Download Taxation Around the World in Profile

1. Introduction
Introduction
The United Nations Millennium Declaration, adopted by the General
Assembly at its fifty-fifth session, notes that the globalization of the world
economy “offers great opportunities” but also states that “the central challenge
that we face today is to ensure that globalization becomes a positive force for
all the world’s people.” The declaration recognizes that “developing countries
and countries with economies in transition face special difficulties in
responding to this central challenge.” One important challenge for developing
and transitional countries is the creation of effective tax systems that will
allow them to mobilize their domestic resources for development purposes.
The forces of globalization create new difficulties for these counties to
overcome as resources are generated.
The Millennium goals of sustainable development, broad-based
economic growth, and poverty eradication will not be achieved easily or
evenly. As the Declaration notes, a concerted effort must be made to
formulate both national and global policies in order to ensure that all countries
and all people gain from the emerging world economic structure. Concerted
efforts on many fronts will be required to improve the economic conditions in
developing and transitional countries. To achieve broad-based growth, these
countries almost certainly will require the continuation and expansion of
international flows of private capital. In some cases, the cancellation or
reduction of official bilateral debts will be a practical necessity. Perhaps most
importantly, these countries must join with the developed countries in
promoting a higher degree of international cooperation on issues of
importance to development, including cooperation on social, financial, fiscal,
and tax policies.
No matter how successful efforts may be to increase the international
flow of capital and to improve its allocation, the adequate mobilization and
1
efficient and equitable use of domestic resources remains critical to the
achievement of the Millennium goals in developing and transitional countries.
As the Secretary-General noted in a recent report (“Road map towards the
implementation of the United Nations Millennium Declaration”), “the
mobilization of domestic resources is the foundation for self-sustaining
development” because such resources in even the poorest countries almost
invariably “play the main role in financing domestic investment and social
programmes, which are essential for economic growth and making permanent
gains in eradicating poverty.”
As the Group of Experts on Public Administration and Finance
recently noted, “effective democratic governance and efficient public
administration are, arguably, among the most important elements in the
promotion of a country’s national development agenda.” The SecretaryGeneral’s road-map report specifically singles out “sound fiscal policy” as one
of “the elements of good governance that are crucial to economic and social
development. The report goes on to note that one of the “strategies for moving
forward” is “disciplined fiscal policy, including clear goals for the
mobilization of tax revenues.” As was stated at a Side Event on Capacity
Building for Effective Tax Revenue Collection held during the Monterrey
International Conference on Financing for Development (2002), “the
mobilization of domestic resources is the foundation for self-sustaining
development.” That document went on to warn that “there are indications that
globalization may reduce tax revenues due to increased tax competition
among developing countries to attract foreign direct investment, exponential
growth in electronic commerce, increased mobility of factors of production,
and growing importance of off-shores and tax havens.”
The passages quoted above set the stage for the present report, which
focuses on strategies for domestic resource mobilization in developing and
transitional countries in the context of globalization. By way of introduction to
taxation in developing and transitional countries, Chapter 2 of the report
presents a brief quantitative overview of the level and structure of taxation
around the world, with special attention to recent trends. As this chapter
2
shows, the phrase “developing and transitional countries ” encompasses a very
wide variety of countries that differ in level of development, command over
resources, and degree of integration into the world economy. It is thus highly
unlikely that precisely the same tax policy will be suitable for all countries.
Chapter 3 of this report reviews several critical aspects of taxation and
domestic resource mobilization in general terms. Some of that discussion is
directed primarily at those developing and transitional countries that are still
struggling to establish effective governance and have not yet developed the
capacity to implement complex instruments of tax policy. Other parts of that
chapter, however, have broad implications for most developing countries.
Chapter 4 considers the implications of globalization for domestic tax policy
in developing and transitional countries. Much of that chapter has more
immediate relevance to those countries that are close to becoming full
members of the global community. Chapter 5 addresses the need for improved
international cooperation in tax matters. It suggests that the Millennium goals
cannot be achieved without substantially greater international cooperation on
tax matters than has seemed possible in the not too distant past. The various
policy recommendation implicit in the report are set forth in chapter 6.
A final cautionary note seems needed. Taxation is never a simple
matter. Many of the issues discussed in this report are both complex and
controversial. No short general discussion can cover all relevant
circumstances and possibilities. The literature on the subject is huge and often
conflicting. Although we have obviously drawn extensively on that literature
in preparing this report, our aim is not to summarize the literature but rather to
attempt to distil from our reading of it, tempered by our own experience, some
lessons for developing and transitional countries. We have, therefore, not cited
references except when specific facts or examples are offered. Similarly, our
aim is not to set out a package of tax policies for any one country but rather to
suggest the kinds of tax policies that all developing and transitional countries,
or at least large subsets of such countries, should consider.
