ArticlePDF Available

Assessing Tax Systems Using a Benchmarking Methodology

Authors:
  • DevTech Systems Inc. and USAID

Abstract and Figures

International institutions, such as the World Bank, the International Monetary Fund, and the U.S. Agency for International Development, have been assessing tax system performance and capabilities for decades without having a solid international comparator basis for undertaking these assessments. This paper provides a series of indicators and benchmarks that can help to put such assessments into an international perspective, set specific targets for performance, reform, and modernization, and monitor progress over time.
Content may be subject to copyright.
ASSESSING TAX SYSTEMS USING A
BENCHMARKING METHODOLOGY
By Mark Gallagher
April 2004
Research Paper
Fiscal Reform in Support of Trade
Liberalization
1250 “Eye” Street
Suite 1115
Washington, DC 20005
Web Site:
www.fiscalreform.net
Development Alternatives, Inc. with the Georgia
State University and the Boston Institute for
Developing Economies implements the Fiscal
Reform in Support of Trade Liberalization project.
The U.S. Agency for International Development
funds the Project, under Task Order under SEGIR:
EP, Contract No. PCE-1-00-00-00015-00
Assessing Tax Systems Using a Benchmarking Methodology
CONTENTS
LIST OF FIGURES ............................................................................................................ 1
TABLES ............................................................................................................................. 1
Introduction......................................................................................................................... 1
Development of the Methodology...................................................................................... 3
The Benchmarks ................................................................................................................. 6
Tax Structure and performance....................................................................................... 6
Organization.................................................................................................................. 13
Legal framework........................................................................................................... 17
Enforcement.................................................................................................................. 19
Receipts and collection................................................................................................. 23
Systems and resources.................................................................................................. 23
Next steps.......................................................................................................................... 25
Annex: The Benchmarks as Developed in the Guatemala Study.................................... 27
LIST OF FIGURES
Figure 1............................................................................................................................. 11
TABLES
Table 1: Administrative Costs of Taxation at National Level (mid-1990s) ....................... 9
Table 2: VAT Gross Compliance Ratio (GCR)................................................................ 10
Table 3: Enterprise Income Tax Productivity................................................................... 12
Table 4: Personal Income Tax Productivity...................................................................... 13
Table 5: Functional organization of tax administration.................................................... 14
Table 6: Countries that have established Semi-Autonomous Revenue Authorities......... 15
Table 7: Tax Administrations with Large Taxpayer Units............................................... 17
Table 8: Public employees in the national tax administration institution......................... 20
Table 9: Active taxpayers per tax official......................................................................... 21
Assessing Tax Systems using a Benchmarking Methodology
Mark Gallagher
,
January 2004
Abstract
International institutions, such as the World Bank, the International Monetary Fund, and
the U.S. Agency for International Development, have been assessing tax system
performance and capabilities for decades without having a solid international
comparator basis for undertaking these assessments. This paper provides a series of
indicators and benchmarks that can help to put such assessments into an international
perspective, set specific targets for performance, reform, and modernization, and monitor
progress over time.
Introduction
Until recently, there has been only limited effort to develop comprehensive tools for
assessing tax systems, despite the fact that national governments and international
organizations or foreign assistance agencies, such as the World Bank, the International
Monetary Fund of the U.S. Agency for International Development, have been assessing
tax systems in developing and emerging market countries for decades.
A recent pamphlet produced by Michael Lane of an international consulting firm, Sandler
Travis Trade Advisory Service in Washington, DC, provides a checklist for customs
operations, but offers no international comparator information.
1
Barbone et alia
(undated) provide a “Framework for Diagnosis of Tax System Weaknesses”, which is a
matrix of policy formulation, accountability, and service delivery by performance,
capacity and institutions.
2
This framework provides a good guide for what analysts
should look for when comparing tax systems across countries, but provides no
Director of the Fiscal Reform in Support of Trade Liberalization worldwide program funded by the U.S.
Agency for International Development and an economist with Development Alternatives, Inc. The
opinions expressed here are solely attributable to the author and should not be taken to represent those of
the U.S. Agency for International Development nor Development Alternatives, Inc.
The author thanks Richard Bird, David Dod, George Guess, Ted Kill, and Art Mann for their helpful
comments. All errors and omissions, of course, are attributable only to the author.
1
Lane, Michael (1997) Workbook for Customs Modernization, Global Customs Advisors.
2
See especially page 6 of Barbone, Luca, Arindam Das-Gupta, Luc De Wulf and Anna Hansson (undated)
Reforming Tax Systems: The World Bank Record in the 1990s, World Bank.
ASSESSING TAX SYSTEMS USING A BENCHMARKING METHODOLOGY -- Gallagher
Fiscal Reform in Support of Trade Liberalization – Research Paper
2
comparator data nor, in most instances, even verifiable indicators. Jit Gill (2000)
presents a very useful framework for assessing revenue administration, with a focus on
process and environment but with very few international comparators.
3
More recently,
the World Bank has developed a website devoted to evaluating tax policy and
administration, but this site does not include any international comparators, and it
provides little basis for establishing specific, quantitative goals and targets, although it
does discuss many of the institutional and resource matters that are discussed in this
paper.
4
Moreover, despite the website’s title, it offers very little substance for evaluating
tax system performance.
Bird and Banta (1999) present a number of indicators of fiscal sustainability covering the
basic economy statistics, tax policy, specifically, VAT, tax administration, public
expenditure management, decentralization, and pensions.
5
This set of indicators is very
comprehensive, but is not meant specifically to assess tax systems, though clearly it can
be used that way. In addition, the paper includes data for 26 countries of East and
Central Europe and the Former Soviet Union. In a second paper, Bird (1999), in the
same volume, presents a number of comments and criticisms that were raised during a
conference regarding the indicators system.
6
Some of these are repeated in the final
section of this paper. Building on the shoulders of giants, several of the Bird and Banta
observations are included in this paper.
Das Gupta (2002) strives “towards a framework for tax system performance” to assess
tax and tax administration reform in India using indicators of performance, capacity,
institutions, and exogenous constraints.
7
He creates a number of specific indicators that
he then applies to the pre-reform period and the post-reform period to assess the quality
of the reform. The framework for Das Gupta’s assessment then is to determine if
improvement in India has taken place. This allows judgments of improvement and status
for seven broad rating indicators, ranging from “external constraints” to “policy research
capacity” and “administrative capacity.” For these seven indicators Das Gupta judges
that they have shown some improvement, no change, or some deterioration. Das Gupta
3
Jit B.S. Gill (2000) A Diagnostic Framework for Revenue Administration, World Bank. See:
http://wbln0018.worldbank.org/prem/ps/iaamarketplace.nsf/075c69a32615405f8525689c0051fb88/4cdd74
a05fd3506b8525689c005333bc/$FILE/DIAGFRMWK-5-2000.doc
4
See the World Bank’s “Tax Policy and Administration” website at:
http://www1.worldbank.org/publicsector/tax/themes.html
.
5
Bird, Richard and Susan Banta (1999) “Fiscal Sustainability and Fiscal Indicators in Transition
Economies,” in Shapleigh, Alexander, Fuat Andic and Susan Banta eds. Transition Economies and Fiscal
Reforms: Proceedings of the Conference on Central and Eastern Europe and the New Independent States,
June.
6
Bird, Richard (1999) “Some Reflections on the Conference,” in Shapleigh, Andic, and Banta op. cit.
7
Das Gupta, Arindam (2002) “ Central Tax and Administration Reform in the 1990s,” in Development,
Poverty, and Fiscal Policy, Rao, M. Govinda editor, Oxford University Press, New Delhi.
ASSESSING TAX SYSTEMS USING A BENCHMARKING METHODOLOGY -- Gallagher
Fiscal Reform in Support of Trade Liberalization – Research Paper
3
also judges that these indicators show the status of the Indian tax system to range from
“poor” to “inadequate” to “adequate.”
The EU has been developing its "Draft Fiscal Blueprints" that provide guidelines for
would-be member-nation tax administrations in many areas of administration and even
provide a listing of performance indicators. The Blueprints are fairly comprehensive, but
lack any discussion of comparative "quantitative goals and targets."
It is interesting to note that the Organization for Economic Cooperation and Development
and the Asian Development Bank have both seen fit to produce checklists, questionnaires
and surveys for assessing public expenditure management systems, yet have not produced
similar tools for tax system assessment.
8
This paper goes beyond these attempts and presents benchmarking as a tool for assessing
both tax system performance and the inputs and systems of any tax administration. It is a
means of comparing a set of specific indicators that capture the essence of most any tax
system to either international best or perhaps most relevant practices. The system also
helps to facilitate establishing goals and specific targets for tax system improvement and
modernization. Specific benchmarks can be tracked over time and can show how reform
or modernization efforts are being implemented and even how they contribute to
performance.
Benchmarking can be used not only to compare a country’s tax system with a regional or
international set of norms or comparators, but it can also be used to compare the
condition and performance of the tax system over a period of time, either discrete
snapshots in time or evolution of the benchmark or indicators over a number of years.
