of tax expenditures can never be used as an
argument against adjustment of other (tax-)expenditures if a ministry is required to
diminish its expenditures or ind compensation for overspending.
Guideline 15
Tax expenditures should be estimated by revenue forgone, corrected by an
equivalent tax margin, if equivalent expenditure transfers are taxed (or by outlay
equivalence).
Guideline 16
he responsibility for tax expenditure estimates should remain with the Ministry
of Finance.
Source: OECD, Working Party of Senior Budget Oicials, Best Practice Guidelines –
Of Budget and Tax Expenditures, 25th Annual Meeting of Senior Budget Oicials,
Madrid, Spain, 9–10 June 2004.
Most recently, in 2010 the OECD released its report entitled Tax
Expenditures in OECD Countries, outlining the use of tax expenditures
through a study of ten OECD countries: Canada, France, Germany,
Japan, Korea, the Netherlands, Spain, Sweden, the United Kingdom and
the United States. he OECD, in line with its objectives of providing its
member countries with the ability to compare policy experiences, iden-
tify good practice and work together towards co-ordinating domestic and
international policies in relation to economic, social and environmental
challenges, designed its tax expenditures project to allow government
oicials and the public to better understand both associated policy and
technical issues. To this end, the OECD study is a useful tool for both
developed and developing nations.
3.2.1.2 he World Bank

he World Bank, in 2004, published Tax Expenditures – Shedding Light on
Government Spending hrough the Tax System,22 highlighting many of the
concerns surrounding tax expenditures. In this report the World Bank out-
lines a framework for evaluating tax measures, highlights methodological
issues and reports on diferent nations’ approaches to the reporting of tax
expenditures focusing on Australia, Belgium, Canada, the Netherlands, the
United States, China and Poland. For its purposes, the World Bank deines
tax expenditures as ‘concessions that fall outside a tax norm or benchmark.
he tax norm includes the rate structure, accounting conventions, deduct-
ibility of compulsory payments, provisions to facilitate tax administration,
and international iscal obligations. Tax expenditures may take a number
of forms: exemptions, allowances, credits, preferential tax rates, tax defer-
rals, and so forth.’23 he World Bank study concentrated on the reporting
processes of developed, developing and transition countries and came to
similar obvious conclusions to previous studies; countries treat tax expen-
ditures diferently, but unless tax expenditures are exposed to adequate
scrutiny there is the opportunity for iscal opportunism.24 A result of this
broader focus, and of particular relevance, is the World Bank’s suggested
framework for evaluating tax measures. It adopts three criteria and specif-
ically recognises that diferent policy objectives can inluence the manner
in which the criteria are addressed.25 It lists the three criteria as:
Relevance: Is the tax measure consistent with policy priorities, and does it realistic-
ally address an actual need?
Efectiveness: Is the tax measure meeting its objectives efectively, within budget,
and without unwanted outcomes?
Eiciency: Is the tax measure the most appropriate and eicient means to achieve
objectives, relative to alternative design and delivery approaches?
Source: Polackova Brixi et al. (eds), Tax Expenditures – Shedding Light on Government
Spending hrough the Tax System: Lessons from Developed and Transition Economies
(2004), 19.
22
Polackova Brixi et al. (eds), Tax Expenditures – Shedding Light on Government Spending
hrough the Tax System: Lessons from Developed and Transition Economies (2004), 227.
23
24
25
In addition, the World Bank proposes four policy options for develop-
ing nations wishing to adopt tax expenditure analysis. hey are:
Policy option 1: he government should periodically compile an inventory of tax

expenditures and report on their nature, legal basis, expected and actual efects,
and past and likely future iscal costs.
Policy option 2: Annual budgetary documents, or other core iscal policy documents,
should discuss tax expenditures and their iscal costs and likely socioeconomic
efects in the context of the government’s overall iscal policy analysis.
Policy option 3: Before deciding on a new tax expenditure, the government should
clearly specify its objective, assess how this objective its with policy priorities,
evaluate alternative instruments for achieving the objective, and design the new tax
expenditure so as to minimize its possible negative efects.
Policy option 4: A new tax expenditure should be approved in the context of the
annual budget process, with the expected cost of the proposed tax expenditure
competing against proposed spending items and the expected cost of proposed pro-
grams of government contingent support.
Source: Polackova Brixi et al. (eds), Tax Expenditures – Shedding Light on

Government Spending hrough the Tax System: Lessons from Developed and
Transition Economies (2004), 227–32.
3.2.1.3 he IMF
he IMF, whilst not speciically addressing tax expenditures in a stand-
alone report, recently produced two documents which comment on the
signiicance of estimating and reporting key tax expenditures: the Manual
on Fiscal Transparency (2007) and the Manual on the Role of the Legislature
in the Budget Process (2010). he Manual on Fiscal Transparency provides
a narrow deinition of tax expenditures as ‘revenues foregone as a result
of selective provisions in the tax code’26 and lists examples such as exemp-
tions from the tax base, allowances deducted from gross income, tax cred-
its deducted from tax liability, tax rate reductions and tax deferrals, while
the Manual on the Role of the Legislature in the Budget Process emphasises
that tax expenditures are second best practice for delivering programmes
but where used, should be reported as part of the normal budget cycle.27
he IMF in its 2007 manual goes on to state:
26
IMF, Manual on Fiscal Transparency (2007), 76.
27
A statement of the main central government tax expenditures should be
required as part of the budget or related iscal documentation, indicating
the public policy purpose of each provision, its duration, and the intended
beneiciaries. Except in particularly complex cases, major tax expenditures
should be quantiies. Ideally, the estimated results of previous tax expendi-
tures compared with their policy purposes should also be presented so that
their efectiveness can be assessed relative to expenditure provisions.28
Again, this statement is an attempt to establish a minimum best practice

for the reporting and management of tax expenditures. he result that the
IMF is attempting to achieve is a reporting regime which allows an assess-
ment of the efectiveness of tax expenditures compared to direct spending
programmes.
3.2.2 What does this all mean?

he fundamental nature of the work of the OECD, the World Bank and the
IMF highlight both the diferences and similarities between various juris-
dictions which adopt tax expenditure analysis, along with a best-practice
approach being a system of robust and comprehensive tax expenditure
reporting, analysis and management in all democratic nations. To this
end, the 2004 World Bank report concludes that:
here are no generally accepted policy options and methodologies for
reporting tax expenditures, for evaluating their beneits and iscal and eco-
nomic costs, or for bringing them into the usual framework for government
decision making and scrutiny. Many industrial countries have standardised
their tax expenditure reporting. Several have taken steps toward analysing
the beneits and broad economic and direct iscal costs of tax expenditures.
Few have tried to manage tax expenditures strategically, in a broader frame-
work, such as through the annual government budget process.29
We agree with this proposition that there are no generally accepted policy
options and methodologies and recognise that it is the current position,
rather than as others do of drawing this proposition from an analysis of
diferent jurisdictions and relying on it as the basis for a conclusion for
their comparative studies. hese studies have revealed that tax expend-
iture budgets are not a neutral examination of the tax system and should
not be read as ofering a value-free analysis of a tax regime. Unlike Kahn
and Lehman, we argue that this does not mean they are exercises of
28
IMF, Manual on Fiscal Transparency (2007), 76.
29
mystiication.30 However, we do agree that tax expenditures operate in

a politicised domain, but we claim that this provides an opportunity for
enhancing the legitimacy of tax systems speciically, and government
more generally, in ‘modern’ democracies where it is commonly suggested
that more civic engagement upon diicult moral questions is desirable.31
As we stated in Chapter 1, this opportunity arises because the normative
aspect of the tax expenditure concept invites critical review of the coun-
try’s contemporary taxation system as well as consideration of myriad
alternate possible taxation systems. Identiication of a country’s norma-
tive taxation system is intrinsically contentious and can only be resolved,
albeit contingently, by a community’s adoption of a unique combination
of normative tax principles that should relect the community’s priorities.
As such, we move on to consider those choices which a community must
make in relation to tax expenditure reporting and management, starting
where we did with the theoretical controversies surrounding tax expen-
ditures; determining the purpose for which tax expenditures are recog-
nised and the function of tax expenditure reporting.
3.3 he impact on reporting of the purpose and function of

recognising tax expenditures
We previously observed that a nation’s purpose for categorising a pro-
vision of its tax regime as a tax expenditure will shape deinition of the
tax expenditure concept, which in turn, will afect the function that tax
expenditure analysis is designed to perform for that nation.32 his func-
tion and purpose will also govern and provide the basis for determining
what information is contained within a nation’s tax expenditure report
because the identiication of the function and purpose is central to deter-
mining and shaping the key elements to that report, especially the bench-
mark. he reporting may be designed to expose tax expenditures as a
form of government spending to inform decision making, enhance gov-
ernment iscal management, promote public consideration as to how to
achieve a nation’s social and economic policy goals or encourage analysis
30
Kahn and Lehman state: ‘While the several existing tax expenditure budgets give an
appearance of being the products of a highly sophisticated, expert, neutral examination
of the tax system, they could just as accurately be characterised as exercises in mystii-
cation. hey create an illusion of value-free scientiic precision in a heavily politicized
domain,’ Kahn and Lehman, ‘Tax Expenditure Budgets: A Critical View’ (1992), 1663.
31
See, for example, OECD, Citizens as Partners (2001).
32
See Chapter 2, section 2.3.
The impact on reporting of tax expenditures 77
of tax policy principles to promote a nation’s views on what it sees as a

fair, eicient and simple tax system. Based on a determination of that pur-
pose, the speciic information and level of detail to be provided in the tax
expenditures report will be ascertained.
3.3.1 he growth and maturity of tax expenditure reporting

Where nations adopt the view that tax expenditures are to be deined as
equivalent to direct spending programmes (which is generally accepted),
the primary purpose of reporting is obvious: to subject tax expenditures to
an acceptable level of scrutiny which ought to apply to both direct spend-
ing and tax spending alike. While emerging economies with democratic
governments have signiicantly improved their tax expenditure reporting
in recent years, many developing countries still only adopt this funda-
mental rationale and use tax expenditure reporting for the purposes of
iscal transparency and eicient resource allocation.33
However, the growth and maturity of tax expenditure reporting in
developed countries suggests that the purpose of this reporting extends
beyond treating tax expenditures in a manner similar to direct spend-
ing. More detailed reporting and a broader deinition of the tax expend-
iture concept relect more sophisticated purposes which, in turn, relect
a deeper concern with assessing the performance of a tax system meas-
ured against stated public policy objectives. Where this is done, nations
will weigh the competing imperatives such as eiciency and equity and
ascertain their priorities as discussed in Chapter 2 . Where eiciency is
seen as a priority, a government may estimate the iscal cost of a social
or economic policy introduced via a tax expenditure to determine
whether this is an eicient way to deliver the expenditure programme.
A nation’s tax expenditure reporting and management function in this
case will be to identify ineicient tax expenditures so that their removal
can lead to the lowering of overall tax rates, consequent reduction in
deadweight loss and hence greater national wealth.34 On the other hand,
where equity is a priority, tax expenditure reporting will be designed
to assess the equitable implications of the tax expenditures such as the
distributional impacts and efectiveness in achieving policy objectives.
A nation will use tax expenditure analysis and management for this
33
34
For consideration of the limitations of the eiciency norm see section 5.5.
purpose to reveal signiicant gaps in the impact of the tax expenditure

on, for example, who the true beneiciaries are, the gender impact and
impact on minority groups, to name a few. To this end, many nations
will state their adopted objectives as part of the process of reporting on
tax expenditures, while others will imply or inherently accept a speciic
purpose or purposes.
here are many country examples which can be cited which expli-
citly state their purpose of adopting tax expenditure reporting. In
Australia, for example, it is explicitly acknowledged that the annual
Tax Expenditures Statement serves three key functions: to allow tax
expenditures to receive a similar degree of scrutiny as direct expen-
ditures; to allow for a more comprehensive assessment of government
activity; and to contribute to the design of the tax system, by promot-
ing and informing public debate on all elements of the tax system.35 In
the United States, where tax expenditure analysis plays a major role in
policy debate, 36 and it is legislated that the President’s annual budget
submission contain a list of tax expenditures, as well as Congress’s Joint
Committee on Taxation produce its own tax expenditures statement,
the purpose is also explicitly stated. he Joint Committee on Taxation
report speciically provides the purpose for which tax expenditure
reporting is undertaken, stating that ‘tax expenditure analysis can help
both policymakers and the public to understand the actual size of gov-
ernment, the uses to which government resources are put, and the tax
and economic policy consequences that follow from the implicit or
explicit choices made in fashioning legislation’.37 Both the Australian
and United States examples demonstrate broader purposes than mere
comparison with direct spending and focus on tax reform. In contrast,
many European countries focus on the budgetary processes relevant to
tax expenditures. For example, Belgium uses its annual report to assess
the impact of tax expenditures on revenue while France uses its annual
statement to facilitate the budget process.38
he World Bank study, in 2004, summarised the purpose for which
ten selected OECD countries reported on tax expenditures, along with
35
Australia, 2011 Tax Expenditures Statement (2012), 14.
36
37
United States of America, Estimates of Federal Tax Expenditures for Fiscal Years 2009–
2013 (2010), 1.
38
hrough the Tax System: Lessons from Developed and Transition Economies (2004), 27–8.
The impact on reporting of tax expenditures 79
whether those countries were faced with a legal requirement to report, the
relationship of that report with the respective national budget, and the
frequency of the reporting.39 It found a mixture of purposes including but
not limited to tax reform,40 facilitating the budget process,41 assessing the
impact of various tax measures on revenue,42 pre-budget consultations,43
subsidies and expenditure reduction,44 shaping tax reform45 and provid-
ing the Parliament with insight into budgetary cost of tax expenditures
and possible budgeting.46
Non-OECD and developing nations such as India, South Africa, Chile
and Brazil are also becoming increasingly sophisticated in their iscal
reporting and have all introduced tax expenditures statements for various
purposes. For example, India’s goal is to monitor and assess tax expen-
ditures for the purpose of identifying tax incentives which increase the
deadweight costs, distort resource allocations and stunt productivity,
while in South Africa, emphasis is placed on iscal transparency. In South
Africa, a nation which by its own admission has faced signiicant hurdles
in the reporting of tax expenditures due to the lack of useful and accur-
ate data and a serious manpower shortage in revenue, has now published
a report several times including 2003, 2006 and 2011. In these reports,
South Africa considers that a tax expenditure statement serves three
broad objectives: it increases iscal transparency and accountability; it
creates comparability between direct and indirect government expendi-
tures; and it assists in the design of tax policy by promoting and inform-
ing public debate.
Surprisingly, there are also developed countries with a very rudimen-
tary approach to tax expenditure reporting. For example, New Zealand,
ater not having reported on tax expenditures since 1984, recently started
voluntarily reporting on them again.47 Its approach is a cautious one
designed to provide additional transparency around what it refers to as
policy motivated expenditures made through the tax system and it spe-
ciically states that the report is not a statement of policy efectiveness or
desirability.48
39
See Polackova Brixi et al. (eds), Tax Expenditures – Shedding Light on Government
(2004), 10–15.
40
Austria. 41 Austria. 42 Belgium.
43
Canada, France, United Kingdom.
44
Germany. 45 United States. 46 Netherlands.
47
he irst tax expenditures report since 1984 was released in 2010.
48
New Zealand, 2012 Tax Expenditure Statement (2012), 1.
3.3.2 he regulatory status of tax expenditure reporting

he regulatory status of tax expenditure reporting within a nation may
also inluence the purpose and function. Diferent nations’ tax expend-
iture reporting will have diferent legal status and frequency, may or
may not be linked to the budget, may cover all or only certain taxes, and
may only cover the central government or additionally report on local
governments. Ideally, a nation’s tax expenditure report is supported by
a legal and institutional framework, whether part of the budget legis-
lation or tax regime, which provides for such factors as methodology,
information management, evaluation, control and oversight. As such,
oten nations will bind their stated or implicit purpose for tax expend-
iture analysis and management to factors such as any regulatory obliga-
tions to report on tax expenditures along with the information required
by that regulation. For example, speciic bodies in countries such as
Australia, Austria, Belgium, France, Germany, Italy and the United
States are all legally obliged under domestic legislation to provide tax
expenditure statements and, while not all of these nations must report on
an annual basis,49 most will.50 Some developing countries such as those
already mentioned of India, South Africa, Chile and Brazil have also
made substantial progress in legislating for the reporting of tax expen-
ditures and also report on an annual basis. On the other hand, nations
like Canada, which has no statutory obligation to provide a tax expen-
ditures report, does so annually for the purposes of pre-budget consult-
ation. In the United Kingdom, while there is no statutory requirement
for the government to produce a report on tax expenditures, estimates
of all major tax expenditures are provided on an annual basis in the
‘Financial Statement and Budget Report’ and reviewed twice yearly dur-
ing the pre-budget and budget process. hen there are other nations,
such as Italy, which only report on a sporadic basis.
3.3.3 Resource and data constraints

A nation’s best practice would theoretically require the determination
of the purpose and function of tax expenditure analysis unfettered by
49
For example, Germany has a legal obligation to report biennially and Italy is not subject
to regular reporting requirements.
50
See Polackova Brixi et al. (eds), Tax Expenditures – Shedding Light on Government
(2004).
Tax expenditures so defined 81
pragmatic constraints. However, each nation will not only ascertain the
information to be contained in its tax expenditures report ater decid-
ing on the purpose of recognising tax expenditures taking into account
its own objectives, but will also moderate that choice in accordance with
resource and data constraints which afect implementation and main-
tenance capacity. As such, the costs associated with attaining theoretical
ideals will be weighed against practical limitations. A review of vari-
ous tax expenditure reports reveals that information contained in those
reports ranges from the simple identiication and basic costing of tax
expenditures to more detailed and sophisticated descriptions and ana-
lysis of a comprehensive list of tax expenditures, along with details of the
benchmark tax system adopted, and the iscal cost of the individual tax
expenditures as the estimated current revenue foregone as well as future
projected revenue forgone. Within a nation’s own tax expenditures state-
ment it may include a deinition of ‘tax expenditure’, the comprehensive
identiication of the types of tax expenditures it adopts, a deinition of the
‘benchmark tax law’, a method for estimating revenue foregone, structure
of the reporting of tax expenditures and information for each tax expend-
iture. We consider the difering approaches to each of these elements of
information below.
3.4 Tax expenditures so deined

It is obvious that a deinition of tax expenditures per se is fundamental
to the reporting and analysis of tax expenditures. It was demonstrated
in Chapter 2 that there is no universally agreed deinition and, as such,
because we argue that it is a matter for each society, based on its moral
philosophy, to determine how it dei nes tax expenditures for its own
determined purpose, deinitions will vary across nations. Further, those
deinitions of tax expenditures invariably link it to another, just as elusive
concept, being a benchmark from which tax expenditures are considered
deviations. No matter what deinition is accepted by a nation, tax expend-
iture reports should ideally contain a deinition of a tax expenditure not
only for income tax but for each type of legislated tax within that juris-
diction. Given countries which report on tax expenditures have generally
adopted their own language and approach for the purposes of a deinition
it is necessary to search for a commonality, with that commonality being
a connection between the loss of revenue to the government which results
from a tax concession or tax preference allowed to a particular class of
taxpayer or activity.
Country examples reveal the diverse range of approaches to the dein-

ition of tax expenditures demonstrating that some tie the deinition of tax
expenditures to a normative tax base, while others do not. For example,
for the purposes of the United Kingdom report, a formal deinition is
not provided but rather tax reliefs are divided into three categories: irst,
reliefs which are considered alternatives to direct spending programmes
are referred to as tax expenditures; second, reliefs which are considered
integral to the tax structure or simplify the tax system are referred to as
structural reliefs; and third, tax reliefs which combine elements of the
irst two categories. his can be contrasted with Australia where Treasury
reports on tax expenditures annually in a separate government docu-
ment and a formal deinition is provided which is attached to a normative
benchmark. Australia deines a tax expenditure as:
a provision of the tax law that provides a beneit to a speciied activity
or class of taxpayer that is concessional when compared to the ‘nor-
mal’ tax treatment that would apply. A negative tax expenditure arises
when arrangements impose an additional charge rather than beneit.
Tax expenditures can be provided in many forms, including tax exemp-
tions, tax deductions, tax ofsets, concessional tax rates or deferrals of tax
liability.51
Tax expenditures in the United States are defined under the

Congressional Budget and Impoundment Control Act of 1994 as ‘revenue
losses attributable to provisions of the federal tax laws which allow a
special exclusion, exemption, or deduction from gross income or which
provide a special credit, a preferential rate of tax, or a deferral of tax liabil-
ity’.52 For the purposes of identiication, the United States then considers a
tax expenditure to be a preferential exception to the baseline provisions of
the tax structure. As highlighted in Chapter 2, this baseline structure has
been subject to a great deal of controversy in the United States recently
with the Joint Committee on Taxation, in 2008, adopting for one year only
and then abandoning a new paradigm for identifying tax expenditures.53
In South Africa, tax expenditures are deined as: ‘In brief, tax expendi-
tures are provisions in tax legislation that reduce the amount of tax rev-
enue that could otherwise have been collected. Tax expenditures can be
deined as deviations from the benchmark of a standard tax legislative
51
52
Congressional Budget and Impoundment Control Act of 1994 (Pub. L. No. 93–344), sec.
3(3).
53
Base?: A Critique of the “New Paradigm” and its Denouement’ (2010).
The benchmark tax law 83
framework.’54 As such, it links this deinition to a benchmark so deined.

he linking of the deinition of a tax expenditure to a benchmark is also
adopted by countries such as Belgium, Canada, France, the Netherlands
and Chile, to name a few.
New Zealand’s recently reinvigorated voluntary reporting of tax
expenditures deines a tax expenditure as ‘an exemption, allowance, pref-
erential tax rate, deferral or ofset that reduce a tax obligation to achieve
a speciic policy objective’.55 In classifying government spending, it delib-
erately avoids using a normative benchmark, further deining tax expen-
ditures as ‘individual features of the tax system that reduce an entity’s tax
obligation in a way that is designed to give efect to policy other than to
raise revenue in the most eicient and economically neutral way’.56 he
stated purpose of avoiding the use of a normative benchmark is to allow a
focus at irst instance on a narrow subset of tax expenditures which ‘bear
a distinct iscal cost and represent a clear policy-motivated exemption to
current practice’.57
We could continue with a list of difering deinitions of tax expendi-
tures with each nation’s choice revealing a diferent approach. What is
common, however, to all of the deinitions, is the need for a benchmark or
baseline from which to determine whether a provision of the tax system is
classiied as a tax expenditure. As such, we move to the benchmark.
3.5 he benchmark tax law

While nations may provide a deinition of tax expenditures, it is the dein-
ition of the benchmark tax law which is the basis for a nation’s identiica-
tion of its list of tax expenditures, as the tax expenditure concept involves
an assessment of those rules or practices which are a departure from the
benchmark so deined. We have demonstrated that theoretical contro-
versies abound; however, it is essential that if a nation wishes to provide a
comprehensive list of tax expenditures it needs to establish the deinition
of the benchmark which it adopts. his choice of an appropriate bench-
mark will always be the most diicult for a nation to agree on at a domes-
tic level and it is unlikely that any form of international consensus will
ever be reached. While one jurisdiction may consider a provision of the
54
Republic of South Africa, 2012 National Budget (2012), Annexure C, 1.
55
New Zealand, 2012 Tax Expenditure Statement (2012).
56
57
tax legislation to be a tax expenditure another jurisdiction may consider

the same type of provision to be part of the benchmark tax base.58 his is
because in practice this benchmark, from which deviations are identiied,
is problematic, controversial and deeply political in nature.
Given there is no generally accepted benchmark, nations will deter-
mine which approach to adopt at a domestic level taking into account the
tradeofs between competing policy objectives embodied in their tax law.
Ideally, a nation will deine and disclose their benchmark tax system, stat-
ing why it is appropriate, and the reason for the choice of its particular
benchmark for each type of tax, to allow for public debate. While advo-
cating for an elitist model, Surrey and McDaniel recognised the role that a
nation’s own views play in determining what constitutes a tax expenditure.
However, they did so only in the context of borderline issues. hey state:
However, in some borderline situations, matters such as the history and
present stage of tax development, the sophistication of the tax adminis-
tration, and the attitudes of the country’s citizenry must be taken into
account in the process of classifying a particular deviation from the norm
as a tax expenditure. he purpose of a tax expenditure list is to give use-
ful information about the extent to which government is intervening in
the economic and social life of a country by running spending programs
through its tax system.59
Previous studies which have considered the diferent national approaches

to benchmark deinitions generally allocate the method adopted by dif-
ferent nations to one of three categories: the conceptual benchmark
approach, the reference benchmark approach or the substitution bench-
mark approach.60 Generally, and for the reasons stated below, the concep-
tual benchmark approach will result in the largest list of tax expenditures
with the reference and substitution benchmark approaches resulting in a
more limited list. Craig and Allan, in their comparative study of the difer-
ent benchmarks adopted,61 and Villela et al., in their study on the concepts
58
For a discussion on the diiculties associated with determining an appropriate national
benchmark for separate jurisdictions because of unique tax bases and economic struc-
tures see: Glenday and Swit , Establishing Tax Expenditure Accounts – Towards Better
Fiscal Accountability and Transparency (2006).
59
(1985), 6.
60
See, for example, Craig and Allan, Fiscal Transparency, Tax Expenditures, and Budget
Processes: An International Perspective (2002) and Villela et al., Tax Expenditure Budgets:
Concepts and Challenges for Implementation (2010).
61
Craig and Allan, Fiscal Transparency, Tax Expenditures, and Budget Processes: An
International Perspective (2002), Appendix 1.
and challenges for the implementation of tax expenditure budgets,62 con-

clude that a mixture of the three approaches is adopted. hese two studies
note that Australia, Belgium, Canada, Chile, Finland, Ireland, Portugal,
Spain, Sweden and the United States all use the conceptual baseline in
some form, while Argentina, Austria, Brazil, Colombia, France, Korea
and the Netherlands use the reference benchmark approach.63 Germany
and the United Kingdom adopt the third method, the substitution bench-
mark approach. Calls for the abandonment of tax expenditure analysis
are founded in the diiculties in determining the benchmark and argu-
ably the diferent approaches support this claim. Some would argue that
the diferent approaches, along with the diferent interpretations of the
benchmark within each of the approaches, means that determining what
is a tax expenditure is arbitrary because formulating the income deinition
is arbitrary.64 We disagree. Rather, we argue these diferent approaches
simply relect the diferent priorities of nations as to the purpose and
function of tax expenditure analysis. he diferent approaches should be
embraced and encouraged through a democratic process.
3.5.1 he conceptual benchmark approach

A conceptual benchmark which is the most common starting point is also
the most controversial as it links the benchmark tax to a normative struc-
ture. For example, for income taxes the Schanz–Haig–Simons concept
of income is used as the starting point. It is also known as an ‘external’
or ‘ideal’ benchmark because of its linking to tax policy principles irre-
spective of whether those principles are embodied in a nation’s tax law.
In essence, tax expenditures are identiied as departures from a policy
ideal. As discussed in Chapter 2, nations normally adopt an economic or
comprehensive income tax benchmark which includes all economic gains
and has only limited modiications. As an alternative, a broad based con-
sumption tax benchmark can be adopted which covers the consumption
of all goods and services and may assume a lat rate of tax.
heoretical work on the normative base as devised by Surrey tends to
always centre on the Schanz–Haig–Simons dei nition of income. Surrey
62
Villela et al., Tax Expenditure Budgets: Concepts and Challenges for Implementation
(2010).
63
he United States uses both the conceptual benchmark and the legal benchmark in lim-
ited circumstances.
64
did however recognise the need for adjustments to be made. McDaniel

and Surrey state: ‘hus, in preparing tax expenditure lists for income
taxes, a normative concept of income is helpful both as a starting point
and a check during the classiication process. he norm, however, is to
be tempered by … considerations … to avoid a rigidity which might
have theoretical appeal but little practical utility.’65 In practice, many
countries do exactly this and vary from the theoretical base by adopt-
ing a base which more closely resembles their actual income tax. Many
nations start with something that resembles the Schanz–Haig–Simons
dei nition but then modify that dei nition to take into account their
own standard taxation treatment as well as standard structural elem-
ents. As such, the manifestation of the dei nitions varies signiicantly.
Some jurisdictions are highly prescriptive, more closely relecting an
objective standard, whilst others are suiciently vague to require sig-
niicant judgemental lexibility. Hence, there are major disagreements
which arise in relation to the benchmark even when this model is used.
An individual nation will decide what the normal structure of its tax
system is, based on policy answers to questions relating to its tax sys-
tem in general.66 It is these diferent answers at a domestic level which
result in variations to the benchmark. h is also extends to other taxes
in many jurisdictions.
In practice, it can be argued that most nations adopt a combination
of the comprehensive benchmark and reference benchmark by start-
ing with an external benchmark to identify tax policy ideals and then
modifying that to take into account internal features that are considered
part of the structural elements or design features of the tax system. No
matter the approach adopted, the normative tax base will generally be
inluenced by what a society considers to be a gain, with nations then
considering structural elements of the tax regime as part of the bench-
mark. hese structural elements include the tax rate structure, the unit
of taxation, the tax period, a minimum tax free threshold or standard
exemption, accepted deductions, jurisdictional rules and administrative
rules. here are also private assets such as imputed rent from living in
your own home and the sale of a home, which are also generally consid-
ered part of the benchmark.
65
(1985), 6.
66
(1985), 6.
3.5.1.1 Country examples

he United States is an example of a nation which primarily relies on the
conceptual benchmark, but also uses the reference benchmark approach
in certain cases. It is also the nation which has, in practice, seen the most
internal controversy over its choice of benchmark. Tax expenditures in the
United States as deined in the Congressional Budget and Impoundment
Control Act of 1974 rely on what is referred to as a normal tax as the bench-
mark; however, no deinition of that benchmark is given. he two sep-
arate bodies, the United States Treasury and the United States Congress
Joint Committee on Taxation, which prepare tax expenditure budgets in
the United States, both use the conceptual benchmark. However, in add-
ition, the United States Treasury provides estimates for the reference or
legal benchmarks as well, which is argued to be closer to existing law, and
it reports estimates of those tax expenditures in terms of both revenue
foregone and outlay equivalent. he United States is a unique example
of a nation having two separate reporting bodies producing separate tax
expenditure reports with those two reports varying in their interpretation
of the tax baseline. For example, the baseline adopted by Treasury excludes
certain tax expenditures which are included by the Joint Committee on
Taxation.67 In Chapter 2, we addressed the 2008 abandonment by the
United States Congress Joint Committee on Taxation, in favour of a two
part methodology for identifying tax expenditures. his was an attempt
to adopt a modiied version of the reference tax approach.68
Unlike the United States, Australia provides a deinition of what it con-
siders the ‘normal’ tax system or benchmark, with that benchmark based
on two principles. First, the benchmark represents the standard taxation
treatment that applies to similar taxpayers or types of activities. Second,
the benchmark incorporates structural elements of the tax system, that
is, those elements of the tax system that depart from the uniform treat-
ment of taxpayers but are integral design features of the regime.69 he
benchmark is then split into three major components: the income tax
benchmark, the consumption tax benchmark and the externalities tax
67
For a summary of the diferences see Ghilarducci, Beyond 401(k)s: Guaranteeing
Retirement Security for all Americans (2010).
68
For criticisms of this adoption see New Zealand, 2012 Tax Expenditure Statement
(2012), and McDaniel, ‘he Staf of the Joint Committee on Taxation Revision of Tax
Expenditure Classiication Methodology: What Is to Be Made of a Change that Makes No
Changes?’, (2011).
69
benchmark. he income tax benchmark uses the Schanz–Haig–Simons

deinition of income but then departs from the standard baseline with the
inclusion of structural features such as the taxation of capital gains on a
realisation basis. he consumption tax benchmark is much more prescrip-
tive and is based on the consumption of speciied categories of goods and
services such as fuel, tobacco, alcohol, and inal consumption of goods
and services by households. Also included is the natural resource bench-
mark which references the extraction and production of Australia’s nat-
ural resources. he externalities benchmark takes into account revenue
generating taxes which have ‘the purpose of ensuring that the private
costs of certain activities align with the social costs of those activities’70
rather than the traditional general revenue collection purpose of taxes on
income or consumption. he Australian approach has been described as
a tripartite division of tax laws: core taxing rules (including provisions
that are central to the operation of the tax system, international tax rules
and generally accepted principles); boundary rules (that is, those which
may not fall within the irst category but are accepted as central to the
operation of the taxation laws); and tax expenditures (whether nega-
tive or positive).71 he consequence of such an approach means that the
reporting of tax expenditures is highly dependent on Treasury identify-
ing the appropriate benchmark, and, by their own admission, within the
Australian context, ‘benchmarks vary over time and across countries and
can be arbitrary’.72 In Australia: ‘Tax benchmarks are not deined in law
or in accounting standards. Instead, a tax benchmark is an abstraction
derived from Treasury’s analysis of the Australian taxation system as it
stands at a given point in time.’73
Tax expenditures in Canada are deined broadly with only the most
fundamental parts of the tax system considered part of the benchmark.74
Canada provides a benchmark for personal and corporate income tax sys-
tems, as well as its goods and services tax. he benchmark for the personal
and corporate income tax systems is deined in Canada’s Tax Expenditures
and Estimated Report 2010 as considering the tax base, existing tax rates
and brackets, the unit of taxation, the time frame of taxation and the
treatment of inlation for calculating income. In addition, the benchmark
70
71
Australia, Preparation of the Tax Expenditures Statement (2008), 49.
72
Australia, 2011 Tax Expenditures Statement (2012), 21 and 201.
73
74
includes measures that reduce or eliminate double taxation, recognise

expenses incurred to earn business income and allow business losses to
be claimed over a number of years. Finally, the constitutional immun-
ity of Canada and the provinces from taxation is recognised as part of
the benchmark system for income taxation.75 Canada, with its highly pre-
scriptive model, describes its base as follows:
he benchmark income tax base is a variant of the Haig–Simons compre-
hensive income base, which requires the taxation of real current additions
to purchasing power, or real increases in wealth. Under a Haig–Simons
base, worldwide income from all sources – labour income, rents, divi-
dends, interest, transfers, capital gains, imputed rent on owner-occupied
dwellings, imputed value of household services, gits and inheritances – is
taxed as it is earned. A strict application of the Haig–Simons base would
make corporate income tax redundant since income earned at the cor-
porate level would be taxed as it accrued to individuals. he benchmark
assumes, however, that corporate net income from all sources is subject
to taxation.
In deining the personal income tax benchmark, the Haig–Simons dei n-
ition is modiied to exclude non-market transfers of money and property
between taxpayers as well as the imputed values of rent and household
services. With this deinition of the tax base, measures that provide pref-
erential tax treatment of savings, such as Registered Retirement Savings
Plans and Tax-Free Savings Accounts, are considered tax expenditures.
Non-market transfers of presumptively tax-paid amounts between tax-
payers, such as gits, inheritances and spousal/child support payments,
are neither taxable in the hands of the recipient nor deductible by the
donor under this benchmark .
3.5.2 he reference benchmark approach

An alternative to the conceptual approach is the reference approach,
adopted by some nations where the tax legislation provides the basis for
deining the benchmark. As noted above, nations are oten considered to
have applied a conceptual benchmark modiied for provisions of its tax
regime, thereby in practice combining the two approaches. he reference
benchmark approach is also known as an ‘internal’, ‘normal’, ‘legal’ or
‘baseline’ benchmark. Using this method, a tax expenditure will be iden-
tiied where there is an explicit exemption from a tax provision, or an
existing law. While this approach, given its pragmatic nature, avoids some
of the controversies associated with the conceptual approach, it generally
75
Canada, Tax Expenditures and Evaluations (2011).
provides a more limited list of tax expenditures because it requires a devi-

ation from a posited tax, and, as such, many hidden tax expenditures will
not be captured. Contributing to its limitations is also the fact that it is not
set independently with reference to objective tax principles.

Despite the perceived limitations of using a reference benchmark, nations
do adopt such an approach. For example, the Netherlands, which adopts
a deinition of a tax expenditure which ‘is a government spending in
the form of a loss or deferment of tax revenue that is due to a tax provi-
sion insofar as that tax provision is not in accordance with the bench-
mark structure of the tax law’,76 then deines its benchmark tax structure
according to the legislated elements of its tax regime. France also adopts
a reference benchmark approach. It deines tax expenditures by reference
to legal or statutory measures which lower tax revenue as compared to
the benchmark or ‘norm’.77 In this case the ‘norm’ is then determined by
considering what the intentions of the legislature were.78
3.5.3 he substitution benchmark approach

A third possible approach is one of substitution or the analogous sub-
sidy approach. hat is, only those tax expenditures which can be replaced
with a direct subsidy are identiied and reported on. Again, this is likely to
produce a limited list of tax expenditures and fail to capture many of the
hidden ones, and in practice will result in a very similar outcome to the
reference approach. his method, which is a very traditional approach to
tax expenditure analysis, is used by very few nations.

he tracing of the origins of the substitution benchmark approach goes
back to the use of tax expenditure reporting by Germany in the 1960s.
Germany does not provide a legal deinition of tax expenditures or its tax
expenditure benchmark but, rather, considers tax expenditures to be pro-
visions which aid enterprises and business sectors of the economy.79 he
German approach has been described by Shannon as follows:
76
77
78
79
Identification of all types of tax expenditures 91
Rather than identifying tax expenditures by comparing individual provi-
sions with the generally accepted structure of an income tax, the German
analysis begins much further back. In the German literature, identify-
ing tax expenditures is a process of interpretation based on the princi-
ples underlying the choice of income as a tax base. hus, German analysis
recognizes that the principles giving rise to our choice of income as a tax
base provide guidance for structuring the income tax. While speciic pro-
visions may not follow a priori from general principles, those principles
serve as guidelines for decision making. hey limit the realm of accept-
able choices.80
he United Kingdom also adopts a similar approach by identifying what it

considers to be the three categories of tax expenditures discussed above.
3.6 Identiication of all types of tax expenditures

Ideally, a nation’s tax expenditure statement will comprehensively report
on all of its tax expenditures. However, in practice, this identiication is
limited by two factors: irst, the deinition adopted by a nation, and second,
resource and data constraints. We argue that ‘any’ reporting is better than
no reporting of tax expenditures as it will go some way towards address-
ing the purpose for which a nation identiies tax expenditures and can
be a useful public policy tool. However, we also argue that ideally, where
there are no constraints, all tax expenditures should be identiied.
3.6.1 he limiting nature of the tax expenditures deinition

We previously argued for the expansion of the deinition of tax expendi-
tures to include implicit and operational expenses.81 To date, this argu-
ment is purely theoretical as nations have not recognised this expanded
dei nition but, rather, retain its more traditional form. An example,
however, can be provided of the types of non-express tax expendi-
tures to which we refer. New Zealand, with its recent restoration of tax
expenditure reporting, takes a highly conservative approach. Further,
New Zealand currently does not impose a tax on capital gains. Under
a comprehensive tax benchmark, the exemption of these capital gains
from being taxed would be treated as a tax expenditure. However, New
Zealand does not identify the exemption of capital gains from taxation
80
81
See Chapter 2, section 2.7.
as a tax expenditure because it is not an expressly legislated exemption
but rather an omission to tax.
3.6.2 Resource and data constraints

In practice, tax expenditure reports only identify the most iscally signii-
cant tax expenditures and report on those in detail. To this end, a nation
will again need to weigh various imperatives to determine which are the
signiicant tax expenditures to be reported, with iscal constraints most
likely to signiicantly afect developing nations.
Countries that report on tax expenditures generally provide coverage
of both personal and income taxes; however, not all nations report on con-
sumption taxes. For example, while the United States reports on federal
corporate income taxes, estate and git taxes and social security contribu-
tion, it does not report on VAT or other various taxes such as excise tax.
Further, all countries list tax expenditures at a central government level
but many do not report on various expenditures at other levels of govern-
ment. Where nations are structured as a federal system under a decentral-
ised model, this is not ideal as oten both state (provincial/regional) and
local (municipal) governments apply numerous taxes, with these systems
also containing numerous tax expenditures. Many of a nation’s tax expen-
ditures will be at this level and will not be reported where the central gov-
ernment only provides information on its own tax expenditures. Yet, even
where sophisticated tax expenditure reporting is undertaken, very little
progress has been made in terms of extending that reporting beyond cen-
tral government which is best placed to manage aggregated reporting to
assist with better management of all of a nation’s tax expenditures as well
as provide the community with a more realistic picture of the true situ-
ation relating to tax expenditures. Even where reporting is undertaken at
a sub-national level, it tends to be poorly done relative to national report-
ing. For example, in the United States where 44 of the States do report on
tax expenditures (6 fail to do so) many of those reports have been criti-
cised for their perceived signiicant shortcomings.82 It has been revealed
that of those 44 States, 10 omit major taxes from their report, 6 fail to pub-
lish a report at least once every two years, and almost every State’s report
omits some essential information, such as the law that mandates a given
tax expenditure or the number of households or businesses that beneit,
82
Measurement of tax expenditures 93
with some even omitting the cost of many tax expenditures.83 Further, 2
States fail to make their report publically accessible.84
In the United Kingdom, tax expenditures are measured only at the cen-
tral government level, however, a wide range of taxes are subject to ana-
lysis. Its statement lists tax expenditures under the headings of income
tax, corporation tax, national insurance contributions, capital gains tax,
inheritance tax, value added tax, landill tax, climate change levy, hydro-
carbon oils duty, stamp duty land tax, passenger duty and vehicle excise
duty.85 Australia is similar in that it reports on tax expenditures relating
to income tax (personal and business), including capital gains tax (CGT)
and income, tax paid on retirement income, fringe beneits tax (FBT),
the goods and services tax (GST), excise duties, customs duty (includ-
ing tarifs), wine equalisation tax, luxury car tax, petroleum resource rent
tax, crude oil excise and other indirect taxes, but it does not report on
state taxes. Germany provides one of the most comprehensive lists of tax
expenditures and identiies tax expenditures relating to income tax, cor-
porate tax, net worth tax, business tax, turnover tax, insurance tax, motor
vehicle tax, excise taxes, betting and lottery tax, property tax and inher-
itance tax.86
he reports of developing countries are continuing to expand to
include not only income taxes and corporate taxes but also other categor-
ies. For example, India reports on tax expenditures in relation to personal
income tax, corporate income tax, excise duty and customs duty, while in
South Africa tax expenditures relating to personal income tax, corporate
income tax and VAT are reported.
3.7 Measurement of tax expenditures

In practice, for the purposes of reporting tax expenditures, it is necessary
to determine a basis for measuring the iscal cost of each expenditure.
Further, whilst not all tax expenditures can be measured in iscal terms,
where it is possible to do so, jurisdictions tend to attempt such an exer-
cise, obtaining information through the direct gathering of data from the
83
Reporting (2011), 3–4.
84
85
www.hmrc.gov.uk/stats/tax _expenditures/table1–5.pdf (last accessed 7 December 2012).
86
hrough the Tax System: Lessons from Developed and Transition Economies (2004).
revenue authority, or the use of microsimulation models. We concluded

the section in Chapter 287 on measurement by highlighting the three
methods for estimating the cost of tax expenditures. In this part, we pro-
vide further detail as to the decisions a nation will need to make as to the
choice of methodology.88
he choice of measurement of tax expenditures will fundamentally
be dependent on the objective being pursued by a nation, along with its
resource and data constraints. he three methods – revenue foregone,
revenue gain and outlay equivalent – each have their advantages and dis-
advantages. Each involves a diferent concept as to how the iscal cost is
determined based on the recognition of changes in taxpayer behaviour.
Whatever the method which a nation adopts for estimating the iscal cost
of its individual tax expenditures, ideally it should be clearly explained in
its tax expenditures report. Oten, nations also state the degree to which
each tax expenditure estimate is reliable, as this is a signiicant factor
afecting the usefulness of the data contained in the statement. Whether
the revenue foregone, revenue gain or outlay equivalent approach is used,
a nation will need to calculate the iscal cost of each tax expenditure where
it is able to do so. In Chapter 4 we review some diiculties with quanti-
fying tax expenditures as well as measuring their impact with respect to
various public policy imperatives.
3.7.1 he revenue foregone approach

he revenue foregone approach is the most common method adopted by
jurisdictions to measure the iscal cost of individual tax expenditures.
A survey of nations reveals that Australia, Austria, Belgium, Canada,
Denmark, Finland, France, Germany, India, Italy, the Netherlands,
Norway, the United Kingdom, the United States, South Africa and
Sweden all estimate the iscal cost of tax expenditures on a revenue fore-
gone approach. Although, some of these nations add to this method with
87
Chapter 2, section 2.6.
88
here are many sources which discuss the advantages and disadvantages of the vari-
ous methods. he discussion below is a summary of those. See, for example, Polackova
Brixi et al. (eds), Tax Expenditures – Shedding Light on Government Spending hrough
the Tax System: Lessons from Developed and Transition Economies (2004); OECD, Tax
Expenditures in OECD Countries (2010); Nordic Working Group, Tax Expenditures in the
Nordic Countries (2010); Villela et al., Tax Expenditure Budgets: Concepts and Challenges
for Implementation (2010); and Burton and Stewart, Promoting Budget Transparency
hrough Tax Expenditure Management: A Report on Country Experience for Civil Society
Advocates (2011).
Measurement of tax expenditures 95
further estimates using the revenue gain approach or outlay equivalence

approach.
his method is also the easiest to apply as it measures the cost of a tax
expenditure by comparing the amount of tax revenue raised under the
existing tax law (including the tax expenditure) with the amount of rev-
enue that would have been raised under the existing tax law, excluding
the tax expenditure. Put simply, it is an ex post calculation of the loss in
revenue incurred by a government.89 No further data is required as this
method assumes that all other tax provisions, and other laws, remain the
same and that taxpayer behaviour is unchanged in response to the hypo-
thetical abolition of the tax expenditure. However, the fact that the revenue
foregone approach does not account for taxpayer behaviour is its greatest
failing. Ignoring behavioural efects can be problematic, and cause sub-
stantial inaccuracy, because taxpayers do modify their behaviour depend-
ing on the particular legal framework and opportunities available to them.
First and foremost, it is unlikely that a repeal of the measured tax expend-
iture would result in tax gains equal to the measured revenue foregone,
resulting in an overestimate of the iscal cost of the tax expenditure. his is
because taxpayers will most likely aim to minimise the impact the removal
of the tax expenditure has on their ater-tax income. Ultimately, taxpayer
behaviour will be afected by the removal of a tax expenditure. he removal
of a tax expenditure itself is also likely to be politically diicult in many
situations, leading to the ‘grandfathering’ of existing concessions and
incentives. Again, this distorts the true value of the revenue foregone in an
upward direction and gives a false impression as to the possible increases
in revenue from the abolition of the tax expenditure.
Distortion of the revenue foregone may not only be overestimated but,
in fact, also be underestimated by not factoring in taxpayer circumstances.
Where progressive tax rates apply and the removal of, for example, certain
allowances and exemptions push a taxpayer into a higher bracket, more
revenue will be raised than predicted. Further complicating the measure-
ment of tax expenditures under this approach is the fact that the removal
of certain allowances and exemptions, which reduce a taxpayer’s taxable
income, once removed, will increase taxable income, afecting other tax
expenditures such as those credits that are means tested.
Why then is the revenue foregone method the predominant approach?
Apart from its relative simplicity to calculate, this method is comparatively
89
hrough the Tax System: Lessons from Developed and Transition Economies (2004).
simple for a government department to apply and for observers to under-

stand and analyse, and provides a parallel between the tax expenditure
budget and direct expenditure budget. For example, the cost of a tax
expenditure which is a deduction for philanthropic activities is simply
the amount of tax foregone on that deduction. his varies from the other
two methods, which whilst potentially more accurate, lead to added com-
plexity in terms of data collection, underlying assumptions and the actual
calculations.
3.7.2 he revenue gain approach

he revenue gain approach measures the anticipated revenue gain that
would arise from abolition of a speciic tax expenditure. Where this
approach varies from the revenue foregone approach is that it is an ex ante
approach which aims to take into account likely taxpayer responses, or
behaviour, to that hypothetical abolition of the tax expenditure. It also
aims to take into account other ‘second order’ efects such as the inter-
action of tax laws with other laws. As such, it is a much more precise esti-
mate of the revenue that would be collected if the tax expenditure was
abolished.
he revenue gain approach is potentially more accurate than the rev-
enue foregone approach on the basis that estimates determined under this
method are thought to be comparable to estimates of the iscal impact of
traditional budget measures. However, this approach is not without its
problems. First, it requires considerable amounts of high quality data to
estimate the behavioural response of taxpayers to the abolition of a tax
expenditure along with the critical elasticities of supply and demand of
income and goods which are subject to the tax expenditure. For example,
by removing tax expenditures on retirement savings we would need to
take into account whether taxpayers are likely to continue to save that
money or spend it on consumables. And, if taxpayers are likely to continue
to save, will they simply place the money in another tax efective invest-
ment, causing the cost of another tax expenditure to increase?90 he cost
of tax evasion would also need to be factored in as a portion of the tax will
likely not be collected. In addition to these obvious problems, the revenue
gain approach is more time consuming and expensive to apply and also
relies on ad hoc policy assumptions, such as the date of removal, the order
90
Burton and Stewart , Promoting Budget Transparency Through Tax Expenditure
Management: A Report on Country Experience for Civil Society Advocates (2011), 63.
Tax expenditures reports 97
of removal if several tax expenditures are involved and the prospective/

retrospective nature of the removal. here are examples of jurisdictions
which use this method for reporting purposes; however, they are limited.
For example, the United States applies this method and the revenue fore-
gone method in its tax expenditure analysis, whilst Australia applies this
method to a limited number of large tax expenditures. However, it is not
applied by most other countries.91
3.7.3 he outlay equivalence approach

he third approach to the measure of tax expenditures is the outlay
equivalence approach or the equivalent direct expenditure method. his
method measures how much it would cost to deliver the same ater-tax
beneit to the beneiciaries of the tax expenditure by way of a cash grant.
It assumes that the cash grant would be subject to the normal applica-
tion of the existing tax law, including being taxable, and without the tax
expenditure. his method ofers the opportunity of assessing the merits
of alternative means of delivering the government beneit and is argu-
ably the most accurate where the purpose of tax expenditure analysis is
to determine the cost of the tax expenditure and whether it should be
replaced with a direct spending programme. Measuring the actual bene-
it in the hands of the beneiciaries, however, requires high quality data
about the taxpayer.
his method is not widely used by nations, although there are a limited
number of examples which can be found. Denmark, for example, while
primarily relying on the revenue foregone method, does use the outlay
equivalence approach,92 while in Sweden, tax expenditures which would
afect the budget balance if abolished are calculated using both the rev-
enue foregone approach and the outlay equivalence approach.93
3.8 Structure and information contained in tax

expenditures reports
We argue that ‘any’ reporting is better than no reporting of tax expen-
ditures as it will go some way towards addressing the purpose for which
91
Although there are limited cases of nations which take into account some behavioural
efects. For example, Chile and Argentina adjust for total constant expenditure taking
into account lower spending which results in lower VAT collected.
92
Nordic Working Group, Tax Expenditures in the Nordic Countries (2010), 28.
93
Nordic Working Group, Tax Expenditures in the Nordic Countries (2010), 48.
a nation identiies tax expenditures. However, the process of simple
tax expenditure reporting falls short of enabling a nation to critically
scrutinise and engage in the management process of tax expenditures
proposed or adopted. he reporting and iscal costing of tax expendi-
tures, even if done on a regular and eicient basis, only forms part of the
management process. To this extent, we argue that current reporting as
described below tends to only focus on spending reform even where a
nation purports to use tax expenditure analysis for broader purposes.
Genuine tax expenditure management requires the establishment of a
framework for identifying, measuring and critically assessing the merits
of both individual and aggregated tax expenditures on an ex ante and
ex post basis. Further, it needs to be done in a timely manner to ensure
the information is both current and relevant. When making a choice
about the type of information that should be contained in a tax expendi-
tures statement, democratic oversight would generally require key elem-
ents to be included such as publicly accessible analysis of any proposed
tax expenditures, the prima facie reporting of existing tax expenditures
as well as frequent and regular updating of those tax expenditures taking
into account changes to the broader economic and social context within
which the tax expenditures operate. Ideally, there will also be some form
of independent regulatory oversight of the reporting and management
of tax expenditures as we have seen with various national audit oices.
We return to this ‘ideal’ in Chapter 6. For now, we consider the current
approach.
3.8.1 he overall structure

here is no internationally accepted approach for presenting or structur-
ing a tax expenditures report, nor is there a speciic format for reporting
on tax expenditures and it will be a matter of choice based on a nation’s
individual decisions as to how the information is presented. For example,
countries such as the United States, Australia, Germany, Italy and France
all have a legal requirement imposed on its government to produce reports
with certain information contained in those reports, while other coun-
tries such as Canada and the Netherlands are without any statutory obli-
gation but elect to voluntarily produce a report. Most developed countries
produce their tax expenditure reports annually, although some produce
reports every two years and a few produce the reports sporadically. Once
the decision to produce a tax expenditure report is made a nation will
either report on tax expenditures within the budget documents such as
Tax expenditures reports 99
Germany does, link or annex a report to budget documents such as is
the case with the United States and France, provide it as a supplement to
the revenue statement as is the case in the United Kingdom or provide the
statement as a separate government document as is the case in Australia
and Canada. When produced, it is imperative that the information is eas-
ily accessible and useable to allow policy debate in all areas and by all
citizens.
A tax expenditures report will need to be structured in such a way to
provide information which can be analysed in a manner consistent with
a nation’s adopted purpose for reporting on tax expenditures and to meet
the needs of the policymakers. Various possible classiications currently
used include budgetary function, industry, type of tax, beneiciary and
the purpose of the tax expenditure.94 Classiication may even be by region
such as is the case in Italy. Currently, the approach considered ideal,
and one usually adopted, is to organise and report on tax expenditures
according to government function, with revenue foregone estimated for
each tax expenditure and aggregated for each function. his not only
allows a comparison with the direct spending of a government accord-
ing to function but also gives great assistance to a nation’s citizens in an
holistic assessment of a government’s approach to its social and economic
policies. For example, Australia classiies its tax expenditure according
to economic function, type of taxpayer afected and within the relevant
benchmark,95 while Canada reports its tax expenditures by reference to
the type of tax (personal income tax, corporate income tax, VAT) and
then subcategorises those expenditures into budgetary function, industry
or category.96 Other nations, such as the United Kingdom, simply provide
a list of tax expenditures. Given tax expenditures are oten multifunc-
tional and intergovernmental, ideally a range of diferent government
agencies are required to be involved in the process of estimation, moni-
toring and control,97 although, oten pragmatic constraints mean that the
less than ‘ideal’ involvement is achieved.
Longitudinal reports of this information are adopted by many nations
and oten both ex ante estimates and ex post projections, which allow for
trends in tax expenditure costs to be observed, are provided (although,
94
95
See Australia, 2011 Tax Expenditures Statement (2012).
96
Canada, Tax Expenditures and Evaluations (2011).
97
Villela et al., Tax Expenditure Budgets: Concepts and Challenges for Implementation
(2010), 16.
variations to the tax law and/or benchmark over the time period should
be clearly stated). For example, in the United States, the Joint Committee
on Taxation provides current and projected estimates for a ive-year
period, while the most recent South African report provides historical
(three years) and current data (one year) for a four-year period. A tax
expenditure report will oten also aggregate the total of all estimated tax
expenditures to allow for a rudimentary estimate of the total revenue fore-
gone due to the reported tax expenditures. Again, longitudinal reporting
allows trends to be observed.
3.8.2 Information on each tax expenditure

