2010 Tax Expenditure Countries Oecd

Tax Expenditures
in OECD Countries
OLICY
TA X P s pe n d
ing X
ment ING TA
S PE N D
n
g ov e r licy Y G
E N T a x po POLIC Y PENDIN
NM G TA X POLIC
t er
ENT S
g
y g ov poli
GOVER nt s p
endin
E N D IN polic G TA X g ta x R N M ve
GOVE
e x in o
g ov e r
n m
T SP nding
ta
NDIN t s pe
n d
olicy
g
ta x p
olicy
O V E R NMEN nmen
t s pe
M E N T S PE g ov e r
nmen
X P OLICY nding
ta x p
IC Y
G r N GT A e
PO L
OVER
v e y s p
OLICY
o l ic
ment
G TA X
g o
TA X P ta x p
olicy
LICY G nding
ta x p
PENDIN policy govern
t s pe
nding
G TA X PO nmen
t s pe
R N M ENT S ta x N T S PENDIN nding
ta x
IN E E
Y GOV
r g e
ER NM
n e p
S PE N D
n m e o v d in ent s
g ov e r y g pe n V G
N T polic
P O L IC ent s
Y G O ver nm N D IN
E a x n m o E
NT SP
g
NM G TA X POLIC
t e r y
nding g ov olic
GOVER n m e n t s pe
S PE N D IN t a x p olicy
ING TA X
s pe n d in g ta x p
OVER N M E
olicy
g ov e r
MENT t s pe
nding
S PE N D ment LICY G
ta x p
IC Y G OVERN g ov e r
nmen
V E R N MENT o licy g
ov e r n
IN G T A X PO
Y GO
p
OL l icy a x
PE N D
TA X P
o t
ta x p nding
nding POLIC t s pe ENT S
nmen
t s pe
N D IN G TA X g ov e r
nmen
G O V ER NM
g o v e r
T SP E o l icy
OLICY
ta x p
NMEN nding TA X P
GOVER r nmen
t s pe
S P E NDING g
g o v e
ENT s pe n d in
ta x p
olicy
Y G O VER NM ov e r n
ment
OLIC y g
olic
TA X P nding
ta x p
t s pe
nmen
g ov e r
ov
icy g
x pol
end ing ta Y
IC
rnme
nt s p X POL
g ov e ING TA
ta x p
olicy
ENT S PE N D g ta x
din
n
nding M t s pe
er n ment
s pe
IC Y G OVERN g ov e r
nmen
P ENDIN
G
ENT S
y
y g ov X POL po l ic
polic T A g ta x R N M ve
ta x G in E o
PENDIN GOV
e n d y g
t sp olic
N M ENT S g ov e r
nmen
A X P OLICY nding
ta x p
L IC Y
OVE R ta x p
o l ic y
ING T nmen
t s p e
XP O
LICY G nding S PE N D g ov e r ING TA
IN G T A X PO r nmen
t s pe
V E R N MENT g ta x p
olicy
E N T S PE N D e nding
ta x
P E N D y g o v e
Y G O pe n d in
V E R N M en t s p
EN
NT SP
m
TS p o l ic
POLIC n t s
Y GO e r n
NMEN g ta x
G TA X g ov e r
nme
POLIC olicy
g ov
R NME
GOVER nt s p
endin
N D IN p olicy
G T A X g ta x p G O V E en
T S PE OLICY
e x in m
NDIN
n m t a s pe n d n
g ov e r nding ov e r
ta x p
olicy
O V E R NMEN nmen
t s pe
M E N T S PE ov e r n
ment
IN G TA X P ta x po
licy g
L IC Y
G v e r
VER N y g
S PE N D g
X PO
OLICY y g o pol ic pe n d in
TA X P g ta x
polic
P O L ICY GO pendin
g ta x
E R N MENT ver nm
ent s
P E N D ING TA x po
t sp e n d in
G TA X er nm
e n t s
Y GO V polic
y g o
ENT S pendin
g ta
g ov e r
nmen
PENDIN y g ov
G TA X
POLIC g ta x
Y GOV
ER NM ent s
G TA X
P
ENT S g ta x
polic
ent s
pendin
IC y g ov
er nm
GOVE R N M
t sp e n d in
T S P E N D IN
g ov e r n m
T A X P O L tax p o l ic
T S P E DIN
N g
MEN NDING MEN
m e n ic y in g icy
n pol pe n d pol
ta x p
olicy
g ov e r
IC Y G OVERN s pendin
g ta x
M E N T S PE v er nm
ent s
IC Y G OVERN pendin
g ta x
IC Y
OL t N g o
O L t s
OL
GOVER
n n
TA X P TA X P TA X P
nm e olic y nme
g ov e r ta x p g ov e r
Tax Expenditures
in OECD Countries
ORGANISATION FOR ECONOMIC CO-OPERATION
AND DEVELOPMENT
The OECD is a unique forum where the governments of 30 democracies work together to
address the economic, social and environmental challenges of globalisation. The OECD is also at
the forefront of efforts to understand and to help governments respond to new developments
and concerns, such as corporate governance, the information economy and the challenges of an
ageing population. The Organisation provides a setting where governments can compare policy
experiences, seek answers to common problems, identify good practice and work to co-ordinate
domestic and international policies.
The OECD member countries are: Australia, Austria, Belgium, Canada, the Czech Republic,
Denmark, Finland, France, Germany, Greece, Hungary, Iceland, Ireland, Italy, Japan, Korea,
Luxembourg, Mexico, the Netherlands, New Zealand, Norway, Poland, Portugal, the Slovak Republic,
Spain, Sweden, Switzerland, Turkey, the United Kingdom and the United States. The Commission of
the European Communities takes part in the work of the OECD.
OECD Publishing disseminates widely the results of the Organisation’s statistics gathering
and research on economic, social and environmental issues, as well as the conventions,
guidelines and standards agreed by its members.
This work is published on the responsibility of the Secretary-General of the OECD. The opinions
expressed and arguments employed herein do not necessarily reflect the official views of the
Organisation or of the governments of its member countries.
ISBN 978-92-64-07689-1 (print)

ISBN 978-92-64-07690-7 (PDF)
Also available in French: Les dépenses fiscales dans les pays de l'OCDE
Corrigenda to OECD publications may be found on line at: www.oecd.org/publishing/corrigenda.

© OECD 2010
You can copy, download or print OECD content for your own use, and you can include excerpts from OECD publications,
databases and multimedia products in your own documents, presentations, blogs, websites and teaching materials, provided
that suitable acknowledgment of OECD as source and copyright owner is given. All requests for public or commercial use and
translation rights should be submitted to [email protected]. Requests for permission to photocopy portions of this material for
public or commercial use shall be addressed directly to the Copyright Clearance Center (CCC) at [email protected] or the Centre
français d’exploitation du droit de copie (CFC) at [email protected].
FOREWORD – 3
Foreword
In all OECD member countries, governments collect revenues through

taxes and redistribute this public money, often by obligatory spending on
social programmes such as education or health care. Their tax systems
usually include “tax expenditures” – provisions that allow certain groups of
people, such as small businessmen, retired people or working mothers, or
those who have undertaken certain activities, such as charitable donations, to
pay less in taxes.
The use of tax expenditures by governments is pervasive and growing.
At a time when many government budgets are threatened by population
ageing and adverse cyclical developments, there is a pressing need to avoid
inefficient government programmes, some of which may utilise tax
expenditures.
This book sheds light on the use of tax expenditures, mainly through a
study of ten OECD countries: Canada, France, Germany, Japan, Korea, the
Netherlands, Spain, Sweden, the United Kingdom and the United States.
This book will help government officials and the public better understand
some of the technical and policy issues behind the use of tax expenditures. It
highlights key trends and successful practices, and addresses a broad range
of government finance issues, including tax policy making, tax and budget
efficiency, fiscal responsibility and rule making.
The book is the result of a project led by the Budgeting and Public
Expenditures Division (BUD) of the OECD Public Governance and
Territorial Development Directorate (GOV), under the auspices of the
OECD Working Party of Senior Budget Officials. The project was co-
ordinated by Barry Anderson, Head of Division (GOV/BUD). The author of
the report is Joseph J. Minarik, a consultant to the OECD who works for the
Committee for Economic Development, an NGO located in
Washington DC. Stephen Matthews and Jens Lundsgaard of the OECD
Centre for Tax Policy and Administration (CTP) and Chris Heady, formerly
of CTP, provided valuable input for the report.
TAX EXPENDITURES IN OECD COUNTRIES © OECD 2010

4 – FOREWORD
The book has benefited from meetings and seminars organised in 2008
and 2009 by both the Working Party of Senior Budget Officials and
Working Party No. 2 on Tax Policy Analysis and Tax Statistics. It includes
results from a questionnaire that was sent to a selection of OECD member
countries. The author is grateful for the participation and discussion at
meetings and for the responses to the questionnaire. Any misinterpretations
from these sources of information are the responsibility of the author.
The OECD Working Party of Senior Budget Officials aims to improve
the effectiveness and efficiency of resource allocation and management in
the public sector. Every year the Working Party organises a number of
meetings on topics of interest to budget officials. Some are organised on a
regular basis – for example, the meetings of the network on financial
management (accrual accounting) and the network on performance and
results. In addition to those meetings, other topics are discussed on an
ad hoc basis, as requested by the Working Party. Such is the case for this
project on tax expenditures.

TABLE OF CONTENTS – 5
Table of Contents
Part I: A look at tax expenditures .....................................................................9
Chapter 1 Introduction .....................................................................................11

What are tax expenditures? .............................................................................12
What are the different types of tax expenditures? ...........................................12
How are tax expenditures measured? ..............................................................13
Trends in tax expenditures ..............................................................................14
Defining tax expenditures ...............................................................................15
Tax expenditure controversy ...........................................................................17
Notes ...............................................................................................................20
Bibliography....................................................................................................21
Chapter 2 Policy background and practices....................................................23
Policy background...........................................................................................24
Tax expenditures and policy-making practices ...............................................43
Notes ...............................................................................................................52
Bibliography....................................................................................................55
Chapter 3 The role of tax expenditures in the budget process ......................59
Types of budget rules ......................................................................................63
Notes …………………………………………………………………………66
Bibliography …………………………………………………………………67
Chapter 4 Country profiles: Methods, institutions and data .........................69
Notes on cross-country data comparisons .......................................................70
Tax expenditures in Canada ............................................................................76
Tax expenditures in France .............................................................................84
Tax expenditures in Germany .........................................................................88
Tax expenditures in Japan ...............................................................................93
Tax expenditures in Korea ..............................................................................99
Tax expenditures in the Netherlands .............................................................105
Tax expenditures in Spain .............................................................................111
Tax expenditures in Sweden .........................................................................120
Tax expenditures in the United Kingdom .....................................................125
Tax expenditures in the United States ...........................................................132
Notes .............................................................................................................141
Bibliography..................................................................................................144

6 – TABLE OF CONTENTS
Chapter 5 Conclusions.....................................................................................147
Definition and measurement .........................................................................148
Reporting .......................................................................................................154
Policy making................................................................................................155
Policy review.................................................................................................157
“Make work pay” tax expenditures ...............................................................158
Number of tax expenditures ..........................................................................159
Amount of tax expenditures ..........................................................................161
Conclusions ...................................................................................................165
Notes .............................................................................................................166
Bibliography..................................................................................................167
Part II: Comparing tax expenditures in OECD countries ..........................169
Explanatory key ............................................................................................171
Table II.1. Tax expenditures in Canada (% of GDP).......................................172

Table II.2. Tax expenditures in Canada (% of central government total tax
and non-tax receipts)................................................................... 176
Table II.3. Tax expenditures in Canada (% of relevant tax revenue).............. 178
Table II.4. Number of tax expenditures in Canada (% of GDP)..................... 180
Table II.5. Tax expenditures in Germany (% of GDP)....................................182
Table II.6. Tax expenditures in Germany (% of central government total
tax revenue)................................................................................. 184
Table II.7. Tax expenditures in Germany (% of relevant tax revenue)........... 186
Table II.8. Number of tax expenditures in Germany (% of GDP)...................187
Table II.9. Tax expenditures in Korea (% of GDP)......................................... 188
Table II.10. Tax expenditures in Korea (% of central government total tax
and non-tax receipts).................................................................. 190
Table II.11. Tax expenditures in Korea (% of relevant tax revenue)............... 192
Table II.12. Number of tax expenditures in Korea (% of GDP)...................... 193
Table II.13. Tax expenditures in the Netherlands (% of GDP)........................ 194
Table II.14. Tax expenditures in the Netherlands (% of central government
total tax and non-tax receipts).................................................... 196
Table II.15. Tax expenditures in the Netherlands (% of relevant tax
revenue)……………………………………………………….. 197
Table II.16. Number of tax expenditures in the Netherlands (% of GDP)....... 198
Table II.17. Tax expenditures in Spain (% of GDP)........................................ 199
Table II.18. Tax expenditures in Spain (% of central government total tax
and non-tax receipts).................................................................. 200
Table II.19. Tax expenditures in Spain (% of relevant tax revenue)................ 201
Table II.20. Number of tax expenditures in Spain (% of GDP)....................... 202
Table II.21. Tax expenditures in the United Kingdom (% of GDP)................ 203

TABLE OF CONTENTS – 7
Table II.22. Tax expenditures in the United Kingdom (% of central

government total tax and non-tax receipts)................................ 205
Table II.23. Tax expenditures in the United Kingdom (% of relevant tax
revenue)...................................................................................... 207
Table II.24. Number of tax expenditures in the United Kingdom (% of
GDP)…………………………………………………………... 208
Table II.25. Tax expenditures in the United States (% of
GDP)........................................................................................... 210
Table II.26. Tax expenditures in the United States (% of central
Table II.27. Tax expenditures in the United States (% of relevant tax
revenue)...................................................................................... 218
Table II.28. Number of tax expenditures in the United States (% of GDP)..... 222
Table II.29. International comparison of tax expenditures (% of GDP).......... 224
Table II.30. International comparison of tax expenditures (% of central
Table II.31. International comparison of tax expenditures (% of relevant
tax revenue)................................................................................ 226
Table II.32. International comparison of number of tax expenditures (% of
GDP) .......................................................................................... 228
Figure II.1. Income tax expenditure by purpose in Canada.............................. 229

Figure II.2. Income tax expenditure by purpose in Germany........................... 229
Figure II.3. Income tax expenditure by purpose in Korea................................ 230
Figure II.4. Income tax expenditure by purpose in the Netherlands................. 230
Figure II.5. Income tax expenditure by purpose in Spain................................. 231
Figure II.6. Income tax expenditure by purpose in the United Kingdom......... 231
Figure II.7. Income tax expenditure by purpose in the United States............... 232
Figure II.8. Number of tax expenditures........................................................... 233
Figure II.9. Number of income tax expenditures.............................................. 233
Figure II.10. Income tax expenditures (% of GDP).......................................... 234
Figure II.11. Income tax expenditures (% of income tax revenue)................... 234
Figure II.12. Income tax expenditures (% of GDP).......................................... 235
Figure II.13. All tax expenditures (% of GDP)................................................. 235
Figure II.14. All tax expenditures (% of total tax revenue).............................. 236
Figure II.15. Canada’s “memorandum items”................................................... 236
Figure II.16. Cost of ten largest tax expenditures............................................. 237
Figure II.17. Intensity of use of tax expenditures............................................. 237
Data sources ………………………………………………………………. 238

8 – TABLE OF CONTENTS
This book has...
StatLinks2
A service that delivers Excel® ﬁles
from the printed page!
Look for the StatLinks at the bottom right-hand corner of the tables or graphs in this book.
To download the matching Excel® spreadsheet, just type the link into your Internet browser,
starting with the http://dx.doi.org prefix.
If you’re reading the PDF e-book edition, and your PC is connected to the Internet, simply
click on the link. You’ll find StatLinks appearing in more OECD books.

PART I: A LOOK AT TAX EXPENDITURES – 9
Part I
A look at tax expenditures

1. INTRODUCTION – 11
Chapter 1
Introduction
This chapter gives a brief introduction and history of tax expenditures. It begins by
attempting to define tax expenditures then proceeds to discuss the different types of tax
expenditures. There is a short discussion on the different ways to measure them. It then
gives several concrete examples of tax expenditures in different countries. It concludes
by discussing some of the controversy concerning tax expenditures.

12 – 1. INTRODUCTION
What are tax expenditures?
Tax expenditures are “provisions of tax law, regulation or practices that

reduce or postpone revenue for a comparatively narrow population of
taxpayers relative to a benchmark tax” (Anderson, 2008). For government, a
tax expenditure is a loss in revenue; for a taxpayer, it is a reduction in tax
liability. Tax expenditures are better known in many OECD countries as tax
reliefs, tax subsidies and tax aids (Schick, 2007).
In practice, defining tax expenditures is difficult because “some tax
measures may not be readily classified as part of the benchmark or an
exception to it” (Whitehouse, 1999). The problem begins with defining the
“basic tax structure”. Most experts would agree that structural elements of a
tax system should not be recorded as tax expenditures, while
“programmatic” features should be.
According to Kraan (2004), the “benchmark tax includes: the rate
structure, accounting conventions, the deductibility of compulsory
payments, provisions to facilitate administration, and provisions relating to
international fiscal obligations”.
Since tax expenditures are not actual outlays, the amounts “spent” are
notional; that is, they are based on assumptions and estimates as to how
taxpayers would behave under particular conditions.
What are the different types of tax expenditures?
Tax expenditures may take a number of different forms:

• allowances: amounts deducted from the benchmark to arrive at the tax
base;
• exemptions: amounts excluded from the tax base;
• rate relief: a reduced rate of tax applied to a class of taxpayer or taxable
transactions;
• tax deferral: a delay in paying tax;
• credits: amounts deducted from tax liability (Anderson, 2008).

Box 1.1. Examples of tax expenditures
• Professional expenses: meals and entertainment expenses, commuting expenses, etc.;

• Interest deduction (housing): tax credit for repayment of mortgage loans and a special
deduction for interest;
• Interest on saving accounts (up to a certain ceiling);

• Corporate investments;
• Tax assistance for childcare expenses;
• Reduced tax rate for small and medium-sized enterprises (SMEs);
• Pension income tax credit;
• Charitable donations tax credit;
• Deductions for energy saving measures (alternative energy, etc.);
• Employer funded health benefits.
How are tax expenditures measured?
Tax expenditures are calculated using the “revenue forgone method

which calculates the tax that would have been payable if the tax concession
were removed, and economic behaviour remained unchanged” (Whitehouse,
1999). As Anderson explains (2008), there are alternative ways to measure
tax expenditures:
• Initial revenue loss (gain): the amount by which tax revenue is reduced
(increased) as a consequence of the introduction (abolition) of a tax
expenditure, based upon the assumption of unchanged behaviour and
unchanged revenues from other taxes.
• Final revenue loss (gain): the amount by which tax revenue is reduced
(increased) as a consequence of the introduction (abolition) of a tax
expenditure, taking into account the change in behaviour and the effects
on revenues from other taxes as a consequence of the introduction
(abolition).

• Outlay equivalence: the direct expenditure that would be required in

pre-tax terms to achieve the same after-tax effect on taxpayers’ incomes
as the tax expenditure if the direct expenditure is accorded the tax
treatment appropriate to that type of subsidy or transfer in the hands of
the recipient.
Trends in tax expenditures
Tax expenditures – defined as “a transfer of public resources that is

achieved by reducing tax obligations with respect to a benchmark tax, rather
than by a direct expenditure” (Kraan, 2004) – have been a serious concern of
budget and tax analysts for almost half a century.1 The concern is that tax
expenditures may have ill effects on both budget and tax policy, and that
both political and policy-making considerations may make tax expenditures
easier to enact, and less likely to undergo rigorous review and repeal, than
equivalent but more straightforward spending programmes. At the same
time, tax expenditures are a part of the tax systems of every developed
country around the world. Particular tax expenditures are defended as sound
tax policy instruments, and there is no visible, serious proposal that tax
expenditures be eradicated anywhere. In the interests of both tax and fiscal
policy, tax expenditures would seem to be a fitting topic of inquiry today.
Though the concept of tax expenditures was first identified and analysed
in the United States, the concern about the issue now extends across
countries. Accounting in many countries suggests that the use of tax
expenditures is pervasive and growing (Polockova Brixi, Valenduc, and
Swift, 2004). At any time, the possibility that a back channel for resource
allocation could lead to inefficient government “spending” would be
troubling. When many government budgets are threatened by population
ageing and adverse cyclical developments, the concern is only greater.
Accordingly, the Organisation for Economic Co-operation and
Development (OECD) has decided to devote its attention to this issue, along
with associated and similar budgetary questions. The Working Party of
Senior Budget Officials discussed a report on Off-budget and Tax
Expenditures at its 2004 meeting in Madrid (Kraan, 2004), following
previous work on tax expenditures by the OECD Centre for Tax Policy and
Administration (OECD, 1984; OECD, 1996; OECD, 2003). The OECD Best
Practices for Budget Transparency (OECD, 2002) contain some basic
guidelines for the treatment of tax expenditures. The World Bank has
evidenced similar concern in Tax Expenditures – Shedding Light on
Government Spending through the Tax System (Polockova Brixi, Valenduc,
Swift, 2004). Such concern and attention has contributed to some improved

and extended procedures of tax expenditure reporting, review and control by

OECD member countries (Koiwa, 2006). Still, there is considerable room
for improvement. There is a perceived need 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.
An important and timely associated issue is that some OECD member
countries have enacted, or are considering, fiscal rules that make use of
expenditure ceilings. The handling of tax expenditures under such rules is
critical because a systematically lesser degree of budgetary control on tax
expenditures, as opposed to spending narrowly defined, could direct
increasing flows of what would – and often should – be “spending” through
the tax systems of the affected countries.
All of these considerations suggest that continued and even greater
attention to the use of tax expenditures would be timely and worthwhile.
This report will address the issue from several directions. A key part will be
a survey of the level, and change, of the number and revenue effect of tax
expenditures across several OECD member countries. An analysis of these
data will suggest the underlying forces that have led to the prevalence of tax
expenditures, as well as the tax, efficiency, and fiscal implications of these
trends.
Further discussion will identify successful practices regarding the
reporting of tax expenditures. Questions regarding the review (such as it is)
of tax expenditures in the policy process will be explored, including some
ideas that have been proposed but not implemented. Finally, additional
analysis will put these successful practices into the particular context of
budget rules, especially spending-based rules. In combination, these
discussions should address a broad range of issues of government finance,
from policy making to tax and budget efficiency, and on to fiscal
responsibility and rulemaking.
Defining tax expenditures
Identification of any particular tax provision as a tax expenditure

requires more than a broad and general definition. Different countries have
identified different specific criteria. In 1987, a working group in the
Netherlands tasked with this mission, compared practice in other countries,
identified five criteria, and in the end rejected three and accepted the other
two. In their particular instance, the group rejected the pursuit of a non-fiscal
policy goal, convertibility of the provision into a direct expenditure, and the
benefit of a limited group of taxpayers, even though those criteria were used
elsewhere. They retained for future analysis the reduction of revenue and the

deviation from a benchmark tax structure (van den Ende, Haberham, and
den Boogert, 2004). One might conclude that there is significant diversity in
working definitions of tax expenditures across countries, but that a frequent
common element is some notion of departure from a tax system benchmark.
In practice, some of the other criteria – particularly the loss of revenue, the
convertibility into a spending programme, and the limited group of
beneficiaries – might be thought to be objective to some degree. In contrast,
the conception of a benchmark tax system might provide the greatest degree
of room for difference of judgment.
In fact, conceptions of the benchmark tax differ from analyst to analyst
and country to country. The World Bank compendium cited above says that
the benchmark or “norm includes the rate structure, accounting conventions,
deductibility of compulsory payments, provisions to facilitate tax
administration, and international fiscal obligations” (Swift, Polockova Brixi,
and Valenduc, 2004), which echoes earlier OECD work (Kraan, 2004).
However, each of these items provides considerable judgmental leeway, and
when examining country practice, each application is in some way unique.
Canada’s benchmark is articulated to a considerable degree of detail:
“the benchmark for the personal and corporate income tax systems includes
the existing tax rates and brackets, the unit of taxation, the time frame of
taxation, the treatment of inflation in calculating income, and those
measures designed to reduce or eliminate double taxation [of corporate
profits]” (Seguin and Gurr, 2004). Particular decisions such as the choice of
the individual rather than the family as the unit of taxation, and the inclusion
of Canada’s particular method of relief for double taxation of dividends,
lead to differences in the identification of tax expenditures relative to other
countries. In contrast to this specificity, Japan and Korea do not yet identify
any specific benchmark tax system, rather identifying tax expenditures (or in
the case of Japan, what are called “special tax measures”) by reference to
deviation from principles which are not so explicitly articulated. Other
countries state their own methods with varying degrees of specificity, and
with unique choices of policy standards.
Because the choice of a benchmark or other measurement yardstick
varies substantially from country to country, identifications of tax
expenditures in any given country can be quite different from those in other
countries. Polackova Brixi, Valenduc, and Swift (2004) believe that the
differences in benchmarks are so severe that they choose not to provide
comparative data in their cross-country survey.

Kraan (2004) says that the problem of disagreement in the choice of a

benchmark tax…
…is rooted in different views of the normative tax base. The
normative tax base is the monetary sum in the hands of private
households to which the tax ought to be applied, for instance: income,
value added, profit, sales… [T]he definition of the normative tax base is
a very political exercise. For this reason, attempts in the past to define
tax expenditures in terms of the normative tax base…have not been very
successful. They have led to neither international nor domestic
agreement about the concept of tax expenditure. Thus an alternative
definition of a tax expenditure abstracts from the normative tax base.
The definition uses rather the more neutral yardstick of the “benchmark
tax”. Tax expenditures in this sense are deviations from the benchmark
tax. The benchmark has no normative significance. Deviations from it in
order to arrive at the normative tax base may be perfectly appropriate.
Tax expenditures may thus also be appropriate.2
Kraan thus defines more specifically the term “benchmark” to provide
that it has no normative content. Presumably, if closer agreement on the
nature of the benchmark did not require equivalent agreement on what the
correct or best tax system is, different observers could come closer to
common ground, but such progress is not yet forthcoming.
Tax expenditure controversy
The concept of tax expenditures has been controversial since its

inception, with much of that criticism following upon Kraan’s concern about
the choice of the benchmark tax. A report by the US Joint Committee on
Taxation (2008) summarises and revises the criticisms, and offers an
alternative framework along the general lines that Kraan has proposed in an
attempt to develop agreement on the usefulness of the concept (Joint
Committee on Taxation, 2008).3 The Joint Committee approach is too new
to evaluate, but the criticisms of current methodology that they glean from
the literature should be understood in considering this report. To paraphrase
some of the key points, in particular:
• Some US critics believe that the normal tax system was not developed
from first principles with sufficient rigour to serve as such a standard,
resulting in errors in the identification of tax expenditures.4 Some would
suggest, along the lines of Kraan, that the differences in values among
analysts are so strong that consensus on the nature of the benchmark
would be impossible to achieve (Burman, 2003).

• Some critics see the normal tax system as a hidden agenda or target for a
particular brand of tax “reform,” such that, for example, an income tax
benchmark would be a roadblock to the development of a consumption-
based tax (Bartlett, 2001).
• The tax expenditure concept’s recent focus on tax policy issues can be
seen as an abdication of its original self-avowed motivation to compare
tax provisions to spending programmes with similar objectives (Shaviro,
2004).
• Yet another line of argument is that the concept of tax expenditures
implies a sense of “exceptionalism” for tax policy – that is, a conviction
that tax policy should remain surgically clean and efficient, while all
messy political compromises go back to the spending side of the budget
where they belong (Logue, 2000).
Relative to these criticisms, the intent of the current report is quite
pragmatic. The aspirational goal is better policy. To this end, differences
between various countries’ tax expenditure methodologies in general, and
their benchmark tax systems in particular, should not prohibit analysis.
Although such differences may prevent a cardinal ranking of the various tax
systems according to the criterion of tax expenditure avoidance, such a
ranking would serve little useful purpose. Because as Kraan suggests, there
is not and should not be any presumption that all tax expenditures are bad,
the counts of tax expenditures in different countries cannot be a measure of
the relative merits of their tax systems.
In this report, 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, or as the actual tax system for the
country in question or for any other country. Rather, the motivation is that,
given the possible policy problems that could be caused by tax expenditures
as described below, a provision so identified in any given country bears
examination, which may or may not suggest its modification or repeal. The
budget and tax policy processes that yield more or fewer identified tax
expenditures might bear consideration as well. For that matter, some tax
expenditure measurement systems might, upon discussion, seem more
conducive to this kind of analysis, and thus be worthy of consideration.
With respect to the criticisms expressed above, other than those that
relate to the choice of the benchmark, there is no intended implication that
tax policy making should be devoid of politics, and that spending policy
making should be mired in it. With budgetary resources scarce, all
government allocation decisions should be as efficient as possible. There is a
presumption that tax expenditures with valid policy objectives should be

compared with possible spending policies that would achieve the same
objective. Tax expenditures that fit the negative profile of the policy type –
those that benefit small and less worthy groups, are non-transparent, etc. –
should be considered for repeal, reduction, or replacement by better
targeted, more open spending policies. Realistically, those decisions will be
made in a political environment that may not be friendly for what could be
characterised as a tax increase to finance a larger government with no
change in policy mission – which is indicative of why the earliest
contributors to this field believed that tax expenditures needed special
attention in the first place.

Notes
1. See Surrey (1973), based on work done by the author in government

service in the late 1960s, may have been the earliest full explication of
this concept.
2. See page 131.
3. The new framework proposed bears some resemblance to the
methodologies used in Japan and Korea, in that rather than identifying a
“normal tax system,” the JCT would choose exceptions to “congressional
intent” as revealed in part in the tax law itself. The Joint Committee paper
notes, however, that the list of provisions that would result, some of
which would be called “tax subsidies” and others “tax-induced structural
distortions” rather than “tax expenditures”, would contain virtually the
same items; more than anything else, the derivation would change, in the
hopes of the authors to a more defensible process.
4. An early articulation of this argument was Boris I. Bittker (1969).

Bibliography
Anderson, Barry (2008), powerpoint presentation at the Asian Senior

Budget Officials meeting, 10-11 January 2008, Bangkok, Thailand,
www.oecd.org/dataoecd/40/6/39944419.pdf.
Bartlett, Bruce (2001), “The End of Tax Expenditures as We Know
Them?”, Tax Notes, Vol. 92, No. 3.
Bittker, Boris I. (1969), “Accounting for Federal ‘Tax Subsidies’ in the
National Budget,” National Tax Journal, Vol. 22, No. 2.
Burman, Leonard E. (2003), “Is the Tax Expenditure Concept Still
Relevant?”, National Tax Journal, Vol. 56, No. 3.
Ende, Leo van den, Amir Haberham, and Kees den Boogert (2004), “Tax
Expenditures in the Netherlands,” in Polockova Brixi, Valenduc, and
Swift (2004), Tax Expenditures – Shedding Light on Government
Spending through the Tax System, The World Bank, Washington DC,
pp. 134-135.
Joint Committee on Taxation (2008), A Reconsideration of Tax Expenditure
Analysis (JCX-37-08), United States Congress, Washington DC.
Koiwa, Tetsura (2006), “Recent Issues on Tax Expenditures in OECD
Countries”, OECD unpublished paper.
Kraan, Dirk-Jan (2004), “Off-budget and Tax Expenditures,” OECD
Journal on Budgeting, vol. 4, no. 1, OECD, Paris, pp. 121-42.
Logue, Kyle (2000),“If Taxpayers Can’t be Fooled, Maybe Congress Can: A
Public Choice Perspective on the Tax Transition Debate,” University of
Chicago Law Review, Vol. 67.
OECD (1984), Tax Expenditures: A Review of Issues and Country Practices,
OECD, Paris.
OECD (1996), Tax Expenditures: Recent Experiences, OECD, Paris.
OECD (2003), Special Feature for the 2003 Edition of Revenue Statistics:
Note by the Secretariat, OECD, Paris.

Polockova Brixi, Hana, Christian M.A. Valenduc, and Zhicheng Li Swift

(2004), Tax Expenditures – Shedding Light on Government Spending
through the Tax System, The World Bank, Washington DC, pp. 1-3.
Schick, Allen (2007), “Off-budget Expenditure: An Economic and Political
Framework”, OECD Journal on Budgeting, Vol. 7, No. 3, OECD, Paris.
Seguin, Marc, and Simon Burr (2004), “Federal Tax Expenditures in
Canada,” in Polockova Brixi, Valenduc, and Swift (2004), Tax
Expenditures – Shedding Light on Government Spending through the Tax
System, The World Bank, Washington DC, p. 99.
Shaviro, Daniel (2004), “Rethinking Tax Expenditures and Fiscal
Language,” Tax Law Review, Vol. 57, No. 2.
Surrey, Stanley S. (1973), Pathways to Tax Reform: The Concept of Tax
Expenditures, Harvard University Press, Cambridge, Massachusetts,
United States.
Swift, Zhicheng Li, Hana Polockova Brixi, and Christian M.A. Valenduc
(2004), “Tax Expenditures: General Concept, Measurement, and
Overview of Country Practices,” in Polockova Brixi, Valenduc, and
p. 3.
Whitehouse, Edward (1999), “The Tax Treatment of Funded Pensions”,
Social Protection Discussion Paper Series, The World Bank,
Washington DC.

2. POLICY BACKGROUND AND PRACTICES – 23
Chapter 2
Policy background and practices
This chapter explains why tax expenditures are adopted and when they might work
well. It then discusses the different theoretical allegations of negative effects of tax
expenditures. Next, it explains the multiplication and growth of tax expenditures. It
continues by discussing the special case of “make work pay” tax expenditures. Finally,
it discusses policy making processes involved in implementing tax expenditures, such as
reporting, review and oversight, and legislative process and enactment.

24 – 2. POLICY BACKGROUND AND PRACTICES
Policy background
The very earliest analyses carried a strong sense that tax expenditures
were an inferior form of budget policy. Tax expenditures were said to be
unfair, distortionary and costly, but also to be prone to rapid growth in both
number and size, and resistant to eradication: in effect, to be a non-native
plant in the garden of government-programme alternatives. And yet, tax
expenditures remain a feature of all tax systems, and many are widely
believed to be effective and efficient as well as politically unassailable. Tax
expenditures must be considered realistically relative to alternative policy
tools – spending programmes and perhaps regulation – which have their
own process deficiencies in enactment and review, and introduce their own
economic and political distortions. Logically, and for purposes of
discussion, it is worth separating the ill effects of existing tax expenditures
from their tendency to multiply and grow regardless of those flaws and
failings. But first, it is important to understand why tax expenditures remain
a fixture of tax systems worldwide.
Why are tax expenditures adopted and when might they work well?
Tax expenditures are enacted because there are perceived legitimate
reasons for their use. Tax expenditures have a role to play; they are
employed widely, and there are few, if any, suggestions that all tax
expenditures should be repealed.1 Assuming in the first instance that there
are valid reasons for government involvement (such as market failures or
merit goods), there are conditions under which tax expenditures are most
likely to be successful, or even the best, policy tools to achieve their
objectives.
Administrative economies of scale and scope

The pursuit of some public objectives might be administratively costly
through conventional government spending programmes. Because tax
expenditures usually deliver their reward through a reduction of tax that
would be paid in any event, government spending agencies need not engage
in the administrative effort to manage a programme and deliver payments. In
instances where relevant information is already reported by the taxpayer
through the tax system, that information can be used at lower marginal
administrative cost than through duplicated reporting to a spending agency.
Alternatively, where the tax expenditure involves not reporting some forms
of income for tax purposes, there are consequent economies of
administration.

Limited probability of abuse or fraud

Spending programme grants to individuals typically involve prior
reporting by individuals and verification by a spending agency before the
grant is paid. Where detailed verification is not necessary, a tax benefit that
is paid solely on the ground of the taxpayer’s filing can be cost-effective.
Such situations may obtain where tax-preferred income or expense is
delivered or paid by an employer, such as in employer benefits for
insurance, child care, or education, or employer withholding of tax paid to
provincial or local levels of government. The employer may report those
payments to the tax authorities along with the information on the
employee’s taxable wages. The availability of employer information serves
as a check on the reporting of the employee, much like the parallel reporting
of expenditures by purchaser and seller under a value-added tax. In other
instances, ready availability of verifying data from a separate entity, such as
for interest or retirement payments to a financial institution, can effectively
deter false reporting without prior verification by a spending agency.
A proper wide range of taxpayer choice

In subsidies for purposes such as retirement saving or health care, there
may be wide ranges of private preferences. In those and other instances, the
distinctions among different activities that properly qualify for governmental
support may not be considered important. The involvement of a spending
agency in such choices might be considered inappropriate or unnecessary,
and a simpler reporting and verification process through the tax system
might be thought to be more efficient.
Measurement of taxpaying capacity

Deductions or exclusions from income can be justified as proper
measurements of the ability to pay tax, or as essential to measure income
accurately. Under many applications of the tax expenditure concept,
however, such deductions or exclusions would be considered structural
features of the tax system rather than tax expenditures.
Theoretical allegations: When and why are tax expenditures bad?

It is alleged that many tax expenditures are not justified by the
administrative advantages above, and have proved to be less than optimal
tools for their designated objectives. One categorisation of the long-alleged
flaws of tax expenditures is the equally long-lived taxonomy of the
objectives of tax policy: fairness, efficiency and effectiveness, simplicity,
and fiscal responsibility.

Fairness
It is alleged that the tax expenditure tool tends toward unfair results,
both in the likelihood that undeserving groups of taxpayers obtain them, and
in the operation of the tax expenditures once they have been enacted.
Selective and lucrative tax expenditures are most usually those that
provide advantages for income from capital2 or for self-employment, rather
than labour. Those who own wealth, including businesses, tend to be those
with higher incomes.3 Upper-income taxpayers are more likely to take
advantage of tax benefits for retirement saving and housing that are
available in many countries.
There is the superficially reasonable argument that the well-to-do are the
most capable of influencing the legislative process, though it is difficult to
judge whether this bias would be more likely to affect the enactment of tax
provisions than spending programmes. Reinforcing this point, there is the
recognition that tax expenditures can be established by practice or in
regulation, as well as by law. It would not be surprising if those most
equipped to look beyond the relatively (though usually not absolutely)
straightforward tax instructions into the detail of regulations and practice
would be those with the most resources.
Once tax expenditures are in place, they are likely to benefit well-off
taxpayers more than the rank and file. The well-to-do simply have more tax
liability in the first instance, and so have more to gain from tax
expenditures. To the extent that tax expenditures are complex and confront
those who wish to claim the tax benefits with complex tasks, the most well-
to-do are most likely to have the financial and technical knowledge, or the
hired assistance, to take advantage of those opportunities.
Also, under a progressive system, any tax expenditure that reduces
taxable income or postpones the recognition of taxable income, will most
benefit those taxpayers who are in the highest tax-rate brackets. Those who
are not taxable do not benefit at all from tax expenditures structured in that
way.4 This effect has been labelled an “upside-down subsidy,” and has been
considered a disadvantage of tax expenditures as a policy tool (Surrey and
McDaniel, 1980; Gravelle, 2005). This upside-down effect can be defeated
by using non-wastable tax credits – that is, tax credits in amounts that are
fixed regardless of income, and that are payable in full to taxpayers even if
the credits exceed the amount of tax liability – at the cost of additional
complexity. For some purposes, this issue can be extremely important, and it
will be close to the heart of later discussions of tax expenditures designed to
“make work pay”.

Empirical estimation of the distributional effects of tax expenditures is a

large and complex undertaking. The quantitative estimates of the revenue
effects of tax expenditures are typically of little help because they are
undertaken one by one; and because they rely on the revenue forgone
method, they incorporate no behavioural response on the part of taxpayers.
The US Joint Committee on Taxation (2007) has periodically provided
distributional estimates for selected individual income tax expenditures, and
in its most recent annual report included estimates for 11 provisions.
Burman, Geissler and Toder (2008) provided joint estimates for “non-
business tax expenditures” which they define as “all tax expenditures
reported on individual income tax returns, with the exception of those that
affect taxes paid by business, such as depreciation allowances and business
tax credits.” Their estimates account for roughly 90% of the revenue loss
from those tax expenditures. They find that “tax expenditures in the
individual income tax benefit taxpayers in all income groups. They benefit
high-income taxpayers more than low-income taxpayers in absolute terms
and relative to their income, but less relative to the taxes they pay. The
distributional effect of eliminating tax expenditures depends on how the
budgetary savings are distributed.” In other words, there are two key points.
First, any distributional assessment of tax expenditures, like for any
structural tax changes, can appear different depending upon the choice of
prism between changes relative to tax liability and changes relative to after-
tax income. Second, the ultimate distributional effects of tax expenditures
are highly dependent on the behavioural responses of both taxpayers and
government policy makers, who could substitute spending programmes for
tax expenditures, distribute revenue raised by cutting tax expenditures
through structural tax cuts, or use the revenue to reduce fiscal deficits.
Efficiency and effectiveness

Existing tax expenditures are difficult to evaluate and trade off with
realistic alternative spending programmes – if any – that could pursue
similar objectives. This is in part due to the customary division of labour in
legislative bodies, where tax policies and spending programmes are often
considered by different committees. It is also true because of the nature of
tax expenditures, with benefits flowing to individuals and corporations
through reduced tax liabilities. The effects of tax expenditures in terms of
induced individual and business behaviour can be difficult to measure and
difficult to compare with the outputs of government spending programmes.
Finally, tax expenditures and spending for an identical purpose are rarely
truly equivalent programmes. Because tax expenditures function through the
tax system, they can have different effects on different taxpayers based upon
their differing marginal tax rates, or the status of the taxpayers as taxable or

non-taxable based on their level of income. Spending programmes rarely

would be designed to rely so explicitly on tax status, and so likely would
function differently.
Similarly, the core competencies of revenue authorities generally do not
include programme management, as opposed to straightforward revenue
collection. There can be administrative efficiencies or economies of scale
and scope in the use of tax expenditures, because relevant data are already
collected on tax forms, and the existing tax-filing process can therefore fulfil
an additional function. However, these efficiencies can result merely from
the absence of truly necessary programme administration. Tax authorities
may lack the programmatic expertise necessary to determine eligibility, and
the premium on quick processing of tax returns may conflict with sufficient
oversight. In this respect, there is an inherent difference between the
administration of a tax expenditure and, for example, a government grant.
The grant, typically, would be based on applications that would be accepted
or rejected before checks are cut and the grant proceeds are used. In
contrast, the beneficiary of a functionally equivalent tax expenditure usually
would make a financial commitment and then claim a tax reduction on the
basis of it. Thus, unlike the spending grant, the taxpayer’s financial
commitment could already be made before his or her claim could ever be
contested – which arguably makes the government’s task to revoke the
claimed tax benefit somewhat more difficult, or in some instances even
impossible. The need for additional data to judge eligibility can conflict with
the value of limiting the number of tax forms and the amount of information
on them. Beyond outright undetected fraud, this can mean that the objectives
of tax expenditure programmes are not truly met, or are met at higher net
cost than would be the case under spending programme administration.
Both the efficiency of the achievement of the objectives of tax
expenditures and the ease of evaluation of their relative success, are reduced
by the upside-down subsidy effect. Perhaps the simplest way to implement a
tax expenditure may be to exclude or deduct from income an item of income
or expenditure. In some instances, it might be appropriate that an item of
income simply not be reported. However, under a progressive income tax
with some unconditional amount of tax relief, the behavioural incentive of
such an exclusion or deduction to earn such excluded income or undertake
such expenditure would be greater for higher income persons, less for
moderate-income persons, and perhaps nil for those with the lowest
incomes. In instances such as a policy inducement to save for retirement or
to purchase medical insurance, this incentive pattern might be judged
absolutely perverse – giving the most inducement to those who need the
inducement least – and yet it is the common practice in at least some
countries.

Though evaluation of tax expenditures may be difficult, a more serious

problem may be the failure to try. Tax expenditures typically are written into
permanent law, in contrast with annual appropriated spending and with
mandatory or entitlement programmes that are sometimes required to be
reauthorised every few years. Thus, there is a perceived opportunity for
potential beneficiaries who approach the political and policy making process
to obtain a long-term, or even a perpetual, benefit by achieving the
enactment of a tax expenditure rather than a spending programme. In
fairness, it should be noted that review and oversight of spending
programmes, especially mandatory programmes, often falls short of the
ideal.
There are also weaknesses in tax expenditure reporting in the budget.
Tax expenditures are seldom presented together with equivalent spending
programmes. Seeing these costs together could make it easier to weigh cost
amounts and consider tradeoffs.5 Especially when combined with the
legislative stovepipe structure of responsibility between outlay and tax
policy, an out-of-sight, out-of-mind attitude can arise and continue to
insulate inefficiencies from scrutiny for periods of years.
Complexity
Tax expenditures, like tax systems themselves, can be complex. Still,
aspects of tax expenditures can cause the resulting complexity of the whole
to exceed the sum of the complexity of the parts, in public perception as
well as reality. As legal provisions, regulations, instructions and forms are
piled upon one another, the body of tax wisdom needed to navigate the
system can grow beyond the capacity of many non-experts. The marginal
added provisions, even if they do not apply to a particular taxpayer, obscure
that taxpayer’s field of vision of what he or she needs to know. From a
simple systems perspective, the potential interactions among additional tax
expenditures could grow geometrically as more are added.6
To the typical taxpayer, as the mass of the tax system’s processes
becomes increasingly forbidding, the perception of unfairness and of being
left out from unknown but assumed benefits for others could be
demoralising. Likewise, the manipulation of unintended interactions among
tax expenditures could reduce the efficiency of the allocation of resources,
and reduce revenue relative to what was intended or needed, as well as
reducing real and perceived fairness. Thus, the lack of transparency of tax
expenditures can have a real cost in the effectiveness of government.

