This paper reviews the published literature on the definition and measurement of the administrative and compliance costs of taxation, with special reference to VAT (including evasion and fraud) in the European Union.
Written by Luca Barbone, Richard M. Bird, and Jaime Vasquez-Caro. Published in March, 2012.
See more on our website: http://www.case-research.eu/en/node/57573
Report
Share
Report
Share
1 of 66
Download to read offline
More Related Content
CASE Network Report 106 - The Costs of VAT: A Review of the Literature
3. The CASE Network is a group of economic and social research centers in Po-land,
Kyrgyzstan, Ukraine, Georgia, Moldova, and Belarus. Organizations in the
network regularly conduct joint research and advisory projects. The research co-vers
a wide spectrum of economic and social issues, including economic effects of
the European integration process, economic relations between the EU and CIS,
monetary policy and euro-accession, innovation and competitiveness, and labour
markets and social policy. The network aims to increase the range and quality of
economic research and information available to policy-makers and civil society,
and takes an active role in on-going debates on how to meet the economic chal-lenges
facing the EU, post-transition countries and the global economy.
The CASE network consists of:
CASE – Center for Social and Economic Research, Warsaw, est.
1991, www.case-research.eu
CASE – Center for Social and Economic Research – Kyrgyzstan,
est. 1998, www.case.elcat.kg
Center for Social and Economic Research – CASE Ukraine, est.
1999, www.case-ukraine.kiev.ua
CASE –Transcaucasus Center for Social and Economic Research,
est. 2000, www.case-transcaucasus.org.ge
Foundation for Social and Economic Research CASE Moldova, est.
2003, www.case.com.md
CASE Belarus – Center for Social and Economic Research Belarus,
est. 2007.
Center for Social and Economic Research CASE Georgia, est. 2011
4. Luca Barbone, Richard M. Bird, Jaime Vázquez Caro
Contents
Summary ................................................................................................................ 8
1. Introduction ..................................................................................................... 10
2. Definition and Measurement of Compliance Costs ...................................... 12
2.1. The Costs of Tax Compliance ......................................................... 12
2.2. Cost Drivers: Administrative and Compliance Costs ...................... 14
2.3. Compliance Costs and the Standard Cost Model (SCM) ................ 16
2.4. The Standard Cost Model ................................................................ 18
3. Estimates of Compliance and Administrative Costs .................................... 21
3.1. Administrative Costs ....................................................................... 21
3.2. Estimates of Compliance Costs ....................................................... 23
3.3. Cross-Border Transactions in the EU and Compliance-
Administrative Burdens ........................................................................... 29
3.4. Private Sector Contribution to Compliance, Compliance Costs and
VAT Fraud Analysis ................................................................................ 31
3.5. What do the Studies Show? ............................................................. 32
3.6. Concluding Remark ......................................................................... 40
4. Compliance Costs and Non-Compliance ....................................................... 41
4.1. Overview .......................................................................................... 41
4.2. Quantitative Evidence of Evasion and Fraud .................................. 46
4.3. Searching for Solutions: Anti-Fraud Strategies ............................... 52
4.4. In Closing - VAT Compliance Costs and Fraud: Is There a Link? . 57
References............................................................................................................. 59
CASE Network Reports No. 106 4
5. THE COSTS OF VAT: A REVIEW OF THE LITERATURE
The authors
Luca Barbone has been President of the CASE Management Board since Sep-tember
2011. He joined CASE in January 2011 upon his retirement from the
World Bank, where he had worked since 1988, holding various leadership posts,
among others Director in the Poverty Reduction and Economic Policy Unit in the
Europe and Central Asia Regional Office (2007-2011), World Bank Director for
Poverty Reduction (2004-2007), and Regional Director for Ukraine, Moldova and
Belarus (2000-2004). Prior to the World Bank, Mr. Barbone worked for the Or-ganisation
for Cooperation and Development (Paris), the International Monetary
Fund, The Planning Institute of Jamaica, and the Bank of Italy. He holds a Ph.D.
in Economics from the Massachusetts Institute of Technology. He has published a
number of articles in professional journals and books. Main areas of personal in-terest
now include: (i) economic crisis and growth prospects; (ii) economic conse-quence
of long-term demographic trends; (iii) migration and development; (iv)
fiscal institutions, fiscal consolidation (v) social cohesion and political economy of
reforms in Europe and Central Asia.
Richard M. Bird is Professor Emeritus of Economics, Rotman School of
Management, and Adjunct Professor and Associate of the Institute for Municipal
Governance and Finance, Munk School of Global Affairs, University of Toronto.
He is also Distinguished Visiting Professor, Andrew Young School of Public Poli-cy
at Georgia State University in Atlanta, and Adjunct Professor of the Australian
School of Taxation and Business Law of the University of New South Wales, Aus-tralia.
He is currently Chair of the Central Advisory Group of the new Internation-al
Centre for Tax and Development at the Institute for Development Studies in the
UK.
5 CASE Network Reports No. 106
6. Luca Barbone, Richard M. Bird, Jaime Vázquez Caro
Abstract
This paper reviews the published literature on the definition and measurement
of the administrative and compliance costs of taxation, with special reference to
VAT (including evasion and fraud) in the European Union.
CASE Network Reports No. 106 6
7. THE COSTS OF VAT: A REVIEW OF THE LITERATURE
In Memoriam
This paper is dedicated to the memory of Luis Jaime Vázquez Caro, who died
unexpectedly on 28 September 2011, during the final stages of this project, on
which he worked as part of the CASE team. Jaime, a 1973 graduate of the Interna-tional
Tax Program at Harvard University, was a well-known professional in the
area of tax policy and administration, having been the deputy tax commissioner of
Colombia, and having spent over fifteen years of his career working on tax admin-istration
reform issues first at the IMF, and then at the World Bank. In the mid-
1990s he was involved in several projects to help in the implementation of tax and
tax administration reforms in central Europe, and was particularly concerned with
the emergence of different types of VAT-based frauds, discussed in this report. At
the time of his passing, he was also working on a project on tax administration
reform in Armenia. Our condolences go to his wife Maria Teresa and his son José
Camilo, who live in Bogotá, Colombia.
7 CASE Network Reports No. 106
8. Luca Barbone, Richard M. Bird, Jaime Vázquez Caro
Summary
This paper provides a review of published economic literature on the definition
and measurement of compliance costs for taxation and regulation in general (with
emphasis on VAT and on the European Union), as well as of VAT evasion and
fraud.
The first section focuses on the compliance costs of VAT and related taxes, and
discusses existing estimates of the level and structure of such costs. It begins by
clarifying the sometimes confusing terminology found in the literature (e.g., com-pliance
vs. administrative costs), and proceeds to review older and more recent
estimates of costs borne by the public and private sectors. The main conclusions
are as follows:
1. With regard to administrative costs (those costs that are borne directly by
the public sector, and indirectly by all taxpayers), the review shows con-siderable
variation of overall tax administration costs among EU countries,
suggesting the potential for efficiency improvements in at least several of
them. However, little specific information is currently available to single
out VAT costs, as most administrations are not organized by single tax,
but rather by functions that cover a multiplicity of taxes. While the com-plexities
of the tax systems undoubtedly add to the administrative costs
(including VAT), most likely other country-specific factors contribute to
country differences.
2. With regard to compliance costs (those that are directly borne by VAT
taxpayers), the review shows that over the past ten years or so the Stand-ard
Cost Model (SCM) has quickly become the gold standard among prac-titioners
and policymakers for assessing such costs, and for setting policy
goals. While the SCM is not immune from criticism (being criticized for
lack of consideration of market failures and imperfections, or for its as-sumptions
on a “normally efficient” firm, etc), it has helped produce sev-eral
country and global studies that allow comparisons over time and
across countries. The general conclusions are that, in the EU as well as in
the many other countries that have adopted VAT taxation, (i) compliance
costs are high and significant for individual businesses (but perhaps less so
at a macroeconomic level); (ii) compliance costs are regressive, in the
sense that small business in particular are more than proportionally bur-dened
by compliance requirement; and (iii) compliance costs are not re-ducing
over time, perhaps with the exception of countries that have ag-
CASE Network Reports No. 106 8
9. THE COSTS OF VAT: A REVIEW OF THE LITERATURE
gressively adopted E-filing procedures, and others that may have adopted
aggressive programs of burden reduction.
The paper also reviews compliance burdens linked to intra-EU trade and relat-ed
reporting requirements, and concludes that the bulk of the costs to taxpayer
engaged in intra-EU trade is attributable to statistical requirements that might exist
even in the absence of VAT. Further research is however warranted in this respect,
given the most recent technological advances in data reporting that could be cap-tured
by up-to-date surveys.
In the second section, the paper takes stock of the existing quantitative and the-oretical
literature with regard to VAT evasion and fraud, with specific focus on
EU countries. VAT evasion is a well-recognized phenomenon, and the most recent
estimates of fraud put the revenue loss for EU countries to a (wide) range of some
2 to 30 percent of potential revenues, with an overall average of about 14 percent.
The literature points to several reasons to explain VAT fraudulent practices.
VAT specific and EU general policies have been recognized as major determinants
of the compliance and enforcement environments that facilitate the emergence of
fraud. These include (i) parametric issues on base, rates, exemptions, zero rating,
registration and return filing thresholds, refunding of VAT specific rules and the
existence of parallel small taxpayers regimes; and (ii) two broadly recognized
general EU policy principles that affect VAT fraud: the intra-European single
market in force since 1993, and the application of the subsidiarity principle to tax
administration which generates a second layer of differentiation in the actual ap-plication
of the laws.
Overall, this literature review points to several avenues for further research that
might help policymaking, for instance in the better understanding of administra-tive
costs for VAT (e.g., by applying the SCM model to tax administrations and
related services); the effect of the increasing prevalence of e-reporting on compli-ance
costs, particularly for intra-EU trade; and the trade-offs between the added
compliance costs of (more) enforcement efforts and the revenue losses associated
with laxer attitudes, just to cite three important examples.
9 CASE Network Reports No. 106
10. Luca Barbone, Richard M. Bird, Jaime Vázquez Caro
1. Introduction
In the province of Germany it is quite clear that goodness and respect for
religion are still to be found in its peoples … When these republics have
need to spend any sum of money on the public account … each person pre-sents
himself to the tax collectors in accordance with the constitutional
practice of the town. He then takes an oath to pay the appropriate sum, and
throws into a chest provided for the purpose the amount which he conscien-tiously
thinks that he should pay; but of this payment there is no witness
save the man who pays1.