The methodology can be generalized to all national-level tax systems throughout the
world.
The next sections provide a discussion of the development of the methodology and the
benchmarks, discussion of many of the benchmarks, and finally, a discussion of how the
benchmarking methodology can be improved.
Development of the Methodology
This benchmarking methodology has been developing over a few years through
application in a number of countries. First, some of the international comparators were
applied in El Salvador, by Julio Piza (1994).
9
Later Gallagher (1995) extended the
application to assess the Guatemalan tax system.
10
Gallagher (1997) applied the
8
See Allen, Richard and Daniel Tommasi (2001) Managing Public Expenditures: A Reference Book for
Transition Countries, OECD, and
9
Piza R., Julio Roberto (1994) Administracion Tributaria de los Impuestos Internos de El Salvador,
USAID, December.
10
Gallagher, Mark (1995) Options for Donor Assistance in Tax Policy and Administration Reform in
Guatemala, USAID/Guatemala, March.
ASSESSING TAX SYSTEMS USING A BENCHMARKING METHODOLOGY -- Gallagher
Fiscal Reform in Support of Trade Liberalization – Research Paper
4
methodology in Nicaragua
11
then later in parts in Tanzania.
12
In 2001, a comprehensive
system was applied and further developed in Guatemala by Mann et al (2001), covering
all tax operations, as well as customs, in 2001.
13
The 2001 Guatemalan case represents
the first time the entire methodology has been applied in a fully systematic way. Most
recently, in December 2003, Gallagher applied the methodology to the Egyptian tax
system and made a presentation of the preliminary results at the USAID Mission to that
country.
Piza (1994) prepared an assessment of tax administration in El Salvador. Piza included a
number of international comparators in his report, including: tax ratio (tax revenue as
percent of GDP), number of tax administrators per 1,000 of national population, and,
comparative tax administration costs.
One indicator that was subsequently developed is called the VAT-Gross Compliance
Ratio, which compares actual VAT collections to potential VAT collections if there were
no evasion, tardiness in payment, or exemptions and exonerations. Other indicators were
developed with time. Some had been developed elsewhere, such as indicators of tax
evasion and efficiency in collections. Tanzi (1996) clearly delineated a number of
criteria of good tax systems, relating to the simplicity of the tax system, which Gallagher
(1997) subsequently converted into specific, quantitative indicators; these were
subsequently developed and applied in Nicaragua.
14
Gallagher (1995) undertook an assessment of tax and customs performance. In this
assessment Gallagher compared a number of other indicators, such as maximum and
minimum income tax rates and collections in percentages of GDP to the other Central
American countries. Gallagher also compared import duty collections and other results
indicators in Guatemala to those in neighboring countries.
The set of indicators was expanded and applied partially in a paper on the Nicaraguan tax
system. Many of the results of the Gallagher (1997) study were incorporated into the tax
and trade reform legislation developed and enacted in the same year. However, the 1997
reforms were almost exclusively focused on policy, especially simplification, and little
attention was paid to administration.
11
Gallagher, Mark (1997) Propuesta de Reforma Tributaria Integral, Núcleo Especial Para Análisis e
Implementación, Managua, January. http://www.fiscalreform.net/library/pdfs/nicaragua--
propuesta_de_una_reforma_tributaria_integral.pdf.
12
No specific report was prepared for Tanzania, rather the methodology helped in the development of the
tax reform agenda of 1998 and 1999.
13
Mann, Arthur, Arturo Jacobs, Mark Gallagher, Harry Alison, James Westrick, Alex Segovia, and Ted
Kill (2001) APLICACIÓN DE MEJORES PRÁCTICAS INTERNACIONALES AL DESEMPEÑO DE LA
ADMINISTRACIÓN TRIBUTARIA DE GUATEMALA: Un Estudio de Benchmarking, USAID/DevTech
Systems, and DAI
.
14
Tanzi, Vito (1991) Public Finance in Developing Countries, Edward Elgar Publishing,England.
ASSESSING TAX SYSTEMS USING A BENCHMARKING METHODOLOGY -- Gallagher
Fiscal Reform in Support of Trade Liberalization – Research Paper
5
In Tanzania the methodology was only partially applied. It showed the need for further
modernization of the tax system in terms of improving the application of the stop or late
filers systems, automated notices, and linking customs and internal revenues information
systems. In addition, one specific indicator – the Gross Compliance Rate for VAT,
discussed later in this paper – was used to project the likely revenues that would be
generated in the first year of the operation of the VAT. This projection, based on the
GCR, was used as the revenue projection in the national budget for 1997/98. As it turns
out, this projection was very accurate, i.e., within one percent of actual collection.
The first full-blown application of the benchmarking methodology was conducted in
Guatemala in 2001, at the request of the Minister of Finance, and funded by the U.S.
Agency for International Development.
15
This study showed the considerable progress
that had been achieved in tax system performance and the institutional development of
the tax administration, and laid out the major areas for continuing support and progress.
Most recently (in December 2003) Gallagher undertook a review of U.S. Government
assistance to the Government of Egypt in tax policy and administration and applied an
abbreviated version of the benchmarking system. The results of the abbreviated
application of the benchmarking methodology were presented in a seminar held at the
U.S. Agency for International Development Mission to Egypt. Several participants in the
seminar expressed the opinion that the methodology would have been useful to establish
a set of specific targets for the U.S. assistance program to monitor progress and
performance.
As mentioned, no single source of comparator data exists for all of the useful indicators
of tax system performance. For some indicators, see tables 1 through 4, data are
available in international sources, such as the IMF’s Government Finance Statistics
Yearbooks. Other data must be extracted from other sources, such as Article IV
Consultations of the IMF, country sources, and from the Spanish Tax Institute.
For the Guatemala project a team of highly skilled and widely experienced specialists
was assembled to develop a useful set of comparator indicators. These are in the Annex
to this paper.
The benchmarks are in a number of important areas, but before going into the structure of
the benchmarks, it would be worthwhile to discuss the make up of the team that was sent
to Guatemala to undertake the benchmarking study.
1. Two public finance economists worked on the study. Their focus was on tax
system performance from the perspective of comparing compliance, collection
efficiency, and changes in tax policy and how these would affect performance.
Between them, these two economists have more than 50 years of experience
working on fiscal reform in developing countries, with many of those years
working on these issues in Latin America, and especially in Central America.
ASSESSING TAX SYSTEMS USING A BENCHMARKING METHODOLOGY -- Gallagher
Fiscal Reform in Support of Trade Liberalization – Research Paper
6
2. One tax administration consultant. This expert had a long and high profile
career with U.S. Internal Revenue Service, has worked on tax administration
reform in several Latin American countries and in other countries of the world,
including Former Soviet Union, Balkans, and Africa.
3. One computer systems consultant, with specialization in tax systems. This
expert has worked on the design and implementation of tax modernization
programs throughout the world, including the U.S. Specifically, in recent years,
he has worked on tax and customs modernization in Central America, specifically
Guatemala, and the Balkans.
4. One customs administration consultant. This expert had a 30 year career with
U.S. Customs and has spent an additional 15 years advising customs officials in
developing countries of Latin America and Asia.
The Benchmarks
Tax system performance indicators come from a variety of sources. Some were created
in the process of developing the benchmarking system itself, whereas others are the result
of other people’s work or standard methodologies. It is also important to keep in mind
that appreciating or assessing tax system performance requires knowledge of the other
factors found in this system.
The benchmarks are discussed in the following sections.
Tax Structure and performance
Number of taxes that comprise the top 75% of receipts: This benchmark provides some of
the degree of dispersion or, on the contrary, concentration of the tax system. A tax
system that relies on very few sources of revenue runs the risks of higher volatility in
revenue receipts from year to year and also faces greater challenges in forecasting
revenue performance. The risks of relying on too few tax sources has been most
obviously demonstrated in countries that rely on a single commodity for most of their
revenues. A case in point is Nigeria’s, which relies on the petroleum sector for 80% of
public revenues that fluctuate wildly from year to year. Relying on too many taxes
represents an administrative burden and creates taxpayer compliance concerns. In short,
if the number is too low, there is too much risk for the fiscal system. If it is too high,
there is too much dispersion, management difficulty, and annoyance to taxpayers.
16
16
Mann, Arthur and Robert Burke (2002) El Gasto Tributario en Guatemala, DevTech Systems, Inc. for
USAID/Guatemala and DAI. See:
http://www.fiscalreform.net/library/pdfs/el_gasto_tributario_en_guatemala--revision_marzo_2002.pdf
.
Also, the Guatemalan Tax Administration now publishes its calculations of tax expenditures on its website.
See: http://www.sat.gob.gt/estadisticas/pdf/EstudioExenciones2002-2004.pdf
.