Comprehensive tax expenditure management requires more than the
simple identiication and iscal cost of each tax expenditure. We recog-
nise that the information contained in the tax expenditure report on
each tax expenditure is likely to be the most afected by resource and data
constraints.98 Yet, depending on the purpose of a nation’s tax expend-
iture reporting, that comprehensive information on each individual tax
expenditure can be the most valuable in terms of tax expenditure analysis.
It has already been discussed that the estimated cost of each tax expend-
iture along with the reliability of the estimate and quality of the data is
usually provided in a report. In addition, the report will at a minimum
ideally identify the source of the tax expenditure, for example, whether
it is contained in domestic tax legislation, is an administrative practice
or is contained in a tax treaty. he type of tax expenditure, for example,
whether it is delivered by way of credit, exemption or additional deduc-
tion, should also be stated.
A comprehensive tax expenditure report which allows for full trans-
parency where the purpose of tax expenditure reporting extends beyond
identifying iscal costs and a direct comparison with direct expenditures
to being a signiicant social and policy instrument, will also contain rele-
vant qualitative and quantitative data which analyses the impact of each
tax expenditure. A democratic government aspiring to achieve account-
ability and transparency with respect to its policy measures should aim to
provide this information to aford the greatest opportunity for debate. We
return to this in the next part of the book.
98
Roin argues that incomplete information should be acknowledged rather than treated as
perfect as this can be a critical factor in decision making. See Roin, ‘Truth in Government:
Beyond the Tax Expenditure Budget’ (2003), 640.
Conclusion 101
3.9 Conclusion
‘Best practice’ enabling tax expenditure analysis and management will
consist of key elements which make up the process. he identiication and
reporting of tax expenditures will provide the foundation for scrutiny;
however, critical analysis of the advantages and disadvantages of each tax
expenditure contrasted with other means of delivery such as regulation of
direct spending along with the policy justiication for each expenditure
is also a key element to management. he regular review of tax expendi-
tures, along with some form of assurance by an independent agency (nor-
mally a national audit agency) of the integrity of the process by which tax
expenditures are reported and managed is also an important component
to tax expenditure analysis. Regular evaluation requires reporting to go
beyond the iscal costs to assess whether the tax expenditure is serving
a valid government purpose and relects a government spending prior-
ity, whether a tax expenditure is the best government policy instrument
and what is the most appropriate design within the tax regime.99 Where
comprehensive tax expenditure reporting is correctly designed it will be
possible to assess whether the beneits are distributed fairly, the intended
targets are in fact the recipients, whether there are any unintended dis-
torting efects, and whether administrative and compliance costs are rea-
sonable, to name a few.
Despite the ongoing academic criticisms aimed at tax expenditure ana-
lysis, along with the controversies surrounding its various key elements,
it has been demonstrated in this chapter that nations are not abandoning
tax expenditure reporting and analysis, but rather are embracing the con-
cept and adapting its usages in line with their adopted purposes. Citizens
of democratic societies are also demanding greater transparency of their
nation’s iscal policy, of which tax expenditures are signiicant. However,
this chapter has also highlighted that, contrary to Surrey’s purported elit-
ist model of tax expenditures, there are no ‘right answers’ in the applica-
tion of tax expenditure reporting and analysis when applied in practice
and at a national level; a claim that is supported by international studies.
We argue for the continued growth of tax expenditure reporting and dis-
agree with the critics who call for the abandonment of tax expenditure
analysis and management on the basis that it is controversial. Currently,
99
See Brooks, ‘he Under-Appreciated Implications of the Tax Expenditure Concept’ (2009)
and Sadiq, ‘Commentary: he Under-Appreciated Implications of the Tax Expenditure
Concept’ (2009), 264.
however, much of tax expenditure management, whilst it varies between

nations, is still implicitly underpinned by substantive and bureaucratic
rationality, both of which have been conceived in terms of elitist political
philosophies. We argue in the remaining chapters that tax expenditure
management is fundamentally a moral question, and that there are no
‘right’ answers to the questions. Rather we should aim for a system which
allows for objective moral tax expenditure management and that ought
to be grounded upon particular moral stances as to the nature of democ-
racy and the nature of the society that we think a democratic community
ought to aspire to.
he thesis of the book is that the irresolvable controversies surround-
ing the tax expenditure concept require consideration of how such deep
moral controversies should be resolved in a modern democratic setting.
We now move to Chapter 4 to highlight the practical signiicance of tax
expenditures from the perspective of competing imperatives confronting
a democratic government and explore the debates regarding each of these
imperatives – competing conceptions of distributive justice for example –
as well as the diiculty of making i ne quantitative measurements of
these impacts.
4
he practical signiicance of tax expenditures
4.1 Introduction
In the preceding chapters we examined the conceptual and empirical dif-
iculties with respect to tax expenditure management – the deinition of
the tax benchmark, rival measures of the quantum of tax expenditures,
the speciication of a tax expenditure reporting standard and also the
process by which tax expenditures ought to be managed.
In response to these diiculties it has been suggested that tax expend-
iture management is a failed enterprise1 because it does not meet the
standards of bureaucratic rationality which ought to be expected of con-
temporary public policy in a ‘modern’ democracy.2 From some quar-
ters it is suggested that the concept of a ‘tax expenditure’ is politicised,3
the measurement of tax expenditures too imprecise4 and the gathering
of tax expenditure data imposes dead weight costs by creating add-
itional tax compliance/administration costs.5 From other quarters it
is argued that the tax expenditure literature exhibits ignorance of the
broader signiicance of tax expenditures to an holistic appraisal of tax
policy principles,6 distributive justice7 or the functional allocation of
1
Zelinsky, for example, contends that tax expenditure management has enjoyed only mod-
est success if it is measured against what Zelinsky states are the ambitions of tax expend-
iture managers to abolish all tax expenditures: Zelinsky, ‘Do Tax Expenditures Create
Framing Efects? Volunteer Fireighters, Property Tax Exemptions, and the Paradox of
Tax Expenditure Analysis’ (2005), n3.
2
Bartlett, ‘he End of Tax Expenditures as We Know hem?’ (2001).
3
Bittker, A “Comprehensive Tax Base” as a Goal of Tax Reform’ (1967); Bruce, ‘Tax
Expenditures and Government Policy: An Introduction to the Conference Volume’
(1988), 11.
4
Wildavsky, ‘Keeping Kosher: he Epistemology of Tax Expenditures’ (1985), 416–17, 420–2.
5
Australia, Preparation of the Tax Expenditures Statement (2007).
6
MacNevin, ‘Issues in Income Tax Accountability: An Analysis of the Rental Housing
Sector’ (1994), 682.
7
So tax expenditures ought to be subsumed within a broader tax/transfer analysis which
focuses upon social outcomes rather than the components by which those outcomes are
achieved.
103
104 The practical significance of tax expenditures
government programmes.8 From any of these perspectives it has been

argued that proposals to engage in tax expenditure management ought
to be abandoned.
he objectives of this chapter are twofold. Firstly, we survey the prac-
tical signiicance of tax expenditures from the perspective of competing
imperatives confronting a democratic government. he object of this
survey is to validate the desirability of the tax expenditure management
project. he thesis of this chapter is that tax expenditure management
cannot be abandoned because tax expenditures are too signiicant in
terms of their impacts upon economic eiciency, distributive justice, the
iscal position of governments (federal and otherwise), the constitutional
framework of countries, public administration in general, tax adminis-
tration in particular and the legitimacy of governments.
he second object of this chapter is to acknowledge that demonstrating
this thesis is not as straightforward as some commentators seem to sug-
gest. here are, for example, competing conceptions of distributive just-
ice. Further, there are considerable diiculties in making ine quantitative
measurements of impacts upon perhaps desirable social objectives such
as eiciency and justice. Why would we want to make our thesis more
challenging by acknowledging such diiculties? he purpose for doing
this is to demonstrate that the cases both for and against tax expend-
iture management are inevitably grounded upon acceptance of contin-
gent propositions. Acknowledgement of the contingent aspects of the tax
expenditure concept (Chapter 2), empirical aspects of tax expenditure
management (Chapter 3) and of the contingent aspects of the cases for
and against tax expenditure management in this chapter comprises the
foundation for our consideration of the politics of tax expenditure man-
agement in Chapter 5.
4.2 Where to start? Holism, atomism and tax

expenditure management
his chapter takes tax expenditures as a discrete subject of study with the
object of identifying their key practical ramiications. To some observers,
this atomistic concentration upon parts of a social complex is irrelevant
and possibly misleading.9 hose observers say that concentrating upon
8
Weisbach and Nussim, ‘he Integration of Tax and Spending Programs’ (2004), arguing
that tax expenditure analysis ought to be subsumed into a whole of government allocation
of functional responsibilities across government agencies.
9
See, for example, MacNevin, ‘Issues in Income Tax Accountability: An Analysis of the
Rental Housing Sector’ (1994), 681–2.
Holism, atomism and tax expenditure management 105
tax expenditures in isolation is misconceived because tax expenditures

are but one part of a public policy whole and so tax expenditures of them-
selves do not produce particular social outcomes. It is tax expenditures
in combination with myriad contextual factors which produce those
outcomes.
As in the physical sciences, the quest for ‘the theory of everything’ in
the ‘social sciences’ confronts this epistemological tension – is knowledge
best obtained through eforts to understand an entire social system or
can ‘the theory of everything’ be constructed atomistically? his tension
between atomism and holism leaves its trace across the tax expenditure
literature.
he discipline of economics focuses upon how individuals and soci-
eties allocate scarce resources in diferent contexts. If the role of govern-
ment is to ensure that the best social outcomes are achieved in a world
of scarcity, concentrating upon tax expenditures in isolation from what
is happening in the rest of the social world could be counterproductive.
he economic theory of the second best relects this tension between hol-
ism and atomism. Taking economic eiciency conceived in terms of max-
imised aggregate social wealth as its starting point,10 the theory of the
second best suggests that modifying or removing tax expenditures which
appear to impede eiciency when viewed in isolation might not enhance
aggregate eiciency.11 Some therefore suggest that comprehensive social
welfare analysis would deliver the desired holistic analysis, although it
is accepted that such analysis is beset with empirical diiculties because
of the diiculty of identifying and computing all efects of a particular
phenomenon.12
Others embrace diferent perspectives that inform their respective
understandings of the merits of the tax expenditure concept and/or its
place in social and political theory. Weisbach and Nussim argue that pur-
suit of a theoretically pure income tax is misconceived from the holistic
standpoint of organisational theory.13 hey argue that ‘tax expenditures’
which most efectively deliver a desired public good should not be classi-
ied as tax expenditures (on the basis that classiication as a tax expend-
iture would jeopardise continuation of that measure).
10
Diferent concepts of eiciency are noted below in section 4.3.3.
11
For discussion of this point see Zelinsky, ‘Eiciency and Income Taxes: he Rehabilitation
of Tax Incentives’ (1986), at 996f.
12
OECD, Inventory of Estimated Budgetary Support and Tax Expenditures for Fossil Fuels
(2011), 38–9. More generally, see McGarity, Reinventing Rationality (1991).
13
Weisbach and Nussim, ‘he Integration of Taxation and Spending Programs’ (2004); see
also Weisbach, ‘Tax Expenditures, Principal-Agent Problems, and Redundancy’ (2006).
Similarly, from the perspective of distributive justice, an holistic

approach suggests that the spending side of government should not be
ignored.14 If there is to be one tax expenditure benchmark against which
all taxes are to be measured, it would be misconceived to specify a dis-
tributively fair tax mechanism (such as, say, an income tax) if it would be
more equitable and more eicient to fund redistributive spending with a
regressive tax. With respect to tax expenditures speciically, Steuerle sug-
gests that distributive analysis of a prospective tax expenditure requires
consideration of how the burden of the reduced tax revenue is to be
borne – by higher taxes on some or by spending cuts, for example.15 Such
an holistic analysis can only be speculative given that hypothecation of
tax revenues to speciic spending measures is rare. Typically, hypotheca-
tion is rejected upon the grounds that the government of the day should
exercise complete budgetary control in setting and meeting the policy
imperatives of the time and (in a democracy) be accountable for its deci-
sions to the electors of the day.16
From an holistic pragmatic perspective, it might be argued that eforts
to install a purer tax base guided by eiciency or fairness might be coun-
terproductive because the current tax system, with all of its laws, is as
good as it can be in the current political environment. It is possible that
the most equitable social outcome is achieved by maintaining a theoret-
ically impure income tax riddled with tax expenditures because those tax
expenditures are an essential component of a tax base in an open dem-
ocracy where tradeofs must be brokered in order to procure legislative
action.17 From this standpoint, abolition of tax expenditures might be
self-defeating if the objective is to install a ‘pure’ tax base, as elimination
of tax expenditures could jeopardise the political compact upon which
the ‘impure’ tax was founded. Better to have an imperfect income tax
than no income tax at all, it might be said.
We agree that an holistic assessment of social outcomes with respect to
such measures as wealth distribution should be the aspiration of social pol-
icy development, rather than an atomistic assessment of particular public
policies being the endpoint of public policy analysis. However, as we noted
in section 1.4.4, the costs of obtaining and analysing such information
14
Kaplow, ‘Taxation and Redistribution: Some Clariications’ (2007); Steuerle, ‘Can the
Progressivity of Tax Changes Be Measured in Isolation?’ (2003).
15
Steuerle, ‘Can the Progressivity of Tax Changes Be Measured in Isolation?’ (2003).
16
See discussion of this issue in McGee, he Philosophy of Taxation and Public Finance
(2004), ch. 18.
17
United States of America, he Moment of Truth (2010), 30.
The economic significance of tax expenditures 107
most likely would be considerable while the measurement of the beneits

will be problematic. Moreover in our view the prospect of achieving such
a state of comprehensive knowledge is slim, at least for the foreseeable
future. In this context, we consider that the absence of such comprehen-
sive public policy assessment does not justify acquiescence with respect
to the current public policy institutions and/or policies. he absence of
perfect public policy knowledge means that there are perhaps unknown
and also unquantiiable risks associated with any proposed public pol-
icy action – whether that action constitutes the continuation of existing
tax expenditures, the insertion of new tax expenditures or other actions
with respect to tax expenditures. Paralysing tax expenditure analysis until
such time that perfect information becomes available does not necessarily
achieve the best social outcome in managing those risks. In section 1.4.4
we argued that tax expenditures were a special case warranting investment
of social resources in order to identify and analyse their impact upon vari-
ous social imperatives. One reason for undertaking the relective review of
the consequences of tax expenditures in this chapter is to suggest that we
should proceed with tax expenditure management while acknowledging
that the epistemological foundations for our action are insecure.
4.3 he economic signiicance of tax expenditures

For our purposes ‘economics’ is the study of what constitutes the wealth
of a community, how that wealth is generated and how that wealth is
distributed.
4.3.1 Well-being indicators

he gross domestic product of a nation is commonly relied upon as the
primary measure of a community’s aggregate social welfare. However, the
limitations of concentrating upon the market value of goods and services
produced over a speciied period are widely acknowledged and need not
be revisited here.18
In recognition of these limitations, some eforts have been directed to
broadening the range of welfare indicators with a view to more accurately
measuring social welfare. Factors which might be taken into account in
this broader welfare accounting include information regarding a country’s
18
Stiglitz et al., Report by the Commission on the Measurement of Economic Performance
and Social Progress (2009).
political institutions and the extent to which those institutions promote

civic engagement by facilitating open and informed deliberation upon
public policy.19 his factor is related to but discrete from the well-being
indicator of good governance, which is concerned with ensuring that a
community’s laws are born of a robust process.20
While it is only part of this broader canvas of political engagement,
tax expenditure management is nevertheless a iscally and politically
signiicant aspect. A government’s budget is its most important policy
document because a budget relects the tradeofs made by government
in the course of raising revenue and spending.21 As a crucial feature of a
budget, it is reasonable to propose, in the absence of convincing empirical
evidence, that a government’s approach to tax expenditure management
inluences perceptions of government openness more generally. here is
a recurrent theme in a segment of the tax expenditure literature that at
least some tax expenditures are one means by which politicians confer
beneits upon ‘special’ interests,22 oten in a less than transparent manner
and sometimes corruptly.23 Looking beyond the direct causes, Howard
argues that tax expenditures may be both the cause of, and the efect of,
low trust in government. He contends that there is a vicious cycle by which
low trust in government prompts the use of tax expenditures which, in
turn, prompts less trust in government. Although Howard suggests that
‘welfare’ tax expenditures are more openly described as such by govern-
ments in recent decades,24 he concludes by speculating that low trust in
government prompts politicians to shield spending behind the veil of
tax spending. his obfuscation, he suggests, is the way politicians man-
age conlicting calls for less welfare, more support for the poor and lower
taxes in a period when trust in government is low. One example of this
discursive alignment manifested in the form of a tax expenditure is the
19
Legatum Institute, he 2010 Legatum Prosperity Index (2010); OECD, How’s Life? (2011),
ch. 9.
20
As we note in Chapter 5, both of these well-being indicators spring from a contingent
conception of the political good. he i rst is framed in terms of deliberative democracy
and the second in terms of public rationality.
21
See, for example, OECD, Best Practices for Budget Transparency (2002), 3; Levi, Of
Rule and Revenue (1988), where Margaret Levi argues that the ability of a government
to impose taxes is a signiicant determinant of public policy; see also Brennan and
Buchanan, he Power to Tax (1980), ch. 2. See also Brennan, ‘Public Choice and Taxation:
Leviathan Ater Twenty Years’ (1997).
22
United States of America, he Moment of Truth (2010), 30.
23
Johnston, Free Lunch (2007); Perfectly Legal (2003).
24
Howard, ‘he Politics of Tax Expenditures in Wealthy Democracies’ (2011), 7:1–7:2.
earned income tax credit, which singles out the ‘deserving’ poor for spe-
cial tax favour by apparently lowering taxes on the earned income of some
low paid workers. However, this tax expenditure also beneits employers
to unknown but presumably varying degrees. Nevertheless, business lob-
byists are most particular in ensuring that business tax expenditures do
not come to be perceived as ‘business welfare’.25
It is reasonable to hypothesise that there is a link between tax expen-
ditures and the civic engagement aspect of social well-being. However,
at this stage the measurement of civic engagement as part of social well-
being is nascent.26 he present limitations of the concept of civic engage-
ment and also the limitations of the data regarding civic engagement
mean that quantifying the signiicance of tax expenditures to this aspect
of social well-being is not possible as yet.
4.3.2 Distributive justice

Distributive justice refers to the moral norms dictating the extent to
which a community has an obligation to ensure a fair distribution of the
community’s resources across its ‘citizens’.
Although it is possible that a tax expenditure will leave the distribu-
tion of resources unchanged, most if not all tax expenditures will have
the efect of redistributing some part of a community’s resources from
one person to another, or from one group to another. Indeed, many tax
expenditures are introduced for the express purpose of redistribut-
ing wealth within a community. Identifying and quantifying the dis-
tributive impact of tax expenditures entails consideration of several
questions:
1. To what extent ought distributive justice be recognised as an inde-
pendent public policy norm?
2. What deinition of distributive justice should underpin the distribu-
tive analysis of the tax expenditure?
3. What beneits should be considered when assessing the distributive
justice of a particular tax expenditure?
4. Who beneits from the tax rule and to what extent?
5. Who bears any burdens imposed as a result of the tax expenditure?
6. When do the beneits and/or burdens identiied in questions #4 and
#5 arise? hat is, to what extent should net measurements of taxes and
25
Australia, Preparation of the Tax Expenditures Statement (2007), 36–7.
26
OECD, How’s Life? (2011), ch. 9.
transfers within an entity’s lifetime be incorporated into distributional

tables?
7. How should those who receive beneits or bear the burdens of a tax
expenditure be categorised? For example, by income, gender, religion,
race/ethnicity, tax residence, age/generation or some other classiica-
tion such as a socio-economic index score?27
8. What would be the net change to aggregate tax collections if the par-
ticular tax expenditure were abolished?
9. Does the allowance of the tax expenditure create a second tax expend-
iture, being the exclusion of the beneit of the tax expenditure from the
tax base?28
4.3.2.1
he independent status of the distributive
justice norm
he extent to which distributive justice stands as an independent moral
norm is the subject of considerable debate. Should distributive justice be
subordinate to pragmatic considerations such as international comity and
the maximisation of wealth, or should it be recognised as an independent
norm which must be weighed against such competing imperatives?
Surrey and McDaniel do not expressly consider the tradeof between
distributive justice and pragmatic imperatives, although they impli-
citly accept that the Schanz–Haig–Simons grounded tax expenditure
benchmark must be relaxed for the purposes of categorising tax rules
with respect to cross-border transactions. his relaxation, they imply,
is necessary to take account of pragmatic factors such as capital lows
across the country’s borders and historical factors.29 Taking up this prag-
matic approach, Avi-Yonah argues for a beneit based benchmark with
respect to non-residents, and an equity based measure for residents.30 It is
not clear that this advances the consideration of distributive justice that
much, as the concept of ‘beneit’ is rather luid, such that ‘beneit’ could be
equated with ‘economic income’.31 Moreover, if distributive justice were
taken to require beneit taxation, the empirical determination of the value
of beneits conferred upon non-residents would be problematic.
27
Boadway. ‘he Annual Tax Expenditure Accounts: A Critique’ (2007), 127.
28
For discussion of this point see Edrey and Abrams, ‘Equitable Implementation of Tax
Expenditures’ (1989).
29
30
Avi-Yonah, International Tax as International Law (2007), 1.
31
Cooper, for example, argues that beneit can be dei ned to equate with the ability of the
taxpayer (i.e. economic income): Cooper, ‘he Beneit heory of Taxation’ (1994). For
As we note below, some deinitions of ‘eiciency’ such as Pareto opti-

mality incorporate a distributive justice element.32 We agree with Cohen
that Pareto based derivations of eiciency are a pale shadow of what dis-
tributive justice ought to be taken to require.33 From this standpoint, we
take the view that an egalitarian interpretation of distributive justice ought
to stand independent from the moral norm of eiciency so that tradeofs
between the two norms might be made more transparent.34 Constructed
in this way, the question of distributive justice becomes one of ‘how does
this tax expenditure afect the equal distribution of resources?’ his is
markedly diferent from the Pareto question ‘does this tax expenditure
leave all citizens better of, albeit to varying degrees?’
What beneits/burdens should be the subject of

4.3.2.2
distributional analysis?
When considering distributive justice, a fundamental question is ‘just,
with respect to what?’ As a community we could consider the distribu-
tion of resources such as income or wealth. Or we could consider whether
social policy achieves a just distribution of capabilities. Or we could con-
sider whether social policy procures just outcomes.35
he literature regarding distributive analysis of tax expenditures typic-
ally focuses upon cash beneits provided by tax expenditures and appar-
ently ignores non-pecuniary beneits/burdens which may be imposed
upon others as a result of entities taking up the tax expenditure incentive.
hat is, the literature concentrates upon the notional cash transfer cre-
ated by the tax expenditure and ignores non-cash distributions of com-
munal resources. his concentration upon cash transfers runs deep in
the economic literature. Kaplow, for example, argues that a proportional
‘tax’ (including a negative tax component with a lat phaseout rate) can
be the most redistributive tax possible, a proposition founded upon the
twin assumptions that distributive justice entails a concentration upon
cash transfers and that a tax that let each person with the same amount
of cash would be perfectly redistributive.36 At issue is the extent to which a
community (at any level, including the level of a global community) bears
further discussion of the beneit principle, see Duf, ‘Private Property and Tax Policy in a
Libertarian World: A Critical Review’ (2005).
32
Cohen argues that this subordinates the concept of justice, a fate from which he argues it
should be rescued: Cohen, Rescuing Justice and Equality (2008).
33
Cohen, Rescuing Justice and Equality (2008).
34
35
For discussion of these alternate approaches see Sen, he Idea of Justice (2009), ch. 11.
36
Kaplow, ‘Taxation and Redistribution: Some Clariications’ (2007).
a responsibility to ensure that individuals have a just share of the relevant

community’s aggregate welfare. Selection of the subject of distributive
justice is interrelated with this subject as, for example, a concentration
upon an equal distribution of a community’s income will not allow for the
disparate needs of community members.
his elision of non-monetary beneits and burdens from distributive
analysis is a signiicant omission. It would be possible, for example, for the
cash beneit of a tax expenditure to be distributed fairly (according to the
particular deinition of ‘fairness’ adopted) whilst the burdens might be
borne by a particular group. Environmentally harmful tax expenditures
and also trade subsidies are cases where burdens in the form of environ-
mental degradation or exclusion from markets might be borne by third
parties without any compensation.37 Moreover, these burdens might
exacerbate existing disadvantage, as those with the least means may have
little capacity to mitigate the additional burden38 and/or lack the capacity
to win ofsetting beneits delivered through public spending. However, if
the distributive analysis of tax expenditures is to be broadened beyond a
concentration upon cash distributions, the analysis of tax expenditures
would become vastly more data intensive and challenging, as tracing both
cash and non-cash impacts of a tax expenditure measure at both domestic
and international levels would be extremely diicult.
Consideration of the distribution of tax expenditure beneits was
prompted by Surrey’s identiication of the upside-down efect.39 hat
efect is illustrated in the following example:
Example – illustration of the ‘upside-down’

efect of a tax expenditure
Assume that the government allows a deduction of $100 for certain private expenses
of a taxpayer, which would qualify as a tax expenditure. Taxpayer A pays zero tax as
he has a low income. Taxpayer B has medium income and has a 20 per cent tax rate
with respect to her marginal income (i.e. the last unit of income received by the tax-
payer). Taxpayer C has a high income and has a marginal tax rate of 40 per cent.
37
Conversely, it is possible that the beneit of an environmentally beneicial tax expenditure
is broadly distributed, reaching beyond those with any direct or (otherwise) indirect con-
tact with the direct recipient. For example, takeup of a tax expenditure for environmental
clean-up costs might spare geographically remote persons from the burdens of pollution
that otherwise would exist (i.e. groundwater contamination, atmospheric pollution).
38
World Bank, Gender and Environment (2010).
39
Surrey, Pathways to Tax Reform: he Concept of Tax Expenditures (1973), ch. 3.
Taxpayer A Taxpayer B Taxpayer C

(low income) (medium income) (high income)
Marginal tax rate (tax 0% 20% 40%

rate imposed on last
$100 of income)
Deduction allowed $100 $100 $100
Tax saved as a result of $0 $20 $40
the deduction
Efective subsidy $0 $20 $40
received
In response to the apparent inequity of awarding a greater proportion of

tax expenditure beneits to those with higher incomes, the refundable tax
credit has been proposed as a more equitable mechanism for delivering
tax expenditures. Taking the prior example, the operation of a refundable
tax credit is illustrated in the following example:
Example – refundable tax credit

Looking again at Example 1, above, instead of a deduction of $100, the govern-
ment allows a refundable tax credit of $40. In that event, Taxpayer A, who owes no
income tax before the tax credit is considered, would receive a cash payment of $40.
Taxpayer B would receive a $20 credit against their tax liability and a cash payment
of $20. Taxpayer C would receive a $40 credit against their income tax liability. As a
result, the refundable tax credit is worth an equal amount of cash to each taxpayer,
in spite of their diferent levels of income.
Taxpayer A Taxpayer B Taxpayer C

(low income) (medium income) (high income)
Marginal tax rate (tax 0% 20% 40%

rate imposed on last
$100 of income)
Refundable tax credit $40 $40 $40
allowed
Tax saved as a result of $0 $20 $40
the deduction
Cash payment received $40 $20 $0
All else being equal, we agree that refundable tax credits can be the most
equitable mechanism for delivering a tax expenditure in a progressive
taxation system.40 However, there is a danger in too readily assuming that
a refundable tax credit is distributively fair and thereby avoiding a wider
inquiry regarding non-cash beneits and burdens of the tax expenditure.
Further, even if we were to focus upon cash beneits, a theoretically fair
refundable tax credit can break down in practice when administration of
tax credit systems is imperfect, a matter to which we return later in this
chapter.41
Moreover, the costs of administering the refundable tax credit mech-
anism can be signiicant and ineicient, particularly where many are enti-
tled to a refund such that there is a churn of gathering revenue only to
refund it.42 his may be diicult for many tax administrations to imple-
ment and might even be counterproductive in terms of distributive fair-
ness. For example, implementing a refundable credit mechanism might
absorb a signiicant proportion of extremely limited administrative
resources where those resources might otherwise have been applied to
gathering considerable sums of revenue which could have been expended
upon targeted social welfare with greater impact in terms of distributive
justice.
Identifying beneiciaries and those sufering burdens

4.3.2.3
of tax expenditures
he distributive efect of tax expenditures is diicult to assess because
the measurement of tax expenditures is problematic, as is the incidence
of any tax expenditure.43 We saw in Chapter 3 that the measurement of
tax expenditures is problematic because imperfect information generally
means that strong assumptions have to be made about the counterfactual
scenario to the existence of a particular tax expenditure.
Informational limitations he lack of information arises for four rea-

sons. he irst is that the particular tax characteristics of each tax entity
may be diicult to ascertain. For example, where a tax system relies
upon self-reporting by an entity of its tax liability to government, it may
40
Edrey and Abrams, ‘Equitable Implementation of Tax Expenditures’ (1989).
41
See section 4.6.
42
Other issues include the risk of fraud and the diversion of administrative resources to
management of this risk: United States of America, 2011 Report of the National Taxpayer
Advocate (2011), 24.
43
See the discussion of incidence below in this section.
be diicult for government to ascertain which entities actually receive

the beneit of tax expenditures. Second, even if those beneiciaries can
be ascertained, classiication of the beneiciary for the purposes of dis-
tributive analysis may be problematic. he distributive efect of a tax
expenditure could be assessed having regard to the income/wealth of the
recipients. However, in such circumstances, relaxed tax reporting require-
ments, tax exemptions, tax underreporting and/or presumptive taxation
rules may mean that the actual economic income/wealth of a recipient
cannot be ascertained.44 hird, there is uncertainty regarding the behav-
ioural response to addition/removal of a tax expenditure. his matter was
noted in Chapter 3, where the behavioural response to abolition of a tax
expenditure was seen to be crucial to quantifying the iscal cost of a tax
expenditure. Fourth, where the tax expenditure provides a timing advan-
tage, there is uncertainty regarding the application of (perhaps altered)
taxation law in future taxation periods.
To some degree these informational shortcomings might be overcome
by obtaining information that is available to those afected (beneicially
and/or detrimentally) by a tax expenditure. However, requiring taxpay-
ers to provide more information oten increases the compliance costs of
the tax.45 Although the tax compliance cost literature has emphasised
the deadweight and regressive characteristics of these costs,46 less or no
attention has been paid to the countervailing public beneit of public
policy grounded upon better information. Indeed, concerns regarding
tax compliance costs have seen some tax administrators accept that the
minimisation of tax compliance costs is an absolute imperative rather
than one susceptible to trade of against other imperatives such as public
administration.47
Conceptualising tax burden Leaving these informational diiculties

to one side, the appropriate measure of burden also has proved con-
troversial, with some arguing that burden increases if the amount of
tax paid by a particular category of taxpayers increases. For example,
a reduction in the tax rate applied to the distortive taxation of realised
(rather than accrued) capital gains/losses could mean that those with
capital wealth will unlock some of that wealth and thereby trigger a
44
Vazquez et al., “Sizing” the Problem of the Hard-to-Tax’ (2004); section 2.5.
45
Noted, for example, in MacNevin, ‘Issues in Income Tax Accountability: An Analysis of
the Rental Housing Sector’ (1994), 683.
46
Sandford et al., Administrative and Compliance Costs of Taxation (1989).
47
tax liability.48 Advocates of capital gains tax reductions rely upon this
‘dynamic’ modelling of response to iscal change in arguing that such
reductions therefore increase the tax burden of the wealthy.
his dynamic approach downplays the redistributive purpose com-
monly taken to underpin the adoption of an income tax in the irst place.
A ‘static’ measure of the cost of a tax expenditure, by contrast, assumes
that there is no behavioural response to the tax change. On this basis, a
static approach measures the diference between the amount of tax that
would be paid on a ‘business as usual’ basis without the change to the law
and compares that amount with the amount of tax that would be paid
regarding the same transactions again on a ‘business as usual’ basis but
assuming the elimination of the tax expenditure.
Incidence of beneits and burdens he beneiciary of a tax expenditure

is diicult to identify for the same reasons that the incidence of a tax is
problematic.
It cannot be assumed that all of those legally entitled to the beneit of
a tax expenditure procure their full beneit as the administration of the
tax expenditure oten will not perfectly identify eligible entities.49 hose
able to aford high quality tax advice may be better placed to take advan-
tage of tax expenditures. his may be because tax advice is obtained pro-
spectively by those with higher incomes, such that a taxpayer’s afairs
might be structured to take best advantage of tax expenditures by what
McBarnett calls ‘creative compliance’.50 Further, there may be little or no
marginal cost for those with higher incomes to keep records which val-
idate their claims for tax expenditures, whereas the compliance costs for
those with lower incomes may be prohibitive such that they are not in a
position to obtain the beneit of tax expenditures to which they are for-
mally entitled.
Further, the legal entitlement to a tax expenditure might not identify
all of those who beneit from the tax expenditure, as beneits and/or bur-
dens might pass to other entities through market and non-market trans-
actions. Where one taxpayer has dependants, it is reasonable to expect
that some portion of the tax burden nominally allocated to the taxpayer
in fact might be borne by each of the dependants respectively. Likewise,
48
Graetz, ‘Paint-by-Numbers Tax Lawmaking’ (1995), 668–72.
49
he signiicance for the legitimacy of the tax administration of this failure to deliver the
tax spending to all eligible entities is considered below in section 4.6.
50
McBarnett, ‘When Compliance Is Not the Solution but the Problem: From Changes in
Law to Changes in Attitude’ (2001).
the tax burden nominally borne by one entity may be passed to another
entity through market mechanisms – a trader may increase the price of
goods to pass the tax on to customers while an employer might reduce the
salary/wage paid to staf in response to a tax liability directly borne by the
employer.
For example, the earned income tax credit in the United States of
America generally is portrayed as an instance of welfare spending.51
However, the extent to which this credit is an indirect subsidy to employers
of low-skilled workers, or indeed the customers acquiring goods/services
from those employers, is not clear. What evidence there is suggests that a
considerable part of the beneit lows to the employer who lowers wages of
eligible employees. Further, there is the possibility that employers might
lower the wages of ineligible, similarly skilled employees.52 his suggests
that care should be taken when assessing the distributive impact of tax
expenditures which supericially have a foundation in relieving poverty,
but which may beneit entities other than the nominal recipients.53
he deinition of distributive justice and

4.3.2.4
international justice
he deinition of the relevant ‘community’ is a crucial aspect of the sim-
ple deinition of distributive justice that we have adopted, as there is no
logical reason for restricting the concept of ‘community’ to those living
within discrete nation states at a particular time rather than, say, regional
or global communities.
Typically the efects of tax expenditures with respect to distributive just-
ice are framed in terms of identifying and categorising those who beneit
51
Howard, he Hidden Welfare State (1997), 64; although it is also portrayed as an ei-
ciency enhancing measure: Batchelder et al., ‘Eiciency and Tax Incentives: he Case
for Refundable Tax Credits’ (2006). he latter characterisation appears to be based
upon the view that promoting the positive externality of higher workforce participation
would generate positive externalities which are not igured into the prospective work-
er’s calculations of whether or not to enter the labour market. hat is, in making their
eiciency-grounded argument for refundable tax credits such as the earned income tax
credit, Batchelder et al. are prepared to assume the eiciency case for such a measure.
52
Azmat, he Incidence of an Earned Income Tax Credit: Evaluating the Impact on Wages
in the UK (2006); Rothstein, he Unintended Consequences of Encouraging Work: Tax
Incidence and the EITC (2008); Leigh, ‘Who Beneits from the Earned Income Tax Credit?
Incidence among Recipients, Co-workers and Firms’ (2010).
53
Eissa and Hoynes, ‘Redistribution and Tax Expenditures: he Earned Income Tax Credit’
(2011). For earlier discussion of the empirical diiculties in the context of the earned
income tax credit see: Hill et al., ‘Evaluating Work Related Cash Beneit Programs: he
Earned Income Tax Credit’ (1998).
directly from the tax relief. hese direct beneiciaries are limited to resi-
dents of the conferring nation state and also those non-residents generat-
ing income in the relevant jurisdiction. Here again the concentration upon
direct beneiciaries receiving cash equivalents is not necessarily consistent
with deinitions of distributive justice framed in universal terms.54
As we have already noted, the detrimental environmental outcomes
associated with the provision of particular tax expenditures typically
are ignored in distributive analyses of tax expenditures. If the burden of
externalities generated by the provision of tax expenditures should be
taken into account, as we have suggested it should, again the universal
nature of distributive justice indicates that there is no reason for exclud-
ing burdens imposed upon those resident outside of the conferring nation
state. For example, any increased cost of staple foods borne by residents
of less developed countries and triggered by excise subsidies for ethanol
should be relected in any distributive analysis of those subsidies.55
Likewise, it is possible that tax expenditures provided in one country
will produce distributively beneicial impacts outside of that country. For
example, the provision of tax expenditures for renewable energy sources
in a large national economy might reduce the world price of renewable
energy technology, making renewable energy more afordable for those
elsewhere.
4.3.2.5 Determining the time at which the tax expenditure

beneits the recipient/imposes a burden upon a person
he lifetime of entities is divided into smaller tax periods (oten annually,
quarterly or monthly) for pragmatic reasons to do with gathering in rev-
enue.56 Distributive analysis of tax expenditures similarly focuses upon
the value of tax expenditure beneits conferred upon a particular class in
a particular year or upon a segment of an entity’s lifetime.57 he reason
54
he relative insigniicance of the intergenerational and interjurisdictional imposition of
tax expenditure burdens in the tax expenditure literature may relect an implicit adop-
tion of a pragmatic imperative. hat imperative is one of enhancing state legitimacy in
a modern democracy where politicians are to some degree accountable to present day
voters in a particular country. As we have already noted, if such tradeofs are to be made,
they should be expressly acknowledged.
55
Determining the extent to which the use of biofuels has increased the price of staple foods
is problematic, given the range of variables at play: United States of America, Why Have
Food Commodity Prices Risen Again? (2011), 9.
56
Arnold, Timing and Income Taxation: he Principles of Income Measurement for Tax
Purposes (1983), 53.
57
Fullerton and Rogers, ‘Distributional Efects on a Lifetime Basis’ (1995).
for this is not clear, although it might have been adopted as a result of
the convention of preparing government budgets annually or for a budget
period.
From the perspective of distributive justice there is no obvious rea-
son for this division of a person’s lifetime. Taking the deontological per-
spective of distributive justice, the divisions of humanity between nation
states and also across time are irrelevant in a fundamental sense.58 he
fundamental rights of a human should not vary depending upon where
and/or when they live.
he division of a person’s lifetime into annual segments for the pur-
pose of distributive tax expenditure analysis might not accurately depict
the distribution of tax beneits across the social spectrum. his is because
the annual reporting of tax expenditures ignores the respective aggregate
value of tax expenditures received by a particular taxpayer over her or
his lifetime. Tax expenditures for those with more assets, for example,
might endure past retirement as capital income continues to enjoy tax
concessions. However, the beneit of earned income tax concessions (for
example) is typically received by those with less income and only for per-
haps some part of their respective working lives. Similarly, as Steuerle
notes, health related beneits might implicitly favour those with higher
incomes who tend to live longer.59
A person may bear a high tax burden in their early years only to be
more than compensated in their later years (assuming that they live
long enough and do not otherwise lose their entitlements to govern-
ment beneits). Even ignoring the provision of government beneits,
the beneit of a tax expenditure in an earlier tax period may be to some
extent recouped in a later year, as where accelerated depreciation deduc-
tions in earlier years are to some degree ofset by greater taxable income
in later years.60
If the purpose of distributive analysis is to assess the provision of
communal resources to diferent groups within a community, dis-
tributive analysis ought to include reporting of expected present day
58
However, this is not to say that the geographical and temporal situs of a human is entirely
irrelevant to application of the deontological concept of justice in the tax expenditure
context. For example, the quantiication of externalised costs arising from provision of
tax expenditures harmful to the environment would include application of a discount
factor across the period between the time of the qualifying behaviour and to the time of
the notional future generation.
59
Steuerle, ‘Can the Progressivity of Tax Changes Be Measured in Isolation?’ (2003).
60
Kahn, ‘Accelerated Depreciation – Tax Expenditure or Proper Allowance for Measuring
Net Income?’ (1979).
aggregate values of tax expenditures provided to the identiied social

groups.61 However, quantifying such lifetime beneits would intro-
duce further considerable complications into the tax expenditure
management framework. Overcoming these diiculties would require
assumptions with respect to a range of matters, including taxpayer lon-
gevity, the characteristics of future taxation systems and the discount
rate to be applied in calculating the net present value of lifetime tax
expenditures.
Intergenerational effects from the perspective of distributive

justice he concentration upon notional cash lows has induced a focus,
in both the statistical work and also the secondary commentary, upon the
distribution of resources amongst a particular population of taxpayers
presently living.
Intergenerational considerations ought to be taken into account in
ensuring that beneits/burdens are equitably distributed across genera-
tions of taxpayers/subjects. his intergenerational aspect is not restricted
to the sharing of the iscal burden relative to the beneits lowing from
the use of public monies (i.e. cash transfers, public assets constructed) –
it also entails consideration of what resources will be depleted now and
what will be preserved and/or created for future users. In this regard tax
expenditures may create a double cost for future generations – by passing
both a iscal and an environmental burden to our successors.
However, here again the empirical difficulties are considerable.
Determining the relative beneits/burdens received/borne by diferent
generations is problematic having regard to the diversity of assets and the
considerable unknowns regarding the true costs that will be borne, and
the beneits that will be enjoyed, at the individual level both now and in
the future. Many assumptions will be necessary with respect to matters
including future population size, technological change and the character-
istics of future tax/transfer systems.
4.3.2.6 Classiication of recipient groups