Curiously, this rather imaginative description of Germany in the 16th century
has a 20th century parallel in the vision of the libertarian thinker Ayn Rand2:
In a fully free society, taxation – or, to be exact, payment for government
services – would be voluntary. Since the proper services of a government –
the police, the armed forces, the law courts – are demonstrably needed by
individual citizens and affect their interests directly, the citizens would (and
should) be willing to pay for such services, as they pay for insurance.
In both these visions of an ideal, law-abiding society – one of a (probably im-aginary)
past and one of a (probably unrealisable) future – people would voluntari-ly
pay the taxes they owe, and the task of the revenue administration would be
little more than to provide the facilities for citizens to discharge this responsibility.
Alas, no such country exists, nor – despite what Machiavelli may have thought –
has it ever existed.
Compliance with tax laws does not occur without effort: it must be created, cul-tivated,
monitored, and enforced in all countries. In economic terms, of course,
effort is just another word for cost, and tax compliance costs have been a concern
for centuries. One of Adam Smith’s famous ‘canons’ of taxation, for instance, was
that “(e)very tax ought to be levied at the time, or in the manner in which it is most
likely to be convenient for the contributor to pay it. 3”
Although a few early attempts were made to measure the cost of tax compli-ance
(e.g. Haig 1935), the “father” of modern compliance cost studies was un-
1 Machiavelli N, The Discourses, ed. Bernard Crick (Penguin Books, 1983), pp. 244–45.
2 Rand, The Virtue of Selfishness, New York, Signet, 1964, p. 117.
3 Smith, Wealth of Nations, quoted in http://www.progress.org/banneker/adam.html.
CASE Network Reports No. 106 10
11. THE COSTS OF VAT: A REVIEW OF THE LITERATURE
doubtedly Cedric Sandford, who amongst many other works on the subject (e.g.,
Sandford 1973, 1994) produced the first detailed study of the costs of VAT (Sand-ford
et al. 1981). In this report, we review much of the modern literature on com-pliance
costs with particular attention to the costs associated with VATs, with
particular reference to the existing VATs in the EU. We also consider some as-pects
of VAT fraud and non-compliance.
This paper is organised as follows. Section 2 provides a review of concepts
with regard to different aspects of administrative and compliance costs; Section 3
reviews quantitative estimates of such costs; and Section 4 provides a discussion
of non-compliance and fraud in the European VAT context, including available
quantitative estimates, proposals for reform to reduce fraud, and some suggestive
evidence on the trade-off between compliance costs and fraud.
11 CASE Network Reports No. 106
12. Luca Barbone, Richard M. Bird, Jaime Vázquez Caro
2. Definition and Measurement of
Compliance Costs
In recent years, a substantial body of literature has been devoted to the defini-tion
and, in many cases, the quantification of the costs of complying with taxation
and with regulation in general. In the last decade or so, particularly but not exclu-sively
in the European Union, an increasing proportion of this literature has taken
the form of cost estimates based on the Standard Cost Model (SCM). As discussed
below, the SCM is in effect a version of a subset of the broader concept(s) used
earlier in the broader Tax Compliance Cost (TCC) literature. This section provides
a brief overview of the different concepts of burdens, drivers and methodologies
found in both the SCM and the broader TCC literature. While these concepts are
generally applicable to all taxes, we highlight VAT-specific issues when appropri-ate.
2.1. The Costs of Tax Compliance
A number of important definitional issues need clarification when approaching
taxation compliance costs, particularly in view of not only the shifting debates
within the EU but also the increasing use of the Standard Cost Model and related
approaches recently popularised through such other influential publications as the
World Bank’s Paying Taxes 2011.
In a recent thorough review of the compliance cost literature, Evans (2008)
provides a clear and broad definition of terms with respect to the costs of taxation:
“Modern taxation systems have the capacity to impose a heavy burden on
taxpayers, and particularly on small business taxpayers. That burden typi-cally
consists of three elements. In the first place there are the taxes them-selves
(…) Secondly, there are the efficiency costs (variously referred to as
deadweight losses or excess burden), involving tax-induced market distor-tions.
And finally there are the operating costs of the tax system: the costs to
the government (ultimately borne by taxpayers) of administering and col-lecting
the taxes (usually referred to as “administrative costs”), and the
CASE Network Reports No. 106 12
13. THE COSTS OF VAT: A REVIEW OF THE LITERATURE
costs expended by taxpayers in complying (or sometimes not complying)
with their tax obligations (usually referred to as “compliance costs”)”.
Evans (2008) goes on to note that “in addition to this generally accepted hard
core of compliance costs, there are a number of other costs that need to be consid-ered.
For example, there is little doubt that there will always be a measure of psy-chological
cost that is induced by the operation of the tax system. Taxpayers suffer
stress, anxiety and frustration as a result of attempting to comply with their tax
obligations. Unfortunately, no studies have yet managed to successfully quantify
these psychological costs, although research in this area is now taking place”. For
this reason, we will not pursue this strand of the literature further in the present
review, although James and Edwards (2010) list several interesting examples of
behavioural and experimental research which appear to offer some promise of
future practical relevance (e.g. Coleman et al., 2003). In particular, it is perhaps
worth noting that at least one such study (Hasseldine and Hansford, 2002) sug-gests
that psychic costs are positively associated with financial costs of compli-ance.
In most of the tax compliance cost (TCC) literature surveyed by Evans (2008)
(and catalogued extensively by James and Edwards, 2010) the term administrative
costs refers to the public budgetary costs associated with collecting taxes (includ-ing,
of course, VAT). Confusingly, however, in such SCM-based studies as SCM
Network (2005), this term has the very different meaning of the direct resource
costs imposed on taxpayers, assuming full compliance with the law.
Equally confusingly to those familiar with the TCC literature, the SCM studies
introduce the term administrative burden (AB) to mean those costs that are direct-ly
attributable to the various “information obligations” imposed on taxpayers by
such regulations as VAT law and procedures, as distinguished from the costs – e.g.
of registering a business – necessary for simple “business as usual” (BAU) opera-tion.
Thus defined, “administrative burden” is of course a major component, but
not the whole, of the “compliance costs” imposed by VAT on the private sector as
discussed and measured in the broader TCC literature. That literature often con-siders
not only the compliance costs imposed on the private sector by taxation but
also the public sector’s administrative costs. While in some instances administra-tive
and compliance costs may be substitutes and in other instances complements,
both constitute real resource costs – the “operating costs” – of a given tax system,
and should be properly accounted for. Table 1 – drawn largely from Sandford,
Godwin and Hardwick (1989) – may help the reader disentangle the overlapping
but distinct measurement approaches found in the compliance literature. Most of
the concepts listed in the table are discussed further in the next section.
13 CASE Network Reports No. 106
14. Luca Barbone, Richard M. Bird, Jaime Vázquez Caro
Table 1. Compliance and Administrative Costs
A. Administrative (or ‘enforcement’) public sector costs
1. Budgetary costs of revenue department(s)
2. Costs incurred by other departments in providing information
3. Judiciary and other costs related to dispute resolution
4. Interest costs (of ‘loans’ extended by legal lags in collection)
B. Compliance costs incurred by private sector
1. Direct costs incurred by taxpayers or ‘taxpayer costs’(time, labour cost, expert ad-vice,
other)
(a) in complying with legal obligations (“involuntary” or unavoidable costs) [1]
(b) in tax planning and attempting to evade (“voluntary” or avoidable costs)
(c) psychic costs (stress, anxiety, frustration)
2. Costs incurred by third parties (information providers, voluntary helpers)
C. Possible offsetting compliance ‘benefits’ to private sector
1. Management benefits (from improved accounting required for tax purposes)
2. Cash flow benefits (the private sector side of A.4)
D. Net compliance costs = B – C (in addition, some costs may be reduced to the extent
they are tax deductible)
E. Operating costs = Administrative + Compliance costs = A + B (or A + D) [2]
Notes:
[1] The SCM model essentially attempts to measure B1(a) – which it calls “administra-tive
burden” – distinguishing it from what is rather confusingly called “administrative
cost”, by which is meant the ordinary costs of running a business as opposed to the nar-rower
concept of the costs of complying with the specific ‘information obligations’ im-posed
by a particular law.
[2] Since there may be substantial ‘start-up’ costs for both the public and private sectors
when tax laws and procedures are changed and even the initial operating costs may be
reduced (‘learning effect’) over time, as discussed later it is sometimes important to dis-tinguish
initial from ongoing costs.
2.2. Cost Drivers: Administrative and Compliance Costs
Taking as given the standard (OECD, 2011) definition of administrative costs
as the resources devoted by governments to administer and enforce taxes and regu-lations
(including VAT)4, a number of studies have looked at what makes coun-tries
more or less efficient and effective in these tasks. OECD (2011) provides a
detailed, and often quantitative, comparison of tax administration practices in EU
member states, among others. Unfortunately for our purposes, though understand-ably,
since modern tax administrations are organised not by tax but by function,
4 The most useful general discussion of defining administrative costs probably remains that
of Sandford, Goodwin and Harwick. (1989).
CASE Network Reports No. 106 14
15. THE COSTS OF VAT: A REVIEW OF THE LITERATURE
increasingly with some segmentation of key taxpayer groups (such as large busi-nesses),
none of this information is provided on a ‘tax’ (e.g. VAT) basis. In any
case, valuable as they are, the OECD data can only be used for comparative pur-poses
with great care owing to the many comparability problems that remain to be
sorted out.
For a first attempt to incorporate some of this information in a more systematic
cross-country study, see Slemrod and Robinson (2010). In recent years, a number
of attempts have been made to compare such costs, mainly in developing coun-tries,
as discussed by Gallagher (2005). In addition, careful studies have been
made of the operational efficiency of tax offices in a number of countries such as
Belgium (Moesen and Persoon, 2002). 5
Although conceptually quite distinct, administrative costs and compliance costs
share certain “drivers”: for example, more complex regulations increase the bur-den
on taxpayers and generally also require higher managerial resources on the
part of enforcement agencies. However, the burden on the two sides of the process
– the taxer and the taxed – is likely to be quite uneven, may differ sharply in dif-ferent
sectors, and at different times and, in the case of VAT, may depend to a
considerable extent on such features as rate structure, thresholds, integration with
other business taxes, etc. (Cnossen, 1994; Evans, 2003). Among the ‘drivers’ of
administrative costs – and to a considerable extent of compliance costs also – iden-tified
in the literature are:
1. The complexity of legislation (the number of ‘lines’ to be drawn – exclu-sions,
exemptions, deductions; rate differences; goods/services distinc-tions,
etc.; frequency and nature of changes; costs involved in explaining
legislation, making rulings and determinations, etc.). Distinguishing set-up
(initial, implementation) costs of changes in these factors from on-going
recurrent costs is not always easy.