To make our points as concrete as possible, we have occasionally
3
made reference to experiences in particular countries. We caution against
generalizations from these experiences. Every country is different, and these
differences matter with respect to tax policy as in other areas. These
difference, nonetheless, do not eliminate all common ground. All countries
today face certain common pressures and problems in designing and
implementing an effective tax system that arise from the increasingly
globalizing world. Many of the policies appropriate for dealing with those
pressures and problems in one country will be appropriate in other countries
as well. We present this report in the firm belief that some important
prescriptions for good tax design, applied with appropriate caution, transcend
the particular circumstances of particular countries.
4
Chapter 1
Taxation Around the World in Profile
No single tax structure can possibly meet the requirements of every
country in the world. The best system for each country must be determined on
the basis of its economic structure, its capacity to administer taxes, its public
service needs, and many other factors discussed later in this report.
Nonetheless, to set the stage for that discussion, it is useful first to stand back
and take a look at some broad patterns of taxation that exist around the world
and how those patterns seem to have changed in recent years. For this
purpose, a short time series of data on tax revenue was collected for 167
countries, representing every region of the world.1
How High Are Taxes?
The tax ratios – taxes as a share of GDP– reported in data for the 167
countries range from well under 10 percent in a number of countries, most of
which are small and low-income (such as Myanmar, Nepal, Guatemala, Haiti,
Niger, Chad, and Central African Republic), to well over 40 percent in a
number of countries, mostly in western Europe (Netherlands, Denmark, Italy,
France, Sweden). Some lower-income countries, and particularly transitional
countries, also had surprisingly high tax ratios. These include countries such
as Algeria, Democratic Republic of Congo, Sudan, Ukraine, and Belarus.
1
GDP data were obtained from the IMF World Economic Outlook Database at
www.imf.org/external/pubs/ft/weo/2002/01/data/index.htm and the World Bank Group World
Development Indicators 2002 CDROM. Revenue data were obtained from IMF Country
Reports at www.imf.org/external/country/index.htm, the OECD Revenue Statistics CDROM,
1965-2000, 2001 and the World Bank Group Development Indicators 2002 CDROM. There
are many problems in assembling such data. For example, since countries report their data in
many different formats, it is not always possible to extract data for national taxes only: for
123 countries, the data cover the central government only, for 29 countries general
government is covered, and for 16 countries it is not clear which governments are covered. In
the discussion below, the data for each country is that for the most recent year available. In
addition, the length of the time series used to investigate the changes in taxing patterns varies
by country, depending on data availability. Since GDP data were not available for five
countries, these countries are excluded from the examination of trends. The time series used
for “trend” analysis has an average length of only 5.81 years, usually covering the mid-1990s.
5
Similarly, some higher-income countries (such as the United States) had
notably lower tax ratios than others in this group, with Hong Kong being the
extreme case in this respect. Excluding social insurance contributions– the
importance of which varies greatly from country to country depending upon
both the magnitude of the social insurance system and the extent to which it is
financed from designated contributions2 – on average taxes accounted for a
little less than one-fifth of GDP (18.8 percent) for the 167 countries in the
sample.3 The median country in the sample has an even lower tax ratio (17.3
percent). If social insurance contributions are included, the average tax ratio
rises to 23.0 percent of GDP, although the data may not include all social
insurance payments for all countries.
Of course, none of these figures adequately depicts the wide diversity in
tax levels around the world. Taxes as a share of GDP vary significantly, with
much of the divergence appearing to derive from differing levels of income
and geographic locations, though of course many other factors may also affect
both the need or desire for public services and the ability to collect taxes. Both
opportunity and choice affect tax levels. For example, countries with access to
rich natural resource revenue bases, such as United Arab Emirates, Venezuela,
and Azerbaijan, tend to have higher tax ratios than otherwise comparable
countries.4 Particularly in countries with higher income levels, tax levels
seem also to reflect more choice than chance: some such countries (e.g.
Sweden and the Netherlands) have very large and centralized governments,
while others (such as Switzerland and the United States) have much smaller
and more decentralized governments.