Typically, the appraisal of tax expenditures from the perspective of dis-
tributive justice concentrates upon the division of the beneits of tax
expenditures having regard to the income of the recipient. Less com-
monly, other discriminators such as gender are considered in deining
61
For discussion of present value accounting in the context of tax expenditure analysis see
Rea, ‘Registered Retirement Savings Plans as a Tax Expenditure’ (1980).
recipient categories.62 As we have noted in Chapter 2, these characteris-

tics of direct taxation in developing countries might be relected in the
tax expenditure report, depending upon the benchmark adopted for tax
expenditure management purposes.
Care needs to be exercised when assessing distributive data as there is the
risk of misrepresenting tax expenditures as distributively unfair. Maktouf
and Surrey note that tax expenditures might appear to favour non-resident
investors in developing countries, but upon further scrutiny this might be
explained upon the basis that few residents are subject to income taxation
because of presumptive taxation rules, non-compliance and poverty.63
4.3.2.7 Concluding remarks regarding the relevance of

distributive justice to tax expenditure management
he preceding discussion indicates that understandings of distribu-
tive justice vary, as does the weight to be accorded distributive justice,
depending upon the political theory adopted. However, assuming that
distributive justice is to be accorded some weight, and that a deonto-
logical conception of distributive justice is adopted, the preceding dis-
cussion suggests that tax expenditures most likely are signiicant for their
efect upon the distribution of the world’s resources between classes of
person within nation states, between persons around the world and also
between generations.
Amongst those who acknowledge the validity of the tax expenditure
concept as a budgetary tool, few if any suggest that the information regard-
ing the distribution of tax expenditure beneits is anything other than
deicient both in terms of its scope and also its accuracy.64 Many coun-
tries ofer no distributional data at all. For countries such as the United
States of America where distributional tables are available, those tables
cover only a limited range of discrete tax expenditures and only limited
information is provided. In particular, the combined efect of diferent
tax expenditures at diferent levels of government is not available.
Although there are signiicant diiculties with identifying and quan-
tifying the distributive impact of tax expenditures, these diiculties do
not warrant the conclusion that distributive analysis of tax expendi-
tures should be ignored. his is particularly so given that the literature
62
Elson, Budgeting for Women’s Rights: Monitoring Government Budgets for Compliance
with Cedaw (2006).
63
Maktouf and Surrey, ‘Tax Expenditure Analysis and Tax Budgetary Reform in Less
Developed Countries’ (1983), 754–5.
64
Sugin, ‘Tax Expenditures, Reform, and Distributive Justice’ (2011).
regarding the distributive efect of tax expenditures in some jurisdic-

tions indicates that a substantial proportion of the beneits of iscally sig-
niicant tax expenditures low to those with higher incomes65 – against
the ‘progressive’ principle upon which many income tax systems are
constructed. Although there are considerable weaknesses in the distri-
butional data available to policymakers, the limited data suggests that
income tax expenditures redistribute resources in favour of those with
greater ability to pay tax.66 his is despite the introduction of the refund-
able tax credit mechanism across a range of measures.67
4.3.3 Allocation eiciency (including compliance costs,

global tax competition)
From a utilitarian perspective, one of the core functions of government in
a market economy is to allocate resources to achieving outcomes which
private mechanisms would not achieve, or would not achieve in such a
way as to maximise social welfare (or ‘utility’). From the perspective of
welfare economics, this utilitarian quest for maximised social welfare
could be taken to suggest that eicient allocation of societal resources is
the lynchpin upon which society should turn.68
here are various deinitions of eiciency according to the degree to
which those afected by wealth maximising moves are incorporated into
the eiciency function.69 Pareto optimality holds that an economy is
65
See the literature referred to in Toder et al., Distributional Efects of Tax Expenditures
(2011), 4:3–4:6.
66
Burman et al., ‘How Big Are Total Individual Tax Expenditures and Who Beneits from
them?’ (2008). Note, however, that this study most signiicantly excluded corporate tax
expenditures, at least some of the beneit of which can be expected to low to those with
direct and indirect (i.e. 401 retirement plans) stakes in the recipient entities. It is reason-
able to expect that those with greater means, who can aford to save, would beneit dis-
proportionately from such trickle-down tax expenditure beneits.
67
For discussion of refundable credits see Batchelder et al., ‘Eiciency and Tax Incentives:
he Case for Refundable Tax Credits’ (2006).
68
For discussion of the changing role of the state vis-à-vis ‘the economy’ see Pusey,
Economic Rationalism in Canberra (1991).
69
For a useful overview of diferent concepts of eiciency see Byrne, ‘Progressive Taxation
Revisited’ (1995), 749–54. Zelinsky, for example, draws a distinction between aggregate
social welfare, sectoral eiciency (being the prevention of exceptional barriers impeding
entry to any particular economic sector) and technical eiciency (being the administra-
tive efectiveness of a particular public policy mechanism). Zelinsky also notes the dis-
tinction between Kaldor–Hicks eiciency, Pareto optimal measures and Pareto superior
measures: Zelinsky, ‘Eiciency and Income Taxes: he Rehabilitation of Tax Incentives’
(1986), 978f.
eicient when no person can be made better of without making at least

one other worse of.70 Pareto superiority holds that a change to the allo-
cation of resources within an economy will be eicient if the welfare of
at least one person can be enhanced while leaving no one worse of.71
Kaldor–Hicks eiciency holds that an economy is eicient if an increase
in the welfare of at least one person is greater than any consequential det-
riment sufered by another, on the assumption that those detrimentally
afected could be compensated whether or not they are actually compen-
sated.72 Eiciency might also be deined in terms of perfect private mar-
kets which are assumed to maximise social welfare, without any regard to
the distribution of aggregate social welfare.73
From the perspective of allocation eiciency, tax expenditures might
reduce social welfare by inducing allocation of resources to less socially
productive activity or tax expenditures might enhance social welfare.
One aspiration of economics as a discipline is to enable us to diferentiate
the former from the latter. Two questions must be resolved when assess-
ing the eiciency efects of a proposed tax expenditure:
1. Is it appropriate for government to intervene by allocating social
resources in pursuit of the particular end (this being determined
according to a speciied social welfare function)?74
2. Assuming that the irst question is answered airmatively, what is the
most eicient means to achieving that end, having regard to the range
of instruments open to the government?75
However, application of the eiciency principle in resolving these ques-
tions is easier said than done76 for a number of reasons:
1. Taking the ‘pure’ form of welfarist eiciency unadulterated by dis-
tributive justice concerns, the aspiration of government should be
70
Byrne, ‘Progressive Taxation Revisited’ (1995), 750–1.
71
Byrne, ‘Progressive Taxation Revisited’ (1995), 751.
72
Byrne, ‘Progressive Taxation Revisited’ (1995), 753.
73
See, for example, Batchelder et al., ‘Eiciency and Tax Incentives: he Case for Refundable
Tax Credits’ (2006), 42.
74
he elements of the social welfare function are open to debate. For example, whether
costs/beneits of future generations ought to be discounted and, if so, the determination
of the discount factor: Weisbach and Sunstein, ‘Climate Change and Discounting the
Future: A Guide for the Perplexed’ (2009).
75
Driessen, ‘A Qualiication Concerning the heory of Tax Expenditures’ (1987), 125–31.
76
Some commentators pass over this challenge, as where MacNevin cursorily refers to
‘appropriate modelling techniques’: MacNevin, ‘Issues in Income Tax Accountability:
An Analysis of the Rental Housing Sector’ (1994), 682.
construction of an environment in which perfect markets allocate

societal resources.
In the real world, this aspiration commonly is thwarted by infor-
mation asymmetry, externalities and concentrations of market power.
Asymmetric information means that market participants cannot deter-
mine their respective best interests and so the bargains they make might
not increase aggregate social welfare. Externalities – beneits and bur-
dens received by or imposed upon third parties to a transaction – oten
cannot be the subject of market transactions because they are widely
distributed and may also be diicult to identify and quantify (i.e. envir-
onmental harms). he failure of markets to capture positive/negative
externalities means that these beneits/burdens oten will not be factored
in to bargains made. Again, this informational shortcoming means that
aggregate social wealth might not be enhanced by a market transaction.
A case for tax expenditures might therefore be constructed upon the
basis that a particular tax expenditure overcomes a market failure77
and thereby rights the ineicient social ship to eicient equilibrium.
Ideally, the overall cost of the tax expenditure would be less than the
overall cost of the targeted market failure so that aggregate social wel-
fare would increase.
However, precisely identifying and quantifying the cost of a par-
ticular instance of market failure is problematic.78 Risks associated
with national security and ecological sustainability, for example, are
not readily measurable79 and valuing such risks will therefore be dii-
cult if not impossible.
2. he theory of the second best indicates that removal of one impedi-
ment to an eicient economy will not necessarily increase aggregate
77
Yorio, ‘he Future of Tax Reform: A Rejoinder to Professor Zelinsky’ (1987), 901–3.
78
See, for example, Easson, Tax Incentives for Foreign Direct Investment (2004), ch. 3.
79
and Social Progress (2009). Metcalf discounts the relevance of national security concerns
when considering energy related tax expenditures, upon the basis that oil is traded on
world markets and so any interruption to supply would afect the price of domestically
produced oil/oil substitutes and therefore encouraging domestic production would make
no diference in terms of insulation from future price shocks. However, this overlooks the
possibility of complete cessation of supply from external sources, as in the time of war or
severe disruption to international shipping: Metcalf, ‘Using Tax Expenditures to Achieve
Energy Policy Goals’ (2008), 91. Moreover, it will be extremely diicult to determine
whether the risk of oil price shocks has been accurately priced into ‘secondary’ transac-
tions – for example, the risk of personal insolvency of a borrower caused by unemploy-
ment triggered by an oil price shock.
social welfare.80 h is complication could be dealt with by making

the simplifying assumption that the economy is otherwise perfectly
competitive.
3. Measuring the welfare efect of any measure must be based upon
assumptions regarding the circumstances at which welfare will be
maximised. For example, comparisons between pre- and post-tax
worlds typically assume that the pre-tax world maximises welfare as
omniscient individuals will have maximised their respective personal
welfare. he assumption of omniscience is a bold one. Imposing a tax
upon motor fuel might induce some to substitute to other means of
travel. Individuals might realise that this ‘distortion’ enhances their
respective welfare because of hitherto unappreciated beneits of the
new mode of transport – the itness beneits of cycling, the social
aspect of public transport/car pooling and so forth.
4. he treatment of externalised costs is problematic, as many such costs
will be diicult to quantify. Even in Weisbach ’s limited concept of
eiciency, being minimisation of deadweight costs (i.e. minimisation
of substituting behaviour which generates welfare costs greater than
the revenue raised from the relevant tax),81 the external costs of alter-
nate policy mechanisms can be diicult to measure. For example, the
impact of diferent line drawing rules upon the legitimacy of a tax sys-
tem, and arguably therefore upon widespread voluntary compliance
with a tax system and even more generally upon public acceptance of
government action, are signiicant factors which complicate the com-
putation of deadweight costs. Such costs might be assumed away, but
the validity of such assumptions need not necessarily be accepted.82
5. he time frame over which the eiciency impacts of measures are to be
assessed is signiicant because policies might diminish aggregate wel-
fare in the short run, but enhance it in the long run. Or vice versa. In
such cases selecting the appropriate discount rate to apply in assessing
the implied future beneit will be problematic.83
here are, then, considerable conceptual and practical challenges with
respect to the imperative of eiciency. he discipline of economics might
80
For discussion of this point see Zelinsky, ‘Eiciency and Income Taxes: he Rehabilitation
of Tax Incentives’ (1986), at 996f.
81
Weisbach, ‘Line Drawing Doctrine and Eiciency in the Tax Law’ (1999), 1650–5.
82
See, for example, the critique of Weisbach ’s framework in O’Reilly, ‘Principles of Eicient
Tax Law: Apocryphia’ (2008), Part V.
83
Weisbach and Sunstein, ‘Climate Change and Discounting the Future: A Guide for the
Perplexed’ (2009).
not provide razor sharp identiication and quantiication of preferential

treatment by government in a market economy. However, mathematical
precision is not always necessary to raise questions around the provision
of public beneits to a particular group, and in such cases the focus should
shit to identify and quantify the wider public good arising from such tax
expenditures.
his point can be illustrated by reference to the example of ethanol
fuel subsidies provided by the federal government of the United States of
America. One of the hallmarks of ‘developed’ economies is their energy
use relative to ‘undeveloped’ economies. his demand for energy, the pro-
spect that the world community has arrived or passed the point of peak oil,
political instability in oil producing regions and in regions through which
energy supply lines pass, strategic concerns in terms of assuring a coun-
try’s energy supply during exigencies such as war, the scale of resource
conglomerates and their ability to shape public policy, climate change
and other environmental concerns all make for a complex socio-political
environment in which energy policy is shaped. he signiicance of the
issue is relected in the taxation treatment of energy. Along with all man-
ner of regulatory and direct spending measures, tax expenditures igure
prominently in respective governments’ energy policies.84
Typically the literature regarding tax expenditures in this ield adopts
the imperative of eiciency as a starting point – the maximisation of
social welfare is the focus rather than the distribution of that aggregate
social welfare across individuals.85
Given the multiple facets of national energy policy, determining
whether social wealth has been maximised as a result of adoption of
these tax expenditures would require many bold assumptions regarding
the respective risks and the costs of alternate risk mitigation strategies.
Nevertheless, the impacts of some energy tax expenditures are such that
84
For a general overview of energy related subsidies, including tax subsidies, see OECD,
Inventory of Estimated Budgetary Support and Tax Expenditures for Fossil Fuels (2011).
For a review of energy related tax expenditures in the United States of America see
Milne, ‘US Climate Change Policy: A Tax Expenditure Microcosm with Environmental
Dimensions’ (2011).
85
Metcalf, ‘Using Tax Expenditures to Achieve Energy Policy Goals’ (2008), 90–4. As
we have already noted, environmentally damaging tax expenditures can meaning-
fully be assessed from the perspective of distributive justice: see section 4.3.2 . See also,
for example, Toder, Eliminating Tax Expenditures with Adverse Environmental Efects
(2007); Metcalf, ‘Environmental Taxation: What Have We Learned in h is Decade?’
(2009); OECD, Environmentally Related Taxes and Charges in OECD Countries: Issues
and Strategies (2001), 21.
their merit must be questionable. Federal biofuel subsidies provided in

the United States are one example which have attracted negative assess-
ments with respect to their eicacy and cost.86
4.3.4 he size of government

To some observers, there is an optimal size of government – big enough to
provide the assurance necessary for private markets to operate while small
enough to enable private operators the greatest liberty possible to maxi-
mise personal wealth, and hence some assume social wealth, through the
market mechanism.87 For some the size of government is a matter of overt
political philosophy while others refer to it as though this factor is signii-
cant without explaining why it might be signiicant.
Simplistic analyses of the size of government ignore tax expenditures
and other means by which government intervenes in allocation decisions
(i.e. regulation, asset ownership) by focusing upon cash transfers.88 hese
analyses overlook the fact that their ‘neo-liberal’ concern is with mini-
mising government interventions in the private allocation of resources
irrespective of how those interventions might arise. More sophisticated
analyses acknowledge that tax expenditures are attractive to politicians
because in one fell swoop a tax expenditure can ofer the appearance of
small government by reducing government revenue while simultaneously
being a lever by which government intervenes in allocation decisions.
‘Neo-liberal’ think tanks therefore see tax expenditures as embodying
two evils. First, tax expenditures reach beyond the neo-liberal’s concep-
tion of the minimalist state by intervening in allocation decisions which
should be let to individual preferences expressed through market and
other (i.e. intra-family) mechanisms. Second, tax expenditures are crucial
86
United States of America, he Impact of Ethanol Use on Food Prices and Greenhouse
Gas Emissions (2009); United States of America, Using Biofuel Tax Credits to Achieve
Energy and Environmental Policy Goals (2010); United States of America, Opportunities
to Reduce Potential Duplication in Government Programs, Save Tax Dollars, and Enhance
Revenue (2011); United States of America, Biofuels – Potential Efects and Challenges of
Required Increase in Production (2009); Searchinger et al., ‘Use of US Croplands Increases
Greenhouse Gases hrough Emissions from Land Use Change’ (2008).
However, such analyses cannot be applied in other jurisdictions without careful consid-
eration of any diferentiating characteristics of biofuel production in those jurisdictions:
Australia, Biofuels in Australia – An Overview of Issues and Prospects (2007), 5.
87
Caribata , ‘From Welfare State to Optimal Size of Government: A Paradigm Shit for
Public Policy’ (1998).
88
Caribata , ‘From Welfare State to Optimal Size of Government: A Paradigm Shit for
Public Policy’ (1998).
to the ‘iscal illusion’ propagated by the welfare state, in that their exist-
ence means that the iscal budget is not an accurate measure of the scale of
government intervention in the allocation of resources. As the iscal illu-
sion account goes, the opacity of public interventions in private decisions
allows the state (‘Leviathan’) to extend its reach beyond the point which
the public would allow if transparent reporting of the size of government
applied.89 Neo-liberal commentators therefore call for retrenchment
of tax expenditures (as well as a reduction in the size of a government’s
budget).90
To this group, tax expenditures are a key illustration of their critique
of the legitimacy of contemporary democratic government. Tax expen-
ditures are portrayed as burdening the ordinary person with excessive
taxation used to pay for ineicient beneits to ‘special interests’, and in a
manner which undermines transparent critical review.
4.4 Public administration
4.4.1 Fiscal signiicance

he existence of tax expenditures is iscally signiicant for several
reasons.
4.4.1.1 Aggregate cost

International comparison of aggregate tax expenditure budgets is prob-
lematic given the host of subjective judgements and assumptions upon
which such budgets are grounded.91 Variations to the deinition of the tax
expenditure concept across jurisdictions, the inclusion of negative tax
expenditures when calculating the aggregate tax expenditure amount,
variations to the deinition inter-temporally within jurisdictions and also
variations to the efectiveness with which tax expenditures (as deined)
are both identiied and measured mean that comparison of aggregated
tax expenditure data is extremely diicult, if not impossible. Moreover,
it is not necessarily correct to assume that the quantum of revenue fore-
gone is the sum of the individual tax expenditure amounts, given that
89
Davidson, Fiscal Illusion: How Big Government Makes Tax Look Small (2007). For an
early review of the literature, see: Misiolek and Elder, ‘Tax Structure and the Size of
Government: An Empirical Analysis of the Fiscal Illusion and Fiscal Stress Arguments’
(1988).
90
Carling, ‘Ten Principles of Tax Reform’ (2009), 11.
91
See the discussion of this point in Chapter 3.
Public administration 129
the behavioural response to complete abolition of all tax expenditures

may entail some revenue leakage and/or the introduction of correspond-
ing direct spending programmes. Further, as Laurie and McDonald note,
variations in the comparator such as the impact of terms of trade upon
nominal GDP, mean that care should be taken with time series data.92
Notwithstanding all of these reservations, from the perspective of pru-
dent iscal and economic management the scale of tax expenditures in
gross and relative terms can be such that they have a signiicant impact
upon a government’s iscal position and so should not be ignored.93
Another perspective upon the scale of tax expenditures is to consider
their signiicance relative to the amount of ‘discretionary’ public expend-
iture in a government’s annual budget.94 he discretionary spending com-
ponent of a government’s budget is quantiied by excluding entitlement
programmes such as social security from the direct spending compo-
nent of the budget. he distinction between discretionary and manda-
tory spending has more or less signiicance within the budgetary context
of diferent countries, depending upon whether direct spending upon
‘entitlement’ programmes must be the subject of an annual appropriation
by the legislature.95 Assuming that tax expenditures equate to ‘discretion-
ary’ spending, quantifying them relative to a government’s discretionary
direct spending serves to highlight the signiicance of tax expenditures
with respect to public policy which truly is funded on an annual basis.96
No matter which way they are framed, in many countries tax expendi-
tures comprise such a substantial component of a government’s spending
as to demand careful management.
4.4.1.2 International assistance

he provision of tax expenditures can also afect the iscal position of
states, as efective tax expenditure management is one consideration taken
into account with respect to assessment of sovereign debt by international
92
Laurie and McDonald , A Perspective on Trends in Australian Government Spending
(2003), 29.
93
See the international comparative data recorded in OECD, Tax Expenditures in OECD
Countries (2010), Part II.
94
See, for example, Kleinbard , ‘How Tax Expenditures Distort Our Budget and Our
Political Processes’ (2009), 930.
95
For discussion of this aspect see: Yin, ‘Temporary-Effect Legislation, Political
Accountability, and Fiscal Restraint’ (2009), 183.
96
As we will note below with respect to the place of tax expenditures in the budget process,
when compared to discretionary spending, the absence of an annual appropriation for
tax expenditures is a striking lacuna in contemporary budget process.
organisations including the World Bank and the International Monetary
Fund.97
Promoting public policy takeup through tax

4.4.1.3
expenditures
By contrast to other public policy mechanisms, tax expenditures can
leverage of the higher proile of an existing and broad based taxation sys-
tem and thereby ofer the twin beneits of a higher takeup rate for the
intended beneiciaries by comparison to other spending mechanisms and
also minimisation of administration costs. Although Zelinsky argues
that this ‘technical eiciency’ of tax expenditures afords a good case for
retaining meritorious public intervention, he does not quantify those ei-
ciency gains and nor does he specify a methodology by which those gains
might be quantiied.98 Further, as we discuss below,99 there may be sig-
niicant costs speciic to the tax expenditure mechanism which can be
diicult to quantify but which should nevertheless be taken into account.
4.5 Legal signiicance
4.5.1 Constitutional power and tax expenditures

he tax expenditure concept embodies a substance over form analysis
of a legislative rule that focuses attention upon the substantive spending
behind the formal tax rule. Surrey and McDaniel considered the signii-
cance of this substance over form approach to ‘tax spending’ to the appli-
cation of a country’s constitution.100 hey noted that a federal constitution
might allocate particular substantive heads of power to a particular level
of government. Under such a constitutional framework, a government
might rely upon its taxing power by using tax expenditures efectively to
extend its reach into otherwise constitutionally proscribed ields.101
97
Stewart and Jogarajan, ‘he International Monetary Fund and Tax Reform’ (2004);
Stewart and Phillipps, ‘Deining Fiscal Transparency: Transnational Norms, Domestic
Laws and the Politics of Budget Accountability’ (2009); Burton, ‘Is Participatory Tax
Transparency Achievable?’ (2006).
98
Zelinsky, ‘Eiciency and Income Taxes: he Rehabilitation of Tax Incentives’ (1986),
1036–7.
99
See section 4.6.
100
Surrey and McDaniel, Tax Expenditures (1985), ch. 5.
101
However, as we have already noted, other commentators have observed that there is
no logical reason for stopping this substantive analysis of constitutional rules with
the tax/spending dichotomy, given that ‘spending’ can be conceived as a ‘negative tax’:
Legal significance 131
4.5.1.1Tax expenditures as a means of achieving objectives

that cannot be the subject of direct spending
A government’s successful deployment of tax expenditures in pursuit of
policy objectives which would be illegitimate in the case of direct spend-
ing will depend upon the role and approach of a constitutional tribunal
or court – whether the court will hear petitions regarding the scope of the
taxation power and whether the tribunal or court will adopt a formalis-
tic or a substance approach when interpreting the scope of the taxation
power. In the United States of America this issue has received some judi-
cial attention as well as attracting secondary commentary.102
Most recently, the formalistic approach was adopted by a bare majority
of the Supreme Court in Arizona Christian School Tuition Organization v.
Winn.103 In that case the majority held that residents of Arizona did not have
standing to challenge, under the Establishment clause of the United States
Constitution, the provision of tax credits under Arizona state law with
respect to contributions to School Tuition Organizations. he Supreme
Court Justices in the minority applied the concept of a tax expenditure
in concluding that a tax beneit is a form of government spending and
that the earlier decision in Flast v. Cohen104 therefore applied. In Flast the
Supreme Court allowed taxpayers standing to challenge the expenditure
of federal funds for the beneit of religious schools. he minority in Winn
considered that Arizona taxpayers had a suicient interest in the provi-
sion of tax beneits to religious schools because the provision of tax ben-
eits necessarily meant that higher taxes would be borne by the remainder
of the taxpaying community. In other words, tax expenditure analysis
informed the minority’s view that tax beneits equate to direct spending
and therefore should be subject to the same constitutional safeguards.
he majority in Winn, however, rejected the proposition that the provi-
sion of tax beneits necessarily meant that the non-beneiciaries amongst
the taxpaying community would bear higher taxes.
Winn illustrates the tension between substance and form which the tax
expenditure concept crystallises. In the Supreme Court, the majority’s
rejection of the tax expenditure concept relects a lawyer’s categorisation
Kaplow, ‘Taxation and Redistribution: Some Clariications’ (2007). Similarly, a spend-

ing programme could be reconstituted into a regulatory programme: Roin, ‘Truth in
Government: Beyond the Tax Expenditure Budget’ (2003).
102
Zelinsky, ‘Are Tax “Beneits” Constitutionally Equivalent to Direct Spending?’ (1999);
Zelinsky, ‘Winn and the Inadvisability of Constitutionalizing Tax Expenditure Analysis’
(2011).
103
131 S. Ct. 1436 (2011). 104 392 US 83 (1968).
of social phenomena having regard to the mechanism by which a particu-

lar object is achieved. In this vein, Zelinsky notes the controversial nature
of the tax expenditure concept and argues that the concept is too impre-
cise to ofer any basis for constitutional adjudication. Zelinsky adopts
a positivist philosophical stance in this respect, only accepting truth
claims which can be demonstrated to be true. hus, Zelinsky contends
that the concept of a tax expenditure cannot be demonstrated to be true
and therefore should be discarded. his positivist stance is at odds with
Zelinsky’s other work, where he is prepared to propose various defences
for tax expenditures without demonstrating that the grounds for those
defences are true.
4.5.1.2 Tax expenditures and iscal federalism

Tax expenditures can also impact upon the federal structures prescribed
under some constitutions.
In a federal system of government tax expenditures allowed under
the tax law of one level of government and which beneit another level of
government may be prescribed by the relevant federal constitution. For
example, taxation of government property may be prohibited – overt rec-
ognition of the signiicance of ‘iscal federalism’.105 he constitutional sta-
tus of such tax expenditures ought not mean that they should pass without
scrutiny. here have been cases where government agencies beneiting
from the tax expenditure have, efectively in return for a fee, entered into
arrangements with the object of providing some of the beneit of their tax
exemption to commercial parties which would not otherwise be entitled
to such beneits.106 he efect of such arrangements is that the revenue lost
to the illegitimate use of the tax exempt status is equivalent to additional
cash grants from one level of government to another. In the absence of
comprehensive tax expenditure management, such ‘grants’ would escape
public scrutiny of their respective merits. Identiication and manage-
ment of such tax expenditures might prompt constitutional reform and/
or prompt closer legislative speciication of tax rules intended to achieve
compliance with the constitutional requirement.
4.5.1.3 International trade law

he Agreement on Subsidies and Countervailing Measures (‘the SCM
Agreement’) is a multilateral treaty which speciically regulates the
105
For elaboration of the concept see Boadway and Shah, Fiscal Federalism (2009).
106
FCT v. Orica Ltd (1998) 98 ATC 4494.
Legal significance 133
actions of World Trade Organization (‘WTO’) members with respect to

prescribed ‘subsidies’. Paragraph 1.1(a)(1)(ii) provides that, for the pur-
pose of the SCM Agreement, a subsidy shall be deemed to exist if:
(a)(i) there is a inancial contribution by a government or any pub-
lic body within the territory of a Member (referred to in this
Agreement as ‘government’), i.e. where: …
(ii) government revenue that is otherwise due is foregone or not col-
lected (e.g. iscal incentives such as tax credits).107
Provided that the subsidy is ‘speciic’, such as where it is explicitly
limited to particular enterprises,108 in general109 the subsidy will
be prohibited110 and also will be subject to compulsory notiication
procedures.111
he efect of the SCM Agreement is that WTO Members must identify
subsidies of the proscribed kind, notify the WTO of their existence and
desist from providing or maintaining such subsidies. hus, for members
of the WTO tax expenditure reporting is not optional as compliance with
the SCM requirements necessitates identiication of subsidies, including
tax expenditures. Moreover, the breadth of the deinition of ‘subsidy’ in
terms of ‘revenue that is otherwise due’ being foregone means that a tax
administrator’s decision to forego disputed amounts under a tax assess-
ment could be subject to the SCM Agreement, particularly where those
disputed amounts are export related such as with respect to transfer pri-
cing adjustments.
In the context of the SCM Agreement, the diiculty of identifying the
benchmark against which ‘revenue foregone’ can be identiied, character-
ised and measured has been acknowledged.112 Nevertheless, the impera-
tive of reducing harmful subsidies has prompted the WTO Appellate
107
Other multilateral treaties might apply within a particular region. For example, Article
90 of the EC Treaty states:
No Member State shall impose, directly or indirectly, on the products of other
Member States any internal taxation of any kind in excess of that imposed directly
or indirectly on similar domestic products. Furthermore, no Member State shall
impose on the products of other Member States any internal taxation of such a nature
as to aford indirect protection to other products.
108
See the SCM Agreement, Paragraph 2.1(a).
109
Agricultural subsidies are speciically excluded: see SCM Paragraph Article 3.1.
110
SCM Agreement, Paragraph 3.1.
111
SCM Agreement, Article 25.
112
World Trade Organization, World Trade Report 2006 – Exploring the Links between
Subsidies, Trade and the WTO (2006), 196f.
Body and also successive Panels to construct a benchmark against which

tax favours have been identiied.113
4.6 Administration of taxation law

From a public administration perspective, the integration of a plethora of
public policy measures within the tax framework engenders challenges
with respect to the administration of tax expenditures and the adminis-
tration of the tax regime itself. hese challenges include:
1. he creation of express tax spending rules within the tax law intro-
duces a ‘spending’ function commonly assigned to tax oicials who
operate predominantly within a ‘revenue raising’ paradigm. he con-
lict between spending and revenue raising creates a risk that the tax
administration will be perceived to be unfair in not applying the ‘tax’
law evenhandedly, by emphasising the risks to revenue over risks to
underachievement of the tax spending rules. In turn, this can impact
upon the voluntary compliance upon which administrators of ‘mass’
tax systems rely.
2. he creation of what we have called ‘operational tax expenditures’
(in Chapter 2) can relect any of a range of causes, including a failure
of the tax law to enable application of the law in accordance with the
benchmark, a lack of administrative resources to execute a law which
supericially applies the benchmark, political pressures imposed upon
tax administrators, weaknesses in the tax legislative process and also
weaknesses in the independence of the tax administration from polit-
ical inluence in those countries where the accountability of the chief
tax oicial to the government is circumscribed.
3. Tax expenditures are oten inserted into the tax law without rigorous
analysis of their purpose and/or of the legislation intended to achieve
that purpose. Poorly constructed legislation increases tax administra-
tion costs in several ways, including:
a. the need to ofer practical guidance to taxpayers in delimiting the
scope of the provisions;
113
Appellate Body Report, United States – Tax Treatment for ‘Foreign Sales Corporations’,
WT/DS108/AB/R, adopted 20 March 2000, DSR 2000:III, 1619; Panel Report, United
States – Tax Treatment for ‘Foreign Sales Corporations’, WT/DS108/R, adopted 20
March 2000, as modiied by Appellate Body Report WT/DS108/AB/R, DSR 2000:IV,
1675; Panel Report, United States – Measures Afecting Trade in Large Civil Aircrat
(Second Complaint), WT/DS353/R, adopted 31 March 2011, 220f.
Administration of taxation law 135
b. the need to manage the risk of gaming of the tax expenditure rules;
c. the risk management mentioned in b. might entail gathering add-
itional information from taxpayers, which may impose additional
taxpayer compliance costs (i.e. by requiring taxpayers to provide
more information) and this additional complexity might heighten
tension with the tax administrator around ‘tax complexity’;
d. the management of impacts upon voluntary compliance arising
from any diminution of the legitimacy of the tax law; and
e. in view of the preceding factors, making recommendations for
reform or repeal of problematic tax expenditures.
hese ‘hidden’ costs are rarely acknowledged in deliberations regarding
proposed tax expenditures, but ought to be assessed when undertaking
an analysis of the comparative merits of delivering spending through the
tax system as compared to other delivery mechanisms.
4.6.1 he management of tax expenditures is an important

aspect of tax compliance management
Although tax expenditures are a form of government spending, as a gen-
eral rule the tax administrators responsible for overseeing tax expen-
ditures do not apply the same performance indicators as are applied by
other government agencies with respect to direct spending programmes.
A government’s management of a spending programme includes con-
sideration of whether the resources are reaching the target population in
the most eicient and efective manner. By contrast, tax administrators
concentrate upon ‘compliance’ with the taxation law, which is taken to
include tax expenditure provisions.
In this section we argue that the misapplication of a ‘compliance’
framework to government spending in the form of tax expenditures may
undercut the relatively recent trend towards adoption of the ‘partnership’
approach by a number of tax administrators.114 By ignoring performance
assessment of tax expenditure administration against spending indica-
tors, we suggest that tax administrators could reinforce an individualist
approach to taxation which undermines the communitarian underpin-
nings of the partnership model.
here are three components to our argument. he irst is that there is
a risk tax administrators will concentrate upon compliance in terms of
114
Dabner and Burton, ‘Lessons for Tax Administrators in Adopting the OECD’s “Enhanced
Relationship” Model – Australia’s and New Zealand’s Experiences’ (2009).
the underpayment of tax, rather than compliance being understood in
terms of the correct application of the taxation law to an entity’s circum-
stances. he second component of our argument is that this concentra-
tion upon a narrow concept of compliance can lead tax administrators
to ignore application of spending performance management indicators
to the spending component of the laws that they administer. he third
component is that tax administrators’ concentration upon underpayment
of tax can jeopardise the legitimacy of the tax administration and hence
voluntary compliance with the tax law, a key aspect of eicient and efect-
ive tax administration.
4.6.2 What is compliance?

he Shorter Oxford English Dictionary dei nes ‘comply’ as to ‘act in
accordance with or with a request, command; consent to or agree to’.
his suggests that the connotation of ‘comply’ is of one person bending
to the will of another or of the irst person accommodating the wishes
of another where to do so is not necessarily in the i rst person’s inter-
est. he general conception of a tax is that it is a compulsory and unre-
quited transfer of property to the state.115 Putting ‘tax’ and ‘compliance’
together, the predominant social discourse is of an entity making com-
pulsory payments to the government because they bend to the will of
government.116
here are various interpretations of ‘tax compliance’ which accept that
non-compliance can include overpayment of tax. hese deinitions pro-
ceed from a construction of ‘compliance’ framed in terms of conduct in
accordance with a legal obligation. hus, for example, Roth et al. state
that ‘[c]ompliance with reporting requirements means that the taxpayer
i les all required tax returns at the proper time and that the returns
accurately report tax liability in accordance with the Internal Revenue
Code, regulations, and court decisions applicable at the time the return is
115
See, for example, the dei nition of ‘tax’ adopted by the OECD for statistical purposes:
http://stats.oecd.org/glossary/detail.asp?ID=2657 (last accessed 7 December 2012).
116
We feel that there is merit in the argument developed by Murphy and Nagel, he Myth
of Ownership (2002), that a tax is not necessarily an extraction from pre-existing private
property but that it could represent that part of a community’s resources not allowed
to the individual concerned. h is approach stands in opposition to the view that a tax
applies to pre-existing private property. In the context of tax expenditures, the iden-
tiication of the Haig–Simons concept of income as the income tax benchmark has
been interpreted by some as equivalent to the state laying claim to all private property:
Wildavsky, ‘Keeping Kosher: he Epistemology of Tax Expenditures’ (1985), 414–15.
iled’.117 For present purposes it is suicient to note that general deinitions
of compliance such as that adopted by Roth, Scholz and Witte implicitly
acknowledge that non-compliance can arise because of both underpay-
ment and overpayment of tax.
Notwithstanding that deinitions of tax compliance can describe cases
of underpayment and cases of overpayment of tax, we suggest that the
literature dealing with tax compliance118 focuses upon underpayment of
tax.119 Much of the literature regarding the tax gap adopts this focus upon
underreporting of income and underpayment of tax. In Australia a dis-
tinction is drawn between compliance with the taxation law and ‘access-
ing beneits’,120 more generally the literature regarding compliance risk
management focuses upon underpayment of tax121 and in practice a tax
auditor’s framing of the ‘risk hypothesis’ focuses upon revenue at risk
rather than tax spending not delivered. At best, it is fair to say that the
existence of taxpayer over-compliance is only sporadically mentioned in
the tax administration literature.122
4.6.3 Evidence of overpayment of tax

Within the tax compliance literature there is suicient indication that
tax overpayment exists and may be signiicant. hus Roth et al. report
that over-compliance was found in a substantial proportion of taxpayer
returns reviewed by the IRS in the 1980s.123 Similarly, in 1987 Long and
Schwartz found that high income individuals tended to adopt conser-
vative tax positions and overpay tax.124 More recently, in Ireland it was
117
Roth et al., Taxpayer Compliance (1989), 21 (original emphasis).
118
For an annotated bibliography of the literature in this area see Richardson and Sawyer,
‘A Taxonomy of the Tax Compliance Literature: Further Findings, Problems and
Prospects’ (2001).
119
Such as delayed payment of taxes owing.
120
Australia, Compliance Program 2010–11 (2010), 3.
121
OECD, Compliance Risk Management: Managing and Improving Tax Compliance
(2010), 7.
122
hus, for example, although the OECD acknowledges that non-compliant taxpay-
ers might be paying more tax than they are obligated to pay, the consideration of
non-compliance proceeds upon the basis that non-compliance entails an underpayment
of tax: OECD, Compliance Measurement – Practice Note (2001), para 72.
123
Roth et al., Taxpayer Compliance (1989), 51–4.
124
Susan B. Long and Richard D. Schwartz , ‘he Impact of IRS Audits on Taxpayer
Compliance: A Field Experiment in Speciic Deterrence’, Annual Law and Society
Association Meeting (1987), cited in OECD, Compliance Measurement – Practice Note
(2001), 5.
found that taxpayers failed to claim tax beneits to which they were
entitled, resulting in overpayment of tax .125 hese slim references to
the existence of over-compliance raise many questions as to the extent
of taxpayer over-compliance, its causes, its impact upon distributive
justice and also with respect to the role of government in monitoring
such matters and developing appropriate strategies to deal with over-
compliance.
4.6.4 Reasons for construction of compliance in terms of

underpayment of tax
While the revenue agency has expertise in managing tax law and systems,
it is less likely to have expertise in implementing public policy in a particu-
lar substantive domain that a tax expenditure is meant to inluence (such
as the environment, housing, or research and development portfolios).126
A substantial taxation bureaucracy might ameliorate this challenge by
creating specialised sections responsible for administering particular tax
expenditure programs and by enhancing its risk management tools to
better identify non-compliance with the particular tax expenditure rules.
However, if compliance with the tax expenditure rules ultimately falls for
generalist tax auditors to oversee, there must be a limit to the ability of
auditors to cope with the additional complexity engendered by additional
tax expenditures.127
Further, taxation intermediaries are central to procuring widespread
voluntary tax compliance. While large professional tax practices may
adopt similar ameliorative strategies to those adopted by revenue agen-
cies, specialisation may not be available to the bulk of tax intermediaries
and the clients of those smaller advisors may be less likely to bear the
marginal costs associated with the acquisition of specialist knowledge by
tax advisors.128
his can lead to a tax expenditure being applied in either a more or less
generous way than was intended by policymakers. Alternatively, oicers
in the revenue agency may interpret a tax expenditure rule narrowly in a
way that cuts out some of the intended beneiciaries.
125
Ireland, Interim Report on the Underclaiming of Tax Credits, Allowances and Reliefs by
Taxpayers (2007).
126
127
See, for example, the discussion of this at United States of America, 2010 Report of the
National Taxpayer Advocate (2010), 4.
128
Brooks, ‘Comment’, (1988), 326.
4.6.5 Performance measurement

In an efort to enhance the performance of public administration, con-
siderable attention has been devoted to performance measurement in
general. Tax administration performance management also has been
examined from this perspective.129 here is no international standard
regarding performance measurement in the taxation context but one per-
formance indicator is tax revenue raised per unit of cost.130 If this factor
is applied without appropriate caution, the legal allowance of tax expen-
ditures would reduce the administrator’s assessment under such per-
formance measures. hus, the revenue agency may have little incentive to
ensure that tax expenditures are obtained by all eligible taxpayers.131
In circumstances where the tax agency does not actively manage the
risk that tax expenditures are not being delivered to all eligible entities,
perverse outcomes might arise. From the perspective of distributive just-
ice, for example, only taxpayers who can aford to pay for the advice of a
tax lawyer or accountant may succeed in accessing all the tax expendi-
tures for which they are eligible.
4.6.6 Administrative fairness, legitimacy and voluntary compliance

Although legitimacy is not the sole source of a government’s power, a
government relies upon its legitimacy to aford it the capacity to gov-
ern. Following Weber, Tyler observed that the substantive fairness of a
law is an insecure foundation for legitimacy, as this source of legitimacy
depends upon a cultural homogeneity which few societies can achieve,
let alone maintain.132 Others have noted that taxpayers view government
holistically, but suggest that voluntary compliance with a tax law is more
closely correlated to the perceptions of the tax administrator rather than
of the government as a whole.133 Tyler therefore focused upon the pro-
cedural aspect of the implementation of the law.134 his suggests that it is
129
Crandall, Revenue Administration: Performance Measurement in Tax Administration
(2010).
130
OECD, Tax Administration in OECD Countries: Comparative Information Series (2004),
23–4.
131
United States of America, 2010 Report of the National Taxpayer Advocate (2010), 5.
132
Tyler, Why People Obey the Law (1990), 25–6.
133
Kornhauser, Normative and Cognitive Aspects of Tax Compliance: Literature Review and
Recommendations for the IRS Regarding Individual Taxpayers (2007), 149–50.
134
Tyler, Why People Obey the Law (1990). Tyler’s study focused upon policing of what
might broadly be described as ‘anti-social’ crime. he extent to which Tyler’s study can
appropriate to focus upon the question of whether the tax administrator

administers the law fairly, or at least is perceived to administer the law
fairly, so that voluntary compliance is maximised.
his indicates that a perception of imperfect administration of tax
expenditures will be signiicant if there is a perception that the tax admin-
istrator is more diligent in prosecuting the revenue side of the taxation
law. Similarly, a perception that some taxpayers are taking advantage of
‘tax loopholes’ may jeopardise voluntary compliance amongst those who
believe that they cannot access the same favours.135
4.6.7 Assessing administrative efectiveness – identifying

operational tax expenditures
In Chapter 2 we noted that the phrase ‘tax expenditures’ traditionally
has been deined in terms of expressly legislated tax concessions. his is
attributable to the dominance of the United States literature with respect
to tax expenditures, and the acceptance by US commentators that the
statutory deinition of tax expenditure applied at the federal level in the
USA exclusively deines the concept. Accordingly, only passing mention
has been made of what we call non-express tax expenditures, a subcat-
egory of which is ‘operational tax expenditures’.136
Operational tax expenditures arise where there is a departure from
the benchmark taxation system notwithstanding the fact that the tax
law accords with the benchmark in the relevant respect. Operational
tax expenditures might arise, for example, as a result of undetected
non-compliance, administrative failure to enforce the law in cases of
non-compliance, administrative failure to collect tax where tax was due
and also cases where the administrator chooses not to enforce a tax law
which appears to comply with the benchmark but where such enforce-
ment would come at too great a cost.137
be generalised to other contexts, such as taxation law, is open to conjecture. Some tax
administrations, such as Australia and New Zealand, have adopted a ‘cooperative com-
pliance’ model informed by Tyler’s work (and the work of others).
135
United States of America, 2010 Report of the National Taxpayer Advocate (2010), vol 1,
365.
136
Reporting (2011), 2; Krever, ‘Analysing Implicit Tax Expenditures’ (2011).
137
hese examples are of positive operational tax expenditures. here could also be nega-
tive operational tax expenditures where, for example, the tax administrator collects
more tax than the benchmark suggests notwithstanding that the law accords with the
benchmark. Speciic examples of problematic application of laws in line with the income
Rei nement of the concept of operational tax expenditures would

enable more speciic reporting, and critical appraisal, of such departures
from the substantive tax benchmark.138 For example, application of the
‘good administration rule’ by a tax administrator in deciding not to pur-
sue a particular tax claim makes sense where the anticipated costs exceed
the anticipated beneits of such action.139 However, recognition of such
foregone revenue as a tax expenditure should prompt closer scrutiny
of the tax rules at the root of the decision not to pursue the claim. he
operation of the transfer pricing rules, and in particular the application
of the arm’s-length principle in cases where there is no readily identiied
comparable uncontrolled price, can be expected to generate cases where
it may be prudent for a tax administrator not to pursue such claims. he
costs of the administrative action may be signiicant, the outcome may
be extremely uncertain, the taxpayer may have ‘moved on’ by adopting a
diferent approach more favourable to the revenue140 and the opportunity
cost of devoting limited administrative resources to such action may be
signiicant. However, from a systemic perspective the uncertainty regard-
ing the operation of the law should be highlighted by reporting revenue
losses against the benchmark and this ‘cost’ to the revenue might prompt
reconsideration of the nature of the relevant rules. For example, although
formulary apportionment would constitute a departure from the ability
principle underpinning the income tax, in the case of transfer pricing a
government might decide that this ‘arbitrary’ allocation of income might
be both administratively expedient and achieve in practical terms a closer
approximation to the ability principle than prevails where the arm’s-
length principle is imperfectly applied.141
benchmark are the arm’s-length principle for transfer pricing cases (see Kobetsky,
International Taxation of Permanent Establishments (2011), 70–93) and also market
value substitution rules which adopt the market value of assets/services with respect to
prescribed transactions. In both of these cases the diiculty of identifying illegitimate
pricing challenges the resources of tax administrators and taxpayers alike.
138
For the sake of transparency, the benchmark for operational tax expenditures would
be legislated tax rules. h is way, legislated tax expenditures would be reported separ-
ately and operational departures from the legislated benchmark would relect positive
or negative departures from that legislated benchmark. h is would also avoid double
counting.
139
Inland Revenue Commissioners v. National Federation of Self-Employed and Small
Businesses Ltd [1982] AC 617.
140
For example, as a result of a change of senior management personnel where the newcom-
ers have a diferent position with respect to management of tax risk.
141
he merits of formulary apportionment are assessed in Kobetsky, International Taxation
of Permanent Establishments (2011), 93–4, 403–28. See also Roin, ‘Can the Income Tax
4.6.8 Integrating spending rules within a taxation framework

Tax expenditures are usually administered by the revenue agency of a
country because tax expenditures comprise part of the country’s ‘tax’ law.
As Zelinsky notes,142 where the information relevant to determining eli-
gibility for a tax expenditure is also relevant to the determination of tax
liability there is an opportunity to minimise administrative and compli-
ance costs by incorporating the spending programme within the taxation
law. Where the tax system applies to most of the relevant community, tax
expenditures also ofer an opportunity to inform intended beneiciaries
of the government spending in an eicient way. However, the incorpor-
ation of spending measures within a tax system poses challenges to the
tax administrator in designing and implementing the administration of
the tax.
An alignment of the information required for the purposes of assessing
tax liability (absent the tax expenditure) and of the information necessary
to determine eligibility for the tax expenditure will be rare by virtue of the
fact that most tax expenditures are targeted having regard to information
that would not otherwise be necessary for assessing tax liability. hus, for
example, in the case of ‘make work pay’ tax expenditures the income of
the recipient/recipient’s family unit may be used for tax assessment pur-
poses while other information such as the responsibility for dependants
will require additional information to be gathered.143
One efect of the tax compliance cost literature has been to raise the
proi le of the cost of administering a tax system borne directly by tax-
payers.144 Tax administrators are conscious of the length and complex-
ity of their tax return forms, responding to calls for reducing ‘red tape’.
However, in the public mind a tax administrator’s administration of tax
expenditures concerns ‘tax’ rather than ‘spending’, and so introducing
additional tax expenditures and their respective attendant informa-
tion requirements compels tax administrators to confront the challenge
of obtaining suicient information to deliver the government spend-
ing without aggravating public concerns regarding taxation red tape.
Be Saved – he Promise and Pitfalls of Adopting Worldwide Formulary Apportionment’

(2008).
142
Zelinsky, ‘Eiciency and Income Taxes: he Rehabilitation of Tax Incentives’ (1986).
143
Eissa and Hoynes , ‘Redistribution and Tax Expenditures: he Earned Income Tax
Credit’ (2011), 695.
144
For the seminal text on this subject see Sandford et al., Administrative and Compliance
Costs of Taxation (1989).
If information gathered by a revenue authority can be optimal in terms

of length/complexity of a tax return form (for example), it is possible that
the introduction of another tax expenditure will compel a tax adminis-
trator to trade of revenue gathering against delivering spending.
he targeted delivery of public spending is one indicator of high quality
public programmes. However, where tax expenditures are integrated into
a mass tax system the administrative assurance that tax expenditures are
being delivered to the intended recipients can be problematic. In general,
a taxpayer does not need to apply speciically to obtain the beneit of a tax
expenditure and so their eligibility for the beneit is not determined prior
to receipt of the beneit. Instead, typically taxpayers claim the beneit of
the tax expenditure when iling their tax return (if one needs to be iled).
his may not occur until ater the end of the iscal year. In this context,
the overall cost of a tax expenditure may be inlated by illegitimate or
erroneous claims made as part of the tax i ling process.145 he capacity
of the tax administration to eiciently manage this risk by identifying
ineligible recipients may require speciic information not available to it
because tax returns do not request this information. Requiring taxpay-
ers to provide this information might impose considerable costs upon the
intended recipients of tax expenditures, increasing the cost of the pro-
gramme which might not have been factored in to the selection of the tax
expenditure as the optimal mechanism for delivery of the ‘spending’.
In the absence of such speciic information, an unworthy recipient may
fall into the general mass of taxpayers and be diicult to identify by the
administration’s general audit and enforcement tools. Where risk man-
agement tools are utilised by the tax administration, the cost of applying
audit and enforcement tools to particular taxpayers might outweigh any
anticipated direct gain to the revenue. Charged with maximising revenue
given a speciied resource level, tax administrators might prudently con-
centrate their resources upon the ‘low hanging fruit’ where the cost of a
tax audit is more than recouped by the tax and penalties recovered. In this
regard, the diiculty of measuring the indirect beneits – such as enhan-
cing public conidence in the administration of the taxation law and hence
of any additional tax revenue attributable to the voluntary compliance
that that legitimacy engenders – may mean that such beneits are not rec-
ognised or at least are not properly factored in to the tax administration’s
145
For a discussion of this in relation to an earlier version of the Earned Income Tax Credit
(US) see Liebman, ‘he Impact of the Earned Income Tax Credit on Incentives and
Distribution’ (1998), 112f.
risk management tools. As a result, tax expenditures obtained by ineli-

gible recipients (intentionally or not) may be let unchallenged.
Moreover, even if ineligible recipients are identiied through tax
audit, the error may not be identiied, if at all, until several years later,
when taxpayers can be faced with large bills for unpaid taxes plus inter-
est and penalties. Where such penalties and interest are mandatory,
their imposition may diminish the legitimacy of the tax system. Where
the imposition of penalties and/or interest is at the discretion of the tax
administrator, this adds another level of complexity to the administra-
tion of the law.
4.7 he legitimacy of government

he impact of tax expenditures upon the legitimacy of a tax administra-
tion system, and its link to voluntary tax compliance, has been addressed
earlier in this chapter.
Less attention has been devoted to the efect of tax expenditures upon
the legitimacy of the tax system overall and the efect any reduction of
that legitimacy may have upon voluntary tax compliance. In his earl-
ier work, Tyler concentrated upon the process by which tax law was
administered. he process by which tax law is created and also the efect
of the substantive fairness of a taxation law were downplayed. here is
a danger that this literature may be taken to suggest that tax legitim-
acy emerges from fair administration of the tax law. Perhaps belatedly,
Tyler observed that the process by which a law is made should also be
included amongst the factors which shape voluntary compliance with
a law.146 Accounts of ‘tax revolts’ emphasise the fact that a government
cannot hide behind a fair tax administrator if the tax legislation itself,
or the process by which that legislation was created, is considered to be
unfair.147
However, at this stage the strength of this correlation between the
process by which tax legislation is made and voluntary compliance is
speculative148 and in any case may vary with context. It may be the case
that taxpayers accept the fairness of the tax administrator while simul-
taneously believing that the tax law is illegitimate. hus, while the link
between voluntary tax compliance and the presence of tax expenditures
146
Tyler, Why People Obey the Law (1990).
147
Burg, A World History of Tax Rebellions (2004), xx–xxi.
148
Torgler and Murphy, ‘Tax Morale in Australia: What Shapes it and Has it Changed Over
Time?’ (2004), 329–30.
Tax expenditures – the weak link in the fiscal chain? 145
within a country’s tax system has been noted by others,149 the extent to
which tax expenditures afect voluntary tax compliance is not known.
However, the absence of proof establishing a link between tax expendi-
tures and voluntary compliance through legitimation does not mean that
tax expenditures can be ignored. Certainly, there is a considerable body of
literature which proceeds upon the basis that tax expenditures are signii-
cant to the legitimacy of a tax system and/or to the legitimacy of govern-
ment. Some suggest that the scale of tax expenditures in the United States
of America is such that they distort the federal law making process.150
Lobby groups,151 government committees,152 scholars153 and journalists154
relect or promote the view that tax expenditures in their various forms
impact upon the legitimacy of government.
Perhaps for this reason some pivotal commentators have ignored con-
sideration of a link between tax expenditures and voluntary compliance,
preferring to concentrate upon the procedural integrity of robust tax
expenditure management as an end in itself.155
4.8 Tax expenditures – the weak link in the iscal chain?