2. Procedural requirements – the number of returns6; requirements for sup-plementary
documentation; treatment of cross-border transactions; and, of
course, registration. The latter is an especially key factor in VAT because
possession of a VAT number carries with it the potential to, in effect,
write a payment order on the Treasury without the Treasury approving it
or even being aware of it.
5 Other relevant country studies of aspects of this issue, with varying degrees of sophistica-tion,
include Hunter and Nelson (1995) on the United States, Klun (2003) on Slovenia,
Serra (2005) on Chile, Forsund et al. (2006) on Norway, and von Soest (2007) on Zambia.
6 For example, the Doing Business and Paying Taxes studies of the World Bank place
considerable weight on the number of returns.
15 CASE Network Reports No. 106
16. Luca Barbone, Richard M. Bird, Jaime Vázquez Caro
3. The size and nature of clientele (number of taxpayers; structure of econo-my
and of business sector; the importance of B2B (transactions between
VAT registrants) relative to B2C (transactions with non-registrants);
cross-border transactions; size of threshold). In this connection, note that
there are ‘marginal costs’ associated with the growth of the taxpayer popu-lation
as well as with policy and procedural changes, and that these cate-gories
should in principle be distinguished.
4. The difficulty of verifying ‘self-assessed’ information, which varies with
such factors as the size of the informal sector; the extent and nature of
links between formal and informal sectors; ‘border effects’ on information
flows; the extent to which efforts are made with respect to verification and
chasing down suspect cases7; extent of e-invoicing; and the role played by
tax professionals – accountants in particular.
2.3. Compliance Costs and the Standard Cost Model (SCM)
There is also debate about what should be included in the measurement of tax
compliance costs. Tax compliance costs are those costs “incurred by taxpayers, or
third parties such as businesses, in meeting the requirements laid upon them in
complying with a given structure and level of tax” (Sandford, Godwin and Hard-wick,
1989, p. 10). Paraphrasing Evans (2008), while this is an area in which there
will always be debate, it is possible to identify a “hard core” of costs that are in-disputably
part of the costs of complying with tax requirements. Typically these
will include:
the costs of labour/time consumed in completion of tax activities. For ex-ample,
the time taken by a business person to acquire appropriate
knowledge to deal with tax obligations such as VAT; or the time taken in
compiling receipts and recording data in order to be able to complete a tax
return;
the costs of expertise purchased to assist with completion of tax activities
(typically, the fees paid to professional tax advisers); and
incidental expenses incurred in completion of tax activities, including
computer software, postage, travel, etc.
7 To illustrate, business surveys (such as KPMG, 2010) often find that the highest and most
troublesome operating costs are those associated with audit. Interestingly, OECD (2011)
shows that such costs are equally prominent on the other side of the taxing equation.
CASE Network Reports No. 106 16
17. THE COSTS OF VAT: A REVIEW OF THE LITERATURE
2.3.1. Involuntary vs. voluntary costs
Evans (2008) also notes that there is contention over other aspects of the pre-cise
boundaries of compliance costs. For example, compliance costs are some-times
divided into computational (unavoidable or involuntary) and tax planning
(avoidable or voluntary) costs (a distinction first made by Johnston, 1963). Many
tax lawyers and policy-makers continue to insist that only computational costs
constitute legitimate measures of tax compliance costs, and some attempts have
been made to disentangle the two (Pope, Fayle and Chen, 1991).
However most major modern studies (for example, Sandford, 1973; Sandford,
Godwin and Hardwick, 1989; Allers, 1994; Evans, Ritchie Tran-Nam and Wal-pole,
1997) have not distinguished computational and tax planning costs in their
estimates of compliance costs – if only for the obvious reason that it is often al-most
impossible to disentangle the one from the other. Moreover, as noted by
Slemrod and Sorum, “both kinds of costs are real resource costs of collecting the
taxes” (1984, p. 461). Despite these sound comments, the SCM approach does
attempt to distinguish these costs essentially, as discussed further below, by as-sumption.
2.3.2. Social vs. private (taxpayer) costs
Evans (2008) also points to the distinction between what have variously been
termed total, gross or social compliance costs and net or taxpayer compliance costs
(Allers, 1994; Evans, Ritchie, Tran-Nam and Walpole, 1997), i.e. the costs to the
economy vs. the costs directly borne by taxpayers. Social compliance costs tend to
be less than taxpayer compliance costs for two reasons:
In the first place there are various offsetting benefits that may be generated
for taxpayers as a result of compliance with their tax obligations. These
include, fairly obviously, certain cash flow benefits that may arise as a re-sult
of the timing difference between receipt of funds and payment of tax
relating to those funds. Most modern empirical studies quantify the value
of these benefits with some certainty. Less obviously, managerial benefits
may also occur as a result of tax compliance. For example, better accounts
and record keeping may lead to improved business decision-making and
reduce the costs of audit for small businesses, resulting in lower ac-counting
fees. Two major studies in the UK attempted, somewhat impreci-sely,
to quantify the managerial benefits generated for business taxpayers,
and concluded that the value of these managerial benefits can be quite sig-
17 CASE Network Reports No. 106
18. Luca Barbone, Richard M. Bird, Jaime Vázquez Caro
nificant (Sandford, Godwin, Hardwick and Butterworth, 1981, p. 96; Nati-onal
Audit Office, 1994, pp. 19–20).8
Secondly, net taxpayer costs – though not social costs – may be reduced to
the extent that they are deductible in computing income tax liability. The
tax deductibility of business taxpayer compliance costs has also been tak-en
into account in a few of the major studies but not in most. The three
major studies that appear to have factored in a value for the tax deductibil-ity
of certain compliance costs are those conducted by Johnston (1963),
Allers (1994), and Evans, Ritchie, Tran-Nam and Walpole (1997).
2.4. The Standard Cost Model
The Standard Cost Model (SCM), developed in the 1990s in the Netherlands
and quickly became the standard bearer of the definition of compliance costs for
practitioners, particularly in Europe. The spread of the reach of the SCM is little
short of phenomenal. Its features are discussed in a “manual” now widely in use
among practitioners (see International Working Group on Administrative Burdens,
2004). Other useful references include Wegrich (2009), from which Box 1 is
adapted, as well as a very vibrant on-line debate, best represented perhaps by the
network Standard Cost Model, which maintains a growing website community of
practice at http://www.administrative-burdens.com/.
As Box 1 discusses, the SCM per se did not introduce particularly innovative
concepts or techniques to estimate compliance costs by taxpayers. Its strength,
which makes it so appealing to spontaneous replication across many administra-tions
and professional circles, lies in its accounting-like methodology, which
promises to quantify costs based on an assumed real-life simulation of what it
takes to comply with legal and administrative reqirements. In some countries, as
discussed in the next section, this approach has produced apparently precise esti-mates
of costs of regulation and taxation (among which VAT features prominent-ly),
which have then been used to set the stage for a public debate on creating bet-ter
business environments. The SCM is also at the root of the World Bank’s Pay-ing
Taxes (2011) methodology, as we discuss later, and this methodology may,
over time, produce even more powerful effects on public policy debate, as have
8 A recent study in South Africa (Smulders, 2012) found that 75 percent of small busines-ses
thought they had received some benefits from the improved record-keeping required by
VAT but most were unable to assign any quantifiable value to such benefits.
CASE Network Reports No. 106 18
19. THE COSTS OF VAT: A REVIEW OF THE LITERATURE
the more general indicators and rankings included in the Bank’s broader Doing
Business (2011) studies.
In addition to being restricted to a subset of compliance costs, the SCM is not
immune from other criticism. In particular, Weigel (2008) has argued that the
model is deficient for a number of reasons:
the lack of explicit consideration of the market failures which gave rise to
information obligations, which may lead to economically flawed results
because the simulation assumes that the costs attributable to the IOs may
be eliminated with no detriment to the efficiency or coverage of the tax;
the disregard of other market imperfections that permit strategic actions
that may lead to errors in the assessment of tax obligations;
The disregard of the variation in costs of compliance by assuming a “nor-mally
efficient” firm – a methodology that to some extent makes the over-all
result of the exercise almost arbitrary, and certainly far from statistical-ly
representative9;
Finally, other reasons may lead different firms to perform differently (such
as those suggested by the theory of X-inefficiency), so that the reactions of
firms to changes in the reduction of compliance burdens may be quite dif-ferent
from those suggested by the SCM numbers (and presumably desired
by policy makers).
Box 1. The Standard Cost Model: Rapid Deployment of a Simple Technique
The SCM policy template was developed starting in the early 1990s in the Netherlands
(cf. WIFO-CEPS, 2006; OECD, 2007). The perception of increasing regulatory burdens
was a recurring theme on the public sector reform agenda in the Netherlands (Larsen,
2006; Toonen and van den Ham, 2007). The idea of measuring and quantifying regulatory
burdens was part of this debate. However, earlier attempts to measure overall costs of
regulation were frustrated by the perceived complexity of such an approach, and also by
difficulties in accounting for benefits of regulations.
Rather than developing increasingly complex solutions to these problems, policy devel-opment
was guided by the idea of reducing complexity by focusing the measurement on a
specific component of regulatory costs, namely what came to be defined as administrative
costs (see text discussion for semantic differences with earlier literature). Administrative
costs in the SCM are defined as those parts of the regulatory (or compliance) costs that are
imposed on firms by specific information obligations (IOs) included in laws or secondary
legislation. Administrative costs thus defined are then distinguished from so-called sub-stantial
compliance costs, i.e costs attributable to compliance with regulatory standards
9 This problem is perhaps less important in the EU than in less developed countries in
which the evidence shows that there is tremendous dispersion in firm behaviour and per-formance
for a variety of reasons (see e.g. Hallward-Driemeier and Prichett, 2011).
19 CASE Network Reports No. 106
20. Luca Barbone, Richard M. Bird, Jaime Vázquez Caro
(such as emission standards). While the boundary between administrative and substantial
costs is difficult to draw and those two types of regulatory costs are clearly related, the key
idea is to quantify those costs that are easy to measure in order to permit the setting of
quantitative targets for reducing administrative burdens.