To some extent, countries seem to have been influenced by the
behaviour of their neighbours. Although such similarity may also reflect many
2
For example, traditionally some countries such as Australia and New Zealand financed
social insurance from general taxation while others, such as many transitional countries, have
relied heavily on wage taxes imposed as part of a social insurance system.
3
The figures cited are simple averages. As is common in such international comparisons in
which the nation-state is the basic unit of analysis, each country is treated as a single
observation, so that a small island such as St. Lucia receives the same weight as a large
country such as the United States.
4
Of course, such revenues are often more volatile, reflecting changes in basic commodity
prices.
6
other factors, such as a common colonial heritage or common cultural or
economic factors, countries may to some extent follow their neighbours
because of the demonstration effect -- they learn what possibilities exist in
terms of services and taxes -- or because high taxes in one country may
loosen the economic and political constraints on raising taxes in its
neighbours. Taxes in Europe, and particularly northern Europe, are, for
example, much higher than in other regions, while taxes in Asia are noticeably
lower than in other continents (see Figure 1). Interestingly, tax levels are more
diverse in Africa than in any other continent. Of the 47 African countries
included in the sample, more than 25 percent are in the lowest one-third of all
countries in terms of tax levels, with tax ratios of 14.6 or less (Chad, GuineaBissau, etc.) and more than 25 percent also in the top one-third, with tax ratios
of 21.5 percent or more (Algeria, Botswana, etc.).
Figure 1: Tax Revenue as a Percentage of GDP by Continent
23.98
25
20
19.44
18.82
17.5
18.78
16.28
14.67
15
10
5
0
Africa
As ia
E urope
7
N.
Am erica
Oceania
S.
Am erica
Total
Tax ratios also vary with income levels. The 164 countries in the
sample for which GDP data were available were divided into three groups.
Sixty-four countries with per capita GDP less than US$1,000 in 1999 were
classified as “low-income”, 76 countries with per capita GDP between
US$1,000 and US$17,000 were classified as “medium-income,” and the 24
countries with per capita GDP greater than $US17,000 were classified as
“high-income”. As earlier studies (Tanzi, 1987) have shown, the data
evidence a general tendency for taxes as a share of the economy to rise as per
capita incomes rise (Figure 2).
Figure 2: Tax Revenue as a Percentage of GDP by GDP per Capita Category
25
23.17
19.97
18.78
20
15.78
15
10
5
0
L ow
M id
High
Total
Several factors may explain why countries tend to collect relatively more
taxes as their incomes grow. The demand for public services may rise faster
than income (the income elasticity for services is greater than one),
8
particularly in countries where income levels are low. Relatively larger shares
of income, for example, are directed through the public sector to activities
such as education and health care as incomes rise. In addition, as a rule
urbanization tends to increase along with income levels, and the demand for
public services is generally higher in urban areas, which have greater needs
for transportation, water and sewer services, and other governmental services.
Fortunately, at the same time as rising incomes give rise to more demands on
the public sector, the capacity to pay taxes out of income also obviously rises.
Indeed, when countries spend wisely, investing in people and productive
infrastructure, public expenditure itself may further expand taxable capacity.
Moreover, the ability of countries to tap that capacity is also generally
improved. It is, for example, easier as a rule to collect taxes in urbanized
areas. Accounting and other information needed to administer modern taxes
effectively is generally more prevalent and accessible in higher-income
countries. Tax administrations have access to more highly skilled workers and
better technology for collecting taxes.
More detailed analysis confirms the broad picture shown in Figure 2:
on average taxes rise with per capita income levels.5 The relationship
between rising income levels and higher taxes is clearly significant within the
group of lower-income countries. As average incomes rise in poor countries,
the size of the public sector almost invariably expands in relative terms. After
some point, however, this “income determinism” of tax levels declines, and in
the group of higher-income countries, there is no longer any clear relation
between income and tax levels. As noted earlier, the rich, it seems, have more
choices, and some rich countries have chosen to levy much lower taxes than
others. The most important conclusion emerging from statistical analysis of
such data, however, is that there is at best a very weak relationship between
5
A simple regression of per capita taxes on per capita GDP has an elasticity of 1.13,
indicating that taxes grow faster than income, this fits the picture seen when countries are
sorted by income levels (as shown in Figure 1). A seemingly more appropriate quadratic
regression on income shows that taxes tend to rise with income, but at a slower rate as income
rises. Indeed, after income reaches high levels (about $22,000 per capita), taxes actually tend
to fall as a share of GDP. Linear regressions were also estimated separately for low, median,
and high-income countries. The coefficient on per capita GDP was only significant in the
equation for low and middle-income countries.