Revenue sufficiency
Though tax expenditures may be judged in some instances to be optimal
uses of public resources, they always reduce revenue,7 and thus always
present a trade-off with general rate reduction. To the extent that tax
expenditures or interactions among them are used opportunistically, or that
the revenue costs of tax expenditures are simply underestimated at their
creation, revenues will fall short of what was intended or needed.
The revenue cost of tax expenditures can be more difficult to estimate
than the cost of government spending programmes. To be sure, the costs of
even annually appropriated spending programmes can be unpredictable.
Take a conspicuous example: large-scale, long-term construction
programmes can experience cost overruns, and delays can add costs due to
cumulative general inflation. However, tax expenditures can confront
unanticipated levels of taxpayer take-up. Mis-estimations of revenue cost
can lead to unanticipated fiscal deficits.
Tax expenditures can also confront cost uncertainties simply because of
measurement difficulties involving changes in utilisation across the
progressive marginal tax rate schedules that are typical of individual income
taxes. A tax expenditure delivered through a deduction or exclusion, and
even some tax credits that are phased out as income increases, can fluctuate
in cost as taxpayers move between tax rate brackets. This phenomenon can
be driven by unexpected changes in inflation as well as by real economic
growth, because adjustments in tax parameters for inflation typically are
made only with a lag. Such measurement unpredictability renders the
measured “cost” of a tax expenditure(s) much less useful as a budgetary
target.
By this same token, however, the case against tax expenditures on fiscal
grounds must be fair. The cost of existing unchanged tax expenditures can
increase because of real economic growth or inflation, in fashions that might
arouse undue concern. For example, accelerated real economic growth could
push taxpayers into higher progressive tax rate brackets, even if those tax
brackets are indexed, thereby increasing the measured cost of tax
expenditures that operate by deduction or exclusion from income – while
increasing net revenues and decreasing fiscal deficits at the same time. And
given the typical nature of tax expenditure cost estimates, which are
undertaken for each tax expenditure separately, income growth pushing a
taxpayer into a higher marginal rate bracket can increase the measured cost
of multiple tax expenditures that operate through exclusion or deduction.
Slower real economic growth might have a disproportionate slowing effect
on tax expenditure growth, as this same mechanism operates in reverse. Or
slower growth might make more taxpayers eligible for wage-supplement or

wage-replacement tax expenditures, thereby increasing the measured cost. In

addition, in the United States, even the number of tax expenditures as well
as their estimated cost, can change from year to year without legislative
action. Some tax expenditures are not specifically created in law, but rather
arise over time through practice as justified by interpretations of regulations.
Thus, there have been occasions in which tax expenditures have been
“discovered” by the analysts who prepare the lists, and so the numbers of tax
expenditures have changed without legislative action. For these reasons, the
measured revenue cost and the number of tax expenditures should be used
only with analysis of the sources of the changes and how they relate to the
underlying policy issues.
A question of definition
Some might consider designation of a revenue provision or practice as a
“tax expenditure” to presume that this provision or practice is bad – and that
this presumption is either wise or unwise. To some degree and in some
instances, such a presumption would depend on the definition used by a
particular country, or by a particular policy analyst. For example, some
might argue that a particular tax expenditure is justified because it adjusts
for the taxpayer’s ability to pay – as would an allowance for disabled
taxpayers. Yet under some definitions, a tax provision that defines ability to
pay would not be considered a tax expenditure. Similarly, a tax provision
that simplifies and facilitates the administration of the tax system – such as
using a fixed-currency amount for an allowance, rather than requiring
precise accounting – would not be considered a tax expenditure in some
countries.8 Such a definition, in effect excluding from the tax expenditure
designation any tax provision that had merit, might be seen by some as
tendentious. However, such examples do highlight that the definition that
any particular policy analyst brings to the subject of tax expenditures might
influence the implicit tone of any discussion. To be explicit, this report will
not presume that designation as a tax expenditure is an implied badge of
demerit, but rather that every tax provision or practice, designated tax
expenditure or not, should be evaluated individually, on its own merit – as
should every spending programme, in a world of scarce resources.
Practical allegations: Why do tax expenditures multiply and grow?
The number of tax expenditures

Tax expenditures arguably can be less difficult to shepherd into law than
spending programmes. This could be true for several reasons. As noted

earlier, tax laws follow their own stovepipe through the legislative process,
using committees with tax rather than substantive programme expertise.
Thus, merit relative to alternative spending programmes to pursue the same
objective might not be well evaluated. It may be easier to add generally
unrelated provisions to a tax bill than to a spending bill that could be
focused on a single programme or objective, thus allowing the number of
tax expenditures included in a single tax bill to grow.
There may be opportunities to create marginally justifiable tax
expenditures where no such opportunities would exist for programmes of
similar merit and purpose on the spending side of the budget. There is
probably a lower perceptual hurdle to surmount to justify a tax cut than a
new spending programme. It may also be easier to advocate reducing the
taxes of someone who engages in a particular meritorious behaviour than it
would be to argue for printing a physical government cheque to the same
individual. Thus, tax expenditures may sometimes be the most attractive
option available to private interests who seek government support for their
chosen activities. Perception of such a double standard could contribute to a
popular sense of unfairness and reduced taxpayer morale. Specific industry
subsidies might be taken to be particularly subject to this criticism. At the
same time, counter-arguments to repeal of such targeted tax expenditures
would be raised on the ground that the value of the subsidies had been
capitalised into the market prices of the assets, and that recent purchasers
would in effect lose part of their investments should the tax benefits be
eliminated. (The same arguments could be made with respect to industry
subsidies in the spending budget, of course.)
In some countries, tax revenues might be perceived to run at some
customary level, in terms of the share of GDP. Such a sense of regularity
may obtain in a country with a relatively high ratio of tax receipts to GDP,
just as easily as in a country with a lower customary ratio. In good times,
when receipts rise and deficits fall, it can easily be perceived that it is time
for a tax cut. At such times, the enactment of a tax bill is more likely, and
given this underlying motivation to cut taxes, the opportunity to enact
additional tax expenditures could be great.9 Such an occasion might be an
opportunity to repeal or reduce some tax expenditures and use the revenue
raised to provide an even greater structural rate cut. But the ease of
mobilising the constituencies that would lose relative advantages in the
repeal of tax expenditures might suggest the simpler road of avoiding those
tough choices and enjoying a smaller structural tax cut that would challenge
no one. Such reasoning suggests that a more far-reaching reduction of
multiple tax expenditures would be more likely as part of a budget
consolidation. There is no firm rule to favour one type of opportunity over
the other.

Both the complexity and the perceived unfairness of the tax code could
be driven as much by the number of tax expenditures as by the size of their
revenue loss. The volume of paperwork in law, regulations, instructions and
forms is not necessarily proportional to the revenue cost of a tax
expenditure. Thus, in the dimensions of administration, complexity and
unfairness, the harm due to several small tax expenditures could be greater
than that from one larger provision.
The growth of tax expenditures

Tax expenditures tend to evade systematic and critical review. As a
result, they can grow over time and avoid reform, reduction or repeal.
Common practice is that the tax law is permanent, and not subject to
regular legislative reauthorisation or review.10 In contrast with appropriated
spending, which must be re-enacted annually, or even those entitlement
programmes that are subject to periodic reauthorisation, this puts tax
expenditures in a much less vulnerable position. To be fair, of course,
budget analysts routinely decry the quality of review of spending
programmes as well – particularly mandatory or entitlement programmes in
permanent law, which are sometimes called “uncontrollable” in policy
discussions. Even oversight of annually appropriated spending is often
considered deficient.
Similarly, tax expenditures tend to be less transparent than spending
programmes. Tax expenditures are typically not displayed side-by-side with
spending programmes with similar objectives. This weakness of
presentation fails to call attention to the inevitable trade-offs between similar
tax and expenditure programmes, reducing the probability that tax
expenditures might be reformed, trimmed back or eliminated to pursue the
same objectives in alternative ways. In fact, in many countries it has
required considerable effort to ensure that tax expenditures are reported at
all, or are reported anywhere in the governments’ budget documents.
The nature of tax expenditures is also different in important respects
than that of spending programmes. A US policy maker once described the
difference by referring to the vernacular reference to budget analysts as
“bean counters,” counting the “beans” that are spent under outlay
programmes. He then drew a contrast between the spending “beans” and the
“might-have-beans” that are not collected by the revenue system because of
tax expenditures, with the important difference being the behavioural
responses to tax expenditures that are fundamentally uncertain and cannot be
accounted for. This different nature has made analysis more difficult,
thereby inhibiting comparative analysis of tax expenditures and outlay
programmes, and making reductions or eliminations of tax expenditures

harder to justify. It also makes it more difficult to design capping

mechanisms for tax expenditures than for appropriated spending
programmes, or even perhaps for mandatory spending programmes.
To the extent that tax expenditures evade review, they contribute to
longer-term budgetary problems – just as do any spending programmes that
are similarly neglected. Any changes in circumstances over time most likely
would render those tax expenditures less, rather than more, efficient and
effective, thereby with progressively lesser returns for the continuing
(perhaps even growing) reduction of government revenue. Tax rates would
have to be higher, or government deficits and debt would be greater, and
other more important priorities would be neglected.
Furthermore, even with a strong efficiency, effectiveness, or equity case
against a tax expenditure, repeal or reform of that provision might need to
surmount an additional hurdle, in that it would be a tax increase – an option
that is anathema in many political quarters. Schick (2007) hints at the
interesting observation that, if the repeal of tax expenditures and the use of
the resulting revenue to fund spending programmes is not politically feasible
because of a terminal aversion to tax increases, then existing tax
expenditures and alternative spending programmes are not really practical
alternatives after all.11
Are “make work pay” tax expenditures an important special case?

To summarise the arguments to this point: Tax expenditures can be
inferior policy instruments, degrading the efficiency and effectiveness,
fairness, and simplicity of the tax system and of government operations
generally, threatening fiscal sufficiency as well. Tax expenditures may
under important circumstances be easier to enact than spending
programmes, and not always because of underlying policy merit. Tax
expenditures are also generally less transparent, and less subject to review
and remedial action despite any policy deficiency. Recent anecdotal
evidence, to be examined in greater depth later in this report, suggests that in
keeping with this analysis, the incidence and magnitude of tax expenditures
has been growing. All of this justifies significant attention to tax
expenditures, including analysis of successful practices on reporting and
review.
With that said and acknowledged, the so-called “make work pay” tax
incentives, which are designed to increase the attractiveness of participation
in the labour force for those who have comparatively low potential wages,
are one class that might require a separate and somewhat different
evaluation.12 This fairly narrow class of tax expenditures anecdotally
represents a significant share of growth over the last 20 years. If that

anecdotal account should prove accurate, it could alter the sense of the
growth of the number and revenue cost of tax expenditures as a problem –
depending upon the assessment of the merit of the non-wastable “make
work pay” tax credit model. It would not reduce the importance of careful
management of all tax expenditures, including the “make work pay”
provisions.
The goal of making work pay has been widely accepted. Concerns that
benefits for non-working adults have deterred labour-market effort, both
because of the generosity of the benefits for the non-working and because of
the high marginal tax rates required to claw them back, have been
widespread. Given an arguable conviction that cutting such benefits would
be inhumane, the only remaining options would be for public policy to
increase the reward for work.
The menu of non-governmental initiatives is very short and arguably
unsatisfying. Increasing statutory minimum wages can be a positive step
from the point of view of equity. However, it also increases business costs.
In some circumstances, that may be an acceptable trade-off. In others, it may
threaten job creation, price stability, or both.
Government may choose to cut income tax liabilities for those earning
low wages, through increased tax-exempt levels or reductions in the lowest
bracket tax rates. That would increase the reward for those workers who do
reach the lowest tax thresholds. However, for those below the thresholds,
there would be no effect; and for those above but close to the thresholds,
positive tax liabilities would be small, and the potential effect on the
worker’s reward would be small.
So at the lowest wage levels, where income tax liabilities are at or near
zero, that leaves some form of publicly financed cash wage supplement –
hence the interest in the kind of “make work pay” policies that we have seen
in recent years. One argument against public wage supplements, which
would apply to either tax or spending vehicles, would be that any such
programme would serve as a subsidy to employers who create low-wage
jobs, and thus result in more low-wage jobs, and fewer better-wage jobs,
than would otherwise occur. However, that argument would seem to suggest
that the marginal product of the worker is determined by the “marginal
product” of the job. If skilled workers somehow found themselves in low-
wage jobs receiving public wage subsidies, and unless the wage
supplements were phased out unwisely in a cliff fashion rather than
marginally, entrepreneurs would find ways to earn larger returns by putting
those persons to work in settings that made full use of their skills. Countries
that have gone the route of “make work pay” programmes have concluded
that their populations of potential workers who would not be attractive

employees at a minimal socially acceptable wage level are large enough that
those subsidies are not disruptive or counterproductive to the labour market.
Thus, in some developed countries, though admittedly not all, the perceived
question is not whether there should be a cash-transfer policy to make work
pay, but rather what form that policy should take.
Although the “make work pay” class of tax expenditures might well be
defined to include provisions that relieve working parents of some of the
costs of childcare of pre-school aged children, and of older children before
and after school hours, those provisions might be argued by some to provide
merely a more accurate measurement of the cost of earning wage income.
The same might be said of policies to relieve the routine costs associated
with work, such as commuting expenses. More prominent among this class
of tax expenditures are wage supplements, generally delivered as a
percentage of wage income phased down once earnings pass a particular
amount.
Origin and policy issues

The pioneer among tax expenditure policies of this type was the earned
income tax credit (EITC) in the United States. That provision was enacted in
1974 and has been increased in stages ever since. It is controversial on both
policy and administrative grounds, but appears secure within the American
system. Its initial history might shed some useful light on the broader issue
of the role of tax expenditures in today’s tax systems, and the helpful
practices for managing those provisions elsewhere.
In 1974, the United States was in recession, in no small part owing to
the disruption of the world petroleum market at the time. The US income tax
was not yet indexed for inflation, and despite the weakness of the economy,
individual income tax receipts remained above their historical average as a
percentage of GDP. There was a broad consensus that the economy needed a
tax cut for macroeconomic stimulus, inflation notwithstanding, and the
taxpayers demanded relief from the effect of high oil prices and rising taxes
on living standards. At the same time, those hardships on low-wage workers
created pressure to increase the federal statutory minimum wage.13 There
was interest in a policy to “make work pay”, although that phrase had not
yet attained the identity that it has today.
A spending programme to the same effect as the EITC would have been
theoretically feasible, but not institutionally acceptable. Existing cash
transfer programmes, notably the Aid to Families with Dependent Children
(AFDC) Programme, were partly federally funded but were managed by the
states. Those state programmes generally did not cover working people
(except in a few states, and then only for those with extremely low wages),

and a federal government mandate to expand existing programmes or create

a new programme, at least to this degree and with this cost, would have been
perceived as overbearing. The alternative, to create a new federal spending
programme and the proportionate bureaucracy, would also have been
thought excessive. In the case of the United States, the additional
administrative burden of either a new federal or an expanded state
programme would have been massive. As of 2004, more than 22 million tax
returns claimed the EITC (Internal Revenue Service, 2004). Fewer than two
million families benefit from the successor to the AFDC programme, called
Temporary Assistance to Needy Families, TANF (Acs and Loprest, 2007).
Even allowing for the decline in the transfer programme caseload and the
increased generosity of the EITC since the 1970s, and for duplication of
participation between the two programmes, it is clear that implementation of
the EITC as an outlay programme would have increased the administrative
workload of the existing transfer programmes by orders of magnitude.
One advantage of the tax expenditure alternative was that it would enlist
the existing tax administrative apparatus, including the reporting obligations
of employers, possibly making efficient double use of those resources rather
than in effect duplicating that effort (as described earlier in this report). The
availability of tax data might assist in compliance monitoring and
enforcement. Furthermore, existing federal programmes, which subsidised
purchases of food (Food Stamps) and housing (public housing and rent
subsidies), could not logically be expanded to the extent desired. Because
these programmes were already intended to provide adequate access to the
“necessities of life,” it might not be deemed appropriate to provide
additional allocations of those goods and services as a reward for work, and
in the latter instance might merely bid up the price of a relatively fixed
supply of housing.
The desire to “make work pay” could not have been fulfilled by an
increase in the minimum wage alone. Given that inflation was already
accelerating, there was risk associated with adding to employer costs. Also,
the minimum wage benefited not only adult workers, but also teenagers with
part-time jobs. There was strong sentiment for targeting the relief more
narrowly, even so far as to restrict it to families with children.
Although anti-tax ideology was not as well developed at that time as it
became several years later, there was a political preference for delivering
relief in the form of a tax cut. And there were preferences for a total tax cut
package that could be demonstrated to give a “fair share” to low-income
taxpayers – even though many low-income families paid no income taxes,
given the relatively significant low-income relief provisions already in the
tax structure.

The sum of this combination of circumstances leaned heavily in the

direction of a non-wastable income tax credit based on wage income. This
combination of circumstances, though obviously different in the particulars,
has persisted in the following years and apparently has been duplicated in
other countries. It appears that these policy imperatives have made the
“make work pay” non-wastable tax credit a standard and enduring item in
the policy arsenals of some, but not all, OECD member countries, at least in
light of the apparent absence of a viable alternative.
Advantages relative to mandatory spending programmes

Given the need for any wage supplement for low-income families to be
a reliable source of income, a spending programme to fulfil the “make work
pay” function would likely be a mandatory or entitlement programme, rather
than an annually appropriated programme.14 Typically, when tax analysts
engage in a hypothetical comparison of the administrative, legislative and
transparency pros and cons of a tax expenditure and a spending programme,
they assume an annually appropriated spending programme. However, in the
instance of a “make work pay” tax expenditure, the more relevant
comparison is to a mandatory spending programme. Upon consideration, the
latter comparison arguably yields different conclusions.
Any operational differences between such a tax expenditure programme
and a mandatory spending programme would be of degree rather than of
kind. So, for example, a mandatory spending programme might or might not
be subject to greater periodic legislative review and oversight than a tax
expenditure. Mandatory spending programmes receive perhaps as much
criticism for lack of transparency, review, oversight, and legislative action
as do tax expenditures. How a non-wastable tax credit is counted in the
budget – as a spending programme, a tax reduction, or a combination of the
two – might have some impact on the degree of oversight, as well as
transparency.
Another concern about tax expenditures is that they can grow
excessively, and are not cut back in times of fiscal stress. On the first count,
it is not clear that a “make work pay” tax expenditure would grow any more
than an equivalent “make work pay” entitlement. If the tax expenditure were
to grow more rapidly because it was more effective in reaching eligible
beneficiaries, that would most likely be considered an advantage, both for
helping deserving families and for increasing the likelihood of the success of
the programme in encouraging work. On the second point, reducing a low-
income support programme in hard times could be both inhumane and
counterproductive from a macroeconomic standpoint, and so that concern
would seem no better placed in the context of “make work pay” tax

expenditures than with respect to the alternative of mandatory spending

programmes.
Because “make work pay” provisions are linked to labour income, and
labour income is reported to tax authorities as a matter of course, choosing a
tax expenditure vehicle might cut out an additional round of processing
wage and employment records by government outlay offices, at least in
some instances. Tax authorities, after the initial investment is made, can
compute the amount of the benefit and the phase out of that benefit, if
applicable, from employer information that they already have. Pulling low-
wage workers into the tax system to obtain a benefit might be thought to
increase the probability of later tax compliance if and when their wage
incomes increase.
A potential argument for a “make work pay” tax expenditure against a
mandatory spending programme would be on the grounds of inclusion. A
humane government that wanted to deliver wage supplements to low-wage
workers might believe that those workers would be more willing to accept
what was couched as a tax refund than what might be perceived as a welfare
payment. On the same grounds, that government might believe that it would
be more likely to find workers who were unaware of their deserved benefit
through the administration of a tax programme than through an outlay
programme. A taxpayer who files a return to claim a refund of wage
withholding, and who is apparently eligible for, but does not claim, the non-
wastable “make work pay” tax credit, can be pursued by the tax authorities
to verify his or her eligibility. It is not clear that there would be such an
opportunity for spending programme administrators to pursue a low-wage
worker who simply does not come forward to claim an outlay wage
supplement.
Potential disadvantages: the growth of “make work pay” non-

wastable tax credits
There has been substantial growth in the cost of “make work pay” tax
expenditures across countries in recent years. Some consideration yields at
least ten possible reasons why the cost of these “make work pay” provisions
might have increased in absolute terms, or relative to past projections of
costs:
1. Existing spending programmes might have been converted to new
“make work pay” tax expenditures.
2. New and additional “make work pay” tax expenditures might have
been created.

3. Existing tax expenditures might have been made more generous.

4. Population growth (including immigration) might have increased the
population eligible for the tax expenditures.
5. Recession might have pushed workers’ incomes down and made
them eligible for the tax expenditures.
6. Secular economic decline might have had the same effect.
7. Improved administration and outreach might find more eligible
people and encourage them to apply.
8. The tax expenditures might work: because they make work pay,
more people may be working more.
9. Manipulation and/or fraud might increase costs.
10. Estimation and projection error might have made the estimates of
future costs too low.
Of all of these reasons for growth in revenue forgone, or in non-
wastable refund outlays, only the last two present unqualified causes for
concern. Obviously, fraud must be combated, certainly involving more
intensive oversight; and estimating errors should be resolved. However, the
other potential causes are more benign – certainly including programme
success, through either greater labour force participation or more successful
outreach to beneficiaries.
Other potential sources of growth are at least somewhat ambiguous. If
these tax expenditures replaced similar mandatory outlay programmes, then
policy observers would want to be sure that these decisions were not made
to evade necessary oversight. However, in the absence of such manipulation,
policy analysts need a straightforward evaluation of the merits of mandatory
spending programmes versus tax expenditures for this particular purpose.
The creation of such tax expenditures, or increases in their generosity, could
be simple efforts to make work pay, which could in the view of some
analysts be commendable. Unexpected population growth, or changes in
macroeconomic conditions, either cyclical or secular, should be factored
into programme plans, but would not necessarily indicate an advantage of
outlay programmes over tax expenditures.
Problems of “make work pay” tax expenditures

Non-wastable tax credits as wage supplements are undeniably complex.
In the United States, the Earned Income Tax Credit is thought to be one of
the most complex provisions of the tax system. Because the most generous
US credit is restricted to workers with children, there are complex

requirements to define families, and to ensure that only one of two separated
parents claims the credit. There are provisions to offset income from capital
against income from labour, so that wealthy persons with temporary
property-income losses and incidental amounts of labour income do not
benefit from the credit. Other countries surely have encountered similar
complications. To some extent, the involvement of both the tax authorities
and employers lifts some of the compliance burden of such non-wastable tax
credits from the individual taxpayers and from spending-programme
administrators, relative to what would be the case if those programmes were
created as spending programmes rather than tax expenditures. However, as
in the instance of other tax expenditures, the “make work pay” non-wastable
tax credit is arguably under-administered and thus subject to incorrect
results and fraud.
“Make work pay” tax expenditures do raise some new and unique
issues. Tax authorities normally collect money, not give it away. Their usual
task is to ensure that individuals report incomes that might otherwise be
concealed. The revenue agencies may not be as well-equipped to deal with
people who come forward claiming that they made more income, not less.
This leaves open the prospect of a new kind of abuse.15 Presumably,
countries with non-wastable “make work pay” tax expenditures have by now
adapted to this new challenge, but with government resources always scarce,
it remains a concern.
In broader administrative terms, non-wastable tax credits add a new
dimension to the workload of tax agencies, in both the interaction with
taxpayers and data demands. Closer working relationships with outlay
agencies may be helpful, but given that an important rationale of non-
wastable credits is that their administration can double-up on normal tax
processes, the potential role of outlay administrators is to some degree
contrary to the rationale of the choice of a tax expenditure in the first place.
Non-wastable “make work pay” tax expenditures raise a second unique
issue. Almost by definition, most of the beneficiaries of such tax credits live
on tight budgets and need the additional cash from the credits day to day.
The United States has never been successful at delivering cash from the
EITC to its beneficiaries in real time, largely because of the annual
accounting basis for income tax. Instead, the credit is delivered as a lump-
sum payment after tax returns are settled in the early months of the
following year. A mechanism has been created by which firms could give
cash to their workers in effect as negative tax withholding, but it is rarely
used. Firms find it administratively burdensome. Some formulations of a
real-time benefit might leave the employer at financial risk if an employee
claim is found to be unjustified. A related complication is that the actual
EITC amount is determined on annual tax returns, but providing the benefits

to individuals in real time would require a weekly or monthly estimate

starting at the beginning of a year. American tax authorities are always
reluctant to pass out tax benefits in advance, for fear that a subsequent
change in circumstances could increase the taxpayer’s legal liability and that
the excessive tax benefits already distributed could never be recovered, or
could be recovered only with significant taxpayer hardship.
Other countries have taken different approaches to this issue. Canada
and France pay a non-wastable tax credit in periodic instalments in a given
year based upon the taxpayer’s income and other eligibility attributes the
prior year. This certainly simplifies the delivery of the tax credit in real time,
but it raises the prospect of tax credits being delivered to a newly
comfortable family (say with a student newly graduated from a professional
school) one year after it really was needed. Germany simply runs a wage-
supplement programme that might have been characterised as a non-
wastable tax credit instead through government benefits offices, and does
not consider the programme to be a tax expenditure (Koiwa, 2006).
Finally, all non-wastable credits raise transparency issues, in that they
confuse measurement of the size of government. If, arguably, there is no
substantive difference between a non-wastable tax credit and a mandatory
transfer payment that delivers the same benefit, then budget analysts must be
concerned that in the former case the government has lower measured
outlays and lower measured revenues than in the latter.16
With respect to both intergovernmental comparisons and analyses of
individual governments, accounting conventions for non-wastable tax
credits are important. The total budgetary cost of non-wastable credits might
be counted as a reduction of revenues, or as an outlay. Alternatively, the cost
might be divided, with the cash refund portion counted as an outlay, and the
reduction of tax liability counted in that fashion. There are advocates of each
treatment. Some have argued for full outlay accounting, on the grounds that
this most accurately reflects the substantive intent of the programme, and
most directly encourages analysis of the trade-offs with similar outlay
programmes (Koiwa, 2006). Others believe that dividing the cost between
tax reduction and outlay cash refunds most accurately measures the impact
on the size of government, holding that the non-refundable portion of the
credit is a reduction of tax revenue, and the non-wastable portion is a cash
benefit. The division of the cost can be uncertain, however, because for
various reasons, including taxpayer morale, cash payments can be deemed
as refunds of non-income taxes such as payroll taxes or social insurance
contributions. Identification of the provision as a tax programme might
argue for full tax treatment. Those who for political reasons would wish to
portray the programme as a tax reduction, or who would wish to minimise
the measured size of government, would choose full tax treatment.

Arguably, a non-wastable tax credit recorded solely as a reduction in tax

liabilities would be less transparent and less visible in the budget process,
because analysts simply would not know how much of the credit offset tax
and how much did not. The programme therefore would be less open to
effective review and oversight. The same might be said of treatment solely
as outlays, although that treatment could be more accurate than would
treatment as revenue reduction, if the non-wastable portion of the credit is
the larger.
To summarise, non-wastable “make work pay” tax credits and
equivalent outlay programmes arguably perform an important, if not an
essential, function in prosperous economies today. Several countries have
chosen the non-wastable tax credit model with apparent success. Where
there is a significant degree of inequality of wage incomes, with potentially
limited incentives for work on the part of less-skilled individuals, a wage
supplement through the tax system can reward effort without adding to
employer costs. The administrative apparatus of the tax system can arguably
be used to manage the programme at lesser incremental cost than would
running it as a spending programme, although that choice is to some degree
a trade-off against reduced real-time programme verification and
administration. There are undeniable complications to tax systems as a
result, but those complications would not be avoided, only handled
differently, than if the programmes were run through spending agencies. The
growing number of such tax expenditures is testimony to the conclusion of
many countries that this vehicle is the policy instrument of choice for
making work pay.
The anecdotal growth in the number and size of tax expenditures in
recent years is troubling, and justifies investigation of the implications of
this growth for the efficiency and effectiveness of government, and for fiscal
sufficiency. To the extent that tax expenditure growth has arisen from efforts
to make work pay, it might be somewhat less troubling, and it would be less
an indication of a new and more perilous situation broadly. Still, the need to
maximise the efficiency and effectiveness of the allocation of scarce public
resources, and to limit higher-than-desirable fiscal deficits, remains; careful
management of all tax expenditures, including “make work pay” initiatives,
must be an important part of that effort.
Tax expenditures and policy-making practices
In theory, it is government’s job to continually look at everything. There

is nothing to physically stop a policy making and political process from
considering all alternative policy tools to achieve each objective, and to
choose the tools that work the best, even without rules and guidelines. There

are numerous instances in recent history when problems were solved with
political will but without a formalised process. There are few instances when
problems have been solved with a formalised process but without apparent
political will. Although this statement might seem a truism, it is worth
remembering that the absence of rules should not inhibit analysis, and even
the strongest rules require analysis to make them work.
This reality is as true of tax expenditures as it is of other policy issues.
Policy analysts should not take a holiday from documenting the flaws in
current institutions because an existing process should solve the problems,
or because the absence of a process should make solution impossible.
Still, in the real world, at the margin, process can help. Sound process
and rules can help to fend off an out-of-sight, out-of-mind mentality that can
keep important issues from ever coming to debate. However, once the
debate begins, it is up to political will and compelling analysis to lead to
action.
One thing that rules can do is put known important issues on the agenda,
at least nominally. Rules can also force policy makers to at least take
explicit action that acknowledges through their votes and statements that
they are violating widely accepted practice. Rules and processes can also
force advocates of one perspective or another to at least articulate the
unstated premises of their arguments. In these important respects, process
can help.
To be sure, agendas determined by rules have been given short shrift,
and sound practice has been violated. It is always within the power of those
who write the law to change carefully considered existing law. But policy
analysts should not ignore process, because it is sometimes the best tool
available to put the facts on the table. Perhaps the ultimate testimony to
process and rules with respect to tax expenditures will occur when advocates
of some particular narrow-interest policy seek to avoid having it classified
as a tax expenditure.
This section will divide practice questions into three categories:
reporting, review and oversight, and the legislative process. The following
chapter will consider the major issue of the role of tax expenditures in the
budget process.
What practices have been successful in reporting and dealing with tax
expenditures, to maximise fiscal responsibility, oversight, transparency, and
administrative efficiency?

Reporting
Again, in theory and in an ideal world, policy makers can add and
subtract, and can compare any data regarding tax expenditures from any
source with the corresponding data on outlays in the budget. They also can
compare data on tax expenditures with other data on receipts to determine
the impact of tax expenditures on the tax base.
However, in that ideal world, policy makers can also compare data on
outlays from any source with any other one, and can schedule oversight and
review of outlay programmes at their discretion. Yet, it is considered basic
good practice to create processes that prompt regular oversight of outlay
programmes (though whether that basic good practice is fully observed is
another question). There is no reason why tax expenditures, which just as
surely allocate scarce national resources, should receive any less intense and
frequent review than outlay programmes.
One basic standard of tax expenditure reporting is that data be included
in the budget. For purposes of comparison, that is merely a convenience.
More importantly beneath the surface is that the tax expenditure data have
the same standing and be of the same level of quality as spending data in the
budget.
There has been discussion over which agency – the revenue agency or
the programmatic agency with related jurisdiction – should prepare the data.
Some might have a preference for data coming from programmatic agencies.
Thinking strictly of the source of the tax expenditure estimates, such a
preference might not be well founded. There are considerable economies of
scale in the estimation of the revenues forgone from multiple tax
expenditures. There are also benefits of consistency of methodology. If
different programmatic agencies created their own estimates of different tax
expenditures, say under the individual income tax, there would likely be
added costs of creation and maintenance of the various models, and there
would be inconsistencies of concept and quality among the different
estimates. It might make sense that all of the estimates be made by the
revenue authorities. A separate and different point, however, is that the
revenue authorities certainly should co-operate with the programmatic
agencies on obtaining and controlling the quality of any estimating input
data that do not come directly from tax returns or other tax documents. Co-
operation at that level would entail no significant additional cost, and should
increase quality.
A second standard is that tax expenditures should be reported in the
budget in proximity to the outlay programmes that have similar objectives.
Such reporting might reasonably be in addition to, rather than instead of,
reporting in a separate tax expenditure section of the budget. The objective

is to invite and prompt comparison of tax expenditure programmes with

outlay programmes, so that choices and trade-offs are confronted. Again,
however, looking at tax expenditures next to outlays is not quite an apples-
and-apples comparison; it is, rather, as was suggested earlier, a comparison
of “beans” with “might-have-beans.” Although it might be realistic to
consider the elimination of one outlay programme to finance the creation (or
the expansion) of another, it might not be equally politically feasible to
eliminate a large tax expenditure and expect to retain the additional revenues
to increase outlays, even for programmes with the same ostensible objective,
simply because there are perceived limits to the level of taxes and the size of
government. However, if there were duplications of objectives, and tax
expenditures were found to be less effective than outlay programmes, it
might be realistic to discuss the elimination of one tax expenditure to
finance another, or to finance tax-rate reduction. Another way to address
duplication would be to eliminate the spending programme. Beyond that, as
noted earlier, a tax expenditure amount does not usually accurately reflect
the change in revenues from the repeal of the tax expenditure in any event.
Thus, comparisons of tax expenditures with outlay programmes might not
be undertaken in quite the same fashion as evaluations of outlay
programmes, but nonetheless should be highly desirable.
Comparisons of outlay programmes with cheques cut for the refundable
portions of non-wastable credits might be a more natural use of tax
expenditure data juxtaposed with outlays in the budget. Policy makers need
to consider whether changing public preferences and technologies might
change the balance of merit between non-wastable credits and outlay
programmes, and a shift from one to the other would be less likely to raise
widespread concern over the level of the tax burden.
The somewhat different context of non-wastable tax credits raises the
question of what should be the reporting practice – whether the entire
amount of the credit should be reported as outlays, or revenues, or divided
between the non-wastable portion as an outlay and the portion offsetting
other tax liability as forgone revenue. Koiwa argues that reporting as
spending as opposed to dividing between spending and outlays highlights
the trade-off between the tax expenditure and outlay programmes, and
therefore yields better budgetary and fiscal control. That may be true.
However, if both portions of the tax expenditure are presented in the budget
along with outlays, the difference in oversight may be minimal; and
counting the portion of the credit that reduces other tax liability as an outlay
arguably overstates the size of government.17 This choice seems arguable.

Review and oversight

Beyond having tax expenditures reported in the budget, in close
proximity to the related outlay data, an even higher objective of process
might be to obtain regular formal evaluation of tax expenditures in the
budget documents or elsewhere. After all, the point of any presentation of
tax expenditures is to weigh their efficiency and effectiveness against
alternative spending programmes or, for that matter, general tax-rate
reduction. Such evaluation could contribute to changes in policy that would
yield a more efficient allocation of public resources.
However, obtaining such analyses, and maintaining such a level of
commitment over time, would not be easy. An elected government with
different priorities would not want to distract attention from its agenda
toward problems elsewhere, including in tax expenditures. Even more so, a
government would not want to antagonise its potential supporters on its
priority issues by identifying problems with tax expenditures that might be
politically unassailable. Government analysts would not want to anger their
political superiors by picking such fights, or to exert considerable effort in
analyses that they might believe to be quixotic wastes of time.
Elected governments have formidable tools at their disposal to discredit
and weaken the analysis of tax expenditures. Analyses in the budget can be
altered fundamentally by changes in the reference tax system, in particular
from an income tax to a consumption tax, which would define away tax
expenditures that reduce the tax burden on income from capital. Tax
expenditures can even show negative revenue forgone if they do not entirely
eliminate the tax on capital income. Merely changing the reference tax
system at frequent intervals, in whatever way, can render the underlying
analyses less useful.
Also disruptive to serious analysis of tax expenditures is a conviction
that tax expenditures reflect an underlying premise that all income belongs
to the government, therefore any tax expenditure, however distortionary and
ineffective, is preferable to its repeal.18 An example of a dispute over this
point concerns the non-partisan auditing and analysis arm of the United
States Congress concluded a recommendation for greater attention to tax
expenditures with the following:
As we move forward in shaping government for this century, the
federal government cannot accept all of its existing programmes,
policies, functions, and activities as “givens.” Outmoded commitments
and operations constitute an encumbrance on the future that can erode
the capacity of the nation to better align its government with the needs
and demands of a changing world and society. Re-examining the base of
all major existing federal spending and tax programmes, policies,

functions, and activities by reviewing their results and testing their

continued relevance and relative priority for our changing society is an
important step in recapturing our fiscal flexibility and bringing the
panoply of federal activities into line with 21st century trends and
challenges (US Government Accountability Office, 2005).
The then-sitting administration submitted the following response, which
by custom was printed in the very same report:
The GAO [Government Accountability Office] analysis in this
report is deeply flawed and it would be unwise for the Administration to
follow its recommendations. GAO believes that the Administration
should pay more attention to tax expenditures as it formulates the
budget, because of “the severity of the nation’s long-term fiscal
imbalance.” The Administration rejects any attempt to address the long-
term fiscal imbalance with tax increases (Office of Management and
Budget, 1998).
An elected government with such principles would of course seem
unlikely to pursue an aggressive analysis of the merits of particular tax
expenditures. In the United States that reasoning would appear to hold, with
two clear manifestations. First, spending programmes in the US budget are
subject to a methodology called the Program Assessment Rating Tool, or
PART. According to the budget, “these reviews have helped ensure that all
program[me]s have clear, specific definitions of success; performance
measures to track that success; and concrete improvement plans” (Office of
Management and Budget, 2008). Tax expenditures are not subject to PART
review. In addition, the report of the Senate Governmental Affairs
Committee on the Government Performance and Results Act (GPRA) of
1993 calls on the executive branch of the government to undertake a series
of analyses to assess the effect of specific tax expenditures on the
achievement of agencies’ performance objectives. The latest full budget,
produced 15 years later, provides barely three pages of general comments on
measurement issues, including a statement that current data are inadequate
for systematic evaluation of tax expenditures.
To be fair to the then-sitting US administration, the prior elected
government of the other major political party did not make much progress
on this front either. There are many institutional reasons why (Burman,
2003). Every tax expenditure has a political constituency behind it, else it
would not exist. The institutions that protect tax expenditures, including the
customary status in continuing law and the lack of regular mandated review,
mean that any discussion of a tax expenditure is easily construed as an
attack. A government that has different priorities will not want to distract
attention or generate ill will for its own proposals by calling attention to the

failings of other politically unassailable programmes. The government’s

policy analysts will not want to alienate their political superiors by
providing rigorous critical analyses of those unassailable programmes under
the same circumstances.
There may be ways to ensure that there is rigorous analysis of tax
expenditures, even though it may fall short of political ownership by the
elected government. In the United States, the non-partisan Congressional
Research Service, a part of the Library of Congress, has produced
approximately every two years a compendium of analyses of every tax
expenditure.19 The compendium includes an estimate of each tax
expenditure’s revenue cost, its legal authorisation, a description of the tax
provision and its impact, the rationale at the time of adoption, an assessment
which reports the arguments for and against the provision, a distributional
analysis where available and relevant, and bibliographic references. In this
way, the compendium makes the latest scholarship on each tax expenditure
available without the interference of political ownership – but by the same
token, without the impetus for reform provided by political ownership
either.
In a somewhat more selective vein, the nonpartisan Congressional
Budget Office (CBO) biannually produces a volume entitled Budget
Options, which includes over 100 potential policy changes, generally to
provide fiscal savings, including both spending and tax items.20 Not
surprisingly, many (but not all) of the revenue options involve reduction,
repeal or reform of tax expenditures. Not all tax expenditures are addressed,
and the presentations do not include the same comprehensive analysis of
those tax expenditures as the Congressional Research Service volume does.
However, this CBO document does hold selected tax expenditures up to
scrutiny they would not otherwise receive.
Other countries have created successful institutions to facilitate review.
Germany has begun a programme to evaluate its largest tax expenditures.
Independent, non-governmental research institutions make multiple analyses
of each provision to provide a check upon objectivity. The results are
reported to the Parliament. The Netherlands has established a five-year
schedule to review all tax expenditures, with the goals and standards of the
reviews publicly established. These processes are underway, although more
time might be needed to measure their ultimate impact. Other countries have
attempted to establish similar processes, but some have achieved limited
apparent success. Canada has engaged in evaluation only on an ad hoc basis,
but with impressive research products.

These experiences reinforce the controversy that surrounds the tax

expenditure concept. Some in the political and policy-making realm reject
policy analysts’ criticism of the inherent weaknesses of the tax expenditure
vehicle; others merely reject any criticism of the particular tax expenditures
that they support. Given this controversy, the ideal of regular, rigorous
analysis of all tax expenditures by governments, feeding back into their
policy decisions and proposals – akin to recognised (though admittedly often
ignored) sound practice for all spending programmes – can be difficult to
achieve. It is still the goal. However, for countries that have sufficient
resources to have quasi-independent and non-partisan governmental
research organisations, or as in Germany non-governmental institutions, a
second-best and more attainable and reliable approach might be to commit
those institutions to the task of providing ongoing evaluation and review, or
at least cataloguing the review undertaken by academics and others outside
of government. Time will tell whether independent analyses can motivate
action by governments with no ownership of those studies.
A remaining question that is much more relevant in presidential (as
opposed to parliamentary) systems is whether the locus of review and
oversight should be in the executive or the legislative branch of government.
Usually, executive review would be more likely to yield proposals for
reform. Precisely for that reason, institutionalised review in the executive
would be highly sensitive. The US example shows how even relatively
explicit requirements for research can be technically fulfilled by statements
about the importance of still further research, but at some time in the
indefinite future. This suggests that the second-best approach of
institutionalised non-partisan legislative review might be a second-best
alternative to blissful ignorance. Especially in the context of process rules
that force consideration of alternatives for budget savings, legislative review
might prove productive; this context is discussed later.
Legislative process and enactment

Some countries have formal budget processes that involve quantitative
disciplinary rules. Many do not. This section presents views regarding
possible legislative practices apart from a budget process – for countries
without formal processes, or for countries with budget processes to consider
in addition to budget rules. Discussion of tax expenditures and the budget
process rules themselves will come in the following chapter.
Apart from the fiscal deficit dimension, tax expenditures arguably can
increase tax system complexity and distort the allocation of resources. Some
techniques have been used in other contexts to try to address such problems
arising from other forms of government action.

For example, the United States requires regulatory consideration of the

paperwork impacts of changes in tax law and in other regulatory decisions
(Graham, 2003). Regulation impact statements have been used in New York
State (Department of Taxation and Finance) and Australia (Australian
Parliament, 1999), and have been proposed for broader use in the field of
taxation (Tran-Nam, 1999). A required assessment of the impact of any
proposed tax expenditure might force policy makers to face up to any
additional administrative or compliance burden.
The dimension of economic distortion might be measured, imperfectly
to be sure, by the revenue cost of a proposed tax expenditure. One way to
focus attention on the distortionary cost of tax expenditures would be to
require reporting of the amount of tax-rate reduction that could be financed
with the revenue loss.
In the United States, some entitlement programmes (although not the
largest and most important ones) are legally authorised for limited periods of
time (typically five years), requiring periodic reauthorisation. Doing the
same for tax expenditures might improve oversight by requiring additional
examination in the course of consideration of the reauthorisation legislation.
However, the benefits of such requirements could come at a heavy price. As
discussed above, the passage of large temporary tax cuts earlier in this
decade, on the presumption that they would be reconsidered upon their
expiration in light of any further developments on the fiscal deficit, now
appears to have been at best ineffective, and at worst highly disruptive and
divisive. In the United States, the existence of tax preferences with
scheduled expiration dates could cynically be explained by a desire to
ensure that tax bills will be taken up periodically by the legislature.
However, there is some evidence (discussed later, in Germany, Japan and
Korea) that such sunset dates might have led to more positive results in
different environments.

Notes
1. Proposals to change from an income tax base to a consumption tax base

are sometimes supported on the ground that they would substantially
reduce the number of tax expenditures. However, past changes of tax laws
generally regarded as significant have left large numbers of tax
expenditures in place. For example, the United States Tax Reform Act of
1986 repealed perhaps 19 of 119 pre-existing tax expenditures (with the
precise numbers sensitive to counting conventions; United States Office
of Management and Budget, 1988). The largest effect on the number of
dollars of tax expenditures in that instance likely came from the reduction
of marginal tax rates, which reduced the values of the many tax
expenditures based on exclusions or deductions from taxable income that
were not repealed.
2. This point, of course, does not address the argument by some that the
appropriate individual tax system, and therefore the appropriate reference
tax system, is a consumption tax that would not tax income from capital at
all.
3. Some tax expenditures, such as tax benefits for contributions to pension
plans, are capped, and therefore the value to upper-income taxpayers is
limited. However, other tax expenditures, like relief for realised capital
gains or corporate dividends, typically are not limited in this fashion.
4. Maximum marginal tax rates have declined over recent decades, reducing
this effect; the numbers of taxpayers with no liability have grown as well.
5. However, tax expenditure amounts are usually different in concept from
outlay amounts, for reasons including, but not limited to, the fact that the
benefits from some outlay programmes are taxed (thus calling for “outlay
equivalent” measures of tax expenditures). See below.
6. At least in the United States, some of the greatest perceived abuses of so-
called “loopholes” in the tax system have arisen not from individual
provisions, but from the opportunistic use of combinations of provisions,
sometimes tax expenditures, whose interactions were never intended or
even considered. Examples include real estate tax shelters, which
combined accelerated depreciation and the capital gains exclusion, among

other tax expenditures (along with the deductibility of interest, which in

the US system is not considered a tax expenditure); and the tax-motivated
sale and leaseback of non-profit and foreign property for the purpose of
claiming deductions for depreciation that otherwise would have accrued
outside of the taxable sector.
7. Ignoring the highly unlikely eventuality that a tax expenditure would have
such beneficial effect that it would increase GDP sufficiently to replace
its own direct revenue loss. In some instances, as well, negative tax
expenditures have been measured, for example, for tax depreciation
slower than the amount necessary to compensate for true depreciation or
for the effect of inflation.
8. This is the case in the definition used by the Netherlands, as one example.
See van den Ende, Haberham, and den Boogert (2004), pp. 136.
9. For example, in the United States, federal revenues have averaged slightly
more than 18% of GDP for many years, and deviations in either direction
have been short lived. In 2001, with recent revenues at record highs,
considerable momentum for a tax cut arose, and the legislation that year
included perhaps three net new tax expenditures (the increase in the
number of tax expenditures reported in the federal budget documents for
FY 2003 over that in FY 2002) while increasing the revenue cost of
several others.
10. Note discussions below of customary or mandatory “sunset” or “phase-
down” requirements in Germany, Japan and Korea.
11. During the fiscal surpluses earlier in this decade in the United States,
temporary tax reductions were enacted with the justification that they
could be allowed to expire if deficits were to reappear. Now, renewal of
those tax reductions is argued on the ground that their expiration would
constitute a tax increase, and no tax increase could be tolerated, even with
renewed fiscal deficits.
12. There are of course other common and widely accepted classes of tax
expenditures, such as deductions for charitable contributions or medical
expenses, accelerated deductions for depreciation of physical business
investments, and so on.
13. The minimum wage was increased in stages in 1966 legislation, and then
increased again in 1974. In the absence of the new tax expenditure, there
might have been pressure for a larger increase in 1974.
14. There has been some United States support for in effect converting
entitlement programmes into discretionary programmes, on the grounds
that they would then demand more attention and oversight. That notion

has not carried the day thus far, probably in part because it would at least
in form put at risk what are considered essential supports to needy
persons. See Antos, Bixby, Butler, et al. (2008).
15. In the United States, there was metaphorical concern about people
claiming labour income from taking in each other’s laundry to become
eligible for larger EITCs. And there were actual cases of fraud, using
forged employer earnings reports to claim the EITC.
16. With respect to intergovernmental comparisons, the importance of this
issue is reduced to the extent that many developed economies use the
same non-wastable tax credit policy tool; the differences among countries
are thus restricted to whether the distributed monies are accounted for as
outlays or negative tax revenues, and to the relative generosities of their
tax expenditures.
17. Admittedly, in the United States, the refundable portion of the credit is far
larger than the amount that offsets other taxes.
18. “The tax expenditure concept relies heavily on a normative notion that
shielding certain taxpayer income from taxation deprives government of
its rightful revenues. This view is inconsistent with the proposition that
income belongs to the taxpayers and that tax liability is determined
through the democratic process, not through arbitrary, bureaucratic
assumptions.” United States Joint Economic Committee (1999),
www.house.gov/jec/fiscal/tax/expend.pdf, accessed 5 November 2007.
19. United States Senate, Committee on the Budget, Tax Expenditures:
Compendium of Background Material on Individual Provisions, Senate
Print 108-54, December 2004, prepared by the Congressional Research
Service, is the latest volume.
20. Congress of the United States, Congressional Budget Office, Budget
Options, Volumes 1 and 2, December 2008 and August 2009 are the most
recent editions.

Bibliography
Acs, Gregory, and Pamela Loprest (2007), “TANF Caseload Composition

and Leavers Synthesis Report”, The Urban Institute, Washington DC,
www.acf.hhs.gov/programs/opre/welfare_employ/tanf_caseload/reports/
tanf_caseload_comp/tanf_caseload_final.pdf accessed 2 November
2007.
Antos, Joseph, Robert Bixby, Stuart Butler et al. (2008), “Taking Back Our
Fiscal Future,” The Brookings Institution and the Heritage Foundation,
Washington DC, www.brookings.edu/~/media/Files/rc/papers/2008/04_
fiscal_future/04_fiscal_future.pdf.
Australian Parliament (1999), “Bills Digest No. 87: A New Tax System
(Fringe Benefits Reporting) Bill 1998”, Department of the Parliamentary
Library, www.aph.gov.au/library/pubs/bd/1998-99/99bd087.htm.
Burman, Leonard E. (2003), “Is the Tax Expenditure Concept Still
Relevant?”, National Tax Journal, Vol. 56, No. 3, pp. 613-28.
Burman, Leonard E., Christopher Geissler, and Eric J. Toder (2008), “How
Big are Total Individual Income Tax Expenditures and who Benefits
from Them?”, American Economic Review, Vol. 98, No. 2, pp. 79-83.
Committee on the Budget (2004), Tax Expenditures: Compendium of
Background Material on Individual Provisions, Senate Print 108-54,
United States Senate, Washington DC.
Committee on Government Affairs (1993), Government Performance and
Results Act of 1993, Report 103-58, United States Senate,
Washington DC.
Congressional Budget Office (2008, 2009), Budget Options, Volumes 1 and
2, United States Congress, Washington DC.
Department of Taxation and Finance (2007), “Regulatory Impact Statement”
New York State Government, New York, United States,
www.tax.state.ny.us/pdf/rulemaking/oct307/part90/sapa/ris.pdf.

pp. 136.
Graham, John D. (2003), “Statement before the Committee on Small
Business, United States House of Representatives”, Office of
Information and Regulatory Affairs, Washington DC,
www.whitehouse.gov/omb/legislative/testimony/graham050103.html.
Gravelle, Jane G. (2005), “Tax Expenditures,” in NTA Encyclopedia of
Taxation and Tax Policy, second edition, edited by Joseph J. Cordes,
Robert D. Ebel, and Jane G. Gravelle, Urban Institute Press,
Washington DC.
Internal Revenue Service (2004), “Returns With Earned Income Credit, by
Size of Adjusted Gross Income, Tax Year 2004”, Washington DC,
www.irs.gov/pub/irs-soi/04in04ic.xls, accessed 2 November 2007.
Joint Committee on Taxation (2007), “Estimates of Federal Tax
Expenditures for Fiscal Years 2007-2011” (JCS-3-07), United States
Congress, Washington DC.
Koiwa, Tetsura (2006), “Recent Issues on Tax Expenditures in OECD
Countries”, OECD unpublished paper.
Schick, Allen (2007), “Off-Budget Expenditure: An Economic and Political
Framework,” OECD Journal on Budgeting, Vol. 7, No. 3, OECD, Paris.
Surrey, Stanley S., and Paul R. McDaniel (1980), “The Tax Expenditure
Concept and the Legislative Process”, in The Economics of Taxation,
edited by Henry J. Aaron and Michael J. Boskin, Brookings Institution
Press, Washington DC, pp. 123-44.
Tran-Nam, Binh (1999), “Tax Reform and Tax Simplification: Some
Conceptual Issues and a Preliminary Assessment”, Sydney Law Review,
Vol. 21, No. 3, www.austlii.edu.au/au/journals/SydLRev/1999/20.html.
United States Government Accountability Office (2005), “Government
Performance and Accountability: Tax Expenditures Represent a
Substantial Federal Commitment and Need to Be Reexamined,” GAO-
05-690, Government Accountability Office, Washington DC, pp. 72-82.
United States Joint Economic Committee (1999), “Tax Expenditures: A
Review and Analysis,” United States Congress, Washington DC,
www.house.gov/jec/fiscal/tax/expend.pdf, accessed 5 November 2007.