ASSESSING TAX SYSTEMS USING A BENCHMARKING METHODOLOGY -- Gallagher
Fiscal Reform in Support of Trade Liberalization – Research Paper
7
Broad tax base with limited exemptions: Obviously, the international standard is that the
tax base should be as broad as possible. This means that tax systems should incorporate
all sectors of their economy and have as few exemptions as possible. We found that in
Central America, in general, and Guatemala, specifically, that the tax base is narrow and
that there are many exemptions. In a later work, “tax expenditures”, i.e., exemptions,
exonerations, deductions, and other special privileges, in Guatemala were equal in value,
in terms of foregone revenues, to the entire total of tax receipts.
Percentage of total taxpayers that provide 75% of tax receipts. In general, throughout
the world, because of disparities in company size, exemptions and exonerations, and poor
administration and inability to reach certain sectors, such as agriculture and informal city
merchants, only a limited percentage of the taxpaying companies pay the majority of a
country’s taxes, be they corporate income tax, sales or VAT, or excises and customs
duties. The more limited this number of major contributors, the more vulnerable the tax
system is to economic change. More important, however, is that the more limited this
number is, the less deeply inserted into the fabric of society is the tax system.
Limited number of tax rates. Consistent with Tanzi’s recommendation that tax systems
not be overcomplicated, it is important that the structure of tax rates for any given tax not
be complicated. Too many tax rates make the system confusing, encourage tax
avoidance schemes, may provide perverse incentives to economic activity, and make tax
administration complicated. The international best practice is to have very limited
number of company and personal income tax rates, as well as to have single or near
single sales or VAT rates. The income tax need not necessarily be “flat” but a relatively
simple table of income tax rates does not detract from the ability to impose a somewhat
progressive income tax system. The import duty schedule would be best if it were
uncomplicated. Indeed, all international trade texts demonstrate that if the import tariff is
to be used for revenue generation, then it should focus on limiting the damage it does to a
country’s international competitiveness and limit protectionism by having as minimal a
divergence between input commodities, capital goods imports, and the duty rates on final
goods.
In Central America and Guatemala, the tax systems are not overly complicated by
complex rate structures. The reforms of the income tax and the sales taxes in Egypt over
the recent decade have greatly simplified tax structures, but there remains much to be
done.
VAT rate. The international benchmark is for a single rate VAT at about 16%. The
benchmark for Central America is 13%, and the actual nominal rate in Guatemala is 12%.
No particular rate should be considered “correct”. More important, however, is whether
there is a single rate. For instance, the best practice is for all goods and services to be
taxed at the same rate, except for exports, which should be zero-rated.
17
Certain items
17
Zero-rating means that when a product is exported, the exporter is refunded all of the VAT paid on prior
stages of production are refunded back to the exporter. This differs from exempted products, where the
ASSESSING TAX SYSTEMS USING A BENCHMARKING METHODOLOGY -- Gallagher
Fiscal Reform in Support of Trade Liberalization – Research Paper
8
may be exempted from VAT, but these should be relatively few. All of the Central
American countries have only one general VAT rate. In Egypt, the General Sales Tax – a
“near VAT” – has a variety of rates, ranging from 5% to 20% of value.
Indirect as percentage of total taxes. There is no right or correct ratio of indirect to total
tax revenues, however, richer countries tend to collect a greater proportion of their tax
revenues from direct taxes, while poorer countries collect a greater share from indirect
taxes. In the U.S., which has no VAT or sales tax at the federal level, direct taxes on
companies and individuals generate the overwhelming proportion of tax revenues at the
federal level, while the property tax (real estate and personal property) is the largest local
revenue generators for many localities. On the other hand, in Europe, the VAT generates
a relatively large share of central government revenues.
VAT collections as percent of total tax collections. Given the importance of the VAT in
most tax systems around the world, this indicator is obvious. The international norm is
for the VAT, where it exists, to generate about 35% of all tax revenues. In Central
America, the average VAT share comes to 45%, and in Guatemala it was 44%, as of
2001.
Tax ratio. This indicator predates any efforts at establishing the benchmarking system. It
is the ratio of actual tax collections to Gross Domestic Product (GDP). Generally, the
higher is per capita income in a country, the higher is the expected tax ratio.
Indeed, the tax ratio for high income countries is about 40%, according to Government
Finance Statistics of the International Monetary Fund. For middle income countries it is
around 25% of GDP and about 18% in low income countries. In Latin America, the tax
ratio tends to be low, given income levels, especially compared to Africa, where it tends
to be high given the low incomes of these countries. In oil producing countries, the tax
ratio is lower than in countries with similar income levels; this is due to the high reliance
on royalties and other revenues related to oil exploitation. For instance, 80% of Federal
Government revenues in Nigeria derive directly from the oil sector. In Egypt, about half
of government revenues come from oil royalties and receipts from the Suez Canal.
Administrative cost of taxation. This is a rather gross indicator of efficiency and it covers
up a number of differences in tax administration, economic structure, and overall societal
modernity. Yet, it also goes directly to the heart of the matter. How much does it cost, in
administrative terms, for a government to impose taxation on its people? Internationally,
the administrative costs vary widely, with the richest countries generally having the
lowest costs with respect to how much they collect, whereas the poorest countries have
the highest costs. However, there is even considerable variance among countries of
similar development levels. That said, the numbers are very interesting and lead to other
questions that need to be answered in any tax system assessment.
final seller of the exempted product is not required to collect the VAT on the final sale, but is not refunded
back the VAT that had been paid on all prior stages of production.
ASSESSING TAX SYSTEMS USING A BENCHMARKING METHODOLOGY -- Gallagher
Fiscal Reform in Support of Trade Liberalization – Research Paper
9
Table 1 presents a cross section of countries with their administrative costs per $100 in
collections.
Table 1: Administrative Costs of Taxation at National Level (mid-1990s)
Country
Administrative costs per $100 in tax collections
Nicaragua
3.86
Guatemala (1995)
Guatemala (2001)*
3.16
2.25
Peru*
3.00
Tanzania*
3.00
El Salvador
2.19
United Kingdom
1.47
Canada
1.18
Spain
0.90
Colombia
0.87
USA
0.83
Sources: Piza (1994) (original from The State Tax Administration Agency of Spain), Gallagher (1995).
18
2001 data for Guatemala are from the Superintendencia de la Administracion Tributaria, see Mann et al.
(2001).
* These tax administrations are Semi-Autonomous Revenue Authorities and their budgets are a fixed, 3%,
of the revenues collected.
Gross Compliance Ratio. One measure for assessing the degree to which the VAT is
applied is the Gross Compliance Ratio (GCR). The GCR is merely the actual VAT
collection in ratio, or as percentage, to the total, potential VAT collection. Total
conceivable VAT collection is that collection that would be achievable in the absence of
evasion or exemptions. This indicator is fairly easy to calculate in countries where only
one VAT rate applies. Our team calculated the international benchmark value of 69%
GCR for advanced countries: indeed, the highest GCRs for Latin America were 64% for
both Costa Rica and Chile.
VAT Productivity. A very commonly used indicator of how well a VAT is applied in a
country is the VAT Productivity rate. This is merely the ratio of VAT collections to GDP
divided by the nominal VAT rate. This is easier to calculate than the GCR, since there is
no need to collect data on aggregate private consumption in an economy. On the other
hand, however, private consumption information is usually readily available from official
sources. In addition, the GCR is a superior measure because it does adjust for this very
important difference in structure of national economies.
18
Gallagher, Mark (1995) Opciones para la Política Fiscal y la Reforma Administrativa de Guatemala,
USAID/Guatemala.
ASSESSING TAX SYSTEMS USING A BENCHMARKING METHODOLOGY -- Gallagher
Fiscal Reform in Support of Trade Liberalization – Research Paper
10
Table 6 presents GCRs and VAT Productivity Rates calculated for a set of countries,
mostly based on data from the mid-1990s.
Table 2: VAT Gross Compliance Ratio (GCR)
Country Private
Consumption
% GDP
VAT
collection
% GDP
Potential
collections
% GDP
Nominal
VAT rate
GCR Productivity
Nicaragua
90-80
5.4
13
15
42 .36
Brazil
81
6.5
13.8
17
47 .38
Madagascar
81
3.2
12.2
15
26 .21
Guatemala
(93)
85
2.7
6.0
7
45 .39
Guatemala
(94)
81
1.6
5.7
7
28 .23
Guatemala
(2000)
-- -- -- 12 49
--
Costa Rica
58
3.7
5.8
10
64 .37
Ecuador
63
1.3
3.8
6
34 .22
Honduras
72
1.5
3.6
5
42 .30
Haiti
98
1.1
9.8
10
11 .11
Dominican
Republic
74
0.9
4.4
6
20 .15
Panama
56
1.9
2.8
5
68 .38
Bolivia
73
0.3
3.7
5
8 .06
El Salvador
(94)
81
4.5
8.1
10
45 .45
El Salvador
(95)
81
5.0
8.1
10
63 .50
Chile
70
9
12.6
18
68 .50
Source: Data are taken from various issues of Government Finance Statistics Yearbook.