here is a body of considerable literature regarding the ‘tax legislative
process’ adopted in various countries and within this literature consider-
able attention is given to the creation of tax expenditures.156 Arguably it is
more appropriate to consider ‘tax legislative processes’ in recognition of
the fact that legislation might emerge from diferent parallel policy cycles.
Some legislation will be the product of lengthy, considered and transpar-
ent deliberation while other legislation will be the product of last minute
149
See, for example, Yorio, ‘he Future of Tax Reform: A Rejoinder to Professor Zelinsky’
(1987), 904–5.
150
Kleinbard, ‘How Tax Expenditures Distort Our Budget and Our Political Processes’
(2009), 932; although Kleinbard then outlines the institutional failings which allow the
distortionary phenomenon of tax expenditures to occur, so it is not clear which came
irst – the tax expenditures or the institutional failings. In any case, Kleinbard’s point is
that the scale of tax expenditures is so great as to indicate that there is a problem which
needs to be addressed.
151
Hodge, Who Beneits from Corporate ‘Loopholes’? (2011).
152
United States of America, he Moment of Truth (2010).
153
See, for example, Witte, h e Politics and Development of the Federal Income Tax
(1985), 62.
154
McQuaig, Behind Closed Doors (1987).
155
OECD, Tax Expenditures in OECD Countries (2010); Surrey, Pathways to Tax Reform:
he Concept of Tax Expenditures (1973), chs 2–3.
156
McLure, ‘he Budget Process and Tax Simpliication/Complication’ (1990), 47–51.
compromise brokered so that a legislative package passes legislative hur-

dles. Indeed, it is arguable that the adoption of various tax legislative proc-
esses is one way that a government seeks to meet what may be competing
demands regarding the way legislation ought to be made.157 Competing
imperatives regarding the legislative process weigh upon government in
a democratic environment. he relative power of interest groups which
may have considerable capacity to conduct both direct and indirect
media campaigns may necessitate meetings with lobbyists ‘behind closed
doors’.158 Legislative deals might be made which are politically expedient
for the government without necessarily being justiiable upon the basis of
substantive policy objectives and/or the information and process under-
pinning the government’s preparedness to strike the deal.
Relecting upon tax minimisation practices, Vito Tanzi constructed the
metaphor of iscal termites.159 Braithwaite reached beyond Tanzi’s speciic
instances to the more general problem of gaming of tax rules.160 Taking
the termite metaphor beyond Braithwaite’s concern with the administra-
tion of taxation law, we suggest that taxpayers with the will to pay less tax,
and with the means to pursue that ambition, can be expected to seek out
the weaknesses in ‘the iscal chain’ – the holistic framework by which gov-
ernment constructs and implements its iscal base. his may be achieved
by exploiting weaknesses in the administration of the law, Braithwaite’s
concern, or else at other points in the iscal chain.
One link in that chain is the legislative construction of the tax law.
hose with the will to minimise their tax payments can be expected to
procure political access either directly or through their respective lobby-
ists with a view to obtaining legislative favour or preserving such favours
already allowed to them. hus Kleinbard chastises Zelinsky for too read-
ily assuming that generalist tax writing committees were less likely to be
captured by ‘special interests’ than specialist committees, noting that lob-
byists know to whom they should turn when seeking preferential taxation
treatment from a tax writing committee.161
157
Burton, ‘Chaos, Rhetoric and the Legitimation of “Democratic” Government – A
Critical Review of Australia’s Tax Legislative Process’ (2007).
158
159
Tanzi, Globalization, Technological Developments, and the Work of Fiscal Termites
(2000).
160
Braithwaite, Markets in Vice, Markets in Virtue (2005), 141.
161
(2009), 936; contra Zelinsky, ‘James Madison and Public Choice at Gucci Gulch: A
Procedural Defense of Tax Expenditures and Tax Institutions’ (1993).
5
he politics of tax expenditure management
5.1 Introduction
In preceding chapters the controversies regarding the tax expenditure
concept, the critical appraisal of tax expenditures and the process for
managing tax expenditures have been considered. hese are moral con-
troversies as they deal with competing conceptions of ‘the good life’
which a government ought to pursue through its policies in the name of
‘the people’. he subject of this chapter is whether those moral contro-
versies can be resolved upon objective grounds, at least with respect to
tax expenditure management. If moral objectivity is possible, the con-
troversies considered in earlier chapters of this book can be explained as
contests of power between advocates of rival views rather than as a contest
of ideas underpinned by a commitment to ascertaining truthful propos-
itions which describe or explain phenomena.1
For our purposes, moral objectivity can exist under either universal-
ist or relativist norms.2 Under the universalist concept of moral object-
ivity, controversies can be resolved by reference to a universally accepted
(or at least acceptable) substantive or pragmatic norm or set of norms.3
hat is, there is a norm or set of norms to which any community (i.e. a
community that exists at any point in time and in any geographical loca-
tion and comprised of any constituents) can refer to resolve moral dis-
putes. Such a universal norm might be a irst order norm, such as the
principles of equity or eiciency respectively. Alternatively, the relevant
universal norm might be a second order or constitutional norm which
objectively resolves conlict between irst order norms. An example of the
1
Habermas, Between Facts and Norms (1996), ch. 1.
2
See Harman and homson, Moral Relativism and Moral Objectivity (1996).
3
At least in the realm of tax policy, the rationalist epistemology of pursuing a universally
acceptable truth claim is typically accepted to be the best vehicle for ascertaining truth.
Surrey’s ‘general consensus’ amongst tax policy experts, noted at various points in this
book, is just one illustration of this point.
147
148 The politics of tax expenditure management
latter are the principles that Rawls derives from his ‘original position’ and
which constrain the process for resolving moral disputes and which also
constrain the outcomes of moral disputes.
By contrast, a relativist concept of moral objectivity is one which holds
that controversies can be resolved by recourse to a substantive or prag-
matic norm that is right for a particular community without necessarily
being universally right in the sense of being right for all communities at
all times.
From both objectivist and relativist standpoints, there are several alter-
nate tax expenditure benchmarks. Contenders for such a norm or set of
norms include libertarianism, distributive justice and procedural justice
conceived in terms of assuring democratic norms of accountability and
transparency. he irst part of this chapter reviews these deontological
(or ‘ought to do’) propositions by reference to the broader literature
regarding liberal democratic political theory and explains the respective
relevance of these norms to tax expenditure management.
he irst purpose of this chapter is to set out our scepticism regarding
the prospects of moral objectivism in this domain. Although we cannot
review all irst order norms, we illustrate our argument by reviewing com-
peting political philosophies which underpin what we suggest are funda-
mentally incommensurable approaches to the question of tax expenditure
management. Our conclusion is that there is little prospect of identifying
a morally objective constitutional rule for resolving moral disagreements
about the weight to be accorded competing irst order norms.
he second purpose of this chapter is to review the ontological
(or descriptive) literature regarding tax expenditure politics in a demo-
cratic setting. his descriptive literature indicates that any admittedly
contingent tax expenditure management framework should be estab-
lished with an eye upon the pragmatic aspect of democratic politics.
his chapter is central to the thesis of the book in that it highlights
the fact that political philosophy must underpin any consideration of tax
expenditure management. In Chapter 6 we argue that the way forward
with respect to the controversial, but socially signiicant, tax expend-
iture concept is to explicitly acknowledge that the nature and purpose
of tax expenditure management is governed by one’s moral and political
philosophy. hat is, one’s moral vision for the nature of society and also
one’s conception of modern government will determine answers to ques-
tions regarding the viability of the tax expenditure concept, whether tax
expenditure management is appropriate and also what tax expenditure
management would entail.
Moral objectivity and tax expenditures politics 149
he iscal and moral signiicance of tax expenditure management mean

that how a society deals with the irresolvable controversies regarding tax
expenditures provides instructive insights into the political constitution
of that society more generally. For example, in a democratic context, these
insights concern the institutional pathways for dealing with these contro-
versies, the capacity of citizens to inluence the shape and activity of these
institutions, the extent to which citizens seek to exercise this capacity and
whether some are more equal than others in advancing their respective
contingent moral view.
5.2 he connection between moral objectivity and the politics of

tax expenditure management
he question of moral objectivity is signiicant because the answer to
the question will shape our understanding of the political aspect of tax
expenditure management. here are three possibilities:
1. he objectively right tax expenditure management framework is pos-
sible and has been implemented in at least one jurisdiction;
2. he objectively right tax expenditure management framework is pos-
sible but as yet that framework has not been implemented in at least
one jurisdiction; or
3. here is no prospect of identifying the right framework for tax expend-
iture management.
5.2.1 Objectively right tax expenditure framework is possible,

but has not yet been implemented
If moral objectivity is possible, a study of the politics of tax expenditure
management ought to focus upon distilling the elements of the object-
ively right framework. Informed by this theoretical work, empirical
research could then proceed to determine whether the one right tax
expenditure management framework has been implemented in a demo-
cratic context.
If the right tax expenditure management framework has been adopted
in at least one democratic jurisdiction, an account of the politics of tax
expenditure management for that jurisdiction would be a ‘democratic
politics success story’ from which lessons might be learned and applied
in other jurisdictions.
If, on the other hand, the correct tax expenditure management frame-
work has not been adopted in any jurisdiction, ‘politics’ assumes a
predominantly negative hue as it focuses upon how the public interest

constituted by the one true path has been subverted by human failing
and/or deliberate obstruction. In the context of democratic politics, this is
a narrative of the failings of democracy to achieve the one true public pol-
icy path which may (or may not) only be discernible to the expert eye.
he review of the literature undertaken in earlier chapters indicates
that the tax expenditure literature can be divided broadly between
advocates who argue that irst order moral objectivity is possible in this
domain and those who disagree with that proposition. he former argue
that there is a right answer to questions regarding the meaning and appli-
cation of the tax expenditure concept.4 he latter argue that there is no
universal or relativist moral principle from which right answers might
be derived.
Typical of the former approach was Surrey’s projection of the Haig–
Simons income tax base as one grounded upon ‘widely accepted deini-
tions of income’.5 Moreover, once he believed that he had established the
benchmark against which tax spending could be identiied, Surrey drew
upon rationalist philosophy in arguing that diferent instances of ‘spend-
ing’ should be subjected to uniform treatment. Surrey and McDaniel con-
tinued this universalist aspiration by laying down what they hoped was
the blueprint for a universal tax expenditure management framework.6
Surrey’s appeal to universal norms has been continued by more recent
tax expenditure advocates. Kleinbard, for example, claims that the Joint
Committee of Taxation’s construction of ‘tax induced structural dis-
tortions’7 allows analysis of some tax provisions purely upon eiciency
grounds ‘and not from any normative perspective’ such that the Joint
Committee of Taxation approach could ‘withstand any allegation of a
political or policy prejudgment of the issues’.8 Here Kleinbard projects the
eiciency norm as the standpoint upon which the morally objective ‘view
from nowhere’ can be constructed.
Some commentators seem to combine inconsistent stances. For
example, Avi-Yonah appears to adopt a relativist stance in suggesting that
4
See, for example, Kleinbard, ‘How Tax Expenditures Distort Our Budget and Our Political
Processes’ (2009), 926 (‘neutral design principles’, ‘a principled and neutral way’).
5
6
7
8
Kleinbard , ‘How Tax Expenditures Distort Our Budget and Our Political Processes’
(2009), 927; implying that the utilitarian maximisation of aggregate welfare that under-
pins the eiciency norm is ‘objective’.
the tax expenditure concept ofers an insecure foundation for tax reform
because ‘what is and what is not a tax expenditure lies very much in the
eye of the beholder’9. Undercutting this apparent relativism, however,
Avi-Yonah proposes a three pronged critical framework to underpin his
argument for a tax code comprising both income tax and consumption
tax elements. Avi-Yonah does not explain why these three goals of taxation
are not subject to the same critique – that the goals and their respective
weights are ‘in the eye of the beholder’. Moreover, if the three goals of tax-
ation – revenue, redistribution and regulation – indicate the adoption of
a tax code that combines income and consumption taxes, why can’t these
goals be applied to identifying and managing tax expenditures? It seems
that Avi-Yonah dismisses this prospect upon the basis that ‘tax expend-
iture purists’10 – who are unnamed – purportedly maintain that the tax
expenditure concept is intended to render the tax code free of regulatory
impurities by prompting the elimination of all tax expenditures.11 his
rhetorical ploy – constructing a straw person – does not accurately relect
the bulk if not all of the tax expenditure literature. Ater all, even Surrey
appeared to accept that the purpose of identifying tax expenditures was
to prompt critical scrutiny of those measures without necessarily expect-
ing that all tax expenditures would be abolished.12
However, the controversies surrounding the tax expenditure con-
cept, tax expenditure management and even the practical signiicance of
tax expenditures discussed earlier in this book suggest that few would
agree that any one tax expenditure management framework adopted in a
9
Avi-Yonah, ‘hree Goals of Taxation’ (2006), 24.
10
11
Other commentators have adopted this ‘all or nothing’ approach in discounting tax
expenditure analysis. For example, Weisbach and Nussim make the point that there is no
a priori reason for abolishing all tax expenditures as some of these measures might be the
most efective means of implementing government policy. For this reason they argue that
tax expenditure analysis is focused too narrowly: Weisbach and Nussim, ‘he Integration
of Tax and Spending Programs’ (2004), at 977f. See also Zelinsky, ‘Eiciency and Income
Taxes: he Rehabilitation of Tax Incentives’ (1986), 1021, observing that Surrey and other
‘proceduralists’ considered that tax expenditures were theoretically acceptable if it were
appropriate for government to intervene in private markets, but suggesting that this group
contends that all tax expenditures ‘must be rejected because the tax system is an imper-
fect instrument with which to implement government policies’. See also Zelinsky, ‘James
Madison and Public Choice at Gucci Gulch: A Procedural Defense of Tax Expenditures
and Tax Institutions’ (1993), 1164: noting the ‘Surrey school’s invariable preference for
direct spending’.
12
Developed Countries’ (1983), 757 (‘it may be ai rmatively desirable to utilize certain tax
expenditures’).
particular jurisdiction can be morally objective from either universalist

or relativist standpoints.
Notwithstanding the controversy regarding the prospect of identifying
the one right tax expenditure management framework, much of the tax
expenditure literature apparently proceeds upon the basis that a univer-
sally right framework is possible but has not been achieved in any juris-
diction. his segment of the tax expenditure literature oten turns to the
politics of tax expenditure management with a view to identifying the
sociological factors that have contributed to the failure to achieve rational
public policy management. To the extent that this literature focuses upon
the public policy process, moral objectivity underpins the quest for iden-
tifying the one right process by which laws (including tax expenditures)
should be made and reviewed. his entails consideration of second order,
or constitutive, moral principles which could underpin the resolution of
such controversial moral questions.
5.2.2 An objectively right tax expenditure management framework

underpinned by one irst order norm is not possible, but there is an
objectively right constitutional rule which resolves such disagreement
In the face of these controversies regarding the identiication of irst
order substantive moral principles that might underpin the one true tax
expenditure management framework, some argue that tax expenditure
management is so controversial that it ought to be abandoned or at least
be accorded less signiicance.13 hese critics ignore the contingent foun-
dations for this abandonment.14
If the one right model of tax expenditure management cannot be
derived from irst order moral principles, the subject of the politics of
13
Bittker, ‘A “Comprehensive Tax Base” as a Goal of Income Tax Reform’ (1967), 985;
Bittker, ‘Income Tax “Loopholes” and Political Rhetoric’ (1973), 1113; Kahn, ‘Accelerated
Depreciation – Tax Expenditure or Proper Allowance for Measuring Net Income?’
(1979); Bartlett, ‘he End of Tax Expenditures as We Know hem?’ (2001).
14
h is confusion of sceptical and rationalist epistemology is evident in Zelinsky, ‘James
Madison and Public Choice at Gucci Gulch: A Procedural Defense of Tax Expenditures
and Tax Institutions’ (1993), where Zelinsky notes (at 1166) that the application of the tax
expenditure benchmark is not always clear while then proceeding to claim (also at 1166)
that in some cases it will be more eicient to provide a tax subsidy. Implicit in the latter
claim is the proposition that social welfare outcomes of alternate policy mechanisms can
be accurately quantiied. Zelinsky adopts a sceptical interpretative stance with respect to
the meaning of the benchmark, but does not explain why the generally acknowledged dif-
iculty of identifying eiciency measures should not lead to a similar scepticism regard-
ing empirical claims regarding social phenomena.
tax expenditure management entails consideration of more fundamental

constitutional questions. Is there any role for government in this morally
indeterminate policy domain? If so, how can democratic government be
constructed in such a way that it can cope with substantive pluralism with
respect to such policy issues?15
Pursuing this inquiry entails consideration of whether one univer-
sally true constitutional framework, by which such political contests can
be resolved in a manner that builds rather than destabilises legitimate
government, can be identiied. Acknowledging that moral objectivity
is problematic in the domain of tax expenditure management therefore
requires consideration of whether moral objectivity is possible with
respect to democratic institutions themselves. If there is no prospect of
one right democratic framework, again there is no bedrock upon which to
found democratic institutions which govern the procedures for shaping
tax expenditure management. In section 5.6 of this chapter we critically
review several procedural and substantive constitutional models which
seek to construct an absolute constitutional rule by which moral contro-
versies might be resolved. However, we argue that no one of these models
is convincing. Further, we note the similarity between the tax expend-
iture literature and liberal political theory and suggest that the existence
of such controversies is unexceptional.
The existence of apparently irresolvable controversies brings
the ‘political’ aspect of tax expenditure management to the fore.
Acknowledging the existence of value pluralism is not new and has
underpinned a substantial body of literature dealing with the ontological
aspect of tax expenditure management. hat is, this literature purports
to present what ‘really happens’ rather than starting from the deonto-
logical standpoint of considering what ought to happen. his ontological
literature includes, for example, the public choice literature, critical stud-
ies and pluralist theories of politics. In more recent times the more opti-
mistic pluralist theories have waned as ‘politics’ has been associated with
unaccountable, unprincipled and sometimes corrupt determination of a
community’s laws.16 he tax expenditure management literature includes
a healthy segment of public choice and critical studies literature present-
ing this critical depiction of ‘democracy’ at work. Although there can be
15
In this chapter we focus upon democratic government because the identiication and
application of a tax expenditure management framework is most problematic in a demo-
cratic context where competing conceptions of the political good must be resolved.
16
For example, as a result of the depiction of self-interested legislators that is central to the
‘hard’ version of public choice theory described in section 5.5 below.
deontological elements within some strands of this ontological work – for

example where the work depicts the human failure to achieve deonto-
logical democratic norms of accountability and transparency – in the
main this body of work merely purports to describe social reality through
a positivist lens. he negative portrayal of tax expenditure management
within this literature can feed the libertarian deontological norm of indi-
vidual freedom by suggesting that all state institutions must be corrupt
and therefore ought to be reduced to a bare minimum.
However, this critical depiction of political institutions is not the only
connotation of ‘political’. Chapter 6 proposes a normative vision of demo-
cratic politics – conceived as a rhetorical building of alliances. In this
model the social signiicance of taxation indicates that the controversial
aspects of a country’s taxation system, including tax expenditures, should
not be ignored if the aspirations of modern governments to bureaucratic
rationality and democracy are to be honoured.
5.3 Deontological norms and their relevance to

tax expenditure management
In this section we survey some of the major irst order moral norms to
demonstrate that the selection of a preferred approach to tax expend-
iture management is shaped substantially by the theory adopted. In this
context, we adopt a deinition of ‘political theory’ which focuses upon
the deontological question of ‘what role should government play?’ his
deontological aspect can be contrasted with the sociological aspect of
social/political theory – the question of what role government actually
plays and how the exercise of government power is efected. his socio-
logical aspect of social/political theory is considered in section 5.5 of this
chapter.
Further, the survey of moral norms undertaken here is intended to
demonstrate that no one moral norm can lay claim to universal accept-
ance. h is leads to the question of whether there is any one neutral
constitutional theory which can be applied in resolving the conl icts
between proponents of alternate deontological political theories. In sec-
tion 5.6 we consider two political theories which grapple with this prob-
lem. We suggest that the literature in the ield of constitutional political
theory indicates that universal acceptance of any such theory is a slim
prospect.
he competing deontological political theories outlined in section
5.3 and the absence of a universally acceptable (and possibly neutral)
Deontological norms and tax expenditures 155
constitutional political theory suggest that the prospect of identifying one

universally accepted tax expenditure management framework is remote.
hat is, the elaboration of a tax expenditure management framework can-
not be undertaken within a critical epistemological framework.
his rejection of critical public policy analysis provides the starting
point for the consideration of ontological approaches to tax expenditure
management, considered in the second part of this chapter. Rather than
setting out to identify a universally acceptable tax expenditure manage-
ment framework, this literature seeks to describe, and sometimes explain,
the social phenomena to do with tax expenditure management. hese
descriptive accounts are critically surveyed in section 5.5.
5.3.1 Libertarianism
For the purposes of our selective survey of irst order moral theories, we
have selected libertarianism17 as our irst subject owing to its signiicance
within contemporary tax policy literature.18 he libertarian challenge to
the welfare state over the latter part of the twentieth century and into the
early twenty-irst century comprises a prominent thread within the tax-
ation discourse of Western democracies. he proliferation and promin-
ence of libertarian think tanks have no doubt contributed to the paring
back both of welfare spending and also the tax regimes which underpin
an interventionist state over these decades.19
5.3.1.1 Individual autonomy

Although there are diferent strands of libertarianism, in broad terms
libertarians adopt the concept of an individual as an entity independ-
ent from society. In this state, the individual is taken to have been ‘free’.
Hayek treats liberty and freedom as synonymous, deining them in nega-
tive terms as ‘that condition of men in which coercion of some by others
is reduced as much as is possible in society’20 and, slightly diferently, ‘the
state in which a man is not subject to coercion by the arbitrary will of
17
Libertarianism was a term which Hayek grudgingly accepted in order to distance his
philosophy from rationalist liberals: Hayek, he Constitution of Liberty (1960), 408.
18
Kornhauser, ‘Legitimacy and the Right of Revolution: he Role of Tax Protests and
Anti-Tax Rhetoric in America’ (2002).
19
For discussion of the emergence of libertarian think tanks initiated by the Trilateral
Commission see Pusey, he Experience of Middle Australia (2003), 7–10. See also Cockett,
hinking the Unthinkable: hink-Tanks and the Economic Counter-Revolution, 1931–1983
(1994); McGarity, Reinventing Rationality (1991), xiv.
20
Hayek, he Constitution of Liberty (1960), 11.
another or others’.21 he implication of this concept of freedom is that the

role of the state is to assure a society that is as free as possible, whilst pre-
serving a core of individual rights to property, contractual promises and
personal safety.22
5.3.1.2 Tax base and redistribution

hese individuals, it is hypothesised, would only forego some of their free-
dom by conferring authority upon a state to monopolise coercive power if
to do so would be consistent with what is perceived to be the individual’s
fundamental right to the product of their labour,23 or a fundamental right
to procure that which is not held by another.24 Libertarians conceive of
taxation as akin to the price paid for state provision of goods and services.
Joining a liberal society, the libertarian contends, should not entail any
ongoing25 commitment to provide private property to another through
some redistributive state mechanism.
hus, in a libertarian society each individual’s contribution to the cost
of the state would be matched by the value of beneits received by that
individual from the state. Libertarians therefore nominally adopt a bene-
it theory of taxation. However, quantifying the value of beneits received
by individuals is both conceptually and practically challenging. he prac-
tical diiculty of matching beneits to taxes at the individual level means
that libertarians difer regarding their preferred tax base – a poll tax, an
income tax or a consumption tax. Duf suggests that libertarians ought
to be agnostic regarding the choice between income and consumption
tax bases.26 Kornhauser suggests that concern regarding the diiculty of
determining the redistributive element of income taxes means that lib-
ertarians prefer a consumption tax base over an income tax, upon the
footing that consumption taxation is less distortive than an income tax.27
21
22
Nozick, Anarchy, State, and Utopia (1974), ix.
23
Locke, ‘he Second Treatise on Civil Government’ (1947), chap. V para 27.
24
Nozick, Anarchy, State, and Utopia (1974), 175.
25
Nozick accepts that there may be a case for correcting past injustice in the pre-liberal
society, although he also observes that this proviso is problematic and most certainly
does not extend to any acceptance of a socialist state: Nozick, Anarchy, State, and Utopia
(1974), 152–3.
26
Duf, ‘Private Property and Tax Policy in a Libertarian World: A Critical Review’ (2005),
32–3.
27
In arriving at this conclusion, the social value of public expenditure is ignored or dis-
counted. Further, assuming that government revenue must remain the same, it is assumed
that the distortive efect upon work/leisure allocation arising from the introduction of a
Further, alternate understandings of ‘beneit’ mean that both income and

consumption tax bases could be rejected by a libertarian in favour of a
poll tax.28
With respect to tax expenditures, libertarians express their preference
for consumption taxation by allowing income tax expenditures which
preference capital income, while rejecting other income tax expenditures
on the basis that they are arbitrarily determined whether as a result of
informational failure or democratic failure. Allowed free hand, libertar-
ians would recast the ‘income tax’ along the lines of a consumption tax
base.29
5.3.1.3 he negative deinition of freedom and redistribution

By contrast to those advocating positive liberty, Hayek’s ‘negative’ dein-
ition of liberty/freedom does not require, and indeed militates against,
an active role of government in shaping the context in which individuals
pursue their respective interests.30 Accepting value pluralism, Hayek con-
tended that government cannot possibly understand the myriad ambi-
tions and capacities of its citizens who pursue their respective unique
concepts of the good life. Government intervention necessarily would be
arbitrary and therefore a breach of liberty/freedom. hus Hayek considers
it better for government to step back and allow individuals themselves to
determine the best means to achieve their respective conceptions of the
good.31
higher consumption tax is more than compensated for by the beneit arising from the
neutral treatment of capital income.
28
Schoenblum, ‘Tax Fairness or Unfairness? A Consideration of the Philosophical Bases
for Unequal Taxation of Individuals’ (1995).
29
Feldstein, ‘A Route to Cutting the Deicit’ (2010).
30
Hayek, he Constitution of Liberty (1960), ch. 2.
31
his vision is grounded upon the proposition that institutions promoting freedom have
evolved rather than being the product of rational insight into the true nature of human-
ity. Hayek, he Constitution of Liberty (1960), 54. Hayek therefore purports to it within
the empiricist tradition of English liberal philosophy rather than the rationalist philoso-
phy of the Continent (he Constitution of Liberty (1960), ch. 4). A contradiction within
Hayek’s work is his conservative preference for liberal institutions which he sees as hav-
ing evolved and his acceptance that no one person can profess to have absolute knowledge
of what is right. his conl ict between monism and pluralism remains unacknowledged
and unresolved – how can Hayek be so sure that social institutions which inhibit freedom
will not be better?
Further, although Hayek considers that departure from his liberal ideology threatens
Western civilisation and that there is a general mood against his liberal individualism, he
suggests that he is doing nothing more than revealing certain fundamental values upon
which there is wide consent (he Constitution of Liberty (1960), 3). However, at other
he onus of proving that the fundamental libertarian principles are

not impinged is borne by the proponent of any public measure, a diicult
proposition given the conceptual and empirical diiculties in the ‘policy
sciences’ in general32 and taxation in particular. hus Hayek emphasises
the arbitrariness of progressive taxation and hence its illegitimacy, 33 a
subject which had been revisited by Blum and Kalven34 shortly before
Hayek’s work. Further, libertarians object to the phenomenon of bracket
creep attributable to the non-indexation of the progressive income tax
rate scale, and so would see this as a negative tax expenditure.35
he libertarian’s negative dei nition of liberty/freedom is a crucial
aspect of this aversion to redistribution. A positive conception of freedom
holds that a person is only free if they can fulil their potential in pursu-
ing their conception of the good life.36 his positive deinition of liberty
requires consideration of a broad range of circumstances to determine
whether a person has been allowed adequate opportunity to lead a fulill-
ing life, including consideration of a person’s education, access to good
health and access to resources which enable them to become what they
want to be (i.e. transport, communication, information, etc.). Acceptance
of a positive deinition of liberty, Hayek states, is a short step from the
evil of government intervention in allocating social resources through
redistribution.37
5.3.1.4 he size of government and iscal illusion – the

libertarian role for tax expenditure management
In Chapter 4 we noted that libertarians see tax expenditures as a grand
deception perpetrated by the state, casting tax expenditures in terms of
the opaque and illegitimate encroachment of the state upon the private
domain.38
points Hayek seems to suggest that principles which are anathema to his negative concept
of freedom, such as redistribution, have come to be widely accepted as a result of error
(he Constitution of Liberty (1960), 309). Hayek does little to explain how the general
population can have been lead so astray from its ‘fundamental values’, although he sug-
gests that the rise of the elitist administrative state has hidden the true nature of arbitrary
and unaccountable administrative rule (he Constitution of Liberty (1960), ch. 17).
32
McGarity, Reinventing Rationality (1991).
33
34
Blum and Kalven, he Uneasy Case for Progressive Taxation (1953).
35
In Canada, for example, the tax brackets are indexed annually.
36
Berlin, ‘Two Concepts of Liberty’ (1969).
37
38
See section 4.3.4.
5.3.1.5 Libertarian tax expenditure ‘management’

For their own reasons, libertarians see tax expenditures as unwarranted
state intervention in private dealings, constructing them as the product of
illegitimate special interest purchase of the state’s monopoly of coercive
power and/or the product of a misguided faith in ‘policy science’ and/or
as symptomatic of cynical obfuscation of ‘Leviathan’s’ reach into private
afairs.
Tax expenditure ‘management’ is not a key concern of the libertarian
given their cynicism regarding state action. he libertarian scepticism
regarding the bureaucratic rationality of the state means that ‘sin’ taxes
and taxes which correct for market failure would be rejected upon the basis
that the quantiication of the tax would be too imprecise to provide assur-
ance that there is no redistributive element of the tax. Under a libertarian
framework, such taxes would be reported as negative tax expenditures.39
Further, for a libertarian the ‘management’ of tax expenditures could
only entail paring back all tax expenditures as one aspect of their battle
with the welfare state. At the individual level, however, libertarians could,
quite consistently with their philosophy, pursue their own interest by
proposing retrenchment of tax expenditures which beneit others while
simultaneously pursuing whatever tax expenditure ‘tax cuts’ they might
procure as a result of special interest pleading during the tax legislative
process. Both strategies procure individual advantage under what is per-
ceived to be an illegitimate tax/transfer system.40 Moreover, the success-
ful pursuit of special interest reinforces the libertarian’s cynical depiction
of the state as the bloated puppet of ‘special’ interests.
5.3.2 State neutrality – optimising allocative eiciency

By contrast to the libertarian’s distrust of the state as a vehicle for social
good, others view the state as the most appropriate mechanism for advan-
cing the social good. he purpose of this section of the chapter is to illus-
trate the proposition that concepts of the social good difer given the
fundamental diferences of opinion regarding the perspective from which
‘the social good’ ought to be constructed. In this section we consider the
view that the role of the state in a market economy is to ensure that those
private markets operate so as to maximise the eicient allocation of social
39
hat is, more tax being collected than the benchmark indicates.
40
Schoenblum, ‘Tax Fairness or Unfairness? A Consideration of the Philosophical Bases
for Unequal Taxation of Individuals’ (1995).
resources. Moreover, we consider the normative relevance of this concept

of eiciency to tax expenditure management.
5.3.2.1 Alternate deinitions of eiciency

he general concept of eiciency connotes maximisation of output from
a given set of inputs. here are several quite diferent deinitions of eco-
nomic eiciency.41 Diferent deinitions relect diferent approaches to:
1. What outputs ought to be considered;
2. he extent to which, if any, the distribution of resources within a com-
munity ought to be included in the deinition of eiciency.
5.3.2.2 Eiciency as wealth maximisation

According to this deinition of eiciency, an economy will be eicient when
aggregate social wealth, as measured by the aggregate value of goods and
services produced, is at its greatest. he value of goods or services is deter-
mined by the price obtained for them in private markets. On this account,
Adam Smith’s ‘invisible hand ’42 within private markets makes the market
mechanism the best means of allocating scarce social resources to the ends
which will be most productive, such that a properly functioning market
mechanism will maximise aggregate social welfare. hus, from the per-
spective of this concept of eiciency, aggregate social wealth will be max-
imised if the tax system does not interfere with efective43 markets.
Assessing tax expenditures from the perspective of wealth maximisa-
tion, the relevant benchmark tax system would be that which maximised
aggregate social wealth. Accordingly, some commentators seem to sug-
gest that this concept of eiciency dictates that there should be no tax
expenditures,44 although it is more likely that the better interpretation
of welfare maximisation would construct the role of the state as one of
ensuring that efective markets exist. It is conceivable that this object
could be pursued by deploying some tax expenditures.
5.3.2.3 Pareto optimal eiciency

Not all accept that the eicient allocation of social resources, by itself,
will produce ‘the good society’. he wealth maximisation deinition of
41
Byrne, ‘Progressive Taxation Revisited’ (1995), 749–54.
42
Smith, he Wealth of Nations (1993), para IV.2.
43
Frank, ‘Progressive Taxation and the Incentive Problem’ (2000), 491.
44
Swit et al., ‘Tax Expenditures: General Concept, Measurement, and Overview of Country
Practices’ (2004), 6.
eiciency leaves the distribution of societal wealth to the distribution aris-

ing from market interactions. By contrast, an economy reaches the state
of Pareto optimal eiciency when no one can be made better of without
making someone else worse of.45 Although this deinition of eiciency
incorporates a distributive element, the distributive norm does not dic-
tate the shares of any new economic wealth between economic winners
and losers under a Pareto move. Similarly, Rawls’ diference principle spe-
ciied some constraint upon the distribution of social resources by requir-
ing social institutions that ‘work as part of a scheme which improves the
expectations of the least advantaged members of society’.46 Some, such
as Cohen,47 prefer to see the distribution of social resources identiied as
a separate imperative of a good society, albeit an imperative that may be
traded of against other imperatives.
5.3.2.4 Welfare maximisation

Rather than focusing upon wealth maximisation, a diferent deinition of
eiciency directs attention to maximising aggregate utility.
As we noted in Chapter 4, measurement of the monetary value of goods
and services is an imperfect proxy for national well-being48 because not all
forms of utility are relected in market exchanges (i.e. leisure). Further, to
the extent that externalities and internalities are not accurately relected
in exchange contract prices, the measurement of aggregate wealth might
not accurately capture aggregate utility.49
However, as we also noted in Chapter 4,50 the deinition of eiciency in
terms of welfare maximisation poses empirical challenges given the dii-
culty of measuring many forms of non-market utility.
5.3.2.5 Technical eiciency

Technical eiciency exists when, for a given resource input, outputs can-
not be increased. Technical eiciency ignores the question of whether
aggregate welfare has been maximised (by, for example, deploying more
resources) and also ignores issues with respect to distributive justice.
45
See the discussion of this concept in Byrne, ‘Progressive Taxation Revisited’ (1995), 750.
46
Rawls, A heory of Justice (1971), 75.
47
48
and Social Progress (2009).
49
Batchelder et al., ‘Eiciency and Tax Incentives: he Case for Refundable Tax Credits’
(2006), 42.
50
See section 4.3.
From this perspective, Zelinsky argues that many tax expenditures

either are or may be consistent with the eiciency principle upon the basis
that tax expenditures leverage of key aspects of a mass income tax sys-
tem and thereby deliver public beneits at lowest input cost.51 However,
the methodology upon which this contention is founded is not clear and
Zelinsky ofers no empirical basis although his claim would seem to
require such a foundation.
5.3.2.6 he rhetorical appeal of wealth maximisation

In recent decades it is fair to say that increased weight has been accorded
to the wealth maximising eiciency principle relative to other tax pol-
icy principles. Repetti suggests that this may be because eiciency can be
measured accurately,52 a point which resonates with Graetz’s observation
regarding the magic of ‘a number’ in tax policy deliberations as part of
preparation of the US federal budget.53 his concentration upon numer-
ical evidence, notwithstanding the limitations of the theory and process
upon which those numbers are founded, can also be seen in the eforts of
the Joint Committee of Taxation to redeine the tax expenditure manage-
ment paradigm54 as well as in the statement of taxation principles evident
in the Australian Henry Review.55
5.3.2.7heoretical disagreements regarding the application

of wealth maximisation
Even if one were to accept that a wealth maximising society is a ‘good’
society, the apparent objectivity of the wealth maximising norm is open
to several areas of dispute. here will be diferences of view as to empirical
matters (i.e. how to measure wealth), the application of the principle in
the real world of the second best,56 the time frame over which eiciency is
to be assessed and also the geographical scope of the question (i.e. max-
imising national or global wealth?).
51
1036.
52
Repetti, ‘Democracy and Opportunity: A New Paradigm in Tax Equity’ (2008), 1130–1,
suggesting that eiciency can be measured, whereas equity can be diicult to measure.
53
Graetz, ‘Distributional Tables, Tax Legislation, and the Illusion of Precision’ (1995), 18.
54
55
Australia, Australia’s Future Tax System (2009), 32.
56
hose who give more weight to distributive justice relected in the ability to pay prin-
ciple adopt a limited concept of eiciency by accepting that a particular tax, rather than
a particular tax system, ought to maximise eiciency where all else is equal: Fleming
and Peroni, ‘Reinvigorating Tax Expenditure Analysis and its International Dimension’
(2008), 461.
5.3.2.8 Empirical matters

here are two elements to the empirical limitations of eiciency conceived
in terms of wealth maximisation. he irst is that the behavioural con-
struct of ‘economic man’ which underpins the wealth maximising con-
cept of eiciency does not necessarily relect observed human behaviour.
Secondly, the behavioural response of a hypothetical economic man to a
tax cannot be predicted with certainty.
he economic literature contemplates two possible responses to the
imposition of a tax upon earned income. hose subject to a tax might sub-
stitute leisure for income earning activity (the substitution efect). Such
substitution might trigger deadweight loss as each individual’s respect-
ive substitution of leisure for income might be founded upon an exces-
sive valuation of leisure relative to income earning, thereby creating a net
social wealth loss. Alternatively, individuals subjected to a tax might work
harder to make up their income to its pre-tax level (the income efect).
As Frank notes, the evidence regarding the degree to which one of these
behavioural responses is triggered in preference to the other by a particu-
lar tax rate at a particular point in time is inconclusive.57 Notwithstanding
these evidential concerns, the supply side emphasis upon the substitution
efect is taken to be the mainstream view.58
5.3.2.9
Application of wealth maximisation in the
real world of the second best
Eiciency deined in terms of wealth maximisation adopts the neoclassical
view that perfectly functioning markets are the best means for a society
to allocate scarce resources. However, we live in an imperfect world where
imperfect information, uneven distributions of market power and external-
ities render markets an imperfect means of real world resource allocation.
Neoclassical economists accept that public interventions to correct
these market weaknesses are appropriate. However, there is considerable
diiculty in determining the point at which a public intervention such as
a tax expenditure ceases to correct for market failure and becomes a drain
on social resources.59
5.3.2.10 Time frame over which eiciency is determined

Eiciency in terms of wealth maximisation is typically considered in a
static form – identifying the wealth maximising settings at any particular
57
Frank, ‘Progressive Taxation and the Incentive Problem’ (2000).
58
Slemrod, ‘he Economics of Taxing the Rich’ (2000), 25.
59
See, for example, Australia, Australia’s Future Tax System (2009), section E2.
point in time. However, the concept of dynamic eiciency acknowledges

that the measurement of wealth needs to be undertaken over the long run.60
Determining the relevant time frame for this purpose is problematic.
Moreover, the intergenerational aspect of maximising aggregate social
wealth is problematic, relected in the debate as to whether future social
costs should be discounted and, if so, what that discount rate ought to
be.61
A compounding factor is how to treat variations in population size
when computing aggregate social wealth over time.62
5.3.2.11 Geographical scope

here is no speciication of the geographical limits when measuring
‘social’ wealth63 nor is there any clear speciication of who should be
included when measuring social wealth.
5.3.2.12Eiciency as wealth maximisation and its relevance

to tax expenditure management
Universal truth may be the aspiration of economics; however, the applica-
tion of the particular ‘mainstream’ concept of eiciency considered here
remains controversial. Adoption of the wealth maximising concept of
eiciency entails acceptance of a particular construct of ‘the good soci-
ety’. Further, pursuit of this contingent construction of the good society
entails assumptions regarding the geographical and temporal scope of
the eiciency objective. Finally, there are considerable empirical diicul-
ties associated with the application of eiciency as wealth maximisation
in any particular context.
Notwithstanding these conceptual and empirical diiculties, some
commentators see eiciency conceived in terms of wealth maximisa-
tion as the basis upon which an ‘objective’ appraisal of tax expenditures
might be undertaken. Libertarians object that such an enterprise can only
be highly speculative,64 and we agree with this scepticism regarding the
prospects of wealth maximisation to procure the good society. However,
60
Abel et al., ‘Assessing Dynamic Eiciency: heory and Evidence’ (1989).
61
See Weisbach and Sunstein, ‘Climate Change and Discounting the Future: A Guide for
the Perplexed’ (2009); see also Harrison, Valuing the Future: he Social Discount Rate in
Cost-beneit Analysis (2010). Ater reviewing the literature, Harrison observes that rec-
ommended discount rates range from 1 per cent to 15 per cent.
62
For discussion of this point with respect to the application of welfare economics, see:
Kaplow, he heory of Taxation and Public Economics (2008), 387.
63
Kaplow, he heory of Taxation and Public Economics (2008), 379.
64
we also agree with ‘economic democrats’ who express concern regarding

the unmitigated pursuit of wealth maximisation without regard to other
social objectives such as fairness. On this we part ways with libertarian
critics of wealth maximisation as a public policy norm.
5.3.3 Democracy – distributive justice

In the preceding section we saw that the mainstream construction of
‘eiciency’ as wealth maximisation is agnostic regarding distribution of
resources which arises from market/non-market actions. A state might
intervene in private allocations of social resources in pursuing a fair dis-
tribution of social resources. he tax expenditure reports of many coun-
tries do not assess the tax base and/or the tax rate against a redistributive
norm. Even where distributive analysis of tax expenditures is undertaken,
as in the United States, only limited information regarding a small num-
ber of tax expenditures is available.
Leaving to one side the atomism/holism issue already considered in
section 4.2, there are three components to the proposition that a govern-
ment ought to examine the distributive impact of tax expenditures in
pursuing a redistribution of social resources:
1. Whether the state ought to assume a role in redistributing social
resources;
2. If so, whether taxation is the appropriate mechanism for pursuing such
a redistribution; and
3. If the answer to both of the preceding questions is airmative, what
the relevant redistributive norm ought to be and its relevance to tax
expenditure management.
5.3.3.1 Does the state have a redistributive role?

he i rst question to be resolved is whether the state ought to redis-
tribute wealth within a community.65 In the context of societies with
democratic aspirations, this entails consideration of what it means to
be ‘equal’ in an economic sense, rather than restricting the concept of
democracy to political equality on the assumption that the two can be
disconnected.66
65
Expressing the proposition in this manner implies that wealth distribution pre-exists the
state, an assumption challenged by Murphy and Nagel, he Myth of Ownership (2002).
66
his would be a bold assumption, as vast inequality of wealth can be expected to manifest
itself in unequal political power.
As we noted in our consideration of the concept of eiciency,67 Rawls’

interpretation of ‘equality’ supported a construction of ‘equality’ which
emphasised equality of opportunity in preference to equality of outcomes.
he diference principle acknowledges the least well of by requiring that
their circumstances be improved, but leaves the relative distribution of
wealth untouched by the public hand. A libertarian considers that the
state ought not to redistribute wealth within a community upon the basis
that the individual is entitled to any wealth that they might ‘legitimately’
muster.68 Some counter this proposition by denying that an individual
can own property in a pre-state state of nature, such that redistribution of
that property misdescribes progressive taxation.69 On this view, progres-
sive taxation is justiied not as a state-orchestrated redistribution of assets
between private parties but as a fee charged for the use of community
assets or as akin to a proit sharing arrangement under which the individ-
ual shares a portion of their proits with the community as co-venturer.70
However, on the basis that redistribution of property by the state exists,
many are prepared to accept that the state ought to redistribute wealth
from ‘rich’ to ‘poor’, although the extent of the redistribution is let rather
vague. his redistributive function of the state has been justiied on the
grounds of ‘economic democracy’,71 fairness,72 justice in terms of eco-
nomic eiciency,73 justice in terms of substantive outcomes being consist-
ent with the most coherent interpretation of a community’s principles,74
67
See section 5.3.2.
68
For example, Locke expressed two provisos regarding the right of individuals to
secure property: Locke , ‘The Second Treatise on Civil Government’ (1947), ch. V,
paras 27, 29, 31.
69
Murphy and Nagel, he Myth of Ownership (2002).
70
Kornhauser, ‘Tax Policy in an Era of Rising Inequality: Choosing a Tax Rate Structure in
the Face of Disagreement’ (2005), 1726–7.
71
Repetti, ‘Democracy and Opportunity: A New Paradigm in Tax Equity’ (2008).
72
Dodge, ‘heories of Tax Justice: Ruminations on the Beneit, Partnership and Ability to
Pay Principles’ (2005).
73
Repetti, ‘Democracy, Taxes and Wealth’ (2001), 837f: arguing that redistribution in the
form of education can enhance economic growth as compared to a non-redistributive
government programme.
74
Kornhauser, ‘Tax Policy in an Era of Rising Inequality: Choosing a Tax Rate Structure in
the Face of Disagreement’ (2005). Kornhauser seeks to induce an ‘ought’ from the ‘is’ –
redistribution is justiied as the most coherent interpretation of the United States’ com-
mitment to fairness understood in terms of equality of opportunity. Although portrayed
as autonomous from ickle community opinion relected in opinion polls and/or from an
existing legislated scheme, Kornhauser’s conclusion would seem to reach no further than
the status quo which incorporates the principle that entrepreneurial endeavour ought to
be rewarded and not penalised by high taxation.
political pragmatism in the sense of keeping mass discontent in check,75

‘intuitive appeal’76 and historical grounds (at least for some countries).77
5.3.3.2 Redistribute by taxation?

If the irst proposition is accepted, the second question to be addressed is
whether the tax system is an appropriate tool for achieving this redistri-
bution objective, and in some circumstances it may be both more efective
and more eicient not to use the tax system for this purpose.78
For example, a non-redistributive tax system might i nance redis-
tributive spending measures. Some advocates of the consumption tax
argue that the redistributive objective can be achieved under both con-
sumption and income taxes but that the consumption tax ofers both
eiciency and simplicity advantages;79 this argument relies upon accept-
ance as a fact that high wealth individuals will consume all of their
wealth in their respective lifetimes such that the consumption value
of wealth itself will be taxed in the hands of the appropriate entity.80
Moreover, bold assumptions must be made regarding the simpliica-
tion savings of a multi-tier value added tax over an income tax, and it
is likely that the administrative costs will be born disproportionately,
75
See, for example, Avi-Yonah, ‘h ree Goals of Taxation’ (2006) 17–18. Of course, prag-
matism entails the selection of a desired objective which imports a deontological aspect.
For example, a welfarist approach might justify redistribution of wealth upon the basis
that to do so enhances the aggregate social utility (Rawls). Alternatively, a pragmatic
approach might value social order as a i rst order priority and see the elimination of
‘excessive’ inequality of wealth through redistribution as central to this objective. On
this latter point, there are competing views regarding the adoption of the US income tax.
Some see it as the triumph of democratic social values: Ventry, ‘Equity versus Eiciency
and the US Tax System in Historical Perspective’ (2002). However, others see a more cyn-
ical manipulation of the public mind in line with Steven Lukes’ third projection of social
power by which the public mind may be subject to manipulation: Lukes, Power: A Radical
View (2005). An illustration of this third dimension of power can be seen in Robert S.
Stanley’s work. Stanley argues that the apparent progressivity of the income tax is a cen-
tral plank in maintaining a social order which beneits the well-of: Stanley, Dimensions
of Law in the Service of Order (1993).
76
Simons, Personal Income Taxation (1938), 18–19; Blum and Kalven, he Uneasy Case for
Progressive Taxation (1953), 135–7.
77
In the United States, for example, the income tax was part of a rat of measures enacted
with the intention of either symbolically or really combating excessive concentrations of
wealth: Howard, he Hidden Welfare State (1997), 48–53.
78
Bird and Zolt, ‘Redistribution via Taxation: he Limited Role of the Personal Income Tax
in Developing Countries’ (2005).
79
Bankman and Weisbach, ‘he Superiority of an Ideal Consumption Tax Over an Ideal
Income Tax’ (2006), 1413.
80
Avi-Yonah, ‘he hree Goals of Taxation’ (2006), 12–13.
particularly given the economies of scale achieved by taxpayers with

larger taxable turnover.
5.3.3.3 Elaboration of normative tax redistribution principle

If both of the irst two propositions are accepted, the third component
is that a normative redistribution principle must be speciied and be
included in the tax system benchmark. he nature of this principle is not
resolved. For present purposes commentators who see the tax system as
a crucial redistributive tool favour income taxation, estate taxation and
also capital transfer taxes.
Although Surrey demurred on this point,81 some tax expenditure advo-
cates explain this preference upon the basis that the Schanz–Haig–Simons
concept of income ought to be adopted both as the starting point for def-
inition of the income tax base and also as at least the starting point for
deining the income tax benchmark for the purposes of tax expenditure
management.82 he SHS advocates claim that it best measures a person’s
ability to contribute to public revenue and therefore plays a crucial role
in furthering a redistributive function which legitimates adoption of the
income tax as at least one signiicant source of state revenue.83
However, as Andrews observed84 and Surrey appeared to agree,85 the
Schanz– Haig–Simons concept of income is problematic in that it assumes
that individuals are equally endowed such that expenditure upon cor-
rective surgery is not allowed as a deduction under an income tax base,
notwithstanding that such expenditure might place the person in a com-
parable position to a ‘normal’ taxpayer. his critique of income as a basis
for comparing a person’s capacity to contribute to public revenue has been
extended by Sen.86
According to Sen and others, the income concept inadequately meas-
ures capacity because equality of opportunity requires a consideration
of a person’s opportunities and capabilities to become the person they
might aspire to become,87 and this might entail a redistribution of societal
81
Surrey and McDaniel, Tax Expenditures (1985), 3–4, 186f.
82
83
Dimension’ (2008), 450–60.
84
Andrews, ‘Personal Deductions Under an Ideal Income Tax’ (1972).
85
86
Sen, he Idea of Justice (2009), ch. 11.
87
Sen, he Idea of Justice (2009), for discussion of which see below.
Democracy as political procedure 169
resources from the well endowed to those who ind themselves in circum-
stances which limit their capacity to achieve their ambitions.88
5.3.3.4 Concluding comments

If all three propositions are accepted (and the preceding discussion indi-
cates that there are many disagreements with respect to how each of
the three questions identiied at the beginning of this section should be
answered), tax expenditure management would be appropriate. From this
standpoint, tax expenditure management would focus upon managing
the tax aspect of the relevant state’s redistributive function by establishing
a benchmark redistribution norm and assessing individual taxes, or an
entire taxation system, against that norm. his approach to tax expend-
iture management difers markedly from the approaches informed by lib-
ertarianism and eiciency as wealth maximisation respectively.
5.4 Democracy as political procedure – accountability and

transparency
In view of the ongoing disagreement concerning the application of sub-
stantive norms of freedom, eiciency and fairness illustrated in the pre-
ceding part of this chapter, some suggest that procedural decision making
rules promise an objective foundation for resolving conlicts between
incommensurable substantive imperatives.89
Even for those who hold to a particular substantive perspective (liber-
tarianism, eiciency, redistribution), democratic procedural rules must
be considered if the taxation system does not accord with the ‘right’ nor-
mative tax system from the particular critical perspective adopted. In this
context, the apparent failure of democratic institutions to reach the right
outcome must be explained by democratic failure, for which the respect-
ive substantive theorists ought to prescribe a procedural remedy.
However, procedural rules are not value-free and the nature of the pro-
cedural rule adopted will shape the approach to tax expenditure manage-
ment. For the purposes of our study, we have focused upon the elaboration
of procedural rules where the nation claims to be a democracy or at least
has democratic aspirations. From the perspective of political justice,
88
As we have seen earlier in this chapter, this is the positive aspect of liberty, which libertar-
ians such as Hayek reject upon the basis that it embroils the state in arbitrary decisions
allocating social resources.
89
he aspiration to a neutral constitutional framework is a vexed issue within liberal con-
stitutional theory: Gray, Two Faces of Liberalism (2000), ch. 1.
democratic government entails acceptance of the equality of individuals

in terms of their capacity to inluence public decisions.90 As we saw in
the case of distributive justice, so in the case of political justice, diferent
procedural rules relect difering constructions of what political equality
means. For example, political equality can be constructed in terms akin
to equal opportunity to participate in the decision making process or it
might be taken to require the equal participation and power of all mem-
bers of the polity with respect to all decisions.91 Indeed, Held identiies
nine models of democratic government, having regard to the means by
which public oicials are appointed, the role of those oicials in making
‘public’ decisions and the degree to which the citizenry is involved in or
capable of scrutinising decision making by public oicials.92
With respect to these matters, two key concepts are accountability and
transparency. In broad terms, accountability entails the ability of one
person to call another to account.93 In the context of public administra-
tion, this means providing an explanation for oicial action (or inaction)
and being susceptible to rewards/penalties with respect to one’s conduct.
Again, in the context of public administration, in broad terms transpar-
ency entails the provision of information regarding how and why oicials
behaved in a particular fashion.
he purpose of this section of the chapter is to illustrate the diversity
of views regarding these core democratic concepts, with the purpose of
demonstrating the absence of any universally accepted procedural rules
which might aford a basis for breaking the deadlock between competing
substantive principles that underpin alternate approaches to tax expend-
iture management.
5.4.1 Transparency – for whom and for what purpose?