United States Office of Management and Budget (1988), Budget of the

United States Government: Special Analyses, Fiscal Year 1988,
Washington DC, pp. G42-46, table G-2.
United States Office of Management and Budget (2008), Budget of the
United States Government, Fiscal Year 2008, Washington DC, p. 25.

3. THE ROLE OF TAX EXPENDITURES IN THE BUDGET PROCESS – 59
Chapter 3
The role of tax expenditures in the budget process
This chapter discusses the role of tax expenditures in the budget process. Allowing the
enactment of new tax expenditures without careful consideration of measurement issues
and of regular review and oversight would make subsequent budget control much
harder. A key question is how budget control processes can be designed to put tax
expenditures on equal footing with spending decisions. It could be argued that a
properly configured spending rule would be more effective than a deficit rule, both in
maintaining fiscal balance and in creating incentives to control tax expenditures.

60 – 3. THE ROLE OF TAX EXPENDITURES IN THE BUDGET PROCESS
One reason why tax expenditures have attracted more attention of late,
beyond their apparent recent growth in number and size, is that fiscal
deficits are large in some countries, and could worsen with current adverse
cyclical developments. Large and growing deficits are now particularly
troublesome because many developed economies will soon face either
accelerating or strongly continuing population ageing, which will tend to
make fiscal deficits even worse. One result of looming fiscal deficits is
growing interest in budget process and disciplines, including restraint of tax
expenditures.
Disciplining tax expenditures is not easy or simple. There are several
discontinuities between the measure of tax expenditures and the reality of
budget outcomes. For example, the most straightforward and widely used
method of measurement of tax expenditures, the revenue forgone method,
does not account for taxpayer behaviour. For that reason, amounts of tax
expenditures explicitly do not equal the revenues that would be gained by
terminating them. The most likely outcome after the repeal of a tax
expenditure is that taxpayers would attempt to minimise the impact on their
spendable incomes, and so would attempt to minimise the increase in their
tax liabilities. Some tax expenditures could likely be eliminated only with
“grandfathering” of existing investments and transactions. Thus, on this one
count, the increase in tax revenue from the repeal of a tax expenditure is
likely to be less than the measured amount of the tax expenditure.
Other data regularities make the picture murkier. Tax expenditures
interact with each other in varying ways. Eliminating multiple tax
expenditures might push taxpayers into higher progressive tax-rate brackets,
raising more additional revenue than the sum of the individual estimates. In
general, it is difficult to predict the revenue effects of changes in tax
expenditure policies from the published estimates, especially considering
that those estimates typically use the revenue forgone method. Under other
circumstances, multiple repeals could raise less than the sum of the
individual items.1 For this reason, revenue agencies routinely warn analysts
against summing the estimates of individual tax expenditures to arrive at a
total. It would raise complications if such an inexact total were
operationalised as a target of budget policy. Of course, estimating the effects
of policy changes for mandatory or entitlement spending programmes is also
difficult.
There are other potential peculiarities. Faster income growth pushes
taxpayers into higher tax-rate brackets, increasing measured tax
expenditures even if the underlying law does not change.2 Tax expenditures
can evolve through changes in taxpayer practice or tax regulations, even
without legal action, such that their revenue costs could increase or decrease
relative to prior estimates.

Yet another question is the tenor of public attitudes toward the level of
taxation, however replete with tax expenditures the tax system may (or may
not) be. Whether a country’s level of taxation is relatively high or relatively
low, when revenues grow above a historical average level, which most
typically occurs in good economic times when fiscal deficits are declining,
there can be strong public sentiment pushing toward a tax cut. If such an
implicit limit on tax revenues should be binding, it is unlikely that tax
expenditures could be repealed, thus increasing receipts without
compensating reductions in tax rates or similar changes in other structural
tax features that would reduce the level of revenue back down toward the
historical range. Of course, this argument could be broadened to hold that
taxes are not likely to be an acceptable tool to narrow the budget gap under
any circumstances. That broader argument may be politically realistic, but it
would be most discouraging from the point of view of fiscal responsibility.
A still broader view may suggest that major government spending
programmes are just as politically entrenched as are tax expenditures,
meaning that any deficit reduction is an uphill climb, and no option should
be discarded.
There is another side to the budget control process, which is preventing
the enactment of policies that worsen the outlook. Here, vigilance against
the expansion of existing tax expenditures or the enactment of new ones
could prevent both short-term decline of public budgets and long-term
creation of implicit property rights in those particular provisions.
Furthermore, because tax expenditures are typically enacted into permanent
law with no guarantee of regular subsequent review and oversight, allowing
the enactment of new tax expenditures without careful consideration would
make subsequent budget control much harder.
In sum and as a broad conclusion, measured tax expenditures are an
imperfect target, at best, for a budget control strategy. On the other hand,
individual tax expenditure policies as a matter of principle should be as
likely candidates for action to reduce a fiscal deficit as any other
government policy, including spending programmes and structural tax
features. Thus, tax expenditure evaluation should be part of the efforts for
fiscal consolidation, which might or might not occur under the influence of a
fiscal rule or a budget process. A key question is how budget control
processes can be designed to put tax expenditures on equal footing with
spending decisions.
To repeat some of the earlier discussion, the creation or expansion of tax
expenditures can be the budgetary path of least resistance, offering multiple
political benefits and advantages. Tax expenditures have the appeal of
reducing taxes, for however narrow a beneficiary population. Relative to a
spending programme to the same effect, tax expenditures result in a smaller

measured size of government, which may be politically attractive. Tax law

can be complex, meaning that benefits for targeted populations can be less
obtrusive, and might be politically acceptable as tax reductions when they
would be unthinkable as spending programmes. Because they are typically
written as permanent law, tax expenditures can be a more secure source of
benefit for favoured populations than spending programmes that require
annual appropriation, or at least periodic reauthorisation. Indeed, tax
expenditures may have advantages in the policy-making process over
general, structural tax cuts. For the revenue cost of a general tax cut so small
as to be imperceptible, a substantial benefit can be directed to a small group
of favoured taxpayers.
The political attraction of tax expenditures can affect even the mix of
receipts by type of tax. If the policy-making system can give favours to
selected constituencies most easily through tax expenditures in the income
tax, there may be a tendency to increase income tax rates, reduce receipts by
giving favours through tax expenditures in the income tax, and then make up
the difference and raise further necessary revenues through other taxes. Such
a distortion of tax policy making is probably undesirable.
To be sure, some of the political advantages of tax expenditures can be
achieved through enactment of permanent mandatory spending programmes.
As noted earlier, for purposes of support of low- and moderate-income
working families, a non-wastable tax credit and a mandatory spending
programme are close alternatives. And realistically speaking, the difference
in merit between the two can be quite small, especially if benefits of
whatever form for low-income working families are unlikely to be cut for
fiscal consolidation. That would leave associated issues, such as: Is the
oversight of mandatory programmes any more rigorous than for tax
expenditures? Are programme management, innovation and improvement
greater or less for tax expenditures than mandatory programmes? Is control
of fraud greater or less for tax expenditures? Is beneficiary service
(including the availability of payments on a periodic basis throughout the
year, as opposed to only once per year) greater or less for tax expenditures?
Are programme features such as implicit marginal tax rates for phasing out
the benefit, treatment of married couples, and the like better or worse for a
tax expenditure programme? In sum, with a non-wastable tax credit, there is
likely to be less real-time programme administration, less-accurate or no
real-time delivery of benefits, and resulting lower programme maintenance
costs; whereas with an entitlement spending programme, there is likely to be
greater real-time programme administration, real-time delivery and review
of benefits, and resulting greater programme maintenance cost. Depending
on programme reporting, there may be greater transparency with an outlay

programme, and a different measured cost of government. It may be

arguable which form a “make work pay” programme should take.
Still, apart from such “make work pay” policies, there is distinct
political attraction to the tax expenditure mechanism as a means to direct
public resources toward relatively narrowly targeted ends. Therefore, any
budget control mechanism must plug the tax expenditure loophole to be
fully successful.
Types of budget rules
At a high level of conceptual aggregation, there are two broad types of

budget rules. A deficit rule sets as its target a level of the fiscal deficit,
either in currency or as a percentage of GDP. Examples of this type of rule
are the European Community’s Stability and Growth Pact and the US
Gramm-Rudman budget process of the late 1980s. On the surface, a deficit
rule would seem to address tax expenditures directly, in that revenue losses
increase the deficit. In practice, however, deficit rules can be relatively
ineffective against tax cuts in general or any other use of government funds,
because such programmes can be created or expanded when the economy is
strong, and the deficit target is not binding; and then, when the economy is
weak and the deficit rises, the weakness in the economy can be used as an
argument against reversing policy – which would have an undesirable pro-
cyclical effect. Thus, a deficit rule is not by its nature an air-tight protection
against the enactment or expansion of tax expenditures, and the task under
such a rule is to make the rule effective generally. There is a case to be made
that this task is extremely difficult (Anderson and Minarik, 2006).
The second type of rule is a spending rule. A spending rule sets a target
of levels or changes of spending, rather than using the fiscal deficit as a
goal. The advantage of a spending rule is that it can require, and allow,
counter-cyclical behaviour: restraint in good times, without enforced
restraint in bad times (Anderson and Minarik, 2006). When the economy is
strong, a spending rule does not allow policy to use any budget windfalls.
However, when the economy is weak, there is no requirement for pro-
cyclical tightening of the budget, and the budget’s automatic stabilisers are
allowed to work to cushion the downturn. American experience with such a
rule in the 1990s was positive, until the rule was first waived in 2001 and
then allowed to expire.
On its face, a “spending” rule would seem to allow tax expenditures free
rein. And in fact, Sweden employed a spending rule that imposed no
restraint on revenue policy. Such a rule may constrain measured spending
and the measured size of government, but unless those are the sole

objectives, its open loophole for tax policies to pursue spending-policy goals
would seem to be a fatal flaw. As the most fundamental analysis of tax
expenditures has explained from the very beginning, preferential exceptions
to broad-based, neutral tax systems are in important ways the equivalent of
spending. A budget process that limited the amount of spending but not the
expansion of tax expenditures could be expected to steer plans for the use of
public resources over the path of least resistance to the revenue side of the
budget, in the form of tax expenditures – as illustrated by the experience in
Sweden. However, the term “spending rule” can, and probably should, be
recognised as a misnomer for such budget disciplines. A “spending” rule can
cover revenues as well as spending narrowly defined. The US application of
a spending rule included “pay-as-you-go” as a logically complete budget
discipline system. In addition to statutory limits on annual appropriated
spending, the pay-as-you-go discipline restricted combined changes of
mandatory spending and taxes, including both tax expenditures and changes
to structural tax parameters. Any change in mandatory spending or tax
policies that increased the deficit over an estimation period of up to ten
future years was required to be fully paid for by other changes in the same
universe of policies.3 Violations of the appropriated spending caps, and of
the pay-as-you-go restrictions, were sanctioned separately by across-the-
board spending cuts in the offending category. This system allowed no profit
from converting an entitlement programme to a tax expenditure. Converting
appropriated spending to a tax expenditure would make room under the
appropriated spending cap, but would require an equal pay-as-you-go
saving.
The pay-as-you-go system as applied in the United States was effective
prospectively, and thus was subject to measurement manipulation such as
underestimating costs or legislating unrealistic policy sunsets. However, a
prospective deficit rule would be subject to the same manipulations. In fact,
one might conclude that a pay-as-you-go rule would be less vulnerable to
manipulations through optimistic mis-estimation than would be a deficit
rule, because under a pay-as-you-go rule any error would be around a total
estimated to have a net zero impact on the budget. In contrast, policies
enacted to dissipate a budgetary windfall under a deficit rule would begin
with an adverse net impact, and any estimating error would proceed from
that point.
It could be argued that pay-as-you-go, as applied in the United States, is
weak in that it restricts only adverse changes in policy, allowing an existing
structural deficit to continue. However, the same system could be initialised
with a requirement for specified amounts of future policy savings to be
achieved from mandatory spending and taxes – including, if desired, a
specified amount of budget savings through reductions and reforms of tax

expenditures. Some have argued for adding a deficit rule to a spending rule
such that the tighter of the two constraints should bind. However, if the
overall system could be satisfied by meeting one or the other rule, in effect
the looser of the two constraints, then all of the weaknesses of a deficit rule
would apply. If the ultimate constraint were the tighter of the two rules, it
would require pro-cyclical budget tightening in bad economic times.
A pay-as-you-go rule could create an incentive to cut or repeal existing
tax expenditures or mandatory spending programmes to finance any new
mandatory spending programmes or general tax reductions. Pay-as-you-go
could also stimulate evaluation of administration and quality improvement
of tax expenditures, to achieve better outcomes from the fixed base of
available mandatory spending and tax programmes.
The spending rule including restraint on taxes in the United States is
widely held to have been instrumental in the progress of the budget from
large deficit in the early 1990s to substantial surplus at the end of that
decade. However, as yet more proof that sound process without political will
is for naught, the spending rule was overridden in the first years of this
decade, and then allowed to expire at the end of 2002, and substantial fiscal
deficit has been the result. Still, based on this experience, it could be argued
that a properly configured spending rule would be more effective than a
deficit rule, both in maintaining fiscal balance and in creating incentives to
control tax expenditures.

Notes
1. In a US example, repealing one itemised income tax deduction might

leave a taxpayer claiming the fixed standard deduction, such that repeals
of additional itemised deductions would have no revenue effect from that
taxpayer. Burman, Geissler, and Toder (2008) provide estimates for non-
business tax expenditures in the United States, which are further
complicated by interactions with the alternative minimum tax (AMT).
2. This would certainly apply to faster real economic growth. It could apply
to faster inflationary economic growth if the tax law is not perfectly
indexed, or if tax-rate brackets and allowances are indexed with a lag (as
they almost certainly would be).
3. Legislative rules also provided a point of order against policies with
measurable net costs beyond the ten-year estimation window.

Bibliography
Anderson, Barry and Joseph J. Minarik (2006), “Design Choices for Fiscal
Policy Rules”, OECD Journal on Budgeting, Vol. 5, No. 4, pp. 159-208.
Burman, Leonard E., Christopher Geissler, and Eric J. Toder (2008), “How
Big are Total Individual Income Tax Expenditures and Who Benefits
from Them?”, The American Economic Review, Vol. 98, No. 2,
American Economic Association, Nashville, Tennessee, United States.

4. COUNTRY PROFILES: METHODS, INSTITUTIONS AND DATA – 69
Chapter 4
Country profiles: Methods, institutions and data
This chapter presents summary comparisons of ten countries’ (Canada, France,

Germany, Japan, Korea, the Netherlands, Spain, Sweden the United Kingdom and the
United States) description of their own concepts and methods in defining and measuring
tax expenditures, in order to provide a greater understanding of how different countries
define, measure, review, and control tax expenditures It goes on to describe each
country’s institutions and practices that are relevant to the making of tax policy in the
budgetary context. For most countries, it gives a count and a tabulation of measured tax
expenditures, as shares of GDP and relative to aggregate revenues. It disaggregates tax
expenditures under income taxes into a standardised set of budget purposes or functions.
It attempts to compensate for differences across countries in categorisations of
provisions as “structural” rather than as tax expenditures.

70 – 4. COUNTRY PROFILES: METHODS, INSTITUTIONS AND DATA
The following sections present summary comparisons of each country’s

description of its own concepts and methods in defining and measuring tax
expenditures. The sections go on to describe each country’s institutions and
practices that are relevant to the making of tax policy in the budgetary
context. A difficult goal is to shed light on some important questions about
the handling of tax expenditures in OECD member countries. What
government offices provide the accounting for tax expenditures? What can
be said about the quality of the measurements? Are tax expenditures
reviewed more or less carefully or often than similar mandatory spending
programmes? How much of the recent growth of tax expenditures is
accounted for by “make work pay” provisions? To what degree are tax
expenditures integrated into the budget process? Is the budget process
effective in disciplining the enactment or growth of tax expenditures?
For most of the countries, availability of data and language issues
allowing, there is also a count and a tabulation of measured tax
expenditures, as shares of GDP and relative to aggregate revenues. Tax
expenditures under income taxes are disaggregated into a standardised set of
budget purposes or functions. There is also an attempt to compensate for
differences across countries in categorisations of provisions as “structural”
rather than as tax expenditures, hopefully without doing violence to the
countries’ own measurement standards. For all of those countries, there
remain questions as to the precise nature of some important tax
expenditures, based upon issues of language and translation as well as the
inherent complexity of tax systems. Collaboration with generous colleagues
in each of these countries is gratefully acknowledged.
The goal is to provide a greater understanding of how different countries
define, measure, review, and control tax expenditures. Based on the
collaborative process thus far, few countries would claim to follow
demonstrably superior processes, or to have achieved great success in
dealing with these policy-making problems. Still, in the course of
discussion, there were some ideas raised by individual countries that
appeared to others to be worth serious consideration, or to have yielded
some apparent desirable results. The report will close with a brief
comparison and conclusion.
Notes on cross-country data comparisons
As was suggested in the discussion of the definition of tax expenditures

at the outset of this report, any comparison of data across countries is subject
to profound limitations. The comparisons in this report should be viewed
more as qualitative than precisely quantitative. Even apparently significant
numerical differences in numbers and amounts of tax expenditures can be

driven by apparently small differences in definition and judgment. To put

the issue briefly at the outset, the point of the data comparisons in this
volume is really not to provide answers, but rather to identify good and
useful questions.
The differences that emerge from the comparisons do not come from the
measurement issues that are most widely discussed at the conceptual level.
For example, a textbook overview of the measurement of tax expenditures
would compare and contrast the initial revenue loss method, the final
revenue loss (or “revenue forgone”) method, and the outlay equivalence
method. However, in practice, all countries use the revenue forgone method
of analysis, and only one country provides outlay-equivalence estimates as
supplementary information; and so this headline measurement issue has no
bearing on actual tax expenditure data. A second textbook focus is on the
choice of an income or a consumption tax benchmark for measurement of
tax expenditures. But as important as that decision arguably is in its
implications for tax policy choices, all countries either explicitly or
implicitly use an income tax benchmark; that major topic of classroom
debate has no bearing on the data presented in this report. However, every
country has some unique and subtle definitional aspects of its benchmark tax
system which all else equal render the estimated costs of otherwise identical
tax expenditures different and non-comparable from what would obtain in
every other country. The discussion that follows will attempt the impossible
task of identifying and explaining at least some of those subtle distinctions.
Even though those distinctions do render every country’s data precisely
non-comparable with every other country’s, if those data are not over-
interpreted, there are some apparent, important questions for further
research, and qualitative lessons for policy may emerge. Careful, judgmental
effort could provide some important and useful insights about how different
approaches to tax expenditure enactment, review and control could lead to
systematically different results. It would seem that research in this field,
even without precise cross-country comparisons, would be worth the effort.
Conceptual issues
As noted above, one of the fundamental issues in the choice of a
benchmark system is the taxation of income from capital. Under a
consumption tax benchmark, any taxation of capital income would be a
negative tax expenditure, and no relief of capital income taxation would be
identified as a tax expenditure. Because no country uses a consumption tax
benchmark, however, the issues in capital income taxation are not so
extreme. However, there are far more subtle associated questions. For
example, under an income tax, depreciation allowances that are more

generous than true depreciation should be identified as a tax expenditure.

There is precisely no agreement on a quantitative measure of true
depreciation. Likewise, depending on the precise interpretation of the
benchmark, there may or may not be an identified tax expenditure for any
relief from the double taxation of corporate dividends. Relief for the effect
of inflation on the real value of interest or dividend income or accrued
capital gains could raise measurement issues, although this issue is
sufficiently complex that all of the countries considered here choose to
ignore it. Countries that have schedular systems for capital-income taxation,
under which income from capital is taxed separately at its own rates, often a
single, flat rate, could choose either the schedular rate or the progressive
rates on other income as the reference system.
Personal income taxation in general raises its own issues. Most
definitions of tax expenditures would categorise provisions used to measure
the ability to pay tax as structural provisions rather than tax expenditures.
That principle raises a series of judgment calls as to whether many common
tax features are subsidies to particular households or rather structural
measurements of the ability to pay. Relief provisions for families with
children could be categorised either way. Even a non-wastable tax credit to
make work pay, especially one that is restricted to families with children,
could perhaps be said to measure the ability to pay. The benchmark unit of
tax expenditure analysis, and of taxation, is typically the family. However,
for tax expenditure purposes, Canada defines the tax unit as the individual.
That raises the prospect of different categorisations of family relief
provisions as either tax expenditures or measurements of ability to pay
under comparisons across countries. Koiwa (2006) argues that the failure to
identify ability-to-pay reliefs puts them on a different plane of analysis than
outlay programmes with the same objective.1 On the other hand, if the point
of income taxation is assessment based on the ability to pay, then
identification of such provisions as tax expenditures would seem misplaced.2
Some countries identify tax expenditures only for their income taxes on
individuals and corporations. Other countries also identify tax expenditures
for value-added or sales taxes, estate taxes, specific excise taxes, and all
other taxes. In this report, tax expenditures on different types of taxes are
considered separately. This decision is vulnerable to the potential criticism
that one country might choose to deliver a particular necessary relief
through a tax expenditure under the income tax, while another country might
decide to deliver relief to the same persons under the same circumstances
through a tax expenditure under the value-added tax.

Categorising tax expenditures

Given the underlying purpose of tax expenditure analysis to compare tax
expenditures with spending programmes, one task for this report was to
categorise tax expenditures by their “purpose,” or “function,” or “budget
function.” From an examination of existing tax expenditures, we derived a
set of purposes as follows:
• low-income, non-work-related;
• make work pay;
• retirement;
• employee benefits (non-health, non-retirement);
• education;
• health;
• housing;
• general business and investment;
• research and development;
• specific industry support;
• capital income relief;
• intergovernmental relations;
• charity;
• other.
These categories were chosen to reflect the typical range of outlay or tax
programme functions used in other contexts, without falling into excessive
detail. This did involve some ambiguity. For example, an issue noted earlier
was the narrow distinction between tax expenditures that deliver relief to
specific population groups and structural provisions that measure the ability
to pay tax, and so are not tax expenditures. Potentially borderline provisions
would include a tax credit for families with children, or a “make work pay”
non-wastable tax credit.3 Another ambiguity is a distinction between a tax
relief or incentive for all general business activity (like accelerated
depreciation for all equipment), and one for specific industries (like a tax
incentive for investment by manufacturing firms). This distinction can
become ambiguous when the incentive is for investment in computer
equipment by all firms, which this report categorises as an incentive for all
businesses (because any business can invest in computer equipment), rather

than an incentive for a specific industry (which manufactures the computer

equipment). This report establishes a category for intergovernmental tax
relief available to all sub-national units of government (such as a federal
income tax deduction for income taxes paid at the provincial or local level),
but categorises incentives for all business investment in selected
underdeveloped sub-national jurisdictions as general business tax
expenditures. Another ambiguity was resolved by categorising tax relief or
incentives for a specific industry in specific provinces as a tax incentive for
the industry, when it was found that all of that particular industry in that
country was located in those provinces. Because of perceived interest in
incentives for research, development and experimentation, there is a
category for such tax expenditures; this report restricts that category to
incentives for the activity of research, not to include investments in
equipment or processes derived from research (which are considered general
business relief where the incentives are available to all businesses). There is
a category for employer-provided benefits to employees; that category does
not include exclusions of the benefits of government programmes that
provide assistance to workers. To provide some measure of comparability
across countries for tax provisions providing relief for income from capital,
this report creates a separate category for all such provisions; it includes
estimates provided by some countries for provisions for relief for capital
gains, or for double taxation of corporate dividends, which those countries
consider to be structural provisions rather than tax expenditures. The list of
ambiguities could go on.
This list of budget purposes was applied flexibly, to reduce to the extent
possible issues of non-comparability across countries. In particular, the
category for capital income relief was chosen to collect provisions of the
type that conceivably could be considered as partial steps toward the
benchmark for a consumption tax. Thus, that category could include
provisions like Canada’s relief from double taxation of corporate dividends,
or capital gains relief, or accelerated depreciation for business investment. In
contrast, the tax expenditures categorised as specific industry relief, or
research and development incentives, or even general business incentives
would be the kinds of subsidies that would more likely be considered as
outside of the benchmark even of a consumption tax.
This report presents two sets of data presentation for each of six
countries. The first part of each table attempts to reflect the tax expenditures
as closely as possible to each country’s own intended presentation. Every
tax expenditure identified by each country is included, and provisions
identified by the two countries (Canada and the United Kingdom) that list
structural provisions, or “memorandum items,” that are not considered as tax
expenditures are not included.

The second part of each table attempts to minimise the degree of non-
comparability across countries by recategorising tax expenditures and
memorandum items to as close to a common standard as possible. There are
distinct limits to what can be done through this process, and the degree of
change in fact is quite small. However, as one example, Canada defines its
dividend relief provision as a part of its benchmark tax, and so does not
consider that provision as a tax expenditure. It is, however, included in the
list of memorandum items with a cost estimate calculated as though it were
a tax expenditure. Examination of other countries’ practices suggested that
they would consider such a provision as a tax expenditure. Therefore, in the
second part of each table, Canada’s memorandum item for reduction of the
double taxation of dividends is included as a tax expenditure in the category
for broad capital income relief. In another example, some provisions that are
identified as tax expenditures that provided relief to large groups of the
population, and that might be interpreted as measurements of the ability to
pay tax, are moved from the list of tax expenditures to separate groups of
“structural items” that are created for every country, not just for Canada and
the United Kingdom that already have themselves created categories for
“memorandum items.”
Because every country measures tax expenditures for their income taxes,
but some countries measure tax expenditures only under the income taxes,
and the degree of coverage of non-income taxes differs from country to
country, this report has not extended the categorisation of tax expenditures
beyond the income tax for any country. Tables in the annex report tax
expenditures under taxes other than the income tax simply by type of tax.
There are four tables for each country: 1) tax expenditures as a
percentage of GDP; 2) tax expenditures as a percentage of total taxes; 3) tax
expenditures by type of tax as a percentage of revenue collected under that
type of tax; and 4) a count of the number of tax expenditures, described
below.
Counting tax expenditures

This report includes counts of tax expenditures as of the latest actual
data, to give some sense of trend. Those counts, like the amounts of tax
expenditures, involve ambiguities. In general, this report considers an
incentive used by corporations and by individuals in unincorporated
businesses as one tax expenditure. However, where very similar incentives
were listed separately, either for corporations, or for individuals, or both, we
count each listing as a separate tax expenditure. Finally, there are of course
potential language issues. These and other ambiguities have the effect that
the counts reported in this report do not necessarily match those reported

elsewhere, although they may be more consistent from country to country

than might be others that were done for one country alone.
In sum, there are difficult, probably intractable, issues in cross-country
comparisons of tax expenditures. There is no way to make such comparisons
unambiguously “right.” The goal of this analysis is to make the comparison
as useful as possible – again, not to provide answers, but rather to identify
questions.
Finally, for reasons presented in earlier sections, the sums of amounts of
different tax expenditures do not necessarily equal the amounts that would
be measured if all of those tax expenditures were considered jointly. Thus,
in theory, tax expenditures should not be added. In the real world, there is
little alternative to adding tax expenditures if there is interest in their total
amounts, because quantitative analysis of any combinations of tax
expenditures is rare or non-existent.
Tax expenditures in Canada
Definition and measurement
Definition
Canada defines tax expenditures as deviations from a benchmark tax
system.
Types of taxes measured

Canada presents tax expenditure estimates covering the individual and
corporate income taxes, and the goods and services tax (GST). (The central
government’s published tax expenditure estimates cover only federal taxes,
but some provinces generate their own estimates.)
Benchmark tax system

As noted earlier, the attributes of the tax system included in the
benchmark are “the existing tax rates and brackets, the unit of taxation, the
time frame of taxation, the treatment of inflation in calculating income, and
those measures designed to reduce or eliminate double taxation” (of
corporate-source income). To resolve those choices that are not already
explicit in this statement of issues with respect to taxation of individuals,
Canada defines the tax unit as the individual, the time frame of taxation as

the calendar year, and the tax base is partially adjusted for inflation. With
respect to corporations, the tax unit is the corporation, the time frame is the
fiscal year, and the tax base is not adjusted for inflation (Seguin and Burr,
2004). Canada’s benchmark tax system for measurement of tax expenditures
under the individual and corporate income taxes is an income tax. Because
of inclusion of double-taxation relief in the benchmark, Canada’s gross-up
and credit tax provisions for corporate dividends, and its non-taxation of
inter-corporate dividends, are not considered to be tax expenditures (but
note below Canada’s extensive presentation of “memorandum items”,
including the provision for relief of double taxation of dividends, in its tax
expenditure reporting). The treatment of inflation in calculating income
relates to the existing system’s indexation of credits and progressive rate
brackets under the individual income tax; the system does not index
amounts of capital income for the effects of inflation. Because the unit of
taxation under the personal income tax is taken to be the individual, some
provisions relating to taxation of the family, which would be considered to
be structural in most other countries, are categorised as tax expenditures.
For the GST, Canada’s reference tax system is a broadly based, multi-
stage (with credit relief for business inputs) value-added tax, collected
according to the destination principle (that is, with relief for exports at the
border and taxation of imports).
Concepts
Canada presents estimates not only for what it considers tax
expenditures, but also for all but the most fundamental structural provisions
of the tax system, which would be considered part of the benchmark tax
system. Those provisions that are considered structural rather than tax
expenditures narrowly defined are listed separately as “memorandum
items.” Also included under this heading are measures where data
limitations do not permit a separation of the tax expenditure and benchmark
components of the measure. This additional information is intended to allow
users of the estimates to make their own judgments as to the proper universe
of tax expenditures for analysis. Because decisions regarding what
provisions do or do not constitute tax expenditures are controversial (an
example is Canada’s decision not to consider its dividend-relief provision as
a tax expenditure), Canada’s approach is helpful in cross-country
comparisons.
Canada uses the revenue forgone method to measure its tax
expenditures; the estimates assume no changes in taxpayer behaviour or
economic activity as a result of the presence of the tax expenditure.
Estimates are for annual cash flows, rather than present values of longer-run

or steady-state effects.4 Each tax expenditure is evaluated independently,

and so there is no attempt to capture interaction effects among any
combinations of tax expenditures; this means that any sum of tax
expenditures does not accurately reflect the combined impact of all of the
relevant provisions. Tax expenditures are also estimated independently of
effects on government spending programmes, and of any possible changes
in other government tax or spending policies that are made because of the
tax expenditures. Therefore, the amount of a tax expenditure is not a precise
estimate of the budgetary effect of its repeal.
Methods
Canada’s measurement methodology is very thoroughly documented
(Seguin and Burr, 2004). The models used for estimation employ statistical
samples of individual and corporate tax returns where the returns provide
the relevant information, and otherwise rely on administrative, survey or
other data from a variety of sources. Estimates for tax expenditures under
the GST are derived from a separate model.
Reporting
Location of estimates
Estimates are displayed in a document separate from the budget, and
therefore separate from the amounts of spending outlays for comparable
purposes (Department of Finance, 2007b).
Frequency of reporting and years covered

Reporting of tax expenditures is not required by law, but Canada reports
its tax expenditures every year. Since 1997, reports have covered the report
year, the five preceding years, and two succeeding years (thus eight years in
total). Each cycle’s tax expenditures for the three earliest years for the
individual income tax, and the four earliest years for the corporate income
tax, are developed using final administrative data; the later years’ figures are
estimates or projections. Tax expenditure figures are produced for calendar
years rather than fiscal years. Every four years, Canada produces a detailed
enumeration and description of all tax expenditures.

Policy making
Adding or expanding tax expenditures in the budget process

The budget process in Canada imposes no special restraints on
enactment of tax expenditures. However, in the early 1980s, Canada
implemented an “envelope system” in the budget formulation process, under
which a total sum for outlays and tax expenditures would be made available
for each policy area. Implicitly, line agencies could spend the amount in
their envelopes either on tax expenditures or on spending programmes. This
system had some initial success, but then was abandoned (Grady and Phidd,
1992). There were perceived problems with treating different policy areas
even-handedly and systematically. Tax expenditures proposed by the
Finance Ministry were not charged to the envelope of any line agency, and
so those agencies had an incentive to try to associate tax expenditures that
they wanted with the Finance Ministry. At the same time, line agencies had
an incentive to ask that tax expenditures in their areas be repealed, in effect
proposing tax increases to finance their own increased spending. From a
theoretical standpoint, that may be precisely what policy analysts would
want: a choice between a tax expenditure and a spending programme to
achieve the same policy goal, with the presumption that more transparent
spending programmes would tend to displace tax expenditures. However,
elected policy makers would likely see the issue differently, given that there
are probably perceived limits to the acceptable level of total revenues, and
that taxpayers and voters might not be receptive to what they would consider
a tax increase to finance higher spending.
In 1988, there was a tax reform aimed at broadening the tax base, and
some tax expenditures were eliminated. Under fiscal austerity pressures in
the mid-1980s and early 1990s, tax expenditures were considered for repeal
or consolidation. In more recent years, Canada has run consistent budget
surpluses, and there has been pressure to increase both spending and tax
expenditures.
Incentives to repeal or reduce existing tax expenditures in the budget

process
Canada has no special restraints on tax expenditures or structural tax
provisions in the annual budget process. There is no requirement that
changes in policy be fiscally neutral, although there are non-binding targets
for the growth of spending, keyed to the growth of the economy. Currently
there are only three tax provisions that are time-limited: two accelerated
depreciation provisions (manufacturing and clean energy technologies) and

a tax credit for investors who buy tax-advantaged shares in mineral

exploration companies. None of the major transfers to persons (elderly
benefits, employment insurance, and the Canada Child Tax Benefit (CCTB))
are scheduled to expire.
Policy review
Review of tax expenditures

There is no formal mechanism for tax expenditure review by Parliament
or Cabinet after provisions have been approved in a budget. However, tax
measures are reviewed on an ongoing basis within the Department of
Finance (and the Canada Revenue Agency, with respect to administrative
matters), with technical input as appropriate from line departments. Some
measures are evaluated more formally on a discretionary basis and the
results are published (Department of Finance, 2007b).
Review of comparable spending programmes

Projections of future programme spending are provided for a five-year
period in the Fall Economic and Fiscal Update (or economic statement, as
the case may be) and for a two-year period in the budget.5
Information about causes of changes in budget results relative to past

projections
The budget and the Fall Economic and Fiscal Update typically report
changes in the budgetary balance arising due to changes in revenues and
expenditures, which reflect economic and fiscal developments since the last
forecast. The impact of any new legislative changes is identified separately
(Receiver General, 2007). The budget and the Fall Economic and Fiscal
Update also present five-year forward estimates of both revenues and
outlays.6
“Make work pay” tax expenditures

Canada has several tax provisions that could be categorised as “make
work pay” tax expenditures, but other outwardly similar provisions should
not be so considered. These provisions were motivated by a perceived
“welfare wall” that imposed high implicit marginal tax rates and thereby
discouraged work effort on the part of public-benefit recipients.

The CCTB is composed of three income-tested benefits: a base benefit,

the National Child Benefit Supplement (NCBS), which is targeted to low-
income families, and the Child Disability Benefit (CDB), which is an
additional benefit for families caring for children with severe disabilities.
The main purpose of these benefits is to provide financial support to families
with children, and only the NCBS could be considered to be a “make work
pay” provision.
Although administered through the tax system, the CCTB is recorded as
an expense in the Public Accounts of Canada.
The CCTB is paid monthly to eligible families provided that each
supporting parent has filed an income tax return for the previous year. The
amount of benefits, which is paid over a July to June period, is based on the
combined net income of the supporting parents for the previous taxation
year. That is, the benefits paid between July 2007 and June 2008 will be
based on 2006 net income.
There are two other tax provisions that can be considered as “make work
pay” measures: the Refundable Medical Expense Supplement and the
Working Income Tax Benefit.
The Refundable Medical Expense Supplement provides assistance for
above-average disability and medical expenses to low-income working
Canadians. The measure improves incentives for disabled Canadians to
participate in the labour force by providing an alternative to disability-
related supports under provincial social assistance arrangements. It
effectively takes the form of a non-wastable credit that is calculated and
deducted against the individual’s income tax liability when the annual tax
return is filed.
The government has introduced a Working Income Tax Benefit (WITB)
(which is analogous to the Earned Income Tax Credit in the United States).
The WITB helps to make work more rewarding and attractive for an
estimated 1.2 million Canadians already in the workforce, thereby
strengthening their incentives to stay employed. In addition, it is estimated
that a WITB will encourage close to 60 000 people to enter the workforce.
A WITB of up to CAD 1 000 is provided to couples and single parents
with family earnings of CAD 3 000 or more and net income less than
CAD 21 167. Couples and single parents with earnings of CAD 8 000 or
more and net family income less than CAD 14 500 will receive the full
CAD 1 000 amount. The WITB is provided as a non-wastable tax credit,
effective for the 2007 tax year, with payments beginning in 2008. Since
2008, families can apply for an advance payment of one-half their estimated
annual entitlements. The WITB is generally available to individuals 19 and
older, not attending school full-time.

An additional supplement will be provided for low-income working

Canadians with disabilities, as these individuals generally face even greater
barriers to workforce participation. Employed individuals who are eligible
for the Disability Tax Credit (DTC) will qualify for the disability
supplement of the WITB. Benefits from the WITB will start when the
earnings of the DTC-eligible individual reach CAD 1 750. The disability
supplement will increase with individual earnings up to a maximum annual
amount of CAD 250. The WITB does not appear as a tax expenditure in the
current listing because it was not yet implemented.7
Number of tax expenditures
Income taxes
In 2004 (the latest year based on final or near-final data), Canada
reported 143 tax expenditures under the income taxes (see Table 4, as
reported by country). That is an increase from 126 in 1994. Over those
years, the CCTB was counted as a tax expenditure, but it has not been so
considered starting in 2006 (although Canada has other, smaller tax
expenditures which this volume would categorise as “make work pay”).
Despite that change, Canada’s tax expenditure count increased to 154 by
2007 (based on not-yet-final data). Of those 143 tax expenditures in 2004,
the largest category, specific industry relief, accounted for 34. General
business relief included 32 tax expenditures. As noted earlier, Canada
considers its unit of analysis for the benchmark tax system to be the
individual, not the family. It therefore identifies two tax expenditures in
each year that likely would not be so considered in all other countries, which
implicitly choose the family as the unit of taxation. On the other hand,
Canada lists several provisions for the relief of double taxation and for relief
from inflation as memorandum items, rather than as tax expenditures. All
other countries covered in this report would most likely count those
provisions as tax expenditures. Under the approaches of other countries,
therefore, Canada most likely would count 148 tax expenditures, rather than
143, in 2004 (see Table 4, with reclassifications by author).8
Other taxes
Canada reported 32 tax expenditures under the GST in 2004, an increase
from 31 in 1994.

Memorandum items
Canada reported data on 38 “memorandum items”, which are considered
to be structural rather than tax expenditures, in 2004. There were
44 memorandum items reported in 1994.
Amount of tax expenditures
Income taxes
In 1994, Canada reported tax expenditures under the income taxes equal
to 8.3% of GDP. By 2004, that had dropped to 5.4% of GDP (see Table 1, as
reported by country). Thus, even though there were more tax expenditures in
2004, their admittedly imprecise arithmetic sum had dropped fairly
substantially. The fundamental measurement problems prevent us from
drawing any firm conclusions, but the figures do suggest that the increase in
the number of tax expenditures may not have led to a greater influence on
the economy or on sums in the tax system. More specific research would be
needed to examine that conclusion. In 2004, the largest reported tax
expenditures were for retirement and for inter-governmental relations, with
general business incentives accounting for about half their amount. The
most numerous tax expenditures, for specific industry relief, actually
accounted for a sum equal to a comparatively small percentage of the GDP.
Under the recategorisations in this report to achieve somewhat greater
comparability with other countries, some general business relief tax
expenditures are moved into the category for capital income relief more
broadly (see Table 1, with reclassifications by author).
Other taxes
Canada reported tax expenditures under the GST that added to an
amount equal to 1.2% of GDP in 2004. That was down from 1.4% of GDP
in 1994.
Memorandum items
Canada’s memorandum items summed to an amount equal to 3.6% of
GDP in 2004. That is down from 4.3% of GDP in 1994, thus following the
decline in the reported arithmetic sum of tax expenditures.

Tax expenditures in France
Definition
Tax expenditures are defined in France as legal or statutory measures
whose implementation induces a lower tax revenue for the State in
comparison with the application of the benchmark or “norm” – that is, the
basic calculation principle of the tax.

Tax expenditures are defined for every tax, including individual and
corporate income taxes, wealth tax, value-added tax, stamp duties, and other
indirect taxes. Some public documents list tax expenditures for social
insurance taxes or contributions. Until recently, tax expenditures were
defined only at the central government level; however, since the Budget Act
for 2007, tax expenditures for local taxes (when they are refunded by the
central government) have been listed.

The “norm” is an interpretation of the intentions of the legislature. One
general criterion used to define a tax expenditure is that it have a narrow
scope. A general measure, which aims to benefit a large majority of
taxpayers, is more likely to be considered a part of the norm. Previously, the
age of a tax provision was also relevant. A measure of long standing could
have been taken to be accepted as a part of the norm, and removed from the
list of tax expenditures. However, the concept of the norm has now been
changed, and long-standing tax expenditures are no longer dropped from the
list and taken to become a part of the norm.
The French benchmark tax system for purposes of calculation of tax
expenditures under its income tax is in fact an income tax, and so provisions
that reduce the tax burden on capital income are considered to be tax
expenditures. However, the norm is conceived as a basic income tax, such
that many provisions that might otherwise be considered “structural” are
counted as tax expenditures; this aspect of the definition of the norm might
be thought to increase the number of provisions that would be considered as
tax expenditures. So for example, although the norm includes progressive
income tax rates and different filing statuses for married couples and single
persons, special allowances for handicapped persons or single parents are
not included in the norm.

Concepts
In France, tax expenditure estimates indicate the revenue directly
forgone by the government, with no accounting for behavioural responses,
changes in economic activity, or interactions among different provisions.
Accordingly, tax expenditure estimates are not precise predictions of the
amount of revenue that would be gained by repeal, and any sum of tax
expenditures does not accurately reflect the combined impact of all of the
relevant provisions. Only provisions judged to be tax expenditures are
estimated; there are no reported estimates for structural provisions that are a
part of the norm.
Methods
Tax expenditures are measured using simulations from a statistical
sample of taxpayers and by other methods. The perceived reliability of many
estimates is also specified, using designations as “very good,” “good,” or
“approximation.” For 2009, approximately 11% of all tax expenditures were
not estimated at all; this was a significant reduction from 44% not estimated
in 2001, and 20% in 2008. Of the remaining 89% of the tax expenditures
that were estimated, 24% of the estimates were judged of “very good”
reliability, 29% were “good,” and 47% were judged as “approximations.”
The number of beneficiaries of a tax expenditure is reported where it is
available.
Reporting
Tax expenditures are reported each year in the Budget Act, as an annex,
“Ways and Means Evaluation,” of the Finance Bill, as well as in the Finance
Bill for Social Security. The presentation in the Finance Bill of the Budget
Act includes a legal reference for the provision; the number of beneficiaries
(when available); the method of evaluation (when available); the reliability
of the evaluation; the year of the creation of the tax expenditure, and of the
last important modification of it; and the cost for the budget year and two
preceding years. In the Finance Bill for Social Security, there is a
presentation of the provision, a legal reference, the number of beneficiaries,
the year of creation, the cost, and whether there is a compensation for social
security for the provision.


As noted above, tax expenditures are reported each year in the Budget
Act and in the Finance Bill for Social Security. Both the Budget Act and the
Budget for Social Security Act report the cost of tax expenditures for the
budget year and the two prior years.
Policy making

At present, France has no restriction upon the consideration of new tax
expenditures, although Article 40 of the Constitution prohibits the
submission of private member bills that would either reduce public revenue
or increase expenditures (Assemblée nationale, 1958).

process
As noted above, there is no restriction on the consideration of tax
expenditures. By the nature of the spending increase prohibitions in the
Constitution, there is also no incentive to propose to decrease a tax
expenditure to finance an increase in spending.
Policy review

The Organic Law requires that the Ways and Means Annex of the
Finance Bill of the budget present an evaluation of each tax expenditure, but
to date such evaluation has been limited to an estimate of cost. A new
process of evaluation of tax expenditures began in 2006, and improvement
of evaluations is a priority. However, at this stage of the process, there is
concern that there is not yet a working set of performance criteria for those
evaluations. The criteria used at present may be too numerous and not fully
relevant. Any sunset dates for tax expenditures would be specified in the
wording of each individual provision; there is no comprehensive list of
expiring measures and the dates of their expirations.


The French budget does contain mandatory spending. It is estimated that
80% or more of the budget is comprised of mandatory spending. There are
no specific sunset dates for mandatory spending programmes.

projections
The Budget Review Act (la Loi de règlement) reports on differences of
actual revenues from projected estimates. It specifies the differences due to
legislative changes on particular taxes or tax expenditures.

France has a non-wastable “make work pay” tax credit. If this tax
expenditure exceeds the amount of tax owed, it can be granted in cash.
However, even the non-wastable amount is accounted for as a decrease in
revenues, not an expense. People receiving cash in one year can be granted a
monthly advance as an anticipation of the benefit in the next year. The
amount of the benefit is regularised once a year after the actual amount of
the credit due is known.
The main goal of this measure is to reduce the tax wedge for people in
the lower part of the income distribution and thus make their return on
labour effort attractive. France considers from a purely economic point of
view that such a measure is a tax credit rather than a public allowance. It
was also seen as more rational, from an administrative point of view, to
create a tax credit on income instead of a new spending allowance that
would have involved administrative complexity. This measure was inspired
by the similar foreign measures (such as the American EITC and the English
WFTC) which are, in the majority, considered tax expenditures. At this time
there is consideration that the insufficient co-ordination between the “make
work pay” tax expenditure and the benefits system creates inefficiencies.

This report includes no independent analysis of tax expenditure data
from France. A communication from France indicates that the number of tax
expenditures, at 486 for fiscal year 2008, is 3.6% greater than for 2007, and
21% greater than the 401 tax expenditures in 2001.9 The latest French
publication indicates that the number of tax expenditures declined to 469 in
2009 (Ministère de l’Economie, des Finances et de l’Industrie, 2007).
Recalling that, at least in the past, France’s evolving norm in effect absorbed

some tax expenditures over time, the growth in the number (and the amount)
of tax expenditures could have been greater if the norm were somehow held
constant.

Communication from France indicates that the total amount of tax
expenditures for fiscal year 2008 was estimated to be 7.1% greater than for
2007, and 16% greater than for 2001. The latest French publication indicates
that the amount of tax expenditures increased by another 4.2% in 2009.
Tax expenditures in Germany
Definition
Germany does not have a legally stated definition of tax expenditures.
The law makes reference to aid to enterprises and business sectors of the
economy. Provisions that benefit households are reported as tax
expenditures only in so far as they are indirect subsidies to private
enterprises or business sectors. Thus, the implicit definition of tax
expenditures used in Germany is somewhat different from that in other
countries.

Germany measures tax expenditures under a wide range of taxes, but not
all of its taxes have tax expenditures. The central government lists the
revenue forgone to the federal budget, to the budgets of the Länder, and to
the local government authorities; that is, tax expenditures are reported for all
levels of government. This follows from the requirement that the legal basis
for most taxes is the federal law. Even taxes (for example, the inheritance
tax) whose revenues are assigned exclusively to the Länder, or taxes whose
revenue is partly assigned to the Länder or to municipalities, must be
adopted by both of the chambers (both the Bundestag and the Bundesrat) of
the German Parliament.