The method for measuring administrative costs was developed by research organisa-tions
and consultancies over a decade and tested in various pilots. Unlike approaches as-sessing
administrative costs accumulating in companies by focusing on single regulations,
the main idea of the SCM is to start from information obligations included in legislation,
calculate the time (hence: costs) of work needed in a company to comply with this obliga-tion,
and then sum up the number of ‘cases’ (frequency of occurrence and number of com-panies
affected by the information obligation). The total cost calculated for all the individ-ual
information obligations of a regulation is regarded as the quantification of the adminis-trative
cost of this regulation. While the calculation of the costs of complying with infor-mation
obligations is based on information gathering activities, such as interviews or, in
some cases, actual time measurement (stopwatch approach), the tool is not meant to pre-sent
either an exact measurement or a representative sample of the actual costs of compli-ance
for companies. Rather, the idea is systematically to assess what the costs would be in
a ‘standard’ process of compliance with the information obligation. Experiences with the
measurement exercise, the development of databases etc. and comparative ‘benchmarking’
are said to enhance the precision of the assessment. Nevertheless, the quantification re-mains
a proxy of a cost measurement that is supposed to allow tracking change (as well as
benchmarking across jurisdictions) over time and hence evaluate the effectiveness of re-duction
measures. The method does not account for different administrative implementa-tion
styles of regulations in terms of over- or under-enforcement by agencies, and of
course takes no account of any possible benefits from the regulations, e.g., by improving
management’s information on operations.
Diffusion of the SCM policy template in Europe
From its very inception, the SCM has enjoyed almost unparalleled popularity among
both practitioners and policymakers, and consequently it has rapidly spread as the standard
tool for quantification of costs of taxation and regulation, particularly in the EU. Taking
for instance the starting point as 2003 when the Netherlands carried out the SCM baseline
measurement (accounting for all regulations as by end of 2002), by 2004, only two other
countries were engaged in any activity of administrative costs measurement and reduction.
However, by the end of 2007, 15 out of 29 EU-25/EFTA member states had developed
such programmes (with two further countries having expressed the intentionto engage in
SCM measurement exercises). Almost all EU-15 member states have adopted this ap-proach.
Moreover, so have two larger EFTA countries (Norway and Switzerland). As yet,
however, smaller EU-15 member states, most of the new member states of the 2004 en-largement
and the two small EFTA countries have done relatively little along these lines.
Still, the scope of diffusion in western Europe is striking – all larger western European
countries are involved in some kind of emulation of the SCM policy template. Interestingly
(but beyond the scope here), South Africa was the first non-European country to adopt the
SCM approach, and other non-European countries (Australia, Canada, US, Australia, New
Zealand) are involved in the OECD’s project related to the SCM method (Red Tape Score-board,
OECD, 2007).
CASE Network Reports No. 106 20
21. THE COSTS OF VAT: A REVIEW OF THE LITERATURE
3. Estimates of Compliance and
Administrative Costs
With the observations just discussed on concepts and definitions, we now re-view
in turn evidence on administrative and compliance costs.
3.1. Administrative Costs
Administrative costs imposed on the public sector by VAT are largely captured
in the budgetary allocation of the revenue department(s). The OECD (2011), in its
biannual publication on comparison of tax administration performances, provides
detailed country-by country data on budgetary allocations for different tax admin-istration
(though not divided by tax, something which is obviously very difficult to
quantify with precision). Figure 1 provides a snapshot for 2009, the latest compa-rable
data period (note that Greece does not report such data).
Figure 1. Tax administration budgetary allocations
1.6
1.4
1.2
1.0
0.8
0.6
0.4
0.2
0.0
% GDP
Austria
Belgium
Bulgaria
Cyprus
Denmark
Estonia
Finland
France
Germany
Hungary
Ireland
Czech Republic
Luxembourg
Netherlands
Italy
Latvia
Lithuania
Poland
Portugal
Romania
Slovakia
Slovenia
United Kingdom
Spain
Sweden
21 CASE Network Reports No. 106
22. Luca Barbone, Richard M. Bird, Jaime Vázquez Caro
Thus defined, tax administration costs in the EU averaged 0.29 percent of
GDP, ranging from a minimum of 0.12 percent in Estonia to 1.3 percent of GDP in
Cyprus (an outlier, with the second-highest country being Hungary with 0.34
percent of GDP). Although the OECD study warns of the pitfalls of international
comparisons of such ratios, it also provides data on other measures of
administrative efficiency of such expenditures, such as the ratio of expenditures to
tax collections.
Of course, not all administrative costs are attributable to VAT. On the other
hand there are additional administrative costs that can and should also be taken
into account e.g., other government departments, judiciary, etc.10 It should also be
mentioned that budgetary numbers often do not convey the economically relevant
facts: for example, many countries do not charge appropriate ‘rents’ for the office
facilities used by tax agencies to the budgets of those agencies and capital outlays
(e.g. not simply buildings and computers but also such outlays as training costs
and ‘taxpayer services’) are seldom depreciated appropriately from an economic
perspective. In sum, it appears that as yet no attempt has been made to develop as
detailed an approach as the SCM to allocating the costs of tax departments to the
various ways and amounts in which real resources are devoted to either on-going
VAT activities or the impact of changes in legislation or procedures.
Partly because of the increasing extent to which tax administration is organised
by function rather than by tax, little information on administrative costs by tax is
available. In the UK, however, a recent report shows that the cost of administering
VAT in that country is 0.7 percent of VAT revenues.11 Occasionally other coun-tries’
budgets, annual reports, etc. provide information on the estimated costs of
changes in various aspects of VAT administration.12 As an example, the recent
incorporation of the provincial sales tax into the national VAT system was report-ed
in the budget of the province of Ontario to reduce the province’s administrative
costs by $500 million in 2010-11: although no details were reported, it is likely
that this outcome reflects the fact that the new tax, unlike the previous provincial
sales tax, would be administered at no cost to the province by the Canada Revenue
Agency.13
10 See for one such wider approach Vaillancourt, Clemens and Palacios (2008), for Canada.
11 See p. 13 of http://www.hmrc.gov.uk/about/annual-report-accounts-1011.pdf.
12 For a detailed consideration of comparative information on the costs and requirements of
administering a tax system, including a VAT, see Australian Government (2007), a re-search
guide which compares approaches to information management, risk management
and internal organisations among several large administrations (Australia, Canada, United
States, the “OECD model”, etc.).
13 See http://www.fin.gov.on.ca/en/budget/ontariobudgets/2010.
CASE Network Reports No. 106 22
23. THE COSTS OF VAT: A REVIEW OF THE LITERATURE
Overall, in this area what has been done in terms of understanding and estimat-ing
administrative costs appears to have been governed far more by the availability
of data than by any rigorous consideration of what should be done to obtain the
most useful or relevant measures for the purpose at hand. Of course, this is both
understandable and acceptable; but perhaps more thought should be given to how
we might obtain the ‘correct’ economic information or at least to thinking about
the possible extent and direction in which the numbers we do have might be al-tered
if we were able to take such information – even if it proves unattainable in
practice – into account.
3.2. Estimates of Compliance Costs
3.2.1. The SCM approach
The spread of the SCM discussed above has produced a flurry of estimates of
compliance costs by countries, some of which are based on very detailed analyses
of business processes and obligations resulting from tax legislations and other
regulations.
KPMG (2006) reports a detailed study of the “administrative burden” (compli-ance
burden in our terminology) for a number of tax and other obligations for the
UK for the year 2005. According to the study, the total compliance burden can be
quantified at £5.1 billion (or about 0.42 percent of GDP), of which costs attributa-ble
to VAT would amount to about £1 billion, or 0.08 percent of GDP. The report
also provides estimates of costs based on types of obligations, as well as on the
size of the business units. As is typical in such studies, smaller businesses (if sub-ject
to VAT obligations) are reported to bear a disproportionately large share of
the total burden.
SCM Network (2005) reports estimates of what it labels the “administrative
burden of VAT” calculated using the SCM methodology for four countries (Den-mark,
Netherlands, Norway and Sweden) and based on the structure of VAT as of
2003 for each country. As shown in Table 1, this SCM concept is roughly equiva-lent
to what other studies term compliance costs (i.e. those costs incurred directly
by taxpayers) and we therefore use the term “compliance costs” for consistency
and simplicity. According to this study, compliance costs on businesses (per aver-age
business unit) range from a low of Euro 180 in Denmark (which, multiplied by
the number of businesses reported by the study, yields a “total cost” amounting to
0.3 percent of VAT collections, or 0.03 percent of GDP), to a high of Euro 807 for
23 CASE Network Reports No. 106
24. Luca Barbone, Richard M. Bird, Jaime Vázquez Caro
the Netherlands (for a total of 2.17 percent of VAT collections and 0.17 percent of
GDP), with Norway at Euro 430 (0.64 percent of VAT collections and 0.06 per-cent
of GDP) and Sweden at Euro 344 (0.75 percent of VAT collections and 0.07
percent of GDP). The authors advance a number of explanations for this wide
range. Inspection of the rate structure, filings, thresholds, registration requirements
etc. reveals a number of differences across countries that without pointing to a
single culprit, give food for thought. For instance, it is notable that Denmark is the
country of the group with the least differentiation in rates, so that the other coun-tries’
businesses are burdened with somewhere between 2 and 48 extra hours per
year for the administration of multiple rates. Similarly, Denmark’s filing proce-dures
are more lenient than in Norway or Sweden, again resulting in cost ad-vantages
for Danish businesses. For example, a business with limited liability and
a turnover of 200 000 euro must file four times per year in Denmark, six times per
year in Norway and 12 times per year in Sweden. The differences between the
countries can also be seen by looking at the proportion of businesses in each re-spective
country that file a different number of times per year. 44 percent of the
businesses obliged to pay VAT in Sweden file 12 times per year, in Denmark the
same proportion is only 10 percent.
For SCM studies of the new EU members and candidate countries, see Klun
(2003) for Slovenia, and Blažić (2004) on Croatia. The latter specifically addresses
the issue of regressivity of taxation (and of VAT in particular). It finds that VAT
compliance costs amount to 3.9 percent of turnover for individual entrepreneurs,
while falling to 1.5 percent for firms with more than 6 employees. The study
comes to the conclusion that “The regressive effect of tax compliance costs is
proven in the case of Croatian small business (businesses that pay personal income
tax), even with respect to micro businesses. In the cost structure the time cost,
predominantly the owner’s time, is predominant” (Blažić, 2004, p. 15). The study
produces an overall estimate of the VAT burden in the range of 16-25 percent of
VAT collections, an astoundingly high figure attributed to the still-recent introduc-tion
of VAT when the study was conducted.14
14 These results are driven largely by the estimated cost of owner’s time. Such estimates
are always difficult. Two studies of compliance cost in South Africa, for example, pro-duced
very different estimates of compliance cost. The first study (Coolidge and Ilic,
2009) used average wages to estimate this cost and produced much lower figures than a
subsequent study (Smulders, 2012) that used figures based on salaries for bookkeepers that
indicated that tax complicance costs (mainly VAT) were over 25 percent of turnover for
the smallest businesses. However, since the survey data also implied that for these firms all
bookkeeping costs amounted to an astounding 60 percent of turnover, these results seem
implausible.