Chameleon-like, the concept of transparency can assume a political and
also an economic hue. he purpose of this section is to illustrate the lexi-
bility of the concept in demonstrating that the adoption of a particular
concept of transparency is founded upon a contingent conception of the
public good. Diferent conceptions of the public good underpin compet-
ing indications of whether transparency advances the public good and, if
90
Held, Models of Democracy (2006), 1.
91
However, if unanimity is required, the last voter is efectively allowed a veto right which
would aford them more power than the irst voter.
92
Held, Models of Democracy (2006).
93
Dowdle, ‘Public Accountability: Conceptual, Historical and Epistemic Mappings’ (2006).
so, how that public good is best procured.94 Further, diferent approaches
to the second issue entail quite diferent procedural rules. In the absence
of an objective means of determining which concept of transparency
ought to prevail, the apparent objectivity of procedural rules designed to
enhance ‘transparency’ is illusory.
he luidity of the transparency concept is relected in the deinition of
the concept adopted by the International Monetary Fund, where trans-
parency is deined as:
openness towards the public at large about the structure and functions,
iscal policy intentions, public sector accounts and projections. It involves
ready access to reliable, comprehensive, timely, readily understandable
and internationally comparable information on government activities
… so that the electorate and i nancial markets can accurately assess the
government’s inancial position and the true costs and beneits of govern-
ment activities, including their present and future economic and social
implications.95
his deinition relects two perceived purposes for transparent public

decision making – informing inancial markets and also informing vot-
ers.96 he two concepts are not co-extensive in terms of the type of infor-
mation that is necessary for their respective fuli lment, and also with
respect to the manner in which information is provided. A choice must be
made as diferent resource inputs will be required.
he International Monetary Fund97 and some countries98 at times seem
to adopt the more limited concept of transparency. Emphasising this aspect
94
Rubin, for example, argues upon pragmatic grounds that public accountability con-
ceived in terms of shit ing the constitutional responsibility for the executive power to
elected oicials is misconceived: Rubin, ‘he Myth of Non-Bureaucratic Accountability
and the Anti-Administrative Impulse’ (2006). Under the Westminster system of govern-
ment, however, at least in theory elected oicials are held accountable for the exercise of
the executive power of government.
95
Kopits and Craig, Transparency in Government Operations (1998), 1. h is dei nition is
adopted in IMF, Manual on Fiscal Transparency (2007), 127.
96
Stewart and Philipps, ‘Dei ning Fiscal Transparency: Transnational Norms, Domestic
Laws and the Politics of Budget Accountability’ (2009); Burton, ‘Is Participatory Tax
Transparency Achievable?’ (2006).
97
See, for example, IMF, Manual on Fiscal Transparency (2007), para 3: ‘in response to
an increased emphasis on ensuring the provision to the IMF and to markets of the best
available economic and i nancial information, it has become apparent that more atten-
tion needs to be paid to the data quality aspect of iscal transparency’; IMF, Assessing
and Promoting Fiscal Transparency: A Report on Progress (2003), para 3; Hameed, Fiscal
Transparency and Economic Outcomes (2005), 6.
98
Note the approach to tax expenditure reporting adopted in New Zealand, described in
Chapter 3.
of Kopits and Craig’s deinition would mean that the imperatives of the
international ‘inancial architecture’99 inform the minimum transparency
standard, rather than any domestic imperative such as the capacity of the
electorate to participate, and have conidence in, public policy formation.
By contrast, at other points ‘democratic transparency’ is acknowledged
to require a broader array of information accessible to the general public
and presented in a manner which would enable an interested member
of the general public to inform themselves regarding the merits of any
particular tax expenditure.100 It is assumed that more information is bet-
ter.101 his would be a far more onerous, resource-demanding standard
for government, but one which we suggest is consistent with contempor-
ary concepts of representative democracy in a ‘post-positivist’ era where
conidence in the science of government is less secure.102
In a world of scarce resources, a choice must be made regarding the
tradeof between these competing concepts of transparency. From the
substantive perspective of allocation eiciency understood in terms of
wealth maximisation,103 the public cost of gathering and disseminating
information in general will be i nanced by taxation. If it is accepted
that taxation triggers deadweight costs, the addition of each extra
unit of information triggers the need for additional taxation which in
turn triggers additional deadweight economic losses.104 If economic
eiciency is understood in terms of aggregate wealth maximisation,
the principle indicates that there is a tradeof between the public bene-
it to be gained from government transparency and the deadweight
costs associated with providing that information.105 As we noted in
99
Managing Director of the IMF, Speech at the 24th Annual Conference at the International
Organisation of the Securities Commission, Lisbon, 25 May 1999, reported in IMF
Survey, 7 June 1999.
100
See, for example, IMF, Manual on Fiscal Transparency (2007), 76: ‘Ideally, the estimated
results of previous tax expenditures compared with their policy purposes should also be
presented so that their efectiveness can be assessed relative to expenditure provisions.’
101
Shaviro, ‘Rethinking Tax Expenditures and Fiscal Language’ (2004), 219. However, the
literature upon the subject is equivocal as to the willingness and capacity of individuals
to process this additional information: Bimber, Information and American Democracy
(2003), ch. 5.
102
deLeon, Democracy and the Policy Sciences (1997); deLeon and Vogenbeck, ‘he Policy
Sciences at the Crossroads’ (2007).
103
See section 5.5.
104
Australia, Architecture of Australia’s Tax and Transfer System (2008), 311.
105
Downs, An Economic heory of Democracy (1957), 215. his point deserves speciic men-
tion, as it is overlooked by some commentators. See, for example, Shaviro, ‘Rethinking
Tax Expenditures and Fiscal Language’ (2004), 219.
Chapter 4,106 the diiculty of identifying and measuring all social well-
being factors oten leads to a concentration upon those factors that can
be measured upon the basis that those factors are a valid proxy for all
social well-being. h is approach may lead some to the view that ‘mar-
ket transparency’ is both necessary and suicient for government to
comply with the democratic concept of transparency, with political
transparency ofering no independent social value.
5.4.2 Elitism
Even if ‘market transparency’ were to be rejected by adopting concepts of
accountability and transparency which concentrate upon the interaction
of voters and government in a democracy, once again there are competing
interpretations of these concepts which relect competing conceptions of
how the public good is best identiied and pursued through government.
Given the pragmatic necessity of representative government in a ‘mass’
democracy, two key democratic issues are whether the political represent-
atives are bound to relect the ‘popular will’ when making public policy
decisions and, if so, how this requirement is to be assured as far as pos-
sible by procedural rules. For our purposes it is suicient to identify what
might loosely be described as elitist and popular concepts of accountabil-
ity which present answers to these questions. Considering contempor-
ary democratic government, Dowdle argues that recent US experience
relects a tension between elitist and republican democratic models, that
these models entail quite diferent concepts of accountability and that the
elitist concept of accountability is predominant.107
According to elitist accounts of democratic government, elected rep-
resentatives are elected to govern for a i xed term and at the end of that
term they are held accountable for their decisions. he mechanics of this
form of accountability are not always overtly stated. One strand of elitism
holds that unelected technical experts within the executive arm of gov-
ernment should have considerable inluence over the decisions of elected
representatives.108
hus, writing in the 1960s, C. Wright Mills observed that the public
participation in the running of government was insigniicant relative to
106
See the discussion of well-being indicators in section 4.3.
107
Dowdle, ‘Public Accountability: Conceptual, Historical and Epistemic Mappings’
(2006), 4.
108
h is version of elitism is discussed in Wildavsky, he New Politics of the Budgetary
Process (1988).
the might of the modern bureaucracy.109 At the time of Mills’ work, there
was a widely held view that the general public was too unintelligent and/
or too disinterested to ofer much if the object of public policy was pur-
suit of the public good.110 Although he was mindful of Weber’s concern
regarding the power of the bureaucratic state, Mills saw this as an oppor-
tunity for a bureaucratic elite to implement public policy formed upon the
basis of a consensus among the intellectual and bureaucratic elite.111
In the context of tax expenditure management one of the key legacies
of Surrey’s work has been the adoption of an elitist vision of democratic
government. It may be that this elitism was prompted by the pragmatic
recognition of the fact that ‘public interest’ groups were not equipped to
tackle tax spending at the time Surrey was developing the tax expend-
iture concept.112 Surrey was conident enough with an elitist model of
bureaucratic rationality that he considered preparation of a tax expend-
iture statement and consideration of that statement by executive gov-
ernment to be suicient to at least bring many tax expenditures under
close scrutiny led by technical experts who, he suggested, would adopt
a consensus view as to the underlying public policy norms. Implicitly,
Surrey doubted that many tax expenditures would survive such scrutiny,
although he clearly acknowledged that there may be a good case for some
tax expenditures.113
he impact of Surrey’s legacy, of tax expenditure reporting simpliciter,
is diicult to assess given the opacity of budget deliberations. It is diicult
to determine the extent to which the relatively rudimentary tax expend-
iture reporting that has operated over the past decades has prompted tax
expenditure analysis, retrenchment of tax expenditures and/or redesign
of tax expenditures. Nevertheless, Surrey’s conidence that a policy elite
had the capacity and the will to right public policy wrongs remains a
widely held view.
Anecdotally, public oicials who profess to act in the public interest
express concern regarding the provision of detailed information to the
general public, lest special interests utilise that information to undermine
109
Mills, Power, Politics and People (1963), 237.
110
See, for example, Schumpeter, Capitalism, Socialism and Democracy (1947), 262.
111
Mills, Power, Politics and People (1963), 246.
112
See, for example, the discussion of nascent public interest group involvement in
the United States budgetary process of the 1970s: Surrey and McDaniel, ‘he Tax
Expenditure Concept and the Budget Reform Act of 1974’ (1976), 724–5.
113
Developed Countries’ (1983), 757.
the apparent universality of a particular public policy and thereby advance
their respective partisan causes. Further, public oicials conident in their
capacity to act in the public interest might see consultation as an unneces-
sary step in public policy formation. Witte, for example, argues that tax
policy ought to be insulated from the representative process rather than
being exposed to representative pressures which, he fears, too readily
become special interest pleading.114 Taking this elitist view one step fur-
ther, Schenk argues that the opacity of tax expenditures would enable
their repeal – the lower their political salience the less likely it is that they
will be missed if repealed.115 Commentators such as Witte and Schenk
seem to accept that the law making process in the United States has been
too democratic, and has therefore let the lobbyist ‘barbarians’ inside the
law making citadel. According to this view, the remedy for this failing is
to centralise law making power in the hands of the few who, it seems to be
assumed, would champion the ‘right’ conception of the public good.
We consider that this conidence in a policy elite willing and able to
right the listing tax expenditure ship is misplaced, for four reasons:
1. From the perspective of liberal democratic political theory, concentra-
tion of law making power in a policy elite is not appealing. According
to this prescription, ‘the people’ are an irrelevance at best and, at worst,
an obstruction to ‘correct’ policy framed upon the basis of one of the
(contingent) social norms discussed in section 5.3 of this chapter.
he elitist prescription seems to assume that an altruistic policy elite
either exists or will spring into the breach at the right time. Public serv-
ants may not be uniformly saints nor sinners, but they are as subject
to human weakness as the remainder of the population. In our view,
exposing their conduct in public oice to public scrutiny would tend
to enhance the rigour of their justiications for public action,116 as they
would conduct themselves knowing that those justiications would be
subjected to critical scrutiny. Experience suggests that self-interested
114
Witte, he Politics and Development of the Federal Income Tax (1985), 381. To similar
efect, see Zelizer, Taxing America: Wilbur D. Mills, Congress, and the State 1945–1975
(1998), 310–11; Steuerle, Contemporary US Tax Policy (2004), 250; McLure, ‘he Budget
Process and Tax Simpliication/Complication’ (1990), 77–9, 89–90.
115
Schenk, ‘Exploiting the Salience Bias in Designing Taxes’ (2011).
116
See, for example, the ‘Demand for Good Governance’ literature which emphasises the
importance of demand for good governance alongside ‘supply-side’ governance strat-
egies (including governance-enhancing institutions): World Bank, Demand for Good
Governance in the World Bank: Conceptual Evolution, Frameworks and Activities
(2010).
public oicials may have little interest in exposing their policy short-
comings to public scrutiny.117
2. From the perspective of normative tax policy, the elitist prescription
seems to assume that there is one right answer which a policy elite will
agree upon, despite the evidence to the contrary apparent from our
consideration of the tax expenditure concept in Chapter 2. Critics of
the tax expenditure concept have therefore hardened their stance over
the years – from Bittker’s grudging acceptance that the concept may be
of some use118 to more recent strident rejections of the concept’s util-
ity.119 Indeed, this perceived failure of a policy elite to agree upon a uni-
versally defensible solution to policy problems is a phenomenon which
has prompted a sense of crisis within the ‘policy sciences’ more gener-
ally.120 If public policy cannot proceed upon a universalist or relativist
foundation, the prospect of a technocratic elite achieving bureaucrat-
ically rational tax expenditure management appears remote at best.
3. From the perspective of political science, the elitist prescription ignores
the possibility that tax expenditures might have come into being by
virtue of the opacity of tax expenditure management processes and
the opacity of the tax expenditure rules. It is quite conceivable that
this opacity has been exploited by lobbyists and politicians and so the
same forces might combine to sustain tax expenditures unless pub-
lic scrutiny is the catalyst which exposes the particular tax expend-
iture to scrutiny. Corruption and political inluence can arise in the
most transparent of governments, but the risks and costs are higher for
those engaging in such behaviour by comparison to less transparent
political arenas. Prudent management of the risk of corruption and
political inluence suggests that Surrey’s faith in this elitist, bureau-
cratically rational model of democratic government is misconceived,
as tax expenditures have grown despite considerable efort devoted to
reporting them.121
4. From the perspective of sociology, the simplistic elitist model of
accountability also ignores the role of second order factors in shaping
the context within which individuals exercise political power. Larry
117
118
Bittker, ‘Accounting for Federal Tax Subsidies in the National Budget’ (1969), 260–1.
119
Bartlett, ‘he End of Tax Expenditures as We Know hem?’ (2001).
120
121
(2009).
Patriquin argues that public media can inluence public perceptions of

taxation law,122 although this may be a case of the media both echoing
and entrenching deeper communal narratives.123
Also from the perspective of sociology, the irrelevance of the general
populace to the exercise of political power by an elite between elections
is not convincing when one considers the tenuous nature of central-
ised power. he history of tax rebellions124 indicates that even if there is
a ruling elite, it does not hold absolute power. Always, power is contin-
gent in a way which means that a ruling elite must always have one eye
on ‘the popular mood’. Even between elections, a hypothesised ruling
elite could not exercise absolute power as the public has the capacity to
ignore the usual electoral cycle by adopting a rebellious posture. In the
main the general public might accept the electoral cycle, but this does
not make the public politically irrelevant.125
here are both normative and descriptive reasons to reject an elitist
approach to tax expenditure management. Normatively, the discussion
of morally incommensurable standpoints in sections 5.2 and 5.3 indicates
that an elite cannot claim to have discovered the one true path to the one
right good life. here are diferent conceptions of the social good and they
are incommensurable. How can libertarianism and egalitarianism be
reconciled within the one conception of the good life? Further, even if a
community could agree upon pursuit of a contingent vision of the social
good, sound representative government requires active review of repre-
sentatives’ conduct undertaken in the name of all. Finally, elitism ofers
a weak descriptive account of social behaviour because it downplays the
ultimate social power held by ‘the general public’ and that is most evident
in times of social unrest.
5.4.3 Elitism with a republican facade: majoritarian

democratic positivism
Although not expressly an elitist theory, a form of elitism with a republican
facade can be found in ‘democratic’ interpretations of legal positivism.
122
Patriquin, Inventing Tax Rage (2004).
123
Kornhauser, ‘Legitimacy and the Right of Revolution: he Role of Tax Protests and
Anti-Tax Rhetoric in America’ (2002).
124
Burg, A World History of Tax Rebellions (2004).
125
For the same reason, the public choice account of the relative insigniicance of ‘ordinary’
voters fails the positivist standard upon which public choice theory is founded.
In general contemporary legal positivists accept that the authority of law

derives from the fact that it is posited in accordance with a community’s
‘rule of recognition’126 – those requirements that must be satisied for a
norm to be recognised as a legal norm.
5.4.3.1 hin and thick legal positivism

As with any label, ‘positivism’ must be used cautiously because there are
diferent strands of positivism. For our purposes it is suicient to draw a
distinction between ‘thin’ and ‘thick’ positivism.
According to ‘thin positivism’, the rule of recognition is restricted
to manner and form requirements necessary to preserve democratic
accountability. Most particularly, thin positivists maintain that the rule
of recognition does not specify compliance with ethical norms.127 By con-
trast, ‘thick’ or ‘inclusive’ positivism holds that the rule of recognition
may import moral criteria.
5.4.3.2 he democratic aspect of democratic positivism

he normative appeal of ‘thin’ legal positivism is couched in democratic
terms.128 According to this view, law making ought to be let to the legis-
lative forum where law making is transparently undertaken by account-
able, elected representatives. hus, the legitimacy of the law, and hence the
foundation of maintaining social order, is assured.129 he debate between
thin and thick positivism is central to the question of whether the validity
of a law is subject to moral constraints and hence to the role of the courts,
and of advisers, in identifying the law.
hin democratic positivists maintain that the role of judges should be
restricted to the impartial interpretation and application of the objective
meaning of those laws because judicial accountability is limited.130 If the
126
Hart, he Concept of Law (1994), 114–17.
127
For example, Waldron argues for a majoritarian procedural rule whereby rules are
authorised as legal rules, although he leaves much to be resolved. For example the ques-
tion of representative/direct democratic models is not dealt with adequately. A thin
positivist accepts that a valid law may import moral elements but does not accept that
the validity of that law itself depends upon satisfaction of moral criteria.
128
Waldron, he Dignity of Legislation (1999), 2; Waldron, ‘Representative Lawmaking’
(2009), 338; Goldsworthy, ‘Preface’, in Legal Interpretation in Democratic States (2002),
xi; Campbell, ‘Grounding heories of Legal Interpretation’ (2002), 33.
129
Goldsworthy, ‘Legal Intentions, Legislative Supremacy, and Legal Positivism’ (2002).
In the context of Australian taxation see, for example, Carmody, ‘Ethics and Taxation’
(1999).
130
Goldsworthy, ‘Preface’, in Legal Interpretation in Democratic States (2002), xi.
rule of recognition imports moral criteria, do unelected judges apply their

construction of a society’s moral norms in determining the status of a
rule that has been created by a democratically elected legislature?
here is debate as to whether a textual131 or an intentionalist interpret-
ation132 of the authorised legal text is appropriate in discovering just what
was resolved by the legislature. In any case, a thin positivist considers that
it is better that the accountable law making forum determine, with all of
its peccadilloes, the instances in which state power is exercised than that
law making be undertaken by the judiciary or the executive.
5.4.3.3Descriptive and prescriptive accounts of

democratic positivism
Some within the mainstream fold appear to adopt this legal positivism as
both a descriptive and a normative model of law,133 while others suggest
that its descriptive power is weak and so this sub-group accepts that legal
positivism is largely normative.134
Descriptive democratic positivism Descriptive democratic positivism

accepts the mainstream portrayal of the judicial role by taking as a given
that contemporary ‘democratic’ communities are in fact democratic and
that judges are currently applying ‘the law’ enacted by a ‘democratic’
legislature.
By inviting acceptance of these propositions, descriptive democratic
positivists avoid debate regarding the democratic legitimacy of positive
law because, they say, there must be an obligation to obey rules which
have the ‘right’ formal pedigree irrespective of any peccadilloes alicting
the democratic legitimacy of the process by which those rules came into
being. To a descriptive democratic positivist, a community that is ignor-
ant of inequitable, ineicient, inefective and complex tax expenditures is
as obligated to obey a tax law as a community which was informed of per-
verse tax rules but which accepted or acquiesced in such perversity.
Prescriptive democratic positivism Prescriptive democratic positiv-

ism acknowledges that there may be a gap between contemporary judicial
practice and the positivist ideal. he prescriptive form of legal positivism
131
Waldron, he Dignity of Legislation (1999), 26–9.
132
Goldsworthy, ‘Legislative Intentions, Legislative Supremacy, and Legal Positivism’
(2002).
133
See, for example, D’Ascenzo, ‘he Rule of Law: A Corporate Value’ (2007), 1.
134
Campbell, he Legal heory of Ethical Positivism (1996), 2–4.
appears to assume a strong form of human agency in taking for granted
the proposition that human actors have a free choice between legal posi-
tivism and other legal paradigms. his standpoint of human agency is
signiicant for two reasons. First, it allows this group of legal positivists
to ignore any sociological explanation for why legal practice departs from
their prescriptive positivist model – the implication is that adoption of
a positivist/non-positivist paradigm is a matter of individual choice by
actors free of sociological factors which may shape and constrain avail-
able choices.135 Second, the normative claim that ‘democracy’ underpins
an obligation to obey legal rules brokered in a ‘democratic process’ begs
the question of what is a ‘democratic process’?
5.4.3.4he slender concept of democracy in democratic

positivism
Although democratic positivists make much of the democratic aspects
of positivist jurisprudence, there is only leeting reference to the type of
‘modern democracy’ envisaged. Notwithstanding the title of their book,
Goldsworthy and Campbell make only brief reference to the democratic
concept in their respective contributions to the book. Goldsworthy labels
judicial activism as ‘undemocratic’ without explaining why legislation is
necessarily ‘democratic’.136 Similarly, Campbell makes only cursory refer-
ence to ‘democracy’:
[I]f we take the view that legislative intent matters only in so far as the
legislature is democratically elected and textual meaning matters because
it is publicly accessible and pragmatically useful for the coordination of
social action, then we have a powerful combination of policy reasons for
justifying the assumption that legislators intend that citizens and adjudi-
cators utilise the texts the legislators enact in accordance with their con-
textually evident meaning.137
Waldron’s consideration of ‘democratic jurisprudence’ focuses upon ‘a

theory of law as it presents itself in democracies’,138 rather than a consider-
ation of the intersection of substantive democratic norms with the process
of law making. hus, Waldron suggests that a legislature will comply with
a transparency norm merely if it overtly makes law – the values applied
135
h is would be what Steven Lukes identiies as a third form of power: Lukes, Power: A
Radical View (2005).
136
Goldsworthy, ‘Preface’ in Legal Interpretation in Democratic States (2002), xi.
137
Campbell, ‘Grounding heories of Legal Interpretation’, (2002), 33.
138
Waldron, ‘Can here Be a Democratic Jurisprudence?’ (2009), 679.
in the making of that law need not be overt for the law to be law and nor
ought the conduct of law making be transparent.139 In his introduction to
he Dignity of Legislation, Waldron hints at a more substantial concept of
transparency:
In this volume, then, I am going to try to recover and highlight ways of
thinking about legislation that present it as a digniied mode of govern-
ance and a respectable source of law. I want us to see the process of legis-
lation – at its best – as something like the following: the representatives of
the community come together to settle solemnly and explicitly on com-
mon schemes and measures that can stand in the name of them all, and
they do so in a way that openly acknowledges and respects (rather than
conceals) the inevitable diferences of opinion and principle among them.
hat is the sort of understanding of legislation I would like to cultivate.140
However, notwithstanding this ‘democratic’ cloak, none of these por-

trayals of positivism incorporates a substantial theory of democratic
government because their proponents are committed to a thin version
of positivism. On one reading, a dictatorship would meet Waldron’s thin
standard of transparency, provided that the general public knew who the
dictator was. Even if Waldron intended to restrict his minimalist trans-
parency norm to cases where the lawmaker was ‘democratically elected’,
that standard invites deliberation upon the criteria to be applied in deter-
mining whether the relevant election was ‘democratic’.
As we have noted, the incorporation of such moral criteria deining
‘democracy’ within the ‘rule of recognition’ is anathema to thin positiv-
ism because such substantive constraints would embroil courts in moral
questions regarding the nature of ‘democracy’ if courts were called upon
to determine whether a rule truly was a ‘law’. For thin positivists, it is
suicient to adopt a limited concept of transparency to ground their criti-
cism of non-transparent law making by ‘activist’ judges.
he positivists’ argument for a majoritarian rule of recognition has noth-
ing to say, for example, with respect to the substantive conditions which
must be satisied before a majority’s decision is validly made. Goldsworthy,
for example, quotes with approval from the work of Justice Scalia, to the
efect that judges need not be troubled with democratic norms because
‘elections take care of that quite well’.141 his preparedness to equate
139
Waldron, ‘Representative Lawmaking’ (2009), 338.
140
Waldron, he Dignity of Legislation (1999), 2.
141
Goldsworthy, ‘Legislative Intentions, Legislative Supremacy, and Legal Positivism’
(2002), 43; citing Scalia, ‘Originalism: he Lesser Evil’ (1989), 862.
‘democracy’ with some form of public election, the elements of which can
vary dramatically (compulsory or voluntary voting, proportional represen-
tation or otherwise, etc.) ignores the many substantive considerations ger-
mane to the identiication of what truly constitutes a democracy.
Although legal positivists profess an interest in transparency, the con-
cept of transparency adopted by democratic positivists is so thin as to
be inefectual in advancing the ‘modern’ concept of democracy to which
they refer. If the principle of transparency is limited to identifying func-
tional allocations, such as assigning the task of law making to the legis-
lature, as Waldron suggests,142 then transparency about the principles
underlying that law making and the process by which the law is made is
immaterial. If democratic positivism truly is to have an ethical founda-
tion in democratic politics, more needs to be done in elaborating that eth-
ical foundation. For present purposes, this would require consideration of
the purpose of tax expenditure reporting (see section 2.3). To what extent
should tax expenditure reporting overcome the challenge identiied in
the public choice literature, of too many with too little ‘immediate’ indi-
vidual interest in both procuring relevant information and constructively
engaging in deliberation upon a community’s tax system?143 Of course,
incorporation of such ethical elements within the positivist’s ‘rule of rec-
ognition’ would contradict the core tenet of democratic positivism – the
exclusion of moral elements from the rule of recognition.
5.5 Ontological accounts of tax expenditure management

In section 5.3 it was argued that none of the deontological theories of
liberal democracy we surveyed underpins an approach to tax expend-
iture management that is necessarily right. In section 5.4 we noted that
‘objective’ procedural principles such as transparency and accountability
can be interpreted in various ways depending upon the vision of democ-
racy sought by applying those principles. Democratic positivists couch
their fundamental, procedural rule in democratic terms but ignore any
substantive justiication of why their approach ofers the most appeal-
ing form of democratic politics. Before examining democratic theories
which incorporate more substantial discussion of what counts as ‘demo-
cratic’, a brief excursus into descriptive accounts of tax expenditure pol-
itics is worthwhile so that we may understand more fully the limitations
142
Waldron, ‘Representative Lawmaking’ (2009), 338.
143
his aspect of the public choice literature is discussed below.
Ontological accounts of tax expenditure 183
of democratic positivism and also the challenges that the alternate demo-
cratic models considered in section 5.6 confront.
Two descriptive accounts of contemporary tax expenditure politics are
ofered by economists and by political scientists
5.5.1 Republican success – the public gets what it wants I: public choice
Informed by mainstream economic presuppositions regarding human
nature, public choice theory takes as a given that self-seeking individuals
will at least strive to maximise their welfare by rationally calculating the
net beneit of a proposed transaction (including opportunity costs).144
From this proposition public choice theory seeks to explain why two or
more such self-seeking individuals would agree to create a coercive state.
Moreover, public choice theory seeks to explain the shape, size and func-
tion of such a state as well as the extent to which individuals will partici-
pate in making public decisions145 – again from the assumption that there
is a community of rational individuals seeking to maximise their respect-
ive personal utilities.
he phenomenon of tax expenditures has been examined from the
public choice perspective. he general tenor of this literature is that
tax expenditures are explicable upon the basis of the core public choice
assumption that each individual will pursue their respective self-interest.
his literature focuses upon identifying the circumstances in which mem-
bers of coalitions might combine to procure beneits by various govern-
ment actions (i.e. tax expenditures) and at the expense of those who are
not members of the particular coalition. Although such a distribution of
public welfare to a relatively small group of individuals is a cost borne by
the remainder of the community, public choice theory suggests that indi-
viduals comprising the larger segment of the community will acquiesce
in bearing these costs because it is rational for them to do so. Although at
144
For critical appraisal of this premise of public choice theory see Rubin, ‘Public Choice,
Phenomenology, and the Meaning of the Modern State: Keep the Bathwater, But h row
Out the Baby’ (2002).
145
See, for example, Mueller, Public Choice III (2003), 1:
Public choice can be dei ned as the economic study of non market decision mak-
ing, or simply the application of economics to political science. he subject matter
of public choice is the same as that of political science: the theory of the state, voting
rules, voter behavior, party politics, the bureaucracy, and so on. he methodology of
public choice is that of economics, however.
See also Farber and Frickey, ‘he Jurisprudence of Public Choice’ (1987).
irst blush it might appear that their collective self-interest would prompt
this larger segment to unite against the plundering of the public cofers by
the minority, several factors combine to make acquiescence the rational
course of action in many cases. In particular, any iscal saving from over-
turning the tax expenditure redistribution to the minority would be dis-
persed across the taxpaying community at large. he transaction costs
incurred by opponents of the tax expenditure, the diiculty of ensuring
that all members of the general community bear those costs equally and
the diiculty of ensuring that the beneits of any tax expenditure retrench-
ment are distributed fairly (equally or otherwise) are key elements noted
in the public choice literature. All of these factors mean that members
of the majority face a prisoner’s dilemma – if the majority cannot agree
to share the costs and beneits of political activism on this issue, no one
individual will take the initiative of bearing those costs to procure widely
distributed beneits for the community. As the costs exceed the beneit
that any member of the majority would receive, in the absence of collect-
ive action each member of the majority rationally would acquiesce in the
minority’s procurement of state resources.146
he corollary of this account is that groups will form where the ben-
eits to be gained from a tax expenditure exceed the cost borne by any
one member of the group. his will depend upon the group’s capacity to
exclude free riders from obtaining beneits won at the expense of members
of the group. For example, tax expenditures might be targeted in such a
way as to beneit only members of the group to the exclusion of all others.
According to the public choice account, the costs borne by the bene-
iciary group comprise the purchase of political favours from decision
makers (i.e. politicians),147 who rationally compare the beneit to be
obtained from the provision of such public wealth to any costs associated
with conferring beneits upon the minority at the expense of the major-
ity. Rather than limiting themselves to cases of corruption, the public
choice literature focuses upon the considerable amounts contributed to
politicians’ respective campaign inance funds.148 Politicians therefore
146
Hartle, he Expenditure Process of the Government of Canada: A Public Choice Rent
Seeking Perspective (1988), 66; Oppenheimer, ‘Public Choice and hree Ethical Properties
of Politics’ (1985), 243.
147
Doernberg and McChesney, ‘Doing Good or Doing Well?: Congress and the Tax Reform
Act of 1986’ (1987), 893, 898; see also Doernberg and McChesney, ‘On the Accelerating
Rate and Decreasing Durability of Tax Reform’ (1987).
148
Act of 1986’ (1987), 901–3.
self-interestedly construct their role as one of purveyors of legislative

favour.149 hus, this strand of public choice literature casts the demo-
cratic process as one in which cynical, small bands of robbers can raid the
public interest by exploiting the very dispersal of political power which
democratic theory oten proclaims to be the best means of preventing
tyranny.150
A second strand of public choice literature casts politicians’ behav-
iour in a more democratic light by applying a principal/agent approach
to the political ‘marketplace’.151 Under this approach, politicians do not
sell legislation for cash and other beneits but rather seek to maximise
their chances of re-election by responding to the preferences of voters.152
By contrast to the strong form of public choice, under this approach
campaign inancing is just one factor to be taken into account as cam-
paign inancing is a means of marketing a politician’s wares to voters.153
his account has also been applied in the context of party government,
where an individual politician adheres to ‘the party line’ upon the basis
that doing so assures party inancial support in election campaigns and
thereby ofers the best chance of re-election.154
Under this principal/agent approach to the political process, politi-
cians will maximise their utility (i.e. prospects of re-election) by adopt-
ing strategies which will ‘lose the fewest votes’155 amongst undecided
voters.156 From this perspective, politicians will prefer taxation measures
which are not salient and so will avoid imposing taxes in an overt way
and/or will adopt decisions which attract key interest group support.157
hus, Blount argues that legislators will construct alliances with key
stakeholders and pass obscure tax legislation embodying ‘hidden’ taxes
which incorporate a revenue multiplier efect.158 However, Blount does
149
Act of 1986’ (1987), 898.
150
Dahl, A Preface to Democratic heory (2006), ch. 1.
151
Peltzman, ‘Constituent Interest and Congressional Voting’ (1984), 184; Hartle, he
Expenditure Process of the Government of Canada: A Public Choice Rent Seeking
Perspective (1988).
152
See also Blount, ‘he Art of Taxation’ (2001).
153
Peltzman, ‘Constituent Interest and Congressional Voting’ (1984), 185.
154
Seeking Perspective (1988).
155
Blount, ‘he Art of Taxation’ (2001), 347.
156
Seeking Perspective (1988), 48f.
157
158
not explain the introduction of the highly visible and divisive Australian
Goods and Services Tax by a government which had not won a major-
ity of the vote although it had won a majority of seats in the lower house
of the Australian legislature. Blount ultimately concedes that this is one
case where the ideology of the political igures concerned was a key driver
which saw this new tax introduced into Australia.159
5.5.1.1 he application of public choice to tax expenditures

Either of the public choice accounts described above suggests that rational
self-maximising voters get what they want.
Under the principal/agent form of public choice voters in a majoritarian
voting system dictate legislative outcomes as legislators maximising their
chances of re-election must capture the will of the majority. Although
this irst strand takes voter preferences as a given, rather than examining
the factors shaping those preferences, this strand accepts that the public
has a role to play in shaping public policy. his strand ofers a relatively
positive depiction of the functioning of the political marketplace within
a representative democracy – the ‘general will’ is at least represented dur-
ing the legislative process by legislators keen to maximise their respective
chances of re-election. his account leaves unexplained why tax expendi-
tures which favour a minority without clearly procuring beneits for the
broader community would express the general will.
he ‘hard’ version of public choice ofers an answer to this question.
his hard version suggests that members of the ‘general public’ ration-
ally decline from political action to retrench tax expenditures because
the costs of such action will outweigh the personal beneits. According to
this account, costs and beneits are measured by individuals in terms of
personal wealth. Further, structural features within a legislative process,
such as the existence of veto points, mean that political power can become
concentrated in the hands of a few. hat concentration of power can be
exacerbated by opaque legislative procedures and also opaque legislation
which diminishes the capacity of individuals to act against tax expendi-
tures. he literature regarding tax salience suggests that opacity obfus-
cates the signiicance of tax expenditures and therefore impedes accurate
159
‘John Hewson and John Howard took a stand on an issue about which they were person-
ally convinced, but which continued to divide both interests and voters. Conviction is
not necessarily incompatible with the maximisation of self-interest. It is, ater all, why
most individuals become politicians in the i rst place’, Blount, ‘he Art of Taxation’
(2001), 356.
computation of the cost/beneit analysis at the individual level.160 As a

result individuals are not in a position to know how much they might reap
by having tax expenditures retrenched. Secondly, opacity raises barriers
for members of the general public who might otherwise actively engage in
the legislative ‘marketplace’.
5.5.2 he limitations of the public choice account

his public choice portrayal of the rationality of general acquiescence has
several limitations.
Although public choice accounts purport to be positivist161 – merely
describing social phenomena – these accounts can assume a normative
character where the deinition of utility is narrowly deined and where
the human capacity to compute self-interest is assumed. his critique has
been undertaken elsewhere, and so it will suice for present purposes to
sketch key elements of that critique.162
A narrow concept of personal utility – measured in terms of direct con-
trol of wealth – clearly does not explain all social behaviour. For example,
concentration upon maximising wealth does not explain the act of vot-
ing in general elections. From the perspective of rational, self-interested
wealth maximisation, the expense of voting exceeds any accretion to
wealth that any individual voter might expect to gain by casting their
vote in secret. Similarly, the public choice account fails to explain those
occasions where a broad swathe of the ‘general public’ breaks free of the
prisoner’s dilemma depicted by public choice scholars. Instances of mass
protest or even civil war sparked by the appeal of democratic government
are cases where homo economicus does not fully capture the drivers of
human behaviour.
It is possible for individuals to challenge the social context in which
they ind themselves in a way which deies explanation from the per-
spective of self-interest. hus, for example, Hartle grapples with broader
deinitions of utility but concedes that some ‘zealots’ – leaders, heroes
160
Galle, ‘Hidden Taxes’ (2009), 70f; Schenk, ‘Exploiting the Salience Bias in Designing
Taxes’ (2011).
161
Seeking Perspective (1988), 36.
162
See, for example, Rubin, ‘Public Choice, Phenomenology, and the Meaning of the Modern
State: Keep the Bathwater, but h row Out the Baby’ (2002); Hovenkamp, ‘Legislation,
Well-Being, and Public Choice’ (1990); Hovenkamp, ‘Positivism in Law and Economics’
(1990).
and martyrs – may choose to act in a manner contrary to the public

choice assumption of rational maximisation of personal utility.163 Hartle
observes that such behaviour is beyond the scope of public choice, sug-
gesting that psychoanalysis is perhaps the more appropriate discipline for
examining such behaviour.164 Casting legislators as purveyors of legis-
lative favour for cash sums assumes a normative character by suggest-
ing that that is the only motivation for legislator behaviour. If the direct
sale of legislative favours is stripped out of the public choice account, as
where Hartle casts elected representatives as intermediaries between vot-
ers, constituency party members, the central party and the executive gov-
ernment,165 the predictive edge sought by public choice theorists166 is lost.
If the critical import of public choice accounts is reduced to a portrayal
of politicians endeavouring to mediate conlicting pressures as agents
of their respective electorates and/or of ‘the public’ more generally, the
public choice account would be far more diicult to diferentiate from an
account which presents politicians as public spirited intermediaries try-
ing their best to represent an oten diverse electorate.
If individual behaviour is not explicable solely in terms of rational
self-seeking wealth maximisation, three questions arise. First, what other
factors drive human behaviour? Second, how do humans manage con-
licts between any or all of these drivers? hird, do exogenous factors such
as education and the public media shape these drivers?
With respect to the i rst question, at the least it is reasonable to
hypothesise that the construction of an individual’s preferences is a func-
tion of morality, peers, material environmental factors and ideology.
With respect to the second question, the behavioural economics litera-
ture illustrates the point that individuals are oten unable to make inite
wealth maximisation calculations because of human frailty.167 If there are
more, less inite, factors at play such as ‘freedom’, it is reasonable to expect
that humans will be far less certain when quantifying the value of these
factors as part of formulating their respective preferences. he review of
163
164
165
Seeking Perspective (1988), 51 – describing the lot of the ‘workaday politician’.
166
167
McCafery and Slemrod (eds), Behavioural Public Finance (2006).
several competing deontological moral theories in section 5.3 illustrated

the point that at present there is no metatheory by which those competing
theories can be reconciled. Conceivably, the same type of incommensur-
ability exists at the individual level.168 In the absence of a metaconcep-
tion of the good by which competing imperatives might be reconciled, an
individual’s sense of the good life at any point in time will neither be inite
nor will it be stable. Rather, at diferent points in time diferent concepts
of the good will be given greater weight and/or will be interpreted in a
manner which implies one course of action over another.
With respect to the third question, the public choice accounts ignore a
broader view of human nature which incorporates the processes by which
individuals’ respective views of the good life are formed. he role of ideol-
ogy in shaping an individual’s conception of the good life, and the invest-
ment of resources by some in shaping that ideology, are ignored. he
commercial media is assumed to relect the pre-existing views of media
consumers rather than having a role in shaping public opinion.169
5.5.3 he public gets what it wants II – the entrepreneurial dream

According to the irst strand of public choice, legislators more or less
efectively express the general will. Meanwhile, according to the second
strand of public choice, the general public rationally declines from inter-
vening to prevent organised bands from purchasing legislative favours
from legislators. Although both strands of public choice suggest that the
general public in a sense gets what it wants, both strands leave questions
unanswered. he irst strand leaves out an explanation of why the public
would want measures which beneit a minority without procuring clear
beneits for the larger community. he second strand takes as given that
the general public would want its legislators to retrench tax expenditures
if those legislators were to cease selling legislative favours to minority
raiders.
he question whether the public would want to retrench tax expendi-
tures were it able to express its voice through legislative action has been
examined both theoretically and descriptively.170 At the theoretical level,
168
For a positivist consideration of a ‘multi-imperative’ model grounded upon neurological
systems, and its application to behavioural economics, see Brocas and Carrillo, ‘heories
of the Mind’ (2008).
169
Seeking Perspective (1988), 64–5.
170
For a review of the literature see Shapiro, he State of Democratic heory (2003), ch. 5.
the construction of individual preferences and the aggregation of those

preferences into a ‘general will’ has been a core theme of political phil-
osophy, political science and sociology, as it raises the question of where
social power truly resides. In broad terms there are two answers to this
question which we will set out in more detail below. he irst is that the
general public subscribes to the individualist ideology of capitalism such
that those in a position to extract legislative favour in the form of tax con-
cessions deserve such rewards. he second ofers a cynical portrayal of
manipulation of the masses through the construction of a contingent but
nevertheless hegemonic discourse which keeps the masses content.171
By contrast to this negative portrayal of democratic tax expenditure
(mis)management, there is a substantial political science literature which
suggests that the American entrepreneurial dream has been so deeply
accepted by the American public that tax incentives and loopholes favour-
ing those with greater means are supported by at least the majority of
those at the lower end of the income spectrum.172 Shapiro suggests that
this can be explained by the belief on the part of the have-nots that they
might make their riches one day and when they do, they want to maximise
those riches by utilising tax concessions.173 his belief is reinforced by the
Protestant ethic which emphasises individual responsibility for being the
person you are174 – if you do not have the wealth to exploit tax expendi-
tures favouring the wealthy that is the result of your own indigence. his
individualist philosophy combines with a human predisposition to over-
rate our own capacities to make a heady framing efect by which we over-
state the chances of our own success.
If the majority of voters accurately relect the will of the population
more generally, then one response to this theory would be to take these
people as they are and accept the majority will as the best available out-
come. But what if the majority are the dupes of the minority, and only an
elite can see this? he second, cynical, approach to the question of social
power springs from several sources. As we have already seen, some are
171
Herman and Chomsky, Manufacturing Consent (2002).
172
See the summary in Shapiro, he State of Democratic heory (2003), 116f.
173
A similar view has been put in the context of git and estate taxation: Blatt , ‘he
American Dream in Legislation: he Role of Popular Symbols in Wealth Tax Policy’
(1996); McCafery, ‘he Uneasy Case for Wealth Transfer Taxation’ (1994), 328; Graetz,
‘To Praise the Estate Tax, Not to Bury it’ (1983), 285; Graetz and Shapiro, Death by a
housand Cuts (2005). For a public choice account of the US federal estate tax repeal see
McCafery and Cohen, ‘Shakedown at Gucci Gulch: he New Logic of Collective Action’
(2005).
174
Weber, he Protestant Ethic and the Spirit of Capitalism (1992).
Deliberative concepts of accountability 191
sceptical of the intellectual power of the ordinary person to accurately

discern their own interest, let alone the general interest.175 Others see that
the unequal distribution of social wealth that, at least on one view, tax-
ation is meant to overcome, means that those with the most wealth have
the greatest capacity to mislead the majority into accepting tax measures
which undermine the redistributive capacity of taxation. he cynical and
selective presentation of information to the public includes the negative
portrayal of taxation in the public media and hence of tax concessions as
hard won limitations upon state power,176 the apparent progressivity of the
income tax when it was never intended to promote substantial redistribu-
tion from rich to poor,177 anecdotal distractions178 such as the occasional
bringing of a wealthy individual to tax justice despite the fact that the tax
system favours the wealthy. Moreover, despite the widely understood con-
cepts of justice embodied in any tax system, tax policy is couched in terms
which alienate the general public and create a ‘cult of impotence’179 which
plays into the hands of those best placed to make spurious claims for yet
more tax concessions.
5.6 Deliberative concepts of accountability

A key aspect of the preceding discussion of democratic positivism, pub-
lic choice and political science accounts of tax expenditures is the imper-
fect information available to the general public regarding government
policy in this domain. A second issue is the capacity and/or willingness
of members of the general public to process information in the course of
contributing to deliberation by a community upon controversial moral
questions. By contrast to an elitist model of democracy, a deliberative
democratic model envisages more interaction between citizens and their
elected representatives (and executive government) by envisaging the
relationship as a partnership, with citizens actively participating in the
ongoing decision making by elected representatives throughout a gov-
ernment’s term.180
If citizens are to be partners, the public choice and political science
accounts indicate that consideration should be given to lowering the
175
Schumpeter, Capitalism, Socialism and Democracy (1947), 262.
176
Hansen, Taxing Illusions (2003).
177
Stanley, Dimensions of Law in the Service of Order (1993).
178
Shapiro, he State of Democratic heory (2003), 132.
179
McQuaig, he Cult of Impotence (1998).
180
OECD, Citizens as Partners (2001).
barriers to more widespread entry to the political market. h is object
might be advanced by enhancing the transparency of the tax legislative
process and also by requiring more transparent reporting of social out-
comes such as the distributive efect of tax/transfer measures.181 As this
information comes at a cost, such enhancements once again entail moral
questions upon which there is disagreement. What types of informa-
tion should be presented? How should it be presented? Answering these
questions will entail greater speciication of the nature of the partnership
pursued in a community. Merely providing more information would
not restore parity across all political subjects, as the costs of processing
and presenting information to policymakers would remain a barrier to
political action for the general public relative to tightly controlled inter-
est groups.182 he asymmetric capacity to understand the information
provided may or may not ofend the relevant concept of ‘partnership’.
With this issue in mind, we turn to two contemporary theories of polit-
ical justice conceived in terms of democratic deliberation with a view to
gaining some guidance regarding these matters.
5.6.1 Amartya Sen’s account of deliberation

he irst account of democratic deliberation is developed by Amartya Sen
in his elaboration of his theory of justice.183
Sen acknowledges that a plurality of world views makes it impossible
to comprehensively deine just outcomes. He accepts, for example, that
reasonable people can reasonably disagree about whether the top mar-
ginal rate of income tax should be 39 per cent or 40 per cent.184 his indi-
cates that he would agree with our proposition that there is fundamental
moral disagreement about the existence of tax expenditures and also
about the signiicance of tax expenditures.
From this pluralist foundation, Sen’s deliberative theory of justice ties
justice through public reasoning to public reasoning in democratic for-
ums.185 hus, Sen focuses upon deining the acceptable mode of public
reasoning regarding just outcomes rather than advancing a theory of
justice framed in terms of rules and institutions best suited to the just
181
Oppenheimer, ‘Public Choice and hree Ethical Properties of Politics’ (1985), 253.
182
Bimber, Information and American Democracy (2003), 241.
183
Sen, he Idea of Justice (2009).
184
Sen, he Idea of Justice (2009), 396.
185
resolution of moral issues. here are several key aspects germane to the
current consideration of tax expenditure management:
• A commitment to rational discourse bounded by philosophical plur-
alism that, in particular, means that ‘a complete theory of justice may
well yield an incomplete ranking of alternative courses of decision, and
that an agreed partial ranking will speak unambiguously in some cases
and hold its silence in others’;186
• A commitment to democracy upon the basis that it can incorporate
the procedural and environmental factors which facilitate rational dis-
course; and
• A concentration upon realised phenomena as the focus of inquiry into
social justice rather than a focus upon what Sen calls the transcen-
dental institutionalism of John Rawls.187 In particular, this means that
Rawls’ speciication of justice in terms of primary goods such as income
is inapposite, as Sen argues that consideration of human capabilities
more meaningfully takes account of social inequality.188
5.6.1.1 Institutional aspects of Sen’s theory