There is no explicit legal definition of the benchmark tax system for
purposes of estimating tax expenditures. The implicit reference tax system is
defined by the legal language described above, and is in a sense considered
to be re-evaluated with every new tax law. Structural provisions of the law,
like the personal exemptions or progressive rates, are considered as part of
the reference tax system and thus not as tax expenditures.
Concepts
Published tax expenditure estimates represent revenue forgone on a cash
basis for the particular year. They do not account for behavioural responses
on the part of taxpayers, and thus are not intended to be estimates of the
additional revenues that would be collected in the absence of the provisions.
Because some tax expenditures are at least partly intended to simplify tax
administration, the additional revenue collected from the repeal of those tax
expenditures could be less than the reported amounts. Each tax expenditure
is evaluated independently, and so there is no attempt to capture interaction
effects among any combinations of tax expenditures; this means that any
sum of tax expenditures does not accurately reflect the combined impact of
all of the relevant provisions. Germany provides estimates only of
provisions that it deems to be tax expenditures, not for any provisions that
are considered to be structural parts of the benchmark tax system.
Methods
The tax expenditure estimates are based on different sources of data
depending on the nature of each tax expenditure. Direct payment data,
estimates based upon official statistics, and business statistics are used. In
some instances, specially developed comprehensive estimating instruments,
such as a micro-analytic income tax simulation model, are used.
Reporting
Tax expenditure estimates are submitted within the federal
government’s subsidy report (which covers both tax expenditures and outlay
subsidies, although tax expenditures are not presented side-by-side with
outlay programmes that pursue similar purposes) biannually, together with
the draft budget. A list of the 20 largest tax expenditures of the central

government is attached to the draft budget every year. These estimates are
not integrated with the information on spending programmes.

The subsidy report is submitted biannually and includes tax expenditure
figures for the current year, the two preceding years, and one future year.
Policy making

Beyond the rules of the European Union’s Stability and Growth Pact,
Germany has its own deficit-limiting budgetary rules in the German Basic
Law (the Grundgesetz) and the Budget Principles Act (the
Haushaltsgrundsätzegesetz), with corresponding regulations. The
fundamental principle is a “golden rule,” under which borrowing is to be
limited to the amount of investment expenditures, such that the
government’s net asset position is maintained. Exceptions are entertained
only in the case of an actual or impending serious and sustained
macroeconomic disturbance. There are also non-binding guidelines from the
Federal Cabinet in 2006 that new subsidies should be given as grants, or
“financial aids,” rather than as tax expenditures, and that they should be
“paid for.” These processes are seen in Germany as successful barriers
against expansion of tax expenditures. Reform of these rules to make them
even stronger is a high priority of the federal government’s fiscal federalism
programme. The goal would be structurally balanced budgets, and therefore
fiscal balance in cyclically adjusted terms.

process
As a further budgetary restraint, the Federal Cabinet decided in 2006
that subsidies (whether as tax expenditures or direct payments) should be
time-limited and should decline over time. The Cabinet also held that
subsidies should be delivered as spending programmes, rather than as tax
expenditures. Tax expenditures should be examined to consider whether
they could be changed to spending programmes. There is also a non-binding
agreement within the Financial Planning Council (which includes the
Federal Minister of Finance, the finance ministers of the Länder, and
representatives of the local authority associations) that expenditures at all
levels of government should rise no more than an annual average of 1% in

nominal terms over the medium-term financial planning horizon. Tax

expenditures are not covered by this agreement, but it is reported that there
are no initiatives to circumvent the spending limitation by increasing tax
expenditures.
Policy review

Germany has begun a process of formal reviews of tax expenditures.
The 20 largest tax expenditures – accounting for 92% of the total cost of all
tax expenditures – are to be evaluated. The evaluations are charged to define
the objective of the tax expenditure, including macroeconomic motivations
or perceived market failures; determine whether the tax expenditures are
effective and efficient, and whether the tax expenditure is the best public-
policy instrument to pursue the objective; and to find any side effects for the
tax system broadly. Several respected outside research institutes perform the
reviews, with the use of multiple reviewers seen as an important guarantee
of unbiased analysis. The Ministry of Finance will comment on the reviews,
and report the findings to the Parliament.

The future costs for public programmes are estimated annually during
the budget preparation by the respective departments in co-operation with
the Federal Ministry of Finance. The estimates cover the financial plan
horizon, which is a five-year period including the current year and the
following four years. The estimates are published in an aggregate form in
the annual Finance Report (Finanzbericht) which is submitted with the
budget bill, and every two years in the subsidy report to the Parliament.
These estimates are publicly available. The Finance Report contains
estimates of revenue and expenditure for the financial plan period, and also
retrospectively. It includes information on public expenditures by function
and economic classification from 1952 to the end of the financial plan
horizon.
There are additional reports and statistics for certain spending areas
(i.e. family, housing, labour market). The Statutory Pension Report and the
Public Service Pension Report have a special long-term focus.10 Finally,
long-term developments of public expenditures are analysed in the
Sustainability Report that is published every four years by the Federal
Ministry of Finance.

In general, mandatory spending programmes and tax law provisions do

not include sunset clauses. For subsidies subject to a time limit, data on
expirations are included in the subsidies report.

projections
Tax revenues are estimated by an independent group of experts
(Arbeitskreis “Steuerschätzungen”) twice a year (in May and November).
The results are included in the federal budget draft for the following year. In
the meeting in May the group of experts reports on the differences between
the actual estimate and the previous estimate. Differences caused by
legislative change are reported separately from other deviations, but there is
no distinction between changes caused by macroeconomic fluctuations and
those due to estimation errors.

In Germany, there are no “make work pay” tax expenditures like the
earned income tax credit in the United States. Those government
programmes to the same effect (especially the employment allowance within
the so-called “Unemployment Benefit II”) are not considered to be tax
expenditures.
Income taxes
In 2006, which is the latest year for which final or near-final data were
available, Germany reported 56 tax expenditures under the income taxes
(see Table 8, as reported by country).11 The largest category by purpose was
specific industry relief, which included 22 tax expenditures; the second-
largest category was housing, with ten. Germany’s make work pay
programme is not considered a tax expenditure, and thus does not affect any
of the counts or totals. This report agrees with Germany’s number of tax
expenditures (see Table 8, with reclassifications by author).
Other taxes
Tax expenditures under all other taxes totalled 30, including 13 under
the fuel tax, and six each under the electricity tax and the sales or value-
added tax (see Table 8).

Total
In total, Germany reported 86 tax expenditures in 2006. That was a very
small increase from 2003, when there were 82.12 However, in 2008, one tax
expenditure for owner-occupied housing is phased out, reducing the total.
Income taxes
Total tax expenditures under the income tax summed to an amount equal
to 0.29% of GDP in 2006 (see Table 5, as reported by country). Tax
expenditures for housing accounted for more than half of the total. The more
numerous tax expenditures for specific industry relief summed to only about
0.1% of GDP.
Other taxes
Tax expenditures under taxes other than the income taxes added to
0.45% of GDP in 2006. About half of the total came from the fuel tax, and
not much less from the electricity tax.
Total
All tax expenditures in 2006 summed to 0.74% of GDP. In 1980, the
equivalent figure was 0.80% of GDP; it fell to as little as 0.49% of GDP in
1995 and 1996. In part because of the phasing out of the tax expenditure for
owner-occupied housing, the total of tax expenditures was expected to
decline to 0.64% of GDP in 2008.
Tax expenditures in Japan
Definition
Japan’s legally defined analogue to tax expenditures is “Special Tax
Measures.” Special Tax Measures are provisions that take exception to
Japan’s fundamental tax principles (equity, neutrality, and simplicity) to
pursue some other policy objective.


Special Tax Measures are estimated for individual income tax,
corporation income tax, and other taxes, including inheritance tax, gift tax,
liquor tax and gasoline tax. Special Tax Measures are estimated for local
taxes as well. Article 84 of the Constitution requires that the basic
framework of any taxes, including local taxes, be passed by the Japanese
Parliament, or Diet. The ministry responsible for designing local taxes is the
Ministry of Internal Affairs and Communications (MIC). Local authorities
also retain some legislative discretion, such as applying their own tax rates
within a range set by national law. However, these autonomies seldom affect
the total amounts of Special Tax Measures in local taxes. Local authorities
usually use direct subsidies or grants to particular groups rather than local
tax expenditures to pursue their policy objectives.

As noted above, Japan defines its Special Tax Measures by comparison
to fundamental tax principles, rather than relative to a benchmark tax system
per se. For income tax, among the items considered “structural” and
therefore not included among the Special Tax Measures are employment
income (that is, salaries) deduction; basic exemption; deductions for
spouses; special exemption for spouses; exemption for dependents; and the
progressive tax rate structure itself. Japan’s concept of Special Tax
Measures is somewhat different from the notion of tax expenditures in some
other countries, and so comparisons must be made with care. However, a
fair generalisation could be that Japan’s benchmark of fundamental tax
principles is somewhat more general and broad than other countries’
reference tax systems, and a result may be the inclusion in Japan’s list of
Special Tax Measures of some provisions that would not be considered tax
expenditures in other countries.
Concepts
Special Tax Measures are based solely on revenue estimates. Outlay
equivalents or the net present value method are not used for tax revenue
estimates for future years. Japan recognises some negative Special Tax
Measures.

Methods
The tax officials of the Ministry of Finance (MOF) in charge of
estimating tax revenue, including the effects of Special Tax Measures,
usually use large volumes of tax statistics compiled by National Tax Agency
(NTA). The basic sources for tax estimates include aggregate tax
information based on the available data from tax returns in the “National
Tax Agency Statistics Almanac”. The NTA also conducts some sample
surveys for tax estimates, including “wages of employees in private
enterprises”, “financial situation of corporate enterprises”, and “situation of
filing income tax returns”. Other economic statistics, such as the System of
National Accounts, are used when necessary and appropriate. In addition,
the tax officials in the Ministry of Finance often conduct special interviews
with major corporations to reflect the ongoing economic trend in their
estimate of future tax revenue.
Reporting
Officially, the revised estimates for the Special Tax Measures are
reported to the Diet annually in the “Summary of Tax Revision” and
“Explanation of Tax Revenue and Stamp Duties Budget”, which are
submitted along with the other budget documents. The aggregate estimates
of all Special Tax Measures are also reported to the Budget Committee of
the Diet annually, though this is not the official report.

Only current fiscal year data of the annual changes and aggregate
estimates are reported to the Diet. The restriction to reporting of changes
leads to some concern that ongoing provisions are “secret subsidies”.
Policy making

There are no specific rules regarding Special Tax Measures in Japan’s
budget process. The Public Finance Act of 1947 defines Japan’s budget
process and basic budget rules, such as limits on government borrowing. It
determines the basic golden rules of the budget. The recent reorganisation of
the fiscal process added some procedural elements outside of the Public
Finance Act.

Under the current process, in June of each year the Council on

Economic and Fiscal Policy (CEFP) deliberates the fundamental issues and
determines the “Basic Policies” for the next fiscal year’s budget. It usually
makes decisions only for the next year’s budget. However, “Basic Policies
for 2006” extended to future years, determining fiscal consolidation targets
of achieving primary surplus by 2011 and then reducing the debt-GDP ratio
by the mid-2010s. These decisions included a five-year spending ceiling.
These Basic Policies were reaffirmed by the Cabinet.
In subsequent years, based on the “Basic Policy,” the Minister of
Finance proposes the detail of the spending ceilings for each spending item
in the “Guideline for Budget Request,” which is affirmed by the Cabinet in
July and August. The ceilings’ quantitative formulas are presented in the
guideline.
There is no specific quantitative target for Special Tax Measures in the
budget process, but any change of the tax system should be in line with the
fiscal consolidation targets stipulated in “Basic Policies for 2006.”

process
As noted above, any proposed change in existing Special Tax Measures,
like any proposal for a new Special Tax Measure, would have to comply
with the fiscal consolidation targets in the “Basic Policies for 2006.”
Presumably, also because the “Basic Policies” impose targets for the deficit
with respect to the GDP, policy makers would contribute to meeting their
targets by removing tax expenditures (or raising structural taxes) as well as
by reducing expenditures.
Policy review

Special Tax Measures are reviewed annually by tax officials of the
Ministry of Finance, mainly focusing on those that expire in the next year
due to sunset clauses. Usually, the majority of the Special Tax Measures at
the national level are stipulated in the Special Tax Measures Laws to have
two- or three-year sunset clauses. These sunset clauses have functioned
effectively, because they force tax officials and other related parties to
review the contents of the Special Tax Measures regularly.

Negotiations between tax officials and the requesting ministries over the
Special Tax Measures expiring in the next spring (usually the end of March)
begin in September, at the same time as the budget expenditure negotiations.
In many cases, each ministry requests the creation of new Special Tax
Measures for their policy objectives. The necessity, effectiveness and
efficiency of the measures are scrutinised in the negotiations. At the same
time, the government Tax Commission, which is an advisory council to the
Prime Minister, deliberates tax policy for the coming fiscal years. From late
November to early December, the tax commissions of the ruling parties
begin their decisions on tax policies for next fiscal year, including the
Special Tax Measures. In this deliberation, the tax officials explain the
discussions among the related ministries. In December, the Ministry of
Finance decides the contents of the tax proposals based on the report
submitted by both the government and the ruling parties’ tax commissions.
The Tax Bill is usually submitted to the Diet in the next January or
February.

The review process for tax expenditures described above is independent
from the expenditure process, though the tax revenue estimate in the next
fiscal year’s budget is based on the tax reform decided, and thus presumably
is implicitly considered along with spending plans to comply with the debt
targets in the “Basic Policies for 2006”.

projections
When the supplementary budget is submitted to the Diet during the
ongoing fiscal year, the tax revenue estimate is reviewed and, if necessary,
revised. The Final Settlement Report shows the difference between the
initial or supplemental revenue estimates and the actual results. There is no
official analysis of the cause or source of the difference between the revenue
projections and actual revenue in the final report for settlement of accounts.

Japan currently has no “make work pay” tax expenditure, and in
particular no provision resembling a non-wastable “earned income tax
credit.” The recent Tax Commission (the Prime Minister’s advisory body)
report noted that further discussions on an EITC-type policy should be
undertaken with respect to its necessity, policy goal and problems to be
solved, with reference to other countries’ experience, considering practical

difficulties in implementing such a policy and coping with possible fraud.

Currently, assistance to low-income families is delivered mainly through
social welfare spending programmes managed by the Ministry of Health,
Labour and Welfare.

As noted above, there is no public, comprehensive listing of Japan’s
Special Tax Measures. However, communication from Japan indicates that
the number of Special Tax Measures related to business enterprises has
declined from 81 in 1998 to 61 in 2007.13
Income taxes
Communication from Japan with respect to central government Special
Tax Measures indicates that Special Tax Measures under the individual
income tax, 52.5% of the total in 2007, come largely from a tax credit for
housing loans (24.0% of the total of all Special Tax Measures), with a tax
credit for dividends and a deduction for life and earthquake insurance
premiums coming next (7.8% and 4.6% of all Special Tax Measures,
respectively). Under the corporate income tax (33.7% of the total), the
largest Special Tax Measures are the special tax credit for R&D (17.9%) and
the tax credit for promoting investment by small- and medium-sized
enterprises (6.8%).
Other taxes
All central government Special Tax Measures under other taxes total to
13.8% of the total of all Special Tax Measures at the central government
level.
Local government special tax measures

Special Tax Measures at the local government level are slightly less than
one-third the amount of central government Special Tax Measures. A little
less than half of the local government total arises because of central
government Special Tax Measures, and slightly more than half comes from
Special Tax Measures that originate at the local government level.

Size of special tax measures

Central government Special Tax Measures are of a size equal to 0.6% of
GDP. Local government Special Tax Measures equal 0.2% of GDP.
Tax expenditures in Korea

Korea is now revising its tax expenditure measurement and reporting
system. Effective in 2010 (for the 2011 budget), the National Fiscal Act
(NFA) requires a “tax expenditure budget” within the budget documents.
Although not all of the details of this new system have yet been determined,
there certainly will be marked changes in the categorisation of tax
expenditures in the reports, and likely further changes in the methodology
employed. The discussion that follows describes the current procedures,
rather than any specific plans for the process in 2011 and beyond.
Definition
Korea does not now provide a formal definition of tax expenditures by
law or regulation. Rather, the NFA specifies that the Ministry of Finance
and Economy (MOFE) shall compile a report, referred to as the “Tax
Expenditure Budget Document,” that “analyses, by function and tax, the
actual amount for the immediately preceding fiscal year and estimates for
the current and following fiscal year for tax reductions and exemptions,
income deductions, tax credits, rate reliefs, and deferrals.” This document
provides a general definition as the “tax-subsidy counterpart to fiscal
expenditures…the reduction of national tax revenues that result from the
application of special provisions, as exceptions to the normal taxation
system, for reducing the tax burden of [a specific target group of]
taxpayers.” By inference, that listing has become the practical definition of
tax expenditures.

Because of the immediately preceding language of the NFA, Korea
measures tax expenditures under all of its taxes. Two of its taxes are in fact
surtaxes on other taxes, and so have no measured tax expenditures of their
own; but all 12 of the other taxes do have identified tax expenditures. Only
national-level taxes are considered.


Korea’s benchmark tax system is not yet fully articulated, but more
detailed documentation is planned as a part of the effort to refine its tax
expenditure measurement for the new process to begin in 2011. To date,
Korea basically follows the outline described in OECD’s 1996 Tax
Expenditure Report.
Concepts
The language of the NFA dictates that Korea’s list of tax preferences
will be long. Because there is no formally established reference tax system,
which could define some tax provisions as a part of a benchmark, more tax
provisions are included in the list of tax expenditures than might be the case
in other countries. Thus, like Canada and Japan, Korea has a long list of tax
provisions in its tax expenditure exercise; but unlike Canada, Korea does not
draw a distinction between tax expenditures narrowly defined, which
provide narrowly focused benefits through exceptions to the general tax law,
and “memorandum items” which follow the structure of a benchmark tax,
are broadly applied, and thus do not provide targeted preferences for small
groups of taxpayers. All of the provisions identified as tax expenditures are
included in a single list.
Methods
Korea does not provide extensive documentation of its models and
procedures for estimating tax expenditures, and in fact is aggressively
refining its process to comply with the 2011 mandate in the NFA.
Reporting
At present, Korea provides its tax expenditure estimates in a document,
the “Tax Expenditure Report,” which is separate from, and released after,
the budget. The “Tax Expenditure Report” is produced by the MOFE, not by
the Ministry of Planning and Budget (MPB), which is the agency that
produces the budget itself. Thus, tax expenditures are not now presented,
and are not likely to be presented in 2010, side by side with corresponding
outlay figures in the budget.


The “Tax Expenditure Report” is released annually, in keeping with the
language of the NFA. The NFA also mandates that estimates be provided for
the year prior to the budget year, the budget year itself, and the year after;
however, the requirement is not considered binding until 2010, and
projections for the succeeding year have not to date been provided. The
retrospective year is based on final data; the current year is a projection. Tax
expenditures hitherto have been reported according to functional areas that
have not aligned with the functions for the reporting of spending.
Recategorising the tax expenditures, and providing estimates for the
succeeding year, are among the leading tasks for the 2010 reporting and
process reform.
Policy making

Korea’s budget process places no special budgetary constraints on the
enactment of tax expenditures (but see below). Non-binding provisions of
the NFA specify that bills that will entail spending or tax reductions shall be
accompanied by a report that includes estimates of changes in revenues and
spending, for the five fiscal years beginning with the year of enactment, to
offset such changes. However, there is no mechanism in the law to require
that such compensating action be taken, or to provide remedies if it is not.
Another provision in the NFA, which became effective in 2007, imposed a
five-year PAYGO restriction only on tax expenditures. It is too early to
judge the success of that provision, but given international interest in
restraining tax expenditures, it bears watching.

process
Korea imposes a non-binding, five-year medium-term fiscal plan, which
specifies annual total spending ceilings and sub-ceilings. The ceilings are set
annually and usually expressed as percentages of GDP, with adjustment for
the economic cycle. They are enforced for the annual appropriations through
a top-down budgeting process.
Korea has two more noteworthy constraints on tax expenditures. In
principle, since the enactment of a law in 1976, tax expenditures are subject
to a five-year sunset, and must be re-enacted to continue in effect. It is
unclear how rigorously this requirement has been enforced, and how highly

it is regarded, because the number of tax expenditures in effect now, and the
number that have been proposed in recent years, are both perceived to be
very high. However, over the years 2002-07, the number of tax expenditures
declined in three of the five years of observed change, from 269 in 2002 (the
highest number for the six years) to 219 in 2007 (the lowest). In 2005, 2006
and 2007, seven, five, and 38 tax expenditures expired respectively. Such
reductions have not been common in the major OECD member countries,
and so this favourable trend may indicate success for the use of a sunset
requirement for tax expenditures.14
Another 2007 provision of the NFA requires that annual increases of
total tax expenditures be limited such that the ratio of tax expenditures to the
sum of tax expenditures and tax revenues grow by no more than 0.5% of the
average of the previous three years. This provision is certainly aggressive. It
raises the issues of the accuracy of the measurement of tax expenditures, the
problems of summing tax expenditures, fluctuations of the amounts of tax
expenditures with the economic cycle, and the non-comparability of tax
expenditures with outlays (the “might-have-bean” problem). This is one
more experiment in Korea that will be watched closely.
Policy review

In addition to the current effort to upgrade reporting on tax expenditures
by 2011, in 1999 the MOFE began to report on tax expenditures to the
National Assembly, based on a procedure prescribed by the Special Tax
Treatment Control Act of 1965.

The above cited reports, in addition to the five-year sunset requirement
and the new reporting requirements for 2011, suggest that the review of tax
expenditures may be more rigorous than the review of mandatory
(entitlement) spending programmes in Korea. Since the fiscal year 2005
budget, the budget office has required that the annual budget requests of all
spending ministries include projections for the upcoming four years for all
spending programmes (appropriations as well as mandatory programmes).
However, there is little evidence that those spending projections have any
meaningful effect. And while tax expenditures are at least in theory subject
to five-year sunsets, which have in fact shown some apparent success, there
are no significant sunset requirements on mandatory spending programmes.


projections
The budget has not explained differences between previous forecasts or
budgets and actual outcomes, for tax receipts or outlays. In fact, with no
projections of future tax expenditures, such a reconciliation would be
impossible.

Korea has recently enacted a “make work pay” tax expenditure.
Eligibility assessments began in 2008, with payments to begin in 2009. It
was anticipated that the programme would be administered exclusively by
the tax authorities. It would utilise an annual accounting process for the
individual taxpayer, as is common with respect to income taxes. Payments
are likely to come only from the government, without any apparent structure
for advance payment from employers.
Income taxes
With the very low number of reclassifications in this report for purposes
of international comparability, Korea had 136 tax expenditures under its
income taxes in 2006, the latest year with final or nearly final data. Korea
reported 143 such tax expenditures in 2007 (see Table 12, with
reclassifications by author).15 The greatest number of tax expenditures in
each year was in the general business incentives category, with the second
largest number in specific industry relief.
Other taxes
Korea reported 82 tax expenditures under other taxes in 2006, and 81 in
2007. The two largest contributing taxes were the VAT, with 26 tax
expenditures in each year, and the securities transaction tax, with 17 in each
year.

Memorandum items
For purposes of international comparability, this report reclassified two
tax expenditures as “structural items,” because they provided comparatively
broad tax relief for a purpose not practically distinguishable from measuring
the ability to pay tax.
Data reported by Korea

Communication from Korea presented numbers of tax expenditures over
time, which are presented here to provide a sense of trend. However, please
note that these counts do not correspond to those undertaken for this
report.16
2002 2003 2004 2005 2006 2007
Number of tax expenditures 269 254 220 226 230 219
New 13 9 27
Expired 7 5 38
Although information is not complete, Korea’s own count of the number

of tax expenditures shows a significant decline from 2002 to 2007. While
the number has shown considerable fluctuation, the decline from 2005
through 2007, and particularly in 2007, appears to owe a great deal to
expirations because of the requirement of a sunset clause imposed by the
Special Tax Provision Limitation Act of 1998.
Income taxes
With the minimal reclassifications described above, this report finds
Korea’s tax expenditures under the income taxes to sum to an amount equal
to 1.76% of GDP in 2006 (see Table 9, with reclassifications by author). The
largest category of tax expenditures is general business incentives, adding to
more than one-third of the total. Tax expenditures for health, the second-
largest category, are less than half as large.
Other taxes
Tax expenditures under other taxes sum to 0.72% of GDP. Almost two-
thirds of that total comes under the VAT.

Data reported by Korea

Communication from Korea again can provide a sense of trend, but
again without matching the numbers computed for this volume.
These figures indicate that Korea’s tax expenditures grew more rapidly
than tax revenues from 2002 through 2005, and then reversed course, with
most of the reduction coming in 2007. This perspective is of practical
importance because of Korea’s statutory limit on the tax expenditure rate as
calculated above, which was effective in 2007.
Billions of Korean won 2002 2003 2004 2005 2006 2007

Tax expenditures 14 726 17 528 18 286 20 017 21 338 22 708
Increase (%) 7.3 18.9 4.4 9.5 6.6 6.4
Tax revenues 103 968 114 664 117 796 127 466 138 044 158 334
Increase (%) 8.5 10.3 2.7 8.2 8.3 14.7
Tax expenditure rate (% 12.4 13.2 13.4 13.6 13.4 12.5

of revenues)
Tax expenditures in the Netherlands
Definition
The Netherlands considers deviations from its benchmark tax system
that reduce tax revenue to be tax expenditures (van den Ende, Haberham,
and den Boogert, 2004).

The Netherlands defines and measures tax expenditures at the national
level under all of its taxes other than its social security premium for
employers and employees.

The Netherlands defines its benchmark using the general rate structure
of its existing tax system – that is, a separate individual and corporate
income tax, with schedular treatment of wage and capital income under the
individual income tax, plus a value-added tax, a motor vehicle tax, and so

on. Although in principle this criterion is subject to interpretation, this

incorporation in the benchmark of different tax treatment of different types
of income likely leads in practice to a relatively smaller number of tax
expenditures identified in the Netherlands than would be the case using the
conventions of other countries. Other features of the tax system that are
incorporated in the benchmark are:
...the possibility of offsetting losses…The fixed rate of imputed
income for owner-occupied housing…and for savings and
investments…The federal tax credit…Exemptions, deductions, and tax
credits that adjust taxable income in line with the ability-to-pay
principle. In general, those provisions relate to personal circumstances,
such as being a single parent, having children, having a disability, or
being ill. Provisions that enhance the efficiency of taxation, such as the
use of fixed amounts to avoid disputes between taxpayers and the
revenue service (Van den Ende, Haberham, and den Boogert, 2004).
Because the benchmark specifies that tax provisions that measure
“ability to pay” are considered structural, and therefore are not tax
expenditures, some provisions that might be deemed to be tax expenditures
in other countries are not so considered in the Netherlands. These criteria
extend to consideration as part of the benchmark the tax advantages for
pension premiums (that is, contributions paid by employer and employee to
the pension funds), mortgage interest, and a tax credit for workers (which
was created to relieve the costs of earning wage income, but has since been
increased to the extent that it now far exceeds that goal, and might be
considered an incentive to work as well).
Concepts
The Netherlands uses the revenue forgone method to measure its tax
Estimates are for annual cash flows, rather than present values of longer run
or steady-state effects. Each tax expenditure is evaluated independently, and
so there is no attempt to capture interaction effects among any combinations
of tax expenditures; this means that any sum of tax expenditures does not
accurately reflect the combined impact of all of the relevant provisions. Tax
expenditures are also estimated independently of effects on government
spending programmes, and of any possible changes in other government tax
or spending policies that are made because of the tax expenditures.
Therefore, the amount of a tax expenditure is not a precise estimate of the
budgetary effect of its repeal.

Methods
Tax expenditures are estimated by the Ministry of Finance. Although
some tax expenditures are based on hard data, the ministry considers some
of the other data that are needed for the estimating process to be of lesser
quality and timeliness.
Reporting
Tax expenditure estimates are presented in the Tax Plan and Budget
Memorandum, which is a part of the budget, but separate from estimates of
outlay programmes with the same purpose as the tax expenditures.

The Tax Plan and Budget Memorandum is presented every year. It
provides tax expenditure figures for the budget year, one prior year, and the
five following years. The tax expenditures reported for the year prior to the
budget year are the final figures for that year. Changes in individual tax
expenditures – repeals, new provisions, increases and decreases – are
presented.
Policy making

Budget enactment in the Netherlands is based on a Coalition Agreement,
formed at the start of a new government, and covering four years. The
Coalition Agreement sets amounts for spending and revenues in currency,
and thus serves as a non-binding spending cap and revenue floor, creating an
informal “pay-as-you-go” system. The revenue floor relates to policy
changes rather than to changes in receipts caused by macroeconomic
fluctuations; thus, the automatic stabilisers in the tax system are allowed to
work, but both structural tax cuts and the creation of new tax expenditures
are restrained. (The operation of the automatic stabilisers in the budget
could, however, trigger a violation of the European Union’s Stability and
Growth Pact.) Although the Coalition Agreement does not have the force of
law, it has been respected and thus has acquired its own moral force.


process
Although the Coalition Agreement does not target tax expenditures
specifically, the revenue component of the Agreement would induce the
consideration of repeals or reductions of existing tax expenditures to finance
any new structural tax cuts or tax expenditures.
Five tax expenditures have their own annual caps. These tax
expenditures include credits for environmental and energy-saving
investments. If applications for those credits reach the annual limit, use of
the credits is closed until the beginning of the next fiscal year.
Policy review

In 2004, the Netherlands began a programme of evaluations of tax
expenditures, with the goal of reviewing each tax expenditure approximately
every five years. Responsibility is held jointly between the Ministry of
Finance and the pertinent spending department. The purpose of the
evaluation is to estimate the effectiveness and efficiency of the tax
expenditure. Questions that are specified for the evaluations to answer
include: Does the tax expenditure accomplish its objective? Can the same
goals be achieved with lower costs through a different policy instrument? Is
the tax expenditure the logical instrument to achieve these objectives? Is the
tax expenditure really the cause of any perceived effect, or would the same
outcomes have occurred without the tax expenditure? This evaluation
programme is fully underway, and evaluations have already been produced.

The budget includes estimates of annual costs of comparable spending
programmes.

projections
There is no ex post analysis of causes of changes in revenues from the
planned or projected levels. Such changes are defined as “endogenous” as a
general rule.


The Netherlands has no non-wastable “make work pay” tax credit. The
most widely held view is that such a programme should be considered a
spending programme. The Netherlands does have a deduction for employees
to compensate for the costs associated with working; this provision started
as a deduction of actual, documented costs, but then was converted to a
fixed amount (although employees could document and deduct higher actual
costs) in the interests of simplicity. This fixed amount has been increased so
that it would serve as a stronger incentive, and will be changed to be
income-related in the future.
Income taxes
In 2006, the latest year based on final or near-final tax data, the
Netherlands reported 55 tax expenditures under the income taxes (see
Table 16, as reported by country). Of those, 16 are categorised by this report
as providing specific industry relief, and 13 as providing general business
incentives.17 For 2007 and 2008, the Netherlands reports 53 and 52 income
tax expenditures, respectively; the decline results from reductions in the two
biggest categories for general and focused business tax incentives. As noted
earlier, there is no non-wastable make work pay tax incentive among these
tax expenditures. For purposes of cross-country comparability, this report
would reclassify one income tax expenditure in each year as a “structural
item,” on the ground that it provides comparatively broad and general tax
relief (see Table 16, with reclassifications by author).
Other taxes
The Netherlands reports 46 tax expenditures under non-income taxes in
each year 2006-2008. Of those, 17 apply under the VAT, and narrower
excises taxes and a tax on the sale of immovable property combined account
for 13 in each year.
Total
The Netherlands reports 101 tax expenditures in 2006, 99 in 2007, and
98 in 2008. This is down from 118 tax expenditures reported in 2001, and
123 in 2002.

Data reported by the Netherlands

Communication from the Netherlands reported data for several years
that provide a sense of trend, although these data do not match precisely the
calculations for this report.18
The data from the Netherlands confirm the reduction in the number of
tax expenditures in the early years of this decade, at least in terms of
government proposals. In 2003, 15 tax expenditures were abolished (and
five others were cut; a new government in 2003 used savings from repealed
and reduced tax expenditures to finance structural tax-rate reductions). In
2004, six tax expenditures were proposed for abolition, against two new
propositions, for a net reduction of four. In 2005, two proposed repeals
counterbalanced two proposed new provisions. In 2006 and 2007, however,
two and one new tax expenditures, respectively, were proposed.
Income taxes
In 2006, with the reclassification for cross-country comparability, this
report finds a total of income tax expenditures of a sum equal to 1.1% of
GDP (see Table 13, as reported by country). Almost half of that total is
categorised among general business incentives; less than one-fifth, the
second-largest category, falls in specific industry relief. Projections through
2012 show the total falling gradually to less than 1.0% of GDP.
Other taxes
Tax expenditures under taxes other than the income taxes sum to 0.9%
of GDP in 2006. About three-fourths of that total falls within the VAT. By
2012, this total is projected to decline to about 0.8% of GDP.
Total
The sum of all tax expenditures for 2006 is about 2.0% of GDP. The
total is projected to decline to less than 1.8% of GDP in 2012. Earlier data
showed tax expenditures under the direct taxes of 1.8% and 1.9% of GDP,
respectively, in 2001 and 2002, and under the indirect taxes of 1.0% and
1.1%, respectively, in the same years. Thus, the current level of tax
expenditures in the Netherlands is a reduction of about one-third from the
beginning of this decade.

Data reported by the Netherlands

Communication from the Netherlands reported proposed reductions of
tax expenditures in 2003 and 2004, mirroring the proposals for reductions in
the numbers of tax expenditures in the same years. From 2005 through 2007,
proposals would increase tax expenditures, by a cumulative amount equal to
about one-sixth of the combined reductions of 2003-2004. The sum of the
estimates by the government fell from 2.7% of GDP in 2003 to 2.2% of
GDP in 2006, and 2.1% of GDP in 2007.
Tax expenditures in Spain
Definition
Tax expenditures are not defined by law or regulation in Spain. For
purposes of the annual Budget on Tax Expenditures mandated by the
Spanish Constitution, tax expenditures are taken to be provisions of the tax
system that reduce tax revenues for the general government and meet other
conditions, the three most important of which are:
• a tax expenditure is an intended departure from the basic tax structure
(or “benchmark”);
• a tax expenditure is intended to attain some economic and social policy
goal; and
• a tax expenditure provides support only to a certain segment of the tax
population or to certain economic sectors, not to the population
generally.
As in other OECD member countries, the borderline between tax
expenditures and other basic elements of the tax system in Spain is complex.
Therefore, designation of tax expenditures is to some degree subjective. A
main reason is that lawmakers sometimes do not state explicitly whether a
tax provision seeks to attain economic and social goals, or rather to improve
the operations or the management of the tax system, which in the latter case
would not be considered a tax expenditure. The designation of tax
expenditures is made more complex by inclusion of many different tax
concepts such as incentives, reductions, allowances, deductions, reduced tax
rates, and exemptions in tax regulations and in the General Tax Law.


In Spain, the annual Budget on Tax Expenditures includes only central
government taxes, and those taxes not entirely ceded to the comunidades
autónomas (regional governments); tax expenditures are not measured or
defined for those taxes effectively administered by the comunidades
autónomas or by local governments. The remaining taxes for which tax
expenditure estimates are made include:
• income taxes: personal and corporate income taxes for both resident and
non-resident taxpayers;
• VAT;
• excise duties;
• tax on insurance premiums; and
• central government fees (revenues included in the central government
budget).
Tax expenditures are estimated for only two manufacturing excise
duties: the tax on hydrocarbons and the tax on alcohol and derivative drinks.
Duties such as the tax on beer, the tax on wine, the tax on the sale of
intermediate products, the tax on electricity, the tax on tobacco products and
all other minor excise duties are not included in the Budget on Tax
Expenditures, because of their relatively low or null taxation, complete
cession of tax yield to the comunidades autónomas, or the lack of reliable
and detailed information to carry out tax expenditure estimates.
Economically significant taxes directly administered by the
comunidades autonónomas, and therefore excluded from the Budget on Tax
Expenditures, include the tax on capital transfers and documented legal acts
(stamp duty), the tax on inheritance and gifts, and the general indirect tax
(Canary Islands). The central government Budget on Tax Expenditures also
excludes the important tax on real estate, which is levied at the local level.
Tax expenditures had been estimated for the wealth tax, but only for the
so-called “real obligation to pay the tax,” that is, for non-resident taxpayers
in Spain, given that the main collection proceeds are totally ceded to the
comunidades autónomas. Nonetheless, tax expenditure estimates for this tax
were removed from the 2009 Budget because of a government decision the
1st January 2008.
Several comunidades autónomas produce tax expenditure reports on
their own taxes. The local governments neither publish reports nor elaborate
on the Budget on Tax Expenditures with respect to their own taxes.


As was noted at the outset, tax expenditures are implicitly defined as
intended departures from the basic tax structure. There is neither a legal
definition of the basic structure of taxation (benchmark) in Spain, nor a list
of all of the components of the tax system, although documentation of the
benchmark is under discussion given its importance for a precise and
objective listing of tax expenditures.
Nevertheless, in practice, the basic tax structure (benchmark) is taken to
be the most permanent structure of taxes. It would include those tax
provisions that are fundamental to the determination of tax liabilities,
reaching a large majority of taxpayers, facilitating the management and
collection of taxes, and preventing double taxation among the different taxes
of the Spanish tax system.
From this standpoint, the basic structure of the personal income tax
includes the following elements:
• the current dual system of taxation, which differentiates between:
i) work and property income (the latter including rental income and
income arising from active business activities) which are subject to a
progressive tax schedule, and ii) the so-called saving income tax base
(capital gains, dividend, interest and insurance income) which is subject
to a flat tax rate (18% in 2008);
• the exemption for dividends received from Spanish companies (up to a
ceiling of EUR 1 500);
• tax deductions to prevent double international taxation;
• personal and family allowances, including allowances for dependent
children, parents and grandparents, and for disability; and
• withholding taxes (for work and capital income) and advance tax
payments (made by individual entrepreneurs and professionals).
The corporate income tax benchmark includes elements such as:
• the general statutory tax rate;
• internal and international double taxation tax deductions;
• amortization tables; and
• advance tax payments (instalments made by companies).

The VAT and excise tax benchmark includes

• general tax rates;
• export exemptions;
• EU intra-community exempt deliveries and other minor deliveries
(international organisations, diplomatic, etc.); and
• some exempt operations for technical and simplification reasons or to avoid
double taxation (insurance operations, land transmissions, second and
subsequent transmission of buildings, some gambling, the VAT special
scheme, etc.).
Concepts
Spain uses the revenue forgone (initial revenue loss) method to measure
its tax expenditures in the annual central government Budget on Tax
Expenditures. This method assumes no changes in taxpayer behaviour or
economic activity in response to the tax expenditures. Tax expenditures are
reported according to the cash accounting method.
Tax expenditure estimates measure annual cash flows at current prices,
rather than a discounted present value of future flows, or a longer-run steady
state. Personal and corporate income tax expenditures are measured one-by-
one, but also in combination for the personal income tax and for the
corporate income tax. The combined estimates eliminate any strictly
numerical, but not behavioural, interactions (for example, the numerical
effect of multiple tax expenditures on effective marginal tax rates, as
opposed to the way taxpayers change what they do because of the tax
expenditures) among different tax incentives for that one tax, so that the
figures add up to one consistent total for that tax.
Tax expenditures for the other taxes are measured independently, with
no attempt to capture interaction effects among any combination of tax
expenditures. Therefore, any sum of tax expenditures under non-income
taxes does not accurately reflect their combined impact, either with respect
to purely computational interactions, or any behavioural effects within the
economy upon production, employment, consumption, and savings, and
their effects in turn and in the longer run on revenues for the treasury. Tax
expenditure assessment is also independent of any potential changes in any
other government tax or spending policies because of the tax expenditures.
Therefore, tax expenditure amounts are not precise estimates of the
budgetary effect if they were repealed. Furthermore, as noted earlier, the
Budget on Tax Expenditures provides estimates only for provisions deemed

to be tax expenditures, not for tax provisions that are deemed to be part of
the basic structure of the tax system.
The inclusion of a particular tax provision in the Budget on Tax
Expenditures is subject to the availability of sufficiently reliable tax and
economic data to allow estimating with acceptable accuracy. Provisions are
included only if they are currently in force, and if a previous analysis
indicates that they meet the definition stated above. Provisions are removed
from the report if they are repealed or are scheduled by law to expire. In any
case, provisions are designated as tax expenditures independently of the
time that has passed from their inception.
Finally, there is no attempt to evaluate so-called negative tax
expenditures that generate revenue increases. Also, provisions such as
advance tax payments, instalments or compensation of negative tax
liabilities from previous years are excluded from the Budget on Tax
Expenditures.
Methods
Spain’s measurement methodology is thoroughly documented in the
Annual Report that has accompanied the Budget on Tax Expenditures since
1996. Several estimation methods are used. The preferred methods use tax
data and micro-simulation models based on taxpayers’ information from
their annual tax returns. Specifically, micro-simulation models for the
personal and corporate income tax use individualised tax data (the whole tax
population, not a sample), extrapolating tax variables such as taxpayers’
income, corporate turnover, population and tax expenditures, to project the
available information (with a gap of two years) to the year to be estimated in
the Budget on Tax Expenditures. Tax expenditures for other taxes such the
tax on non-resident income and central government fees are estimated with
information from administrative records, and other economic and tax
sources.
Excise duties tax expenditures are estimated using univariate time series
methods applied to monthly tax data. VAT estimates come from National
Accounts data sources, which after necessary adjustments yield VAT
collection forecasts, considering VAT assessment peculiarities, special VAT
schemes, and tax fraud. The tax on insurance premiums estimates are based
mainly on information from the insurance industry.

Reporting
The final tax expenditure amount is included in the annual Budget,
which is part of the annual general government Budget Law. Detailed
information about covered taxes, tax expenditure concepts, estimation
methods, information sources, and the like appear in the Annual Report on
Tax Expenditures produced by the Ministry of Economy and Finance, which
is available to the general public, attached to the Budget Law, and sent to
Parliament. A summary of the most important items in the Budget on Tax
Expenditures is presented in the so-called “Yellow Book” (a summary report
of the annual Budget Law).

In accord with a mandate in the Spanish Constitution of 1978, and later
in the General Budget Law, the Budget on Tax Expenditures has been
released every year since 1979. Since 1996, as noted above, there is also a
legal obligation to present an explanatory memorandum (the Annual Report
on Tax Expenditures). This report includes a complete list of tax
expenditures and associated tax regulations, as well as changes from the
previous tax year. Each Annual Report compares tax expenditure amounts
between the current and the preceding year, classified by the type of tax, the
type of tax expenditure, and the public budget purpose. There is no
comprehensive tax expenditure time series.
Policy making

There is no legal restriction on adding or expanding tax expenditures in
the budget process. Also, there is no legal limit on the total amount of tax
expenditures included in the annual budget. Thus, the inclusion of a new tax
expenditure does not necessarily require changes or reductions in grants or
public expenditure programmes. However, the General Stability Budget
Law sets procedures and ceilings for both public expenditure programmes
and public deficit growth figures. The government establishes annual goals
for spending and deficits according to the macroeconomic multiannual
framework for the annual Budget Law.

Finally, the introduction of a new tax provision, whether it is considered

a tax expenditure or not, must be supported by a legal provision (normally a
law) proposed by the government, and therefore is subject to discussion in
Parliament (except in emergency cases for which the legal instrument
normally used is a Royal Decree-Law, which later must be confirmed by
Congress). The government is required to present an economic report
including cost estimates (revenue forgone) of the proposed tax measures,
which often are included in the Budget on Tax Expenditures, at least during
the year of introduction (until later information confirms or modifies the
initial estimates).

process
In Spain, there is no explicit rule in the budget process to require or
encourage repeal or reduction of existing tax expenditures.
There are no explicit “sunset” or expiration dates for tax expenditures
except in some minor cases, such as for extraordinary tax provisions to
address natural disasters or adverse economic conditions (e.g. the recent fuel
or raw materials price increases). In such instances, the government may
adopt temporary tax measures (such as tax rebates) for particular economic
sectors (e.g. and agriculture), which may later be extended for the next year.
Other exceptions have been tax expenditures supporting the celebration of
certain cultural and sport events (e.g. America’s Cup, Expo Zaragoza 2008,
IV Centennial Quixote, etc.), which generally have had lifetimes of a
maximum of three years.
Policy review

Beyond the release of the annual Budget on Tax Expenditures, its
inclusion in the annual Budget Law, and its later delivery to the Parliament
(according to the Spanish Constitution before 1 October every year), there is
no legal requirement of further review of tax expenditures. As noted above,
only a few small tax expenditures are time-limited. There is no requirement
for later quantitative review of estimates, although the Ministry of Economy
and Finance does carry out informal internal ex post reviews to identify any
deviations from the initial estimates, and to analyse potential improvements
in estimation for the future.


Unlike for tax expenditures, Spain has very strict controls for public
expenditure programmes. Final data from both the expenditure and the
revenue sides of the budget are deeply analysed in a report published each
year by the Ministry of Economy and Finance. This report reflects only the
final data on direct public expenditure programmes, not those of an indirect
nature, or tax expenditures.

projections
Each October, when the government sends Parliament a new Budget
Bill for the coming year, the Ministry of Economy and Finance provides
advance estimates of revenues for the current year in comparison with
amounts initially budgeted for the most important taxes, with some
categorical detail. Likewise, the government provides information about
deviations between ex post and initial revenue estimates, including possible
effects from changes in the law and the economy, or from changes in
taxpayers’ behaviour as a consequence of prior policy changes.
Similarly, the new Budget Bill provides information on the effects of
new tax measures in the budget document, especially on new tax incentives
(whether considered as tax expenditures or not) and tax reforms.
Finally, the Budget on Tax Expenditures provides detailed information
on tax expenditures in comparison with previous budgets. It may also
provide some insight about changes in estimation methodology that might
affect the budget, and about possible estimation errors in the previous year’s
forecast.

Spain has several tax provisions to stimulate labour market participation
and support workers through the personal income tax, which therefore could
be classified as “make work pay” tax expenditures. These provisions address
mainly low-income workers and working women with children.
The “work-related income allowance” is a deduction based on income
from labour that provides its greatest relief to low-wage workers. In 2009, it
granted a maximum annual allowance of EUR 4 080 for wage-income
earners whose yearly wage did not exceed EUR 9 180. The allowance is
phased down as the taxpayer’s salary increases up to EUR 13 260, where the
allowance reaches its lower ceiling of EUR 2 652 – thus, the allowance is
not phased down to zero for any taxpayer. In case of disabled workers, the

amounts above are increased by EUR 3 264 and EUR 7 242, respectively,
according to the worker’s degree of disability. Personal income taxpayers
extending their labour market participation beyond retirement age (65
years), and unemployed workers accepting a job in a different location, may
increase their prior work-related income allowance amounts by 100%. This
tax expenditure has been utilised in the Spanish personal income tax, with
modifications, since 1978. The 2009 Budget on Tax Expenditures shows
EUR 8.4 billion in revenue forgone, which is 30.4% of tax expenditures
within the personal income tax, and 14.1% of all tax expenditures. This tax
expenditure was claimed by 19.7 million working taxpayers.
A second “make work pay” tax expenditure is the so-called “personal
income tax maternity tax credit”, a non-wastable tax credit for working
women with children under three years of age (or older in case of adoption)
of EUR 1 200 each year. This tax credit, introduced in 2003, was intended to
raise the low fertility rates in Spain (among the lowest in OECD member
countries), and to increase women’s participation in the labour market (one
of the lowest participation rates among OECD member countries), by
providing direct economic support to women wanting to reconcile working
and family life. This tax expenditure was expected to reach about 1.1 million
working women in 2009, and the revenue forgone estimate was projected to
rise to EUR 0.9 billion, according to the figures included in the 2009 Tax
Expenditure Report.
Income taxes
In 2008 (the latest year for which final data are available), with the
recategorisation performed in this volume for purposes of cross-country
comparability, Spain has 75 tax expenditures under the income tax (see
Table 20, with reclassifications by author). Of those, 24 provide general
business incentives, and ten provide specific industry relief. There are five
make work pay tax expenditures.
Other taxes
Spain has 64 tax expenditures in its taxes other than the income tax. Of
those, 48 are under the VAT, with small numbers distributed among five
other taxes.