International Monetary Fund and from International Financial Statistics, International Monetary
Fund.
Guatemala’s GCR of 49% in 2001 shows a marked improvement from 1994, when it had
only reached about 28%.
In El Salvador, the GCR rose significantly from 45% in 1994 to 63% in 1995. This
increase resulted from strenuous tax efforts, including the incarceration of tax evaders
ASSESSING TAX SYSTEMS USING A BENCHMARKING METHODOLOGY -- Gallagher
Fiscal Reform in Support of Trade Liberalization – Research Paper
11
and shutting down of businesses for non-compliance. On the other hand, the
deterioration in Guatemala from 1993 to 1994 resulted from a demoralized tax
administration, lax enforcement, climbing corruption, and weak leadership. See
Gallagher (1995) for further discussion. The improvements to 2001 result from a
thoroughly reformed tax administration, with a professionalized staff, and the proper
personnel incentives and systems in place.
The GCR is only related to private consumption rather than total consumption. This
neglects government consumption, which is usually nominally subject to the VAT. The
rationale for this is that government consumption in most developing countries is for the
most part expenditure on wages and salaries of government employees, which are not
subject to the tax.
There is a clear relationship between the GCR and the VAT Productivity measure as can
be easily seen from the following figure.
GCR and Productivity: VAT
0
10
20
30
40
50
60
0 20406080
Productivity
GCR
Figure 1
VAT Efficiency Rate. The GCR differs from the more commonly used VAT Efficiency
Rate.
19
The GCR relates the VAT rate to private sector consumption in the economy,
whereas the Efficiency Rate relates the VAT rate to VAT collections in the entire
economy. Since VAT is usually on consumption, it seems a better fit to relate VAT
collections to consumption rather than the entire economy. For instance, VAT is
generally zero-rated on exports and creditable against investments.
VAT Evasion Rate. Evasion is similar to the calculation of the GCR, but it makes an
attempt to net out of the taxable base those goods and services that are legally exempt.
Hence, the Evasion Rate calculation is superior, in concept, to the GCR since it is an
attempt to measure only evasion, whereas the GCR bundles evasion with exemptions.
19
See Bird and Banta (1999) for a discussion of the VAT Efficiency Rate.
ASSESSING TAX SYSTEMS USING A BENCHMARKING METHODOLOGY -- Gallagher
Fiscal Reform in Support of Trade Liberalization – Research Paper
12
Enterprise Income Tax Productivity. The productivity of the enterprise income tax is
calculated in a way that is quite similar to the calculation of VAT productivity, that is, the
ratio of the revenue yield to GDP is divided by the top enterprise income tax rate. As
Stotsky and WoldeMariam (2002), although most countries apply only a single enterprise
income tax, when there are multiple rates, the upper rate should be used as this is the rate
that is applied to most companies, at least when weighted by collections.
20
Table 3: Enterprise Income Tax Productivity
Maximum enterprise
income tax rate
Revenues as
% of GDP
Productivity Productivity
(about 2001)
Argentina 35 1.2 0.034 0.043
Bolivia 25 1.1 0.044 0.059
Brazil 15 1.5 0.100 0.060
Colombia 35 3.9 0.111 0.114
Costa Rica 30 0.5 0.017
Dominican
Republic
25 1.0 0.040 0.044
El Salvador 25 1.9 0.076 0.078
Guatemala 31 1.1 0.035 0.037
Panama 30 2.0 0.067 0.049
Peru 30 2.2 0.073 0.061
Uruguay 30 2.0 0.067 0.076
Sources: Data extracted from Stotsky and WoldeMariam (2002).
Most base data are from late 1990s, latest productivity data are from up to 2001.
Personal Income Tax Productivity Measure. The personal income tax is an important
aspect of most tax systems throughout the world, but in most developing countries it is
usually a very minor revenue source. This stems from a variety of factors, including: low
compliance, high degree of exemptions and exonerations, and usually large, unreachable
sectors of the economy, such as the urban informal sector and smallholder agriculture.
The nature of the personal income tax is different from place to place. The treatment of
capital gains, corporate dividend income, and the levels and types of tax deductions and
credits that are available differ from region to region and country to country.
Despite these difficulties, the personal income tax productivity measure is an attempt to
compare the revenue productivity of the person income tax across a number of countries.
The following table presents the most basic structural aspects of the personal income tax
in a number of countries in Latin America. The factors include: the zero-income tax
bracket, which is the level of income that is not subject to the income tax; the top income
20
Janet Stotsky and Asegedech WoldeMariam (2002) Central American Tax Reforms: Trends and
Possibilities, Working Paper, International Monetary Fund.
ASSESSING TAX SYSTEMS USING A BENCHMARKING METHODOLOGY -- Gallagher
Fiscal Reform in Support of Trade Liberalization – Research Paper
13
tax bracket, which is the level of annual income where the top marginal tax rate takes
effect; and, the maximum and minimum marginal income tax rates. These income
brackets are expressed in multiples of per capita GDP.
The Personal Income Tax Productivity Measure is similar to the Enterprise Tax
Productivity Measure. It is calculated by dividing the personal income tax revenue as
percent of GDP by the top marginal tax rate, expressed in integers, and multiplied by the
top income tax bracket value.
Table 4: Personal Income Tax Productivity
Zero-
bracket
Top-
bracket
Marginal tax rates Revenue as
% of GDP
Productivity
multiple of pc GDP lowest highest
A B C D E F
Argentina 1.4 16.5 35 35 0.6 0.28
Bolivia 0 0 13 13 1.9 0.15
Brazil 1.5 3.1 15 27.5 0.3 0.03
Colombia 4.1 16.6 35 35 0.2 0.09
Costa Rica 0.8 3.7 10 25 1.7 0.25
Dominican
Republic
2.3 5.8 15 25 1.5 0.35
El Salvador 1.2 11 10 30 1.3 0.48
Guatemala 5 22.5 15 31 0.2 0.15
Nicaragua 7.7 61.2 10 25 0.3 0.73
Panama 0.9 57.8 2 30 0.3 0.58
Peru 2.9 22.3 15 20 1.2 1.34
Sources: Central American Tax Reform: Trends and Possibilities, Janet Stotsky and A.
WoldeMariam, FMI, 2002, as well as other IMF and Ministry publications.
This table is telling in a number of ways. According to this measure, the Peruvian
personal income tax system is the most productive of all the represented countries. The
yield of 1.2% of GDP is the second highest among the sample countries. The Bolivian
tax has the highest revenue yield at 1.9% of GDP and the lowest maximum rate, which is
also its highest rate that is applied to all earned income.
A similar comparative table can be constructed for countries in other regions.
Organization
Functional organization. Modernized tax administrations have found that organization
around functions rather than according to taxes or types of taxes leads to greater
integration of operations, better management of staff, and improved compliance and
enforcement. For instance, it is generally considered suboptimal to have the tax
administration organized into the VAT, Income, and Excise Departments, and better to
have them organized into Audit, Services, Archiving, and Enforcement Departments.
ASSESSING TAX SYSTEMS USING A BENCHMARKING METHODOLOGY -- Gallagher
Fiscal Reform in Support of Trade Liberalization – Research Paper
14
Table 5: Functional organization of tax administration
Country Functional organization
Bosnia (Federation) Yes
Bosnia (Republica Srpska) Yes
El Salvador Yes
Guatemala Yes
Hungary* Yes
Kazakhstan Yes
Kyrgyzstan* Yes
Latvia* Yes
Lithuania* Yes
Moldova* Yes
Nicaragua Yes
Belarus* No
Bulgaria* No
Croatia* No
Czech Republic* No
Egypt No
Estonia* No
Georgia* No
Macedonia* No
Romania* No
Turkmenistan* No
* These observations are from Bird and Banta (1999).
Autonomy. A trend among tax administrations in developing countries is the
establishment of the tax department into a semi-autonomous revenue authority (SARA).
SARAs generally are able to provide better pay and other incentives to their staff while
also imposing greater accountability for performance. They generally are outside of the
normal institutional setup of government and may have their own budgetary authorities,
perhaps linked to performance. The jury is still out, though many practitioners have been
recommending SARAs as one step toward institutional reform.
The following countries have established SARAs.
ASSESSING TAX SYSTEMS USING A BENCHMARKING METHODOLOGY -- Gallagher
Fiscal Reform in Support of Trade Liberalization – Research Paper
15
Table 6: Countries that have established Semi-Autonomous Revenue Authorities
Country Date of establishment
Argentina 1988
Bolivia 1987, defunct in 1988, reestablished in 2000/01
Bulgaria --
Colombia 1991
Ecuador 1997-99
Ethiopia --
Ghana 1996
Guatemala 1998/99
Guyana 2000
Jamaica 1981
Kenya 1995
Lesotho 2001-03
Malawi 1995-2000
Malaysia 1994
Mexico 1997
Peru – national 1988—91
Peru – municipality of Lima 1996/97
Rwanda 1998
Sierra Leone 2002
Singapore 1992
South Africa 1996/97
Tanzania 1995/96
Uganda 1991
Venezuela 1993
Zambia 1993/94
Zimbabwe 2000
Data from ongoing research being conducted by Arthur Mann, DAI/Fiscal Reform Project.
www.fiscalreform.net
Customs and tax integration. A number of countries, such as Guatemala and Ecuador,
have integrated their tax and customs departments into a single agency. Peru is in the
process of integrating domestic tax administration with customs.