CASE Network Reports No. 106 24
25. THE COSTS OF VAT: A REVIEW OF THE LITERATURE
Another approach to the use of the SCM (again, not limited to VAT taxation)
concerns the performance of public institutions in reducing the cost of doing busi-ness
through streamlined/reduced/abolished regulation. As an example, see
Agence pour la Simplification Administrative (2009), which reports on detailed
cost reductions by government departments in Belgium, in the context of a multi-year
programme for simplification (no such studies seem to have concentrated on
VAT).15
The World Bank/International Finance Corporation (IFC) (in collaboration
with PriceWaterhouse Coopers/PWC) has developed and publicised an ambitious
and wide-ranging effort to quantify compliance costs using the SCM methodolo-gy.
16 The main value added from this work comes from the comparability of re-sults
across countries, and over time. The methodology employed produces cost
estimates in the form of man-hours required to fulfil different types of tax obliga-tions.
It thus shares with other SCM studies the shortcoming of not estimating
actual as opposed to theoretical costs. Nor does it yield cost estimates that can be
related to tax avoidance or tax evasion activities. Paying Taxes is not concerned at
all with administrative costs as discussed in the preceding section. It does, howev-er
produce figures on consumption tax compliance burdens (of which VAT is the
overwhelming component) that are, by design, comparable, as shown in Figure 2.
As one can see, the Paying Taxes methodology results in a tremendous range of
“potential” burden to taxpayers (and by extension to the national economy) from
compliance with taxation requirements, with some of the more recent members
imposing very high hourly requirements (Bulgaria being a stupendous outlier), and
the more advanced/older member economies being at the low end of the
spectrum.17
15 For a discussion of the potentials for application of the SCM to the case of Italy, see
Cavallo et. al. (2007). FIAS (2009) provides a detailed comparison of the SCM and the
World Bank’s ‘compliance survey’ results for regulatory burden measures in Serbia and
Bosnia, and shows, as Hallward-Driemeier and Pritchett, 2011) do in more detail in a
comparison of the Doing Business indicators with the World Bank’s ‘enterprise surveys’,
that there is far more variation among firms in a country than between countries.
16 For the latest Paying Taxes 2011 Report, see
http://www.doingbusiness.org/~/media/fpdkm/doing%20business/documents/special-reports/
paying-taxes-2011.pdf. Also available on the website are a number of more detai-led
country studies, e.g. Ukraine, Armenia, South Africa.
17 Note also that the figure for the Netherlands in Figure 2 is sharply lower than the one
reported by the just-quoted SCM study, presumably reflecting at least in part the simplifi-cation
programme embarked upon in 2005 by the Dutch authorities, which has resulted in
steadily declining hours for complying with taxes (as documented by the various Paying
Taxes reports).
25 CASE Network Reports No. 106
26. Luca Barbone, Richard M. Bird, Jaime Vázquez Caro
Figure 2. Consumption tax compliance burdens
350
300
250
200
150
100
50
0
Hours for
Compliance
Bulgaria
Poland
Slovakia
Latvia
Belgium
Hungary
Portugal
Greece
Slovenia
Spain
Austria
Romania
Lithuania
Germany
Cyprus
Denmark
Netherlands
Czech Republic
Sweden
United Kingdom
Italy
Ireland
Luxembourg
Estonia
France
Finland
Source: Paying Taxes 2011.
The Paying Taxes 2011 study also offers six general lessons and one interesting
observation on the relative importance of the drivers for compliance costs (in
terms of time requirements), based on the universe of all the 145 countries where
VAT (or a VAT-equivalent tax) is present. These lessons are:
It takes less time on average when VAT is administered by the same tax
authority as corporate income taxes (a similar lesson on the benefits of
different taxes being administered by the same authority is drawn by a
recent World Bank study on costs and benefits of integrating tax
administration (World Bank, 2010));
It takes less time on average in countries where business uses online filing
and payment (see on this OECD (2010), esp. Tables A8 and A12);
The frequency at which VAT returns are required impacts the time to
comply;
The more information is required in the VAT return, the more time is
needed18;
18 To cite a rather old example (Bird, 1999), in the early 1990s at around the same time the
VAT return in the UK, a country with perhaps the most complex VAT in the EU in some
respects, was simplified to one page, Poland, then a relatively new VAT adopter, increased
the number of items required on VAT forms from 61 to 105 on a form that called for 12
separate arithmetical manipulations. The design of tax forms – the direct interface (wheth-er
in paper or web form) between taxpayers and the administration – and in particular not
asking for information that is not directly relevant and is seldom used, remains an im-portant
and too often inadequately considered driver of compliance costs.
CASE Network Reports No. 106 26
27. THE COSTS OF VAT: A REVIEW OF THE LITERATURE
The requirement to submit invoices or other documents with the return
adds to compliance time19;
Changes to the rules and regulations can increase compliance time.
An interesting observation is that there is a positive correlation between the
VAT compliance burden and the time delay in receiving a VAT refund.
3.2.2. Comparison between SCM measures and the TCC literature
The definition of compliance costs by Evans (2008) cited earlier in many ways
approximates more or less what the SCM administrative cost measure attempts to
measure through its survey-cost allocation procedure. In addition, however, as
mentioned above taxes may occasion both psychic and social costs that are obvi-ously
not included in such measures. Such costs, however, seem sufficiently polit-ically
relevant to be recognised in some EU-related work in general terms as costs
of irritation or perceptual aspects of taxation that should be taken into account in
developing ways of redressing problems with the present VAT system. Obviously,
such “soft” notions are difficult to quantify and even harder to interpret meaning-fully.
A more important difference between most compliance cost studies and the
SCM work mentioned earlier is that the latter explicitly excludes three components
of compliance cost included in most other studies in the TCC literature:
costs (and benefits) not directly reflected in outlays or attributable to
simply being in business rather than being taxed,
costs incurred by others than direct taxpayers,
and costs related to activities facilitating not tax payment but tax non-payment
through (legal) tax avoidance or (illegal) tax evasion.
The first of these exclusions is presumably in accordance with the mandate of
the EU studies to measure the direct administrative burden – as defined earlier –
on taxpayers. While the EU studies do clearly try to disentangle tax compliance
costs from business-as usual or core accounting costs they inevitably do so, as did
earlier studies (like Plamondon, 1993) on the basis of expert judgments that are
inherently rather arbitrary. On the other hand, this approach deliberately omits
19 Even if invoices are not required to be submitted with returns – as is the case, for examp-le,
with e-filing, in countries (like Ukraine, for example) in which taxpayers perceive that
they continue to face a high probability of having to present their full documentation to the
tax department in any case they are likely to incur as high compliance costs as if they still
had to file invoices with returns (IFC, 2009).
27 CASE Network Reports No. 106
28. Luca Barbone, Richard M. Bird, Jaime Vázquez Caro
some relevant resource costs (and benefits) of VAT compliance. Again, most of
these factors were set out fairly clearly in the pioneering book by Sandford et. al.
(1981) such as the opportunity cost of cash-flow benefits (and costs) and the pos-sible
managerial benefits accompanying the requirement for better accounting in a
VAT system. Evans (2008) refers to estimates including such factors as estimates
of ‘social’ rather than ‘taxpayer’ costs.
One reason such omissions matter is because they may affect the significance
of the results emerging from the SCM approach. As an illustration, note that the
cash-flow aspects of public and private costs do not cancel out because the two
sectors can borrow at different rates. Moreover, within the private sector any gains
from such interest-free loans are presumably much more valuable to smaller busi-nesses
facing higher borrowing rates. Smaller firms are also of course those most
likely to gain from having ‘better accounting’ practices forced upon them for tax
compliance purposes. Both these factors may to some extent mitigate the market
regressivity of the “gross” VAT compliance costs reported in most studies.
A point that is not often mentioned in the literature is also related to the nature
and size of businesses. Consider two businesses, both of equal size but one en-gaged
in manufacturing and one in services. Both have the option (common in
many countries) of paying a ‘presumptive’ (flat-rate output tax) or being in the
VAT system. The service business, which purchases little from other firms, has
little to gain by recouping input VAT and, if it is mainly B2C, much to gain by
being subjected to a lower output tax (and of course even more if it is completely
outside the system, e.g. in the informal sector). The manufacturing business by
definition is more dependent on purchased inputs and is also more likely to sell to
e.g. distributors rather than final consumers directly. Hence it has much more B2B
on both sides of the sales-purchase journal. Its calculations in choosing to opt out
of VAT are thus more difficult than those of the service firm and depend in part on
how its payment terms to its suppliers and its customers are related to each other
and to the payment (and grace) periods of the VAT system as well as on the rela-tive
compliance costs of the full VAT vs. the simplified systems provided in most
EU countries for small businesses. In principle, if small manufacturers sell mainly
to VAT registered firms, they would presumably choose to register voluntarily
even if their level of operation is below the VAT threshold. However, in countries
with large and persistent ‘informal’ sectors, in which considerable trade takes
place among non-registered firms, in principle the choice may be much less clear
although in practice the overwhelming evidence is that in such countries firms do
their best either to stay outside the VAT system completely or, at least, to do as
CASE Network Reports No. 106 28
29. THE COSTS OF VAT: A REVIEW OF THE LITERATURE
much of their business as possible in cash to avoid being subjected to the VAT
regime.20
3.3. Cross-Border Transactions in the EU and Compliance-
Administrative Burdens
As discussed, the theoretical literature and some empirical evidence point to the
multiplicity of requirements of VAT (as for other taxes and administrative rules)
as a direct driver of the compliance costs firms have to bear (and, to some extent,
also of the costs borne by tax administrations). Multiple VAT rates and exemp-tions
oblige firms to keep more complex accounting codes and records. Further-more,
EU-based taxpayers face additional burdens when they engage in interna-tional
trade, both intra-EU and outside the EU. In addition to having to comply
with domestic regulations, exporters to other EU members have to accommodate
importing countries’ specific sets of rules affecting their cross-border transactions.