Total
In total, Spain has 139 tax expenditures in 2008, with 149 identified as
of 2009. For purposes of cross-country comparability, we have reclassified
two of Spain’s identified tax expenditures for 2008 (and three for 2009) as
structural memorandum items rather than tax expenditures.
Income taxes
Spain’s tax expenditures under the income tax total to 2.3% of GDP (see
Table 17, with reclassifications by author). The largest individual category is
the “make work pay” tax expenditures, which at 0.7% accounts for about
one-third of the total. The next largest categories are the tax expenditures for
housing and health, at 0.5% and 0.4% respectively.
Other taxes
Tax expenditures under taxes other than the income tax equal 2.2% of
GDP, only slightly less than those under the income tax. Tax expenditures
under the VAT nearly exhaust this total.
Total
All tax expenditures sum to 4.6% of GDP.
Tax expenditures in Sweden
Definition
Sweden uses an informal definition of tax expenditures as provisions
that reduce revenue relative to a pre-defined norm, either to pursue a
specific policy objective or to facilitate the efficient operation of the tax
system.


The income tax (for individual income from labour and capital, and for
income from business), social security contributions, value-added tax, excise
duties (but only those on energy and carbon dioxide), and tax credits and
surtaxes are analysed to identify tax expenditures. Tax expenditures are not
measured at the local government (municipal) level, but tax expenditures at
the central government level may affect tax revenues at the local
government level. For example, income tax for individuals in Sweden is a
municipal tax, in that the individual pays a municipal income tax, and then,
for income over a certain threshold, the individual also pays the national
income tax. There are, however, some payments which the central
government has decided not to tax at all (e.g. for donation of blood). This is
considered a tax expenditure at the central government level according to
the norm, but it affects the revenues of both local governments and the
central government.

Sweden’s guideline for its benchmark or norm is a document dating
from about ten years ago. Sweden allows for different norms for different
types of taxes, all of which are based on uniform taxation. One norm applies
to income tax (for income from both capital and labour), and follows the
Haig-Simons comprehensive definition of income. However, within that
norm for income tax, Sweden accepts different tax rates for different tax
bases (that is, different sources of income). Hence Sweden’s different tax
rates for capital income and labour income are considered compatible with
the norm, and the different rates are not considered tax expenditures.
Sweden allows for structural tax provisions to be considered as a part of the
norm, and hence not to be tax expenditures. Thus, Sweden has an earned
income tax credit that is not considered a tax expenditure.
In addition to the norm for the income tax, there is one special norm for
social security contributions, one for value-added tax, and one for excise
duties.
Concepts
Sweden uses the revenue forgone method to measure its tax

expenditures are estimated independently of effects on government spending
programmes, and of any possible changes in other government tax or
spending policies that are made because of the tax expenditures. Therefore,
the amount of a tax expenditure is not a precise estimate of the budgetary
effect of its repeal. Sweden recognises and measures negative tax
expenditures (“tax penalties”) as well. Sweden also provides estimates of its
tax expenditures on an outlay equivalent basis (both net and gross).
Methods
There are some concerns about the absence of data in some areas. There
is a general sentiment that tax expenditure estimates can be of a lesser
quality than spending estimates.
Reporting
Data are presented in the Spring Fiscal Policy Bill appendices, and in
the annual Budget Bills. The Budget Bills repeat the estimates from the prior
Spring Fiscal Policy Bills, and report public expenditure programmes within
each policy objective as well.

Tables including all defined tax expenditures are reported in the
appendices to the annual Spring Fiscal Policy Bills for three years (the
budget year, one year prior, and one succeeding year.) The annual Budget
Bills present figures for the budget year and one succeeding year.
Retrospective data for the period 1992-2008 are available through current
and prior Spring Fiscal Policy Bill appendices and Budget Bills.
Policy making

Sweden has articulated a goal of an annual structural surplus of 1% of
GDP, ultimately to reduce net debt to meet the burdens of an ageing
population. To that end, Sweden imposed a spending cap, defined in units of
currency, and extending over three years. In principle, with a surplus goal
and a spending cap, there is an implicit target for revenues. However, in past

years, there was no explicit restraint of any kind on the implementation of

tax expenditures, including the fact that there was no adjustment to the
expenditure ceiling when a tax expenditure was enacted. As a result, the
spending rule provided a strong incentive to create non-wastable tax credits
for purposes that should have been addressed with spending programmes. In
recent years, that problem was recognised, and those tax credits have been
repealed or re-enacted as spending programmes, in the interest of budgetary
transparency. Public attention, aided by the clarity and prominence of the
spending cap, is thought to have been important in this process.
Furthermore, the Swedish budget process has been reformed to correct
the past flaws going forward. The new Swedish fiscal policy framework
includes three elements. The first is a target of an average general
government surplus of at least 1% of GDP over the course of a business
cycle. Second, to support this target, there is a multi-year expenditure ceiling
for the central government, set three years in advance, which makes it more
difficult to increase spending in good economic times when revenues are
strong. In a significant departure from the prior system, new tax credits that
work as spending programmes lead to an adjustment to the expenditure
ceiling. And third, there is a balanced budget requirement for the local
government sector.

process
The principles of the Swedish budget process require that any tax cut
must be financed through a spending cut, a revenue increase, or the use of a
projected surplus above the surplus target. This last option provides an
opportunity for expansion of tax expenditures that is not available on the
spending side of the budget, because with the spending cap, a spending
increase cannot be financed through a projected surplus above the surplus
target.
Policy review

The primary aim of reporting tax expenditures is to make those indirect
subsidies in the form of tax expenditures on the income side of the budget
more visible. However, tax expenditures are not integrated into the budget
process, and there is no formal evaluation for tax expenditures in the budget
process. There are very few examples of tax credits and tax reductions
which have been enacted for limited time periods through sunset provisions.


For mandatory spending programmes, future cost estimates are usually
reported for the period from year t-1 to year t+3 (thus for two additional
projected years beyond the estimates for tax expenditures). The reports are
presented twice each year: in the Budget Bill (specifying every
appropriation) and in the Spring Fiscal Policy Bill (on a more aggregated
level). Furthermore, all expenditures are specified in detail in the Yearly
Annual Report for the Central Government. Every year there are time series
published for the central government balance, revenues and expenditures
going back ten years. The latest was from 1995 to 2006, in a publication
called Tidsserier statsbudgeten. Sunset clauses for mandatory spending (as
for tax law provisions) are not common in the Swedish context, and there is
no explicit integration of consideration of taxation and spending in the
budget process. Greater and more transparent integrated consideration of tax
expenditures and spending programmes in the budget is a goal of future
reporting, although spending estimates are considered to be more reliable
than tax expenditure estimates, sound data for tax expenditure estimates can
be difficult to find, and the tax benchmark is considered by some to be
outdated.

projections
All legislative changes to the tax system are taken into account in the
budget projections. The revenue effects of changed tax rules are reported
separately. The budget reports the difference between projected estimates
and actual revenues, and explains the deviations from the original
projections in as transparent a manner as possible, although it may be
difficult to differentiate among estimating errors, macroeconomic
fluctuations and changes in behaviour.

Sweden has a so-called in-work tax credit, which was implemented in
2007. Because most benefits in Sweden (pensions or sick leave, for
example) are taxed, the government, in order to make work pay,
implemented this provision to tax income from labour differently from
incomes emanating from benefit systems. Before this change these two
types of income were taxed in a uniform way. However, this provision is not
considered a tax expenditure. The largest tax expenditures aimed at making
work pay include a tax reduction for household services, deductions for
commuting costs and deductions for double housing expenses due to work at

geographical locations too far from home for commuting. These provisions
do not follow the model of the non-wastable tax credit.

This report includes no independent analysis of tax expenditure data
from Sweden. Communication from Sweden provides some information on
the amount of tax expenditures.19 Sweden reported that its 2007 tax
expenditures that fall within 60 specific defined policy areas sum to an
amount equal to 10% of total tax revenue. That figure is net of measured
negative tax expenditures (which might otherwise be called tax penalties).
Tax expenditures under the capital income tax sum to 36% of revenues, and
tax penalties under that tax are 15%; tax expenditures under excise taxes
come to 48%; and those under the VAT add up to 19%. All tax expenditures
sum to 12% of total tax revenue (again net of negative tax expenditures).
Using this measurement approach, tax expenditures under the capital income
tax add up to 61% of revenues (and tax penalties sum to 15%); tax
expenditures under the excise taxes sum to 49% of revenues; and those
under the VAT come to 22%. Total tax expenditures add up to 5.7% of
GDP; those tax expenditures that are directly comparable to spending
programmes come to 4.7% of GDP.
Tax expenditures in the United Kingdom
Definition
The United Kingdom divides tax reliefs into three categories. Those
reliefs that are alternatives to, and have consequences similar to, public
spending are referred to as tax expenditures. Those forms of tax relief that
are either an integral part of the tax structure or that simplify administration
or compliance are called structural reliefs. Structural reliefs include
measures such as the personal allowance and relief from double taxation of
dividends. Tax expenditures include measures such as the exemption of
capital gains on the sale of a principal residence and the exemption of the
first GBP 8 000 of reimbursed relocation packages provided by employers.
However, the government acknowledges that the distinction between
structural reliefs and tax expenditures is not always straightforward, and
includes a third category of tax reliefs, which consists of tax concessions
that combine elements of both the structural and expenditure categories. Into

this less well-defined category they put tax concessions such as age-related
allowances and the tax exemptions for child benefits and disability living
allowances.

Tax expenditures are measured only for the central government. Taxes
analysed include income tax, corporation tax, VAT, national insurance
contribution tax, capital gains tax, inheritance tax, petroleum revenue tax,
stamp duty land tax, and vehicle excise duties. In each tax category covered,
the United Kingdom reports “tax expenditures,” “reliefs with tax
expenditure and structural components,” and “structural reliefs.”
Benchmark tax system(s)

For analysis of the income tax, the United Kingdom identifies tax
expenditures for relief of tax on capital gains and corporate income, which
suggests a reference income tax system.
Concepts
The United Kingdom uses the revenue forgone method to measure its
tax expenditures; the estimates assume no changes in taxpayer behaviour or
economic activity as a result of the presence of the tax expenditure. Each tax
expenditure is evaluated independently, and so there is no attempt to capture
interaction effects among any combinations of tax expenditures; this means
that any sum of tax expenditures does not accurately reflect the combined
impact of all of the relevant provisions. Tax expenditures are estimated
independently also of effects on government spending programmes, and of
any possible changes in other government tax or spending policies that are
made because of the tax expenditures. Therefore, the amount of a tax
expenditure is not a precise estimate of the budgetary effect of its repeal.
The United Kingdom uses accrual accounting for their budget expenditures,
as well as for estimating the cost of tax expenditures.
Structural provisions are included in the “Tax Expenditure and
Structural Relief” report. As noted above, the government acknowledges
that the distinction between structural reliefs and tax expenditures is not
always clear.

Methods
The loss of revenue associated with tax reliefs and allowances cannot be
directly observed. Estimation methods include calculating the amount of tax
that individuals or firms would have had to pay if there were no exemptions
or deductions for certain categories of income or expenditure, and
comparing it with the actual amount of tax due.
Reporting
The data are reported in “Chapter A: Budget Policy Decisions” within
the government’s “Financial Statement and Budget Report” (HM Treasury,
2007a). More details on individual tax allowances and reliefs can be found
in Tax Ready Reckoner and Tax Reliefs, published alongside the pre-budget
report. Estimates are not presented directly alongside outlays for comparable
purposes.

Although there is no statutory requirement to produce a report on tax
expenditures, the government still estimates and reports all major tax
expenditures in the Tax Ready Reckoner every autumn. And Chapter A of
the annual Budget, “Budget and Policy Decisions of Financial Statement
and Budget Report,” contains a list of proposed tax expenditures. No
comprehensive historical report exists, but the “Financial Statement and
Budget Report” was first reported following approval of Parliament (for the
purposes of Section 5 of the European Communities Amendments Act) in
1993. The “Financial Statement and Budget Report” has been published
online since 1997.
Policy making

The government has committed not to take policy measures which are
likely to increase social security or other parts of the budget, including tax
expenditures, without offsetting steps to accommodate the government’s
“Golden Rule” and “Sustainable Investment Rule.” The Golden Rule states
that over the economic cycle, the government will borrow only to invest and
not to fund current spending. The Sustainable Investment Rule states that net

public debt as a proportion of GDP will be maintained below 40% over the
economic cycle. In addition, a section of the 1998 Finance Act requires that
HM Treasury lay before the House of Commons a Code for Fiscal Stability
(CFS). This code emphasises five principles for fiscal policy and requires
HM Treasury, on behalf of the government, to prepare reports outlining past
and prospective developments in fiscal and debt management, including
adherence with the government’s fiscal rules mentioned above.
The United Kingdom has a budget law that limits total debt. There is
also an explicit prudence factor built into the economic assumptions which
reduces the final economic estimates by a set amount. This is informally
done, not legally required. The United Kingdom does not follow a strict
PAYGO process. It does, however, as outlined above, try to follow the
“Golden Rule” and to obey the “Code for Fiscal Stability”.

process
If observed, the “Golden Rule” would provide an incentive to reduce or
eliminate existing tax expenditures. Unanticipated surpluses at year-end can
either be treated as a dividend to the budget carried over to next year, offset
against next year’s budget, or used as price reductions.
Policy review

Tax expenditures are reviewed twice a year by the HM Treasury as part
of the budget and pre-budget report process. This, however, is not a legal
requirement.

The United Kingdom’s budget contains spending that is mandatory in
nature. It is estimated that 40-60% of the budget is comprised of mandatory
spending. There are specific sunset dates for some of these mandatory
expenditure programmes, at which point spending ceases if they are not
renewed.


projections
The United Kingdom does report differences of actual revenues from
projected estimates. Forecasts are based on “cautious assumptions” (a trend
economic growth rate of 0.25% lower than the government’s view) audited
on a three-year rolling basis. An alternative scenario where trend growth is
1% lower than the central case is published to illustrate the risks. Fiscal rules
are assessed in both the central and cautious cases.

In the United Kingdom, due to high marginal effective tax rates, the net
income result from a small increase in gross earnings could be only slightly
positive or possibly even negative. This is particularly true of single parents
and one-earner families. This factor led to early interest in “make work pay”
tax expenditures.
The Married Couple’s Allowance (MCA), whose estimated cost
(GBP 4 600 million in the 1992-93 period) made it the sixth largest tax
expenditure, was abolished in April 2000. The revenue thereby saved was
allocated to funding the Children’s Tax Credit, which replaced MCA and
came into effect at the same time. The Children’s Tax Credit was a wastable
tax credit available to families with one or more children.
In April 2003, two new credits – the Child Tax Credit (CTC) and the
Working Tax Credit (WTC) – replaced the Working Families Tax Credit
(WFTC), the Disabled Person’s Tax Credit (DPTC) and the Children’s Tax
Credit. The CTC replaced the existing, income-related elements of support
for children in WFTC, DPTC, the Children’s Tax Credit, income support
and the income-based Jobseeker’s Allowance. Also, the CTC is non-
wastable, so that people paying no tax could receive the support. Similarly,
the WTC replaced existing elements of support for adults and their child-
care costs in the WFTC, DPTC, and the New Deal Employment Credit for
those aged 50 or older. The WTC also will provide support for working
households without children where at least one adult is aged 25 or over.
The government has enriched these tax credits almost every year since
the introduction of the WFTC. Although their estimated costs are modest –
GBP 3 300 million and GBP 1 100 million, respectively, in 2005 – those
costs do not include the non-wastable payments that exceed liability. Such
payments are currently treated as expenditures and amount to as much as
GBP 15 billion in 2004-05.

HM Revenue and Customs, formed in 2005, pay and administer all tax
credits, including the Working Tax Credit (WTC). The United Kingdom
also subsidises employers by reducing employers’ social security
contributions. These subsidies are given to all those with low earnings,
without any attempt to identify and treat differently sub-groups of the
population. The United Kingdom also subsidises employees, paying cash
transfers to those who separately apply for the benefit and meet the
qualifying conditions.
Income taxes
In 2006-2007, the last measurement period using final or near-final data,
the United Kingdom reported 189 tax expenditures under the income tax,
and 151 under other taxes.20 It also reported 42 “reliefs with tax expenditure
and structural components, and eight “structural reliefs” (see Table 24, as
reported by country).
On the basis of reporting practices observed in the rest of the sample of
OECD member countries, this volume judgmentally classifies 39 of the 42
“reliefs with tax expenditure and structural components,” and three of the
“structural reliefs,” as tax expenditures. The other three of the “reliefs with
tax expenditure and structural components,” and five of the “structural
reliefs,” are classified as “structural items.” Thus, for purposes of greater
cross-country comparability in this volume, the United Kingdom has the
same number of 208 tax expenditures under the income taxes in 2006-2007
as under the United Kingdom’s own count – but they are not the same 208
provisions; three are different (see Table 24, with reclassifications by
authro).21 The United Kingdom also reported 208 tax expenditures for 2007-
2008. The category of general business incentives, with 38 tax expenditures,
was the largest; there were 37 tax expenditures for work-related employee
benefits (other than retirement and health), and 29 for specific industry
relief. These tax expenditures include a non-wastable “make work pay” tax
credit.
Other taxes
The United Kingdom reported 173 tax expenditures for 2006-2007
under all taxes other than the income taxes. In 2007-2008, the number
increased to 175. The VAT was subject to 43 tax expenditures in 2006-2007,
and 44 in 2007-2008. The inheritance tax had 44 tax expenditures in both
years. The stamp duty land tax had 22 tax expenditures in 2006-2007, and
23 in 2007-2008.

Memorandum items
Like Canada, the United Kingdom enumerates and estimates the cost of
some provisions that it considers to be structural, and therefore not tax
expenditures. Uniquely, the United Kingdom also classifies a group of
provisions that it believes have attributes of both structural provisions and
tax expenditures. After categorising these provisions for what is judged to be
the greatest possible cross-country comparability, we have classified eight as
structural in both 2006-2007 and 2007-2008 – five from the United
Kingdom’s own list of eight structural provisions, and three from the United
Kingdom’s list of 42 provisions with both structural and tax-expenditure
attributes.
Income taxes
With this volume’s recategorisation of tax expenditures under the
income taxes, the sum of these provisions equals 8.3% of GDP in 2006-2007
(see Table 21, with reclassifications by author). It was projected to fall to
8.1% of GDP in 2007-2008. Tax expenditures for retirement equal 2.3% of
GDP in 2006-2007; those for capital gains are 0.5% of GDP; those for
accelerated depreciation add to 1.4% of GDP, and dividends to 1.1%, and
those for housing come to 1.2% of GDP. The portion of the make work pay
tax credit that is counted as a reduction of receipts comes to 0.3% of GDP.
The numerous general business incentive tax expenditures add up to 0.8% of
GDP; those for work-related employee benefits sum to 0.2% of GDP; and
the specific industry provisions equal 0.1% of GDP.
In 2001-2002, tax expenditures under the income taxes, using the United
Kingdom’s own categorisation but including the “reliefs with tax
expenditure and structural components,” added to 8.2% of GDP, or not very
different from the current level. In 2002-2003, the corresponding sum was
8.6% of GDP, surely higher in part because of downward pressure on GDP
from the weak global economy at that time.
Other taxes
Tax expenditures under all taxes other than the income taxes sum to
4.5% of GDP in both 2006-2007 and 2007-2008. The VAT tax expenditures
add to 3.2% of GDP, and those under the inheritance tax come to 1.0% of
GDP, so these two taxes virtually exhaust the total.

In both 2001-2002 and 2002-2003, the sum of non-income-tax tax

expenditures was 2.1% of GDP, or less than half the current level.
Memorandum items
The sum of the few provisions we have categorised as structural items in
2006-2007 and 2007-2008 added up to 4.2% of GDP. These are provisions
that provide comparatively broad tax relief, and one (the personal allowance
under the income tax) is quite large. In 2001-2002 and 2002-2003,
provisions that the United Kingdom identified as structural added to 5.9%
and 5.8% of GDP, respectively.
Tax expenditures in the United States
Definition
The statutory definition of tax expenditures in the United States is
“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
liability.”

Tax expenditures in the United States are identified only for central
government taxes. They do include many items that benefit state and local
governments, including exemptions for interest earned on municipal bonds.
Tax expenditures have generally been limited to individual and corporate
income taxes. In principle they could be defined for other taxes as well, but
this has not been done except for a brief period in the 1990s when tax
expenditures were measured for estate and gift taxes.

In general, the tax expenditures in the United States budget are
deviations from a comprehensive income tax in which income is defined as
consumption plus the change in net worth – the Haig-Simons definition. The
statutory definition of tax expenditures given by the 1974 Budget Act does
not specify the precise reference tax system, and the choice of a baseline is

somewhat arbitrary. In recent years, the presentation of tax expenditures in

the budget has discussed the ambiguities in the tax expenditure concept,
pointing out how the list of tax expenditures would change if a pure income
tax or a pure consumption tax were used as a benchmark. Two benchmark
tax systems are used in the budget: the normal tax and the reference tax.
Both are patterned on a comprehensive income tax. The reference tax is
closer to existing tax law, limiting designation as tax expenditures to special
exemptions that serve programmatic functions. For that reason, there are
fewer tax expenditures under the reference tax, and it might be seen in very
broad terms by some observers as more objective, whereas the normal tax
system might be criticised by some as being more prescriptive.
Measurement
Concepts
The United States uses the revenue forgone method to measure its tax
expenditures are estimated independently of effects on government spending
programmes, and of any possible changes in other government tax or
spending policies that are made because of the tax expenditures. Therefore,
the amount of a tax expenditure is not a precise estimate of the budgetary
effect of its repeal. There are structural provisions of the income tax that are
not considered to be tax expenditures. These include personal exemptions,
the standard deduction, and the graduated tax rates for the individual income
tax (the graduated rates in the corporate tax are considered a tax
expenditure). Also, income is generally considered taxable only when it is
realised in exchange. Tax rates vary by marital status, and these variations
are not considered to be tax expenditures. Taxes on purely nominal gains
resulting from inflationary changes in asset values or the effects of higher
expected inflation on interest rates are not regarded as negative tax
expenditures. The corporate income tax is not regarded as a tax penalty on
income, even though that income is also taxed at the individual level. Only
estimates for provisions deemed to be tax expenditures are presented; no
“memorandum items” are presented for provisions that are judged to be
structural parts of the benchmark tax.

Methods
The tax expenditure estimates are an analytical construction prepared in
the Department of Treasury’s Office of Tax Analysis (OTA). This is the
same group that prepares the revenue estimates for the budget and which
analyses the revenue impact of proposed changes in tax laws. Some of the
tax expenditure estimates rely on the same large sample of tax returns that is
used in preparing revenue estimates. This sample is extrapolated for the year
of the budget and the years that follow using the administration’s
macroeconomic forecast and technical assumptions by OTA. Unlike the
revenue estimates, however, the tax expenditure estimates are based on the
previous mid-year forecast, to save time in making the necessary
calculations. Another difference is that the tax expenditure estimates assume
no change in behaviour as a result of varying the tax law; the revenue
estimates generally do allow for microeconomic behavioural changes. The
estimates are defined as “revenue losses” from the tax expenditure
provision. A prior practice of identifying the outlay equivalents of these
revenue loss estimates has been discontinued. Present-value estimates also
are presented for selected tax expenditures where current revenue losses
potentially give a misleading impression of the net impact of the tax
provision. No attempt is made using actual tax return data to verify the
accuracy of the estimates made earlier for the same fiscal year.
Reporting
Tax expenditures are presented in the annual budget, but in a section of
a budget annex volume (called Analytical Perspectives) that is devoted to
revenue issues. Prior to the FY 1990 Budget, they were issued separately in
a volume accompanying the budget called Special Analyses. The estimates
for particular tax expenditures are thus separate from the figures for
spending programmes directed toward similar purposes.

Tax expenditures were presented for the first time in the FY 1976
Budget issued in 1975. Since the late 1970s, the tax expenditure tables show
seven years of estimates: two years prior to the year of the budget, the year
of the budget, and the four years following the year of the budget. The
estimates are currently based on the economic forecast used for the mid-year
estimates of the budget and they are not retrospectively revised or updated.
Each year’s budget includes listings of all new tax provisions enacted in the

preceding year, but no separate listing of new tax expenditures. When new
tax expenditures are enacted, they are included in the annual presentation,
but only a close comparison of the current presentation with that in previous
budgets would reveal which of the listed provisions are new. The tables do
not identify the new tax expenditures. In the past, the budget chapter that
presents all of the revenue proposals also listed the new tax expenditures,
but that practice has been discontinued, because it was difficult to include
the proposed tax expenditures in a timely manner. These proposals were
often determined at the end of the budget process, making it difficult to
prepare estimates for them before the budget was scheduled to print.
Policy making

The most recent US statutory budget disciplines, which were enacted in
1990 and extended in 1993 and 1997, expired at the end of 2002. Those
disciplines included “pay-as-you-go” rules, which required that any tax
reduction (including both new and expanded tax expenditures, and any
structural tax cuts; a new or increased mandatory spending programme
would be subject to the same disciplines) must be fully offset (by repeal or
reduction of an existing tax expenditure or tax expenditures, a structural tax
increase, and/or repeal or reduction of an existing mandatory spending
programme or programmes; an unanticipated budgetary “windfall” of either
higher revenues or lower outlays would not qualify as an offset for these
purposes). The pay-as-you-go process was enforced in law by an automatic
across-the-board reduction (called “sequestration”) in a specified subset of
mandatory spending programmes. This pay-as-you-go rule was fully
observed over a substantial part of its history, and has been considered
responsible for some part of the improvement in the United States budget
during the 1990s. It was, however, waived several times in the legislative
process in 2001 and 2002, immediately before it was allowed to expire.
Separate multi-year disciplinary pay-as-you-go rules (without the force of
law) on all tax and mandatory spending legislation are now imposed in the
House and in the Senate, although these have been relaxed for some
measures. These past and current pay-as-you-go rules are in the nature of
“new spending rules,” which is to say that there is no fixed target for total
spending or revenue amounts, and automatic stabilisers are allowed to
function unimpeded. The past administration proposed modified pay-as-you-
go rules that would be limited to new spending without applying to taxes;
the new administration has proposed a general renewal of the pay-as-you-go
rule. The House and Senate pay-as-you-go rules, if enforced, would not

allow current unanticipated revenues to be spent in the future (or require that
current unanticipated revenue shortfalls be made up in the future).

process
As noted above, the expired pay-as-you-go law, and the current pay-as-
you-go rules, would prohibit the expansion of an existing tax expenditure
without an offset through the repeal or reduction of an existing tax
expenditure, a structural tax provision, or a mandatory spending programme.
The pay-as-you-go process also creates an incentive for the repeal or
reduction of an existing tax expenditure as a possible offset for any proposed
new or expanded mandatory spending programme.
Policy review

There is no required review of existing tax expenditures. However,
many tax provisions (including tax expenditures and structural provisions) –
many more than was the case eight years ago – now have sunset dates, and
will expire in the next few years (many at the end of 2010). This will require
some measure of “reconsideration,” if not “review.” There has been a
biannual volume of analyses of tax expenditures produced by the
governmental but non-partisan Congressional Research Service of the
Library of Congress; that review does not reflect the views of either
executive or legislative policy makers. Also, the governmental but non-
partisan Congressional Budget Office produces a biannual volume of
potential policy changes to reduce the deficit; the ideas considered
inevitably included some reductions or repeals of existing tax expenditures.
Tax expenditures receive considerable attention whenever tax reform is on
the political agenda. In 2005, the President’s Advisory Panel on Federal Tax
Reform issued a report calling for the comprehensive overhaul of the tax
system, which would have drastically altered many of the largest tax
expenditures. This effort at tax reform did not lead to legislation, but the
central place of tax expenditures in the reform options is typical of what a
general tax reform would produce. In the FY 2008 budget, and again in the
FY 2009 budget, the President proposed major changes in the tax
expenditure for private health insurance.


The budget (like the Congressional Budget Office’s comparable annual
report) includes multi-year projections of the costs of entitlement
programmes. The projections cover the same time frame as the other budget
projections: the year of the budget and four years beyond. Social Security
and Medicare both are reviewed annually by their trustees who issue annual
reports which present 75-year projections for these programmes. The
trustees’ projections are usually based on different assumptions from those
used for the budget. For more than a decade, the budget has included a
“Stewardship” chapter in Analytical Perspectives which reports long-run
projections for the budget as a whole including the major entitlement
programmes. It is arguable whether such analyses make legislative review
and revision more likely. Some mandatory spending programmes – but not
the very largest – are subject to periodic expiration, which requires
legislative reauthorisation and, in theory, attendant review in the Congress.
It should be noted that such review has not always resulted in the past, nor
has there always been careful and timely review of the other mandatory
spending programmes, which are in permanent law. There is no readily
accessible comprehensive listing of the required reauthorisations of
mandatory spending programmes.

projections
Both the budget and reports by the Congressional Budget Office identify
three classes of causes of deviations of revenues (and outlays) from
projected estimates after the fact: legislative action, changes in the economy,
and “technical” factors (which can include specific issues regarding
programme details, but often denote simple estimating errors). These
assessments are not updated after they are first made. There are no such
retrospective re-estimates that pertain specifically to tax expenditures.

The US Earned Income Tax Credit (EITC) was one of the first, if not the
first, non-wastable tax credits designed to “make work pay”. It was enacted
in 1974. It is managed solely by the revenue authorities, and typically is paid
to beneficiaries annually by the central government tax authorities. (More
frequent distribution is permitted, but is rarely done, largely because of
perceived complexity for employers.) Amounts of the credit that offset tax
liability (the smaller part) are counted as reductions of revenues; amounts in
excess of tax liability are counted as increased outlays. This programme was

structured as a tax programme rather than an outlay programme because of

an apparent preference to administer it through the existing tax system rather
than through a new spending bureaucracy.
Income taxes
With recategorisations to achieve somewhat greater cross-country
comparability, we have counted 164 tax expenditures under the US income
tax in 2008 (the latest year based on final or near-final tax data; see
Table 28, with reclassifications by author).22 That is an increase from 135 in
2002. The US Treasury projects that two tax expenditures will drop in cost
to zero by 2010. In 2008, there were 54 tax expenditures, more than one-
third of the total, in the category of specific industry relief. We have
categorised 18 as general business incentives, and 16 for education.
The United States communicated historical counts of tax expenditures
that are not directly comparable with the calculations in this volume, but that
give a longer sense of trend.23 In 1985, there were 104 tax expenditures.
That number was reduced somewhat by the Tax Reform Act of 1986, but by
1990 the number was back up to 116. Under the methodology used in the
US communication, there were 130 tax expenditures in 2000 and 161 in
2006, so the growth since the 1986 Act has been continuous and significant.
Other taxes
As noted earlier, the United States does not identify tax expenditures for
any taxes other than income taxes.
Income taxes
With the recategorisation for cross-country comparability, US tax
expenditures under the income tax sum to 6.0% of GDP in 2008 (see
Table 25, with reclassifications by author). That is down from 7.0% in 2002
(although because 2002 was a recession year, GDP was cyclically low, and
tax expenditure estimates were cyclically affected). Health, housing and
retirement purposes each individually account for more than 1.0% of GDP.

Relief for income from capital through accelerated depreciation, capital

gains, interest and dividends accounts for 0.7% of GDP (and tax cuts for
capital gains and dividend income enacted in 2001 and 2003 are not counted
as tax expenditures by the current United States administration). This is a
decline from 0.8% in 2007, and 0.9% in 2006. In the coming years, this
amount is projected to fluctuate wildly. As was noted earlier, the United
States does not in concept recognise negative tax expenditures. However,
the tax expenditure for accelerated depreciation includes both a positive
component in the early years of new investments (when depreciation
deductions are larger than what is deemed to be true economic depreciation)
and a negative component in the later years of those investments (when
depreciation deductions are correspondingly smaller than the neutral
amount). With the current economic downturn, new investment has slowed
sufficiently that the accelerated depreciation deductions on new purchases
are estimated to be smaller than the reduced later deductions on prior
investments. Accordingly, the measured tax expenditure for accelerated
depreciation declined substantially for 2008, and is projected to be negative
for 2009 through 2012. The swing from the larger prior positive estimates to
the projected negative estimates will reduce the total by as much as 0.5% of
GDP. At the same time, with the weakness in financial markets, the tax
expenditure for capital gains has dropped by another 0.3% of GDP. Even
with an assumed economic recovery, the total of tax expenditures for relief
from capital income taxation will barely return to half its 2007 level by
2014.
The numerous provisions for specific industry relief add to only 0.2% of
GDP, general business incentives for 0.3% of GDP, and education for 0.1%.
Because the accounting for the “make work pay” non-wastable tax credit,
the Earned Income Tax Credit (EITC), includes only the portion that offsets
tax liability as a reduction of revenues, the impact of that provision on the
total is only about 0.1% of GDP. Furthermore, because the provision dates
back almost to the dawn of the tax expenditure concept – it was enacted in
1974 – it will have little impact on any measure of the growth, as well as the
level, of tax expenditures. Total income tax expenditures are projected to
grow to 6.8% of GDP by 2013.
Data communicated by the United States show tax expenditures
declining unevenly as a percentage of GDP over a longer term – from 8.8%
of GDP in 1985 to 6.0% in 1990 (after the Tax Reform Act of 1986), but
then rising to 6.4% in 2000, before declining slightly to 6.2% in 2006. These
figures suggest that the reduction of marginal income tax rates in 1986 has
had a long-term effect of holding down the revenue impact of tax
expenditures. The data might also suggest that new tax expenditures added
since 1986 have been somewhat smaller on average than those that were

repealed in the 1986 tax reform. That hypothesis would need to be verified
carefully, but it is true that the tax cuts in this decade have included very
large structural rate reductions, rather than being dominated by new or
expanded tax expenditures.

Notes
1. See Section 3.2.

2. It might be argued that ability-to-pay tax provisions are different from
most spending programmes, in that the former apply to persons with
income from labour or capital, and the latter often apply to those with no
other means of support. Even that distinction clearly is imperfect,
however. All of these issues might suggest that it would be helpful to
identify the revenue costs of ability-to-pay reliefs, even if they were not
formally categorised as tax expenditures. Many countries are far from that
degree of detail in reporting, and such a practice would add to workloads.
3. For one specific example, the United States provides a personal
exemption for all persons including children, which is considered
structural and not a tax expenditure; and an additional tax credit for
children, which is considered a tax expenditure. It certainly is true that the
population of children is smaller than the population of all persons, and so
the tax credit for children might be considered a tax benefit to a narrow
population; but the population of all children still would seem quite broad.
If the benefit were delivered through an additional personal exemption for
children, or a larger personal exemption for children than for older
persons, this categorisation might be even more ambiguous.
4. Present-value estimates, in addition to the annual cash-flow estimates, are
presented for the tax-deferred savings programmes.
5. See the chapters entitled “Fiscal Projections” or “Fiscal Outlook.” The
most recent projections are in Chapter 3 of the 2008 Economic Statement:
www.fin.gc.ca/ec2008/pdf/EconomicStatement2008_Eng.pdf.
6. The most recent projections are in Chapter 3 of the 2008 Economic
Statement, www.fin.gc.ca/ec2008/pdf/EconomicStatement2008_Eng.pdf.
7. For further details on the WITB, see Department of Finance (2007a).
8. All counts and categorisations are by the author. See earlier discussion for
limitations.

9. Communication from Georges-Henri Lion to the OECD conference on

tax expenditures, 10-11 December 2007. All responsibility for
interpretation rests with the current author.
10. For the statutory pension system, expenditure, revenue and the federal
grants to the system are estimated by an independent working group. The
group meets quarterly and their estimates cover the short- and medium-
term. The annual Statutory Pension Report (Rentenversicherungsbericht)
covers revenues, expenditures, sustainability, and the expected
contribution rates for the next 15 years. The Public Service Pension
Report (Versorgungsbericht) is published every four years. It presents
data on expenditures for the previous year and estimates for the next 15
years and analyses the main determinants of expenditures. The most
recent report, published in 2005, includes estimates through 2050.
limitations.
12. These totals attempt to represent central government tax expenditures
only.
13. Communication from Takao Shiraishi to the OECD conference on tax
expenditures, 10-11 December 2007. All responsibility for interpretation
rests with the current author.
14. The effectiveness of a sunset requirement must be judged carefully. In the
United States, a small number of tax expenditures had been subject to
periodic expirations over the 1980s and 1990s, but they were
continuously re-enacted, albeit on occasion after some legislative drama.
Political analysts went so far as to suggest that those sunsets actually had
the effect of increasing the number and size of tax expenditures and
structural tax cuts, because the periodic expiration of those provisions
created repeated occasions for consideration of tax bills when there was
no other pressing motivation. Thus, other countries might want to
consider the Korean experience as evidence of the potential success of
mandatory sunsets, but also whether their own legislative environments
are conducive to such favourable results, more than the apparently less
attractive US experience.
limitations.
16. Communication from John Kim to the OECD conference on tax

limitations.
18. Communication from Wilhelmus van Tol to the OECD conference on tax
19. Communication from Ragnar Olofsson to the OECD conference on tax
20. For purposes of this analysis, tax expenditures under the income tax
include those that apply to national insurance contributions.
limitations.
limitations.
23. Communication from Robert B.Anderson to the OECD conference on tax

Bibliography
Assemblée nationale (1958), Constitution of October 4, 1958, La

Documentation française, www.assemblee-nationale.fr/english/8ab.asp,
Paris.
Department of Finance (2007a), Budget 2007, A Stronger, Safer, Better
Canada, Government of Canada, Ottawa.
Department of Finance (2007b), “Corporate Income Taxes and Investment:
Evidence from the 2001-2004 Rate Reductions,” in Department of
Finance, Tax Expenditures and Evaluations, Government of Canada,
Ottawa, pp. 41-56.
Department of Finance (2007c), “Fiscal Projections”, in Strong Leadership.
A Better Canada, Economic Statement 2007, Government of Canada,
Ottawa, Chapter 2, www.fin.gc.ca/ec2007/ec/ecc2e.html.
Department of Finance (2007d), Tax Expenditures and Evaluations,
Government of Canada, Ottawa.
pp. 134-136.
Grady, Patrick and Richard W. Phidd (1992), “Budget Envelopes, Policy
Making and Accountability,” prepared for the Government and
Competitiveness Reference, Economic Council of Canada, Discussion
Paper 93-16, http://global-economics.ca/budgetenvelopes.pdf, pp. 259-
268.
HM Treasury (2007a), “Financial Statement and Budget Report”, The
Stationary Office, London, and available at www.hm-
treasury.gov.uk/media/D/0/bud07_chaptera_235.pdf.
HM Treasury (2007b), Tax Ready Reckoner and Tax Reliefs, HM Treasury,
London.

Ministère de l’Economie, des Finances et de l’Industrie (2007), Évaluation

des Voies et Moyens, Tome II, Dépenses Fiscales, annex to the Loi des
finances, République française, Paris, pp. 11-12.
Poddar, Satya (1988), “Integration of Tax Expenditures into the Expenditure
Management System: The Canadian Experience”, in Neil Bruce (ed.),
Tax Expenditures and Government Policy, John Deutsch Institute for the
Study of Economic Policy, Kingston, pp. 259-268.
Receiver General of Canada (2007), Public Accounts of Canada, Vol. 1,
Section 1, Government of Canada, Ottawa, www.tpsgc-
pwgsc.gc.ca/recgen/txt/72-eng.html.
Seguin, Marc, and Simon Burr (2004), “Federal Tax Expenditures in
Canada,” in Polackova Brixi, Valenduc, and Swift, Tax Expenditures –
Shedding Light on Government Spending through the Tax System, The
World Bank, Washington DC, pp. 99-100.

5. CONCLUSIONS – 147
Chapter 5
Conclusions
This chapter draws conclusions of the comparison of the ten countries under review.
It discusses the differences between countries in defining tax expenditures and concludes
that there are some commonalities. It then turns to the differences between countries
concerning what types of taxes are measured. It continues by explaining the differences
in benchmark tax systems each country uses in order to define their tax expenditures. It
includes an analysis of the different concepts, methods, reporting, policy making and
review, make work pay programmes, and the number and amount of tax expenditures,
among the nine different countries studied in this report.

148 – 5. CONCLUSIONS
There are several global conclusions from the analysis of tax

expenditures and policy-making institutions. One is certainly that
international comparisons of tax expenditures are difficult, and careful
judgment must be employed. Some useful conclusions may be drawn, but
they should not over-reach the comparability of the data. Still further
research beyond this report would be helpful; at the extreme, it would
involve analysis of the various countries’ tax systems in details far deeper
than the lists of tax expenditures themselves. A second, more immediately
practical conclusion is that there are no policy rules or procedures that
guarantee success in controlling tax expenditures, but that some ideas now
in use among the sampled countries may help. The following discussion
expands on these themes.
Definition
Definitions of tax expenditures share a common core across countries,
but differ widely around that core.
Every country defines “tax expenditures,” either explicitly or implicitly
and sometimes using a minor variation on the title, as exceptions to some
baseline standard for the entire tax. For most countries one criterion to
identify a deviation from the baseline is a loss of revenue, although a few
countries formally recognise “negative tax expenditures,” or “tax penalties.”
Some countries establish a criterion as providing a benefit to a narrow group
of taxpayers. Some look for pursuit of a policy objective other than the core
goals of the tax system itself. Some look for some measure of similarity to a
hypothetical alternative outlay programme toward the same end. However,
all of these criteria seem close enough that they should cause identified tax
expenditures to diverge little from one country to another.
However, countries differ in the purpose of the exercise of the
identification of tax expenditures, and these differences can cause what may
be enormous divergences in the meaning of the tax expenditure statistics.
Most of these differences are somewhere in the nexus between the definition
of tax expenditures and the specification of the benchmark tax system
against which possible tax expenditures are tested. Discussion related to the
benchmark will continue below.
However, as one example of the differences among countries, Germany
explicitly focuses its interest in tax expenditures on their use to deliver
subsidies to business, whether organised in the corporate sector or in

households. Accordingly, Germany identifies and measures tax expenditures

under its household income tax, but they are generally related to business
activity. Germany’s measured number of tax expenditures under its
individual income tax would be expected to be low for this reason, but there
could be other provisions that would be considered tax expenditures but for
the measurement focus on business subsidies. Determining whether there are
such unidentified subsidies delivered through the tax system would require
extensive analysis not of Germany’s list of tax expenditures, but rather of
the rest of its tax system – which would be a considerable exercise.1

Many countries extend their analyses of tax expenditures to central
government taxes beyond income tax. Some, depending on the nature of
their federal fiscal systems, measure tax expenditures for taxes received at
lower levels of government as well. Of course, countries that identify tax
expenditures of non-central governments may list more and greater amounts
of tax expenditures than countries that consider the federal level only. The
identified tax expenditures may serve different governmental functions as
well, given the typical division of labour between central and lower levels of
government, under which central governments deliver larger subsidies and
services keyed to national objectives such as economic growth, while lower
levels of government more often pursue social objectives, and often have
smaller taxes whose tax expenditures might be correspondingly smaller. In
the end, it is essential to recall that a country that does not measure tax
expenditures outside of its income tax may well have unmeasured tax
expenditures outside of its income tax; and a country that does not measure
tax expenditures at the provincial or local levels of government may well
have unmeasured tax expenditures at those levels (if those governments
have taxing powers).
Benchmark tax system(s)

The selection of a benchmark tax system has important leverage on the
outcome of measurements of tax expenditures. Some countries have very
elaborately specified benchmarks, while others have only implicit
definitions of tax expenditures from which their benchmark systems are
inferred. However, whether precise or impressionistic, the choice of a
benchmark is potentially important.
The literature pays considerable attention to the choice between an
income tax and a consumption tax as the benchmark for tax-expenditure
purposes under its income tax.2 In theory, of course, this choice is of

enormous importance. Under a comprehensive income tax benchmark, any

subsidy for income from capital is a tax expenditure. Under a consumption
tax benchmark, any taxation of income from capital is a negative tax
expenditure or a tax penalty.
Each country examined here uses what is basically an income tax as its
benchmark (or uses other criteria that lead to the listing of tax preferences
for income from capital as tax expenditures). And yet there are significant
differences in benchmarks that apparently lead to equally significant
differences in the measurement of tax expenditures, independent of the
nature of the underlying reference tax.
At the most fundamental level, each country has a choice of how general
or detailed its benchmark system will be. A country with a very general
benchmark could consider many provisions of the actual law to be tax
expenditures. In another country, a more elaborate benchmark might include
some of those same kinds of provisions, which therefore would not be
considered tax expenditures. Countries at opposite ends of this spectrum
could be Japan and the Netherlands. Japan expresses its benchmark only in
terms of its basic principles of taxation: equity, neutrality, and simplicity.
Any provision to pursue any other objective is considered a “Special Tax
Measure,” the term that Japan uses to denote a provision that other countries
might call a “tax expenditure.” From this benchmark, Japan might be
expected to have a large number of tax expenditures. In fact, Japan considers
its Special Tax Measures to be qualitatively different from the tax
expenditures recognised and measured in other countries. In sharp contrast,
the Netherlands considers its benchmark to be the “primary structure” of the
actual tax system in place. As a result, that benchmark might be expected to
include some provisions that might otherwise be considered tax
expenditures, and the Netherlands might be expected to have relatively few
recognised tax expenditures, all else equal.
France’s approach is a variation on the definitional end of the scale that
would tend to have fewer tax expenditures. In the past, France’s concept of
its benchmark or “norm” was assumed to evolve over time, and there was a
presumption that a tax provision that had been long-lived could for that
reason become accepted into the “norm.” As a result, some tax expenditures
over time could come to lose that status. It was only recently that France
ceased that practice.
Canada and the United Kingdom employ measurement concepts toward
the end of the spectrum that presents a more restrictive benchmark, and
which thus tends toward the identification of numerous tax expenditures.
Furthermore, in addition to provisions that it considers tax expenditures,
Canada provides estimates of other provisions that it considers to be

structural, and therefore not tax expenditures. Thus, Canada could be

expected to list estimates for many tax provisions, like Japan, but unlike
Japan it divides those estimates into two classes: those that are considered
tax expenditures, and those that are not. The United Kingdom goes one step
further, and separates a third class of provisions, which are identified as on
the borderline between being tax expenditures and being structural.
Thus, through its approach to the definition of tax expenditures and the
benchmark, a country can significantly affect the number and size of tax
expenditures that it identifies. Yet for all of its significance in terms of
measurement outcomes, this choice has much less economic salience than
the more commonly discussed decision between an income tax and a
consumption tax benchmark.
Concepts
As noted elsewhere in this report, there is a fundamental conceptual
choice among measurement by “initial revenue loss,” more commonly
known as “revenue forgone”, “final revenue loss”, and “outlay equivalence.”
In practice, only Sweden provides outlay equivalent measures, in their case
as a supplement to their primary presentation, rather than as the primary
presentation itself; the United States once did the same as Sweden, but has
discontinued the practice.
Thus, the main choice that each nation has confronted is between the
revenue forgone method, and the final revenue loss method. The revenue
forgone method is inevitably discussed as having several important
theoretical drawbacks. It does not take into account interactions among
different tax expenditures; it does not take into account behavioural changes
on the part of taxpayers because of the existence of the tax expenditure; it
does not take into account behavioural changes on the part of government
because of the tax expenditure, such as enacting or repealing other tax
expenditures or outlay programmes; it therefore does not provide an
accurate estimate of the revenue effect of the repeal of the tax expenditure,
and the amounts of multiple tax expenditures cannot be added to obtain an
accurate sum. In contrast, the final revenue loss method has none of these
flaws.
Every government examined here has chosen the revenue forgone
method, despite all its flaws. The reason almost surely is that the final
revenue loss method is totally impractical. The domain of governments is
almost exclusively the production of data, not contestable scientific
estimates. Given only basic assumptions, a revenue forgone estimate is
determinate. In contrast, a final revenue loss estimate is almost totally
subjective. A revenue forgone estimate does not depend on estimates of

predicted human behaviour; a final revenue loss estimate does. Thus, the
task of generating a final revenue loss estimate is infinitely more difficult
and more involved. Furthermore, the number of potential interactions among
tax provisions and other government programmes, when the number of
relevant tax provisions and programmes may rise into triple digits, is for all
practical purposes infinite. Governments cannot generate such estimates,
which would amount to multiple replications of the work of providing an
estimate for a complex tax bill, over any feasible production cycle.
Accordingly, no one should be surprised that governments choose the
revenue forgone method.
However, one implication of the choice of the revenue forgone
estimation method is usually ignored. Separate revenue forgone estimates of
multiple tax expenditures cannot be added together accurately. And yet,
virtually every government (and virtually every private researcher) adds tax
expenditure estimates routinely. The reason is that there is no alternative
way to compare the prevalence and size of tax expenditures over time and
across countries. This report falls into line and follows this questionable
procedure.
There is no way to judge even whether adding multiple tax expenditures
yields a sum that is too high or too low, let alone by what margin. As was
noted earlier, sums of tax expenditures for itemised deductions in the United
States likely overstate the true total; sums of tax expenditures for exclusions
of income likely understate the true total. Sums of a combination of the two
are unpredictable. And that is true even if the objective of the sum is
another, albeit combined, revenue forgone estimate. Once changes of
taxpayer and government behaviour are added into the exercise, the result is
even more uncertain.
The best course is to forewarn policy analysts about the problems of
sums of tax expenditures, and to use them as indicative, rather than as
precise indicators. With appropriate caution, small differences in sums can
be discounted, and large differences can be interpreted as qualitative signs.
Under some circumstances, such indications may prove useful, and may
direct further, more-focused research in important directions. Only Spain
has attempted to provide an estimate of the combined effect of all tax
expenditures; and even that effort would yield no guidance toward the
combined effect of any subset of tax expenditures.