Sen goes to some lengths to diferentiate his work from the transcenden-
tal institutionalists who prescribe rules/institutions for the just resolution
of disputes.189 Sen outlines some conditions for what would count as valid
deliberation upon what is just and seems to suggest that appropriately
conditioned discursive practice will engender agreement upon some par-
ticular instances of just conduct. he conditions for such agreement are
not speciied. For example, Sen seems to suggest that there will be agree-
ment upon some matters, such as the abolition of slavery and the provision
of a ‘basic’ system of health care. Implicitly, this seems to be unanimous
agreement, although Sen does not expressly specify this. However, he also
seems to suggest that ‘democracy’ can facilitate just decision making,
without specifying that decisions must be reached unanimously. When
discussing the validity of claims regarding human rights, Sen observes:
A claim that a certain freedom is important enough to be seen as a human
right is also a claim that reasoned scrutiny would sustain that judgement.
Such sustaining may indeed take place in many cases, but not when-
ever such claims are made. Sometimes we may be quite close to a general
186
187
188
189
Sen, he Idea of Justice (2009), 82–3.
agreement, without getting universal acceptance. he advocates of par-
ticular human rights can be involved in active work to get their basic ideas
accepted as widely as possible. No one, of course, expects that there will
be complete unanimity in what everyone in the world actually wants,
and there is little hope that, say, a dedicated racist or sexist will be invari-
ably reformed by the force of public argument. What sustainability of a
judgement demands is a general appreciation of the reach of reasoning in
favour of those rights, if and when others try to scrutinize the claims on
an impartial basis.190
Shortly aterwards, Sen observes:

It can be reasonably argued that any general plausibility that these ethical
claims – or their rejection – have is dependent on their survival when they
encounter unobstructed discussion and scrutiny, along with adequately
wide informational availability.191
he references to the institutional aspects of Sen’s theory of justice are

insuiciently speciied to enable critical consideration of this aspect of
his theory. Sen draws a link between one understanding of democracy,
‘government by discussion’192 and the open deliberation upon justice
that his theory calls for. Moreover, he notes that a ‘free and independent
press’ are central to this connection. What constitutes a free press is not
speciied, although Sen considers that freedom from censorship is one
hallmark of press freedom. his and other aspects, however, are impre-
cisely elaborated. For example, Sen notes the importance of the press in
procuring ‘informed and unregimented’ formation of values, without
specifying what the modiiers entail. Concentration of press ownership,
for example, is ignored, as is the literature regarding the possibility of
media bias193 and/or the shaping of public discourse.194 Worryingly, Sen
lauds the experience of the European and North American press.195 his
presents something of a conundrum. Earlier in his work Sen bemoaned
attempts by the Executive government in the USA to concentrate power
in its hands.196 If the USA press was ‘free’ during this period, how can this
concentration of power have come to pass?
190
191
192
193
Patriquin, Inventing Tax Rage (2004); Hansen, Taxing Illusions (2003).
194
As countenanced in Lukes’ third concept of power: Lukes, Power: A Radical View (2005),
25f.
195
196
In terms of tax expenditure management, it is not clear what type of
information will be necessary for the requisite portion of the community
to ‘generally appreciate’ the justice of a tax rule upon adequately informed
grounds. In particular, the extent to which government agents are obli-
gated to provide information regarding its programmes is not clear. Sen
notes that investigative journalism may unearth information relevant to
the appraisal of public policy,197 but the extent to which investigative jour-
nalism supplements government transparency and accountability is not
clear.
Sen’s account of justice in terms of norms which are justiied upon ‘gen-
erally appreciated’ grounds also is vague. Sen emphasises that impartiality
dictates that views from perspectives external to the relevant community
must be considered,198 but he does not specify how these views are to be
given voice. Further, it is not clear whether ‘general appreciation’ entails
support of a bare majority or whether some other numerical threshold is
required.
5.6.2 Deliberative democracy

he institutional aspects of Sen’s theory of justice leave much to be desired
in terms of adequately identifying the circumstances in which moral dis-
putes may be justly resolved. here is greater attention to the procedural
aspects of their account of deliberative theory in the work of Gutmann
and hompson.
In common with Sen, Gutmann and hompson accept the pluralist
premise that moral disagreement is not necessarily resolvable by recourse
to some version of moral monism arrived at by deliberation.199 hey also
share with Sen the view that this need not mean that moral disagree-
ment can only be resolved by an exercise of power.200 By contrast to Sen,
Gutmann and hompson prescribe a theory of justice with a stronger
emphasis upon the necessary and suicient conditions for just decision
making. However, it must be stressed that Gutmann and hompson
accept that all of their elements of deliberative practice are susceptible
to change by deliberation.201 h is acknowledged contingency of their
account means that they accept that what might appear to be statements
197
198
Sen, he Idea of Justice (2009), 402f.
199
Gutmann and hompson, Why Deliberative Democracy? (2006), 28.
200
Gutmann and hompson, Why Deliberative Democracy? (2006), 46–7.
201
Gutmann and hompson, Democracy and Disagreement (1996), 348.
of fundamental deliberative principle are no more than expressions of
preference from their perspective.202 his contingency of what constitutes
deliberative democracy raises fundamental questions which we will pass
over as they are beyond the scope of this chapter.203 However, it does serve
to emphasise the luid nature of Gutmann and hompson’s speciication
of procedurally just decision making.
5.6.2.1 Neutrality and deliberative democracy

In much of their work Gutmann and hompson strive to ensure that their
construction of deliberative theory does not overstep the line between
irst order moral theories and second order theories.204 hat is, they pre-
sent deliberative democracy as a relatively ‘thin’ second order theory of
moral deliberation which is as neutral as it can be as between rival irst
order moral theories. On this basis they reject Fish’s sceptical proposition
that moral deliberation is ultimately a contest between competing views
ultimately resolved by an exercise of power.205 Deliberation is conceived
in terms of an exchange between moral agents who respect the moral
agency of others with diferent viewpoints, as opposed to treating others
as objects of power or merely as objects of moral reasoning.206
To Gutmann and hompson, the necessary elements of deliberative
democracy are the three procedural requirements of reciprocity, publi-
city and accountability and the three substantive principles of liberty,207
basic opportunity208 and fair opportunity.209 We are concerned to see
whether Gutmann and hompson ofer a plausible model by which
202
Gutmann and hompson indicate that the only limitation upon the capacity of a delib-
erative forum to redei ne deliberation is that deliberation could never describe a pro-
cess which did not incorporate moral argument: Democracy and Disagreement (1996),
352–3.
203
For example, if a ‘deliberative’ forum which complies with Gutmann and hompson’s
deinition of deliberative practice were to adopt non-deliberative (according to Gutmann
and hompson) procedures, presumably it could still call itself a deliberative forum?
204
For discussion of the distinction, which we do not necessarily accept for reasons not ger-
mane to our task, see Gutmann and hompson, Why Deliberative Democracy? (2006), 13.
205
206
207
Gutmann and hompson, Democracy and Disagreement (1996), ch. 7; for the scope of the
limitation upon deliberative decisions imposed by the principle of liberty see Democracy
and Disagreement, 237.
208
Being a minimal welfare standard requiring the distribution of societal resources sui-
cient to ensure that all members of society have an adequate standard of living: Gutmann
and hompson, Democracy and Disagreement (1996), ch. 8; 272.
209
Gutmann and hompson, Democracy and Disagreement (1996), ch. 9.
competing moral standpoints might be resolved, and so we focus upon
the ‘procedural’ elements of their framework.
Reciprocity entails the giving of reasons which are neither purely
procedural nor purely substantive, which ‘appeal to principles that indi-
viduals who are trying to ind fair terms of cooperation cannot reason-
ably reject’210 and which ‘should be accepted by free and equal persons
seeking fair terms of cooperation’.211 Further, these reasons should be
supported by ‘reliable methods of inquiry, as these methods are avail-
able to us here and now’212 such that the absence of a perfect application
of a universally defensible methodology remains consistent with this
requirement.
Public reason underpins their consideration of what counts as a moral
issue susceptible to political resolution.213 In general, Gutmann and
hompson specify that deliberative democracy requires a commitment to
reasoned deliberation. Reasoned deliberation is underpinned by mutual
respect214 and this imports a commitment to ‘openness’ – a preparedness
to adjust one’s views in light of the evidence. here is some ambiguity
regarding the outcome of this deliberative process. In general Gutmann
and hompson propose that disagreement regarding irst order principles
and/or the application of any one principle will remain,215 in which case
other decision making procedures such as majoritarian voting procedures
will be applied in reaching a inite resolution. However, at other points
they suggest that those adopting competing principles are ‘required to
reach some accommodation with one another and with the claims that
express other fundamental values’.216
To encourage mutual respect between proponents of competing view-
points, the process must not begin with voting, it must entail engagement
between a broad cross-section of the community, be moderated, enlist
expert opinion and assistance when necessary, allow for the provision
of extensive information to participants before the deliberative pro-
cess commences217 and adhere to the principle of the economy of moral
210
211
212
213
Gutmann and hompson, Why Deliberative Democracy? (2006): the three principles of
preclusion are summarised on p. 72.
214
215
Gutmann and hompson, Democracy and Disagreement (1996), 18 (‘[citizens] express
and respect their status as political equals even as they continue to disagree’).
216
217
disagreement – participants ought to seek justiications for their deci-

sions which minimise diferences between opposing views.218
Accountability requires that the reasons must be accessible to all the
citizens to whom they are addressed, and the reasons must be ‘public’ in
the sense that the deliberative process must be open to the public and also
the reasons must be understandable by those to whom the reasons are
addressed.219
5.6.2.2 Power imbalance and the deliberative minimisation

of disagreement
Participation in a deliberative forum entails commitment to the principle
of reciprocity to seek accommodation with those expressing a diferent
view:
he underlying assumption is that we should value reaching conclusions
through reason rather than force, and more speciically through moral
reasoning rather than force, and more speciically through moral reason-
ing rather than through self-interested bargaining. Citizens and oicials,
we assume, can learn how to take each other seriously as moral agents.
hey can enter the discussion in the political forum with the purpose
of discovering principles on which the society as a whole can act, rather
than with the aim of devising arguments by which they can advance only
their own interests.220
Gutmann and hompson concede that at irst blush deliberation may

favour those endowed with the greatest advantage in terms of wealth and
social power. However, they suggest that the moral reasoning at the core
of deliberation is best suited to identifying and eliminating such inequal-
ities. Further, as discussed below, they suggest that impassioned advocacy
on behalf of disadvantaged groups is appropriate in a deliberative forum.
he nature of ‘power’ is not overtly considered by Gutmann and
hompson, although it appears that they contemplate Lukes’ irst and
second forms of power – the ability to exert direct authority over another
and also the ability to control what is subject to deliberation.221 he third,
more subtle form of power – the construction of an ideological frame-
work which shapes the way in which issues are framed and addressed222 –
is not considered expressly. It may be that reason within a deliberative
218
219
220
221
Lukes, Power: A Radical View (2005), 16–24.
222
Lukes, Power: A Radical View (2005), 25f.
forum ofers the best chance of piercing the ideological veil and winning
a majority to a diferent conception of justice, but the impact of external
phenomena upon the nature and subject of deliberation, combined with
the luidity of Gutmann and hompson’s deliberative principles, suggest
that substantive redistribution initiated within deliberative forums can-
not be assumed.
5.6.2.3 Non-deliberative strategies warranted by ‘just’ causes

Gutmann and hompson clearly are concerned about the position of the
less well of in any community. In their concern for social justice they
accept that it may be necessary to underpin deliberative democracy with
distributive justice.
Gutmann and hompson acknowledge that the participation of the
least well of in deliberative decision making is desirable, but the practi-
calities of this engagement are not considered at length.223 To those in the
minority who are most disadvantaged by existing social arrangements
and any associated ideological hegemony, the requirement to ind com-
mon ground with those most advantaged by the status quo elides the
depth of disagreement upon social justice issues.224
It may be because of this inequality of deliberative power that Gutmann
and hompson accept that deliberative principles may be set aside where
‘justice’ warrants a more combative approach, notwithstanding the com-
mitment to deliberation:
Deliberative democrats should recognize that in the political arena pas-
sionate rhetoric can be as justiiable as logical demonstration. hose who
speak on behalf of the disadvantaged can ill aford to ignore the need
to be efective. heorists as well as politicians, at least since the days of
223
Gutmann and hompson, Democracy and Disagreement (1996), 303–6.
224
Testimony of James C. Miller before the Joint Economic Committee: United States of
America, Economic Growth hrough Tax Cuts: What’s the Best Approach? (1999) 106th
Cong. 157, at 153; cited in Kornhauser, ‘Legitimacy and the Right of Revolution: he Role
of Tax Protests and Anti-Tax Rhetoric in America’ (2002). Gutmann and hompson
defend their limited redistributive norm upon the basis that an unqualiied entitlement
to a share of social wealth is not politically practical and nor is it morally defensible. he
irst has nothing to do with the concept of justice. he second is defended upon the basis
that egalitarianism ofers no convincing justiication for unearned beneits. here is lit-
tle diference here between the faith based claims, which Gutmann and hompson reject
as inappropriate for deliberation, and the requirement to see social norms from the per-
spective of distribution rather than from the perspective of eicient resource allocation.
Further, the impassioned pleading to which Gutmann and hompson refer is most likely
to be dismissed as just that, as the onus of proving existing social arrangements as wrong
is borne by proponents of reform rather than by the proponents of the status quo.
Athenian politics and Aristotelian rhetoric, have recognized the legitim-
acy of modes of persuasion in politics that combine reason and passion.
Furthermore, rhetoric may properly have to tip toward passion in some
circumstances. Some issues cannot even reach the political agenda unless
some citizens are willing to act with passion, making statements and dec-
larations rather than developing arguments and responses. When non-
deliberative politics – antiwar marches, sit-ins, and workers’ strikes – are
necessary to achieve deliberative ends, deliberative theory consistently
suspends its requirements for deliberation. We should also observe that
these activities oten provoke more deliberation than would otherwise
occur. But even when they do not, they can be justiied if they lead to
future occasions for deliberative criticism of injustice.225
It is not clear when non-deliberative means are ‘necessary’ to achieve

deliberative ends. Gutmann and hompson argue that ‘any legal means’
are appropriate where they are necessary to advance ‘reasonable perspec-
tives’ that have been neglected.226 here is no elaboration of what consti-
tutes a reasonable perspective, although most likely one which complies
with the requirement of reciprocity would qualify.
Further, if the disadvantaged are able to dispense with deliberative
principles in certain circumstances, presumably the principle of reci-
procity would allow their interlocutors to adopt similar tactics, oten with
greater efect given the greater resources that the privileged can access.
Once a non-deliberative approach has been adopted, it is not clear how
the transition from non-deliberative to deliberative practice will emerge.
If those utilising non-deliberative practice have achieved their objectives
by non-deliberative means, why would they change their tactics other
than to continue their instrumental pursuit of their objectives under the
veil of deliberation?
5.6.2.4 Alternate procedural rules where deliberation fails to

resolve the moral controversy
Gutmann and hompson concede that moral deliberation will not
resolve all disputes. It is diicult, for example, to envisage a Cohen com-
ing to some accommodation with a Hayek. Rather than reaching some
middle ground, such staunch opponents might be expected to engage in
moral debate with a view to shoring up their respective moral arguments
without giving ground to the other side. It is diicult to see how a libera-
tarian might moderate their view in the course of deliberation with an
225
226
egalitarian, notwithstanding that Gutmann and hompson are optimis-

tic regarding such a prospect.227
In such cases, Gutmann and hompson accept that bargaining and
other procedural rules are legitimate means to resolve moral disputes
that would otherwise remain unresolved.228 Determining when a dispute
would otherwise remain unresolved is a question which invites moral
argument itself, but in any case it is fair to suggest that it is not possible
to identify a knockdown argument in favour of any of the competing
approaches to the grand questions regarding the role of the state in pri-
vate markets, political justice and distributive justice which are at the root
of much social policy. his suggests that on these grand questions the par-
ties who are turning over old public policy chestnuts might simply accept
that deliberation is extremely unlikely to be of any advantage, in which
case bargaining and other means will be adopted more or less from the
outset. Although Gutmann and hompson acknowledge that diferent
processes for arriving at decisions are consistent with deliberative deci-
sion making, underpinning all is the requirement either that a decision
meet the requirements of deliberative decision making or that the process
by which a decision is made has been subjected to deliberative review.229
Would deliberate review of a bargain merely repeat the bargaining?
5.6.2.5 Information requirements

Gutmann and hompson consider one of the strengths of deliberative
democracy to be the elucidation of more information prior to and during
the course of the deliberative process as participants exchange diferent
views and explore previously unexplored options. Information and rea-
sons for a decision may also be provided retrospectively.230
However, the extent to which information must be available before
the commencement of the deliberative process relative to the amount of
information gleaned during the course of deliberation is not speciied and
227
Gutmann and hompson, Why Deliberative Democracy? (2006), 57. Gutmann and
hompson refer to Nozick’s acceptance of the proposition that redistribution must
be accepted unless it can be shown that the redistribution is not rectifying injustice;
Gutmann and hompson, Democracy and Disagreement (1996), 206 n8. Given the prior-
ity accorded to personal autonomy by libertarian philosophers, it might be that Nozick
considered that the onus of proving any such justiication rested with the proponent of
the redistributive measure.
228
229
230
Gutmann and hompson, Democracy and Disagreement (1996), 173–4; Gutmann and
hompson, Why Deliberative Democracy? (2006), 18–20.
most probably would not be capable of estimation prior to the completion
of the deliberative process.
In the context of tax expenditure management, for example, Gutmann
and hompson give no guidance regarding how much information ought
to be available in advance of any policy deliberation regarding any par-
ticular tax expenditure or of tax expenditures more generally. Gathering
relevant information presumably is preceded by deliberative assessment of
what empirical questions should be answered, what information is neces-
sary, how that information is to be obtained and what agency or entity is
to gather, analyse and present the information at the deliberative forum.
Gathering, analysing and presenting the relevant information might take
some years231 and entail administrative and taxpayer compliance costs.
Moreover, Gutmann and hompson do not specify who bears the
responsibility of providing the information. Greater detail on this issue
would test the claim that deliberative democracy is relatively neutral. If
the obligation for providing adequate information rested with the state,
a libertarian would object that value-laden answers to the construction
of ‘relevant’ information and its presentation would predispose the delib-
erative process to a particular outcome. On the other hand, if obtaining
information were a private afair, there would be a contest between equal-
ity of opportunity to ind relevant information as opposed to actually
having equal access to information. he resolution of this contest must be
grounded upon a particular, subjective construction of democratic polit-
ics. A libertarian, for example, might argue that the key issue is whether
individuals are free to choose to devote private resources to obtain infor-
mation whether or not many individuals actually have resources to devote
to such action.
231
For example, determining the actual incentive efects of a particular tax concession
would require the gathering of information from those who take advantage of the con-
cession ater the beneit of the concession has been obtained and/or ater a particular
legal structure has been adopted to bring the subject within the scope of the concession.
6
Managing tax expenditure controversies
6.1 Introduction – epistemic and political tensions

In the preceding chapters we have argued that tax expenditure man-
agement relects the epistemological and political tensions encountered
by contemporary democratic governments. he range of approaches to
dei ning the purpose of tax expenditure management, the range of tax
expenditure concepts, the range of tax expenditure benchmarks and
the range of approaches to tax expenditure management indicate the
‘uneasy’ status of tax expenditure management.
However, as we have also noted in those chapters, the adoption of tax
expenditure management by many countries and international organi-
sations suggests that these tensions have not proved to be insurmount-
able hurdles for those advocating an expanded role for tax expenditure
management. he question which remains unresolved is whether all
of this additional reporting efort is having substantive impact with
respect to the efects of tax expenditures described in Chapter 4 . In
the United States of America, for example, multiple tax expenditure
reports prepared by independent public bodies have not prevented the
dire state of afairs painted by Kleinbard. Is this a case of governments
going through the tax expenditure reporting motions without really
being committed to spending reform and/or broader public policy
reform?
A synopsis of the epistemic and political tensions described in earl-
ier chapters is appropriate before we consider ‘what is to be done’ in
light of the challenges that these epistemological and political tensions
pose.
6.1.1 Epistemic tensions

he literature debating conceptual aspects of tax expenditure management
relects the tension between those conident in the capacity of humankind
203
204 Managing tax expenditure controversies
to achieve the aspiration of the ‘social sciences’ – identifying the logic of

social interaction – and those sceptical of this possibility. he latter claim
that the quest for universal truth may be misguided (at best) or, more cyn-
ically, believe that it is a facade masking a self-interested agenda.1
Tax expenditure proponents typically have approached the subject on
the basis that the taxation law has not achieved a speciied ideal. To ground
this criticism, proponents have sought critical bedrock in various formu-
lations of the tax expenditure concept and its purpose. Surrey claimed
broad support among tax policy experts,2 Swit et al. proclaim that neu-
trality is the benchmark against which tax expenditures ought to be iden-
tiied and measured,3 McIntyre inds bedrock in the legislative intention,4
Weisbach and Nussim in a functional allocation test,5 Fleming and Peroni
in tax policy principles,6 the Joint Committee of Taxation in a hybridised
1
Brennan and Buchanan, he Power to Tax (1980), 26; Hayek, he Constitution of Liberty
(1960), 291; Hayek, he Road to Serfdom (2007), 138–9, ch. 13.
2
3
Swit et al., ‘Tax Expenditures: General Overview’ (2004), 6 (‘If a government chooses to
pursue neutrality in its tax system, it will eschew tax expenditures’). Notably, the authors
cite New Zealand as the best example of a neutral tax system, overlooking the exclusion
of many capital gains from income taxation – what we call an implicit tax expenditure
(see Chapter 2). It is clear that this is the second of the two senses of neutrality adopted
by Kahn, ‘he Two Faces of Tax Neutrality: Do hey Interact or Are hey Mutually
Exclusive?’ (1990).
In any case, some tax expenditures could be justiied on neutrality grounds – nega-
tive tax expenditures might be directed towards internalising externalised costs while
positive tax expenditures might be directed to overcoming market failure: Wiedenbeck,
‘Paternalism and Income Tax Reform’ (1985).
Other commentators have sought to elevate neutrality to fundamental status. See, for
example, Zelinsky, ‘Eiciency and Income Taxes: he Rehabilitation of Tax Incentives’
(1986).
4
McIntyre, ‘A Solution to the Problem of Dei ning a Tax Expenditure’ (1980–81), 101. To
similar efect, Edgar argues that the policy justiication provided for a particular provi-
sion goes a long way towards resolving the contentious classiication of tax rules between
technical tax rules and tax expenditures: Edgar, ‘Financial Instability, Tax Policy and the
Tax Expenditure Concept’ (2010), 29.
5
Weisbach and Nussim, ‘he Integration of Tax and Spending Programs’ (2004).
6
With a focus upon ability to pay in the sense conveyed by the Schanz–Haig–Simons con-
cept of income: Fleming and Peroni, ‘Reinvigorating Tax Expenditure Analysis and its
International Dimension’ (2008). Fleming and Peroni argue that ability to pay, the SHS
deinition of income and neutrality are the core tax policy principles which underpin tax
expenditure analysis. Neutrality is narrowly dei ned as internal to the income tax, so as
to maintain consistency with their expressed preference for ability to pay and SHS. hat
is, neutrality is taken to require that all forms of income should be treated uniformly,
and thus the debate regarding the ineiciency of taxing income as opposed to consump-
tion is ignored: Fleming and Peroni, ‘Reinvigorating Tax Expenditure Analysis and its
International Dimension’ (2008), 461.
Introduction – epistemic and political tensions 205
purposive/neutrality classiication7 and others in some other pragmatic

classiication.8
Critiques of each of these approaches focus upon the proposition that
the respective approach relects the subjective preferences of its author(s).
he critics of the tax expenditure concept therefore take aim at the uni-
versalist claims of the tax expenditure proponents. hese critics maintain
that there is no one universally right approach to deining and managing
tax expenditures,9 and no one universally right process for selecting from
amongst several contending approaches to tax expenditure management.
his critique is part of the wider postmodern assessment of the depiction
of the ‘science’ of government and also of public policy. For example, Crane
suggests that the ‘increasingly futile’ quest for rational perfection should be
abandoned.10 he efect of this abandonment would be to delegitimise the
universalising discourse that underpins the normative tax benchmark –
a fundamental element of tax expenditure management. Examining tax
expenditure management speciically, Bartlett contended that the Bush
administration was right to scale back tax expenditure management on
the basis that the normative foundations for identifying and measuring tax
expenditures were too uncertain. However, he set this normative indeter-
minacy aside when expressing his preference for a consumption tax upon
eiciency grounds.11 Similarly, Zelinsky challenges the validity of the tax
expenditure concept upon the sceptical ground that it is of indeterminate
meaning. However, contrary to this scepticism, Zelinsky defends the state
of the tax law with all of its tax expenditures upon the grounds of adminis-
trative eiciency,12 but he does not specify a methodology for quantifying
the eiciency costs and gains arising from any particular taxation rule.
Leaving universalism to one side, it would also be possible to ground
tax expenditure management upon the hermeneutic construction of a
moral framework that is ‘right for a particular community’ even if that
framework is not necessarily right for all people at all times.13 However,
7
United States of America, A Reconsideration of Tax Expenditure Analysis (2008) – pro-
posing that tax expenditures comprise both deliberate departures from core tax rules
and also structural rules which impose substantial eiciency costs.
8
Kraan, ‘Of-budget and Tax Expenditures’ (2004), 131.
9
Bittker, ‘A “Comprehensive Tax Base” as a Goal of Tax Reform’ (1967).
10
Crane, ‘he Income Tax and the Burden of Perfection’ (2006), 184.
11
Bartlett , ‘he End of Tax Expenditures as We Know hem?’ (2001). As we noted in
Chapter 5, the selection of eiciency as the premiere social imperative is a moral matter.
12
Zelinsky, ‘Eiciency and Income Taxes: he Rehabilitation of Tax Incentives’ (1986).
13
For an exposition of hermeneutic theory see Gadamer, Truth and Method (1975),
esp. xiv.
relativist critics observe that no one of the deinitions and/or approaches

to tax expenditure management is ‘right’ for their respective commu-
nity, upon the basis that members of a community adopt multiple incom-
mensurable conceptions of the good life and so the deinition of the tax
expenditure concept is ‘in the eye of the beholder’.14
Further, critics observe that even if there were one ‘right’ approach,
whether universalist or hermeneutic, the empirical veriication of social
phenomena involved in applying that approach is challenging, or even
beyond the capacity of humankind.15
If this abandonment of ‘the science of government’ is underpinned by
scepticism regarding the empirical/rationalist endeavour more generally,
which it seems to be, then logically such scepticism would carry with it
abandonment of the identiication and comparison of efects of tax expen-
ditures with alternate public policy mechanisms. hat is, the postmodern
critique of the ‘science’ of public policy reaches beyond tax expenditure
management to all aspects of government.16 Following this path leads us
to the libertarian scepticism of the legitimacy of government, as govern-
ment would have no clear basis upon which it impinges upon what the lib-
ertarian takes to be the fundamental rights of the individual.
he continuing debate within the literature suggests, at the least, that
there is little prospect of these controversies being resolved in the near
future. At every turn our collective inability to glimpse moral objectivity
through the veil of subjectivity which constitutes us as humans shapes
our approach to tax expenditures.
6.1.2 Political tensions

Politically, the literature regarding tax expenditure management relects
the tension between competing concepts of democracy. he desirability
of tax expenditure management, and what tax expenditure management
would entail, vary as the purpose of this focus upon tax expenditures is
adapted to any one of several competing perspectives upon the nature
of democracy. One key aspect of this tension within democratic political
theory is the role of a ‘tax policy elite’ in formulating a country’s tax laws
relative to the role of the general public. Prior to what some have called
14
15
1021–3.
16
Introduction – epistemic and political tensions 207
the crisis in the policy sciences,17 the divide between experts and the gen-
eral public was more starkly drawn as one between the informed and the
uninformed. Democracy was perceived to be a risk to achievement of sci-
entiic government rather than a catalyst for it.18
As we observed in Chapter 5, a libertarian challenges the ability of a
technical elite to both deine and achieve ‘the public good’. A libertarian
would argue that tax expenditure management should be such that the
individual can act in their own interest when considering the degree to
which they will involve themselves in public decisions.19 Taking a difer-
ent tack, a ‘democratic positivist’ might answer the same question regard-
ing elitist politics by questioning whether tax expenditure management
as a discrete process is necessary, given that a ‘democratic’ legislature is
accountable to the general community for the laws passed through the
legislature.20 Similarly, as we also saw in Chapter 5, one strand of public
choice theory and some political science literature suggests that the public
gets what it wants – either by rationally acquiescing in the conferral of tax
beneits upon a few or in the misguided belief that they themselves will
beneit from the tax expenditures one day.21 To these observers, specify-
ing a benchmark tax system against which the democratically authorised
tax system should be assessed would be ‘undemocratic’.22
Typically, proponents of tax expenditure management have implicitly
resolved (or ignored) this democratic tension by persistent recourse to
their intellectual elitism – the disparity between the ideal tax system and
the legislated tax system is explained by way of democratic failure. he
typical account is one of institutional shortcomings23 in the context of
17
deLeon and Vogenbeck, ‘he Policy Sciences at the Crossroads’ (2007).
18
See review of the literature regarding the limitations of popular understanding in Spicer,
In Defense of Politics in Public Administration (2010), ch. 1.
19
See section 5.4.
20
In this regard democratic positivism overreaches its ‘positivist’ heritage in assuming
the normative claim that such aggregative models are suicient for a ‘democratic’ con-
stitution. he key element of this normative claim is the minimal speciication of what
constitutes accountable government in that democratic positivists assume that there
is no political market failure arising from imperfectly informed market participants,
externalities (such as the incapacity of voters to reach a polling booth as a result of their
employment or other obligations) and inequalities of social power.
21
See section 5.5.2.
22
here is a hint of populism in this rejection of the elitist speciication of what is good
for the country. For discussion of the signiicance of populism in contemporary United
States politics see Frank, What’s the Matter with Kansas? (2004).
23
Steuerle, Contemporary US Tax Policy (2004); Kleinbard, ‘How Tax Expenditures Distort
Our Budget and Our Political Processes’ (2009); McCafery, ‘he Missing Links in Tax
Reform’ (1999).
general apathy/ignorance on the public’s part.24 he elitist argues that the

public needs to be saved from itself. Indeed, some argue that an open pro-
cess will only open the door to interest groups, concluding that tax expend-
iture management is better undertaken behind closed doors.25 hus, some
prefer a ‘small target’ speciication of the benchmark in the form of a min-
imalist budgetary accounting framework.26 Others, including Surrey,
have faith that tax policy experts can hold lawmakers to account. Others
pursue consensus politics by limiting the number of decision makers and
through minimising the dissemination of information.27
Even where elitists seek to address institutional shortcomings with pro-
posals for reform, ‘the public’ is mentioned in passing without close con-
sideration being given to what a commitment to accountability of elected
representatives means in this context. hus, for example, Kleinbard sug-
gests some institutional reforms which would enhance the transparency
and accountability of tax expenditure management in the United States,28
apparently upon the assumption that an altruistic lawgiver29 would imple-
ment such procedural reforms rather than those reforms being driven by
an empowered public. In the absence of such a lawgiver, Kleinbard does
not explain why those with the lawmaking power would willingly give
up that power, and nor does he explain who would champion his insti-
tutional reforms over the protests of vested interests. What Kleinbard’s
account overlooks is an explanation of the forces which have produced
the parlous state that he describes and so his proposals for reform lack any
account of how those forces might be overcome.
6.2 What is to be done?

Critics of tax expenditure management leave the subject at this point,
content that they have demonstrated that there are no right answers to
24
Caplan, he Myth of the Rational Voter (2007).
25
See the discussion of this point in section 5.4.2.
26
See Edgar, ‘Financial Instability, Tax Policy and the Tax Expenditure Concept’ (2010),
citing Shaviro on this point.
27
Fitts, ‘Can Ignorance Be Bliss? Imperfect Information as a Positive Inluence on Political
Institutions’ (1990), arguing that in some contexts it is better for information to be cen-
tralised so that a real world Rawlsian veil of ignorance promotes consensus rather than
self-interested prudential bargaining.
28
Kleinbard , ‘How Tax Expenditures Distort Our Budget and Our Political Processes’
(2009), 936–8.
29
Rousseau created the oice of the lawgiver in recognition of the fact that oten ‘the people’
do not know what their will is: Rousseau, he Social Contract (1968), Bk 2 ch. 7.
What is to be done? 209
be found for fundamental tax expenditure management questions –

including the deinition of the benchmark, methodologies for quantify-
ing tax expenditures and the institutional framework for managing tax
expenditures. However, this postmodern conclusion also indicates that
the current state of the taxation law is beset with the same limitations.
We are in no position to determine whether the current law relects the
‘right’ public policy if the right public policy is illusory given the absence
of a foundational norm from which public policy perfection might be
derived. Moreover, while Zelinsky argues that the tax writing commit-
tees in the United States federal government ofer the best level of scrutiny
within the fractured law making process there,30 this does not make that
process ‘right’. he public choice and deliberative politics literature, for
example, points to criticisms which might be made with respect to the
limited information and education available to the general public regard-
ing the impacts and eicacy of tax expenditures. Why, then, should we
survey this landscape and, as Zelinsky seems to suggest we ought, agree
with Voltaire’s Pangloss that everything must be for the best in the best of
possible worlds?31
here is no normative bedrock which can underpin a precise identi-
ication and operation of a tax expenditure management architecture.
Libertarians and pluralists seem to overlook this when they construct an
‘ought’ from the ‘is’ of moral pluralism. In asserting the autonomy of the
individual, libertarians assert the primacy of one moral construct over
others. Pluralists trust that public policy will be formulated in a process
by which rivals ‘hear the other side’ and strike a bargain.32 hese plural-
ists seem to assume that socio-political power is more or less uniformly
distributed so that the bargains struck are taken to balance competing
interests and thereby achieve ‘the public interest’. Deliberative democrats
such as Sen do not specify the informational requirements of their the-
ory,33 or in the case of Gutmann and hompson appear to accept that
deliberative politics will be of limited relevance when dealing with hoary
political chestnuts such as the incommensurable principles embodied in
‘tax policy’. heir lexible approach regarding the duration and reach of
the deliberative phase seems to accept that in such domains deliberative
30
Zelinsky, ‘James Madison and Public Choice at Gucci Gulch: A Procedural Defense of
Tax Expenditures and Tax Institutions’ (1993).
31
Voltaire, Candide.
32
Spicer, In Defense of Politics in Public Administration (2010). Spicer draws upon the work
of Hampshire, Justice is Conlict (2000).
33
See section 5.6.
discussion will move quickly to bargaining.34 If justice is synonymous

with self-interested bargaining in at least some cases, incorporation of
the ethic of self-interest within the deinition of ‘just’ rules undermines
the claim that individuals are morally obligated to follow a rule which
emerges from self-interested bargaining. Surely, self-interest will under-
pin any compliance with the rule as well?35
As Gutmann and hompson observe,36 reason, knowledge and deliber-
ation in various forms can only take democratic politics so far – at some
point power must be exercised if government is to function. However, the
overt exercise of power can jeopardise the legitimacy of the process. he
tension between principles such as eiciency and justice, for example, is
such that any such ‘consensual’ bargain can only be unstable at best.37
However, acknowledging the absence of right answers does not mean
that the framing of tax policy in general and tax expenditure management
in particular is completely arbitrary. he determination of public policy is
not a haphazard afair – determined by the roll of a dice one moment and
upon some randomly selected factor the next.
6.2.1 Descriptive accounts of tax expenditure management practice

Acceptance of the fact that public policy is neither arbitrary nor a matter
of precise calculation with inite concepts is a starting point from which
a more compelling descriptive account of tax expenditure management
can be developed. hat descriptive account would incorporate elements
of rhetoric,38 power (in its various forms39 and emanations – recognising
that power may be both concentrated and dispersed across social inter-
actions), political institutions40 and the rational/empirical pursuit of the
best possible knowledge.
Consideration of these factors, and of the variability of their relative
signiicance across time and across jurisdictions, would move the descrip-
tive account beyond the narrative of democratic failure to achieve the elit-
ist vision of the social good. As can be seen in the public choice and some
34
See section 5.6.2.4.
35
See Brian Barry’s critique of bargain theory: Barry, Justice as Impartiality (1995), 33.
36
37
Ofe, Contradictions of the Welfare State (1984); Laclau and Moufe, Hegemony and
Socialist Strategy (2001), xii.
38
McCloskey, Knowledge and Persuasion in Economics (1994).
39
Lukes, Power: A Radical View (2005).
40
Steinmo, Taxation and Democracy (1993).
political science literature, such accounts typically are framed in terms of

the dichotomy between self-interest and the public good. Incorporation
of the wider range of factors noted above would inform an explanation
of:
1. How tax expenditures come into being, and how some survive while
others are repealed or substantially revised, by allowing a more
nuanced consideration of the roles of diferent constructions of tax
policy principles and also various pragmatic factors which combine in
various ways to uniquely shape any tax system.
Howard’s account of the introduction and growth of four signii-
cant tax expenditures in the United States of America suggests that, at
least in the United States, any ad hoc combination of policy champions,
ideas, institutions and interests might create a new tax expenditure.41
he continued viability of the tax expenditure is similarly contingent –
the factors that engendered the tax expenditure may change so that the
tax expenditure prospers, withers or dies;
2. Why international comparative studies of the tax expenditure con-
cept, and of tax expenditure management practices, reveal such diver-
gent approaches in diferent jurisdictions.42 Rather than evidencing a
tendency to move towards a common point which might be univer-
sally right or ‘right for this world’ (as a hermeneutic observer might
say), this divergence invites a broader sociological examination of the
social drivers of tax expenditure management practices; and
3. he role of the ‘right answers’ thesis in legitimating contemporary gov-
ernment in terms of rational/empirical practice in the domain of pub-
lic policy.
By incorporating a wider range of social factors, this descriptive account
would be more compelling than, for example, the public choice and the
public interest accounts described in Chapter 5. By including the inter-
action of social power and its inluence upon the dissemination of infor-
mation regarding tax expenditures, this account would reach beyond
taking individuals as it purports to i nd them, as the public choice
account does. Incorporating the social dimension of power, this descrip-
tive account would consider the ways by which individuals have come to
frame themselves as political participants and also the way that this fram-
ing changes with time.
41
Howard, he Hidden Welfare State (1997), ch. 9.
42
OECD, Tax Expenditures in OECD Countries (2010).
6.2.2 Normative deliberation upon tax expenditure management

Acknowledging the absence of a right answer towards which the tax
expenditure discourse is directed, and also the absence of a right process
by which moral disagreement might be resolved, is signiicant also for nor-
mative consideration of the process of tax expenditure management. Just
as the descriptive account of tax expenditure management would show
that the creation and continuation of tax expenditures is not entirely arbi-
trary, notwithstanding the host of factors which make tax expenditures
contingent, so the normative consideration of tax expenditures need not
collapse to a resigned acceptance of the arbitrariness of social policy and
hence a resigned acceptance of any social policy.
Acknowledgement of the contingent nature of normative propositions
with respect to tax expenditure management afords an opportunity to
provide more transparent accounts of the process by which tradeofs
must be made between competing principles and of the process by which
power must be wielded in settling upon a particular course of action.
his approach to the normative element of tax expenditure management
would have two signiicant implications:
1. Rather than being dictated by a ‘view from nowhere’, the speciication
of the tax expenditure concept would prompt a community’s consider-
ation of the type of democracy that the community wanted to be, with-
out recourse to universalist language. Oten, deliberation upon this
issue is conducted in universalist terms – Fleming and Peroni adopt
a norm of distributive justice while others proceed from an eiciency
norm; and
2. Elaboration of the tax expenditure concept can be framed in more
pragmatic terms tailored to the principles considered important by the
community.
We will consider these in turn.
he irst implication of our approach to the normative aspect of tax
expenditure management is that it would focus attention upon the type
of democracy a particular country aspires to being, as manifested in the
approach adopted with respect to the framing of the national budget in
general and tax expenditure management in particular. Tax expenditure
management is an important aspect of a government’s budgetary frame-
work and is therefore an important aspect of a government’s public pol-
icy framework. A country’s tax expenditure management framework can
be seen as one manifestation of a country’s approach to resolving moral
disputes more generally. Elaboration of that framework must start from

the foundation of the desired form of democratic government to which a
community aspires.
Crucial aspects of this constitutional matter are whether there is a role
for a technocratic elite in managing tax expenditures and, if so, the nature
of that role. Surrey appeared to proceed upon the basis that a technocratic
elite would undertake tax expenditure analysis once tax expenditures
had been identiied.43 Similarly, Steinmo and Witte, amongst others, have
expressed a preference for an elitist exclusion of what are assumed to be the
‘uninformed but rationally self interested’44 from the tax legislative process.
Even self-professed ‘democrats’ can be considered to ofer a thin concept of
participatory democracy, as was noted in our consideration of democratic
positivism.45 Similarly, Zelinsky expresses ‘the desirability of maximizing
participation in government processes’.46 However, like the democratic
positivists, Zelinsky appears to accept that the contemporary federal polit-
ical process in the United States meets his participatory standard.
If we accept that ‘the right process’ cannot arrive at a ‘right’ answer
with respect to any public policy question, a diferent dynamic arises. he
emphasis shits from proving that a proposition is correct to the consid-
eration of why others should adopt that proposition, notwithstanding the
range of perspectives they might adopt. h is may entail rhetorical elem-
ents, which is not to say that the reasoning would necessarily be capri-
cious.47 Moral pluralism can inform a deliberative theory upon the basis
that the deliberative process is best adapted to genuinely acknowledging
the contest of competing views that democratic law making entails.
To begin down this path, we suggest that an obligation upon govern-
ment to actively engage with the general community through accessible
tax expenditure reporting can be constructed from the standpoint of sev-
eral normative perspectives, for example:
• Libertarianism – overcoming the ‘iscal illusion’ of government requires
readily accessible and transparent reporting of tax expenditure infor-
mation by government;
43
See the discussion of this elitism in section 5.4.2.
44
Steinmo, Taxation and Democracy (1993), 193, 209; Witte, he Politics and Development
of the Federal Income Tax (1985), 381.
45
See section 5.4.3.1.
46
Tax Expenditures and Tax Institutions’ (1993), 1180.
47
McCloskey, Knowledge and Persuasion in Economics (1994), ch. 8; McCloskey, he
Rhetoric of Economics (1998).
• Utilitarianism – maximising social wealth is oten misconstrued
restrictively in terms of maximising gross national product. Broader
measures of social wealth incorporate political engagement factors;48
and
• Contemporary deliberative theories – recognise that accountability
within a representative democracy requires the provision of informa-
tion suicient to allow the general community to critically assess the
conduct of elected representatives.
We suggest that on this matter there is some shared ground for those
expressing these fundamentally diferent moral theories. In our view, the
minimal democratic standard with respect to the general availability of
public policy information should dictate that suicient information be
available as to allow interested members of the public to inform them-
selves such that they can engage with lawmakers in critically assessing
tax expenditures. hose with diferent moral perspectives may adopt
this proposition, but might then difer regarding their respective pur-
poses for critically assessing tax expenditures. Diferent purposes will
indicate difering informational standards. A libertarian concentrating
upon tax expenditure retrenchment will be satisied with a minimal
identiication of tax expenditures, while a utilitarian will require that
information dictated by their respective utility aggregation formula. In
section 6.3 of this chapter we set out what we consider this standard
would require with respect to tax expenditure reporting. Of course, in
doing so, we accept that this elaboration of the standard in this context
is contingent upon adoption of a particular approach to democratic pol-
itical theory.
In our view, ‘democracy’ should be taken to require at least the capacity
of individuals to ‘meaningfully’ engage in various aspects of the polit-
ical process, while recognising the pragmatic necessity of representative
agents charged with the responsibility of making decisions for which they
are held accountable. What constitutes ‘meaningful engagement’ should
be the subject of speciic elaboration within particular communities. For
both members of the public and accountable lawmakers, in our view the
role of the technocratic bureaucracy is to exercise its skill and knowledge
in providing suicient information in an accessible format to enable input
into, and critical appraisal of, public policy decisions made by account-
able representatives of the general public.
48
and Social Progress (2009); OECD, How’s Life? (2011).
he requirement that information be suicient is an independent pub-

lic policy imperative which cannot be absolute in a world where tradeofs
must be made.49 he direct costs associated with gathering, analysing and
presenting all information relevant to public policy decisions could be
prohibitive. he information task could also be counterproductive if, for
example, public resources were redirected from public services to fulil-
ment of the information objective. Further, there can be compliance costs
borne by the public in providing information to the tax administration.50
Oten the minimisation of tax compliance costs is stated to be an absolute
principle of tax policy, without adding the caveat ‘all else being equal’.51
However, the review of some rival political theories undertaken in
Chapter 5 indicates that the prospects of reaching a consensus on the
generation and provision of information by the government are remote.
Contrary to the concerns of libertarians, some utilitarians and delibera-
tive democrats, for example, democratic positivists logically would reject
the incorporation of any normative element in a community’s ‘rule of
recognition’ which would require a minimal information standard before
a community would be ‘democratic’ such that its laws truly were law. As
we noted in Chapter 5, incorporation of such a standard would require
courts to engage in moral deliberation for the purpose of determining
whether a rule is law. Such antecedent moral deliberation is anathema to
the democratic positivist’s exclusion of moral elements from the rule of
recognition.52
he second implication of acknowledging the absence of a ‘right’ tax
expenditure framework is that, within the speciic domain of tax expend-
iture management, acknowledging the absence of right answers would
allow a community to adopt a pragmatic approach to the deinition of ‘tax
expenditure’. Under this approach a tax expenditure could be deined to
be any departure from a deined, but admittedly contingent, normative
benchmark that had been adopted for the purposes of managing the tax
system with the object of achieving a community’s concept of the public
interest.
Acknowledgement of the contingent nature of the tax benchmark
would free the tax expenditure discourse from Surrey’s legacy of seeking
49
See section 1.4.
50
Gunz et al., ‘Measuring the Compliance Cost of Tax Expenditures: he Case of Research
and Development Incentives’ (1995).
51
See, for example, Batchelder et al., ‘Eiciency and Tax Incentives: he Case for Refundable
Tax Credits’ (2006), 42.
52
See the discussion of this point in section 5.4.2.
to curtail disagreement by adopting the ‘small target’ approach of a nar-

rowly speciied tax expenditure benchmark. In Chapter 2 we noted that
the tax base speciic deinition of a benchmark, such as that adopted in
the United States, could ignore legislative inaction with respect to pro-
spective tax bases which might otherwise be recognised as a tax expend-
iture were an holistic tax benchmark speciied. For example, the omission
of estate taxation in Australia53 arguably creates a tax expenditure if the
relevant benchmark were deined along redistributive lines (and could do
so if the benchmark were deined along eiciency lines as well). However,
this legislative omission would only be identiied if a broad tax system
benchmark were speciied.
Development of the tax expenditure reporting benchmark along the
lines developed here – a global benchmark for the tax system overall and
a speciication of tax speciic benchmarks – would place tax expenditure
management on a footing more closely tied to a country’s expression
of its aspirations for its taxation system as part of a broader public pol-
icy framework. In this way, tax expenditure management would begin
from a community’s own political values rather than the starting point
being the universalist ‘view from nowhere’ that Surrey and McDaniel
envisaged.54
Further, against this benchmark other taxes such as customs duties
and excises might constitute positive or negative tax expenditures.55 he
same community might acknowledge that the present balance of trade-
ofs between competing imperatives warranted continuation of these
‘imperfect’ taxes. For these taxes, the community ought to begin with
the best justiication for the tax under consideration. Amongst the mass
taxes, we agree with Fleming and Peroni that the best justiication of the
income tax is framed upon redistributive grounds, 56 notwithstanding
the limitations of adopting income as a proxy for well-being.57 Similarly,
the best justiication for a consumption tax is framed upon eiciency
grounds. he benchmark for ‘sin’ taxes such as alcohol, tobacco and
53
Duf, ‘he Abolition of Wealth Transfer Taxes: Lessons from Canada, Australia and New
Zealand’ (2005).
54
Surrey, Pathways to Tax Reform: he Concept of Tax Expenditures (1973), 18; Surrey and
McDaniel, Tax Expenditures (1985), 157.
55
Metcalf, ‘Using Tax Expenditures to Achieve Energy Policy Goals’ (2008): observing that
fuel tax excise relief in the United States is ignored at present because of the restrictive
federal deinition of ‘tax expenditure’ adopted.
56
Dimension’ (2008), 450f.
57
Towards democratic tax expenditure management 217
gambling taxes might be framed in terms of harm minimisation. And

so forth.
his broader but contingent foundation for the tax expenditure concept
would therefore enhance the integrity of tax expenditure management, as
it is consistent with the ‘substance over form’ approach which underpins
the tax expenditure concept and thereby precludes artiicial classiication
of rules with substantively the same efect. here is no diference between
an income tax concession and an excise tax concession, for example, in
terms of the substantive outcome of delivering a beneit to a particular
class of taxpayer through the tax system.
Provided that the admittedly contingent deinition of the benchmark
is expressly deined in the relevant tax expenditure statement, concerns
of some commentators58 regarding the subjective ring fencing of some
tax rules from tax expenditure analysis in our view are not that signii-
cant. By highlighting the contingent nature of the benchmark the tax
expenditure statement would invite deliberation upon the speciication of
‘benchmark tax rules’ in the iterative process of tailoring public policy to
the community’s aspirations in a manner consistent with the ‘meaningful
engagement’ standard expressed above.
6.3 Towards a democratic tax expenditure

management framework
If the ‘right’ concept of ‘tax expenditure’ and right processes for man-
aging tax expenditures are illusory, how do we envision a better world
with respect to these matters? In the remainder of this chapter we set out
how we believe key aspects of tax expenditure management ought to be
addressed. Given the thesis of this book, this elaboration of our views can
only be contingent and should be seen as a proposed starting point for
the rhetorical discursive practice that we suggest should be undertaken
within particular communities. Decisions taken at many points will shape
what follows, and so this section can only begin to identify the subjects of
future deliberations in diferent socio/political contexts where those con-
texts will both shape and be shaped by those deliberations.59
58
Weisbach and Nussim, ‘he Integration of Tax and Spending Programs’ (2004); Kahn
and Lehman, ‘Tax Expenditure Budgets: A Critical View’ (1992).
59
h is is a broader point than that of Steinmo, who observed that tax law making insti-
tutions both shape the interactions of political ‘players’ and those players also shape
the institutional terrain upon which they interact: Steinmo, Taxation and Democracy
(1993), 12.
6.3.1 Deinition of tax expenditures

he fundamental component of a tax expenditure management frame-
work is the deinition of ‘tax expenditure’. In Chapter 2 we considered
the range of purposes that might underpin tax expenditure management
and we also considered the range of alternate deinitions of the bench-
mark against which tax expenditures might be determined. We noted
the diferent moral starting points from which those deinitions proceed.
We also noted that a substantial portion of the tax expenditure literature
is inluenced by the restrictive statutory deinition of ‘tax expenditure’
adopted in the United States of America.
In section 6.2 we suggested that diferent benchmarks could be adopted
in promoting consideration of how a tax system might best express a com-
munity’s aspirations for the nature of the society it wishes to create. We
observed that diferent tax system benchmarks might be adopted. hose
benchmarks might be framed in terms of competing imperatives such
as particular approaches to redistribution and eiciency respectively. At
the level of particular taxes we also observed that tax expenditure bench-
marks might be framed in terms of the best justiication for that tax.
Our proposed dei nition of the tax expenditure concept therefore
is broader than that adopted legislatively in the United States in that it
is not restricted to income taxation,60 does not depend upon the legis-
lative intention,61 does not depend upon substitutability of the relevant
rule or practice for a direct spending measure62 and is not restricted to
expressly legislated rules. hus, for example, the limitations of the arm’s-
length principle in transfer pricing cases will be relected in settlement
outcomes, which should be reported as an operational tax expenditure.63
6.3.2 Categorisation of tax expenditures as spending or as investment?