Differential requirements for dealing with different tax administrations are the
determinants of the intra-European-generated additional transaction costs: to take
an extreme example, even a small number of transactions with a country can im-pose
a large cost, if it obliges a firm to set up and maintain a separate accounting
code and recording system. So long as the application of the VAT rules across the
27 member States varies, small businesses will undoubtedly continue to have con-siderable
difficulty in understanding, let alone complying with, intra-EU trade. For
such firms, intra-EU trade may thus be at least as burdensome, and perhaps even
more so, than trading with countries outside the EU. Unfortunately, there appears
to be little empirical evidence bearing on this issue
The introduction of the Single Market was meant to result in a reduction in
compliance costs from intra-EU trade, chiefly through the abolition of customs
declarations21. Verwaal and Cnossen (2002) have however argued that the statisti-
20 For example, while maintaining perfectly good ‘VAT books’ on turnover, they may
under-report much of their sales (to people who do not report the purchases to the VAT
authorities), pay much of their rent and payroll in cash (to people who do not report their
income), and purchase smuggled or under-invoiced imported goods with foreign exchange
obtained in the black market and smuggled out to pay their suppliers (who are also pro-bably
not reporting the income in their own countries). As in the case of cross-border trade
discussed next, it generally takes a ‘network’ of evaders for firms successfully to evade
VAT on domestic sales.
21 We can also surmise that, following the adoption of the Single Market, administrative
costs for national tax agencies may have increased as tax administrations had to quickly
29 CASE Network Reports No. 106
30. Luca Barbone, Richard M. Bird, Jaime Vázquez Caro
cal requirements that were put in place to allow identification of VAT-taxable
transactions and to help record trade among member states (the Intrastat system)
have resulted in a substantial burden for exporters, which they estimate at 5 per-cent
of the value of trade, with wide variation according to size (and country).
Interestingly but not surprisingly, the availability of e-filing systems is a major
reducer of compliance costs.22 These findings (based on a survey conducted in
1996) are quite sobering compared to previously-published studies, such as the
1997 assessment by the European Commission (Commission of the European
Communities (1997) which argued that compliance costs for firms engaging in
intra-EU transactions had been reduced by approximately two-thirds. Unfortunate-ly,
again there seems to have been little subsequent empirical examination of these
questions. A partial exception is the European Tax Survey of 2004 (Commission
of the European Communities, 2004a). This work consisted of a survey of roughly
700 European enterprises subject to VAT taxation, some of which engaged in in-tra-
European trade. It found that “VAT compliance costs appear particularly high
for companies that undertake activities in other EU member states without having
a permanent establishment there and companies that incur VAT on inputs in other
EU member states” – this being a subset of all companies engaging in intra-EU
trade. However, due to the nature of the survey instrument, it was difficult to point
to specific factors, other than “administrative complexity” that could be addressed
by policy.
Verwaal and Cnossen indeed offered policy suggestions (including the aboli-tion
of the Intrastat System for VAT-liable persons with intra-EU transactions, and
a system of compensation for firms with small amount of intra-EU transactions).
The EU, in turn, modified the Intrastat system in 2004, with a view to making it
more transparent (but not, apparently, less onerous for firms), see European Par-liament
(2004). In the absence of follow-up surveys, it is difficult to gauge the
provide access to the new ex-post filing and IT systems to deal with the new procedural
dimensions of the tax. But again, no cost accounting of tax administrations was found in
the literature to substantiate or refute this hypothesis.
22 Firms engaging in EU trade beyond certain thresholds are expected to file EC Sales lists
for VAT purposes. These lists include details on individual transactions and VAT identifi-ers
of corresponding traders, and are to be used by tax authorities of the trading countries
to verify the legitimacy of the VAT claims that may arise (see the following discussion on
the problem of Carousel Trade). These Sales Lists can be filed, depending on the individu-al
countries, manually, electronically or via the internet. See for an example of the UK
system:
http://customs.hmrc.gov.uk/channelsPortalWebApp/channelsPortalWebApp.portal?_nfpb=
true&_pageLabel=pageImport_InfoGuides&propertyType=document&id=HMCE_PROD
_009770.
CASE Network Reports No. 106 30
31. THE COSTS OF VAT: A REVIEW OF THE LITERATURE
extent to which the Verwaal and Cnossen findings have been superceded by the
subsequent reforms in administrative requirements.
It should also be noted that at least some of the compliance costs identified by
Verwaal and Cnossen are not directly related to the existence of the VAT per se,
but rather to statistical requirements which presumably might exist even in the
absence of a VAT (though linked to VAT reporting in the present institutional
context). There is an interesting perspective on the issue of the statistical burden
coming from Intrastat itself: “For all trade operators involved, Intrastat meant a
lighter workload compared with the previous system before 1993 where any intra-
Community trade transaction had to be declared and presented to Customs. But in
these times the respondents were often not aware of the fact that their reporting
obligations for foreign trade statistics were fulfilled when lodging a Customs dec-laration.
With the introduction of the Intrastat system the statistical reporting bur-den
became apparent.” (European Commission – Eurostat, 2007).
Overall, there is surprisingly little recent empirical evidence on the actual com-pliance
costs borne by firms in EU countries attributable to cross-border trade
within the EU or with non-EU countries. As an example, the SCM (and Paying
Taxes) studies typically do not consider firms engaged in exporting activities.
Although extending coverage to such firms would raise no new conceptual prob-lems,
it could be potentially costly to implement, since cross-border costs may
vary from country to country (both within the EU and outside).
3.4. Private Sector Contribution to Compliance, Compliance Costs
and VAT Fraud Analysis
It is appropriate in our review to touch even briefly on an important (and grow-ing)
element in compliance (and fraud analysis) practice–the presence of private-sector
advisors. Casual web searches reveal substantial offerings by practitioners.
Indeed, judging from the great amount of possibly good advice found on the Web,
the “compliance market” is probably larger than the “fraud market”.23
23 An alternative interpretation might perhaps be that the prevalence of private providers of
tax advice on how to comply may reflect to some extent the inadequacy of – or lack of
trust in – official advice: to the extent there is any truth in this argument, increased private
compliance costs are clearly to some extent at least substituting for public administrative
costs, although no one seems to have considered seriously the costs and benefits of such a
substitution (though see the general considerations in Shaw, Slemrod and Whiting, 2010).
31 CASE Network Reports No. 106
32. Luca Barbone, Richard M. Bird, Jaime Vázquez Caro
The compliance market offers complete and partial solutions to compliance and
compliance costs by making available compliance alternatives claimed to be both
legal and more efficient. For example, SAP’s well known Integrated Management
Information System (IMIS) offers a VAT add-on to SAP24, fusing VAT compliance
with accounting compliance. Others – such as IBM – have developed VAT country-specific
software and made it part of their IMIS.25 Still others have developed paral-lel
systems like central banking transactions that may in the end be useful for VAT
tracing.26 Finally, the private sector offers ethical perspectives on compliance and
fraud.27 This work complements and broadens some of the perspectives of the stud-ies
commissioned by the EC from other private tax advisory firms.
3.5. What do the Studies Show?
Table 2 (based largely on Evans (2003), supplemented with additional studies
from Vaillancourt, Clemens and Palacios (2008) and augmented with the SCM VAT
studies discussed above) gives a bird’s eye picture of existing quantative studies of
VAT-related compliance costs for various classes of taxpayers in a number of coun-tries.
All these studies (along with those earlier reviewed by Cnossen (1994), agree
to different extents with three ‘big lessons’ put forth by Evans (2008), which apply
to all compliance costs and certainly to those associated with VAT:
1. Compliance costs are high and significant;
2. Compliance costs are regressive;
3. Compliance costs are not falling over time.
Regarding the first conclusion, Evans (2008), referring to overall taxation,
sums up by saying that “the studies suggest that compliance costs of such taxes are
typically anywhere between two percent and ten percent of the revenue yield from
those taxes; up to 2.5 percent of GDP; and usually a multiple (of between two and
24 A turn-key solution to overcome the pitfalls and shortcomings within SAP’s VAT deter-mination
and reporting logic. http://www.meridianglobalservices.com/vat-add-on-for-sap/.
25 IBM - Sterling Commerce June 24, 2010 http://eeiplatform.com/2350/survey-companies-
risk-fines-for-non-compliance-with-cross-border-invoicing-regulation/.
26http://www.europa-nu.
nl/id/vil6ib65lmzi/nieuws/nadere_toelichting_op_enkele_europese?ctx=vg9pk7ho53zu.
27 Clients: Increased emphasis on corporate compliance. Calling time on international
bribery. Clifford Chance. Available at:
http://www.cliffordchance.com/about_us/annual_reviews/annual_review_2010/clients/calli
ng_time_on_international_bribery.html.
CASE Network Reports No. 106 32
33. THE COSTS OF VAT: A REVIEW OF THE LITERATURE
six) of administrative costs”. Thus, as shown in Table 2, Sanford et. al. (1981) and
Sanford et. al. (1989) estimated costs equivalent to 9 and 3.7 percent of VAT col-lections
for the UK, Pope, Fayle and Chen (1993) showed compliance costs of
over 2 percent of collections for the Australian WST, and Hasseldine (1995) esti-mated
VAT compliance costs of 7.3 percent of revenues in New Zealand. The
picture is more mixed for more recent studies, particularly those using the SCM
methodology. As discussed above and shown in Table 2 we find studies at the
lower end of the spectrum indicated, and some that are much higher. Witness for
instance the notably low estimates, in percentage of GDP and of VAT collections,
for Denmark (SCM, 2005) at 0.3 percent, but also the considerably higher esti-mates
for Slovenia (Blazic, 2004) and Croatia (Klun, 2003) (respectively 25 per-cent
and 8 percent of VAT collections), and presumably for Bulgaria and other
new member states, based on the “hourly burden” data of the Paying Taxes study
discussed above. This wide spectrum of estimates most likely reflects different
country circumstances and, as discussed, the different methodologies applied by
authors at different points in time.
On the whole, as one might expect, the extant studies also suggest that adminis-trative
costs are absolutely and relatively less burdensome than compliance costs.
Those studies that do address administrative costs suggest that they rarely exceed
one percent of revenue yield, and more usually come in well below one percent. As
noted, few reliable estimates can be obtained for VAT administrative costs only.
The regressivity of the compliance burden of taxation, and VAT in particular,
which can be taken as definitively established in the literature, in particular stems
from the large diseconomies of scale involved in complying with tax requirements,
together with the learning curve effect that militates strongly against small firms
(Evans, 2008; see also DeLuca et. al., 2007 for the USA, etc.). To quote Cnossen
(1994), “…all studies emphasise that the compliance costs of the VAT, as a per-centage
of sales, fall with exceptional severity on small businesses”. Of course, as
has also been shown in the literature, much the same can be said with respect to
most if not all taxes since most involve some fixed costs, and such costs invariably
decrease as the size of a business expands.