Similarly, the number of tax expenditure provisions, rather than their

revenue forgone, may not be a definitive indicator (because of differences in
definition across countries) but may suggest avenues for more detailed
research. For example, if the number of a country’s tax expenditures has
increased over time, that may motivate a review of tax policy and the policy
process.
Some particular individual tax expenditures can raise their own
conceptual issues. Tax expenditures are typically, but not always, measured
as annual cash flows. In some instances, governments produce discounted
present-value estimates. Such estimates may be useful, although
governments are not necessarily indifferent between different streams of
costs with the same present value, and often will want to see the year-by-
year figures. Where policy choices must be made with respect to deferrals of
tax liability, like pension programmes where contributions and earnings are
tax deferred but later redemptions are taxable, even annual net tax
expenditure flows can be inadequate. Policy makers may want to see both
revenues collected from taxation of redemptions, and revenues postponed
from deferral of tax on contributions and earnings, for each year. Such detail
is not routinely reported in the public tax expenditure statements.3
Methods
Some countries have more mature tax expenditure reporting systems
than others, which means more publicly available and complete
documentation of methods. However, such documentation as there is
suggests that methods from one country to another are basically similar,
whether the tax expenditure in question is from income tax or some other
tax. Tax expenditures such that the current tax collection processes generate
the necessary data – such as a deduction from reportable income, which
must be claimed on a tax return – are estimated from those data, through
statistical samples of the tax returns themselves. Tax expenditures that do
not generate such data – such as an exclusion from reportable income, which
does not even appear on the tax return – are estimated from whatever data
can be found, whether from public survey statistics, or administrative data of
central, provincial or local government, or even data obtained from private
entities. Government agencies other than the tax policy or collection
authorities may well be involved. Where necessary, econometric estimates
are made using those data.

Reporting
It is widely believed that tax expenditures are best reported in the budget
alongside outlay programmes with the same purpose. However, in some
countries, even the reporting of tax expenditures is not required by law.
Some countries report biannually, rather than every year. Some report
outside of the budget, many annexed to the budget. None of the countries in
this sample report tax expenditures alongside similar outlay programmes;
Germany’s subsidy report may be the closest to such reporting. Depending
on the perspective of the observer, this may be taken as a reason why tax
expenditure policy is made so poorly, or a paradox given that tax
expenditure policy is made as well as it is. As was pointed out earlier, there
may not be realistic trade-offs between tax expenditures and outlay
programmes, particularly if there are firm political constraints against higher
taxes and higher spending. Still, going forward, it would seem desirable to
pursue the goal of side-by-side reporting of tax expenditures and similar
outlay programmes – even if there are no examples of that practice in OECD
member countries at this time.

To keep tax expenditures in the focus of policy makers, annual reporting
might be seen as an ideal, and in fact most countries follow an annual
schedule. Typically, reporting is limited to one, two or three years of
estimates. Canada, the Netherlands and the United States provide the longest
estimation windows, with Canada showing estimates for eight years
(including the budget and two succeeding years) and the Netherlands and
the United States for seven years (in the Netherlands, including the budget
year, one prior year, and five succeeding years, and in the United States,
including the budget year, two prior years, and four succeeding years). The
generally small number of years reported may be an indication of the
priority that is assigned to tax expenditures, or to the perceived complexity
of the task of generating the estimates, although once the estimating
infrastructure is created, there should be economies of scale in replicating
the estimates for multiple years, and in performing the estimates year after
year. No country has made a practice of maintaining a consistent historical
database of tax expenditure estimates, which makes trend analysis difficult
and time-consuming; however, changes of methodology and data limitations
may make such an historic compilation a low priority.

Policy making

Few countries have explicit limits on tax expenditures in their budget
processes, which leaves the widely discussed incentives to create tax
expenditures to act with full force. Sweden’s spending cap, with no
restriction on tax expenditures, pushed this incentive to the extreme; but that
policy arrangement has since been corrected with a spending cap adjustment
for non-wastable tax credits that are close substitutes for alternative
spending programmes. However, Sweden allows new tax expenditures – but
not spending programmes – to be “paid for” with unexpected increases in
the budget surplus. Still, Sweden’s overall budget process has proved highly
successful, at least as measured by a sustained favourable fiscal balance.
The EU Stability and Growth Pact, and the United State’s budget
process (which has expired in its strongest form), impose limits on the
budget deficit, and through it, on tax cuts and spending in general. Several
other countries (including the Netherlands, with its Coalition Agreement,
and Sweden) have non-binding restrictions on the deficit. Germany has its
non-binding directive to deliver business subsidies on the spending side of
the budget. Only Korea imposes limits on tax expenditures narrowly
defined, and those limits are so new as to be as yet untested. One such limit
is the pay-as-you-go restriction on tax expenditures. The other, the
restriction on the total of tax expenditures based on a moving average of
totals over the previous three years, raises questions about the imprecision of
sums of separate revenue forgone estimates, among other conceptual issues.
It remains to be seen whether Korea will attempt to enforce this limit
through more elaborate estimates of all of its tax expenditures taken
together, so as to avoid the problem of the inaccuracy of the sum of multiple
estimates. However, even if Korea does calculate a joint estimate, that will
leave the further problem of fluctuations of estimates based on the state of
the economy, and issues regarding measurement and taxpayer behaviour.
Even if tax expenditure policy does not change, tax expenditure amounts
under the progressive income tax will grow if the economy grows in real
terms. (Conversely, in a recession, tax expenditure amounts could go down,
thus rendering the limit not operative.) It may be that the intent of the
provision was to impose a downward pressure on tax expenditures based
solely on the growth of the economy. However, with such an instrument
based on an imperfect measure, there could well be unintended and
awkwardly timed consequences.

The Canadian envelope system was an interesting attempt to drive

policy makers to make the explicit choice between spending and tax
programmes. Canadian policy analysts were concerned that it was not
implemented systematically. Furthermore, elected officials might not be
indifferent between a budget with fewer tax expenditures, and thus an
apparently larger total revenue and larger spending, and an alternative
budget that actually commands the same amount of the economy’s resources
but has more tax expenditures and thus an apparently smaller government.
Canada’s line agencies reportedly attempted to trade their own tax
expenditures for implicit tax increases to allow more spending. However,
this experiment does at least suggest the potential merit of making line
agencies play a constructive role in choices between spending and
equivalent tax expenditure programmes, and perhaps even some workable
incentive to move any less efficient tax expenditures toward more
transparent and perhaps more frequently reviewed direct spending
programmes.
Incentives to repeal or reduce existing tax expenditures in the

budget process
Several countries provide encouraging experience, and perhaps some
ideas that might be emulated to improve budgetary control. Sweden’s
correction of its incentive to increase tax expenditures involved the
conversion of several non-wastable tax credits, which reportedly had been
chosen solely to take advantage of asymmetric budget process constraints,
into more-transparent spending programmes. The United State’s. pay-as-
you-go system allowed credit for the repeal of tax expenditures to finance
other tax cuts or spending increases. Korea has enacted a similar system.
The EU Stability and Growth Pact, Japan’s Basic Policies, and the United
Kingdom’s Golden Rule in theory provide incentives to reduce or eliminate
tax expenditures to achieve budget deficit targets.4 Korea’s moving-average
target to limit the growth of tax expenditures is too new to evaluate. It does
raise questions about the practicality of the use of sums of revenue-forgone
estimates of tax expenditures. The sunset requirements in Germany, Japan
and Korea show signs of success, but further research and evaluation, and
perhaps experience, are needed.

Policy review

Some countries have formal schedules for review of their tax
expenditures. Germany’s system has the interesting attribute of a mandated
role for multiple non-governmental research institutions, producing
competing evaluations. That process is underway, and bears watching for its
performance both for the quality of the reviews and for their impact within
government. The new five-year cycle of reviews in the Netherlands may
also show results. In contrast, France’s statutorily required evaluations thus
far have produced only quantitative estimates of tax expenditures, although
that process has been revised and apparently improved.
Other countries require evaluations without firm schedules, and have
shown limited results. The United States has a statutory requirement, but
results have been paltry. Greater review, although with limited political
impact, has come from the governmental but non-partisan institutions of the
Congressional Budget Office and the Congressional Research Service of the
Library of Congress. Korea will begin a new statutory process in a few
years. Japan undertakes reviews as its Special Tax Measures approach their
sunset dates; concrete results are unclear.
Canada has no requirement for evaluations, but has an apparently
successful track record of publication of reviews in some detail. Other
countries have no specific requirements and have demonstrated little in the
way of reviews.

It is worth remembering that budget analysts have long railed against
mandatory or entitlement spending programmes, the most likely substitutes
for many tax expenditures, as “uncontrollable” and unreviewed spending
(Schick, 2007). Discussion by OECD member countries does not suggest
any significantly more robust review of entitlement spending than of tax
expenditures. In particular, Korea’s newly established regimen of tax-
expenditure review, though not yet in effect, seems far more robust than that
country’s own review of mandatory spending programmes. The United
States requires periodic reauthorisation of some spending programmes, but
there is no consistent pattern of rigorous review as a result, and the largest
entitlement programmes are not subject to such reauthorisation. Except for
the United Kingdom, it appears that more tax expenditures than spending
programmes are subject to statutory sunset, which might provide some
greater occasion for review.

In other words, although replacing tax expenditures with spending

programmes might result in greater transparency, it is by no means a
guaranteed path to more regular and more rigorous review. The political
energy expended to convert tax expenditures to spending programmes
arguably may yield greater improvement if applied directly to obtaining
review.
Information about causes of changes in budget results relative to

past projections
The United State’s system of semi-annual calculation of an exhaustive
taxonomy of changes to prior budget projections – with changes divided
among effects of new legislation, changes in economic assumptions, and
“technical re-estimates” (usually unexplained forecasting errors, but
sometimes changes in regulations or other identifiable non-legislative, non-
economic factors) – is unique. Other countries provide much less
information, with only Canada, France, Germany and Sweden providing
separate figures for the effects of legislative changes. More specific
accounting could help in considering policy alternatives to tax expenditures.
Over a long historical time horizon, the enactment and expansion of

“make work pay” provisions has had an important cumulative effect on the
total amount of tax expenditures. Koiwa (2006) lists 11 OECD member
countries with make work pay tax expenditures and the budgeting
techniques they report to the OECD.5 However, the sample of countries in
this survey report only limited recent changes. Korea has enacted a
provision like the US Earned Income Tax Credit that will begin making
payments in 2009. The United Kingdom has modified pre-existing
provisions of this type.
Accounting treatment of non-wastable make work pay tax expenditures
varies from country to country. The United States and the United Kingdom
split their provisions between revenues (for the portion that offsets tax
liability) and outlays (for the non-wastable portion). France accounts for its
provision fully as revenues. Canada scores its provision as outlays. Germany
and Sweden do not consider their make work pay provisions to be tax
expenditures. Japan and the Netherlands do not have such provisions. Korea,
as noted above, has not yet begun making payments, and accounting
treatment is not yet determined.

If the make work pay function is truly necessary in contemporary

developed economies, however, then this role will be played on one side of
the budget ledger or the other. The issue is choosing the better of the two
more than a concern that the tax expenditure vehicle would lead to the
creation of such programmes.
Income taxes
The numbers of tax expenditures vary widely from country to country
(see Figure 8 and Table 32). As noted earlier, this could have much more to
do with measurement details as well as broad questions in the definition of
tax expenditures, and of the benchmark tax, than with the issues in the actual
tax laws and regulations of the various countries. The greatest number of tax
expenditures is in the United Kingdom. However, the listing of United
Kingdom’s tax expenditures shows considerable detail, to the extent that
what might be one tax expenditure under the methodologies of other
countries could be counted as several in the United Kingdom. Furthermore,
we have included as tax expenditures many of the provisions listed by the
United Kingdom as having attributes of both structural provisions and tax
expenditures. Other countries with similarly ambiguous provisions might
not have listed them at all.
At the other end of the scale, Germany and the Netherlands list
comparatively few tax expenditures. However, those results do seem to rest
on approach and methodology as well as substance. Germany’s focus on tax
expenditures that deliver subsidies to business could lead to the
identification of fewer provisions under the individual income tax that might
be called tax expenditures in other countries, though Germany has fewer
business-subsidy-oriented tax expenditures than many other countries. The
broad reference tax in the Netherlands may well encompass some provisions
that would be identified as tax expenditures in other countries. This pattern
includes the incorporation of the “three-box” system of taxation of income
from different countries, but also the inclusion in the benchmark of
numerous provisions designed to simplify tax administration.
In short, it would be a mistake to over-interpret these data as even an
imprecise measure of the prevalence of tax expenditures in the tax systems
of the various countries. It would be more useful to use them instead along
with the information about policy practices and institutions in the different
countries, and the trend information that has been collected, to consider
ways to improve policy making in every country.

One result that is to type with the political-economic view of tax

expenditures is the large number of provisions in several countries to
provide specific industry relief (see igure 9 and Table 32). Those provisions
do not comprise substantial revenue losses in any country. Still, part of the
cost of large numbers of narrowly targeted provisions may well be
accumulated administrative and compliance complexity and perceived
unfairness. In addition, although those tax expenditures may not be large,
they may allocate resources far more narrowly and specifically than would
be ideal.
Other taxes
Different countries have different reporting practices with respect to
non-income taxes. For those countries that do report broadly, it is clear that
non-income taxes are hardly devoid of tax expenditures. Value-added taxes
and inheritance taxes do in some instances have significant numbers and
sizes of tax expenditures. And yet again, comparisons from one country to
another are potentially vulnerable to benchmark issues, such as the
identification of provisions as structural measures of the ability to pay tax.
Generally speaking, those countries with the most tax expenditures
identified under income tax have the most tax expenditures identified under
the value-added tax and other taxes (see Figure 8). This finding could
indicate that countries that make the most use of tax expenditures do so
under all of their taxes. However, it could also suggest once more that some
countries are more meticulous than others in defining and cataloguing their
tax expenditures. Under this interpretation, comparisons of numbers of tax
expenditures may say more about measurement methodologies than about
the underlying tax systems, and so must be considered carefully.
Memorandum items
The presentation of estimates for structural measures in Canada and the
United Kingdom helps to present a comparable view of those countries
relative to others. However, because those provisions may not be considered
tax expenditures in a reasonable international comparison, they should not
be counted in a total number of tax expenditures without careful
consideration. The comparison in this report is based on a country-by-
country and provision-by-provision judgment of which tax features should
be counted as tax expenditures on the most consistent basis possible. Thus,
many of the memorandum items in Canada and the United Kingdom are not
counted.

Income taxes
The diversity in the data across countries carries forward from the
numbers of tax expenditures to their amounts as percentages of GDP, both
in absolute terms and in shares of the total within any given country. In
general, countries with more tax expenditures tend to have larger currency
totals of tax expenditures (see Figure 10 and Table 29).
Perhaps the most noteworthy departure from this general pattern is
Korea. Korea identifies a comparatively large number of tax expenditures
from income tax, but the sum of those tax expenditures is a comparatively
small percentage of GDP. Thus, while Korea may have fairly many tax
expenditures, and the number of tax expenditures carries its own cost of
complexity and economic distortion independent of money, the sizes of
those tax expenditures may be comparatively small, or the sum of those tax
expenditures may be particularly affected by some measurement and
methodology issues. One hypothesis could be that Korea provides a
disproportionately large share of its relief through tax expenditures under
taxes other than income taxes, but the data here do not seem to support that
explanation. Because measurement patterns across countries with respect to
non-income taxes do differ, this question would require further research.
This apparent correlation between the number of tax expenditures and
the sum of their costs generally but not precisely translates from tax
expenditures as percentages of GDP to tax expenditures as percentages of
income tax revenues (see Figure 11 and Table 31), with variations
depending on each country’s relative reliance on the income tax.
Looking at patterns of amounts of tax expenditures across countries,
there are some other impressions, if not conclusions. The sums of tax
expenditures in Canada, the United Kingdom and the United States are
significantly larger than those in Germany, Korea and the Netherlands.
Close comparisons are of course impossible because revenue-forgone
estimates do not account for behavioural reactions, and the sums of
individual tax expenditures do not equal the combined effects of change.
However, these large orders-of-magnitude differences do suggest significant
differences in measurement methodologies, institutions, or both.
For example, as noted earlier, Germany’s identification of tax
expenditures is focused on business subsidies, and one hypothesis would be
that Germany may not identify what other countries would call tax
expenditures to benefits households for social purposes. However, all of

Germany’s identified tax expenditures account for a significantly smaller

percentage of GDP that the tax expenditures for capital income tax relief,
general business incentives, and housing alone in Canada, the United
Kingdom, or the United States. Thus, the differences between Germany and
this group of countries likely are more than just that Germany does not
identify social tax expenditures. The differences may be methodological, or
they may be substantive. But these data do suggest that there are significant
differences, and that they could be worth further investigation.
Another question is the public purpose for which tax expenditures are
used. There are only a few sketchy patterns for tax expenditures under the
income tax (see Figure 12, Table 29, and Figures 1-7). Canada, the United
Kingdom and the United States devote comparatively large shares of their
tax expenditures toward retirement. Germany, the United Kingdom and the
United States have relatively large shares for housing. Canada, Germany,
the United Kingdom and the United States put meaningful shares of their
totals toward relief for income from capital (capital gains, interest,
dividends, or depreciation). But beyond that, it is hard to identify more than
a pair of countries that would suggest a degree of similarity that looks other
than purely random. As mentioned earlier, one pattern identifiable by its
absence is the low share of GDP devoted to specific industry benefits,
despite the large number of provisions. Thus, the many specific industry tax
expenditures have apparently low money costs while absorbing larger
volumes of law, regulation, and perhaps taxpayer attention – which can be
real costs to society.
This result as much as any is potentially sensitive to measurement
issues. Tax benefits for broad categories of income from capital are a prime
example. The measure of the tax benefit for depreciation in excess of true
economic depreciation is totally dependent upon the chosen measure of true
economic depreciation, which is a subject of competing volumes of
controversy and ignorance. Canada defines its provision to relieve double
taxation of corporate dividends as part of the benchmark, but reports its
revenue cost as a memorandum item. Because other countries do not
explicitly state their benchmarks in those terms, this report counts that
memorandum item as a tax expenditure. However, in 2003 the United States
enacted ad hoc tax relief for dividend income, and chose not to count it as a
tax expenditure. Without a listing as a memorandum item, which the United
States does not provide for any tax provision, there is no opportunity to
provide treatment comparable to that of Canada’s dividend relief.
Measurement of capital gains provisions could be similarly uneven across
countries.

None of these countries shows a substantial percentage of GDP devoted

to a make work pay non-wastable tax credit. There may be other countries
that do. The countries analysed thus far either do not have such provisions,
or count only the portion that offsets other tax liability as a reduction of
revenues. The largest cost from all make work pay tax expenditures in any
given country by a significant margin is in Spain, at 0.7% of GDP; Spain’s
largest tax expenditure in this category is a deduction (and therefore it is
wastable, that is, not refundable). Second is the United Kingdom, at 0.3% of
GDP, which is not an overly small category but is small relative to the
overall differences between the tax expenditures of the various countries.
Further research could present data for France, which does consider all of
the effect of its “make work pay” provision as a revenue cost; but for this
group of countries, such tax expenditures are not the key element of
numerical differences.
Yet again, the data at this stage do not lend themselves to firm
comparisons or conclusions. The United Kingdom’s sum of tax expenditures
is inexact just as any sum of revenue-forgone estimates would be. The
United Kingdom’s elaborate list of tax expenditures may simply be more
inclusive than is that of any other country. These data should be used to
discover good questions, not final answers.
Other taxes
Just as for income-tax tax expenditures, the United Kingdom has the
largest sum of non-income-tax tax expenditures, and therefore, of course,
the largest total of tax expenditures, as a percentage of GDP (see Figure 13
and Table 29). Again, careful further research would be needed to determine
the sources of the numbers, and whether they reflect programmatic
differences, or rather highly inclusive accounting in the United Kingdom
relative to less-inclusive accounting in other countries. As with tax
expenditures under the income tax, non-income tax expenditures in the
continental European countries and Korea are smaller than those in Canada
and the United Kingdom.. When viewed relative to tax revenue, however,
total tax expenditures in Canada are larger than those in the United Kingdom
(see Figure 14 and Table 30). This follows arithmetically from the larger
share of tax revenues in the GDP in the United Kingdom. Relative to tax
expenditures under the income tax, non-income tax expenditures are slightly
greater in Germany, and not quite equal in the Netherlands. Non-income tax
expenditures are significantly smaller relative to income tax expenditures in
United Kingdom, Korea, and Canada (in descending order of proportion).
The United States, as noted earlier, does not measure tax expenditures other
than in its income tax (but because the United States does not have a
national-level value-added tax, and its income taxes are a higher proportion

of its total revenue, tax expenditures outside of the income tax likely would
be less important in any event).
One question that might arise from the numerical comparisons across
countries is whether any one nation might have a total driven solely by one
very large tax expenditure. To answer that question, Figure 16 ranks each
country’s ten largest tax expenditures as percentages of GDP.6 If one
country had a larger total than another only because of one very large tax
expenditure, then the lines for the two countries in Figure 16 would cross. In
fact, with six countries represented in the figure, lines cross only once –
between Korea and the Netherlands, with both countries having
comparatively small totals of tax expenditures, and the two lines lying quite
close together for all ten tax expenditures. Thus, it appears that the tax
expenditure totals of the sample countries represent relatively consistent
patterns of behaviour over all of their tax expenditures, rather than policy
and measurement anomalies driven by individual outlying legal provisions.
Another question is the relative use of tax expenditures under different
taxes in different countries. To explore this question, Figure 17 compares
the ratio of tax expenditures under the income tax and the value-added tax in
the six countries in the sample that use both taxes. The figure reveals
patterns that are quite different from one country to another. Spain and the
United Kingdom have significantly larger tax expenditures under their
VATs than under their income taxes; Germany and Korea are the reverse;
Canada and the Netherlands have ratios that are more equal. Both ratios are
lower in Germany and the Netherlands than in Canada, Spain and the United
Kingdom; Korea is approximately in the middle on that scale.
Memorandum items
As noted earlier, Canada and the United Kingdom provide substantial
additional information in the form of memorandum items. After
reclassification for purposes of cross-country comparability, the
memorandum items for those two countries remain in excess of 3% and 4%
of GDP, respectively. In the case of Canada, the sum of the memorandum
items comes to well over one-third of the total amount of tax expenditures
(see Figure 15). This information has proved useful in this report. For
example, Canada does not consider its provision for relief of double taxation
of corporate dividends to be a tax expenditure, whereas some other countries
with similar provisions do. Adding Canada’s memorandum item to its list of
tax expenditures provides greater comparability with the tax expenditures
reported in those other countries.

Conclusions
This report presents an exploration of tax expenditure practices across a

sample of OECD member countries.
The definition of tax expenditures can steer analysis in one particular
direction or another. For example, a country that accepts as a tax
expenditure a provision that measures the ability to pay tax (or another
provision that facilitates the efficient operation of the tax system) will likely
take a different perspective on the issue than another that does not. But such
determinative ground rules aside, a tax expenditure will at times be the
policy instrument of choice to achieve a legitimate objective. In other
circumstances, a tax expenditure will be the path of least resistance toward
less well justified ends. Thus, in a time of fiscal challenges, the topic of tax
expenditures is clearly important.
This volume found no fool-proof process to distinguish between good
tax policy and bad in enactment, reporting or review. Different countries
employ different tools, and have had different experiences. Still, this
analysis may have identified policy or process ideas that will prove helpful
in some particular context. If one country’s practice can galvanise another’s
political will, better outcomes will result.
Numerical comparisons may measure differences in reporting practices
rather than in substantive tax policy. However, the presentation in this report
shows an apparently wide range of country practices, and may suggest
productive avenues for further research – again, perhaps motivating one
country to employ sound tools identified in another.
Ideally, these institutional comparisons will suggest some ideas for
policy makers to consider. The data comparisons were expected to raise
questions rather than answers, and they appear to have achieved the more
reasonable goal. Economic and budgetary pressures should motivate further
inquiry and data analysis.

Notes
1. Note that both the number and the size of Germany’s business tax
expenditures are significantly lower than those for most other countries,
suggesting that Germany may simply make less use of tax expenditures,
comparably measured.
2. No country currently employs a so-called “consumed-income” or
consumption tax at the household level (although many countries use a
value-added or sales tax collected at the time and place of routine
purchases by households). However, any country could use a
consumption-tax benchmark to measure tax expenditures under its income
tax. The United States has discussed such a possibility in its tax
expenditure reports. See Office of Management and Budget (2008).
3. Canada does provide such separate estimates.
4. Anderson and Minarik (2006) question the effectiveness of these deficit
rules.
5. See Table 10, p. 49.
6. Spain is not included in Figure 16 because it reports tax expenditures
under its value-added tax in groups rather than individually, which would
distort the results in the figure.

Bibliography
Anderson, Barry and Joseph J. Minarik (2006), “Design Choices for Fiscal
Policy Rules”, OECD Journal on Budgeting, Vol. 5, No. 4, pp. 159-208.
Koiwa, Tetsuro (2006), “Recent Issues on Tax Expenditures in OECD
Countries”, OECD unpublished paper, table 10, p. 49.
Office of Management and Budget (2008), “Tax Expenditures”, in Budget of
the United States Government, Fiscal Year 2009, Analytical
Perspectives, Executive Office of the President, Washington DC,
Chapter 19, p. 287.
Schick, Allen (2007), “Off-Budget Expenditure: An Economic and Political
Framework”, OECD Journal on Budgeting, Vol. 7, No. 3, OECD, Paris.

PART II: COMPARING TAX EXPENDITURES IN OECD COUNTRIES – 169
Part II
Comparing tax expenditures in OECD countries

Explanatory key
This part presents the data for each country. The data are presented in
four different tables for each country. They are:
• Tax expenditures as a percentage of GDP;
• Tax expenditures as a percentage of total taxes;
• Tax expenditures by type of tax as a percentage of revenue collected
under that type of tax; and
• The number of tax expenditures.
The left-hand side of each country table reflects the tax expenditures as
closely as possible to each country’s own intended presentation. Every tax
expenditure identified by each country is included. Provisions identified by
two countries (Canada and the United Kingdom) as “memorandum items”,
which are not considered to be tax expenditures, are listed separately.
The right-hand side of each country table attempts to minimise the
degree of non-comparability across countries by re-categorising tax
expenditures and memorandum items to approach a common standard. This
includes: i) reclassifying some “memorandum items” identified by Canada
and the United Kingdom to be considered as tax expenditures, where
prevailing practices in the other countries would so identify such provisions;
and ii) reclassifying some tax expenditures identified by the other countries
to a category called “structural items”, if prevailing practices in this group of
countries would tend not to consider these provisions to be tax expenditures.
The category of “structural items” includes those “memorandum items” of
Canada and the United Kingdom that continue to be considered not to be tax
expenditures after this reclassification. The number of tax expenditures and
other provisions so reclassified is small.
The tables on international comparisons present the latest actual year of
tax expenditure data for each country, using the reclassifications developed
for the right-hand part of each individual country’s tables.
The sources for the tables and figures that follow are listed on pages
238-240.

172 – PART II: COMPARING TAX EXPENDITURES IN OECD COUNTRIES
Table II.1. Tax expenditures in Canada (% of GDP)
As reported by country
2001 † 2002 2003 2004 2005 ‡ 2006 ‡ 2007 ‡ 2008 ‡ 2009 ‡
Purpose of tax expenditure, income tax*
General tax relief [4] 0.16 0.16 0.15 0.14 0.14 0.14 0.23 0.22 0.23
Low-income non-work related [1] [2] 0.02 0.02 0.02 0.02 0.02 0.02 0.02 0.02 0.02
Retirement [1] [2] 0.65 0.96 1.35 1.68 1.87 2.03 2.07 2.06 2.03
Work related [1] [2] 0.12 0.10 0.09 0.10 0.11 0.12 0.11 0.12 0.12
Education 0.12 0.12 0.12 0.12 0.11 0.12 0.11 0.11 0.12
Health 0.24 0.25 0.26 0.27 0.25 0.25 0.26 0.27 0.27
Housing 0.08 0.12 0.15 0.20 0.25 0.28 0.27 0.27 0.27
General business incentives [1] [2] 1.00 0.82 0.76 0.85 0.97 1.15 1.12 1.12 1.10
Research & development [2] 0.22 0.21 0.20 0.24 0.25 0.27 0.28 0.30 0.33
Specific industry relief [1] [2] 0.21 0.17 0.08 0.02 0.02 0.04 0.07 0.07 0.07
Intergovernmental relations [1] [2] 1.63 1.56 1.55 1.55 1.56 1.57 1.56 1.58 1.60
Charity [1] [2] 0.20 0.18 0.19 0.21 0.20 0.21 0.21 0.21 0.20
Other [1] [2] 0.02 0.02 0.02 0.02 0.02 0.03 0.03 0.03 0.03
Make work pay 0.00 0.01 0.01 0.01 0.01 0.04 0.16 0.16 0.16
Total 4.67 4.71 4.95 5.44 5.77 6.26 6.51 6.54 6.54
Capital income taxation
Accelerated depreciation [2]
Interest
Dividends
Capital gains [2]
Subtotal
Total
2001 † 2002 2003 2004 2005 ‡ 2006 ‡ 2007 ‡ 2008 ‡ 2009 ‡
Make work pay
Total
GST tax related [1] [2] [3] 1.11 1.12 1.12 1.16 1.17 1.10 1.02 0.89 0.90
Total 5.78 5.82 6.07 6.60 6.93 7.36 7.52 7.43 7.44
Memorandum items [1] [2] 3.94 3.86 3.66 3.56 3.50 3.52 3.47 3.40 3.42
Structural items [1] [2]
Grand total 9.72 9.69 9.73 10.16 10.44 10.88 10.99 10.83 10.86
Income tax expenditures by type*
Credits [2] 1.21 1.20 1.19 1.23 1.20 1.32 1.53 1.56 1.59
Deductions, exemptions & exclusions [1] [2] 2.52 2.33 2.31 2.50 2.65 2.86 2.83 2.85 2.85
Deferrals [1] [2] 0.47 0.77 1.15 1.50 1.71 1.85 1.86 1.85 1.83
Reduced rates [1] 0.47 0.40 0.30 0.21 0.20 0.24 0.28 0.28 0.27
† Tax expenditures are reported by calendar year rather than fiscal year.
‡ Projections.
* Classification of income tax expenditures by purpose and by type is to some degree arbitrary.
[1] At least one provision in this category is not estimated because adequate data are not available.
[2] At least one provision in this category is not estimated because it cost less than CAD 2.5 million.
[3] Projections begin in 2006 for GST tax-related provisions.
[4] There are no tax expenditures in this category.
# Based on “budgetary revenues” as defined by the Canadian Government.

Table II.1. Tax expenditures in Canada (% of GDP) cont'd

With reclassifications by author
2001 † 2002 2003 2004 2005 ‡ 2006 ‡ 2007 ‡ 2008 ‡ 2009 ‡
General tax relief
Retirement [1] [2] 0.65 0.96 1.35 1.68 1.87 2.03 2.07 2.06 2.03
Work related [1] [2] 0.46 0.43 0.40 0.39 0.37 0.36 0.34 0.34 0.34
Education 0.12 0.12 0.12 0.12 0.11 0.12 0.11 0.11 0.12
Health 0.24 0.25 0.26 0.27 0.25 0.25 0.26 0.27 0.27
Housing 0.08 0.12 0.15 0.20 0.25 0.28 0.27 0.27 0.27
Charity [1] [2] 0.20 0.18 0.19 0.21 0.20 0.21 0.21 0.21 0.20
Other [1] [2] 0.02 0.02 0.02 0.02 0.02 0.03 0.03 0.03 0.03
Make work pay
Total 4.39 4.54 4.78 5.16 5.36 5.66 5.72 5.76 5.78
Accelerated depreciation [2] 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00
Interest [4]
Dividends 0.29 0.27 0.22 0.27 0.29 0.39 0.39 0.40 0.40
Capital gains [2] 0.42 0.29 0.27 0.35 0.44 0.55 0.51 0.48 0.45
Subtotal 0.71 0.56 0.49 0.62 0.73 0.94 0.91 0.88 0.85

2001 † 2002 2003 2004 2005 ‡ 2006 ‡ 2007 ‡ 2008 ‡ 2009 ‡
Total 5.09 5.10 5.27 5.77 6.09 6.60 6.62 6.64 6.63
Make work pay 0.00 0.01 0.01 0.01 0.01 0.04 0.16 0.16 0.16
Total 5.10 5.10 5.27 5.78 6.10 6.64 6.78 6.80 6.79
GST tax related [1] [2] [3] 1.11 1.12 1.12 1.16 1.17 1.10 1.02 0.89 0.90
Total 6.21 6.22 6.40 6.94 7.26 7.74 7.80 7.70 7.68
Memorandum items [1] [2]
Structural items [1] [2] 3.51 3.46 3.33 3.22 3.17 3.14 3.19 3.14 3.17
Grand total
Credits [2] 1.46 1.43 1.40 1.44 1.40 1.53 1.64 1.67 1.69
Deductions, exemptions, & exclusions [1] [2] 2.71 2.50 2.42 2.64 2.78 3.02 3.01 3.00 3.00
Deferrals [1] [2] 0.47 0.77 1.15 1.50 1.71 1.85 1.86 1.85 1.83
Reduced rates [1] 0.47 0.40 0.30 0.21 0.20 0.24 0.28 0.28 0.27
‡ Projections.
# Based on “budgetary revenues” as defined by the Canadian Government.
12 http://dx.doi.org/10.1787/746862313611

Table II.2. Tax expenditures in Canada

(% of central government total tax and non-tax receipts)
As reported by country ǈ
2001 † 2002 2003 2004 2005 ‡ 2006 ‡ 2007 ‡ 2008 ‡ 2009 ‡
Purpose of tax expenditure, income
tax*
General tax relief [1] 0.93 0.98 0.96 0.92 0.89 0.87 1.45 1.45 1.51
Retirement [2] [3] 3.77 5.95 8.49 10.72 11.96 13.01 13.30 13.34 13.64
Work related [2] [3] 0.68 0.60 0.58 0.67 0.73 0.77 0.74 0.75 0.78
Education 0.67 0.75 0.76 0.78 0.69 0.74 0.68 0.73 0.78
Health 1.39 1.58 1.64 1.70 1.61 1.63 1.68 1.74 1.81
Housing 0.46 0.76 0.95 1.29 1.59 1.77 1.76 1.76 1.79
Charity [2] [3] 1.15 1.13 1.18 1.32 1.26 1.33 1.33 1.34 1.37
Other [2] [3] 0.11 0.12 0.11 0.13 0.12 0.16 0.20 0.21 0.22
Make work pay 0.03 0.03 0.04 0.04 0.04 0.25 1.02 1.03 1.06
Total 26.97 29.23 31.15 34.75 36.97 40.13 41.85 42.37 43.89
Interest [1]
Dividends
Capital gains [2]
Subtotal
Total
Make work pay
Total
GST tax related [2] [3] [4] 6.43 6.95 7.06 7.43 7.48 7.06 6.54 5.79 6.02
Total 33.40 36.18 38.21 42.18 44.45 47.19 48.39 48.16 49.91
Structural items [2] [3]
Memorandum items [1] [2] 22.76 24.00 23.06 22.79 22.46 22.56 22.32 22.04 22.94
Grand total 56.16 60.18 61.27 64.96 66.91 69.75 70.71 70.20 72.85
Credits [3] 7.01 7.45 7.47 7.88 7.70 8.43 9.86 10.12 10.65
Deductions, exemptions & exclusions
14.56 14.46 14.55 15.97 16.98 18.31 18.23 18.44 19.14
[2] [3]
Deferrals [2] [3] 2.69 4.81 7.26 9.56 10.98 11.86 11.98 11.99 12.30
Reduced rates [2] 2.72 2.50 1.88 1.34 1.31 1.53 1.77 1.82 1.81
Ǉ Based on “budgetary revenues” as defined by the Canadian Government.

‡ Projections.

Table II.2. Tax expenditures in Canada

(% of central government total tax and non-tax receipts) cont’d
With reclassifications by author ǈ

2001 † 2002 2003 2004 2005 ‡ 2006 ‡ 2007 ‡ 2008 ‡ 2009 ‡
Purpose of tax expenditure,
income tax*
General tax relief
Low-income non-work related [2]
0.14 0.14 0.14 0.13 0.12 0.12 0.13 0.13 0.13
[3]
Retirement [2] [3] 3.77 5.95 8.49 10.72 11.96 13.01 13.30 13.34 13.64
Work related [2] [3] 2.66 2.64 2.52 2.47 2.40 2.32 2.20 2.21 2.27
Education 0.67 0.75 0.76 0.78 0.69 0.74 0.68 0.73 0.78
Health 1.39 1.58 1.64 1.70 1.61 1.63 1.68 1.74 1.81
Housing 0.46 0.76 0.95 1.29 1.59 1.77 1.76 1.76 1.79
General business incentives [2][3] 3.05 2.93 2.63 2.64 2.72 2.91 2.97 3.07 3.15
Charity [2] [3] 1.15 1.13 1.18 1.32 1.26 1.33 1.33 1.34 1.37
Other [2] [3] 0.11 0.12 0.11 0.13 0.12 0.16 0.20 0.21 0.22
Make work pay
Total 25.34 28.18 30.11 32.97 34.39 36.28 36.78 37.35 38.79
Interest [1]
Dividends 1.70 1.68 1.37 1.70 1.86 2.47 2.53 2.59 2.70
Capital gains [2] 2.40 1.81 1.71 2.23 2.79 3.56 3.29 3.10 2.99
Subtotal 4.10 3.49 3.08 3.93 4.66 6.03 5.82 5.69 5.69
Total 29.44 31.68 33.19 36.90 39.05 42.31 42.60 43.05 44.48
Make work pay 0.03 0.03 0.04 0.04 0.04 0.25 1.02 1.03 1.06
Total 29.47 31.71 33.23 36.94 39.09 42.56 43.63 44.08 45.54
GST tax related [2] [3] [4] 6.43 6.95 7.06 7.43 7.48 7.06 6.54 5.79 6.02
Total 35.89 38.66 40.29 44.37 46.57 49.62 50.17 49.87 51.55
Structural items [2] [3] 20.27 21.52 20.98 20.59 20.34 20.13 20.54 20.32 21.29
Memorandum items [1] [2]
Grand total
Credits [3] 8.41 8.89 8.82 9.18 8.98 9.78 10.54 10.82 11.33
Deductions, exemptions &
15.65 15.51 15.27 16.86 17.82 19.39 19.33 19.45 20.10
exclusions [2] [3]
Deferrals [2] [3] 2.69 4.81 7.26 9.56 10.98 11.86 11.98 11.99 12.30
Reduced rates [2] 2.72 2.50 1.88 1.34 1.31 1.53 1.77 1.82 1.81
Ǉ Based on “budgetary revenues” as defined by the Canadian Government.

‡ Projections.
12 http://dx.doi.org/10.1787/746862313611

Table II.3. Tax expenditures in Canada (% of relevant tax revenue)
As reported by country † ‡
2001* 2002 2003 2004 2005** 2006** 2007** 2008** 2009**
tax***
General tax relief 1.46 1.58 1.58 1.48 1.42 1.37 2.23 2.18 2.22
Retirement [1] [2] 5.95 9.65 13.96 17.23 19.15 20.49 20.42 20.08 20.00
Work related [1] [2] 1.07 0.97 0.95 1.08 1.16 1.54 2.28 2.27 2.28
Education 1.05 1.22 1.24 1.25 1.10 1.16 1.05 1.09 1.14
Health 2.20 2.56 2.69 2.73 2.58 2.57 2.58 2.61 2.65
Housing 0.73 1.23 1.56 2.07 2.54 2.79 2.71 2.65 2.62
Charity [1] [2] 1.81 1.84 1.94 2.13 2.02 2.09 2.05 2.01 2.00
Other [1] [2] 0.17 0.19 0.19 0.20 0.19 0.25 0.30 0.31 0.32
Make work pay 0.05 0.06 0.06 0.06 0.07 0.07 0.43 0.41 0.41
Total
Interest [1]
Dividends
Capital gains [1]
Subtotal
Total
Make work pay
Total 42.57 47.43 51.20 55.84 59.18 63.22 64.26 63.80 64.34
GST tax related [1] [2] [3] 49.49 49.54 48.13 52.38 52.48 48.85 50.18 47.78 53.09
Income tax expenditures by type***
Credits [2] 11.07 12.09 12.27 12.66 12.32 13.28 15.15 15.24 15.61
Deductions, exemptions & exclusions
22.97 23.47 23.92 25.66 27.18 28.84 28.00 27.77 28.06
[1] [2]
Deferrals [1] [2] 4.24 7.81 11.93 15.36 17.58 18.69 18.40 18.05 18.03
Reduced rates [1] 4.29 4.06 3.09 2.15 2.09 2.41 2.72 2.74 2.65
† Percent of tax revenue by type of tax.

‡ Individual and corporate income taxes are considered together.
* Tax expenditures are reported by calendar year rather than fiscal year.
** Projections.
*** Classification of income tax expenditures by purpose and by type is to some degree arbitrary.

Table II.3. Tax expenditures in Canada (% of relevant tax revenue) cont’d
With reclassifications by author † ‡

2001* 2002 2003 2004 2005** 2006** 2007** 2008** 2009**
tax***
General tax relief [4]
Retirement [1] [2] 5.95 9.65 13.96 17.23 19.15 20.49 20.42 20.08 20.00
Work related [1] [2] 4.19 4.29 4.14 3.96 3.84 3.97 4.52 4.47 4.47
Education 1.05 1.22 1.24 1.25 1.10 1.16 1.05 1.09 1.14
Health 2.20 2.56 2.69 2.73 2.58 2.57 2.58 2.61 2.65
Housing 0.73 1.23 1.56 2.07 2.54 2.79 2.71 2.65 2.62
Charity [1] [2] 1.81 1.84 1.94 2.13 2.02 2.09 2.05 2.01 2.00
Other [1] [2] 0.17 0.19 0.19 0.20 0.19 0.25 0.30 0.31 0.32
Make work pay
Total 39.99 45.73 49.49 52.97 55.04 57.48 57.63 57.39 58.01
Interest [1]
Dividends 2.68 2.73 2.25 2.73 2.98 3.89 3.88 3.91 3.96
Capital gains [1] 3.79 2.94 2.82 3.59 4.47 5.60 5.06 4.67 4.38
Subtotal 6.47 5.67 5.07 6.32 7.46 9.50 8.94 8.58 8.34
Total 46.46 51.40 54.56 59.30 62.50 66.98 66.57 65.96 66.35
Make work pay 0.05 0.06 0.06 0.06 0.07 0.07 0.43 0.41 0.41
Total 46.51 51.46 54.62 59.36 62.57 67.05 67.00 66.38 66.75
GST tax related [1] [2] [3] 49.49 49.54 48.13 52.38 52.48 48.85 50.18 47.78 53.09
Credits [2] 13.28 14.42 14.50 14.76 14.37 15.41 16.19 16.30 16.60
24.69 25.17 25.10 27.09 28.52 30.54 29.69 29.29 29.47
exclusions [1] [2]
Deferrals [1] [2] 4.24 7.81 11.93 15.36 17.58 18.69 18.40 18.05 18.03
Reduced rates [1] 4.29 4.06 3.09 2.15 2.09 2.41 2.72 2.74 2.65
* Tax expenditures are reported by calendar year rather than fiscal year.
** Projections.
12 http://dx.doi.org/10.1787/746862313611

Table II.4. Number of tax expenditures in Canada (% of GDP)
2001 † 2002 2003 2004 2005 ‡ 2006 ‡ 2007 ‡
General tax relief 2 2 2 2 2 2 3
Low-income non-work related 4 4 4 4 5 5 6
Retirement 13 13 13 13 13 13 14
Work related 8 8 8 8 8 8 9
Education 8 9 9 9 9 10 10
Health 5 5 5 5 5 5 6
Housing 1 1 1 1 1 1 1
General business incentives 32 32 32 32 31 32 32
Research & development 5 5 5 5 5 5 5
Specific industry relief 32 32 34 34 33 35 35
Intergovernmental relations 8 8 8 8 8 8 8
Charity 13 13 13 13 13 13 13
Other 7 7 7 8 8 9 9
Make work pay 1 1 1 1 1 2 3
Total
Accelerated depreciation
Interest
Dividends
Capital gains
Subtotal
Total
Make work pay
Total 139 140 142 143 142 148 154
GST tax related [1] 32 32 32 32 32 32 32
Total 171 172 174 175 174 180 186
Memorandum items 39 39 39 38 38 38 38
Structural items
Grand total 210 211 213 213 212 218 224
Credits 30 30 32 32 33 37 41
Deductions, exemptions & exclusions 66 67 67 68 68 69 71
Deferrals 35 35 35 35 35 36 36
Reduced rates 8 8 8 8 6 6 6
‡ Projections.

Table II.4. Number of tax expenditures in Canada (% of GDP) cont’d

2001 † 2002 2003 2004 2005 ‡ 2006 ‡ 2007 ‡
Retirement 13 13 13 13 13 13 14
Work related 11 11 11 11 11 11 12
Education 8 9 9 9 9 10 10
Health 5 5 5 5 5 5 6
Housing 1 1 1 1 1 1 1
Charity 13 13 13 13 13 13 13
Other 7 7 7 8 8 9 9
Make work pay
Total 137 138 140 141 140 145 149
Accelerated depreciation 1 1 1 1 1 1 1
Interest 0 0 0 0 0 0 0
Dividends 3 3 3 3 3 3 3
Capital gains 3 3 3 3 3 3 3
Subtotal 7 7 7 7 7 7 7
Total 144 145 147 148 147 152 156
Make work pay 1 1 1 1 1 2 3
Total 145 146 148 149 148 154 159
GST tax related [1] 32 32 32 32 32 32 32
Total 177 178 180 181 180 186 191
Memorandum items
Structural items 33 33 33 32 32 32 33
Grand total
Credits 31 31 33 33 34 38 41
Deductions, exemptions, and exclusions 71 72 72 73 73 74 76
Deferrals 35 35 35 35 35 36 36
‡ Projections.
12 http://dx.doi.org/10.1787/746862313611

Table II.5. Tax expenditures in Germany (% of GDP) †

As reported by country With reclassifications by author
2005 2006 2007 ‡ 2008 ‡ 2005 2006 2007 ‡ 2008 ‡
Retirement 0.03 0.03 0.03 0.03 0.00 0.00 0.01 0.01
Work related [2] 0.04 0.04 0.04 0.04 0.03 0.03 0.03 0.03
Education [3] 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00
Health [1]
Housing [2] 0.20 0.18 0.18 0.15 0.20 0.18 0.18 0.15
General business incentives [2] [3] 0.00 0.00 0.02 0.01 0.00 0.00 0.02 0.01
Research & development [1]
Specific industry relief [2] [3] 0.02 0.01 0.01 0.01 0.02 0.01 0.01 0.01
Intergovernmental relations 0.05 0.03 0.01 0.01 0.05 0.03 0.01 0.01
Charity [1]
Other [2] 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00
Make work pay [1]
Total 0.34 0.29 0.29 0.26 0.30 0.26 0.27 0.23
Interest [1]
Dividends 0.04 0.04 0.02 0.02
Capital gains [1]
Subtotal 0.04 0.04 0.02 0.02
Total 0.34 0.29 0.29 0.26
Make work pay [1]
Total 0.34 0.29 0.29 0.26
Beer tax [1]
Electricity tax [2] 0.17 0.17 0.15 0.15 0.17 0.17 0.15 0.15
Fuel tax [2] 0.20 0.23 0.18 0.17 0.20 0.23 0.18 0.17
Inheritance tax [1]
Insurance tax [2] [3] 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00
Motor vehicle tax [1]
Sales tax [2] 0.05 0.05 0.06 0.06 0.05 0.05 0.06 0.06
Spirits tax 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00
Tobacco tax 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00
Subtotal 0.43 0.45 0.40 0.38 0.43 0.45 0.40 0.38
Grand total 0.77 0.74 0.69 0.64 0.77 0.74 0.69 0.64
Structural items
Income tax expenditures by type* 0.00 0.00 0.00 0.00
Credits [2] 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00
Deductions, exemptions & exclusions [2] [3] 0.33 0.28 0.28 0.25 0.33 0.28 0.28 0.25
Deferrals [2] [3] 0.01 0.00 0.00 0.00 0.01 0.00 0.00 0.00
Reduced rates [2] [3] 0.01 0.01 0.00 0.00 0.01 0.01 0.00 0.00
† Only costs to the federal government are included.
‡ 2007 is an estimate, 2008 is a projection.
[1] There are no federal tax expenditures in this category.