9
economic development and the level of taxation.6 Even the poorest countries,
while more constrained than the rich, evidently have considerable choice as to
how much they raise in taxation.
Who Taxes What?
The tax structures used to generate revenues differ as widely around
the world as do tax levels. The particular pattern of taxes found in any
country, like the tax level, depends upon many factors, such as the tax
structures found in neighbouring countries, the economic structure of the
country, and its history. In addition, of course, different countries may place
varying importance on such commonly accepted characteristics of a good tax
system as fairness, revenues, economic effects, collection costs, and so on.
Nonetheless, within limits, it can be instructive to look at world averages as
one approach to understanding tax structures in countries with different
characteristics.
6
A regression of taxes as a percent of GDP on per capita GDP has an R square of only 0.064,
though it rises to 0.174 when the equation is run in double log form. The constant term is
highly significant in the linear equation with a value of 16.7, which is only a little below the
average share paid in taxes.
10
Figure 3: Tax Compone nts as a Pe rce ntage of Tax Revenue
Figure 3: Tax Components as a Percentage of Tax Revenue
Ex tractio n
6%
M is c.
5%
D ire ct
3 3%
Int. T rade
18 %
Co ns umptio n
3 8%
As Figure 3 shows, for example, for our sample as a whole, consumption
taxes (38.7 percent of the total) – including both excises and general sales
taxes such as the value-added tax -- and direct taxes (33.2 percent) -- taxes on
income and property taxes -- account for by far the largest share of tax
collections. Taxes on imports and exports account for another 18.7 percent,
with the balance largely consisting of “extraction taxes.”7 Within the
consumption tax category, value-added taxes (VATs) account for 40 percent
and excises for 38 percent of the total, and within the direct tax category,
personal income taxes are by far the most important, accounting for 47
percent, compared to 36 percent for corporate taxes and 17 percent for
property taxes.
7
Extraction taxes may alternatively be thought of equivalent to profits taxes levied on specific
industries. If so, in total “income” taxes would then amount to 39.1 percent of all taxes on
average. Thirty-eight countries report extraction taxes, with United Arab Emirates (90
percent) and Algeria (78 percent) reporting the highest share of revenues collected from
extraction taxes.
11
Once again, the data suggest that a country’s revenue structure depends
to some extent upon its location and economic structure. Small island
countries such as the Bahamas and Saint Vincent and the Grenadines, for
example, largely account for the surprisingly high share of taxes from
international trade. More generally, and unsurprisingly, trade taxes tend as a
rule to be more important in lower-income countries which, as a group, on
average get 24.6 percent of their tax revenue from this source compared to
only 2.6 percent in high-income countries (Figure 4). Taxes on international
trade – mainly customs duties – seem almost always to decline as countries
become more developed.8
An interesting exception is the transitional
countries, which, although many of them fall into the low-income group as
defined here, have traditionally relied relatively lightly on taxation of
international trade (Martinez-Vazquez and McNabb, 2000).
Figure 4: Tax Components by GDP per Capita Category
60
50
40
Low
30
M id
20
High
10
0
Direc t
Co ns umptio n
Int. Trade
8
M isc .
E xtra ction
The coefficient is negative and statistically significant in a regression of per capita GDP on
international trade taxes as a share of GDP.
12
Several other consistent statistical relationships exist between per
capita income and tax patterns.9 First, the higher the level of per capita
income, the greater the reliance on direct taxes, especially those on personal
income. Secondly, although the use of consumption taxes rises more slowly
with respect to income than that of income taxes, developed countries on
average also rely more heavily on consumption taxes than less developed
countries. Thirdly, in contrast, trade taxes fall as both a share of GDP and of
total taxes as income levels rise.10 As earlier studies have found, trade taxes
clearly decline in importance as income rises (Burgess and Stern, 1993).
The observed differences in tax structures in large part appear to
reflect the characteristics of higher- as compared to lower-income countries.
Direct taxes generally require a more effective tax administration and
taxpayers who are more sophisticated, conditions that are more likely to exist
in developed countries. Low-income countries tend to raise more revenues at
the border, where relatively fewer collection points must be controlled. For
similar reasons, consumption taxes in high-income countries are more likely
to take the form of broad-based sales taxes, such as the VAT, while lowerincome countries are likely to rely more heavily on excise taxes on tobacco
products, alcoholic beverages, and motor fuels—taxes which can again be
collected from a relatively small number of vendors, and in some cases at the
customs border.
Has Much Changed Recently?