And finally, Evans (2008) makes the strong point that compliance costs are
perceived to be an on-going cause for concern, and a problem not improving over
time. It is interesting to note, however, that several of the studies surveyed here in
the spirit of the SCM, and applied to specific strategies of burden reduction by
national governments and EU member states perhaps give hope of the possibility
of seeing a reduction in such burdens (witness the Belgian case cited above, as
well as the indications for the UK, Denmark, the Netherlands). Given that many
such programmes of burden reduction have only recently been put into operation,
time will tell whether they can be successful and successfully sustained.
33 CASE Network Reports No. 106
34. Luca Barbone, Richard M. Bird, Jaime Vázquez Caro
Table 2. Summary of major published studies of VAT taxation operating costs since 1980
European Studies
Year of
publication
(Years un-der
review)
Author(s) Country (popu-lation
studied)
Taxes
studied
1. Methodology
2. Sample frame
3. Respondents
4. Response rate
Major outcomes
Compliance costs Administrative costs
1981
(1978-1980)
Sandford,
Godwin,
Hardwick &
Butterworth
UK
(UK VAT regis-tered
traders and
their advisers)
Value added
tax
1.Documentary anal-ysis
for
administrative costs;
for compliance costs
(a) postal
survey, followed by
(b) telephone and
personal interviews
and (c) interviews
with advisers (sample
and responses not
published for this
element)
2. (a) 9,094 (b) 445
3. (a) 2,799 (b) 263
4. (a) 31% (b) 59%
Gross compliance costs for VAT estimated as £392m in 1977-78,
and administrative costs £85m.
Total operating costs of c. £480m represented 11% of VAT reve-nue;
VAT compliance costs
“exceptionally regressive in their incidence” (and administrative
costs also likely to be regressive);
cash flow benefits (£73m) and managerial benefits (difficult to quan-tify)
exacerbate the regressiveness; net compliance costs affected
by size of firm, sector (relatively lower compliance costs in
primary production and higher in financial and services sector),
payment or repayment situation
1989
(1986-87)
Sandford, God-win
&
Hardwick
UK
(UK VAT regis-tered
traders)
Value added
tax
1. Postal survey
2. 3,000
3. 680
4. 24%
Aggregate compliance costs were
£791m (3.69% of revenue yield)
and cash flow benefits (dispropor-tionately
enjoyed by larger firms)
were £580m; net compliance
costs were 1% of revenue yield;
compliance costs very regressive;
compliance costs fallen since
1977-78
Administrative costs of £220m
in 1986 -87 were 1.03% of reve-nue
yield
1989
(1987)
Bannock &
Albach
UK & Germany
(UK and German
Value added
tax
1. Postal survey (a)
UK and (b)Germany,
Dissatisfaction with VAT system
was much greater among smaller
Not addressed
CASE Network Reports No. 106 34
35. THE COSTS OF VAT: A REVIEW OF THE LITERATURE
Year of
publication
(Years un-der
review)
Author(s) Country (popu-lation
studied)
Taxes
studied
1. Methodology
2. Sample frame
3. Respondents
4. Response rate
Major outcomes
Compliance costs Administrative costs
businesses) with very limited
telephone follow up
(15 calls in each
country)
2. (a) 600 (b) 800
3. (a) 262 (b) 197
4. (a) 44% (b) 25%
firms in the UK than in Germa-ny,
and compliance costs for
smaller traders were significantly
higher in the UK than in Germa-ny
1994
(1992-93)
National Audit
Office
UK
(UK VAT trad-ers)
Value added
tax
1.Update of earlier
VAT
surveys conducted by
Sandford et al (1981
& 1989)
2. Not relevant
3. Not relevant
4. Not relevant
Compliance costs of VAT were
£1.6b offset by compliance bene-fits
(cash & management) of
£750m; compliance costs regres-sive
Administrative costs were
£399m
2002
(2000)
Hasseldine &
Hansford
UK
(business taxpay-ers)
Value added
tax
1. Postal survey
2. 6,232
3. 1,449
4. 23%
Increased compliance costs are
associated with increased turno-ver,
newly registered businesses,
increased complexity and per-ceived
psychological costs; no
significant differences in patterns
of core compliance costs and
planning costs; businesses with
computerised systems faced rela-tively
higher compliance costs
than businesses with manual
procedures
Not addressed
2002 Verwaal and
Cnossen
Netherlands VAT, Trade
statistical
requirements
1. Postal Survey of
firms
2. 2998
Statistical requirements for intra-
EU trade linked to VAT reporting
system impose on average a 5
Not addressed
35 CASE Network Reports No. 106
36. Luca Barbone, Richard M. Bird, Jaime Vázquez Caro
Year of
publication
(Years un-der
review)
Author(s) Country (popu-lation
studied)
Taxes
studied
1. Methodology
2. Sample frame
3. Respondents
4. Response rate
Major outcomes
Compliance costs Administrative costs
3. 642
4. 21.5%
percent cost on firms with wide
variation. E-filing contributes to
reduce costs.
2003 Klun Slovenia Value Added
tax
Compliance costs between 1.7
and 2.5 percent of GDP.
Not addressed
2004 Blazic Croatia Taxation for
small business-es
(including
VAT)
1. Interviews Total compliance cost 0.8 percent
of GDP. Total compliance cost for
VAT 0.2 percent of GDP.
Not addressed
2004 Commission of
the European
Communities
EU VAT and other
taxes
1. Survey of 700
enterprises
VAT compliance costs appear
particularly high for companies
that undertake activities in other
EU member states without having
a permanent establishment there
and companies that incur VAT on
inputs in other EU member states.
Not addressed
2005 SCM Network Denmark, Neth-erlands,
Nor-way,
Sweden
Value Added
Tax
SCM methodology CC (percent of GDP):
Denmark 0.03%
Netherlands 0.17%
Norway 0.06%
Sweden 0.07%
Not addressed
2006 KPMG UK Value Added
Tax (together
with assess-ment
of overall
administrative
burden)
SCM Methodology The “administrative” (compli-ance)
burden of UK tax regulation
is £5.1 billion. VAT accounts for
£1 billion, or 20 percent of total.
(respectively 0.3 and 0.1 percent
of GDP).
Not addressed
CASE Network Reports No. 106 36
37. THE COSTS OF VAT: A REVIEW OF THE LITERATURE
North American Studies
Year of
publication
(Years un-der
review)
Author(s) Country (popu-lation
studied)
Taxes
studied
1. Methodology
2. Sample frame
3. Respondents
4. Response rate
Major outcomes
Compliance costs Administrative costs
1993
(1993)
Plamondon Canada
(Canadian small
businesses)
Goods and
services tax
1. Interviews (face to
face) conducted by
accountants with
questionnaire
2. 200
3. 200
4. 100%
Compliance costs were not as
high as previous studies had
shown, but were regressive;
businesses using computers for
accounting routines had compli-ance
costs 20% to 40% lower
than those operating manually
Not addressed
1993
(1995)
General Ac-counting
Office (US)
USA
(Federal admin-istration)
Value added
tax
1. Estimate of admin-istrative
costs of a
value added tax
2. Not relevant
3. Not relevant
4. Not relevant
Not addressed Recurrent administrative costs
of a value added tax would be
between US$1.22b and
US$1.83b, with 70% of those
costs related to audit work;
transitional costs of introducing
a value added tax would be
US$800m; costs would vary
with key design features of the
tax, and a simple single rate,
broad-based VAT would mini-mise
administrative costs
1995
(1995)
Plamondon Canada (Canadi-an
small busi-nesses)
Goods and
services tax
(Quick meth-od
of ac-counting
for
GST)
1. Interviews (face to
face) conducted by
accountants with
questionnaire
2. 200
3. 200
4. 100%
Small businesses were not using
the Quick method of accounting
for GST due to a lack of aware-ness;
those who knew of it but
did not use it were not overly
concerned about compliance
costs; savings in tax were
more important than savings in
compliance costs
Not addressed
37 CASE Network Reports No. 106
38. Luca Barbone, Richard M. Bird, Jaime Vázquez Caro
Year of
publication
(Years un-der
review)
Author(s) Country (popu-lation
studied)
Taxes
studied
1. Methodology
2. Sample frame
3. Respondents
4. Response rate
Major outcomes
Compliance costs Administrative costs
2008 Government
Accountability
Office (US)
Australia, New
Zealand, Canada,
France, United
Kingdom
VAT Mixed methodology Some available data indicate a
VAT may be less expensive to
administer than an income tax.
Australasian and South East Asian Studies
Year of
publication
(Years un-der
review)
Author(s) Country (pop-ulation
studied)
Taxes
studied
1. Methodology
2. Sample frame
3. Respondents
4. Response rate
Major outcomes
Compliance costs Administrative costs
1993
(1990-91)
Pope, Fayle &
Chen
Australia
(Australian
businesses)
Wholesale
sales tax
1. Postal survey
2. 2,467
3. 593
4. 24%
Net compliance costs of WST
were $201m, or 2.1% of revenue
yield; compliance costs were
highly regressive; WST generat-ed
a cash flow cost overall rather
than a benefit
Not addressed
1995 Hasseldine New Zealand VAT Compliance costs amount to 7.3
percent of VAT collecations
Not addressed
2002
(1998-2000)
Rametse &
Pope
Australia
(Western Aus-tralian
business
taxpayers)
Start -up costs
of the Goods
and Services
Tax (GST)
1. Postal survey
2. 3,199
3. 868
4. 27%
Estimated GST start -up compli-ance
costs for small businesses
were AUD$7,600; this included
owner/manager time of 131
hours; start -up costs were consid-erably
higher than official gov-ernment
estimates
Not addressed
2002
(Jun 1999 -
Jun
2001)
Tran -Nam &
Glover
Australia
(small business
taxpayers)
Transitional
costs of the
Goods and
Services Tax
1. Case study
2. 31
3. 31
4. Not relevant
Small businesses incurred net
transitional compliance costs of
AUD$4,853 (mean) or
AUD$2,393 (median); (median
Not addressed
CASE Network Reports No. 106 38
39. THE COSTS OF VAT: A REVIEW OF THE LITERATURE
Year of
publication
(Years un-der
review)
Author(s) Country (pop-ulation
studied)
Taxes
studied
1. Methodology
2. Sample frame
3. Respondents
4. Response rate
Major outcomes
Compliance costs Administrative costs
(GST), Aus-tralian
Busi-ness
Number
(ABN), Pay
As You Go
(PAYG) and
Business
Activity
Statement
(BAS)
was preferred); in addition to
monetary costs, small business
taxpayers appeared to suffer sub-stantial
psychological costs during
the transitional period
Source: An extended and updated table from Evans, 2003.