[3] At least one provision in this category is estimated to equal zero because its cost is
< EUR 0.5 million.
12 http://dx.doi.org/10.1787/747017860140

Table II.6. Tax expenditures in Germany (% of central government total tax revenue) †

2005 2006 2007 ‡ 2005 2006 2007 ‡
Retirement 0.41 0.39 0.28 0.05 0.05 0.08
Work related [2] 0.51 0.44 0.42 0.40 0.36 0.36
Education [3] 0.00 0.00 0.00 0.00 0.00 0.00
Health [1]
Housing [2] 2.31 2.01 1.89 2.31 2.01 1.89
General business incentives [2] [3] 0.05 0.04 0.17 0.05 0.04 0.17
Research & development [1]
Specific industry relief [2] [3] 0.23 0.14 0.15 0.23 0.14 0.15
Intergovernmental relations 0.56 0.30 0.15 0.56 0.30 0.15
Charity [1]
Other [2] 0.00 0.00 0.00 0.00 0.00 0.00
Make work pay [1]
Total 4.06 3.33 3.05 3.59 2.91 2.79
Interest [1]
Dividends 0.47 0.42 0.26
Capital gains [1]
Subtotal 0.47 0.42 0.26
Total 4.06 3.33 3.05
Make work pay [1]
Total 4.06 3.33 3.05
Beer tax [1]
Electricity tax [2] 2.05 1.92 1.62 2.05 1.92 1.62
Fuel tax [2] 2.38 2.65 1.88 2.38 2.65 1.88
Inheritance tax [1]
Insurance tax [2] [3] 0.00 0.00 0.00 0.00 0.00 0.00
Motor vehicle tax [1]
Sales tax [2] 0.63 0.59 0.68 0.63 0.59 0.68
Spirits tax 0.00 0.00 0.00 0.00 0.00 0.00
Tobacco tax 0.00 0.00 0.00 0.00 0.00 0.00
Subtotal 5.07 5.16 4.18 5.07 5.16 4.18
Grand total 9.13 8.48 7.23 9.13 8.48 7.23
Structural items 0.00 0.00 0.00
Credits [2] 0.00 0.00 0.00 0.00 0.00 0.00
Deductions, exemptions & exclusions [2] [3] 3.88 3.24 2.95 3.88 3.24 2.95
Deferrals [2] [3] 0.06 0.02 0.04 0.06 0.02 0.04
Reduced rates [2] [3] 0.11 0.07 0.05 0.11 0.07 0.05
† Only costs to the federal government are included.



< EUR 0.5 million.
12 http://dx.doi.org/10.1787/747017860140

Table II.7. Tax expenditures in Germany (% of relevant tax revenue) † ‡ *

2005 2006 2007** 2005 2006 2007**
Purpose of tax expenditure, income tax***
Retirement 1.14 1.04 0.75 0.13 0.14 0.23
Work related [2] 1.42 1.16 1.14 1.13 0.96 0.96
Education [3] 0.00 0.00 0.00 0.00 0.00 0.00
Health [1]
Housing [2] 6.47 5.33 5.11 6.47 5.33 5.11
General business incentives [2] [3] 0.13 0.12 0.45 0.13 0.12 0.45
Research & development [1] 0.00 0.00 0.00
Specific industry relief [2] [3] 0.65 0.36 0.41 0.65 0.36 0.41
Intergovernmental relations 1.56 0.80 0.40 1.56 0.80 0.40
Charity [1]
Other [2] 0.00 0.00 0.00 0.00 0.00 0.00
Make work pay [1]
Total 11.38 8.81 8.27 10.07 7.71 7.57
Interest [1]
Dividends 1.31 1.10 0.70
Capital gains [1]
Subtotal 1.31 1.10 0.70
Total 11.38 8.81 8.27
Make work pay [1]
Total 11.38 8.81 8.27
Electricity tax [2] 60.20 62.27 58.69 60.20 62.27 58.69
Fuel tax [2] 11.31 13.52 11.08 11.31 13.52 11.08
Sales tax [2] 1.62 1.54 1.68 1.62 1.54 1.68
Spirits tax 0.32 0.28 0.31 0.32 0.28 0.31
Tobacco tax 0.05 0.05 0.05 0.05 0.05 0.05
Credits [2] 0.01 0.01 0.01 0.01 0.01 0.01
Deductions, exemptions & exclusions [2] [3] 10.88 8.58 8.00 10.88 8.58 8.00
Deferrals [2] [3] 0.18 0.04 0.12 0.18 0.04 0.12
Reduced rates [2] [3] 0.31 0.18 0.14 0.31 0.18 0.14
† Only costs to the federal government are included. Percent of tax revenue by type of tax.
‡ Current assumptions: that electricity tax, spirits tax, fuel tax, and the tobacco tax are all federal taxes
with 100% of the revenue going to the federal government. It is also assumed that the fuel tax
mineralÖlsteuer and the energy tax Energiesteuer are one and the same.
* Individual and corporate income taxes are considered together.
** Estimate.
< EUR 0.5 million.
12 http://dx.doi.org/10.1787/747017860140

Table II.8. Number of tax expenditures in Germany (% of GDP) †

2005 2006 2007 ‡ 2008 ‡ 2005 2006 2007 ‡ 2008 ‡
General tax relief 0 0 0 0 0 0 0 0
Low-income non-work related 0 0 0 0 0 0 0 0
Retirement 2 2 2 2 1 1 1 1
Work related 4 4 4 4 2 2 2 2
Education 1 1 1 1 1 1 1 1
Health 0 0 0 0 0 0 0 0
Housing 9 10 10 10 9 10 10 10
General business incentives 9 9 10 10 9 9 10 10
Research & development 0 0 0 0 0 0 0 0
Specific industry relief 22 22 23 23 22 22 23 23
Intergovernmental relations 5 7 5 5 5 7 5 5
Charity 0 0 0 0 0 0 0 0
Other 1 1 1 1 1 1 1 1
Make work pay 0 0 0 0
Total 53 56 56 56 50 53 53 53
Accelerated depreciation 0 0 0 0
Interest 0 0 0 0
Dividends 3 3 3 3
Capital gains 0 0 0 0
Subtotal 3 3 3 3
Total 53 56 56 56
Make work pay 0 0 0 0
Total 53 56 56 56
Beer tax 0 0 0 0 0 0 0 0
Electricity tax 5 6 5 5 5 6 5 5
Fuel tax 12 13 13 13 12 13 13 13
Inheritance tax 0 0 0 0 0 0 0 0
Insurance tax 3 3 3 3 3 3 3 3
Motor vehicle tax 0 0 0 0 0 0 0 0
Sales tax 6 6 6 6 6 6 6 6
Spirits tax 1 1 1 1 1 1 1 1
Tobacco tax 1 1 1 1 1 1 1 1
Subtotal 28 30 29 29 28 30 29 29
Grand total 81 86 85 85 81 86 85 85
Structural items 0 0 0 0
Credits 2 2 2 2 2 2 2 2
Deductions, exemptions & exclusions 43 46 47 47 43 46 47 47
Deferrals 4 4 3 3 4 4 3 3
Reduced rates 4 4 4 4 4 4 4 4
† Only tax expenditures which relate to federal government tax revenues are included.
12 http://dx.doi.org/10.1787/747017860140

Table II.9. Tax expenditures in Korea (% of GDP)

2006 2007 † 2006 2007 †
General tax relief 0.08 0.09 0.05 0.05
Low-income non-work related 0.03 0.04 0.03 0.04
Retirement 0.02 0.01 0.02 0.01
Work related [1] 0.03 0.03 0.03 0.03
Education 0.12 0.12 0.12 0.12
Health [2] 0.29 0.31 0.29 0.31
Housing 0.05 0.06 0.05 0.06
General business incentives [1] [2] 0.69 0.65 0.68 0.65
Research & development [1] 0.15 0.12 0.15 0.12
Specific industry relief [1] [2] 0.18 0.24 0.18 0.24
Intergovernmental relations [3]
Charity [2] 0.13 0.14 0.13 0.14
Other [1] [2] 0.02 0.02 0.02 0.02
Make work pay 0.01 0.01
Total 1.79 1.83 1.75 1.79
Accelerated depreciation 0.00 0.00
Interest [3]
Dividends [3]
Capital gains [3]
Subtotal 0.00 0.00
Total 1.75 1.79
Total 1.76 1.80
Stamp tax 0.01 0.01 0.01 0.01
Inheritance tax 0.00 0.01 0.00 0.01
Educational tax 0.03 0.03 0.03 0.03
Security transaction tax 0.02 0.02 0.02 0.02
Special excise tax 0.03 0.03 0.03 0.03
Liquor tax 0.01 0.01 0.01 0.01
Customs duties 0.04 0.03 0.04 0.03
Transportation tax 0.14 0.15 0.14 0.15
Value added tax [2] 0.45 0.44 0.45 0.44
Subtotal 0.72 0.72 0.72 0.72
Grand total 2.52 2.55 2.48 2.52
Structural items 0.03 0.03
Credits 0.02 0.01 0.02 0.01
Deductions, exemptions & exclusions [1] [2] 1.73 1.78 1.70 1.75
Deferrals [1] [2] 0.00 0.00 0.00 0.00
Reduced rates 0.04 0.05 0.04 0.05
† Prospect.

[1] At least one provision in this category was established in 2006 or 2007 and was not estimated in
the year it was established.
12 http://dx.doi.org/10.1787/747025408334

Table II.10. Tax expenditures in Korea


2006 2007 † 2006 2007 †
General tax relief 0.47 0.49 0.29 0.31
Retirement 0.10 0.08 0.10 0.08
Work related [1] 0.16 0.15 0.16 0.15
Education 0.67 0.67 0.67 0.67
Health [2] 1.67 1.78 1.67 1.78
Housing 0.29 0.35 0.29 0.35
Charity [2] 0.76 0.82 0.76 0.82
Other [1] [2] 0.09 0.09 0.09 0.09
Total 10.34 10.41 10.09 10.17
Accelerated depreciation 0.02 0.01
Interest [3]
Dividends [3]
Capital gains [3]
Subtotal 0.02 0.01
Total 10.11 10.18
Total 10.16 10.24
Stamp tax 0.04 0.04 0.04 0.04
Inheritance tax 0.03 0.04 0.03 0.04
Educational tax 0.16 0.16 0.16 0.16
Security transaction tax 0.14 0.12 0.14 0.12
Special excise tax 0.18 0.18 0.18 0.18
Liquor tax 0.03 0.03 0.03 0.03
Customs duties 0.20 0.20 0.20 0.20
Transportation tax 0.83 0.85 0.83 0.85
Value added tax [2] 2.59 2.48 2.59 2.48
Subtotal 4.18 4.10 4.18 4.10
Grand total 14.52 14.51 14.34 14.33
Credits 0.11 0.05 0.11 0.05
Deferrals [1] [2] 0.02 0.01 0.02 0.01
Reduced rates 0.24 0.26 0.24 0.26
† Prospect.

[1] At least one provision in this category was established in 2006 or 2007 and was not estimated in the
year it was established.
12 http://dx.doi.org/10.1787/747025408334

Table II.11. Tax expenditures in Korea (% of relevant tax revenue) † ‡

2006 2006
General tax relief 1.15 0.72
Low-income non-work related 0.45 0.45
Retirement 0.23 0.23
Work related [1] 0.39 0.39
Education 1.64 1.64
Health [2] 4.06 4.06
Housing 0.71 0.71
General business incentives [1] [2] 9.65 9.61
Research & development [1] 2.12 2.12
Specific industry relief [1] [2] 2.56 2.56
Charity [2] 1.85 1.85
Other [1] [2] 0.22 0.22
Make work pay 0.13
Total 25.17 24.56
Accelerated depreciation 0.05
Interest [3]
Dividends [3]
Capital gains [3]
Subtotal 0.05
Total 24.60
Make work pay 0.13
Total 24.73
Stamp tax 8.76 8.76
Inheritance tax 1.64 1.6 [4]
Educational tax 6.80 6.81
Security transaction tax 7.93 7.93
Special excise tax 5.28 5.28
Liquor tax 1.90 1.90
Customs duties 4.39 4.39
Transportation tax 12.64 12.64
Value added tax [2] 9.98 9.98
Credits 0.27 0.27
Deductions, exemptions & exclusions [1] [2] 24.27 23.84
Deferrals [1] [2] 0.05 0.05
Reduced rates 0.57 0.57
[1] At least one provision in this category was established in 2006 and was not estimated in the year it
was established.
12 http://dx.doi.org/10.1787/747025408334

Table II.12. Number of tax expenditures in Korea (% of GDP)

2006 2007 † 2006 2007 †
General tax relief 3 3 1 1
Low-income non-work related 2 2 2 2
Retirement 2 2 2 2
Work related 4 4 4 4
Education 5 5 5 5
Health 3 3 3 3
Housing 12 12 12 12
General business incentives 50 55 49 54
Research & development 7 8 7 8
Specific industry relief [1] 34 34 34 34
Intergovernmental relations 0 0 0 0
Charity 4 4 4 4
Other 11 12 11 12
Make work pay 1 1
Total 138 145 134 141
Accelerated depreciation 1 1
Interest 0 0
Dividends 0 0
Capital gains 0 0
Subtotal 1 1
Total 135 142
Make work pay 1 1
Total 136 143
Stamp tax 6 6 6 6
Inheritance tax [1] 2 2 2 2
Educational tax [1] 3 3 3 3
Security transaction tax 17 17 17 17
Special excise tax [1] 11 11 11 11
Liquor tax 1 1 1 1
Customs duties [1] 13 12 13 12
Transportation tax [1] 3 3 3 3
Value added tax [1] 26 26 26 26
Subtotal 82 81 82 81
Grand total 220 226 218 224
Structural items 2 2
Credits 2 2 2 2
Deductions, exemptions & exclusions 122 [1] 125 [1] 120 123
Deferrals 7 10 7 10
Reduced rates 7 [1] 8 [1] 7 8
† Prospect.
[1] At least one provision in this category applies under at least two different taxes. These provisions are
therefore counted under each tax. If these provisions were counted only once, the total number of tax
expenditures in 2007 would be 215 rather than 226. The corporate and individual income taxes are
considered to be one tax.
12 http://dx.doi.org/10.1787/747025408334

Table II.13. Tax expenditures in the Netherlands (% of GDP)
2006 2007 † 2008 † 2009 † 2010 † 2011 † 2012 †
General tax relief 0.00 0.00 0.00 0.00 0.00 0.00 0.00
Low-income non-work related 0.00 0.00 0.00 0.00 0.00 0.00 0.00
Retirement 0.06 0.06 0.05 0.05 0.05 0.05 0.05
Work related 0.06 0.05 0.05 0.05 0.05 0.05 0.05
Education 0.06 0.06 0.06 0.06 0.06 0.06 0.06
Health [1]
Housing 0.05 0.04 0.04 0.04 0.04 0.04 0.04
General business incentives 0.48 0.45 0.43 0.43 0.42 0.41 0.40
Research & development 0.07 0.07 0.07 0.07 0.08 0.08 0.08
Specific industry relief 0.18 0.15 0.14 0.14 0.14 0.14 0.14
Charity 0.09 0.08 0.08 0.08 0.08 0.08 0.08
Other 0.01 0.01 0.01 0.01 0.01 0.01 0.01
Make work pay 0.04 0.04 0.04 0.04 0.04 0.05 0.05
Total 1.11 1.02 0.99 0.99 0.98 0.98 0.97
Interest
Dividends
Capital gains
Subtotal
Total
Make work pay
Total
Excises 0.08 0.07 0.07 0.07 0.07 0.07 0.08
Heavy motor vehicle tax 0.00 0.00 0.00 0.00 0.00 0.00 0.00
Motor vehicle tax 0.02 0.02 0.03 0.03 0.03 0.03 0.03
Regulating energy tax [2] 0.03 0.03 0.00 0.00 0.00 0.00 0.00
Special excise on motor vehicles 0.01 0.02 0.01 0.01 0.01 0.01 0.01
Tax on the sale of immovable property 0.02 0.02 0.02 0.02 0.02 0.02 0.02
VAT [3] 0.73 0.72 0.71 0.71 0.70 0.69 0.68
Subtotal 0.90 0.88 0.85 0.84 0.83 0.82 0.81
Grand total 2.00 1.90 1.83 1.83 1.81 1.80 1.78
Structural items
Credits 0.06 0.06 0.06 0.07 0.07 0.08 0.08
Deductions, exemptions & exclusions 0.81 0.73 0.70 0.70 0.68 0.67 0.66
Deferrals 0.05 0.06 0.05 0.05 0.05 0.05 0.05
Reduced rates 0.19 0.17 0.17 0.17 0.17 0.18 0.17
† 2007 and 2008 are initial estimates. 2009-2012 are projections.
[2] At least one provision in this category is not estimated because its cost is small.

Table II.13. Tax expenditures in the Netherlands (% of GDP) cont’d

2006 2007 † 2008 † 2009 † 2010 † 2011 † 2012 †
Retirement 0.06 0.06 0.05 0.05 0.05 0.05 0.05
Work related 0.06 0.05 0.05 0.05 0.05 0.05 0.05
Education 0.06 0.06 0.06 0.06 0.06 0.06 0.06
Health [1]
Housing 0.05 0.04 0.04 0.04 0.04 0.04 0.04
Charity 0.09 0.08 0.08 0.08 0.08 0.08 0.08
Other 0.01 0.01 0.01 0.01 0.01 0.01 0.01
Make work pay
Total 1.06 0.98 0.95 0.94 0.93 0.93 0.92
Interest [1]
Dividends [1]
Capital gains [1]
Subtotal 0.00 0.00 0.00 0.00 0.00 0.00 0.00
Total 1.06 0.98 0.95 0.94 0.93 0.93 0.92
Make work pay 0.04 0.04 0.04 0.04 0.04 0.05 0.05
Total 1.10 1.01 0.98 0.98 0.97 0.98 0.96
Excises 0.08 0.07 0.07 0.07 0.07 0.07 0.08
Heavy motor vehicle tax 0.00 0.00 0.00 0.00 0.00 0.00 0.00
Motor vehicle tax 0.02 0.02 0.03 0.03 0.03 0.03 0.03
Regulating energy tax [2] 0.03 0.03 0.00 0.00 0.00 0.00 0.00
Special excise on motor vehicles 0.01 0.02 0.01 0.01 0.01 0.01 0.01
Tax on the sale of immovable property 0.02 0.02 0.02 0.02 0.02 0.02 0.02
VAT [3] 0.73 0.72 0.71 0.71 0.70 0.69 0.68
Subtotal 0.90 0.88 0.85 0.84 0.83 0.82 0.81
Grand total 2.00 1.89 1.83 1.83 1.81 1.80 1.78
Structural items 0.00 0.00 0.00 0.00 0.00 0.00 0.00
Credits 0.06 0.06 0.06 0.07 0.07 0.08 0.08
Deductions, exemptions & exclusions 0.80 0.73 0.70 0.69 0.68 0.67 0.66
Deferrals 0.05 0.06 0.05 0.05 0.05 0.05 0.05
Reduced rates 0.19 0.17 0.17 0.17 0.17 0.18 0.17
† 2007 and 2008 are initial estimates. 2009-2012 are projections.
12 http://dx.doi.org/10.1787/747044711308

Table II.14. Tax expenditures in the Netherlands

As reported by country With reclassiications by author

2006 2007 † 2008 † 2006 2007 † 2008 †
General tax relief 0.01 0.01 0.01 [1] [1] [1]
Low-income non-work related 0.00 0.00 0.00 0.00 0.00 0.00
Retirement 0.16 0.15 0.14 0.16 0.15 0.14
Work related 0.17 0.15 0.13 0.17 0.15 0.13
Education 0.16 0.18 0.17 0.16 0.18 0.17
Health [1]
Housing 0.12 0.12 0.11 0.12 0.12 0.11
General business incentives 1.23 1.22 1.14 1.23 1.22 1.14
Research & development 0.19 0.20 0.19 0.19 0.20 0.19
Specific industry relief 0.47 0.40 0.37 0.47 0.40 0.37
Charity 0.22 0.23 0.22 0.22 0.23 0.22
Other 0.02 0.02 0.02 0.02 0.02 0.02
Make work pay 0.10 0.10 0.09
Total 2.85 2.79 2.60 2.74 2.68 2.50
Accelerated depreciation [1] [1] [1]
Interest [1] [1] [1]
Dividends [1] [1] [1]
Capital gains [1] [1] [1]
Subtotal 0.00 0.00 0.00
Total 2.74 2.68 2.50
Make work pay 0.10 0.10 0.09
Total 2.84 2.78 2.59
Excises 0.21 0.19 0.20 0.21 0.19 0.20
Heavy motor vehicle tax 0.00 0.00 0.00 0.00 0.00 0.00
Motor vehicle tax 0.06 0.06 0.07 0.06 0.06 0.07
Regulating energy tax [2] 0.08 0.08 0.00 0.08 0.08 0.00
Special excise on motor vehicles 0.04 0.04 0.02 0.04 0.04 0.02
Tax on the sale of immovable property 0.05 0.06 0.06 0.05 0.06 0.06
VAT [3] 1.87 1.97 1.89 1.87 1.97 1.89
Subtotal 2.31 2.41 2.23 2.31 2.41 2.23
Grand total 5.16 5.20 4.84 5.15 5.19 4.83
Structural items 0.01 0.01 0.01
Credits 0.14 0.16 0.16 0.14 0.16 0.16
Deductions, exemptions & exclusions 2.08 2.01 1.86 2.07 2.00 1.85
Deferrals 0.14 0.15 0.14 0.14 0.15 0.14
Reduced rates 0.49 0.47 0.44 0.49 0.47 0.44
† 2007 and 2008 are initial estimates.
12 http://dx.doi.org/10.1787/747044711308

Table II.15. Tax expenditures in the Netherlands (% of relevant tax revenue) † ‡

2006 2007* 2008* 2006 2007* 2008*
Purpose of tax expenditure, income tax**
General tax relief 0.03 0.03 0.03 [1] [1] [1]
Low-income non-work related 0.01 0.01 0.00 0.01 0.01 0.00
Retirement 0.55 0.52 0.46 0.55 0.52 0.46
Work related 0.58 0.50 0.42 0.58 0.50 0.42
Education 0.58 0.61 0.55 0.58 0.61 0.55
Health [1]
Housing 0.42 0.42 0.37 0.42 0.42 0.37
General business incentives 4.32 4.18 3.69 4.32 4.18 3.69
Research & development 0.67 0.68 0.62 0.67 0.68 0.62
Specific industry relief 1.63 1.37 1.21 1.63 1.37 1.21
Charity 0.78 0.79 0.70 0.78 0.79 0.70
Other 0.07 0.08 0.07 0.07 0.08 0.07
Make work pay 0.34 0.34 0.30
Total 9.98 9.53 8.40 9.60 9.15 8.07
Accelerated depreciation [1] [1] [1]
Capital gains [1] [1] [1]
Subtotal 0.00 0.00 0.00
Total 9.60 9.15 8.07
Make work pay 0.34 0.34 0.30
Total 9.95 9.49 8.37
Excises 4.41 3.96 4.12 4.41 3.96 4.12
Heavy motor vehicle tax 0.00 0.00 0.00 0.00 0.00 0.00
Motor vehicle tax 4.96 4.55 4.63 4.96 4.55 4.63
Regulating energy tax [2] 3.61 3.62 0.23 3.61 3.62 0.23
Special excise on motor vehicles 2.18 2.36 1.42 2.18 2.36 1.42
Tax on the sale of immovable property 2.02 2.33 2.06 2.02 2.33 2.06
VAT [3] 9.74 9.51 9.66 9.74 9.51 9.66
Income tax expenditures by type**
Credits 0.51 0.54 0.51 0.51 0.54 0.51
Deductions, exemptions & exclusions 7.30 6.86 6.00 7.27 6.82 5.97
Deferrals 0.47 0.52 0.46 0.47 0.52 0.46
Reduced rates 1.70 1.60 1.43 1.70 1.60 1.43
* 2007 and 2008 are initial estimates.
** Classification of income tax expenditures by purpose and by type is to some degree arbitrary.
12 http://dx.doi.org/10.1787/747044711308

Table II.16. Number of tax expenditures in the Netherlands (% of GDP)

2006 2007 † 2008 † 2006 2007 † 2008 †
General tax relief 1 1 1 0 0 0
Low-income non-work related 1 1 1 1 1 1
Retirement 2 2 2 2 2 2
Work related 6 6 6 6 6 6
Education 2 2 2 2 2 2
Health 0 0 0 0 0 0
Housing 2 2 2 2 2 2
General business incentives 13 13 12 13 13 12
Research & development 2 2 2 2 2 2
Specific industry relief 16 14 14 16 14 14
Intergovernmental relations 0 0 0 0 0 0
Charity 6 6 6 6 6 6
Other 2 2 2 2 2 2
Make work pay 2 2 2
Total 55 53 52 52 50 49
Accelerated depreciation 0 0 0
Interest 0 0 0
Dividends 0 0 0
Capital gains 0 0 0
Subtotal 0 0 0
Total 52 50 49
Make work pay 2 2 2
Total 54 52 51
Excises 7 6 6 7 6 6
Heavy motor vehicle tax 1 1 1 1 1 1
Motor vehicle tax 8 8 8 8 8 8
Regulating energy tax 3 3 3 3 3 3
Special excise on motor vehicles 4 4 4 4 4 4
Tax on the sale of immovable property 6 7 7 6 7 7
VAT 17 17 17 17 17 17
Subtotal 46 46 46 46 46 46
Grand total 101 99 98 100 98 97
Structural items 1 1 1
Credits 7 7 7 7 7 7
Deductions, exemptions & exclusions 35 34 33 34 33 32
Deferrals 6 6 6 6 6 6
Reduced rates 7 6 6 7 6 6
† 2007 and 2008 are initial estimates.
12 http://dx.doi.org/10.1787/747044711308

Table II.17. Tax expenditures in Spain (% of GDP)

2008 2009 † 2008 2009 †
General tax relief 0.28 0.82 [3] [3]
Retirement 0.17 0.19 0.17 0.19
Work related 0.01 0.01 0.01 0.01
Education 0.00 0.00 0.00 0.00
Health 0.00 0.00 0.00 0.00
Housing 0.41 0.59 0.41 0.59
General business incentives [1] 0.68 0.56 0.52 0.38
Specific industry relief [1] 0.04 0.04 0.04 0.04
Charity 0.02 0.02 0.02 0.02
Other [1] 0.17 0.18 0.17 0.18
Total 2.58 3.37 1.41 1.48
Accelerated depreciation [3] [3]
Interest [3] [3]
Dividends [3] [3]
Capital gains [5] 0.16 0.18
Subtotal 0.16 0.18
Total 1.57 1.66
Total 2.31 2.56
VAT [1] [2] 2.08 2.20 2.08 2.20
Tributes 0.01 0.01 0.01 0.01
Insurance tax 0.03 0.04 0.03 0.04
Alcohol and by-product beverages tax [1] 0.01 0.01 0.01 0.01
Non-residents equity tax [4] 0.00 0.00 0.00 0.00
Hydrocarbons tax [1] [2] 0.12 0.14 0.12 0.14
Subtotal 2.25 2.39 2.25 2.39
Total 4.83 5.76 4.55 4.95
Credits 0.51 0.53 0.32 0.36
Deferrals [3] 0.00 0.00 0.00 0.00
Reduced rates 0.36 0.21 0.36 0.21
† 2009 is a projection.
[4] This tax expenditure is no longer current.
[5] Tax deductions for corporate reinvestment and reserves.
12 http://dx.doi.org/10.1787/747112580363

Table II.18. Tax expenditures in Spain


2008 2009 † 2008 2009 †
Retirement 0.46 0.51 0.46 0.51
Work related 0.03 0.04 0.03 0.04
Education 0.01 0.01 0.01 0.01
Health 0.00 0.00 0.00 0.00
Housing 1.12 1.62 1.12 1.62
Charity 0.04 0.06 0.04 0.06
Other [1] 0.46 0.49 0.46 0.49
Total 7.08 9.35 3.86 4.09
Subtotal 0.44 0.50
Total 4.30 4.59
Total 6.32 7.08
VAT [1] [2] 5.70 6.10 5.70 6.10
Tributes 0.02 0.02 0.02 0.02
Insurance tax 0.09 0.11 0.09 0.11
Non-residents equity tax [4] 0.00 0.00 0.00 0.00
Subtotal 6.16 6.62 6.16 6.62
Total 13.24 15.97 12.48 13.70
Credits 1.38 1.47 0.87 0.99
Deferrals [3] 0.00 0.00 0.00 0.00
Reduced rates 0.99 0.59 0.99 0.59
12 http://dx.doi.org/10.1787/747112580363

Table II.19. Tax expenditures in Spain (% of relevant tax revenue)

2008 2009 † 2008 2009 †
Retirement 2.07 2.60 2.07 2.60
Work related 0.12 0.20 0.12 0.20
Education 0.05 0.05 0.05 0.05
Health 0.00 0.00 0.00 0.00
Housing 4.98 8.24 4.98 8.24
Charity 0.20 0.31 0.20 0.31
Other [1] 2.05 2.47 2.05 2.47
Total 31.56 47.38 17.21 20.73
Subtotal 1.97 2.54
Total 19.18 23.27
Total 28.16 35.89
VAT [1] [2] 59.69 66.12 59.69 66.12
Tributes 6.78 5.05 6.78 5.05
Insurance tax 23.65 25.07 23.65 25.07
Non-residents equity tax [2] [4] 0.00 0.00 0.00 0.00
Credits 6.17 7.44 3.89 5.00
Deferrals [3] 0.00 0.00 0.00 0.00
Reduced rates 4.40 2.97 4.40 2.97
12 http://dx.doi.org/10.1787/747112580363

Table II.20. Number of tax expenditures in Spain (% of GDP)

2008 2009 † 2008 2009 †
Retirement 3 3 3 3
Education 2 2 2 2
Health 1 1 1 1
Housing 3 5 3 5
Specific industry relief 10 10 10 10
Charity 5 5 5 5
Other 10 9 10 9
Make work pay 5 5
Total 77 81 68 71
Interest
Dividends
Capital gains [1] 2 2
Subtotal 2 2
Total 70 73
Make work pay 5 5
Total 75 78
VAT 48 55 48 55
Tributes 3 3 3 3
Insurance tax 5 6 5 6
Alcohol and by-product beverages tax 3 3 3 3
Non-residents equity tax [1] 1 0 1 0
Hydrocarbons tax 4 4 4 4
Subtotal 63 71 63 71
Total 140 152 138 149
Credits 14 14 13 13
Deductions, exemptions & exclusions 60 64 59 62
Deferrals 0 0 0 0
Reduced rates 3 3 3 3
12 http://dx.doi.org/10.1787/747112580363

Table II.21. Tax expenditures in the United Kingdom (% of GDP)

2006-07 2007-08 † 2006-07 2007-08 †
General tax relief [1] 0.00 0.00
Low-income non-work related [1] [2] 0.00 0.00 0.09 0.09
Retirement [1] [2] 2.13 2.05 2.32 2.24
Work related [1] [2] 0.15 0.15 0.15 0.15
Education [2] 0.00 0.00 0.00 0.00
Health [1] [2] 0.00 0.00 0.00 0.00
Housing [1] [2] 1.20 1.17 1.20 1.17
Intergovernmental relations [2] 0.00 0.00 0.00 0.00
Charity [2] 0.09 0.09 0.09 0.09
Other [2] 0.12 0.11 0.12 0.11
Make work pay [2] [4] 0.35 0.34
Total 4.16 4.00 4.90 4.72
Accelerated depreciation [3] 1.41 1.34
Interest [2] [3] 0.02 0.02
Dividends [3] 1.13 1.07
Subtotal 3.08 2.99
Total 7.98 7.71
Make work pay [3] [4] 0.35 0.34
Total 8.32 8.06
VAT [1] [2] 2.33 2.31 3.19 3.18
Inheritance tax related [1] [2] 0.08 0.08 0.98 1.03
Stamp duty reserve tax [2] 0.00 0.00 0.00 0.00
Stamp duty [2] 0.00 0.00 0.00 0.00
Stamp duty land tax [1] 0.01 0.01 0.19 0.17
Petroleum revenue tax [1] 0.00 0.00 0.09 0.06
Excise taxes 0.00 0.00 0.00 0.00
Landfill tax [1] [2] 0.00 0.00 0.00 0.00
Climate change levy [1] [2] 0.00 0.00 0.00 0.00
Aggregates levy [1] [2] 0.01 0.00 0.01 0.00
Air passenger duty [2] 0.00 0.00 0.01 0.01
Hydrocarbon oils [2] 0.00 0.00 0.00 0.00
Vehicle excise duty [3] [3] 0.01 0.01
Subtotal 2.43 2.41 4.47 4.48
Total 6.59 6.41 12.79 12.54
Reliefs with tax expenditure & structural
5.10 5.08
components
Structural reliefs 5.34 5.26


2006-07 2007-08 † 2006-07 2007-08 †
Grand total 17.03 16.75
Credits [2] 0.39 0.37 1.52 1.44
Deferrals [1] [2] 0.00 0.00 1.47 1.40
Reduced rates [2] 0.00 0.00 0.41 0.40
† Preliminary estimate.
[4] One provision in this category, called “Personal Tax Credits”, includes both the Child Tax Credit
(structural item) and Working Tax Credit (make work pay). The costs of those two parts cannot be
separated.
12 http://dx.doi.org/10.1787/747113565065

Table II.22. Tax expenditures in the United Kingdom

2006-07 2007-08 † 2006-07 2007-08 †
Retirement [1] [2] 5.85 5.63 6.38 6.15
Work related [1] [2] 0.42 0.41 0.42 0.41
Education [2] 0.00 0.00 0.00 0.00
Health [1] [2] 0.00 0.00 0.00 0.00
Housing [1] [2] 3.30 3.22 3.30 3.22
Charity [2] 0.25 0.25 0.25 0.25
Other [2] 0.32 0.30 0.32 0.30
Make work pay [2] [4] 0.95 0.92
Total 11.44 11.01 13.47 13.00
Interest [2] [3] 0.05 0.06
Dividends [3] 3.11 2.94
Subtotal 8.45 8.22
Total 21.92 21.22
Make work pay [3] [4] 0.95 0.92
Total 22.87 22.14
VAT [1] [2] 6.42 6.35 8.78 8.75
Inheritance tax related [1] [2] 0.21 0.21 2.68 2.84
Stamp duty reserve tax [2] 0.00 0.00 0.00 0.00
Stamp duty [2] 0.00 0.00 0.00 0.00
Stamp duty land tax [1] 0.02 0.02 0.51 0.47
Petroleum revenue tax [1] 0.00 0.00 0.24 0.17
Excise taxes 0.01 0.01 0.01 0.01
Landfill tax [1] [2] 0.00 0.00 0.00 0.00
Climate change levy [1] [2] 0.01 0.01 0.01 0.01
Aggregates levy [1] [2] 0.01 0.01 0.01 0.01
Air passenger duty [2] 0.00 0.00 0.01 0.03
Vehicle excise duty [3] [3] 0.04 0.03
Subtotal 6.68 6.61 12.30 12.32
Total 18.12 17.62 35.17 34.46
Reliefs with tax expenditure and structural
14.04 13.97
components


2006-07 2007-08 † 2006-07 2007-08 †
Structural reliefs 14.68 14.44
Grand total 46.84 46.04
Credits [2] 1.06 1.03 4.18 3.97
Deferrals [1] [2] 0.00 0.00 4.03 3.86
Reduced rates [2] 0.00 0.00 1.12 1.11
† Preliminary estimate.
separated.
12 http://dx.doi.org/10.1787/747113565065

Table II.23. Tax expenditures in the United Kingdom (% of relevant tax revenue) † ‡
As reported by country With reclassifcations by author

2006-07 2007-08* 2006-07 2007-08*
Retirement [1] [2] 9.94 9.42 11.03 10.55
Work related [1] [2] 0.71 0.69 0.72 0.70
Education [2] 0.01 0.01 0.01 0.01
Health [1] [2] 0.00 0.00 0.00 0.00
Housing [1] [2] 5.61 5.39 5.71 5.53
Charity [2] 0.42 0.41 0.43 0.42
Other [2] 0.55 0.51 0.56 0.52
Make work pay [2] [4] 1.62 1.54
Total 19.43 18.40 23.27 22.29
Interest [2] [3] 0.08 0.10
Dividends [3] 5.38 5.05
Subtotal 14.61 14.10
Total 37.88 36.39
Make work pay [3] [4] 1.65 1.58
Total 39.53 37.97
VAT [1] [2] 39.93 40.20 54.69 54.74
Stamp duties [1] [2] 0.71 0.73 18.48 15.91
Other [1] [2] [5] 2.43 2.43 46.79 49.23
Income tax expenditures by type**
Credits [2] 1.81 1.72 7.22 6.81
Deferrals [1] [2] 0.00 0.00 6.96 6.62
Reduced rates [2] 0.00 0.00 1.94 1.90
‡ Individual, corporate, and capital gains taxes as well as National Insurance Contributions are
considered together.
* Preliminary estimate.
[4] One provision in this category, “Personal Tax Credits”, includes both the Child Tax Credit
separated.
[5] “Other” includes all taxes with annual total revenue less than GBP 10 billion.
12 http://dx.doi.org/10.1787/747113565065

Table II.24. Number of tax expenditures in the United Kingdom (% of GDP) †

2006-07 2007-08 ‡ 2006-07 2007-08 ‡
Retirement 15 15 16 16
Education 4 4 4 4
Health 4 4 4 4
Housing 7 7 7 7
Specific industry relief 28 28 29 29
Charity 6 6 6 6
Other 33 33 27 27
Make work pay 3 3
Total 189 189 187 187
Accelerated depreciation 2 2
Interest 6 6
Dividends 3 3
Capital gains 7 7
Subtotal 18 18
Total 205 205
Make work pay 3 3
Total 208 208
VAT 34 35 43 44
Inheritance tax related 42 42 44 44
Stamp duty reserve tax 5 5 5 5
Stamp duty 8 8 8 8
Stamp duty land tax 17 18 22 23
Petroleum revenue tax 4 4 9 9
Excise taxes 2 2 2 2
Landfill tax 5 5 5 5
Climate change levy 9 9 9 9
Aggregates levy 18 18 18 18
Air passenger duty 5 5 6 6
Hydrocarbon oils 1 1 1 1
Vehicle excise duty 0 0 1 1
Subtotal 150 152 173 175
Total 339 341 381 383
Reliefs with tax expenditure & structural components 42 42
Structural reliefs 8 8


2006-07 2007-08 ‡ 2006-07 2007-08 ‡
Grand total 389 391
Credits 4 4 5 5
Deductions, exemptions & exclusions 177 177 186 186
Deferrals 7 7 11 11
Reduced rates 1 1 6 6
† Given reporting practices, some of these tax expenditures may have gone into effect only in 2007.
‡ Preliminary estimate.
(structural item) and Working Tax Credit (make work pay).
12 http://dx.doi.org/10.1787/747113565065

Table II.25. Tax expenditures in the United States (% of GDP)
As reported by the country

2002 2003 2004 2005 2006 2007 2008 2009 † 2010 † 2011 † 2012 † 2013 † 2014 †
General tax relief 0.21 0.35 0.19 0.34 0.23 0.23 0.20 0.19 0.18 0.13 0.06 0.05 0.05
Low-income non-work related 0.14 0.14 0.13 0.12 0.11 0.11 0.11 0.11 0.11 0.11 0.10 0.10 0.10
Retirement 1.47 1.50 1.19 1.04 1.00 1.02 1.02 1.06 1.04 1.09 1.11 1.09 1.07
Work related 0.12 0.07 0.07 0.07 0.07 0.07 0.07 0.07 0.07 0.07 0.07 0.07 0.07
Education 0.12 0.13 0.13 0.14 0.15 0.12 0.13 0.13 0.12 0.12 0.12 0.12 0.11
Health [1] 1.04 1.04 1.01 1.08 1.06 1.08 1.05 1.15 1.21 1.26 1.30 1.33 1.37
Housing 0.91 0.89 1.10 1.21 1.20 1.06 1.05 1.02 1.10 1.23 1.31 1.33 1.36
General business incentives [1] [2] 1.51 1.06 1.03 0.94 1.24 1.11 1.10 0.60 0.73 0.72 0.77 0.80 0.83
Research & development 0.08 0.03 0.02 0.08 0.08 0.11 0.09 0.08 0.06 0.05 0.05 0.05 0.05
Specific industry relief [1] 0.25 0.26 0.26 0.24 0.25 0.24 0.23 0.26 0.26 0.26 0.26 0.26 0.26
Intergovernmental relations 0.91 0.90 0.79 0.67 0.67 0.59 0.63 0.57 0.48 0.65 0.73 0.72 0.70
Charity 0.38 0.35 0.30 0.30 0.35 0.35 0.33 0.38 0.40 0.40 0.41 0.42 0.42
Other 0.04 0.03 0.03 0.04 0.04 0.04 0.10 0.10 0.09 0.09 0.09 0.08 0.08
Make work pay 0.07 0.08 0.07 0.07 0.07 0.06 0.06 0.07 0.06 0.06 0.07 0.06 0.06
Total 7.26 6.83 6.33 6.34 6.51 6.19 6.17 5.80 5.92 6.25 6.46 6.49 6.52
Capital income taxation [3]
Interest
Dividends
Capital gains
Subtotal
Total
As reported by the country

2002 2003 2004 2005 2006 2007 2008 2009 † 2010 † 2011 † 2012 † 2013 † 2014 †
Make work pay provisions
Total
Non-income tax [3]
Grand total 7.26 6.83 6.33 6.34 6.51 6.19 6.17 5.80 5.92 6.25 6.46 6.49 6.52
Structural items
Credits [1] 0.54 0.67 0.47 0.62 0.49 0.53 0.54 0.50 0.45 0.38 0.31 0.29 0.28
Exemptions & allowances [1] 5.35 5.12 5.04 5.01 5.03 4.73 4.63 4.77 4.83 5.25 5.51 5.53 5.56
Deferrals [1] 0.77 0.77 0.57 0.46 0.58 0.49 0.80 0.34 0.41 0.43 0.46 0.48 0.49
Reduced rates [1] 0.60 0.28 0.25 0.25 0.41 0.44 0.20 0.19 0.23 0.19 0.17 0.18 0.19
† Projections.
[1] At least one provision in this category was not estimated in 1994 because it cost USD 2.5 million.
[2] Beginning in 2003 lower rates of taxation for dividends and capital gains on corporate equity are not considered tax expenditures.

Table II.25. Tax expenditures in the United States (% of GDP) cont’d

1994 1995 1996 1997 1998 1999 2000 2001 2002 2003 2004
Purpose of tax
expenditure,
income tax*
General tax relief
[3]
Low-income non-
0.15 0.16 0.15 0.15 0.15 0.14 0.14 0.14 0.14 0.14 0.13
work related
Retirement 1.15 1.13 1.14 1.30 1.37 1.36 1.37 1.33 1.47 1.50 1.19
Work related 0.12 0.11 0.08 0.07 0.13 0.12 0.12 0.12 0.12 0.07 0.07
Education 0.04 0.04 0.04 0.04 0.04 0.11 0.11 0.10 0.12 0.13 0.13
Health [1] 0.88 0.88 0.91 0.89 0.85 0.83 0.86 0.90 1.04 1.04 1.01
Housing 1.16 1.07 1.01 1.03 0.93 0.96 0.95 0.97 0.91 0.89 1.10
General business
0.18 0.18 0.17 0.17 0.22 0.22 0.23 0.21 0.22 0.22 0.22
incentives [2]
Research &
0.06 0.04 0.01 0.01 0.03 0.04 0.03 0.07 0.08 0.03 0.02
development
Specific industry
0.23 0.23 0.24 0.24 0.22 0.23 0.22 0.23 0.25 0.26 0.26
relief [1]
Intergovernmental
0.74 0.76 0.78 0.75 0.82 0.90 0.90 0.91 0.91 0.90 0.79
relations
Charity 0.31 0.33 0.27 0.27 0.28 0.27 0.27 0.38 0.38 0.35 0.30
Other 0.04 0.04 0.03 0.03 0.03 0.03 0.03 0.03 0.03 0.03 0.03
Make work pay
Total 5.06 4.96 4.83 4.96 5.06 5.21 5.22 5.39 5.68 5.56 5.25
Capital income
taxation [3]
Accelerated
0.33 0.44 0.42 0.38 0.41 0.31 0.35 0.42 0.44 0.43 0.36
depreciation
Interest 0.02 0.02 0.02 0.01 0.01 0.01 0.00 0.00 0.00 0.00 0.00
Dividends 0.05 0.06 0.06 0.06 0.06 0.07 0.07 0.05 0.05 0.03 0.02
Capital gains 0.47 0.49 0.49 0.41 0.73 0.72 0.70 0.94 0.81 0.38 0.43
Subtotal 0.87 0.99 0.98 0.85 1.21 1.11 1.12 1.42 1.30 0.84 0.81
Total 5.92 5.95 5.81 5.81 6.28 6.32 6.34 6.81 6.98 6.40 6.06
Make work pay
0.10 0.11 0.10 0.11 0.10 0.08 0.08 0.08 0.07 0.08 0.07
provisions
Total 6.03 6.06 5.91 5.92 6.38 6.40 6.42 6.89 7.05 6.47 6.13
Non-income tax [3]
Grand total 6.03 6.06 5.91 5.92 6.38 6.40 6.42 6.89 7.05 6.47 6.13
Structural items 0.00 0.00 0.00 0.00 0.04 0.21 0.20 0.20 0.21 0.35 0.19
Income tax
expenditures by
type*
Credits [1] 0.26 0.26 0.23 0.23 0.27 0.29 0.29 0.31 0.33 0.32 0.27
Exemptions &
4.85 4.78 4.67 4.55 4.90 4.96 4.95 5.07 5.35 5.12 5.04
allowances [1]
Deferrals [1] 0.77 0.86 0.84 0.77 0.69 0.63 0.68 0.78 0.77 0.77 0.57
Reduced rates [1] 0.14 0.16 0.16 0.37 0.51 0.51 0.49 0.74 0.60 0.28 0.25

Table II.25. Tax expenditures in the United States (% of GDP) cont’d

2005 2006 2007 2008 2009† 2010† 2011† 2012† 2013† 2014†
income tax*
Low-income non-work related 0.12 0.11 0.11 0.11 0.11 0.11 0.11 0.10 0.10 0.10
Retirement 1.04 1.00 1.02 1.02 1.06 1.04 1.09 1.11 1.09 1.07
Work related 0.07 0.07 0.07 0.07 0.07 0.07 0.07 0.07 0.07 0.07
Education 0.14 0.15 0.12 0.13 0.13 0.12 0.12 0.12 0.12 0.11
Health [1] 1.08 1.06 1.08 1.05 1.15 1.21 1.26 1.30 1.33 1.37
Housing 1.21 1.20 1.06 1.05 1.02 1.10 1.23 1.31 1.33 1.36
General business incentives [2] 0.32 0.33 0.28 0.41 0.40 0.43 0.42 0.42 0.40 0.40
Research & development 0.08 0.08 0.11 0.09 0.08 0.06 0.05 0.05 0.05 0.05
Specific industry relief [1] 0.24 0.25 0.24 0.23 0.26 0.26 0.26 0.26 0.26 0.26
Intergovernmental relations 0.67 0.67 0.59 0.63 0.57 0.48 0.65 0.73 0.72 0.70
Charity 0.30 0.35 0.35 0.33 0.38 0.40 0.40 0.41 0.42 0.42
Other 0.03 0.03 0.03 0.09 0.09 0.08 0.08 0.08 0.08 0.07
Make work pay
Total 5.30 5.29 5.06 5.21 5.33 5.37 5.75 5.97 5.96 5.97
Accelerated depreciation 0.16 0.27 0.16 0.35 -0.12 -0.07 -0.06 -0.01 0.02 0.04
Interest 0.01 0.01 0.01 0.01 0.01 0.01 0.01 0.01 0.01 0.01
Dividends 0.03 0.03 0.04 0.02 0.02 0.02 0.02 0.02 0.02 0.02
Capital gains 0.44 0.61 0.63 0.33 0.31 0.34 0.34 0.34 0.36 0.37
Subtotal 0.63 0.92 0.84 0.70 0.21 0.30 0.31 0.36 0.41 0.44
Total 5.93 6.21 5.90 5.91 5.54 5.67 6.06 6.33 6.37 6.42
Make work pay provisions 0.07 0.07 0.06 0.06 0.07 0.06 0.06 0.07 0.06 0.06
Total 6.00 6.27 5.96 5.97 5.61 5.73 6.12 6.39 6.43 6.47
Non-income tax [3]
Grand total 6.00 6.27 5.96 5.97 5.61 5.73 6.12 6.39 6.43 6.47
Structural items 0.34 0.23 0.23 0.20 0.19 0.18 0.13 0.06 0.05 0.05
Credits [1] 0.28 0.25 0.30 0.34 0.31 0.26 0.25 0.25 0.24 0.23
Exemptions & allowances [1] 5.01 5.03 4.73 4.63 4.77 4.83 5.25 5.51 5.53 5.56
Deferrals [1] 0.46 0.58 0.49 0.80 0.34 0.41 0.43 0.46 0.48 0.49
Reduced rates [1] 0.25 0.41 0.44 0.20 0.19 0.23 0.19 0.17 0.18 0.19
† Projections.
[2] Beginning in 2003 lower rates of taxation for dividends and capital gains on corporate equity are
not considered tax expenditures.
12 http://dx.doi.org/10.1787/747140815638

Table II.26. Tax expenditures in the United States (% of central government total tax and non-tax receipts)
2002 2003 2004 2005 2006 2007 2008 2009 † 2010 † 2011 † 2012 † 2013 † 2014 †
Retirement 8.24 9.10 7.29 5.89 5.39 5.42 5.77 7.01 6.58 6.30 5.93 5.78 5.65
Work related 0.69 0.43 0.46 0.40 0.38 0.36 0.38 0.47 0.46 0.42 0.37 0.36 0.35
Education 0.68 0.79 0.77 0.77 0.78 0.65 0.76 0.88 0.78 0.70 0.65 0.62 0.60
Health [1] 5.84 6.33 6.17 6.14 5.70 5.73 5.93 7.57 7.64 7.27 6.97 7.04 7.21
Housing 5.12 5.37 6.74 6.88 6.48 5.63 5.90 6.75 6.92 7.10 7.01 7.04 7.16
Charity 2.13 2.10 1.84 1.71 1.91 1.84 1.88 2.49 2.50 2.33 2.20 2.21 2.23
Other 0.20 0.20 0.21 0.25 0.23 0.23 0.55 0.67 0.56 0.51 0.46 0.44 0.44
Make work pay 0.41 0.47 0.44 0.38 0.35 0.32 0.36 0.47 0.39 0.34 0.35 0.32 0.31
Total 40.67 41.37 38.73 36.01 35.17 32.87 34.78 38.31 37.36 36.05 34.58 34.34 34.47
Interest
Dividends
Capital gains
Subtotal
Total
2002 2003 2004 2005 2006 2007 2008 2009 † 2010 † 2011 † 2012 † 2013 † 2014 †
Total
Non-income tax [3]
Grand total 40.67 41.37 38.73 36.01 35.17 32.87 34.78 38.31 37.36 36.05 34.58 34.34 34.47
Structural items
Credits [1] 3.05 4.04 2.87 3.54 2.63 2.81 3.05 3.28 2.82 2.18 1.66 1.56 1.49
Deferrals [1] 4.32 4.64 3.48 2.62 3.14 2.62 4.53 2.25 2.62 2.45 2.49 2.52 2.58
Reduced rates [1] 3.35 1.68 1.51 1.41 2.24 2.34 1.11 1.27 1.43 1.09 0.92 0.97 1.02
† Projections.