21
This has a number of
advantages; primary among them are the better flow of information and the ease of
conducting more integrated and integral audits.
21
Canada had integrated customs and domestic tax administration during the 1990s, but has just recently
separated these functions into two separate agencies. This makes sense for Canada, since customs is more
a border security issue than it is a revenue agency.
ASSESSING TAX SYSTEMS USING A BENCHMARKING METHODOLOGY -- Gallagher
Fiscal Reform in Support of Trade Liberalization – Research Paper
16
In developing countries the VAT, which is a tax on domestic consumption, is collected as
much in customs as it is in the domestic tax administration system. Indeed, in many
countries half of all gross VAT collections occur in customs. The integration of domestic
tax agencies and the customs operations makes more and more sense as countries move
toward ever greater reliance on VAT as a revenue source.
Fiscal projections unit. An important function in a Ministry of Finance and in the Tax
Administration organization is the regularized revenue and receipts projections function.
This is a core function of the fiscal management system. Receipts should be projected by
tax and by region, department or organization, as may be appropriate.
Governments should project annual revenues in order to adequately manage their overall
system of public finance, as well as to manage the macroeconomy. Receipts projections
on monthly or quarterly basis are required for setting revenue targets and tracking
progress in meeting these targets.
Tax fraud unit. Tax administrations should have an office specifically set up to prosecute
cases of tax fraud. This requires special skills beyond those of the rest of the tax
administration staff, including knowledge of the tax fraud legislation, knowledge of the
courts and appeals systems, and law enforcement expertise and ability to liaise with other
governmental offices, such as the Internal Ministry.
Large Taxpayer Unit. There is some difference of opinion about whether tax
administrations should separate out the largest taxpayers from the rest of the taxpayer
public. The argument being that this then allows the tax administrators to ignore or not
dedicate enough resources and effort to the non-large taxpayers. Nonetheless, most
advisors, the IMF, and our Team all feel that it is a good practice to provide special
attention, especially with regard to audit and enforcement to the largest taxpayers.
Silvani and Baer find:
In most countries which have established Large Taxpayer Units the
compliance of this taxpayer group has improved. In Uruguay, Bolivia
and Sri Lanka, for example, where the large taxpayers represent a high
percentage of total tax collection, the percentage of stopfilers among
the approximately 1,000 largest taxpayers dropped from 1987 to 1991.
At the same time, in many countries payments from this group
increased significantly (about 20 percent in real terms) after they
began to be monitored by a LTU.
22
The following table shows countries with and without large taxpayer units.
22
Silvani, C. and K. Baer (1997) “Designing a Tax Administration Reform Strategy: Experiences and
Guidelines” WP/97/30, International Monetary Fund, March.
ASSESSING TAX SYSTEMS USING A BENCHMARKING METHODOLOGY -- Gallagher
Fiscal Reform in Support of Trade Liberalization – Research Paper
17
Table 7: Tax Administrations with Large Taxpayer Units
Country Exists
Guatemala Yes
El Salvador Yes
Nicaragua Yes
Egypt (as of 2003) Yes (experimental)
Bosnia (Republica Srpska) Yes
Bosnia (Federation) Yes
Hungary* Yes
Latvia* Yes
Lithuania* No
Slovenia* Yes
Bulgaria* Yes
Armenia* Yes
Azerbaijan* Yes
Georgia* Yes
Kazakhstan* Yes
Kyrgyzstan* Yes
Moldova* Yes
Russia* No
Turkmenistan* No
Ukraine* No
* These observations are from Bird and Banta (1999).
The international benchmark is that there should be a large taxpayer unit. Guatemala,
Nicaragua, El Salvador, Tanzania, and most recently, Egypt, have all established special
units that cater to the needs and undertake audit of their largest taxpayers.
Legal framework
Tax administration code. The ideal tax code is a single, comprehensive piece of
legislation that defines all the legal rights, requirements, and recourses for taxpayers and
the tax administration, alike. The tax code defines all terms that are to be used in the tax
system, and establishes overall procedures, such as filing and retention of information,
organizational setup of the tax administration, establishment and roles of various
organizations, such as the appeals tribunal. The tax code in many countries either exists
as a large variety of laws, often contradicting each other, with roles, rights, and
responsibilities of taxpayers and tax authorities often not clearly established.
ASSESSING TAX SYSTEMS USING A BENCHMARKING METHODOLOGY -- Gallagher
Fiscal Reform in Support of Trade Liberalization – Research Paper
18
There have been many efforts underway over the past decade and a half to define ideal or
model tax codes. For instance, the IMF has developed a model tax code that it calls
“taxastan.” In addition, the Inter-American Center for Tax Administrations (CIAT) in
Panama has produced a model tax code.
Existence of tax fraud felony. Tax administrators in many countries are stymied in their
enforcement activities by the lack of a tax fraud law. Such a law should impose
appropriate sanctions for fraudulent declarations and preparation of purposely false
documentation. While a tax fraud law is not a panacea, tax enforcement without such a
law is very difficult and it will not lead to voluntary compliance, the cornerstone of
modern taxation.
Application of the tax felony authorities. The idea of voluntary compliance is that
taxpayers will comply with tax laws for a number of reasons, one of which is their desire
to not have tax fraud punishments applied to them. This means that for voluntary
compliance to be effective, tax authorities must from time to time impose the criminal
sanctions that are in tax fraud felony legislation. Too many such applications of this law
mean that something is wrong, probably that the tax fraud laws are being used for reasons
other than to encourage compliance, such as for prosecution of political enemies.
On the other hand, too little application of the tax fraud law, especially in light of open
fraud, means that the tax authorities have little power and the law is without teeth. It is
very difficult to encourage voluntary compliance under these circumstances.
In developed countries, the tax fraud felony is only applied sporadically and often its
application is given high profile in the news media. In many countries in the developing
world, tax fraud has only recently been made a criminal offence, still application is very
weak, limited, and in some cases, these sanctions have never been applied despite having
the law on the books.
Two Central American cases are illustrative. In El Salvador, which enacted its tax fraud
felony legislation about ten years ago, only a handful of cases have been brought to
courts for enforcement. Usually, however, the threat of imposing the weight of the tax
fraud law is enough to force corrections to forms and declarations and renewed
compliance from fraud committing tax filers. On the other hand, since its enactment in
the mid-1990s in Guatemala, the tax fraud law has not once been applied.
The Government of Egypt has recently submitted a revamped sanctions law for
enactment by its Parliament. This sanctions law includes large and escalating financial
penalties for tax fraud, but its structure indicates that the application of jail time for tax
fraud, although contemplated, will never transpire.
Appeals tribunal. Most countries of the world have an appeals process where taxpayers
are able to dispute the decisions of tax authorities. This is an important institutional
arrangement that helps to ensure the protection of taxpayer rights, lends credibility to the
overall tax system, and helps to keep tax authorities under external review.
ASSESSING TAX SYSTEMS USING A BENCHMARKING METHODOLOGY -- Gallagher
Fiscal Reform in Support of Trade Liberalization – Research Paper
19
Enforcement
This section discusses a number of indicators related to enforcement of the tax laws.
Current account. A current account is an accounting of all the taxes that a taxpayer is
responsible to file and pay. The current account shows on an always current basis the
taxes that a person owes and the monies that may be owed the taxpayer, say as a rebate
on VAT payments. The current account is an easily conceived instrument, but its
implementation and maintenance are not easy to set up. A current account requires an
automated payments or receipts system, an up-to-date taxpayer registration system, and
careful monitoring of taxpayer liabilities.
Automated audit selection. Audit selection, whether for income tax, VAT or customs
duties should be based on unbiased risk assessment based on statistically determined
parameters. Such a system will help to select for audit those firms and individuals that
are more likely concealing information and therefore under-declaring their tax obligation.
An automated selection process also is useful for reducing the discretion of tax
authorities and should be an integral part of any tax modernization program, and
especially any program that seeks to reduce politicization of the governmental institutions
and to reduce corruption.
Auditors as percent of administrative staff. Obviously, the entire professional staff of the
tax administration cannot be dedicated to audits. Staff are needed for recording and
information technology, legal analysis, and other administrative processes. However, an
appropriate portion of the staff should be dedicated to audit. The international
benchmark indicates that about 30% of professional, administrative staff of the tax
administration should be dedicated to audit.