Over the relatively short time period covered in our data – for most
countries, five or six years in the 1990s – the level of taxation has risen only
slightly on average, from 18.0 percent to 18.8 percent of GDP. Although taxes
9
Three types of regressions were estimated to quantify the relationship between income and
tax structure more fully. Relationships between GDP per capita and each type of tax as a share
of GDP, each type of tax as a share of total revenues, and taxes per capita were analyzed. The
results in all cases were consistent.
10
The coefficient of international trade taxes as a share of GDP is negative, with an elasticity
of -0.22. The income elasticity of international trade taxes is .78 compared to an income
elasticity for direct taxes of 1.23 and for consumption taxes of 1.11.
13
appear to have gone up in most regions of the world, they actually declined
slightly in Asia over this period. Taking a somewhat longer perspective, Tanzi
(1987) reported for the end of the 1970s an average tax ratio of 17.8 percent
for the 86 developing countries in his sample. The comparable ratio over a
decade later for the 75 countries for which overlapping data are available was
18.6 percent, suggesting that tax levels have been rising gradually over time
for the developing countries as a whole. Closer analysis of our sample
confirms that taxes rose for each income category and most substantially for
the middle-income category – by about 1.0 percent on average. (Figure 5).
Figure 5: Change in Tax as a Percentage of GDP by GDP per Capita
Category
25
20
Firs t Ye a r
15
L a s t Ye a r
10
5
0
Low
M id
H ig h
T o ta l
A useful way to summarize revenue growth over time is in terms of
what is called “tax buoyancy,” which is the percent change in tax revenue
divided by the percent change in GDP. A buoyancy greater than 1.0 means
revenues are growing more rapidly than GDP; a buoyancy equal to1.0 means
14
revenues and GDP are growing at the same pace; and a buoyancy of less than
1.0 means revenues are growing more slowly than GDP. Since, as already
noted, revenues on average have been growing more quickly than GDP, the
average buoyancy for all countries in our sample period was 1.03. Although
buoyancies were roughly equal in our three income categories, in general they
tended to be lower in Africa and especially in Asia, than elsewhere.
The relative importance of different taxes has changed in recent years.
The most striking feature has been a pronounced increase in the share of
revenues generated through consumption taxes. One reason for this increase
has no doubt been the continued move to the adoption of broad-based valueadded taxes, which rose from 34.6 to 40.6 percent of all consumption taxes
during the short period we studied. About 70 percent of the world’s
population lives in the 123 or more countries that now levy VATs (Ebrill, et
al., 2001). Interestingly, the share of revenues raised from direct taxes has also
increased slightly, particularly taxes on personal income and property. In
contrast, although the change is small, reliance on corporate taxation has
fallen. Taxes on international trade have dropped much more dramatically,
decreasing by 4.3 percent of total collections – a decline that has been more or
less offset by the 4.1 percent increase in consumption taxes. The use of trade
taxes has dropped even more over the longer term. In 1981, for example, trade
taxes accounted for 30.6 percent of developing country revenues (Tanzi,
1987), compared to only 24.3 percent for our comparable countries in 1998.
Conclusion
The focus of the present report is on developing and transitional
countries. A very wide variety of countries, with very disparate tax levels and
structures, are obviously encompassed by this term. The transitional countries,
particularly in central and Eastern Europe are clearly one group, diverse
within itself but sharing certain common characteristics and to some extent
common problems. The special tax problems facing such countries have been
discussed in detail elsewhere (Martinez-Vazquez and McNabb, 2000) and are
not, for the most part, rehearsed here in detail.
15
The all too many countries in sub-Saharan Africa and elsewhere that
have not managed as yet to create effective basic governance structures
constitute another important group. These countries generally need to develop
much more effective and efficient tax systems than now exist if they are to be
able both to provide for the needs of their people and to engage more
effectively in the world economy. Chapter 3 is concerned essentially with
setting out some of the conditions needed for more effective mobilization of
domestic resources through taxation in those countries. A third distinct
category consists of countries, like many in Latin America and Asia, that have
progressed some distance – widely varying distances in different cases –
towards development goals but still have some way to go. Such countries in
particular are facing increasing problems in coping with some fiscal aspects of
globalization, as discussed further in Chapter 4. Of course, even countries not
so far down the development path may also face significant tax problems as a
result of globalization, in particular because of their frequently heavy
dependence on trade taxation, as highlighted in the present chapter. Countries
closer to the cusp of development, however, also have to cope with potentially
troublesome and important problems in the income tax area.