Surrey’s purpose in framing the oxymoronic ‘tax expenditure’ nomen-
clature was to jolt policymakers into recognising that the intentional
60
A point noted by Surrey and McDaniel, Tax Expenditures (1976), 724.
61
Compare, for example, Maktouf and Surrey, where tax expenditures are deined in terms
of departure from a normative tax base for the purpose of encouraging private behav-
iour or relieving hardship: Maktouf and Surrey, ‘Tax Expenditure Analysis and Tax
Budgetary Reform in Less Developed Countries’ (1983), 741.
62
63
here is a vast amount of literature regarding the limitations of the arm’s-length prin-
ciple and the relative merits of alternate presumptive income measures such as formu-
lary apportionment: Roin, ‘Can the Income Tax Be Saved – he Promise and Pitfalls of
Adopting Worldwide Formulary Apportionment’ (2008).
non-collection of tax revenue was equivalent to direct spending64 – an

application of the rationalist consistency principle.65 he legacy of Surrey’s
purpose is that the tax expenditure literature proceeds upon the basis
that a tax expenditure is a one-of budgetary expense which procures no
enduring beneit. his is consistent with the treatment of direct govern-
ment aid in the form of capital grants and subsidies under a modiied
accruals accounting system applied to public budgeting.66
here is room for debate regarding the equivalence of tax expenditures
to direct spending. We agree with commentators who see the compar-
ability issue as peripheral to the central question. In our view, the most
important question is whether the publicly available information is such
that elected representatives can be held accountable for their public policy
decisions by voters. hat standard applies to all public policy decisions
and entails a vastly diferent question to that of asking whether the critical
assessment of tax expenditures is equivalent to the critical assessment of
direct spending.
hat said, it could be useful to apply idealised spending criteria to tax
spending rules for the purpose of determining whether it would be bet-
ter to provide the beneit in the form of direct spending. Consideration
of spending criteria is not the only perspective upon which the merits of
a tax expenditure might be critically assessed. It is possible that some tax
expenditures could be converted into a public investment into a public
private partnership (PPP) structure – tax expenditures for public hous-
ing and plantation forestry are two examples where the PPP structure
could be a plausible alternative to the tax expenditure. In such cases the
notional investment constituted by the tax expenditure might remain
64
See, for example, Boadway, ‘he Annual Tax Expenditure Accounts: A Critique’ (2007),
107.
65
Some commentators suggest that this is not a high standard, as humans being what we
are one does not have to look too hard to i nd examples of waste in supposedly more
highly scrutinised direct spending programmes. h is critique of tax expenditures ana-
lysis is not telling. If anything, it merely supports the proposition that a higher stand-
ard of accountability ought apply across the board – to both direct spending and tax
spending.
66
h is is because there is no asset in the relevant sense which arises from the tax expend-
iture. he International Monetary Fund dei nes an asset as ‘any economic resource
controlled by an entity as a result of past transactions of events and from which future
economic beneits may be obtained’: IMF, Manual on Fiscal Transparency (2007), 125.
Although a tax expenditure might be allowed with respect to the acquisition of an asset
and for which there is provision for clawback in cases where the asset is not used in a
prescribed way, it is debatable whether there is suicient government control of the asset
to warrant the conclusion that it is a public asset. See also Montesinos, ‘Government
Budgeting and Accounting Reforms in Spain’ (2002), 342.
of the government balance sheet, but might nevertheless be treated as

a public asset. Adopting this approach to particular tax expenditures
might inform a diferent critical perspective from which to undertake tax
expenditure analysis.67 hat analysis might proceed by comparing pro-
vision of the tax expenditure with conversion of the tax expenditure into
direct public investment by way of an aggregated present value amount
of public funds invested in the relevant sector and an assessment of the
actual or anticipated beneits from that investment. his approach might
indicate that closer controls ought to be imposed with respect to the
ongoing conduct of the project by those beneiting directly from the tax
expenditure. Failure to comply with these controls might trigger claw-
back rules which efectively reverse the beneit formerly allowed by the
tax expenditure.
6.3.3 Ex ante tax expenditures analysis – as part of omnibus budget

legislation or incorporated into discrete tax legislation?
Even if we were to accept that tax expenditures can only arise by way of
expressly legislated rule, there are diferent processes by which a rule may
come to be considered by legislators. Some proposals will pass through
detailed scrutiny while others will be inserted at the last minute in garner-
ing suicient support amongst legislators for an omnibus measure.68 Even
expressly legislated tax expenditures are passed into law without detailed
ex ante cost beneit analysis being undertaken. Once they are passed
into law, the onus for repeal implicitly shits to the critics who are let to
make the empirically diicult case that repeal of a subject tax expend-
iture would enhance public policy.69 We suggest that such adoption of tax
expenditures is contrary to the democratic norm regarding the provision
of public policy information we have set out in section 6.2.
In the context of the federal legislative process in the United States of
America, Zelinsky argues that this criticism is misconceived. He argues
that the salience of tax writing committees and the diversiication of
67
For discussion of these ‘new governance’ tools see Salamon, ‘he New Governance and
the Tools of Public Action: An Introduction’ (2001).
68
Burton, ‘Chaos, Rhetoric and the Legitimation of “Democratic” Government – A Critical
Review of Australia’s Tax Legislative Process’ (2007).
69
Testimony of James C. Miller before the Joint Economic Committee: United States of
America, Economic Growth hrough Tax Cuts: What’s the Best Approach? (1999) 106th
Cong. 157, at 153; cited in Kornhauser, ‘Legitimacy and the Right of Revolution: he Role
of Tax Protests and Anti-Tax Rhetoric in America’ (2002).
lobbyists attending tax writing committee meetings means that those

committees aford greater scrutiny of legislative proposals than would be
the case if tax expenditures were retrenched and comparable spending
measures were considered by specialist committees. Zelinsky argues that
specialist committees are more likely to be captured by specialist interest
groups and so the committee members would be less likely to exercise
their respective expertise.70 his procedural defence of tax expenditures
denies the greater likelihood that specialist lawmakers will subject legis-
lative proposals to more rigorous review outside of the oten pressured
budgetary context.71 Further, in arguing that there is little practical dife-
rence between specialist lawmakers and tax specialists, Zelinsky misses
the point that there is a question of degree. No one would reasonably
expect a tax specialist also to be expert in, for example, nuclear safety or
environmental regulation, so Zelinsky’s claim that members of tax com-
mittees can ‘develop comparable proiciencies in specialized topics’72 is
unrealistic.
Most likely, it is politically infeasible to adopt a blanket prohibition
upon the introduction of tax expenditures that have not been subjected to
ex ante tax expenditure analysis. If there were a tax measure which, over-
all, would enhance redistribution from rich to poor but which depended
upon incorporation of tax expenditures brokered by key decision makers
at legislative veto points, most probably ‘in the name of the greater good’
we would support those tax expenditures notwithstanding their short-
comings. However, the absence of tax expenditure analysis for such tax
expenditures would warrant their categorisation in the ‘red’ class of tax
expenditures for reporting purposes – identifying them as most in need
of closer scrutiny. In this way, the absence of any ex ante analysis would be
readily identiied to the general public in the next tax expenditure report.
6.3.4 Procedural accountability rules

A considerable body of literature deals with various proposals to amend
procedural budgetary rules with the object of compelling critical
70
71
hus Steuerle notes that tax writing committees are seen as the ‘go to’ committees as the
path of a legislative proposal to legislation is less convoluted: Steuerle, Contemporary US
Tax Policy (2004), 154.
72
assessment of tax expenditures. For some, such rules are part of a broader
efort to enhance the transparency of decisions with respect to tax expen-
ditures and therefore strengthen the accountability of elected repre-
sentatives with respect to decisions in this area. hus, for example, the
respective merits of wholesale abolition,73 sunset clauses,74 Pay As You Go
(PAYGO) requirements75 and budget deicit caps76 have been subjected to
critical scrutiny. Given our approach to the politics of tax expenditure
management, we consider that such procedural rules should only be con-
sidered to be a small part of eforts to reform tax expenditure manage-
ment. If tax expenditure management is directed towards ensuring that
members of the general public have the capacity to meaningfully review
the justiications for tax expenditure decisions, then many procedural
rules can be seen to focus upon the wrong question.
6.3.4.1 Sunset provisions

Sunset provisions prescribe an arbitrary time by which the merits of the
particular provision are to be reassessed. he approach outlined here
would focus attention upon ongoing review of the provision. Deliberation
upon overturning a sunset provision asks the wrong question, as the tax
expenditure is implicitly assumed to have been meritorious at some point.
As the sunset draws near, this assumption can be critical as attention can
be drawn to national imperatives as justiication for extension of the sun-
set rather than consideration being given to whether the particular tax
expenditures are the most meritorious allocation of social resources.77
6.3.4.2 ‘Pay As You Go’ (PAYGO) budgetary rules

PAYGO measures were intended to prompt critical scrutiny of tax expen-
ditures by requiring performance analysis of existing tax expenditures
by comparison to the opportunity cost of not funding an additional tax
expenditure. PAYGO requirements present a supericial appearance of
73
United States of America, he Moment of Truth (2010).
74
Kysar, ‘he Sun Also Rises: Political Economy of Sunset Provisions in the Tax Code’
(2006).
75
Staudt , ‘Redundant Tax and Spending Programs’ (2006); Yin, ‘Temporary-Efect
Legislation, Political Accountability and Fiscal Restraint’ (2009); Block, ‘Pathologies at
the Intersection of the Budget and Tax Legislative Processes’ (2002); Block, ‘Congress
and Accounting Scandals: Is the Pot Calling the Kettle Black? (2003); Kleinbard, ‘How
Tax Expenditures Distort Our Budget and Our Political Processes’ (2009), 934.
76
77
United States of America, Economic Efects of Reducing the Fiscal Restraint hat Is
Scheduled to Occur in 2013 (2012).
sound budget management, but in truth they draw focus to the wrong
questions. By focusing upon inding a tax spending cut to pay for a new
measure, PAYGO rules draw attention to the accounting treatment of the
new measure and away from asking whether the new measure has merit.
he Madisonian contest between competing interests that underpins
PAYGO rules might merely lead to an unseemly contest of power between
rival interests or it might induce selection of that tax expenditure per-
ceived to cause the least damage to the wider public. In either event, the
justiication of the outcome to a hypothetical member of the public is not
the centre of the political bargaining that PAYGO promotes.
6.3.4.3 Budget envelope approach

In many cases responsibility for tax expenditures is not formally assigned
to the government agency responsible for the substantive area of the
‘spending’ measure and nor is it clear from the public record as to whether
those agencies have had input into the creation and ongoing management
of tax expenditures in their policy domain. To overcome this institutional
barrier to performance management of tax expenditures relative to other
government programmes, in Canada during the 1980s an ‘envelope’
approach was adopted.78 Under that approach each government agency/
department was given an ‘envelope’ igure for expenditure on achieving
outcomes within its responsibility. Savings could be made by abolishing
tax expenditures within the department’s area of responsibility. By focus-
ing attention on least cost delivery of speciied public policy objectives, it
was hoped that responsible departments would be incentivised to under-
take tax expenditure analysis as part of the usual budgetary cycle.
However, measures introduced by Finance Canada were not assigned
a department, and so there was an incentive for government agencies to
inlate their envelope by having Finance Canada introduce tax expen-
ditures which in truth ought to have belonged to a particular agency.79
Similar gaming of the allocation of tax expenditures to particular depart-
ments could be expected where the tax expenditure arguably incorpo-
rated several policy elements – which many tax expenditures do.80 For
78
For background to this measure amidst wider budgetary reform see Poddar, ‘Integration
of Tax Expenditures into the Expenditure Management Experience: he Canadian
Experience’ (1988); Hartle, he Expenditure Process of the Government of Canada: A
Public Choice Rent Seeking Perspective (1988), 107f, esp. 113.
79
80
And this oten explains why the tax expenditure received suicient political support to be
enacted: Howard, he Hidden Welfare State (1997), 179.
example, the Earned Income Tax Credit could fall across several depart-
mental ‘stools’ – social security, labour, and also departments responsible
for some key sub-groups including single women with dependent children
(returning to the workforce) and particular ethnic groups.81 In applying
the envelope approach, to which government department/agency would
the cost of the credit be assigned?
he abolition of this measure, however, may be attributable to a con-
cern held by elected representatives that the conversion of tax expendi-
tures to direct spending would be too successful, in that the iscal illusion
regarding the true size of government would disappear.82
6.3.4.4 Budget deicit cap

he logic of a budget deicit cap is that it serves to constrain public spend-
ing – both direct spending and also indirect spending.83 he problem is
that tax expenditures can be created during buoyant economic periods.
When the economic cycle enters a downturn phase and government
revenues fall, the government can be faced with the diicult task of cut-
ting public sector ‘spending’ at a time when Keynesian theory indicates
the opposite course of action would reduce the severity of the economic
downturn. In this predicament, governments might adopt the short-term
path of raising the deicit cap – leaving future governments free to con-
tinue expanding the tax expenditure list during more buoyant periods of
the economic cycle.
6.3.5 How should tax expenditures be assessed?

Tax expenditure management entails the diicult task of comparing
alternate public policy outcomes from the perspectives of oten compet-
ing, incommensurable standpoints. Choosing whether to raise an add-
itional dollar of tax revenue in order to spend it on child health, defence,
education or some other alternative cannot be reduced to the comparison
of inite outcomes. While we know that some defence spending is neces-
sary, exact answers to questions regarding what the spending should
be devoted to, where the spending should be directed in a geographical
81
he multiple policy elements underpinning ‘make work pay’ programmes have been
noted by many commentators. See, for example: Howard , he Hidden Welfare State
(1997), chs 3 and 7; Liebman, ‘he Impact of the Earned Income Tax Credit on Incentives
and Distribution’ (1998).
82
83
sense, when the spending should be expended, how the spending alloca-
tion should be managed and how much should be allocated to any par-
ticular objective are not certain. In this penumbral zone it is diicult for
the public to critically assess why a particular tax expenditure ought to
continue. What is needed is more contextual information. Aggregated
information such as distributional tables provides some information
regarding the availability of the tax expenditure. What is needed is more
information regarding the functional performance of the particular tax
expenditure relative to other public policy options. Providing this infor-
mation for all tax expenditures would be a gargantuan task. However,
steps could be taken along this path by focusing attention upon those tax
expenditures at the margin.
Governments could categorise tax expenditures as those which, in its
view, are most entrenched (i.e. green) through to those which are most
tenuous (red). Further, within the red category, government could iden-
tify the most tenuous of the tenuous tax expenditures, and compare the
evidence in support of maintaining that tax expenditure relative to other
options. For example, not spending that revenue and allowing a general
tax reduction or spending that revenue on the next most likely policy
choice (i.e. listing another drug on the subsidised list of drugs, building
another hospital, acquiring more defensive capability, etc.). he classiica-
tion of tax expenditures would be transparent and so appropriateness of
the classiication could be revisited at any time to take account of chan-
ging circumstances. he identiication of the most tenuous of the tax
expenditures and its defence would be transparent and would invite scru-
tiny of this tax expenditure as the most tenuous and also invite scrutiny
of why the tax expenditure was allowed to continue in the face of alternate
taxing/spending measures.
6.3.6 Performance analysis

In Chapter 3 we noted that typically the presentation of tax expenditure
information exhibits the elitist approach to tax expenditure management
championed by Surrey and many ater him. Tax expenditure reports
typically comprise a relatively lengthy document with tables presenting
information which is of limited value in assessing the merits of particu-
lar tax expenditures. Oten it seems that little attempt is made to pro-
vide tax expenditure information in a readily accessible way, an approach
which stands in stark contrast to the eforts of governments to publicise
key elements of their annual budgets with media releases, summary
documents and so forth. In section 6.2 we suggested that the purpose of

tax expenditure reporting was not merely to notify a policy elite of the
existence and scale of tax expenditures by identifying a list of tax expen-
ditures and providing an admittedly imperfect estimate of the cost of
each item on the list. We argued there that the purpose of a tax expend-
iture report is grounded in the democratic governance concepts of open-
ness, transparency and accountability. hat is, tax expenditure reporting
ought to prompt active and transparent management of tax expenditures
by assembling the information to facilitate tax expenditure analysis and/
or to verify to the general public that such an analysis has been under-
taken. his information might be incorporated by reference, as where the
government’s audit oice has undertaken a critical appraisal of the oper-
ation of a particular tax expenditure or where a government agency has
conducted a public review of the measure.
Having regard to the variety of tax expenditures contemplated by
our broad deinition of ‘tax expenditure’, being a rule or practice which
departs from the stated benchmark (rather than an express legislative
concession along the lines of the statutory deinition of ‘tax expenditure’
adopted in the United States of America), it is doubtful that there can be
a template for performance analysis for all tax expenditures. Assessing
the merits for an operational tax expenditure will entail a considerably
diferent approach in key respects to assessing the merits of an express tax
expenditure. For example, oten it will be inappropriate to consider the
purpose of an operational tax expenditure as they might arise because of
the allocation of limited administrative resources or because of legislative
ambiguity.
In any case, as a minimum, the information necessary for performance
analysis of a tax expenditure comprises:
1. If relevant, the stated purpose of the tax expenditure.
2. How the tax expenditure interacts with other government assist-
ance. In the case of countries with more than one level of government,
this would entail consideration of assistance provided by any level of
government.
3. he expected impact of the tax expenditure upon market decisions
and the basis upon which this impact has been determined, including
an assessment of the limitations of that modelling in the particular
context of the tax expenditure.
In our view, the tax expenditure report should expressly state
whether consideration has been given to second order efects, such as
impacts upon those not eligible for the relevant tax expenditure. For
example, efects arising from the pricing of similar goods/services pro-
vided by eligible and ineligible recipients.84
In undertaking this allocative eiciency analysis, the preferred
structure of the tax expenditure must irst be identiied85 before then
considering the allocative efects of that tax expenditure relative to
alternate mechanisms available to government to achieve the same
objective.
4. Explanation of how each measure is to be administered, and, if rele-
vant, why the revenue collection agency is the best equipped to admin-
ister the measure.86
5. The expected fiscal cost of the tax expenditure over the period of
the budget cycle, including information upon which this costing
is based and an assessment of the accuracy of the costing having
regard to factors such as the reliability of the information relied
upon.
6. he anticipated or actual administrative (revenue agency and/or
other government agency) and compliance costs (taxpayer) of the
measure.
7. Information regarding the efective tax rates borne by particular eco-
nomic sectors. here may not be a ‘typical’ taxpayer within any par-
ticular sector, but hypothetical typical operations might be described
with a view to identifying the efective tax rate of the operator of the
hypothetical enterprise. Information of this kind typically is used in
tax administration compliance risk management.
84
Azmat considers the impact of make work pay tax expenditures upon ineligible workers:
Azmat, he Incidence of an Earned Income Tax Credit: Evaluating the Impact on Wages
in the UK (2006). By focusing upon those capable of working, the earned income credit
does not necessarily deliver additional aid to those most in need. Such efects might con-
l ict with the express purpose of the measure, as where the earned income credit pro-
vides assistance only to those engaged in the private market for labour and excludes
those unable to participate in this market: Staudt, ‘he Hidden Costs of the Progressivity
Debate’ (1997), 972f. For example, to ensure the viability of make work pay schemes the
category of welfare recipients eligible for welfare may be restrictively dei ned and the
level of welfare support to those theoretically capable of working must be constrained, in
order to maintain the justiication for the earned income credit being the inancial incen-
tive to work.
85
Batchelder et al. argue that, in certain closely deined circumstances, the preferred mech-
anism for delivery of a tax expenditure justiied upon allocative grounds is a uniform
refundable tax credit: Batchelder et al., ‘Eiciency and Tax Incentives: he Case for
Refundable Tax Credits’ (2006).
86
8. Information regarding the eicacy of delivery through the taxation

system, where several factors might afect the eicacy of the spending
programme:
a. Non-lodgement of a tax return by eligible taxpayers might lead to
underachievement in terms of non-delivery of the beneit to all of
the target group.87
b. Behavioural responses – it is possible that more of the target group
will willingly claim a tax beneit rather than sufering the actual or
perceived stigma of claiming welfare support.
c. How the beneit is to be delivered through the tax system – at regular
withholding tax payment points in the course of the relevant income
year (i.e. salary payments for individuals, tax instalment payment
dates for business and corporate taxpayers) or in a lump sum upon
assessment for tax ater the relevant income year has ended? he
former might be of more use to the beneiciaries in managing cash
lows (i.e. individuals, small business) and also might enable closer
matching of eligibility to receipt of the beneit, but might also add to
the compliance/administration costs of the measure.
d. Whether the tax expenditure will be lost upon taxpayers who
receive an annual global tax assessment which incorporates the
hidden beneit of the tax expenditure, such that desired behav-
ioural responses might be less than if there were a direct cash grant
or some other form of assistance.88
e. he targeting of the tax expenditure, particularly where the tax
expenditure does not take the form of a refundable tax credit and
where the tax expenditure is delivered through a progressive income
tax. For example:
i. It is possible that some of the beneit of the government spending
will be captured as a windfall by high tax rate beneiciaries.89
87
For discussion of this in the context of the earned income tax credit (US) see Liebman,
‘he Impact of the Earned Income Tax Credit on Incentives and Distribution’ (1998), 83
at 109f.
88
Liebman, ‘he Impact of the Earned Income Tax Credit on Incentives and Distribution’
(1998), 83 at 109f.
89
For discussion of this point in the context of tax exemption for state and municipal
bond interest see Fleming and Peroni, ‘Reinvigorating Tax Expenditure Analysis and
its International Dimension’ (2008), 465–8. For more general consideration of the con-
cept of implicit taxation see Weisbach, ‘Implications of Implicit Taxes’ (1999); see also
Johnson, ‘Why Have Anti-Tax Shelter Legislation? A Response to Professor Zelenak’
(1989), 618–19.
ii. Where presumptive tax rules apply, the application of pre-

sumptive tax rules may mean that a portion of the intended tax
expenditure beneiciaries is missed.90
9. he distributional efect of the tax expenditure, including at least
some consideration of the likely second order efects of the measure.
10. In the case of a recurrent tax expenditure, whether the performance
of the tax expenditure has been assessed against its original/amended
objectives.
11. Speciication of the performance outcomes which would mean that
this tax expenditure would be placed in the red (most tenuous) cat-
egory of tax expenditures.
6.3.7 Collaboration between tax policy agencies

In the taxation domain governments continue to explore avenues for more
efectively achieving their respective goals through their respective tax
systems. Cooperation through agencies such as the OECD, renegotiation
of tax treaties to facilitate enhanced information exchange and greater
cooperation between tax administrators to the point of conducting joint
audits relect a need to respond to global tax challenges cooperatively.
Tax expenditures constitute another opportunity for closer cooperation
between governments. Although the tax expenditure concept has been
with us for at least one half of a century, there is considerable room to
expand the knowledge regarding tax expenditure management. As noted
by the United States Government, pragmatic responses are necessary
if the limited data, resource constraints and technical diiculty of per-
formance appraisal are to be overcome with meaningful tax expenditure
management.91
6.3.8 Independent performance review of tax expenditure

management
Tax expenditure statements are commonly prepared by a branch of gov-
ernment in circumstances where the independence of the authors from
90
Developed Countries’ (1983), 754.
91
United States of America, Analytical Perspectives, Fiscal Year 2013, Budget of the United
States Government (2012), 274–5.
the government of the day cannot be assured. Under the Westminster
doctrine, for example, there is a tension between government oicials
conceived as servants of ‘the public’ and government oicials conceived
as servants of the government of the day. Each approach implies a vastly
diferent approach to the publication of information which might prompt
critique of government policy – such as a tax expenditure statement which
identiies wasteful government ‘spending’.
An efective approach to tax expenditure management would man-
age this risk to democratic deliberation upon tax expenditure policies by
instituting mechanisms which assure the independence and quality of the
tax expenditure statement. he role of public audit oices, for example, in
routinely overseeing government management of tax expenditures is one
means by which this objective might be achieved.92
6.4 Into the political future

In the preceding chapters of this book we have examined a great many
controversies to do with tax expenditure management and also contro-
versies within liberal democratic political theory regarding how contro-
versies regarding such matters as tax expenditures might be resolved. he
depth of these controversies indicates that there is little prospect that any
of them will be resolved any time soon.
In our view, the existence of such controversies does not warrant aban-
donment of the practice of tax expenditure management but it does war-
rant open acknowledgement of those controversies as communities shape
the way in which they deal with such controversies while dealing with
those controversies.
Tax expenditure management is deeply political in many if not all of
its aspects – a proposition that Surrey and many ater him have sought
to downplay by various strategies. Transparent acknowledgement of the
political aspect of tax expenditure management creates an opportunity
for enhancing the legitimacy of tax systems speciically, and government
more generally, in ‘modern’ democracies where it is commonly suggested
that more civic engagement is desirable.93 However, with this opportunity
there are risks. he public choice and political science literature challenges
blind faith in participatory/deliberative constructions of democratic pol-
itics. here is also the risk of creating instability and hence delegitimising
92
See EuroSAI, Report on the Coordinated Audit of Tax Subsidies (2008).
93
See, for example, OECD, Citizens as Partners (2001).
Into the political future 231
democratic government in this ield – just as some libertarians proclaim.
Moreover, providing the information to underpin ‘democratic’ engage-
ment does come at a cost.
hese controversies mean that there can never be a last word on tax
expenditure management. To our minds, the scale and signiicance of
tax expenditures make them too important to ignore, even if there is no
‘right’ way to identify, analyse and actively manage them.
BI BL IO GR A PH Y
Abel, Andrew B., N. Gregory Mankiw, Lawrence H. Summers and Richard

J. Zeckhauser, ‘Assessing Dynamic Eiciency: heory and Evidence’, 56
Review of Economic Studies 1 (1989).
Alm, James, Jorge Martinez Vazquez and Sally Wallace, Taxing the Hard-to-Tax
(Elsevier, 2004).
Andrews, William, ‘Personal Deductions Under an Ideal Income Tax’, 86 Harvard
Law Review 309 (1972).
Arnold, Brian, Timing and Income Taxation: he Principles of Income Measurement
for Tax Purposes (Canadian Tax Foundation, 1983).
Australia, 2007 Tax Expenditures Statement (Treasury, 2008).
2011 Tax Expenditures Statement (Treasury, 2012).
Architecture of Australia’s Tax and Transfer System (Treasury, 2008).
Australia’s Future Tax System (Treasury, 2009).
Biofuels in Australia – An Overview of Issues and Prospects (Rural Industries
Research and Development Corporation, 2007).
Compliance Program 2010–11 (Australian Taxation Oice, 2010).
Making Transparency Transparent (AGPS, 1999).
Preparation of the Tax Expenditures Statement (Australian National Audit
Oice, 2007).
Preparation of the Tax Expenditures Statement (Treasury, 2008).
Avi-Yonah, Reuven S., International Tax as International Law (Cambridge
University Press, 2007).
‘hree Goals of Taxation’, 60 Tax Law Review 1 (2006).
Azmat, Ghazala, he Incidence of an Earned Income Tax Credit: Evaluating the
Impact on Wages in the UK (CEP Discussion Paper 724, Centre for Economic
Performance, 2006).
Bankman, Joseph and David Weisbach, ‘he Superiority of an Ideal Consumption
Tax Over an Ideal Income Tax’, 58 Stanford Law Review 1413 (2006).
Barry, Brian, Justice as Impartiality (Clarendon Press, 1995).
Bartlett, Bruce, ‘he End of Tax Expenditures as We Know hem?’, 92 Tax Notes
413 (2001).
Batchelder, Lily, Fred Goldberg and Peter Orszag, ‘Eiciency and Tax Incentives:
he Case for Refundable Tax Credits’, 59 Stanford Law Review 23 (2006).
232
Bibliography 233
Berlin, Isaiah, ‘Two Concepts of Liberty’, in Four Essays on Liberty (Oxford
Bimber, Bruce, Information and American Democracy (Cambridge University
Press, 2003).
Bird, Richard M. and Eric M. Zolt, ‘Redistribution via Taxation: he Limited Role
of the Personal Income Tax in Developing Countries’, 52 UCLA Law Review
1627 (2005).
Birnbaum, Jefrey H. and Alan S. Murray, Showdown at Gucci Gulch (Vintage, 1988).
Bittker, Boris, ‘Accounting for Federal Tax Subsidies in the National Budget’, 22
National Tax Journal 244 (1969).
‘A “Comprehensive Tax Base” as a Goal of Tax Reform’, 80 Harvard Law Review
925 (1967).
‘Income Tax “Loopholes” and Political Rhetoric’, 71 Michigan Law Review 1099
(1973).
Blatt, William, ‘he American Dream in Legislation: he Role of Popular Symbols
in Wealth Tax Policy’, 51 Tax Law Review 287 (1996).
Block, Cheryl D., ‘Congress and Accounting Scandals: Is the Pot Calling the Kettle
Black?’, 82 Nebraska Law Review 863 (2003).
‘Pathologies at the Intersection of the Budget and Tax Legislative Processes’, 43
British Columbia Law Review 863 (2002).
Blount, Simon, ‘he Art of Taxation’, 16 Australian Tax Forum 345 (2001).
Blum, Walter J. and Henry Kalven, he Uneasy Case for Progressive Taxation
(University of Chicago Press, 1953).
Boadway, Robin, ‘he Annual Tax Expenditure Accounts: A Critique’, 55 Canadian
Tax Journal 106 (2007).
Boadway, Robin and Anwar Shah, Fiscal Federalism (Cambridge University Press,
2009).
Bradford, David F., Distributional Analysis of Tax Policy (AEI Press, 1995).
Braithwaite, John, Markets in Vice, Markets in Virtue (Oxford University Press,
2005).
Brennan, Geof rey, ‘Public Choice and Taxation: Leviathan Ater Twenty Years’,
in Tax Conversations: A Guide to the Key Issues in the Tax Reform Debate
(Richard Krever (ed.), Kluwer, 1997).
Brennan, Geof rey and James Buchanan, he Power to Tax (Cambridge University
Press, 1980).
Brocas, Isabelle and Juan D. Carrillo, ‘heories of the Mind ’, 98 American Economic
Review Papers and Proceedings 175 (2008).
Brooks, Neil, ‘Comment’, in Tax Expenditures and Government Policy (Bruce (ed.),
1988).
‘he Under-Appreciated Implications of the Tax Expenditure Concept’, in
Australian Business Tax Reform in Retrospect and Prospect (Chris Evans and
Richard Krever (eds), homson Reuters, 2009).
234 Bibliography
Brooks, Neil, Jinyan Li and Lisa Philipps, ‘Tax Expenditure Analysis: State of the
Art’, in Tax Expenditures: State of the Art (Lisa Philipps, Neil Brooks and
Jinyan Li (eds), Canadian Tax Foundation, 2011).
Brooks Gann, Pamela, ‘What Has Happened to the Tax Legislative Process?’, 86
Michigan Law Review 1196 (1988).
Bruce, Neil (ed.), Tax Expenditures and Government Policy (Queen’s University, 1988).
‘Tax Expenditures and Government Policy: An Introduction to the
Conference Volume’, in Tax Expenditures and Government Policy (Bruce
(ed.), 1988).
Burg, David F., A World History of Tax Rebellions (Routledge, 2004).
Burman, Leonard , Christopher Geissler and Eric Toder, ‘How Big Are Total
Individual Tax Expenditures and Who Beneits from them?’, 98 American
Economic Review 79 (2008).
Burton, Mark, ‘he Australian Small Business Tax Concessions – Public Choice,
Public Interest or Public Folly?’ 21 Australian Tax Forum 91 (2006).
‘Capturing Contemporary “Democracy”: he Shortcomings of Australian
Tax Expenditure Management and heir Ideological Foundations’, in Tax
Expenditures: State of the Art (Lisa Philipps, Neil Brooks and Jinyan Li (eds),
Canadian Tax Foundation, 2011).
‘Chaos, Rhetoric and the Legitimation of “Democratic” Government – A Critical
Review of Australia’s Tax Legislative Process’, 29 Sydney Law Review 613
(2007).
‘Is Participatory Tax Transparency Achievable?’ 4(3) Tax Notes International
333 (2006).
Burton, Mark and Miranda Stewart, Promoting Budget Transparency hrough
Tax Expenditure Management: A Report on Country Experience for Civil
Society Advocates (2011), available at: http://papers.ssrn.com/sol3/papers.
cfm?abstract_id=1864324 (last accessed 21 August 2012).
Byrne, Donna M., ‘Progressive Taxation Revisited ’, 37 Arizona Law Review 739
(1995).
Campbell, Tom, ‘Grounding heories of Legal Interpretation’, in Legal Interpre-
tation in Democratic States (Goldsworthy and Campbell (eds), 2002).
he Legal heory of Ethical Positivism (Dartmouth, 1996).
Canada, Tax Expenditures and Evaluations (Department of Finance, 2011).
Tax Expenditures and Evaluations, Notes to the Estimates/Projections (2010),
www.i n.gc.ca/taxexp-depisc/2010/taxexp1004-eng.asp (last accessed 7
December 2012).
Caplan, Bryan, he Myth of the Rational Voter (Princeton University Press, 2007).
Caribata , Patrick J., ‘From Welfare State to Optimal Size of Government: A
Paradigm Shit for Public Policy’, 5 Agenda 277 (1998).
Carling, Robert, ‘Ten Principles of Tax Reform’, 25 Policy 9 (2009).
Carmody, Michael, ‘Ethics and Taxation’, speech delivered at Edmund Rice
Business Ethics Forum (28 October 1999).
Bibliography 235
Carroll, Robert, David Joulfaian and James Mackie, ‘Income Versus Consumption
Tax Baselines for Tax Expenditures’, in Economic Analysis of Tax
Expenditures (James M. Poterba (ed.), NBER, 2011).
Cavanaugh, Maureen B., ‘On the Road to Incoherence: Congress, Economics, and
Taxes’, 49 UCLA Law Review 685 (2002).
Cockett , Richard , hinking the Unthinkable: hink-Tanks and the Economic
Counter-Revolution, 1931–1983 (Fontana, 1994).
Cohen, Gerald A., Rescuing Justice and Equality (Harvard University Press ,
2008).
Cooper, Graeme, ‘he Beneit heory of Taxation’, 11 Australian Tax Forum 397
(1994).
Craig, Jon, and Allan, William, Fiscal Transparency, Tax Expenditures, and Budget
Processes: An International Perspective (OECD, 2002).
Crandall, William, Revenue Administration: Performance Measurement in Tax
Administration (IMF, 2010).
Crane, Charlotte, ‘he Income Tax and the Burden of Perfection’, 100 Northwestern
University Law Review 171 (2006).
D’Ascenzo, Michael, ‘he Rule of Law: A Corporate Value’, speech delivered at Law
Council of Australia Rule of Law Conference (2007).
Dabner, Justin and Mark Burton, ‘Lessons for Tax Administrators in Adopting the
OECD’s “Enhanced Relationship” Model – Australia’s and New Zealand’s
Experiences’, 63 Bulletin for International Taxation 316 (2009).
Dahl, Robert A., A Preface to Democratic heory (University of Chicago expanded
edn, 2006 (originally published 1956)).
Davidson, Sinclair, Fiscal Illusion: How Big Government Makes Tax Look Small
(CIS Policy Monograph No 81, Centre for Independent Studies, 2007).
deLeon, Peter, Democracy and the Policy Sciences (State University of New York,
1997).
deLeon, Peter and Danielle M. Vogenbeck, ‘he Policy Sciences at the Crossroads’,
in Handbook of Public Policy Analysis (Frank Fischer, Gerald J. Miller and
Mara S. Sidney (eds), CRC Press, 2007).
Dodge, Joseph M., ‘heories of Tax Justice: Ruminations on the Beneit, Partnership
and Ability to Pay Principles’, 58 Tax Law Review 399 (2005).
Doernberg, Richard L. and Fred S. McChesney, ‘Doing Good or Doing Well?:
Congress and the Tax Reform Act of 1986’, 62 New York University Law
Review 891 (1987).
‘On the Accelerating Rate and Decreasing Durability of Tax Reform’, 71
Minnesota Law Review 913 (1987).
Dorf, Michael C., ‘Legal Indeterminacy and Institutional Design’, 78 New York
Dowdle, Michael W., Public Accountability (Cambridge University Press, 2006).
‘Public Accountability: Conceptual, Historical and Epistemic Mappings’, in
Public Accountability (Dowdle (ed.), 2006).
236 Bibliography
Downs, Anthony, An Economic heory of Democracy (Harper, 1957).
Driessen, Patrick, ‘A Qualiication Concerning the heory of Tax Expenditures’,
33 Journal of Public Economics 125 (1987).
Duf, David G., ‘he Abolition of Wealth Transfer Taxes: Lessons from Canada,
Australia and New Zealand ’, 3 Pittsburgh Tax Review 71 (2005).
‘Private Property and Tax Policy in a Libertarian World: A Critical Review’, 18
Canadian Journal of Law and Jurisprudence 23 (2005).
Easson, Alex, Tax Incentives for Foreign Direct Investment (Kluwer, 2004).
Edgar, Tim, ‘Financial Instability, Tax Policy and the Tax Expenditure Concept’,
63 SMU Law Review 969 (2010).
Edrey, Yoseph and Howard Abrams , ‘Equitable Implementation of Tax
Expenditures’, 9 Virginia Tax Review 109 (1989).
Eisenstein, Louis, he Ideologies of Taxation (Harvard University Press, 1961).
Eissa, Nada and Hilary Hoynes, ‘Redistribution and Tax Expenditures: he Earned
Income Tax Credit’, 64 National Tax Journal 689 (2011).
Elson, Diane, Budgeting for Women’s Rights: Monitoring Government Budgets for
Compliance with Cedaw (United Nations Development Fund for Women,
2006).
Ende, Leo van den, Amir Haberham and Kees den Boogert, ‘Tax Expenditures
in the Netherlands’, in Tax Expenditures – Shedding Light on Government
Spending hrough the Tax System: Lessons from Developed and Transition
Economies (Polackova Brixi, Valenduc and Swit (eds), 2004).
Epstein, Richard A., ‘ Taxation with Representation: Or, The Libertarian’s
Dilemma’, 18 Canadian Journal of Law and Jurisprudence 7 (2005).
EuroSAI, Report on the Coordinated Audit of Tax Subsidies (Madrid, 2008),
available at: www.eurosai.org/VIIC/ENG/Report_Coordinated%20Audit_
en.pdf (last accessed 24 August 2012).
Farber, Daniel and Philip Frickey, ‘he Jurisprudence of Public Choice’, 65 Texas
Feldstein, Martin, ‘A Contribution to the heory of Tax Expenditures: he Case
of Charitable Giving’, in he Economics of Taxation (H. J. Aaron and
M. Boskins (eds), Brookings Institution, 1980).
‘he Eiciency of Tax Expenditures: A Reply’, 33 Journal of Public Economics
133 (1987).
‘A Route to Cutting the Deicit ’, he Washington Post, p. A15 (29 November
2010).
Fitts, Michael A., ‘Can Ignorance Be Bliss? Imperfect Information as a Positive
Inluence on Political Institutions’, 88 Michigan Law Review 917 (1990).
Fleming, J. Cliton and Robert A. Peroni, ‘Can Tax Expenditure Analysis Be
Divorced from a Normative Tax Base?: A Critique of the “New Paradigm”
and its Denouement’, 30 Virginia Tax Review 135 (2010).
‘Reinvigorating Tax Expenditure Analysis and its International Dimension’, 27
Virginia Tax Review 437 (2008).
Bibliography 237
Fox, John, If Americans Really Understood the Income Tax (Westview, 2001).
Frank, Robert H., ‘Progressive Taxation and the Incentive Problem’, in Does Atlas
Shrug? (Slemrod (ed.), 2000).
Frank, homas, What’s the Matter with Kansas? (Owl Books, 2004).
Fullerton, Don and Diane Lim Rogers, ‘Distributional Efects on a Lifetime Basis’,
in Distributional Analysis of Tax Analysis (David Bradford (ed.), AEI Press,
1995).
Gadamer, Hans-Georg, Truth and Method (Garrett Barden and John Cumming
(eds and trans.), Seabury, 1975).
Galle, Brian, ‘Hidden Taxes’, 87 Washburn University Law Review 59 (2009).
Ghilarducci, Teresa, Beyond 401(k)s: Guaranteeing Retirement Security for all
Americans (Schwartz Centre for Economic Policy Analysis, 2010).
Glenday, Graham and Zhicheng Li Swit, Establishing Tax Expenditure Accounts –
Towards Better Fiscal Accountability and Transparency (World Bank, 2006).
Goldberg, Fred T. Jr, Peter R. Orszag and Lily L. Batchelder, ‘Eiciency and Tax
Incentives: he Case for Refundable Tax Credits’, 59 Stanford Law Review
23 (2006).
Goldsworthy, Jef rey, ‘Dworkin as an Originalist’, 17 Constitutional Commentary
49 (2000).
‘Legal Intentions, Legislative Supremacy, and Legal Positivism’, 42 San Diego
‘Legislative Intentions, Legislative Supremacy, and Legal Positivism’, in Legal
Interpretation in Democratic States (Goldsworthy and Campbell (eds),
2002).
‘Preface’, in Legal Interpretation in Democratic States (Goldsworthy and
Campbell (eds), 2002).
Goldsworthy, Jef rey and Tom Campbell, Legal Interpretation in Democratic States
(Ashgate, 2002).
Graetz , Michael J., ‘Distributional Tables, Tax Legislation, and the Illusion of
Precision’, in Distributional Analysis of Tax Policy (David F. Bradford (ed.),
AEI Press, 1995).
‘Paint-by-Numbers Tax Lawmaking’, 95 Columbia Law Review 609 (1995).
‘To Praise the Estate Tax, Not to Bury it’, 93 Yale Law Journal 259 (1983).
Graetz, Michael J. and Ian Shapiro, Death by a housand Cuts (Princeton University
Press, 2005).
Gray, John, Two Faces of Liberalism (he New Press, 2000).
Gunz, Sally, Alan MacNaughton and Karen Wensley, ‘Measuring the Compliance
Cost of Tax Expenditures: he Case of Research and Development Incentives’,
43 Canadian Tax Journal 2008 (1995).
Gutmann, Amy and Dennis hompson, Democracy and Disagreement (Belknap
Press, 1996).
Why Deliberative Democracy? (Princeton University Press, 2006).
Habermas, Jurgen, Between Facts and Norms (Polity, 1996).
238 Bibliography
Hameed, Farhan, Fiscal Transparency and Economic Outcomes (International
Monetary Fund, 2005).
Hampshire, Stuart, Justice is Conlict (Princeton University Press, 2000).
Hansen, Phillip, Taxing Illusions (Fernwood, 2003).
Harman, Gilbert and Judith Jarvis homson, Moral Relativism and Moral
Objectivity (Blackwell, 1996).
Harrison, Mark, Valuing the Future: he Social Discount Rate in Cost-beneit
Analysis (Australian Government, Productivity Commission, 2010).
Hart, Herbert L. A., he Concept of Law (Oxford University Press, 1994 (irst pub-
lished 1961)).
Hartle, Douglas G., he Expenditure Process of the Government of Canada: A Public
Choice Rent Seeking Perspective (Canadian Tax Foundation, 1988).
Hayek, Friedrich, he Constitution of Liberty (University of Chicago Press, 1960).
he Road to Serfdom (Bruce Caldwell (ed.), University of Chicago Press, 2007
(originally published 1944)).
Heady, Christopher, ‘Tax Expenditures: Dei nitional and Policy Issues’, in Tax
Expenditures: State of the Art (Lisa Philipps, Neil Brooks and Jinyan Li (eds),
Canadian Tax Foundation, 2011).
Held, David, Models of Democracy (Polity, 2006 (3rd edn)).
Herman, Edward S. and Noam Chomsky, Manufacturing Consent (Pantheon, 2002
Hill, Carolyn J., V. Joseph Hotz and John Karl Scholz, ‘Evaluating Work Related
Cash Beneit Programs: he Earned Income Tax Credit’, in Evaluating Tax
Expenditures: Tools and Techniques for Assessing Outcomes (Lois-Ellin Datta
and Patrick G. Grasso (eds), Jossey Bass, 1998).
Hodge, Scott A., Who Beneits from Corporate ‘Loopholes’ (Tax Foundation, Fiscal
Fact, 2011).
Hovenkamp, Herbert, ‘Legislation, Well-Being and Public Choice’, 57 Chicago
‘Positivism in Law and Economics’, 78 California Law Review 815 (1990).
Howard, Christopher, he Hidden Welfare State (Princeton University Press,
1997).
‘he Politics of Tax Expenditures in Wealthy Democracies’, in Tax Expenditures:
State of the Art (Lisa Philipps, Neil Brooks and Jinyan Li (eds), Canadian Tax
Foundation, 2011).
Howe, John and Ingrid Landau, ‘Do Investment Attraction Incentives Create
Decent Jobs? A Study of Labour Conditions in Industry Assistance Contracts’
(Working Paper, Melbourne Law School, 2008).
Infanti, Anthony C., ‘Tax Equity’, 55 Bufalo Law Review 1191 (2008).
International Monetary Fund, Assessing and Promoting Fiscal Transparency: A
Report on Progress (IMF, 2003).
Manual on Fiscal Transparency (IMF, 2007).
Bibliography 239
Ireland , Interim Report on the Underclaiming of Tax Credits, Allowances and
Reliefs by Taxpayers (Joint Committee on Finance and the Public Service,
2007).
Johnson, Calvin H., ‘Why Have Anti-Tax Shelter Legislation? A Response to
Professor Zelenak ’, 67 Texas Law Review 591 (1989).
Johnson, Nicholas, Jeremy Koulish and Jason Levitis, Promoting State Budget
Accountability hrough Tax Expenditure Reporting (Center on Budget and
Policy Priorities, 2009).
Johnson, Ronald W., Philip G. Joyce and Robert D. Lee Jr, Public Budgeting Systems
(Jones and Bartlett, 2008 (8th edn)).
Johnston, David Cay, Free Lunch (Portfolio, 2007).
Perfectly Legal (Portfolio, 2003).
Joulfaian, David, Robert Carroll and James Mackie, ‘Income Versus Consumption
Tax Baselines for Tax Expenditures’, 64 National Tax Journal 491 (2011).
Kahn, Douglas A., ‘Accelerated Depreciation – Tax Expenditure or Proper
Allowance for Measuring Net Income? ’, 78 Michigan Law Review 1
(1979).
Kahn, Douglas A. and Jef rey S. Lehman, ‘ Tax Expenditure Budgets: A Critical
View’, 54 Tax Notes 1661 (1992).
Kahn, Jef rey A., ‘Personal Deductions – A Tax “Ideal” or Just Another “Deal”?’,
Law Review of Michigan State University – Detroit College of Law 1 (2002).
‘ he Two Faces of Tax Neutrality: Do hey Interact or Are hey Mutually
Exclusive?’, 18 Northern Kentucky Law Review 1 (1990).
Kaplow, Louis, ‘ Taxation and Redistribution: Some Clariications’, 60 Tax Law
Review 57 (2007).
he heory of Taxation and Public Economics (Princeton University Press,
2008).
Kleinbard, Edward D., ‘How Tax Expenditures Distort Our Budget and Our
Political Processes’, 123 Tax Notes 925 (2009).
Tax Expenditure Framework Legislation (University of Southern California
Center in Law, Economics and Organization Research Papers, 2010).
Kobetsky, Michael, International Taxation of Permanent Establishments
(Cambridge University Press, 2011).
Kopits, George and Jon Craig, Transparency in Government Operations (IMF
Occasional Paper No 158, 1998).
Kornhauser, Marjorie E., ‘Legitimacy and the Right of Revolution: he Role of Tax
Protests and Anti-Tax Rhetoric in America’, 50 Bufalo Law Review 819 (2002).
Normative and Cognitive Aspects of Tax Compliance: Literature Review and
Recommendations for the IRS Regarding Individual Taxpayers (National
Taxpayer Advocate, 2007).
‘Tax Policy in an Era of Rising Inequality: Choosing a Tax Rate Structure in the
Face of Disagreement’, 52 UCLA Law Review 1697 (2005).
240 Bibliography
Kraan, Dirk Jan, ‘Of-budget and Tax Expenditures’, 4 OECD Journal on Budgeting
121 (2004).
Krever, Richard, ‘Analysing Implicit Tax Expenditures’, 35 Melbourne University
‘ Taming Complexity in Australian Income Tax’, 25 Sydney Law Review 467
(2003).
Kysar, Rebecca M., ‘he Sun Also Rises: Political Economy of Sunset Provisions in
the Tax Code’, 40 Georgia Law Review 335 (2006).
Laclau, Ernesto and Chantal Moufe, Hegemony and Socialist Strategy (Verso, 2001
Laity, Eric, ‘he Corporation as Administrative Agency: Tax Expenditures and
Institutional Design’, 28 Virginia Tax Review 411 (2008).
Laurie, Kirstie and Jason McDonald , A Perspective on Trends in Australian
Government Spending (Australian Treasury, 2003).
Leachman, Michael, Dylan Grundman and Nicholas Johnson, Promoting State
Budget Accountability hrough Tax Expenditure Reporting (Center on
Budget and Policy Priorities, 2011).
Legatum Institute, he 2010 Legatum Prosperity Index (2010).
Leigh, Andrew, ‘Who Beneits from the Earned Income Tax Credit? Incidence
among Recipients, Co-workers and Firms’, 10 he B.E. Journal of Economic
Analysis and Policy 45 (2010).
Levi, Margaret, Of Rule and Revenue (University of California Press, 1988).
Liebman, Jef rey B., ‘he Impact of the Earned Income Tax Credit on Incentives
and Distribution’, in Tax Policy and the Economy (James Poterba (ed.), MIT
Press, 1998).
Lienert , Ian, Manual on the Role of the Legislature in the Budget Process
(International Monetary Fund, 2010).
Locke, John, ‘he Second Treatise on Civil Government’, in On Politics and
Education (Walter Black, 1947 (originally published 1690)).
Lukes, Steven, Power: A Radical View (New York University Press, 2005 (originally
published 1986)).
MacNevin, Alex S., ‘Issues in Income Tax Accountability: An Analysis of the
Rental Housing Sector’, 32 Alberta Law Review 668 (1994).
Maktouf, Loti and Stanley S. Surrey, ‘Tax Expenditure Analysis and Tax Budgetary
Reform in Less Developed Countries’, 15 Law and Policy in International
Business 739 (1983).
McBarnett, Doreen, ‘When Compliance Is Not the Solution but the Problem: From
Changes in Law to Changes in Attitude’, Working Paper No 18 (Center for
Tax System Integrity, 2001).
McCafery, Edward J., ‘he Missing Links in Tax Reform’, 2 Chapman Law Review
233 (1999).
‘he Uneasy Case for Wealth Transfer Taxation’, 104 Yale Law Journal 283 (1994).
Bibliography 241
McCafery, Edward J. and Linda R. Cohen, ‘Shakedown at Gucci Gulch: he New
Logic of Collective Action’, 84 North Carolina Law Review 1159 (2005).
McCafery, Edward, J. and Joel Slemrod (eds), Behavioural Public Finance (Russell
Sage, 2006).
McCloskey, Deidre N., he Rhetoric of Economics (University of Wisconsin Press,
1998 (2nd edn)).
McCloskey, Donald N., Knowledge and Persuasion in Economics (Cambridge
McDaniel, Paul R., ‘Identiication of the “Tax” in “Efective Tax Rates”, “Tax
Reform” and “Tax Equity” ’, 38 National Tax Journal 273 (1985).
‘he Staf of the Joint Committee on Taxation Revision of Tax Expenditure
Classiication Methodology: What Is to Be Made of a Change that Makes No
Changes?’, in Tax Expenditures: State of the Art (Lisa Philipps, Neil Brooks
and Jinyan Li (eds), Canadian Tax Foundation, 2011).
McDaniel, Paul R. and Stanley S. Surrey, International Aspects of Tax Expenditures:
A Comparative Study (Kluwer, 1985).
McGarity, homas, Reinventing Rationality (Cambridge University Press, 1991).
McGee, Robert W., he Philosophy of Taxation and Public Finance (Kluwer, 2004).
McIntyre, Michael J., ‘A Solution to the Problem of Deining a Tax Expenditure’, 14
UC Davis Law Review 79 (1980 –81).
McLure, Charles, ‘he Budget Process and Tax Simpliication/Complication’, 45
Tax Law Review 25 (1990).
McManus, Jacqui and Neil Warren, ‘he Case for Measuring Tax Gap’, 4 eJournal
of Tax Research 61 (2006).
McQuaig, Linda, Behind Closed Doors (Penguin, 1987).
he Cult of Impotence (Viking, 1998).
Metcalf, Gilbert , ‘Environmental Taxation: What Have We Learned in h is
Decade?’, in Tax Policy Lessons from the 2000s (Alan D. Viard (ed.), AEI
Press, 2009).
‘Using Tax Expenditures to Achieve Energy Policy Goals’, 98 American Economic
Review 90 (2008).
Mills, C. Wright, Power, Politics and People (Oxford University Press, 1963).
Milne, Janet E., ‘US Climate Change Policy: A Tax Expenditure Microcosm with
Environmental Dimensions’, in Tax Expenditures: State of the Art (Lisa
Philipps, Neil Brooks and Jinyan Li (eds), Canadian Tax Foundation, 2011).
Misiolek, Walter S. and Harold W. Elder, ‘Tax Structure and the Size of Government:
An Empirical Analysis of the Fiscal Illusion and Fiscal Stress Arguments’, 57
Public Choice 233 (1988).
Montesinos, Vicente, ‘Government Budgeting and Accounting Reforms in Spain’,
2 OECD Journal on Budgeting 333 (2002).
Moufe, Chantal, Deconstruction and Pragmatism (Routledge, 1996).
Mueller, Dennis C., Public Choice III (Cambridge University Press, 2003).
242 Bibliography
Murnane, Susan, ‘Selling Scientiic Taxation: he Treasury Department’s Campaign
for Tax Reform in the 1920s’, 29 Law and Social Inquiry 819 (2004).
Murphy, Liam and homas Nagel, he Myth of Ownership (Oxford University
Press, 2002).
Nordic Working Group, Tax Expenditures in the Nordic Countries (2010).
Nozick, Robert, Anarchy, State, and Utopia (Blackwell, 1974).
OECD, Best Practice Guidelines – Of Budget and Tax Expenditures (OECD, 2004).
Best Practices for Budget Transparency (OECD, 2002).
Citizens as Partners (OECD, 2001).
Compliance Measurement – Practice Note (OECD, 2001).
Compliance Risk Management: Managing and Improving Tax Compliance
(OECD, 2010).
Corporate Governance and Tax Risk Management (OECD, 2009).
Environmentally Related Taxes and Charges in OECD Countries: Issues and
Strategies (OECD, 2001).
How’s Life? (OECD, 2011).
Inventory of Estimated Budgetary Support and Tax Expenditures for Fossil Fuels
(OECD, 2011).
Journal on Budgeting (OECD, 2002).
Special Feature for the 2003 Edition of Revenue Statistics: Note by the Secretariat
(OECD, 2003).
Tax Administration in OECD Countries: Comparative Information Series
(OECD, 2004).
Tax Expenditures: Recent Developments (OECD, 1996).
Tax Expenditures: A Review of Issues and Country Practices (OECD, 1984).
Tax Expenditures in OECD Countries (OECD, 2010).
Ofe, Claus, Contradictions of the Welfare State (MIT Press, 1984).
Oppenheimer, Joe A., ‘Public Choice and hree Ethical Properties of Politics’, 45
Public Choice 241 (1985).
O’Reilly, Terrance, ‘Principles of Eicient Tax Law: Apocryphia’, 27 Virginia Tax
Review 583 (2008).
Patriquin, Larry, Inventing Tax Rage (Fernwood, 2004).
Pechman, Joseph, Federal Tax Policy (he Brookings Institution, 1987).
Options for Tax Reform (he Brookings Institution, 1984).
Who Paid the Taxes, 1966–85? (he Brookings Institution, 1985).
Peltzman, Sam, ‘Constituent Interest and Congressional Voting’, 27 Journal of Law
and Economics 181 (1984).
Poddar, Satya, ‘Integration of Tax Expenditures into the Expenditure Management
Experience: he Canadian Experience’, in Tax Expenditures and Government
Policy (Bruce (ed.), 1988).
Bibliography 243
Polackova Brixi, Hana, Christian Valenduc and Zhicheng Li Swit (eds), Tax
Expenditures – Shedding Light on Government Spending hrough the Tax System:
Lessons from Developed and Transition Economies (World Bank, 2004).
Pollack, Sheldon D., he Failure of US Tax Policy (University of Pennsylvania Press,
1996).
Pusey, Michael, Economic Rationalism in Canberra (Cambridge University Press,
1991).
he Experience of Middle Australia (Cambridge University Press, 2003).
Rawls, John, A heory of Justice (Belknap Press, 1971, revised edn 2005).
Rea, Samuel A. Jr, ‘Registered Retirement Savings Plans as a Tax Expenditure’, 28
Canadian Tax Journal 459 (1980).
Reese, homas J., he Politics of Taxation (Quorum Books, 1980).
Repetti, James R., ‘Democracy and Opportunity: A New Paradigm in Tax Equity’,
61 Vanderbilt Law Review 1129 (2008).
‘Democracy, Taxes and Wealth ’, 76 New York University Law Review 825 (2001).
Republic of South Africa, 2012 National Budget (2012).
Richardson, Maryann and A. J. Sawyer, ‘A Taxonomy of the Tax Compliance
Literature: Further Findings, Problems and Prospects’, 16 Australian Tax
Forum 137 (2001).
Roin, Julie, ‘Can the Income Tax Be Saved – he Promise and Pitfalls of Adopting
Worldwide Formulary Apportionment’, 61 Tax Law Review 169 (2008).
‘Truth in Government: Beyond the Tax Expenditure Budget’, 54 Hastings Law
Journal 603 (2003).
Roth, Jef rey, John Scholz and Ann Witte, Taxpayer Compliance, vol. I (University
of Pennsylvania Press, 1989).
Rothstein, Jesse, he Unintended Consequences of Encouraging Work: Tax Incidence
and the EITC (Center for Economic Policy Studies, Princeton University,
2008).
Rousseau, Jean-Jacques, he Social Contract (Penguin, 1968).
Rubin , Ed , ‘The Myth of Non-Bureaucratic Accountability and the
Anti-Administrative Impulse’, in Public Accountability (Dowdle (ed.),
2006).
‘Public Choice, Phenomenology, and the Meaning of the Modern State: Keep the
Bathwater, but hrow Out the Baby’, 87 Cornell Law Review 309 (2002).
Sadiq, Kerrie, ‘Commentary: he Under-Appreciated Implications of the Tax
Expenditure Concept’, in Australian Business Tax Reform in Retrospect and
Prospect (Chris Evans and Richard Krever (eds), homson Reuters, 2009).
Salamon, Lester M., ‘he New Governance and the Tools of Public Action: An
Introduction’, 28 Fordham Urban Law Journal 1611 (2001).
Sandford, Cedrick, Michael Godwin and Peter Hardwick, Administrative and
Compliance Costs of Taxation (Fiscal Publications, 1989).
244 Bibliography
Scalia, Antonin, ‘Originalism: he Lesser Evil ’, 57 Cincinnatti Law Review 849
(1989).
Schattschneider, E. E., he Semisovereign People (Harcourt Brace, 1975).
Schenk, Deborah, ‘Exploiting the Salience Bias in Designing Taxes’, 28 Yale Journal
on Regulation 253 (2011).
Schoenblum, Jef rey A., ‘ Tax Fairness or Unfairness? A Consideration of the
Philosophical Bases for Unequal Taxation of Individuals’, 12 American
Journal of Tax Policy 221 (1995).
Schumpeter, Joseph A., Capitalism, Socialism and Democracy (Harper, 1947
(2nd edn)).
Searchinger, Timothy, Ralph Heimlich, R. A. Houghton et al., ‘Use of US Croplands
Increases Greenhouse Gases h rough Emissions from Land Use Change’,
319 Science 1238 (2008).
Sen, Amartya, he Idea of Justice (Harvard University Press, 2009).
Shannon, Harry A., ‘ he Tax Expenditure Concept in the United States and
Germany: A Comparison’, 33 Tax Notes 201 (1986).
Shapiro, Ian, he State of Democratic heory (Princeton University Press, 2003).
Shaviro, Daniel N., ‘Rethinking Tax Expenditures and Fiscal Language’, 57 Tax
Simons, Henry, Personal Income Taxation (University of Chicago Press, 1938).
Slemrod, Joel B., Does Atlas Shrug? (Russell Sage, 2000).
‘he Economics of Taxing the Rich’, in Slemrod (ed.) Does Atlas Shrug? (2000).
Tax Policy in the Real World (Cambridge University Press, 1999).
Smith, Adam, An Inquiry into the Nature and Causes of the Wealth of Nations
(Kathryn Sutherland (ed.), Oxford University Press, 1993 (originally pub-
lished 1776)).
Spicer, Michael, In Defense of Politics in Public Administration (University of
Alabama Press, 2010).
Stanley, Robert, Dimensions of Law in the Service of Order (Oxford University
Press, 1993).
Staudt, Nancy, ‘he Hidden Costs of the Progressivity Debate’, 50 Vanderbilt Law
Review 919 (1997).
‘Redundant Tax and Spending Programs’, 100 Northwestern University Law
Review 1197 (2006).
Stead, Merideth M., ‘Implementing Disaster Relief hrough Tax Expenditures: An
Assessment of the Katrina Emergency Tax Relief Measures’, 81 New York
Steinmo, Sven, Taxation and Democracy (Yale University Press, 1993).
Steuerle, C. Eugene, ‘Can the Progressivity of Tax Changes Be Measured in
Isolation?’, 100 Tax Notes 1187 (2003).
Contemporary US Tax Policy (Urban Institute Press, 2004).
‘Efects of the Budget Process on Tax Legislation’, 10 American Journal of Tax
Policy 141 (1992).
Bibliography 245
Stewart, Miranda and Sunita Jogarajan, ‘he International Monetary Fund and
Tax Reform’ 2 British Tax Review 146 (2004).
Stewart, Miranda and Lisa Philipps, ‘Deining Fiscal Transparency: Transnational
Norms, Domestic Laws, and the Politics of Budget Accountability’, 34
Brooklyn Journal of International Law 797 (2009).
Stiglitz, Joseph E., ‘Of the 1%, for the 1%, by the 1%’, Vanity Fair (May 2011), available
at: www.vanityfair.com/society/features/2011/05/top-one-percent-201105
(last accessed 24 August 2012).
Stiglitz, Joseph, Amartya Sen and Jean-Paul Fitoussi, Report by the Commission
on the Measurement of Economic Performance and Social Progress (Paris,
2009), available at: www.stiglitz-sen-itoussi.fr/documents/rapport_anglais.
pdf (last accessed 24 August 2012).
Sugin, Linda, ‘ Tax Expenditures, Reform, and Distributive Justice’, 3 Columbia
Tax Journal 1 (2011).
Surrey, Stanley S., Excerpts from Remarks Before the Money Marketeers on the US
Income Tax System – he Need for a Full Accounting in the Annual Report of
the Secretary of the Treasury on the State of the Finances for the Fiscal Year
Ended June 30, 1968 (United States Department of the Treasury, 1968).
Pathways to Tax Reform: he Concept of Tax Expenditures (Harvard University
Press, 1973).
Surrey, Stanley S. and Paul R. McDaniel, ‘he Tax Expenditure Concept and the
Budget Reform Act of 1974’, 17 Boston College Industrial and Commercial
Tax Expenditures (Harvard University Press, 1985).
Swift , Zhicheng Li, Hana Polackova Brixi and Christian Valenduc , ‘Tax
Expenditures: General Concept, Measurement, and Overview of Country
Practices’, in Tax Expenditures – Shedding Light on Government Spending
through the Tax System: Lessons from Developed and Transition Economies
(Polackova Brixi, Valenduc and Swit (eds), 2004).
Swift , Zhicheng Li, Christian Valenduc and Hana Polackova Brixi, ‘Tax
Expenditures: General Overview’, in Tax Expenditures – Shedding Light on
Government Spending hrough the Tax System: Lessons from Developed and
Transition Economies (Polackova Brixi, Valenduc and Swit (eds), 2004).
Tanzi, Vito, Globalization, Technological Developments, and the Work of Fiscal
Termites, International Monetary Fund Working Paper WP/00/181 (IMF,
2000).
huronyi, Victor, ‘ Tax Expenditures: A Reassessment ’, Duke Law Journal 1155
(1988).
Toder, Eric, Eliminating Tax Expenditures with Adverse Environmental Efects
(The Brookings Institution, 2007), available at: http://pdf.wri.org/
Brookings-WRI_TaxExpenditures.pdf (last accessed 20 August 2012).
Toder, Eric, Benjamin H. Harris and Katherine Lim, Distributional Efects of Tax
Expenditures (Urban Brookings Tax Policy Center, 2011).
246 Bibliography
Torgler, Benno and Kristina Murphy, ‘Tax Morale in Australia: What Shapes it and
Has it Changed Over Time?’, 7 Journal of Australian Taxation 298 (2004).
Tyler, Tom, Why People Obey the Law (Princeton University Press, 1990, revised
edn 2006).
United States of America, 2010 Report of the National Taxpayer Advocate (National
Taxpayer Advocate, 2010).
2011 Report of the National Taxpayer Advocate (National Taxpayer Advocate,
2011).
(2001).
(2012).
Biofuels – Potential Efects and Challenges of Required Increase in Production
(Government Accountability Oice, 2009).
Economic Efects of Reducing the Fiscal Restraint hat Is Scheduled to Occur in
2013 (Congressional Budget Oice (ed.), Congressional Budget Oice, 2012).
Estimates of Federal Tax Expenditures for Fiscal Years 2009–2013 ( Joint
Committee on Taxation, US Government Printing Oice, 2010).
he Impact of Ethanol Use on Food Prices and Greenhouse Gas Emissions
(Congressional Budget Oice, 2009).
The Moment of Truth (Fiscal Commission, 2010), available at: www.
fiscalcommission.gov/sites/fiscalcommission.gov/files/documents/
heMomentofTruth12_1_2010.pdf (last accessed 24 August 2012).
Opportunities to Reduce Potential Duplication in Government Programs, Save
Tax Dollars, and Enhance Revenue (Government Accountability Oice,
2011).
A Reconsideration of Tax Expenditure Analysis (Joint Committee on Taxation,
2008).
Using Biofuel Tax Credits to Achieve Energy and Environmental Policy Goals
(Congressional Budget Oice, 2010).
Why Have Food Commodity Prices Risen Again? (United States Department of
Agriculture (ed.), United States Department of Agriculture, 2011).
Vazquez , Jorge Martinez , Friedrich Schneider and James Alm, ‘ “Sizing” the
Problem of the Hard-to-Tax’, in Taxing the Hard-to-Tax (Alm, Vazquez and
Wallace (eds), 2004).
Ventry, Dennis J. Jr, ‘Equity versus Eiciency and the US Tax System in Historical
Perspective’, in Tax Justice (Joseph J. horndike and Dennis J. Ventry Jr
(eds), Urban Institute Press, 2002).
Viard, Alan D., Tax Policy Lessons from the 2000s (AEI Press, 2009).
Villela, Luiz, Andrea Lemgruber and Michael Jorratt, Tax Expenditure Budgets:
Concepts and Challenges for Implementation (Inter African Development
Bank, 2010).
Bibliography 247
Waldron, Jeremy, ‘Can here Be a Democratic Jurisprudence?’, 58 Emory Law
Journal 675 (2009).
he Dignity of Legislation (Cambridge University Press, 1999).
‘Representative Lawmaking’, 89 Boston University Law Review 335 (2009).
Weber, Edward P., Pluralism by the Rules (Georgetown University Press, 2007).
Weber, Max, he Protestant Ethic and the Spirit of Capitalism (Routledge, 1992
Weisbach, David, ‘Implications of Implicit Taxes’, 52 SMU Law Review 373 (1999).
‘Line Drawing Doctrine and Eiciency in the Tax Law’, 84 Cornell Law Review
1627 (1999).
‘Tax Expenditures, Principal-Agent Problems, and Redundancy’, 84 Washington
Weisbach, David and Jacob Nussim, ‘ he Integration of Tax and Spending
Programs’, 113 Yale Law Journal 955 (2004).
Weisbach, David and Cass R. Sunstein, ‘Climate Change and Discounting the Future:
A Guide for the Perplexed ’, 27 Yale Law and Policy Review 433 (2009).
Wiedenbeck, Peter J., ‘Paternalism and Income Tax Reform’, 33 University of
Kansas Law Review 675 (1985).
Wildavsky, Aaron, ‘Keeping Kosher: he Epistemology of Tax Expenditures’, 5
Journal of Public Policy 413 (1985).
he New Politics of the Budgetary Process (Scott, Foresman and Co., 1988).
Witte, John F., he Politics and Development of the Federal Income Tax (University
of Wisconsin Press, 1985).
Woellner, Andrew, ‘Spending on an Empty Wallet: A Critique of Tax Expenditures
and the Current Fiscal Policy’, 7 Houston Business & Tax Law Journal 201
(2006).
Workman, hom, Banking on Deception (Fernwood Publishing, 1996).
World Bank, Demand for Good Governance in the World Bank: Conceptual
Evolution, Frameworks and Activities (World Bank, 2010).
Gender and Environment, Issues Paper (World Bank, 2010), available at: http://
siteresources.worldbank.org/EXTENVSTRATEGY/Resources/6975692–
1289855310673/20101210-Gender-and-Environment-Issues.pdf (last
accessed 24 August 2012).
World Trade Organization, World Trade Report 2006 – Exploring the Links between
Subsidies, Trade and the WTO (World Trade Organization, 2006).
Yin, George K., ‘Temporary-Efect Legislation, Political Accountability, and Fiscal
Restraint’, 84 New York University Law Review 174 (2009).
Yorio, Edward, ‘he Future of Tax Reform: A Rejoinder to Professor Zelinsky’, 55
Fordham Law Review 899 (1987).
Zelenak, Lawrence, ‘When Good Preferences Go Bad: A Critical Analysis of the
Anti-Tax Shelter Provisions of the Tax Reform Act of 1986’, 67 Texas Law
Review 499 (1989).
248 Bibliography
Zelinsky, Edward , ‘Are Tax “Beneits” Constitutionally Equivalent to Direct
Spending?’, 112 Harvard Law Review 380 (1999).
‘Do Tax Expenditures Create Framing Efects? Volunteer Fireighters, Property
Tax Exemptions, and the Paradox of Tax Expenditure Analysis’, 24 Virginia
Tax Review 797 (2005).
‘Eiciency and Income Taxes: he Rehabilitation of Tax Incentives’, 64 Texas
‘James Madison and Public Choice at Gucci Gulch: A Procedural Defense of
Tax Expenditures and Tax Institutions’, 102 Yale Law Journal 1165 (1993).
‘Winn and the Inadvisability of Constitutionalizing Tax Expenditure Analysis’,
121 he Yale Law Journal Online 25 (2011).
Zelizer, Julian E., Taxing America: Wilbur D. Mills, Congress, and the State 1945–
1975 (Cambridge University Press, 1998).
I N DE X
Abel, Andrew B., 164 revenue gain approach, 97