Use of external data. For tax enforcement to function efficiently, tax authorities need to
have access to information on taxpayers about such things as their ownership of other
companies, vehicle registrations, and land or real estate holdings. This type of
information provides an important source for determining when a tax declaration seems
inconsistent with other assets. The international benchmark is that this type of
information is usually available to tax authorities on-line.
“Stop filers” as percent of active filers. In many developing countries, seemingly
unexplainable drops in fiscal revenues have resulted due solely to the fact that taxpayers
have found that they can simply stop filing their monthly VAT declarations with
impunity. Tax administrations should have automated notification systems that
immediately remind taxpayers of their responsibility to file and pay their tax obligations.
Keeping a lid on stop filers cannot be overemphasized.
Crossing information among taxes. Many tax administrations in developing countries
miss the opportunity to improve their understanding of their taxpayers by simply not
having immediate and useful crossing of information among the different types of taxes
that their taxpayers are paying and declaring. Such information might include: VAT paid
ASSESSING TAX SYSTEMS USING A BENCHMARKING METHODOLOGY -- Gallagher
Fiscal Reform in Support of Trade Liberalization – Research Paper
20
in customs and VAT paid on domestic transactions; real estate taxes and income taxes;
import duties and VAT.
Number of tax administrators per 1,000 of the national population. This is merely an
indicator of the size of the tax administration and its extension into the overall
population. The international benchmark, or norm, is about one tax administrator per
1,000 population. In Central America the benchmark is only 0.27 tax administrators per
1,000 population. In Guatemala it is even lower, at only 0.17
Table 6 presents this indicator for a number of countries. These data, for the most part,
date to the mid-1990s, although the data for Guatemala are from 2001.
Table 8: Public employees in the national tax administration institution
County
Number of tax staff per 1,000 of national population
United Kingdom 2.36
Germany 2.10
Belgium 1.98
France 1.79
Canada 1.49
Egypt 1.00
United States 0.92
Bosnia (Federation) (2004) 0.82
Bosnia (Republica Srpska) (2004) 0.67
Spain 0.66
Uruguay 0.55
Armenia 0.53
Argentina 0.50
El Salvador 0.32
Costa Rica 0.27
Nicaragua 0.22
Colombia 0.19
Chile 0.15
Guatemala (2001) 0.17
Guatemala (1994) 0.15
Bolivia 0.14
Ecuador 0.13
Peru 0.13
Brazil 0.11
Dominican Republic 0.11
Tanzania 0.10
Source: Some of these data are from Piza (1994). Data from Guatemala are from the Ministry of
Finance (1994) and the Superintendencia de la Administracion Tributaria (2001). Data from
Nicaragua are from the Ministry of Finance in 1995. Data for Tanzania were collected from the
Tanzania Revenue Authority in 1998. Data for Bosnia are from records of the respective tax
administrations in 2004.
Active taxpayers per tax administrator. For Central America, in general, the number is
81 per tax administrator, and in Guatemala, it is only 51. In El Salvador, there are about
ASSESSING TAX SYSTEMS USING A BENCHMARKING METHODOLOGY -- Gallagher
Fiscal Reform in Support of Trade Liberalization – Research Paper
21
200 taxpayers per tax and customs official, while in Tanzania the ratio is about 100 to 1.
For Egypt, there are 71 taxpayers per tax official. In the U.S., the number is about 1,000
taxpayers per Internal Revenue Service employee.
The following table presents some observations on the number of taxpayers per tax
official.
Table 9: Active taxpayers per tax official
Country No. of taxpayers per official (approximate)
US 1,000
Armenia* 616
Moldova* 520
Bosnia (Republica Srpska) 462
Ukraine* 327
Hungary* 308
Uzbekistan* 208
El Salvador 200
Tanzania 100
Lithuania* 100
Egypt 71
Guatemala 51
Bosnia (Federation) 50
Kazakhstan* 36
Georgia* 34
Bosnia (Republica Srpska) 33
Latvia* 22
Kyrgyzstan* 20
* These observations are from Bird and Banta (1999).
The ratio of taxpayers to tax administration staff varies very widely. This has to do with
mainly four factors. First, many countries do not actually register all those individuals
who are incorporated in their income tax systems. For instance, if tax is withheld from
wages and the wage earner does not have multiple sources of income, that taxpayer is not
required to prepare a tax declaration and probably need not be included in the tax
registry. In addition, tax systems that rely heavily on VAT and VAT registration is only
required for relatively large firms, then registration will be very low. For example, in the
mid-1990s, there were only 29 firms registered in the Mongolian VAT system.
Second, many tax administrations have made less effort to incorporate tax registrants in
their tax rolls. A case in point, in Bosnia the Republica Srpska has registered all those
individuals with incomes, including pensioners, regardless of their tax liabilities. This
compares with the Federation in Bosnia, where tax authorities have decided not to
register taxpayers who are only subjected to local taxes (these are collected by the tax
departments in both Republica Srpska and the Federation.)
Third, many tax authorities have simply not made adequate effort to extend their presence
into their societies. For instance, comparing Guatemala to El Salvador, which have
ASSESSING TAX SYSTEMS USING A BENCHMARKING METHODOLOGY -- Gallagher
Fiscal Reform in Support of Trade Liberalization – Research Paper
22
similar tax systems, legal systems, and socio-economic structures, it becomes pretty clear
that Guatemalan officials have ample room to increase the number of tax registrants.
Performance Indicators System for Audit and Auditors. The international best practice,
although not very widespread, is to evaluate the tax administration’s audit and its auditors
by developing performance monitoring systems. These can have many different forms
and not one way is necessarily superior to another.
Such performance indicators might include comparisons of ex ante and ex post valuation
of income tax revenues that would be generated by a specific audit case. Systems for
such valuation are based on statistical analyses and good databases. Monitoring audit
cases per auditor or team can be developed to assess overall performance.
The international benchmark is that tax administrations should have such performance
indicator systems for their audit function. Such systems have been worked on in Costa
Rica and El Salvador, but they are not fully developed and cannot be considered the
norm. The Guatemalan tax administration, in 2001, had begun to develop the statistical
systems that could be developed for audit selection and analysis, but this is very much
lagging in its development. Tanzania had not developed such a system. Nor had El
Salvador, as of 1995, although it was in the stages of designing such a system.
Clean taxpayer registry. One of the very first efforts in tax modernization is the cleaning
up of the taxpayer registry. This registry should be a directory of all taxpayers in the
country, along with their addresses, economic activities, and links to other asset
ownership, such as vehicles and bank accounts, and legal residence. The taxpayer
registry is the backbone of all tax administrations.
It should be automated, easy to enter data and to register taxpayers, and easy for the rest
of the tax administration to link to.
Percent of taxpayers subject to annual audit. Too much auditing implies inefficiency and
harassment. Too little calls into question the enforcement efforts of the tax
administration. In developed countries about one percent of taxpayers are subject to
audits in any given year. In Central America the audit rate has reached to about two
percent of the taxpayer population. Audit data were not available to the benchmarking
team for Guatemala. In some countries, such as Bosnia and Egypt today or Tanzania a
few years ago, the attempt at 100% audit is actually not audit at all, but rather official tax
assessment. Such assessment is completely inconsistent with the concepts of “voluntary
compliance” and is not generally recommended.
Bird (1999) shows audit at 45% of taxpayers in Kyrgyzstan, and 50% in Moldova and
Turkmenistan.
Unified audits. An old shibboleth is “do not audit the tax, audit the person.” A unified
audit, for our purposes, combines the audit of companies for their VAT or sales tax,
income tax and taxes on imports. This is an important international trend, though not
widely practiced yet. It is not done in Guatemala or Central America, or Egypt or
Bosnia, or Tanzania, yet.
ASSESSING TAX SYSTEMS USING A BENCHMARKING METHODOLOGY -- Gallagher
Fiscal Reform in Support of Trade Liberalization – Research Paper
23
Receipts and collection
The payment of taxes should be as simple and low cost as possible. Keeping this
compliance cost down is an important way to encourage voluntary compliance. It is also
necessary that the tax authorities have a payment system that will provide accurate data
on a very timely basis and that this information directly feeds into the other information
management systems of the tax authorities, such as the revenue tracking system and the
taxpayer current account.
Bank system payments. An important innovation over the last ten years in many
developing countries has been the move to tax payment through the banking system.
Compared to payment at government offices, banking system payment can be
substantially more efficacious. It is usually more convenient for taxpayers, provides
fewer errors since it is often introduced with effective automation systems, and is often at
lower cost for the tax or treasury administrations.
Percent of large taxpayers paying via the internet. In both developed and developing
countries payment of tax obligations via the internet is a very important step in reducing
compliance costs and improving tax accounting.
Improved tax administration, especially with lower corruption and compliance costs, can
come about by reducing the opportunities for direct, face-to-face contact between
taxpayers and tax administrators. This is one reason why bank or internet payment
systems are to be preferred to payment at government collection offices or cashiers. Of
course, in developed countries, where the banking system functions well, check payment
by mail has long been the preferred means for payment of taxes, although more and more
taxpayers have been opting to file and pay their taxes via the internet in rich countries.