16
Chapter 2
Mobilizing Domestic Resources for
Development
A tax system, properly designed and utilized, generates sufficient
revenue to finance public sector activities in a non-inflationary way. As
indicated in chapter 2, different countries accomplish this task in very
different ways. Exactly how countries choose to structure their tax systems
depends upon the factors discussed there – the level of development, the need
and desire for increased public services, and the capacity to levy taxes
effectively. It also depends upon the weights that different countries attach to
such public policy goals as attaining a desired distribution of income and
wealth and increasing the rate of national (and perhaps regional) economic
growth. Taxes influence every aspect of the economy -- the willingness to
work in the formal sector, the extent to which people save, the way they hold
their assets, the choice of business structures, the start-up and location of
businesses, and so on. Levying taxes in a manner that fosters desired
responses or at least that generates the least perverse reactions possible can be
one important way to achieve policy goals.
Although governments have policy instruments other than taxes to
achieve their goals, the relative importance attached to different goals by those
making policy decisions is an important determinant of tax policy. In some
cases, measures intended to achieve one goal make the achievement of some
other goal more difficult. The classic, although often overstated, example of
conflicting goals is the choice between redistribution and growth. It is thus
important to understand the real nature of such policy trade-offs if countries
are to establish the best feasible tax system for their objectives and
circumstances.
No one likes to pay taxes. People often seek to avoid paying them, and
governments are wary of the political fallout from imposing them. But taxes
17
are always and everywhere a necessary feature of good government. They are
necessary both to finance desired public spending in a non-inflationary way
and also to ensure that the burden of paying for such spending is fairly
distributed. Despite their necessity and the benefits they provide, taxes almost
always impose real costs on society. Minimizing those costs is thus as
essential an element in good tax policy as meeting revenue needs fairly.
The importance and meaning of the costs of taxation is discussed further
in the next few sections, which summarize some of what researchers have
learned about taxation from the extensive research and experience around the
world over the last few decades, the allocation of resources, and economic
growth. As already mentioned, however, tax policy is not just about
economics. In every country, tax policy also reflects political factors,
including concerns about fairness. A subsequent section of the present
chapter, therefore, considers the surprisingly scanty evidence on the
distributive effect of taxes and the issue of equity and fairness in taxation.
Finally, since no matter what a particular country may want to do with its tax
system, or, viewed from some perspective or other, what it should do with
respect to taxation, what it can do is often constrained in reality, especially in
the short run, by its administrative capacity.
Developing countries face particular problems in administering a tax
system effectively and in meeting some traditional fairness goals. As noted in
chapter 2, such countries on the whole rely more heavily on excise and trade
taxes, and correspondingly less heavily on income taxes, than do developed
countries. Tax structures in developing countries differ in this way largely
owing to differences in economic structure.11 In most developing countries,
for example, the agricultural sector is relatively large and has low
productivity. Even in the urban sector, the “informal” economy tends to be
relatively much larger than in more developed countries.12 Since wage and
other formal sector income is relatively less important and record-keeping is
11
A century ago, the now developed countries also relied heavily on excises and trade taxes, in
part again reflecting the nature of their economies at that time.
12
The informal economy is also relatively large in a number of transitional countries (e.g.
Ukraine) as well as in some relatively advanced developing countries (e.g. Argentina).
18
often of lower quality, income taxation in developing countries is on the
whole both different and more difficult than in more developed countries.
Even the formal urban sector in many countries consists largely of
small firms, with the larger firms often being state-owned or owned in whole
or part by foreigners. It is generally more costly and difficult to collect a given
amount of taxes from many small firms than from a few large ones. On the
other hand, larger firms may sometimes be more sophisticated in tax matters
than the tax administration, particularly with respect to cross-border
transactions. The first of these problems makes it important to keep tax
structures affecting small firms relatively simple. The second problem,
however, requires tax structures to cope with the complex transactions of large
transnational firms. It is obviously difficult for countries with limited
resources to cope with both problems at once. A later section of this chapter
discusses further the critical question of the relation between tax
administration and tax policy.13
Appropriate tax strategies for developing and transitional countries
are, for the most part, discussed in this chapter without explicit attention to the
important global developments discussed in detail in chapter 4 below.
Although this division between domestic and global factors is somewhat
artificial (and is not adhered to strictly, since some international factors are
considered briefly even in the present chapter), it seems appropriate initially to
consider national tax policies apart from the global context both to make the
topic more manageable and also because, as pointed out in chapter 1, the
primary role in achieving developmental goals such as those set out in the
Millennium declaration is inevitably played by resources raised domestically.