Abrams, Howard, 114 Austria
accountability, 170 deinition of benchmark tax in
deliberative concepts of, 191–2, law, 85
see also elitism regulatory status of tax expenditures
Agreement on Subsidies and reporting, 80
Countervailing Measures revenue foregone approach, 94
(SCM Agreement), 132–4 Avi-Yonah, Reuven S., 110, 150–1, 167,
alcohol, taxes on, 22 206
Allan, William, 84 Azmat, Ghazala, 117, 227
allocation eiciency, 122–7
allocation of resources, 6 Bankman, Joseph, 167
Alm, James, 115 Barry, Brian, 210
Andrews, William, 168 Bartlett, Bruce, 103, 176, 205
Argentina, deinition of benchmark Batchelder, Lily L., 117, 122, 123, 161,
tax in law, 85 215, 227
Arizona Christian School Tuition Belgium
Organization v. Winn 131 S. Ct. deinition of benchmark tax in
1436 (2011), 131–2 law, 85
Arnold, Brian, 118 deinition of tax expenditures, 83
Australia functions of tax expenditures
conceptual benchmark approach, reporting, 78
87–8 regulatory status of tax expenditures
deinition of benchmark tax in reporting, 80
law, 85 revenue foregone approach, 94
deinition of tax expenditures, 82 benchmark tax
distinction between compliance and deinition in law, 83–5
accessing beneits, 137 OECD guidelines, 71
functions of Tax Expenditures Berlin, Isaiah, 158
Statement, 78 Bimber, Bruce, 192
regulatory status of tax expenditures Bird, Richard M., 167
reporting, 80 Bittker, Boris, 9, 42–3, 103, 152, 176,
reporting structure, 98, 99 205
restrictions on reporting, from Blatt, William, 190
resource and data Block, Cheryl D., 222
constraints, 93 Blount, Simon, 185–6
revenue foregone approach, 94 Blum, Walter J., 158
249
250 Index
Boadway, Robin, 110, 132, 219 comprehensive income tax
Boogert, Kees den, 25 benchmark, 34
Braithwaite, John, 146 conceptual benchmark approach,
Brazil 85–6
deinition of benchmark tax in law, 85 examples, 87–9
regulatory status of tax expenditures Congressional Budget and
reporting, 80 Impoundment Control Act (US
Brennan, Geof rey, 108, 204 1974), 87
Brocas, Isabelle, 189 constitutional power, circumvention of
Brooks, Neil, 64, 101, 138 restrictions, 130–2
Bruce, Neil, 103 consumption, as tax baseline, 54–5
Buchanan, James, 108, 204 Cooper, Graeme, 110
Burg, David F., 144, 177 corporate tax, 53–4
Burman, Leonard, 122 Craig, Jon, 84, 171–2
Burton, Mark, 6, 20, 69, 94, 96, 135, Crandall, William, 139
146, 220 Crane, Charlotte, 205
Byrne, Donna M., 122, 123, 160, 161 creative compliance, 116
Campbell, Tom, 178, 179, 180 D’Ascenzo, Michael, 179

Canada Dabner, Justin, 135
budget envelope approach, 223 Dahl, Robert A., 185
conceptual benchmark approach, Davidson, Sinclair, 128
88–9 deLeon, Peter, 172, 176, 206, 207
deinition of benchmark tax in deliberation, 192–3
law, 85 institutional aspects, 193–5
deinition of tax expenditures, 83 deliberative democracy, 195–6
regulatory status of tax expenditures alternate procedural rules where
reporting, 80 deliberation fails, 200–1
reporting structure, 98, 99 information requirements, 201–2
revenue foregone approach, 94 just causes and non-deliberative
Caplan, Bryan, 208 strategies, 199–200
Caribata, Patrick J., 127 neutrality and, 196–8
Carling, Robert, 128 power imbalance and minimisation
Carmody, Michael, 178 of disagreement, 198–9
Carrillo, Juan D., 189 democracy, 169–70
Carroll, Robert, 55 importance of type aspired to,
Cavanaugh, Maureen B., 18 212–15
Chile meaningful engagement, 214, 216
deinition of benchmark tax in democratic positivism
law, 85 democratic aspect, 178–9
deinition of tax expenditures, 83 democratic concept, 180–2,
regulatory status of tax expenditures see also descriptive democratic
reporting, 80 positivism; prescriptive
Chomsky, Noam, 190 democratic positivism
Cockett, Richard, 155 democratic transparency, 172
Cohen, G. A., 111, 161, 190 Denmark
Colombia, deinition of benchmark tax outlay equivalence approach, 97
in law, 85 revenue foregone approach, 94
Index 251
deontological norms, relevance to tax as wealth maximisation, 160, 164–5
expenditure management, application in real world of the
154–5 second best, 163
descriptive democratic positivism, 179 empirical limitations, 163
developing nations, policy options geographical scope, 164
for, 74 rhetorical appeal, 162
discretion by revenue collecting theoretical disagreements, 162
authority, generating tax as welfare maximisation, 161
expenditures, 61–2 Eisenstein, Louis, 60
distribution of wealth, 6–7 Eissa, Nada, 117, 142
distributive justice, 165 elitism, 173–7
classiication of recipient groups, Elson, Diane, 121
120–1 Ende, Leo van den, 25
deinitions, 109 excise tax breaks, outside narrow
determining time of impact, 118–20 deinition of tax
environmental outcomes, 118 expenditures, 2
identifying beneiciaries, 114–17 expenditure tax benchmark, 34
impacts to be subjected to external benchmark approach, 85
distributional analysis, 111–14
implementation by taxation, 167–8 FCT v. Orica Ltd (1998) 98 ATC 4494,
independent status of the 132
distributive justice norm, Feldstein, Martin, 157
110–11 Fiekowsky, Seymour, 50
international justice, 117–18 Finland
normative tax redistribution deinition of benchmark tax
principle, 168–9 in law, 85
overview, 121–2 revenue foregone approach, 94
questions identifying distributive iscal federalism, 132
impacts, 109–10 iscal illusion of welfare state, 127–8
role of state, 165–7 Fitoussi, Jean-Paul, 107, 161, 214
Dodge, Joseph M., 166 Fitts, Michael A., 208
Doernberg, Richard L., 184, 185 Flast v. Cohen 392 US 83 (1968), 131
Dowdle, Michael W., 170, 173 Fleming, J. Cliton, 6, 10, 27, 28, 42,
Downs, Anthony, 4, 172 49–51, 55, 57, 78, 82, 162, 168,
Driessen, Patrick, 123 204, 212, 216, 228
Duf, David G., 111, 156, 216 Fox, John, 7
France
Easson, A. J., 124 deinition of benchmark tax in law, 85
EC Treaty, Article 90, 133 deinition of tax expenditures, 83
Edgar, Tim, 204, 208 functions of tax expenditures
Edrey, Yoseph, 114 reporting, 78
eiciency reference benchmark approach, 90
deinitions of, 160 regulatory status of tax expenditures
time frame used, 163–4, reporting, 80
see also allocation eiciency; reporting structure, 98, 99
Kaldor-Hicks eiciency; Pareto revenue foregone approach, 94
optimal eiciency; technical Frank, Robert H., 160, 163, 207
eiciency Fullerton, Don, 118
252 Index
Gadamer, Hans-Georg, 205 Hovenkamp, Herbert, 187
Galle, Brian, 187 Howard, Christopher, 5, 108, 117, 167,
Geissler, Christopher, 122 211, 223, 224
Germany Howard, John, 186
deinition of benchmark tax in Howe, John, 31
law, 85 Hoynes, Hilary, 117, 142
regulatory status of tax expenditures human rights, 193–4
reporting, 80
reporting structure, 98–9 ideal benchmark approach, 85
restrictions on reporting, from identiication (of tax expenditures)
resource and data approaches to, 24–5
constraints, 93 functions of, 28–9
revenue foregone approach, 94 accountability and transparency,
substitution benchmark approach, 31–3
90–1 reporting, 29–31
Goldberg, Fred T., Jr., 123, 161 issues arising, 35–7
Goldsworthy, Jef rey, 178, 179, 180, 181 purpose of, 3–4, 77–9
good tax policy, 27–8 IMF
government assessment of sovereign debt,
size of, 127–8, see also state 129–30
neutrality identifying best practice, 68
libertarianism, 158 reporting of tax expenditures, 74–5
government interventions, assessment transparency, deinitions, 171
of, 3–4 implicit tax expenditures,
Graetz, Michael J., 116, 162, 190 see operational/implicit tax
Gray, John, 169 expenditures
Grundman, Dylan, 63, 92, 93 implied legislative tax expenditure, 62
Gunz, Sally, 215 India
Gutmann, Amy, 195–202, 209–10 purpose of tax expenditures
reporting, 79
Haberham, Amir, 25 regulatory status of tax expenditures
Habermas, Jurgen, 147 reporting, 80
Hameed, Farhan, 171 restrictions on reporting, from
Hansen, Phillip, 191, 194 resource and data constraints, 93
Harman, Gilbert, 147 revenue foregone approach, 94
Harris, Benjamin H., 122 individual autonomy, 155–6
Hart, H.L.A., 178 information asymmetry, 124
Hartle, Douglas G., 184, 185, 187–8, Inland Revenue Commissioners
189, 223 v. National Federation of
Hayek, Friedrich A., 155–6, 157–8, 164, Self-Employed and Small
169, 204 Businesses Ltd [1982] AC 617,
Heady, Christopher, 34, 48, 53, 54 141
Held, David, 13, 170 International Aspects of Tax
Herman, Edward S., 190 Expenditures (Surrey and
Hewson, John, 186 McDaniel), 42
Hill, Carolyn J., 117 international trade law, 132–4
Hodge, Scott A., 145 inverted distributional efect/
holistic approach, 104–7 upside-down efect, 55, 57–8, 112
Index 253
Ireland liberal-democratic political theory,
deinition of benchmark tax in and tax expenditures, 11–13
law, 85 libertarianism, 155, 159
overpayment of tax, 137–8 negative deinition of freedom and
Italy redistribution, 157–8
regulatory status of tax expenditures size of government and iscal
reporting, 80 illusion, 158
reporting structure, 98, 99 tax base and redistribution, 156–7,
revenue foregone approach, 94 see also individual autonomy
Liebman, Jef rey B., 143, 224, 228
Jogarajan, Sunita, 130 Lienert, Ian, 33, 68, 74, 75
Johnson, Nicholas, 63, 92, 93, 140, 228 Lim, Katherine, 122
Johnston, David Cay, 108 linguistic approach to identifying tax
joint taxation, 53 expenditures, 44
Jorratt, Michael, 20, 69, 84–5, 94, 99 Locke, John, 156, 166
Joulfaian, David, 55 Long, Susan B., 137
judicial decisions, generating tax loopholes, 1
expenditures, 61–2 Lukes, Steven, 167, 180, 194, 198, 210
Kahn, Douglas A., 43, 45, 76, 119, 217 Mackie, James, 55
Kahn, Jef rey A., 204 MacNevin, Alex S., 103, 104, 115, 123
Kaldor-Hicks eiciency, 123 Maktouf, Loti, 121, 151, 174, 218, 229
Kalven, Henry, 158 management of tax expenditures, 29
Kaplow, Louis, 4, 106, 111, 131, 164 Manual on Fiscal Transparency (IMF),
Kleinbard, Edward D., 49, 50, 129, 74–5
145, 146, 150, 176, 203, 208, Manual on the Role of the Legislature in
222 the Budget Process (IMF), 74
Kobetsky, Michael, 141 market transparency, 172–3
Kopits, George, 171–2 McBarnett, Doreen, 116
Korea, deinition of benchmark tax in McCafery, Edward J., 188, 190
law, 85 McChesney, Fred S., 184, 185
Kornhauser, Marjorie, 139, 155, 156, McCloskey, Deidre N., 9, 210, 213
166, 177, 199, 220 McDaniel, Paul R., 1, 2, 9, 15–16, 23, 33,
Koulish, Jeremy, 140 34, 35, 36, 39–42, 52, 64, 65, 69,
Kraan, Dirk Jan, 70–1, 205 84, 85–6, 87, 110, 130, 138, 150,
Krever, Richard, 61, 140 168, 174, 216, 218
Kysar, Rebecca M., 222 McDonald, Jason, 129
McGarity, homas, 4, 105, 155, 158
Laclau, Ernesto, 210 McGee, Robert W., 106
Laity, Eric, 60 McIntyre, Michael J., 19, 44, 204
Landau, Ingrid, 31 McLure, Charles, 145, 175
Laurie, Kirstie, 129 McQuaig, Linda, 145, 146, 176, 191
Leachman, Michael, 63, 92, 93 measurement
Lehman, Jef rey S., 43, 45, 76, 217 approaches, 58, 93–4
Leigh, Andrew, 117 issues, 10, 55–8, see also outlay
Lemgruber, Andrea, 20, 69, 84–5, 94, 99 equivalence approach; revenue
Levi, Margaret, 108 foregone approach; revenue
Levitis, Jason, 140 gain approach
254 Index
Metcalf, Gilbert E., 124, 126, 216 OECD
Miller, James C., 199, 220 reporting of tax expenditures, 69–72
Mills, C. Wright, 173–4 studies by, 34–5
Milne, Janet E., 126 Ofe, Claus, 210
Montesinos, Vicente, 219 operational/implicit tax expenditures,
moral objectivity 60–1, 63, 134, 140–1
connection to politics of tax Oppenheimer, Joe A., 192
expenditure management, 149 O’Reilly, Terrance, 125
desire for, 147 Orszag, Peter R., 123, 161
relativist concept, 148 outlay equivalence approach, 58, 97
universalist concept, 147–8 over-compliance, 137–8
Moufe, Chantal, 210
Mueller, Dennis C., 183 Pareto optimal eiciency, 160–1
Murnane, Susan, 9 Pareto optimality, 122–3
Murphy, Kristina, 144 and distributive justice, 111
Murphy, Liam, 136, 165, 166 Pareto superiority, 123
Pathways to Tax Reform (Surrey), 38
Nagel, homas, 136, 165, 166 Patriquin, Larry, 176–7, 194
negative tax expenditures, 22 Peltzman, Sam, 185
Netherlands Peroni, Robert A., 10, 27, 28, 42, 49–51,
deinition of benchmark tax in law, 85 57, 168, 204, 212, 216, 228
deinition of tax expenditures, 83 Philipps, Lisa, 130, 171
reference benchmark Polackova Brixi, Hana, 160, 204
approach, 90 policy neutral tax provisions, 60
reporting structure, 98 political tensions, 206–8
revenue foregone approach, 94 political transparency, 172–3
New Zealand politicians
deinition of tax expenditures, 83 principal/agent approach to the
non-comprehensiveness of political process, 185–6
reporting, 91–2 as purveyors of legislative favour,
non-reporting of tax 184–5
expenditures, 79 Portugal, deinition of benchmark tax
non-enaction of a taxation law, 10 in law, 85
non-express tax expenditures, power, 198–9
22–3, 59–63, 140, prescriptive democratic positivism,
see also operational/implicit tax 179–80
expenditures press freedom, 194–5
normative tax base public administration
criticism of, 42–51 performance measurement, 139
need for, 51 signiicance of tax expenditures for,
process to establish, 51–5 7, 134–5
Surrey’s work on, 37–42 public choice
Norway, revenue foregone acceptance of American
approach, 94 entrepreneurial dream, 189–91
Nozick, Robert, 156 application to tax expenditures, 186–7
Nussim, Jacob, 3–4, 43, 48–9, 56, 63, individuals pursue own self interest,
104, 105, 151, 204, 217 183–6
Index 255
limitations of theory, 187–9 Schanz–Haig–Simons deinition of
Pusey, Michael, 122, 155 income, 27–8, 39–40, 41–2,
50–1
Rawls, John, 148, 161, 166, 167, 193 in conceptual benchmark, 85
Rea, Samuel A., Jr., 120 Schenk, Deborah, 175, 187
reciprocity, 197 Schoenblum, Jef rey A., 157, 159
reference benchmark approach, 85, Scholz, J., 136–7
89–90 Schumpeter, Joseph A., 174, 191
examples, 90 Schwartz, Richard D., 137
refundable tax credits, 113–14 Searchinger, Timothy, 127
Repetti, James R., 162, 166 second best, theory of the, 105, 124–5
reporting (on tax expenditures) Sen, Amartya, 107, 111, 161, 168, 192–5,
best practice, 67, 68–9, 75–6, 97–8, 209, 214
101–2 Shah, Anwar, 132
comprehensiveness of, 91 Shannon, Harry A., 2, 23, 24, 28, 32, 55,
limiting nature of tax 85, 90–1
expenditures deinitions, 91–2 Shapiro, Ian, 189, 190, 191
resource and data constraints, Shaviro, Daniel N., 4, 19, 20, 45–8, 60,
92–3 172, 208
impact of purpose and function, Simons, Henry, 40, 167
76–7 Slemrod, Joel, 163, 188
on growth and maturity of Smith, Adam, 6, 160
reporting, 77–9 South Africa
on regulatory status, 80 deinition of tax expenditures, 82–3
on resource and data constraints, purpose of tax expenditures
80–1 reporting, 79
information per expenditure, 100 regulatory status of tax expenditures
international aspects, 68–9 reporting, 80
OECD, 69–72 reporting structure, 100
rise of, 65–7 restrictions on reporting, from
structure, 98–100 resource and data
revenue foregone approach, 58, 94–6 constraints, 93
revenue gain approach, 58, 96–7 revenue foregone approach, 94
Richardson, Maryann, 137 Spain, deinition of benchmark tax in
Rogers, Diane Lim, 118 law, 85
Roin, Julie, 3, 28, 100, 131, 218 Special Feature for the 2003 Edition of
Roth, J., 136–7 Revenue Statistics (OECD), 70
Rothstein, Jesse, 117 Spicer, Michael, 207, 209
Rousseau, Jean-Jacques, 208 Stanley, Robert, 6, 167, 191
Rubin, Ed, 171, 183 state neutrality, 159–60
rule of recognition, 178, 215 Staudt, Nancy, 222, 227
Stead, Merideth M., 18
Sadiq, Kerrie, 101 Steinmo, Sven, 210, 213, 217
Salamon, Lester M., 220 Steuerle, C. Eugene, 19, 106, 119, 175,
Sandford, Cedric, 115, 142 207, 221
Sawyer, A.J., 137 Stewart, Miranda, 20, 69, 94, 96, 130, 171
Scalia, Justice, 181 Stiglitz, Joseph, 7, 107, 161, 214
256 Index
substitutable tax provisions proposed
methodology, 44–5 assessment of TEs, 224–5
substitution benchmark approach, budget deicit cap, 224
85, 90 budget envelope approach, 223–4
examples, 90–1 categorisation as spending or
Sugin, Linda, 121 investment, 218–20
Sunstein, Cass R., 123, 125, 164 collaboration between tax policy
Surrey, Stanley S., 1, 2, 3, 9–10, 15–16, agencies, 229
18–19, 24, 32, 33, 34, 35, 36, deinition of TEs, 206–8
37–42, 52, 60, 64, 65, 69, 84, ‘Pay As You Go’ (PAYGO)
85–6, 110, 112, 121, 130, 138, budgetary rules, 222–3
150, 151, 168, 174, 204, 208, 216, performance analysis of TEs,
218–19, 229 225–9
death of, 42 performance review of TE
Sweden management, 229–30
deinition of benchmark tax in place for ex-ante TE analysis,
law, 85 220–1
outlay equivalence approach, 97 procedural accountability rules,
revenue foregone approach, 94, 97 221–4
Swit, Zhicheng, 160, 204 sunset provisions, 222
way forward
Tanzi, Vito, 146 descriptive accounts of tax
tax aids, 1 expenditure management
tax base, 54–5 practice, 210–11
tax breaks, 1 normative deliberation upon tax
tax compliance management, 135–6 expenditure management,
deinitions of compliance, 136–7 212–17
in terms of underpayment of tax, possibility of, 208–10
138 starting point, 217
evidence of overpayment, 137–8 tax expenditures
fairness, legitimacy and voluntary aggregation of, 58, 128–9
compliance, 139–40 benchmark apolitical ideal tax
integrating of spending rules, system, 8–10
142–4 case for examining in isolation, 7–8,
legitimacy of, 144–6 104–7
tax credits, 20–1 categorisation, purpose of, 24
tax entity, 53–4 coining of term, 2, 18–20, 24–5, 37,
tax expenditure analysis, as objective, 65, 218–19
32–3 concept, 1, 17–18
tax expenditure management elusiveness of, 26–7
framework original intent, 20–2, 59
epistemic tensions, 203–6 controversies, 8–10, 15–17
ideal, 1 irresolvable nature of, 10–11,
objectively right framework as not 66, 67
possible, 152–4 conversion to public-private
objectively right framework as partnership structure, 219–20
possible, 149–52 deinitions, 2, 19, 81–3
political tensions, 206–8 in the eye of the beholder, 205–6
Index 257
OECD, 69–70 of tax expenditures, 130
pragmatic approach to, 215–17 termites, iscal, 146
proposed, 218 thick legal positivism, 178
efectiveness, 203 thin legal positivism, 178
and liberal-democratic political hompson, Dennis, 195–202, 209–10
theory, 11–13 homson, Judith Jarvis, 147
international comparative analysis, huronyi, Victor, 19, 44–5
33–5, 39 tobacco, taxes on, 22
political future, 230–1 Toder, Eric, 122, 126
signiicance, 4–5 Torgler, Benno, 144
economic, 5–7, 107–28 transparency
iscal, 128–30 deinitions, 170, 171
legal, 130–4 purpose of, 170–3
for public administration, 7, Tyler, Tom, 139, 144
134–5
political, 5 United Kingdom
stigmatising consequences of deinition of benchmark tax in
labelling as, 63–4 law, 85
technical eiciency, 130 deinition of tax expenditures, 82
thesis, 13–14, see also identiication; regulatory status of tax expenditures
negative tax expenditures; reporting, 80
non-express tax expenditures reporting structure, 99
Tax Expenditures – A Review of Issues restrictions on reporting, from
and Country Practices resource and data
(OECD), 70 constraints, 93
Tax Expenditures – Recent revenue foregone approach, 94
Developments (OECD), 70 substitution benchmark
Tax Expenditures – Shedding Light on approach, 91
Government Spending hrough United States
the Tax System (World Bank), 73 conceptual benchmark approach, 87
Tax Expenditures (Surrey and deinition of benchmark tax in
McDaniel), 40 law, 85
Tax Expenditures and Estimated deinition of tax expenditures, 82
Report (Canada, 2010), 88 earned income tax credit, 117
Tax Expenditures in OECD Countries ethanol fuel subsidies, 126–7
(OECD), 34–5, 72 functions of tax expenditures
tax incentives, 41 reporting, 78
tax period, 54 regulatory status of tax expenditures
tax preferences, 1 reporting, 80
tax rate, inclusion in benchmark, 54 reporting structure, 98, 99, 100
Tax Reform Act 1969 (US Pub.L. restrictions on reporting, from
91–172), 38 resource and data constraints,
tax reliefs, 1 92–3
tax subsidies, 1, 19, 41, 50, 60 revenue foregone approach, 94
tax-induced structural distortions, 50 revenue gain approach, 97
taxpayers, modifying their behaviour, upside-down efect/inverted
56–7, 115 distributional efect, 55, 57–8,
technical eiciency, 161–2 112
258 Index
Valenduc, Christian, 160, 204 Woellner, Andrew, 26
Ventry, Dennis, 167 World Bank
Villela, Luiz, 20, 69, 84, 85, 94, 99 assessment of sovereign debt,
Vogenbeck, Danielle M., 172, 176, 206, 129–30
207 comparative studies on tax
Voltaire, 209 expenditures, 68
purpose of tax expenditures
Waldron, Jeremy, 178, 179, 180–1, 182 reporting, 78
Weber, Edward P., 139, 190 reporting of tax expenditures, 73–5
Weisbach, David, 3–4, 43, 48–9, 56, 63,
104, 105, 123, 125, 151, 164, 167, Yin, George K., 129, 222
204, 217, 228 Yorio, Edward, 124, 145
well-being indicators, 107–9
‘whole of government’ approach, 3–4, Zelinsky, Edward A., 59, 60, 103, 105,
7–8 125, 130, 131, 132, 142, 146, 151,
Wiedenbeck, Peter J., 204 152, 162, 204, 205, 206, 209–13,
Wildavsky, Aaron, 103, 136, 173 220–1
Witte, A., 136–7 Zelizer, Julian E., 175, 213
Witte, John F., 19, 145, 175 Zolt, Eric M., 167

Tax Expenditure Management A Critical Assessment

Uploaded by

Document Information

Original Title

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

Tax Expenditure Management A Critical Assessment

Uploaded by

Copyright:

Available Formats

TA X E X PE N DI T U R E M A NAGE M E N T

A tax expenditure is a ‘tax break’ allowed to a taxpayer or group of

m a r k b u r t o n is an Associate Professor in Law at the University

k e r r i e s a d i q is a Professor at the Queensland University of

Tax law is a growing area of interest, as it is included as a subdivision in many

© Mark Burton and Kerrie Sadiq 2013

h is publication is in copyright. Subject to statutory exception

First published 2013

Library of Congress Cataloguing in Publication data

ISBN 978-1-107-00736-9 Hardback

Cambridge University Press has no responsibility for the persistence or

Acknowledgements page viii

3 Reporting on tax expenditures 65

6 Managing tax expenditure controversies 203

1.1 he subject of this book

1.2 What are tax expenditures?

Where a taxpayer is entitled to a tax expenditure, the outcome is com-

relaxation of some of Surrey’s tax expenditure criteria.12 Reaching beyond

1.3 Why single out tax expenditures?

which are most efective in achieving a desired outcome.14 Similarly, a

1.4 Explaining the focus upon tax expenditures – the

discrete subject. If there are insuicient resources to critically assess all

1.4.1 he pragmatic politics of tax expenditures in

1.4.2 Economic signiicance

Assuming that the iscal needs of a government remain constant, the

a community’s aspirations to economic democracy.26 To the extent that

1.4.3 Public administration

1.4.4 he uneasy case for examining tax expenditures in isolation

1.5 What else is controversial about tax expenditures?

1.5.1 he tax expenditure benchmark

One means by which eforts might be focused is by separating less

have attempted to construct an apolitical ‘view from nowhere’,37 but their

1.5.2 Tax expenditure measurement

1.6 he irresolvable nature of these tax expenditure controversies

he deinition of tax expenditures, the appropriate benchmark and the

1.7 Tax expenditure controversies and liberal-democratic

Liberals are uncomfortable with democracy because an unrestrained

In Chapter 5 we argue that diferent approaches to the subject of tax

1.8 he thesis of the book

In Chapter 6 we set out what we consider to be core elements of a tax

1.9 Outline of the book

he tax expenditures concept

We aim, as Surrey suggests, to contribute to the necessary rethinking of

explicitly acknowledge that the nature and purpose of tax expenditure

2.2 he concept of tax expenditures

discussion on tax expenditures of the type taken in this book warrants

2.2.1 he phrase ‘tax expenditures’

controversy and an increasing number of governments adopting compre-

language used to describe a group of laws within a tax regime. Within

2.2.2 he original intent of the ‘tax expenditure’ concept

meet criteria of itself raises issues such as gaming by taxpayers, which we

2.2.3 Negative tax expenditures

2.2.4 A useful starting point

now, we turn to the purpose of categorising a provision of the tax regime

2.3 he purpose of categorisation

a fundamental efect on the benchmark it adopts to identify tax expendi-

2.3.1 Framing the question of what is a tax expenditure

increasingly popular amongst developed jurisdictions. he phrase ‘tax

2.3.2 he elusive underlying concept

2.3.3 ‘Good’ tax criteria

2.4 he function of identifying tax expenditures

2.4.1 Reporting on tax expenditures

here is no doubt that the most signiicant advances in tax expenditure