39 CASE Network Reports No. 106
40. Luca Barbone, Richard M. Bird, Jaime Vázquez Caro
3.6. Concluding Remark
While it is difficult to summarise the lessons of the vast literature just briefly
reviewed, perhaps one quote from the recent Mirrlees Review is appropriate to end
this section:
“Administrative and compliance costs depend on a wide range of factors,
including the complexity of the tax, characteristics of the tax base, structure
of tax rates, frequency of reform, and organisation and efficiency of the tax
authority. Taxes should therefore be kept as simple and stable as possible.
In other areas, there is a trade-off between administrative and compliance
costs: for example, whether it is the tax authority or taxpayers who have re-sponsibility
for calculating tax liability. Providing help and guidance in-creases
administration costs, but reduces compliance costs.” (Shaw et. al.,
2010).
CASE Network Reports No. 106 40
41. THE COSTS OF VAT: A REVIEW OF THE LITERATURE
4. Compliance Costs and Non-
Compliance
“Missing Trader Intra-Community (MTIC) VAT fraud is a large-scale or-ganised
criminal attack on the EU VAT system. The most serious form of the
fraud – known as carousel fraud – involves a series of contrived transac-tions
within and beyond the EU, with the aim of creating large unpaid VAT
liabilities and fraudulent VAT repayment claims”.28
4.1. Overview
A large and growing literature has focused in recent years on the increasing ev-idence
of VAT-associated fraudulent practices in the EU. This is a major concern
for businesses that see themselves at a competitive disadvantage, as well as na-tional
and EU policy makers, as evidenced in recent communications on the matter
(Commission of the European Communities, 2007). This section reviews several
contributions to the debate on evasion and fraud, and places them in the context of
the discussion on compliance and administrative costs reviewed in the previous
sections.
VAT fraud was recognised by the EC as an objective problem and made part of
the EC strategy in 2003. In 2007 the fight against VAT fraud became a major con-cern
of EC strategic thinking. The recognition of VAT fraud as “large-scale organ-ised
criminal attack on the EU VAT system” as defined in the British report cited
above signals both awareness and a notion of magnitude.
The literature points to several institutional reasons to explain VAT fraudulent
practices (in addition to the behavioural variables which we will review below in
the econometric studies of the VAT gap). VAT specific and EU general policies
have been recognised as major determinants of the compliance and enforcement
28 Measuring Indirect Tax Losses–2007 HM Revenue and Customs, available at
http://s3.amazonaws.com/zanran_storage/www.hmrc.gov.uk/ContentPages/13568349.pdf.
41 CASE Network Reports No. 106
42. Luca Barbone, Richard M. Bird, Jaime Vázquez Caro
environments that facilitate the emergence of fraud29. These include (i) parametric
issues on base, rates, exemptions, zero rating, registration and return filing thresh-olds,
refunding of VAT specific rules and the existence of parallel small taxpayers
regimes; and (ii) two broadly recognised general EU policy principles that affect
VAT fraud: the intra-European single market in force since 1993, and the applica-tion
of the subsidiarity principle to tax administration which generates a second
layer of differentiation in the actual application of the laws (See Keen and Smith,
2007; and Cnossen, 2009).
4.1.1. VAT parametric issues
The parameters of individual countries’ VAT systems can create a number of
fiscal complexities and risks to VAT compliance attitudes. This can result from
either the way in which bases, taxpayers and rates are defined or the way in which
the compliance process is structured. Purchases and refund entitlements, for in-stance,
are a most important matter in the case of intra-EU trade fraud.
This type of VAT fraud emphasised in the recent EU literature occurs when
registered sellers charge VAT and buyers request a refund or simply include the
input VAT in their declarations and the seller does not declare and pay the tax.
Bogus traders, in fact, issue what are in effect deferred “cheques” or payment or-ders
in the form of invoices that may be used as input credits in a future VAT re-turn
and may even generate refunds from countries’ treasuries; then they simply
disappear (see Cnossen, 2009; and Harrison and Krelove, 2005). See Table 3 for
information on VAT refund practices from the latter paper.
Table 3. VAT Refunds: A Review of Country Experience
Refund Levels In many countries, levels exceed 40% of gross VAT collections.
40% of survey respondents repay a third or more of gross VAT
collections.
Countries with refund levels under 20% are mostly in Africa,
Asia and Latin America.
Within regions, refund levels are similar among countries with
similar VAT systems and economic conditions.
29 The claim of VAT advocates about “self-enforcing” in the early stages of VAT imple-mentation
around the world (1960s through 1980s) was questioned by Hemming and Kay
(1981) in the early 80s. Michael Keen and Stephen Smith (2007) rejected it because of the
implicit assumption that both buyer and seller were compliant taxpayers. These authors
questioned the claimed “self-enforcing” feature of VAT because of another possible out-come:
both seller and buyer gain by cheating in a context of poor targeting and control. For
this reason the argument is illusory.
CASE Network Reports No. 106 42
43. THE COSTS OF VAT: A REVIEW OF THE LITERATURE
Time frame for
refunding tax
credits
In most developed countries, refunds are paid within four
weeks of a refund claim being made.
In developing and transitional countries, it often takes several
months and sometimes more than a year to process refund
claims. This can seriously undermine the competitiveness of
the export sector.
Notwithstanding that most countries have statutory deadlines
for making refunds, these are often not met by tax authorities.
Reasons for delay-ing
payment of
refunds
Prevalence of fraudulent claims and differences in strategies to
cope with them:
o In less advanced tax administrations: Pursuance of time-consuming
and labor-intensive processes to verify claims
before approving refunds.
o Effective and efficient tax administrations: Refund related
fraud is tackled as part of a broader VAT-compliance
strategy based on risk management principles and limit
pre-refunding verifications to high-risk claims.
When state budgets are under pressure and tax collection tar-gets
are not being made. This happens when administrations
lack suitable forecasting and monitoring systems to anticipate
refund levels and do not set aside sufficient funds.
Although these delays are more likely in transitional or devel-oping
countries, it is not confined to them.
VAT refund abuse
and fraud
All countries report VAT refund abuse, but most have difficulty
estimating the scale of associated revenue losses.
While the nature of VAT refund abuse is similar across coun-tries,
the environment in which it occurs and the approaches to
counteract it vary between countries.
VAT refund abuse is only a component of VAT fraud; in many
countries audit resources focus mainly on VAT refunds and do
not pay adequate attention to other related risks.
Strategies for
controlling late
payment of re-funds
90% of the countries reported that their tax authorities are
bound by law to making refunds within a prescribed timeframe,
generally 30 days.
40% go further, providing by law for interest to be paid on late
refunds. These measures also demand safeguards from fraud-sters,
which range from providing tax officials with statutory
powers to conduct audits and verification checks to requiring
security or bank guarantees from traders who seek refund. In
60% of surveyed countries there is a mandatory carry-forward
period to limit the number of refund claims.
Source: Harrison and Krelove (2005). Countries that responded to the survey were Algeria,
Azerbajan, Bolivia, Bulgaria, Cambodia, Cameroon, Canada, Chile, Colombia, El Salva-dor,
France, Hungary, Indonesia, Ireland, Italy, Kazakhstan, Kenya, Latvia, Mexico, Mon-golia,
Morocco, Mozambique, Netherlands, New Zealand, Pakistan, Peru, Poland, Roma-nia,
Russia, Singapore, Slovak Republic, South Africa, Sweden, Tanzania, Ukraine, United
Kingdom.
43 CASE Network Reports No. 106
44. Luca Barbone, Richard M. Bird, Jaime Vázquez Caro
Similar frauds may also happen within a country, particularly in the presence of
a large informal sector in which difficult-to-track sellers issue invoices and disap-pear
(see Ainsworth, 2008 for some Canadian examples) In a recent careful micro-econometric
study in Brazil, de Paula and Scheinkman (2009) have shown that
there are ‘networks’ of evaders who deal with each other so that when ‘holes’
appear in the VAT chain, whether as a result of legislation (exempt sectors like
transport in Mexico) or administrative problems – whether arising from corruption
or ‘informality’ domestically or cross-border trade as discussed below – they may
spread and eat away more and more of the potential tax base over time unless cor-rected.
4.1.2. Cross-border transactions in a free movement environment
Intra-European free-trade free-movement rules break the domestic VAT chain
on exported goods to other member countries. This generates a VAT collection
loss in exporting countries (through the rebate mechanism) but leaves destination
countries completely open to the behaviour of the taxpayer importing the goods
(Cnossen, 2009). Cnossen and other authors point to lack of proper tax administra-tion,
enforcement and audit practices and/or capacities (see for instance the thor-ough
review reported in GAO (2008).
In intra-EU transactions a common type of fraud involves the trading of VAT
rebate rights in the so-called Carousel (see Box 2, reproduced from Smith, 2007).
The Carousel, the most distinctive fraud for VAT, is a false claim for creditable
VAT paid on inputs or, most dramatically, for a refund based on zero-rated ex-ports,
which break the VAT collection chain at a vulnerable point, the border be-tween
domestic and foreign tax administrations. Carousel fraud exploits the com-bination
of the zero-rating of exports and the deferred payment situation of VAT
periodic return procedures (See Keen and Smith, 2007, p. 13).
An operational challenge these authors identify relates to VAT refunds: en-forcement
has to find an appropriate balance between lax and stringent attitudes
toward refunds, because erring either way creates problems. If too lax, there is too
much incentive for fraud, but an excessively stringent attitude creates high costs
and may end up turning the VAT into a tax on production and exports, defeating
its economic purpose.30 Harrison and Krelove (2005) provide estimates on per-
30 This conundrum is well recognised by the European Commission, see for instance the
2007 document: “It should be kept in mind that all measures that are discussed in the con-text
of the fight against VAT fraud have to respect other EU policies, and in particular the
CASE Network Reports No. 106 44
45. THE COSTS OF VAT: A REVIEW OF THE LITERATURE
centage of refunds over gross VAT collections. A strategy to help lessen this prob-lem
is sometimes called the “gold card scheme” which promises businesses with
good compliance records prompt refund payment.
Box 2. The basic carousel fraud: an illustration
general target of the European Council to achieve, by 2012, a reduction by 25% of the
existing administrative burden (…)” p. 6.
45 CASE Network Reports No. 106