Late payments as percent of total domestic tax receipts. The management of delinquent
taxpayers is a key function of any tax administration. Tax administrations need to make
every effort to keep these delinquents under control. The Guatemala benchmarking team
set a five percent standard, although this is based on the experience only of one of the
team members and no readily available international data are available.
Systems and resources
Internal procedures and manuals. All public entities need to clearly specify in terms of
internal regulations or rules the procedures that are to be followed in carrying out their
functions. Manuals are the handbooks that explain how these procedures are to be
carried out. These manuals should cover basic organizational functions, such as
personnel policies and financial control, as well as document and explain the
requirements for the implementation of all of the tax administration processes, such as
audit, document receipt and handling, archiving, and the other main tax administration
business processes.
While internal procedures and manuals are generally well documented in developed
countries, many developing countries have not documented their internal procedures or
their documentation of these procedures has lagged the changes in these procedures. By
ASSESSING TAX SYSTEMS USING A BENCHMARKING METHODOLOGY -- Gallagher
Fiscal Reform in Support of Trade Liberalization – Research Paper
24
the same token, many developing countries have only inadequate manuals or handbooks,
if they have any at all.
Planning and monitoring systems. Developed country tax departments generally have a
corporate planning department that sets performance targets, monitors the attainment of
these targets, plans for capital programs, and leads the overall direction of the
development of the tax department. In most developing countries the planning and
monitoring systems are very rudimentary and usually not paid much attention.
Coordinate information with ministries and others in government. Of course, the rights
of taxpayers must be respected. However, certain information should be shared that will
not violate these rights. The tax department should have information sharing
arrangements with the central bank, the Ministry of Finance, customs, and local
governments. Information to be shared can include, total revenue receipts by type of tax,
macroeconomic and sectoral economic data, public investment, imports, exports and
international capital flows, data related to criminal behavior, and data related to
international transfer pricing.
General use of automation. All modern tax administrations employ automated systems
for most major processes, such as document receipt and management, issuance of notices,
filing and imaging, taxpayer services-related information sharing, and so on.
Interconnectivity between headquarters and local revenue offices. In many developing
countries information is shared with considerable lag. In some countries information is
not included into modern technological systems and paper trails and physical transfers
and photocopying are the order of the day. In other countries, data transfer may take
place either with the physical delivery of data disks or tapes. Still, in other countries,
taxpayer and tax payment information is transferred, usually on a one-way basis, between
the local office and the headquarters, on a batch basis a few times per day. All of these
methods lead to errors, untimely delivery of information, and may leave information
loopholes that fraudulent taxpayers can exploit.
Data and systems backups. All modern public institutions should ensure that their data
and computer systems are backed up on a daily basis. Such backups are particularly
important for the tax administration as the taxpayer database they manage is the absolute
core of the tax system. The value of this database and its daily backups cannot be
overstated. Such backup systems are routine in industrialized countries, but in many
emerging market countries, the effort and expense for backing up these systems is just
not sufficiently appreciated, leaving these tax administrations at great risk.
The next indicators relate to the quantity and quality of human resources of the tax
administration.
Percentage of employees with university degrees. University graduates tend to have a
much higher representation among tax administration staff in developed compared to
emerging market countries.
ASSESSING TAX SYSTEMS USING A BENCHMARKING METHODOLOGY -- Gallagher
Fiscal Reform in Support of Trade Liberalization – Research Paper
25
Ratio of “director” level salaries to that of tax auditors. Generally, in industrialized
countries the top paid person in the tax administration may have about twice the salary as
does the average tax auditor. This seems to vary considerably among emerging market
countries.
Ratio of average tax administrator’s salary to average GDP per capita. In industrialized
countries the average tax administrator’s salary may be equivalent to about twice the
average per capita income in the country. This is not a particularly high salary, per se,
but it is enough to attract sufficient professional staff to operate the tax administration. In
emerging market countries tax administration staff salaries are several multiples of the
average per capita income, yet these salaries are still too low to attract high caliber
professionals.
Next steps
Bird (1999) lists a number of criticisms of the Bird and Banta (1999) indicators. Some of
these criticisms have some validity and could apply to this set of international
benchmarks as well. In particular, the most appropriate criticism was that these tax
systems are complex and 1) they cannot be represented by simplifying indicator systems,
and 2) they should not be represented in this way. Since Bird and Banta (1999)
specifically were presenting a set of indicators for fiscal sustainability that were meant to
be used in international comparisons, perhaps the critique has some relevance. For this
benchmarking exercise, however, this must be rejected. The point of establishing
benchmarks is to help counsel tax authorities on what needs to be done to improve their
tax administrations and international comparison information is a useful, but not the only,
point of departure.
That said, Bird (1999) puts paid the critique in the following comment.
… life is complex and there is not and cannot be any
perfect way to attempt such a task. All that can be done is
take as comprehensive and consistent approach as possible,
to set out the evidence and methodology as fully and
clearly as possible, and to be as open as possible to
comments, corrections, criticisms, and suggestions for
change.
Page 234 in Shapleigh, Andic, and Banta
Future work on this benchmarking system should include broader and a more
comprehensive application to other countries, comparative case applications, and a
systematic data collection effort.
This system has been developing over a number of years. In each year an indicator or
two has been added, corrected, or removed from the overall system. Application of the
methodology to other countries, as well as follow up applications in the countries that
have already been studied, would help to validate the indicators used, demonstrate their
ASSESSING TAX SYSTEMS USING A BENCHMARKING METHODOLOGY -- Gallagher
Fiscal Reform in Support of Trade Liberalization – Research Paper
26
usefulness as performance or target indicators, and incorporate ever more useful
indicators.
Performing a number of comprehensive benchmarking studies in specific regions and
around the world would be useful to enrich the system and provide better guidance to
individual countries with regard to their own performance enhancement and tax
organization and administration efforts. Regional comparisons are useful since tax
administrators, at least in my experience, seem to want to compare their own systems
with those of neighboring or regional competitors.
A broader set of comparative studies can be helpful since there is much that can be
learned about innovation and reform from outside of a particular region. An example of
this is the so-called trend towards the creation of autonomous revenue agencies, when in
fact, this is only a trend in Africa and Latin America. In the Asia Pacific region, for
example, only New Zealand and Singapore have established autonomous revenue
authorities, while the Philippines government has been studying the case of Peru in the
design of its own autonomous revenue authority. Clearly, lessons can be got from
outside of one’s own backyard.
Finally, in an ideal situation, data for all the benchmarking indicators would be collected
for a large number of countries, regional and taxonomic groupings and averages or norms
could be calculated, and a book produced, perhaps on a periodic basis.
ASSESSING TAX SYSTEMS USING A BENCHMARKING METHODOLOGY -- Gallagher
Fiscal Reform in Support of Trade Liberalization – Research Paper
27
Annex: The Benchmarks as Developed in the Guatemala Study
Indicator International
Benchmark
Central America
Benchmark
Guatemala in 2001
Tax Structure
Number of taxes making up 75% of
collections
6 4 4
Broad tax base with limited
exemptions
Yes No No
Percent of all taxpayers that pay the
top 75% of revenues
5% 1% 1%
Limited number of tax rates Yes Yes Yes
Domestic VAT as percent of VAT on
imports
n.a. 100% 70%
Indirect taxes as % of total tax
revenues
50% 70% 76%
VAT collection as percent of total tax
take
35% 45% 44%
VAT rate 16% 13% 12%
Tax ratio: high income countries 40% n.a. 10%
Tax ratio: middle income countries 25% n.a. 10%
Tax ratio: low income countries 18% 14% 10%
Enforcement
VAT evasion 10% 25% 33%
VAT productivity 0.58 0.39 0.41
VAT Gross Compliance Rate 69% 46% 49%
Use of performance indicators for
audits and auditors
Yes No No
Number of tax administrators per
1000 national population
1.00 to 2.00 0.27 0.17
Ratio of active taxpayers to tax
administrators
150 to 250:1 81:1 51:1
Audited taxpayers as % of total
taxpayers, per year
1% 2% n.a..
Unified domestic and import audits Trend No No
Ex post customs audits Trend No No
Separation of taxpayers by size or
nature
Yes Yes Yes
Payments and Collections
Banking system payments Yes Yes Yes
Percent of large taxpayers declaring
via Internet
100% n.a. In process
“Stop-filers” as % of active taxpayers 5% n.a. n.a.
Late payments as % of total tax
receipts
5% n.a. n.a.
Administrative cost as % of total
receipts
1% 1.5% 3%
Share of adjustments and fines
collected
80% n.a. n.a.