13
In recent years, an additional factor that has become important in developing good tax
policy in many developing and transitional countries has been decentralization, but discussion
of this complex issue has been confined to Appendix A in order to keep the discussion in the
present chapter more tightly focused.
19
The Costs of Taxation
Economists are sometimes accused of over-emphasizing the costs of
taxation and the importance of efficient resource allocation. The reality is,
however, that taxes do, almost inevitably, impose such costs, and it is surely
in the interests of poor countries in particular to minimize such “deadweight
losses,” which typically reduce the resources available to achieve socially
desired objectives. Countries in which, by definition, resources are scarce
should be particularly concerned to use as efficiently as possible the resources
that they have and hence to ensure that the taxes they impose are as efficient
as possible. “Efficient resource allocation” may sound like a theoretical
abstraction of interest only to economists, but it is clearly better than the
alternative of, in effect, simply wasting scarce resources and hence making
society as a whole worse off.
The confusion apparent in some public discussion of these matters
may arise from a misunderstanding what is meant by the “costs” of taxation.
Most people probably think of such costs in terms of the taxes collected. In
economic terms, however, taxes are simply a means of transferring resources
from private to public use and are not themselves a cost in economic terms.
Economic costs are incurred only when the total volume of resources
available for society’s use, whether for public or private purposes, is reduced
by taxes. Costs in this sense arise from the process of taxation for several
reasons.
First, and most obviously, taxes cost something to collect. On average
in developed countries, perhaps 1 percent or so of tax revenues must be
devoted to covering the budgetary costs of tax collection. The costs of tax
administration may often be relatively higher in developing countries – a
recent study in Guatemala, for example, estimated them at 2.5 percent of
collections (Mann, 2002) – but the costs of administering the tax system are
seldom a significant factor in shaping tax policy.
Another real economic cost involved in collecting taxes are the
“compliance costs” incurred by taxpayers in meeting their tax obligations,
20
over and above the actual payment of tax. Tax administration and tax
compliance interact in many ways. Often, administration costs are reduced
when compliance costs are increased, e.g., when taxpayers are required to
provide more information thus increasing compliance costs, but making tax
administration easier and less costly. There may not always be a tradeoff
between administration and compliance costs, however. Compliance costs
may increase along with administration costs when, for instance, a more
sophisticated tax administration requires more information from taxpayers,
undertakes more audits, and so forth. Conversely, administrative and
compliance costs may fall if the tax laws are simplified or interpretive issues
that have generated conflicts get resolved.
Compliance costs are incurred not only by taxpayers but by all who are
involved in the tax process – for example, employers withholding income
taxes from employees and banks who collect and remit taxes to government.
Such costs include the financial and time costs of complying with the tax law,
such as acquiring the knowledge and information needed to do so, setting up
required accounting systems, obtaining and transmitting the required data,
payments to professional advisors and so on. The measurement of such costs
is still in its infancy, but a number of studies of their magnitude and
distribution have been carried out in several developed countries (Sandford,
1995). On average, compliance costs in developed countries are perhaps four
to five times larger than the direct administrative costs incurred by
governments. One of the very few studies of such costs in a developing
country (India) suggests that they may in some instances be considerably
higher than for the developed countries (Box 1). In terms of the costs borne by
business, compliance costs are typically much higher with respect to taxes
collected from smaller, formal market firms than from larger firms.
21
In terms of the costs
borne by business,
compliance costs are
typically much higher
with respect to taxes
collected from smaller,
formal market firms
than from larger firms.
Box 1 — Compliance Costs
A recent careful study of compliance costs with respect to the personal
income tax in India (Chattopay and Das Gupta, 2002) suggests that
compliance costs in that country are especially high for non-salary earners,
whose costs appeared to be up to 10 times higher than those for salaried
employees (whose taxes are essentially withheld). Even for salary income,
estimated compliance cost as a percentage of collections in this study was
found to be up to ten times higher than in developed countries. The
authors of the report characterize their findings as “preliminary and subject
to error,” due to sampling issues. Notwithstanding this caveat, the study
provides highly credible evidence of serious problems with India’s income
tax. These results do not mean that compliance costs are always likely to be
so high for the personal income tax in developing countries. But they do
suggest that a tax like that in India is an inappropriate model for other
developing countries to follow. As reported in the study, India’s personal
income tax is exceptionally complex, is shot full of exemptions and
concessions, has a highly progressive nominal rate schedule, and is
administered in an apparently capricious and dishonest fashion. In brief, the
tax is badly designed for the circumstances in which it must operate and
cries out for major reform.