Table II.26. Tax expenditures in the United States


1994 1995 1996 1997 1998 1999 2000 2001 2002 2003 2004
income tax*
Low-income non-work related 0.15 0.16 0.15 0.15 0.15 0.14 0.14 0.14 0.76 0.84 0.78
Retirement 1.15 1.13 1.14 1.30 1.37 1.36 1.37 1.33 8.24 9.10 7.29
Work related 0.12 0.11 0.08 0.07 0.13 0.12 0.12 0.12 0.69 0.43 0.46
Education 0.04 0.04 0.04 0.04 0.04 0.11 0.11 0.10 0.68 0.79 0.77
Health [1] 0.88 0.88 0.91 0.89 0.85 0.83 0.86 0.90 5.84 6.33 6.17
Housing 1.16 1.07 1.01 1.03 0.93 0.96 0.95 0.97 5.12 5.37 6.74
General business incentives [1] 0.18 0.18 0.17 0.17 0.22 0.22 0.23 0.21 1.21 1.36 1.32
Research & development 0.06 0.04 0.01 0.01 0.03 0.04 0.03 0.07 0.46 0.17 0.12
Specific industry relief [1] 0.23 0.23 0.24 0.24 0.22 0.23 0.22 0.23 1.39 1.55 1.56
Intergovernmental relations 0.74 0.76 0.78 0.75 0.82 0.90 0.90 0.91 5.10 5.46 4.86
Charity 0.31 0.33 0.27 0.27 0.28 0.27 0.27 0.38 2.13 2.10 1.84
Other 0.04 0.04 0.03 0.03 0.03 0.03 0.03 0.03 0.17 0.20 0.21
Make work pay
Total 5.06 4.96 4.83 4.96 5.06 5.21 5.22 5.39 31.78 33.69 32.12
Accelerated depreciation 0.33 0.44 0.42 0.38 0.41 0.31 0.35 0.42 2.48 2.59 2.20
Interest 0.02 0.02 0.02 0.01 0.01 0.01 0.00 0.00 0.03 0.00 0.00
Dividends 0.05 0.06 0.06 0.06 0.06 0.07 0.07 0.05 0.26 0.17 0.13
Capital gains 0.47 0.49 0.49 0.41 0.73 0.72 0.70 0.94 4.51 2.31 2.64
Subtotal 0.87 0.99 0.98 0.85 1.21 1.11 1.12 1.42 7.28 5.08 4.97
Total 5.92 5.95 5.81 5.81 6.28 6.32 6.34 6.81 39.06 38.77 37.10
Make work pay provisions 0.10 0.11 0.10 0.11 0.10 0.08 0.08 0.08 0.41 0.47 0.44
Total 6.03 6.06 5.91 5.92 6.38 6.40 6.42 6.89 39.48 39.24 37.53
Non-income tax [3]
Grand total 6.03 6.06 5.91 5.92 6.38 6.40 6.42 6.89 39.48 39.24 37.53
Structural items 0.00 0.00 0.00 0.00 0.04 0.21 0.20 0.20 1.20 2.13 1.19
Income tax expenditures by
type*
Credits [1] 0.26 0.26 0.23 0.23 0.27 0.29 0.29 0.31 1.85 1.91 1.68
Exemptions & allowances [1] 4.85 4.78 4.67 4.55 4.90 4.96 4.95 5.07 29.96 31.00 30.87
Deferrals [1] 0.77 0.86 0.84 0.77 0.69 0.63 0.68 0.78 4.32 4.64 3.48
Reduced rates [1] 0.14 0.16 0.16 0.37 0.51 0.51 0.49 0.74 3.35 1.68 1.51

Table II.26. Tax expenditures in the United States


2005 2006 2007 2008 2009† 2010† 2011† 2012† 2013† 2014†
income tax*
Low-income non-work
0.69 0.61 0.57 0.61 0.73 0.67 0.61 0.55 0.54 0.53
related
Retirement 5.89 5.39 5.42 5.77 7.01 6.58 6.30 5.93 5.78 5.65
Work related 0.40 0.38 0.36 0.38 0.47 0.46 0.42 0.37 0.36 0.35
Education 0.77 0.78 0.65 0.76 0.88 0.78 0.70 0.65 0.62 0.60
Health [1] 6.14 5.70 5.73 5.93 7.57 7.64 7.27 6.97 7.04 7.21
Housing 6.88 6.48 5.63 5.90 6.75 6.92 7.10 7.01 7.04 7.16
General business incentives
1.83 1.77 1.48 2.29 2.63 2.74 2.42 2.24 2.13 2.09
[1]
Charity 1.71 1.91 1.84 1.88 2.49 2.50 2.33 2.20 2.21 2.23
Other 0.19 0.18 0.18 0.50 0.61 0.51 0.46 0.41 0.40 0.39
Make work pay
Total 30.11 28.59 26.87 29.36 35.19 33.91 33.16 31.95 31.57 31.56
Interest 0.06 0.05 0.05 0.05 0.06 0.06 0.05 0.05 0.05 0.04
Dividends 0.15 0.17 0.21 0.10 0.11 0.12 0.12 0.10 0.10 0.10
Capital gains 2.48 3.27 3.37 1.84 2.04 2.16 1.95 1.84 1.90 1.97
Subtotal 3.59 4.97 4.48 3.94 1.39 1.90 1.80 1.95 2.15 2.33
Total 33.69 33.56 31.35 33.30 36.57 35.81 34.96 33.90 33.72 33.90
Total 34.07 33.91 31.67 33.65 37.05 36.20 35.30 34.25 34.05 34.21
Non-income tax [3]
Grand total 34.07 33.91 31.67 33.65 37.05 36.20 35.30 34.25 34.05 34.21
Structural items 1.94 1.26 1.20 1.13 1.26 1.16 0.75 0.33 0.29 0.26
type*
Credits [1] 1.60 1.37 1.60 1.92 2.02 1.66 1.43 1.33 1.27 1.23
Deferrals [1] 2.62 3.14 2.62 4.53 2.25 2.62 2.45 2.49 2.52 2.58
Reduced rates [1] 1.41 2.24 2.34 1.11 1.27 1.43 1.09 0.92 0.97 1.02
† Projections.
12 http://dx.doi.org/10.1787/747140815638

Table II.27. Tax expenditures in the United States (% of relevant tax revenue) † ‡
2002 2003 2004 2005 2006 2007 2008 2009* 2010* 2011* 2012* 2013* 2014*
Retirement 15.17 17.53 13.72 10.53 9.28 9.08 10.04 13.75 12.48 11.15 10.38 9.96 9.68
Work related 1.26 0.84 0.86 0.71 0.66 0.61 0.66 0.92 0.88 0.74 0.65 0.62 0.61
Education 1.25 1.52 1.46 1.38 1.35 1.09 1.32 1.72 1.49 1.24 1.14 1.07 1.04
Health [1] 10.75 12.19 11.63 10.97 9.82 9.60 10.33 14.84 14.49 12.87 12.19 12.13 12.35
Housing 9.43 10.35 12.69 12.28 11.15 9.44 10.27 13.24 13.12 12.56 12.26 12.13 12.26
Charity 3.92 4.05 3.46 3.06 3.29 3.09 3.27 4.89 4.73 4.13 3.84 3.81 3.82
Other 0.37 0.38 0.39 0.46 0.40 0.39 0.96 1.32 1.07 0.90 0.81 0.76 0.75
Make work pay 0.76 0.90 0.82 0.68 0.61 0.54 0.62 0.93 0.73 0.59 0.61 0.56 0.53
Total 74.91 79.68 72.94 64.34 60.58 55.04 60.55 75.13 70.83 63.81 60.45 59.17 59.01
Interest
Dividends
Capital gains
Subtotal
2002 2003 2004 2005 2006 2007 2008 2009* 2010* 2011* 2012* 2013* 2014*
Total
Total
Non-income tax [3]
Credits [1] 5.61 7.79 5.40 6.33 4.53 4.70 5.31 6.43 5.34 3.87 2.90 2.68 2.55
Deferrals [1] 7.96 8.93 6.55 4.68 5.41 4.39 7.89 4.40 4.96 4.34 4.35 4.34 4.41
Reduced rates [1] 6.17 3.24 2.85 2.52 3.85 3.91 1.94 2.50 2.72 1.93 1.61 1.68 1.75
* Projections.

Table II.27. Tax expenditures in the United States (% of relevant tax revenue) † ‡
cont’d

1994 1995 1996 1997 1998 1999 2000 2001 2002 2003 2004
income tax**
Low-income non-work
0.15 0.16 0.15 0.15 0.15 0.14 0.14 0.14 1.40 1.62 1.48
related
Retirement 1.15 1.13 1.14 1.30 1.37 1.36 1.37 1.33 15.17 17.53 13.72
Work related 0.12 0.11 0.08 0.07 0.13 0.12 0.12 0.12 1.26 0.84 0.86
Education 0.04 0.04 0.04 0.04 0.04 0.11 0.11 0.10 1.25 1.52 1.46
Health [1] 0.88 0.88 0.91 0.89 0.85 0.83 0.86 0.90 10.75 12.19 11.63
Housing 1.16 1.07 1.01 1.03 0.93 0.96 0.95 0.97 9.43 10.35 12.69
General business
0.18 0.18 0.17 0.17 0.22 0.22 0.23 0.21 2.23 2.61 2.49
incentives [2]
Research & development 0.06 0.04 0.01 0.01 0.03 0.04 0.03 0.07 0.85 0.32 0.24
Specific industry relief [1] 0.23 0.23 0.24 0.24 0.22 0.23 0.22 0.23 2.57% 2.98 2.94
Intergovernmental relations 0.74 0.76 0.78 0.75 0.82 0.90 0.90 0.91 9.38 10.52 9.15
Charity 0.31 0.33 0.27 0.27 0.28 0.27 0.27 0.38 3.92 4.05 3.46
Other 0.04 0.04 0.03 0.03 0.03 0.03 0.03 0.03 0.32 0.38 0.39
Make work pay
Total 5.06 4.96 4.83 4.96 5.06 5.21 5.22 5.39 58.53 64.90 60.50
Accelerated depreciation 0.33 0.44 0.42 0.38 0.41 0.31 0.35 0.42 4.57 5.00 4.15
Interest 0.02 0.02 0.02 0.01 0.01 0.01 0.00 0.00 0.05 0.00 0.01
Dividends 0.05 0.06 0.06 0.06 0.06 0.07 0.07 0.05 0.48 0.33 0.25
Capital gains 0.47 0.49 0.49 0.41 0.73 0.72 0.70 0.94 8.31 4.45 4.96
Subtotal 0.87 0.99 0.98 0.85 1.21 1.11 1.12 1.42 13.41 9.78 9.36
Total 5.92 5.95 5.81 5.81 6.28 6.32 6.34 6.81 71.94 74.68 69.87
Make work pay provisions 0.10 0.11 0.10 0.11 0.10 0.08 0.08 0.08 0.76 0.90 0.82
Total 6.03 6.06 5.91 5.92 6.38 6.40 6.42 6.89 72.70 75.57 70.69
Non-income tax [3]
type**
Credits [1] 0.26 0.26 0.23 0.23 0.27 0.29 0.29 0.31 3.41 3.69 3.16
Exemptions & allowances
4.85 4.78 4.67 4.55 4.90 4.96 4.95 5.07 55.17 59.71 58.13
[1]
Deferrals [1] 0.77 0.86 0.84 0.77 0.69 0.63 0.68 0.78 7.96 8.93 6.55
Reduced rates [1] 0.14 0.16 0.16 0.37 0.51 0.51 0.49 0.74 6.17 3.24 2.85

Table II.27. Tax expenditures in the United States (% of relevant tax revenue) †‡
cont’d

2005 2006 2007 2008 2009 * 2010 * 2011* 2012* 2013* 2014*
income tax**
Low-income non-work related 1.23 1.04 0.95 1.06 1.43 1.28 1.08 0.96 0.92 0.90
Retirement 10.53 9.28 9.08 10.04 13.75 12.48 11.15 10.38 9.96 9.68
Work related 0.71 0.66 0.61 0.66 0.92 0.88 0.74 0.65 0.62 0.61
Education 1.38 1.35 1.09 1.32 1.72 1.49 1.24 1.14 1.07 1.04
Health [1] 10.97 9.82 9.60 10.33 14.84 14.49 12.87 12.19 12.13 12.35
Housing 12.28 11.15 9.44 10.27 13.24 13.12 12.56 12.26 12.13 12.26
General business incentives
3.28 3.04 2.47 3.99 5.15 5.20 4.28 3.92 3.68 3.58
[2]
Charity 3.06 3.29 3.09 3.27 4.89 4.73 4.13 3.84 3.81 3.82
Other 0.34 0.31 0.30 0.87 1.20 0.96 0.81 0.72 0.69 0.67
Make work pay
Total 53.79 49.24 45.00 51.10 69.00 64.30 58.70 55.86 54.41 54.04
Interest 0.11 0.09 0.08 0.09 0.12 0.11 0.09 0.08 0.08 0.07
Dividends 0.26 0.29 0.35 0.17 0.22 0.23 0.20 0.18 0.17 0.16
Capital gains 4.43 5.64 5.64 3.21 3.99 4.10 3.46 3.22 3.27 3.37
Subtotal 6.41 8.56 7.50 6.86 2.72 3.60 3.19 3.40 3.71 3.99
Total 60.20 57.79 52.49 57.97 71.72 67.90 61.89 59.27 58.11 58.03
Total 60.88 58.40 53.03 58.59 72.65 68.63 62.48 59.88 58.67 58.56
Non-income tax [3]
type**
Credits [1] 2.86 2.36 2.68 3.35 3.95 3.14 2.53 2.33 2.18 2.10
Deferrals [1] 4.68 5.41 4.39 7.89 4.40 4.96 4.34 4.35 4.34 4.41
Reduced rates [1] 2.52 3.85 3.91 1.94 2.50 2.72 1.93 1.61 1.68 1.75
* Projections.
12 http://dx.doi.org/10.1787/747140815638

Table II.28. Number of tax expenditures in the United States (% of GDP)
2002 † 2003 2004 2005 2006 2007 2008 2009‡ 2010‡
tax*
General tax relief 1 1 1 1 1 1 1 1 1
Low-income non-work related 11 11 11 11 11 11 11 11 11
Retirement 11 11 11 10 10 10 10 10 10
Work related 10 9 9 9 9 9 10 10 10
Education 14 15 15 16 16 16 16 16 16
Health 8 8 8 8 8 9 9 9 9
Housing 8 8 9 9 9 9 11 11 11
General business incentives 22 22 22 23 24 24 24 24 24
Research & development 2 2 2 2 2 2 2 2 2
Specific industry relief 34 35 35 43 50 52 54 54 54
Intergovernmental relations 3 3 3 3 3 3 3 3 3
Charity 4 4 4 4 4 4 5 5 5
Other 4 4 4 4 5 5 5 5 5
Total 132 133 134 143 152 155 161 161 161
Interest
Dividends
Capital gains
Subtotal
Total
Make work pay provisions 4 4 4 4 4 4 4 4 4
Total 136 137 138 147 156 159 165 165 165
Non-income tax related 0 0 0 0 0 0 0 0 0
Grand total 136 137 138 147 156 159 165 165 165
Structural items 0 0 0 0 0 0 0 0 0
Credits 29 29 29 32 36 37 39 39 39
80 81 82 88 91 92 96 96 96
exclusions
Deferrals 22 22 22 22 24 25 25 25 25
Reduced rates 5 5 5 5 5 5 5 5 5
† In fiscal years: fiscal year 2006 is from 1 October 2005 to 30 September 2006.
‡ Projection.
* Classification of tax expenditures by purpose and by type is to some degree arbitrary.

Table II.28. Number of tax expenditures in the United States (% of GDP) cont’d

2002 † 2003 2004 2005 2006 2007 2008 2009‡ 2010‡
General tax relief 0 0 0 0 0 0 0 0 0
Low-income non-work related 11 11 11 11 11 11 11 11 11
Retirement 11 11 11 10 10 10 10 10 10
Work related 10 9 9 9 9 9 10 10 10
Education 14 15 15 16 16 16 16 16 16
Health 8 8 8 8 8 9 9 9 9
Housing 8 8 9 9 9 9 11 11 11
General business incentives 16 16 16 17 18 18 18 18 18
Research & development 2 2 2 2 2 2 2 2 2
Specific industry relief 34 35 35 43 50 52 54 54 54
Intergovernmental relations 3 3 3 3 3 3 3 3 3
Charity 4 4 4 4 4 4 5 5 5
Other 3 3 3 3 4 4 4 4 4
Total 124 125 126 135 144 147 153 153 153
Accelerated depreciation 2 2 2 2 2 2 2 2 2
Interest 1 1 1 1 1 1 1 1 1
Dividends 1 1 1 1 1 1 1 1 1
Capital gains 3 3 3 3 3 3 3 3 3
Subtotal 7 7 7 7 7 7 7 7 7
Total 131 132 133 142 151 154 160 160 160
Make work pay provisions 4 4 4 4 4 4 4 4 4
Total 135 136 137 146 155 158 164 164 164
Non-income tax related 0 0 0 0 0 0 0 0 0
Grand total 135 136 137 146 155 158 164 164 164
Structural items 1 1 1 1 1 1 1 1 1
Credits 28 28 28 31 35 36 38 38 38
Deductions, exemptions & exclusions 80 81 82 88 91 92 96 96 96
Deferrals 22 22 22 22 24 25 25 25 25
Reduced rates 5 5 5 5 5 5 5 5 5
† In fiscal years: fiscal year 2006 is from 1 October 2005 to 30 September 2006.
‡ Projection.
* Classification of tax expenditures by purpose and by type is to some degree arbitrary.
Source: Budget of the U.S. Government, Fiscal Years 2009 and 2010, Analytical Perspectives,
Chapter 19, Table 19-1.
12 http://dx.doi.org/10.1787/747140815638

Table II.29. International comparison of tax expenditures (% of GDP) †

Latest actual year available
United United
Canada Germany Korea Netherlands Spain
Kingdom States
(2004) (2006) (2006) (2006) (2008)
(2006) (2008)
tax*
Retirement 1.68 0.00 0.02 0.06 0.17 2.32 1.02
Work related 0.39 0.03 0.03 0.06 0.01 0.15 0.07
Education 0.12 0.00 0.12 0.06 0.00 0.00 0.13
Health 0.27 0.00 0.29 0.00 0.00 0.00 1.05
Housing 0.20 0.18 0.05 0.05 0.41 1.20 1.05
Intergovernmental relations 1.55 0.03 0.00 0.00 0.00 0.00 0.63
Charity 0.21 0.00 0.13 0.09 0.02 0.09 0.33
Other 0.02 0.00 0.02 0.01 0.17 0.12 0.09
Total 5.16 0.26 1.75 1.06 1.41 4.90 5.21
Accelerated depreciation 0.00 0.00 0.00 0.00 0.00 1.40 0.35
Interest 0.00 0.00 0.00 0.00 0.00 0.02 0.01
Dividends 0.27 0.04 0.00 0.00 0.00 0.00 0.02
Capital gains 0.35 0.00 0.00 0.00 0.16 1.65 0.33
Subtotal 0.62 0.04 0.00 0.00 0.16 3.07 0.70
Total 5.77 0.29 1.75 1.06 1.57 7.97 5.91
Make work pay provisions 0.01 0.00 0.01 0.04 0.74 0.35 0.06
Total 5.78 0.29 1.76 1.10 2.31 8.32 5.97
Non-income tax related 1.16 0.45 0.72 0.90 2.25 4.47 0.00
Total 6.94 0.74 2.48 2.00 4.55 12.79 5.97
Credits 1.44 0.00 0.02 0.06 0.34 1.52 0.34
2.64 0.28 1.70 0.80 1.61 4.92 4.63
exclusions
Deferrals 1.50 0.00 0.00 0.05 0.00 1.47 0.80
Reduced rates 0.21 0.01 0.04 0.19 0.36 0.41 0.20
† For every country except for Canada and Spain, fiscal years rather than calendar years are used. For
the United Kingdom, fiscal year 2006-07 is used (from 6 April 2006 to 5 April 2007).
12 http://dx.doi.org/10.1787/747181561388

Table II.30. International comparison of tax expenditures

(% of central government total tax and non-tax receipts) †
Latest year available
United United
Kingdom States
(2004) (2006) (2006) (2006) (2008)
(2006) (2008)
income tax*
Retirement 10.72 0.05 0.10 0.16 0.46 6.38 5.77
Work related 2.47 0.36 0.16 0.17 0.03 0.42 0.38
Education 0.78 0.00 0.67 0.16 0.01 0.00 0.76
Health 1.70 0.00 1.67 0.00 0.00 0.00 5.93
Housing 1.29 2.01 0.29 0.12 1.12 3.30 5.90
Charity 1.32 0.00 0.76 0.22 0.04 0.25 1.88
Other 0.13 0.00 0.09 0.02 0.46 0.32 0.50
Total 32.97 2.91 10.09 2.74 3.86 13.47 29.36
Interest 0.00 0.00 0.00 0.00 0.00 0.05 0.05
Dividends 1.70 0.42 0.00 0.00 0.00 0.00 0.10
Capital gains 2.23 0.00 0.00 0.00 0.44 4.54 1.84
Subtotal 3.93 0.42 0.02 0.00 0.44 8.45 3.94
Total 36.90 3.33 10.11 2.74 4.30 21.92 33.30
Total 36.94 3.33 10.16 2.84 6.32 22.87 33.65
Non-income tax related 7.43 5.16 4.18 2.31 6.16 12.30 0.00
Total 44.37 8.48 14.34 5.15 12.48 35.17 33.65
Credits 9.18 0.00 0.11 0.14 0.92 4.18 1.92
16.86 3.24 9.79 2.07 4.41 13.54 26.09
exclusions
Deferrals 9.56 0.02 0.02 0.14 0.00 4.03 4.53
Reduced rates 1.34 0.07 0.24 0.49 0.99 1.12 1.11
† For every country except for Canada, fiscal years rather than calendar years are used. For the United
Kingdom, fiscal year 2006-07 is used (from 6 April 2006 to 5 April 2007).
12 http://dx.doi.org/10.1787/747181561388

Table II.31. International comparison of tax expenditures

(% of relevant tax revenue)†‡*
Lastest actual year available
United United
Kingdom States
(2004) (2006) (2006)** (2006) (2008)
(2006) (2008)
income tax***
Retirement 17.23 0.14 0.23 0.55 2.07 10.83 10.04
Work related 3.96 0.96 0.39 0.58 0.12 0.71 0.66
Education 1.25 0.00 1.64 0.58 0.05 0.01 1.32
Health 2.73 0.00 4.06 0.00 0.00 0.00 10.33
Housing 2.07 5.33 0.71 0.42 4.98 5.61 10.27
Charity 2.13 0.00 1.85 0.78 0.20 0.42 3.27
Other 0.20 0.00 0.22 0.07 2.05 0.55 0.87
Total 52.97 7.71 24.56 9.60 17.21 22.86 51.10
Interest 0.00 0.00 0.00 0.00 0.00 0.08 0.09
Dividends 2.73 1.10 0.00 0.00 0.00 0.00 0.17
Capital gains 3.59 0.00 0.00 0.00 1.97 7.72 3.21
Subtotal 6.32 1.10 0.05 0.00 1.97 14.35 6.86
Total 59.30 8.81 24.60 9.60 19.18 37.22 57.97
Total 59.36 8.81 24.73 9.95 28.16 38.84 58.59
VAT or sales tax 52.38 1.54 9.98 9.74 59.69 54.66
Excises [1] 4.41
Heavy motor vehicle tax [1] 0.00
Motor vehicle tax [1] 4.96
Regulating energy tax [1] 3.61
Special excise on motor vehicles
2.18
[1]
Tax on the sale of immovable
2.02
property [1]
Electricity tax [2] 62.27
Fuel tax [2] 13.52
Spirits tax [2] 0.28
Tobacco tax [2] 0.05
Stamp tax [3] 8.76
Inheritance and gift tax [3] 1.64
Educational tax [3] 6.81
Security transaction tax [3] 7.93
Special excise tax [3] 5.28
Liquor tax [3] 1.90
Customs duties [3] 4.39
Transportation tax [3] 12.64

United United
Kingdom States
(2004) (2006) (2006)** (2006) (2008)
(2006) (2008)
Stamp duties [4] 18.47
Hydrocarbon oils [4] 0.00
Other [4] 3.08
Tributes [5] 6.78
Insurance tax [5] 23.65
Alcohol and by-product
10.05
beverages Tax [5]
Non-residents equity tax [5] 0.00
Hydrocarbons tax [5] 19.45
type***
Credits 14.76 0.01 0.27 0.51 4.11 7.09 3.35
27.09 8.58 23.84 7.27 19.65 23.00 45.41
exclusions
Deferrals 15.36 0.04 0.05 0.47 0.00 6.84 7.89
Reduced rates 2.15 0.18 0.57 1.70 4.40 1.91 1.94
‡ Percent of tax revenue by type of tax.
* Individual and corporate income taxes are considered together. For the United Kingdom, capital
gains taxes and National Insurance Contributions are also included in this grouping.
** For Korea, fiscal year 2006 is used.
[1] Only reported in the Netherlands.
[2] Only reported in Germany.
[3] Only reported in Korea.
[4] Only reported in the United Kingdom.
[5] Only reported in Spain.
12 http://dx.doi.org/10.1787/747181561388

Table II.32. International comparison of number of tax expenditures (% of GDP) †

Latest actual year available
United United
Kingdom States
(2004) (2006) (2006) (2006) (2008)
(2006) ‡ (2008)
income tax*
Retirement 13 1 2 2 3 16 10
Work related 11 2 4 6 3 37 10
Education 9 1 5 2 2 4 16
Health 5 0 3 0 1 4 9
Housing 1 10 12 2 3 7 11
Charity 13 0 4 6 5 6 5
Other 8 1 11 2 10 27 4
Total 141 53 134 52 68 187 153
Accelerated depreciation 1 0 1 0 0 2 2
Interest 0 0 0 0 0 6 1
Dividends 3 3 0 0 0 2 1
Capital gains 3 0 0 0 2 8 3
Subtotal 7 3 1 0 2 18 7
Total 148 56 135 52 70 205 160
Make work pay provisions 1 0 1 2 5 3 4
Total 149 56 136 54 75 208 164
Non-income tax related 32 30 82 46 64 173 0
Total 181 86 218 100 139 381 164
Structural items 32 0 2 1 2 8 1
Credits 33 2 2 7 15 5 38
73 46 120 34 57 186 96
exclusions
Deferrals 35 4 7 6 0 11 25
‡ Given reporting practices, some of these tax expenditures may have gone into effect only in 2007.
12 http://dx.doi.org/10.1787/747181561388

Figure II.1. Income tax expenditure by purpose in Canada

Percent of total, latest actual year available
Make Work
Pay; 0 Capital
Work Other; 6 Income
related; 7 Taxation; 11
General
Specific business
industry incentives; 7
relief; 1
Health; 5
Housing; 3
Retirement;
29
Inter-
governmental
R&D; 4 relations; 27
12 http://dx.doi.org/10.1787/746827562747
Figure II.2. Income tax expenditure by purpose in Germany

Make Work Pay Capital income

provisions; 0 taxation; 13
Other; 0
Specific Work General
industry related; 11 business
relief; 4 incentives; 1
Retirement; 2 Health; 0
R&D; 0
Inter-
governmental
relations; 9
Housing; 60
12 http://dx.doi.org/10.1787/746827562747

Figure II.3. Income tax expenditure by purpose in Korea

Capital income
taxation; 0
Other; 20
General business
Make Work Pay incentives; 39
provisions; 1
Work related; 2
Specific industry
relief; 10
Retirement; 1
R&D; 9
Intergovernmental
relations; 0 Health; 16
Housing; 3
12 http://dx.doi.org/10.1787/746827562747
Figure II.4. Income tax expenditure by purpose in the Netherlands

Capital income
taxation; 0
Other; 14
Make Work Pay

provisions; 3
General business
incentives; 43
Work Related; 6
Specific industry
relief; 16
Retirement; 6 Health; 0
R&D; 7 Housing; 4
Intergovernmental
relations; 0
12 http://dx.doi.org/10.1787/746827562747

Figure II.5. Income tax expenditure by purpose in Spain

Capital
income
taxation; 7
Other; 10 General
business
incentives;
23
Make Work
Pay; 32
Health; 0
Housing; 18
Work
related; 0 Intergovern-
Specific R&D; 2 mental
industry relations; 0
Retirement; 7
relief; 2
12 http://dx.doi.org/10.1787/746827562747
Figure II.6. Income tax expenditure by purpose in the United Kingdom

Make Work Pay

provisions; 4 Other; 4
Work related; 2
Specific
industry relief; Capital income
1 taxation; 37
Retirement; 28
R&D; 0
General
Intergovern- business
mental incentives; 9
relations; 0 Housing; 14 Health; 0
12 http://dx.doi.org/10.1787/746827562747

Figure II.7. Income tax expenditure by purpose in the United States

Other; 11 Capital income

Make Work taxation; 12
Pay; 1
General
Work related; 1 business
incentives; 7
Specific industry
relief; 4
Retirement; 17
Health; 18
R&D; 1
Intergovern-
mental relations; Housing; 18
11
12 http://dx.doi.org/10.1787/746827562747

Figure II.8. Number of tax expenditures
450
400 Other taxes

350
300 Income tax

250
200
150
100
50
12 http://dx.doi.org/10.1787/746827562747
Figure II.9. Number of income tax expenditures
250
Specific industry relief
200
150
Other
100
50
0
7
8
6
8
6
4
0
00
00
0
00
0
00
6-
20
20
2
2
,2
,2
00
s,
y,
s,
n,
ea
a
,2
te
nd
an
ai
ad
or
om
a
Sp
m
rla
an
St
K
er
gd
C
he
d
G
te
in
et
ni
K
N
U
d
te
ni
U
12 http://dx.doi.org/10.1787/746827562747

Figure II.10. Income tax expenditures (% of GDP)
9
8
7
Percent of GDP
6
5
4
3
2
1
0
8
6
6
8
4
00
00
0
00
0
00
6-
20
20
,2
2
,2
,2
00
y,
s,
n,
es
a
a
,2
an
nd
re
ai
ad
at
om
Sp
Ko
m
r la
an
St
er
gd
C
he
d
G
te
in
et
ni
K
N
U
d
te
ni
U
12 http://dx.doi.org/10.1787/746827562747
Figure II.11. Income tax expenditures (% of income tax revenue )

Percent of income tax revenue
70
60
50
40
30
20
10
0
6
07
8
6
4
00
00
00
0
0
00
6-
20
20
,2
2
,2
,2
00
s,
n,
y,
es
a
a
,2
nd
an
re
ai
ad
at
m
Sp
Ko
m
rla
an
St
do
er
C
he
ng
G
te
et
ni
Ki
N
d
te
ni
U
12 http://dx.doi.org/10.1787/746827562747

Figure II.12. Income tax expenditures (% of GDP)
9
8
Percent of GDP
7
6
5
4
3
2
1
0
General Relief Low-Income Retirement Work Related Make Work Pay

Education Health Housing Capital Taxation General Business
Specific Industry R&D Intergovernmental Charity Other
12 http://dx.doi.org/10.1787/746827562747
Figure II.13. All tax expenditures (% of GDP)
14
12 Other taxes
Percent of GDP
10
8
Income taxes
6
4
2
0
7
8
6
6
8
4
00
00
00
0
0
00
6-
20
20
,2
2
,2
,2
00
y,
s,
n,
es
ea
a
,2
an
nd
ai
ad
at
or
om
Sp
m
an
rla
St
K
er
gd
C
he
d
G
te
et
n
Ki
ni
N
U
d
te
ni
U
12 http://dx.doi.org/10.1787/746827562747

Figure II.14. All tax expenditures (% of total tax revenue)
50
Percent of total tax revenue
45
40
35
30
25
20
15
10
5
0
8
6
8
4
00
0
00
00
0
00
6-
20
20
2
2
,2
,2
00
s,
s,
y,
n,
ea
a
,2
te
nd
an
ai
ad
or
a
m
Sp
m
la
an
St
K
do
er
er
C
d
ng
h
G
te
et
Ki
ni
N
U
d
te
ni
U
12 http://dx.doi.org/10.1787/746827562747
Figure II.15. Canada's “memorandum items”
9
8
7
Percent of GDP
6
5
4
3
2
1
0
T ax expenditures Memorandum items
12 http://dx.doi.org/10.1787/746827562747

Figure II.16. Cost of ten largest tax expenditures
1,6
1,4
1,2 United Kingdom

Canada
P e rc e n t o f G D P
1,0
0,8 United States
0,6
0,4 Netherlands
Germany
0,2 Korea
0,0
1 2 3 4 5 6 7 8 9 10
12 http://dx.doi.org/10.1787/746827562747
Figure II.17. Intensity of use of tax expenditures

Tax expenditures as percent of tax revenues
70
60
50
40
30
20
10
0
Canada, Germany, Korea, 2006 Netherlands, Spain, 2008 United
2004 2006 2006 Kingdom
2006-07
12 http://dx.doi.org/10.1787/746827562747

Data sources
Canada
Ministry of Finance (2007), Tax Expenditures and Evaluations 2007, Table 1 – Personal Income Tax
Expenditures, Table 2 – Corporate Income Tax Expenditures, Table 3 – Goods and Services Tax
Expenditures, www.fin.gc.ca/purl/taxexp-e.html.
Ministry of Finance (2007), 2007 Economic Statement, Chapter 2: “Fiscal Projections,” Table 2.4
Revenue Outlook, www.fin.gc.ca/ec2007/ec/ecc2e.html.
Ministry of Finance (2007), “Fiscal Reference Tables,” Tables 3 and 6,

www.fin.gc.ca/frt/2007/frt07_2e.html.
Germany
Ministry of Finance, 21st Subsidy Report of the Federal Government: Development of Financial Aid
and Tax Relief Measures of the Federal Government from 2005-2008, pp. 71-92, “Appendix 2:
Overview of the Development of Tax Advantages in the Years 2005 to 2008”.
Ministry of Finance, 20th Subsidy Report of the Federal Government: Development of Financial Aid
and Tax Relief Measures of the Federal Government from 2003-2006, pp. 79-109, “Appendix 2:
Overview of the Development of Tax Advantages in the Years 2003 to 2006.”
Ministry of Finance, 18th Subsidies Report of the Federal Government (Summary): Development of
Financial Aid and Tax Relief Measures of the Federal Government from 1999 to 2002, pp. 2-4.
Ministry of Finance (2008), “Development of Tax Revenue: Overview of the Development of Tax
Revenue,” Tables 3 and 4.
Korea
Republic of Korea (2007), 2007 Tax Expenditure Report, pp. 23-82.
Ministry of Strategy and Finance, Annual Tax Revenues FY 98-FY 2007,

www.mpb.go.kr/eng/mpb_data/statistics/list.jsp?board_no=129.

National Tax Service (2007), Tables 1-1 and 2-1-2,

www.nts.go.kr/eng/resources/resour_31.asp?minfoKey=MINF7520080211223206.
Netherlands
Ministry of Finance, 2008 Budget Memorandum, Chapter 5, Tax Expenditures, Tables 5.3.1 and 5.3.2.
Ministry of Finance, 2003 Budget Memorandum, table: Estimates of Tax Expenditures in the Taxes on
Income, Profits and Property, and table: Estimates of Tax Expenditures in Indirect Taxes.
Ministry of Finance, “Facts and Figures National Finance Annual Report 2006,”
www.minfin.nl/binaries/minfin/assets/pdf/engelse-site/key-topics/budget/facts-and-figures-national-
finance-anual-report-20.pdf.
Ministry of Finance (2007), “Total Government Income in 2008,”

www.minfin.nl/en/subjects,budget/facts-and-figures/Government-income.html.
Ministry of Finance (2007), “Income and Expenditure by the Public Sector in 2008,”
www.minfin.nl/en/subjects,budget/facts-and-figures/Public-sector.html.
Ministry of Finance, “Budget Memorandum 2007: EMU surplus 0.2% GDP” news release,
19 September 2006, www.minfin.nl/en/actual/newsrealeases,2006/09/Budget-Memorandum-2007--
EMU-surplus-0-2--GDP.html.
Spain
Ministerio de Economía y Hacienda, Presupuestos Generales del Estado, Memoria de Beneficios

Fiscales 2009, www.sgpg.pap.meh.es/Presup/PGE2009Proyecto/MaestroDocumentos/PGE-
ROM/N_09_A_A_1B.htm.
Ministerio de Economía y Hacienda, Presupuestos Generales del Estado, Memoria de Beneficios

Fiscales 2008, www.sgpg.pap.meh.es/Presup/PGE2008Proyecto/PGE-ROM/N_08_S_A_1B.htm.
Ministerio de Economía y Hacienda, Presupuestos Generales del Estado, Variaciones en la Estructura

por Políticas de Gastos de los Presupuestos Generales del Estado,
www.sgpg.pap.meh.es/Presup/PGE2009Proyecto/MaestroDocumentos/PGE-ROM/doc/3/3/2/1/N_
09_A_A_2_2_0_1.PDF.
United Kingdom
HM Treasury, Financial Statement and Budget Report 2008, Chapter A: “Budget Policy Decisions,”
pp. 134-137, Table A3.1 – Estimated Costs of Principal Tax Expenditures and Structural Reliefs,
www.hm-treasury.gov.uk/media/2/5/bud08_chaptera.pdf.
HM Revenue and Customs, “Table B.1 – Cost of Minor Tax Allowances and Reliefs,” October 2007,
www.hmrc.gov.uk/stats/tax_expenditures/menu.htm.

A corrigendum has been issued for this page. See: http://www.oecd.org/dataoecd/48/15/44439825.pdf
United Kingdom, HM Revenue and Customs, “Table B.2 – Tax Allowances and Reliefs in Force in
2006-07 or 2007-08, Cost Not Known,” www.hmrc.gov.uk/stats/tax_expenditures/menu.htm.
United Kingdom, HM Treasury, Financial Statement and Budget Report 2008, Chapter C, p. 187, Table
C6: Current Receipts, www.hm-treasury.gov.uk/media/7/3/bud08_chapterc.pdf.
United Kingdom, HM Treasury, Financial Statement and Budget Report 2003, Chapter A: Budget
Policy Decisions, Table A3.1 – Estimated Costs of Principal Tax Expenditures and Structural Reliefs,
www.hm-treasury.gov.uk/budget/bud_bud03/budget_report/bud_bud03_repa.cfm.
United States
United States, Office of Budget and Management, Analytical Perspectives, Budget of the U.S.
Government, Fiscal Year 2009, Chapter 19, pp. 293-296, Table 19-2 – Estimates of Tax Expenditures
for Corporate and Individual Income Taxes, www.whitehouse.gov/omb/budget/fy2009/apers.html.
for Corporate and Individual Income Taxes, www.whitehouse.gov/omb/budget/fy2008/apers.html.
for Corporate and Individual Income Taxes, www.whitehouse.gov/omb/budget/fy2007/pdf/spec.pdf.
Government, Fiscal Year 2004, Chapter 6, pp. 106-109, Table 6-2 – Estimates of Tax Expenditures for
Corporate and Individual Income Taxes, www.whitehouse.gov/omb/budget/fy2004/pdf/spec.pdf.
Government, Fiscal Year 2002, Chapter 5, pp. 66-70, Table 5-2 – Estimates of Tax Expenditures for
Corporate and Individual Income Taxes, www.whitehouse.gov/omb/budget/fy2002/spec.pdf.
United States, Office of Budget and Management, Historical Tables, Budget of the U.S. Government,
Fiscal Year 2009, Section 2, pp. 30-31, Table 2.1 – Receipts by Source: 1934-2013,
www.whitehouse.gov/omb/budget/fy2009/.
United States, Office of Budget and Management, Historical Tables, Budget of the U.S. Government,
Fiscal Year 2009, Section 1, pp. 24-25, Table 1.2 – Summary of Receipts, Outlays, and Surpluses or
Deficits (-) as Percentages of GDP: 1930-2013, www.whitehouse.gov/omb/budget/fy2009/.

OECD PUBLISHING, 2, rue André-Pascal, 75775 PARIS CEDEX 16
PRINTED IN FRANCE
(42 2010 04 1 P) ISBN 978-92-64-07689-1– No. 57055 2010
Tax Expenditures in OECD Countries
In all OECD countries, governments collect revenues through taxes and redistribute this
public money, often by obligatory spending on social programmes such as education
or health care. Their tax systems usually include “tax expenditures” – provisions that
allow certain groups of people, such as small businessmen, retired people or working
mothers, or those who have undertaken certain activities, such as charitable donations,
to pay less in taxes.
The use of tax expenditures by governments is pervasive and growing. At a time when
many government budgets are threatened by population ageing and adverse cyclical
developments, there is a pressing need to avoid inefficient government programmes,
some of which may utilise tax expenditures.
This book sheds light on the use of tax expenditures, mainly through a study of ten
OECD countries: Canada, France, Germany, Japan, Korea, the Netherlands, Spain,
Sweden, the United Kingdom and the United States. This book will help government
officials and the public better understand some of the technical and policy issues
behind the use of tax expenditures. It highlights key trends and successful practices,
and addresses a broad range of government finance issues, including tax policy making,
tax and budget efficiency, fiscal responsibility and rule making.
The full text of this book is available on line via these links:
www.sourceoecd.org/governance/9789264076891
www.sourceoecd.org/taxation/9789264076891
Those with access to all OECD books on line should use this link:
www.sourceoecd.org/9789264076891
SourceOECD is the OECD online library of books, periodicals and statistical databases.
For more information about this award-winning service and free trials, ask your librarian, or write to
us at [email protected].
ISBN 978-92-64-07689-1
www.oecd.org/publishing
42 2010 04 1 P -:HSTCQE=U\[]^V:

2010 Tax Expenditure Countries Oecd

Uploaded by

Copyright:

Available Formats

2010 Tax Expenditure Countries Oecd

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

2010 Tax Expenditure Countries Oecd

Uploaded by

Copyright:

Available Formats

Tax Expenditures

ISBN 978-92-64-07689-1 (print)

Corrigenda to OECD publications may be found on line at: www.oecd.org/publishing/corrigenda.

In all OECD member countries, governments collect revenues through

TAX EXPENDITURES IN OECD COUNTRIES © OECD 2010

TAX EXPENDITURES IN OECD COUNTRIES © OECD 2010

Part I: A look at tax expenditures .....................................................................9

Chapter 1 Introduction .....................................................................................11

TAX EXPENDITURES IN OECD COUNTRIES © OECD 2010

Table II.1. Tax expenditures in Canada (% of GDP).......................................172

TAX EXPENDITURES IN OECD COUNTRIES © OECD 2010

Table II.22. Tax expenditures in the United Kingdom (% of central

Figure II.1. Income tax expenditure by purpose in Canada.............................. 229

Data sources ………………………………………………………………. 238

TAX EXPENDITURES IN OECD COUNTRIES © OECD 2010

This book has...

TAX EXPENDITURES IN OECD COUNTRIES © OECD 2010

A look at tax expenditures

TAX EXPENDITURES IN OECD COUNTRIES © OECD 2010

TAX EXPENDITURES IN OECD COUNTRIES © OECD 2010

What are tax expenditures?

Tax expenditures are “provisions of tax law, regulation or practices that

What are the different types of tax expenditures?

Tax expenditures may take a number of different forms:

TAX EXPENDITURES IN OECD COUNTRIES © OECD 2010

Box 1.1. Examples of tax expenditures

• Professional expenses: meals and entertainment expenses, commuting expenses, etc.;

• Interest on saving accounts (up to a certain ceiling);

How are tax expenditures measured?

Tax expenditures are calculated using the “revenue forgone method

TAX EXPENDITURES IN OECD COUNTRIES © OECD 2010

• Outlay equivalence: the direct expenditure that would be required in

Trends in tax expenditures

Tax expenditures – defined as “a transfer of public resources that is

TAX EXPENDITURES IN OECD COUNTRIES © OECD 2010

and extended procedures of tax expenditure reporting, review and control by

Defining tax expenditures

Identification of any particular tax provision as a tax expenditure

TAX EXPENDITURES IN OECD COUNTRIES © OECD 2010

TAX EXPENDITURES IN OECD COUNTRIES © OECD 2010

Kraan (2004) says that the problem of disagreement in the choice of a

Tax expenditure controversy

The concept of tax expenditures has been controversial since its

TAX EXPENDITURES IN OECD COUNTRIES © OECD 2010

TAX EXPENDITURES IN OECD COUNTRIES © OECD 2010

TAX EXPENDITURES IN OECD COUNTRIES © OECD 2010

1. See Surrey (1973), based on work done by the author in government

TAX EXPENDITURES IN OECD COUNTRIES © OECD 2010

Anderson, Barry (2008), powerpoint presentation at the Asian Senior

TAX EXPENDITURES IN OECD COUNTRIES © OECD 2010

Polockova Brixi, Hana, Christian M.A. Valenduc, and Zhicheng Li Swift

TAX EXPENDITURES IN OECD COUNTRIES © OECD 2010

TAX EXPENDITURES IN OECD COUNTRIES © OECD 2010

Administrative economies of scale and scope

TAX EXPENDITURES IN OECD COUNTRIES © OECD 2010

Limited probability of abuse or fraud

A proper wide range of taxpayer choice

Measurement of taxpaying capacity

Theoretical allegations: When and why are tax expenditures bad?