Business days for VAT refunds 25 30 to 90 30 to 90
Institution that establishes revenue
targets
Ministry Ministry Ministry
Automated Systems
Use of automated systems for daily
use
Yes Yes Yes
ASSESSING TAX SYSTEMS USING A BENCHMARKING METHODOLOGY -- Gallagher
Fiscal Reform in Support of Trade Liberalization – Research Paper
28
Indicator International
Benchmark
Central America
Benchmark
Guatemala in 2001
Interconnectivity between HQ ad
local tax offices
Yes Yes Yes
Backup systems for all uses Yes Limited Almost all
Operating taxpayer current account
(also under enforcement)
Yes Yes In preparation
Clean and operating taxpayer registry Yes Yes In preparation
Automated audit case selection Yes Yes In preparation
Tax declaration entry with automatic
error correction
Yes Yes n.a.
Use of exogenous information
(filers>vehicles>real estate
Yes Trend No
Use of third party databases Yes Trend In preparation
Data crossing among taxes Yes Trend In preparation
Late or stop filers system Yes Trend No
Planning and Coordination
Appropriate use of planning,
monitoring, and evaluation systems
for tax organization
Yes Limited No
Coordination of data flows among tax
administration, Ministry, and other
agencies
Yes Trend No
Human Resources
% of employees with university or
college degrees
70% 40% 40%
Ratio between director and auditor
salaries
2:1 4:1 5:1
Ratio between average tax
administrator’s salary and average
GDP per capita
2:1
5:1
5.5:1
Existence of administrative career
plan
Yes Trend In preparation
Existence of formal retirement plan Yes Trend No
Sanctions and Penalties
Tax code Trend Some, new Yes
Tax fraud felony Trend Some, new Yes
Application of tax fraud felony
sanctions
Little Very little No
Appeals tribunal Yes Yes Yes
Organization, Institutional Credibility and Public Confidence
Stability of top-level leadership Fixed appointment Variable Variable
Professionalism of top-level staff Excellent Good Good
Tax fraud unit in tax administration Yes n.a. No
Unit for investigation of internal
corruption
Yes n.a. No
Diversity and quality of taxpayer
services
Yes Limited Limited but
improving
Internal regulation Yes Trend n.a.
... A broad debate has arisen on the design of the optimal organizational structure for tax administrations. Different alternatives have been tested, among which can be highlighted a structure "by types of taxes," a "functional structure," or a structure "by types of taxpayers" (Gallagher, 2004;Jacobs, 2013b;Junquera et al., 2017). ...
... A research period of 10 years was tested due to the fact that the last scholarly articles found, which researched methodologies etc., specifically in the field of taxation, were by Shevlin (1999) and Gallagher (2004). ...
Conference Paper
Full-text available
~ Research in the fields of tax law and taxation in general, is mostly dependent on a qualitative approach in terms of data found in the form of primary sources such as court cases and secondary sources such as legislation and other literature sources. Often scholars in the field of tax law and taxation studies tend to be very ambiguous with regard to the research methodology and specific research technique applied by them. Very limited information on tax law related literature was found, which specifically discussed the applicable research methodology applicable to this field of study. The main objectives of this study were: Firstly, to obtain evidence of the most appropriate research designs, methods and techniques that are applied in research projects in the field of tax law and taxation in the past decade. Secondly, to determine and scrutinise the most appropriate research design, techniques and methods applicable to this field of study. Lastly, to highlight to novice researchers, the importance of selecting an appropriate research design, method (s) and technique (s) before a research project is conducted. Unobtrusive techniques such as documentary analysis, thematic and cluster analysis, and content analysis, among others, were considered in order to address the objectives stated above.The methodological approach included a documentary and content analysis and the results were documented by way of a systematic literature review. The methodologies endowed in literature were ranked according to their frequency and conclusions are drawn from the most frequently used technique. In addition, to assist novice researchers, different research techniques were compared, summarised and categorised, by way of considering the applicability of each technique's main objective and function and practical application. The study found that documentary analysis and thematic analysis are the most commonly used techniques in literature reviews in the field of taxation and tax law studies.
... In particular, tax auditing systems tend to be outdated and underfunded. Because smaller firms are particularly costly to monitor, there tends to be a high concentration of tax collection on a small number of large firms (Gallagher, 2004). At the same time, larger firms usually have the political leverage to affect tax policy in ways that benefits them. ...
Article
Full-text available
It is hard for tax administrations to levy and collect taxes anywhere and any time. However, taxing certain kinds of activities, sectors, or individuals - the so-called "informal sector" - is an additional challenge for tax administrations in both developing and developed countries, and the "fiscal gap" that arises from the failure to tax this sector can be quite large. This issue is especially pressing in Latin America and Caribbean (LAC) countries, where often over half of the workforce is found in the informal sector. In this paper we examine taxation and tax compliance in LAC countries - and beyond - focusing on several main questions. What is meant by the "informal sector"? What is the size of informal sector in LAC countries? What are some effects from an informal sector, including the size of the "fiscal gap"? What are the reasons for this fiscal gap? What can be done to address these various issues? Using the size of the informal sector as a proxy for the amount of tax evasion, we find that tax evasion is quite prevalent in LAC countries, a prevalence that is due both to tax administrative problems and to a societal culture that often generates low "tax morale". Measures to combat evasion include administrative improvements consistent with the "punishment paradigm" and also with the "service paradigm" views of compliance, as well as changes in societal institutions that address tax morale.
... Smaller firms are particularly costly to monitor. As a result, there tends to be a high concentration of tax collection on a small number of large firms (Gallagher, 2004). At the same time, larger firms usually have the political leverage to affect tax policy in ways that shift part of the tax burden to the non-rich. ...
Article
Full-text available
We experimentally investigate social effects in a principal-agent setting with incomplete contracts. The strategic interaction scheme is based on the well-known Investment Game (Berg et al., 1995). In our setting four agents (i.e., trustees) and one principal (i.e., trustor) are interacting and the access to choices of peers in the group of trustees is experimentally manipulated. Overall, subjects are positively influenced by peer's choices they observe. However, the positive interaction between choices is not strong enough to raise the reciprocity of those observing at the same level of those whose choices are observed.
... According to a benchmarking study (Gallagher: 2004) tax authorities need to have access to other information on taxpayers, relating to their ownership of vehicles, real estate ownership and their interest in other companies. The international benchmark is that this type of information is usually available to tax authorities on-line. ...
Article
The vision of probably all revenue authorities is to promote compliance with the provisions of the taxation laws and to ensure responsible enforcement by the revenue authorities, thereby contributing to the economic well being of the country. As with virtually all revenue authorities the South African Revenue Services has to a large extent implemented the self-assessment approach to tax assessments. Because this system depends on this process of selfassessment, an effective risk-based audit approach is required to ensure that tax compliance and responsible enforcement is adhered to. An effective case selection methodology is required for revenue authorities to make informed choices on how best to direct their activities in order to address areas of greatest risk. Given these imperatives, the purpose of this study is to examine the case selection methodologies used by certain revenue authorities, including the South African Revenue Services, and to focus on the key elements of case selection: the use of computerised database systems, industry profiles, third party data and the role of the risk profiler. The results of the study indicate that the case selection methodology of the South African Revenue Services is lacking in some areas. Computerised risk analysis is limited to a certain classes of taxpayers and other aspects of concern are also highlighted in this study.
Article
Globalization is eroding the efficiency of conventional taxes, such as VAT´s (value added taxes). Meanwhile, a new form of taxation, levied on bank transactions, was successfully used in Brazil (1993-2007). Based on digital technology, this tax innovation proved to be evasion-proof, more efficient and less costly than orthodox tax models. Furthermore, the significant revenue-raising capacity of bank transactions taxation revived the centuries old ideal of the Single Tax. This book carries out a qualitative and quantitative in-depth comparison of the efficiency, equity and compliance costs of a bank transactions tax relative to orthodox tax systems, and opens new perspectives for the use of modern banking technology in tax reform across the world.
Book
Value-added tax (VAT) now dominates tax systems around the world. But should every country have a VAT? Is the current VAT always as good as it could be in economic, equity and administrative terms? In developing and transitional countries the answers to such questions are critical to stability, growth and development. VAT is a critical fiscal tool in most countries. But VAT can sometimes be better designed and almost always better administered. The key questions that must be answered in designing and implementing VAT are essentially the same in all countries. But different tax designs may best suit different countries facing different circumstances. This book reviews experiences with VATs around the world and assesses how the choice of particular design features may affect outcomes in particular contexts.
Article
Full-text available
Pakistan’s tax system has undergone significant reforms over the last two decades, leading to the modernization of direct and indirect taxes. More recent times have seen the rationalization of income tax rates, the introduction of self-assessment for filing income taxes, some expansion of consumption taxes, and the rationalization of the customs tariff structure with a reduction of tariff bands and maximum rates. Currently, the Central Board of Revenue (CBR) is engaged in a comprehensive plan to re-structure and modernize the entire tax administration and customs operations. In addition, the CBR has taken a number of steps in the recent past to increase the number of taxpayers and broaden tax bases. From a macroeconomic perspective, fiscal performance has improved as measured by the reduction in the federal budget deficit and the overall level of debt in terms of GDP.
ResearchGate has not been able to resolve any references for this publication.