Finally, and most importantly, taxes may give rise to what economists
call “deadweight” or “distortion” costs. Almost every conceivable tax has the
potential to alter decisions made by businesses and individuals as the relative
prices they confront are changed.14 In some cases, these costs are the nearly
14
There are a few exceptions. Lump-sum taxes, where the tax burden is the same regardless of
any behavioural responses by taxpayers, are one example, much favoured by theorists. More
importantly in practice, to the extent that taxes fall on economic “rents” – payments to factors
above those needed to induce them into the activity concerned -- they too may not affect
economic activity. Well-designed taxes on natural resources and land, for example, may thus
to some extent produce revenue without economic distortion. Finally, in certain instances,
taxes – again, if properly designed – may even induce desirable behaviour rather than impede
efficient, market-enhancing decisions. Certain environmental levies, for example, or even
crude proxies such as taxes on fuel, may to some extent have such effects. Such instances of
“good” taxes – those with no bad economic effects – should of course be exploited as fully as
possible, just as well-designed charges should to the extent possible, given public policy
objectives, be used to finance certain public sector activities that specifically benefit
identifiable individuals. In the end, however, most of the taxes needed to finance government
will have to come from other sources and will hence give rise to the efficiency costs discussed
in the text.
22
inevitable result of pursuing particular tax policy goals. As a rule,
nevertheless, the resulting changes in behaviour tend to reduce the efficiency
with which resources are used, resulting in lower output and a decline in the
potential well-being of the country as a whole. It is only sensible for a
government to limit the negative impact of such changes in behaviour
whenever feasible. Although the transfer of resources to the government
through taxation may enhance national well-being, that benefit does not
justify the use of inappropriate instruments of taxation.
Decisions to work, for example, are affected by taxes on wages
(personal income taxes, wage taxes, social security taxes, and so forth) that
reduce the incentive to work. The higher the tax rate on wages in the formal
sector, the less attractive it is to work in that sector and the more attractive
untaxed informal sector activity becomes. Consumption taxes too may
discourage work, because they raise the amount of time one must work to pay
for goods and services through the market place. Since taxes are not imposed
on leisure, it is not surprising that, at the margin, the effect of both income and
consumption taxes is likely to discourage (taxed) work.
Of course, taxes not only alter relative prices – in the case of a wage
tax, the net (after-tax) wage – but also income. Since people who work will
have less net income after a wage tax is imposed, they may choose to work
more to offset the income loss. Indeed, in many cases, they may have to work
more simply to keep themselves and their families alive. The net effect on
work of any tax change reflects both this income effect and the effect of the
change in relative prices (the substitution effect).
Voluminous studies of the impact of taxes on work exist for many
countries, with results varying from country to country depending upon the
structure of taxes and the nature of the economy.15 The important point in the
present context, however, is that, regardless of the income effect, the
substitution effect – the change in relative prices – of taxation in itself leads
people always and everywhere to reduce their work decisions and hence, if
15
For one such recent study, of Colombia, for example, see Alm and López-Castaño (2002).
23
those decisions had been economically efficient before the tax, reduces the
potential output of the nation. Such substitution effects are the source of
efficiency losses from taxation. They may be counterbalanced to greater or
lesser degrees by non-fiscal factors or by putting the tax revenues to good use,
but they nonetheless exist. An important concern in designing a good tax
system is thus, to the extent consistent with other policy objectives, to
minimize such effects and hence to minimize the efficiency costs of taxation.
Almost all taxes may have such effects on resource decisions.
Consumption taxes, such as the value added tax may discourage the
consumption of taxed as opposed to untaxed goods (e.g. housing in some
countries). Taxes on gasoline, alcohol, and cigarettes can reduce the
consumption of these items.16 Income taxes, since they tax the return to
savings, may alter the amount of savings or the form in which savings are
held. For example, failure to tax capital gains until they are realized (taken in
cash) encourages the holding of assets that implicitly retain the capital gains
without realization in cash. Taxes may also affect investment, and such effects
may be especially important when economies are more open to trade and
investment. Firms may, for example, choose to locate their activities in any of
a number of countries for many reasons -- the relative costs of production,
access to markets, etc. – but taxes too may affect their choice of location. To
the extent taxes lower the after-tax return on investments in a country or a
region, the level of investment and hence growth may be lower than it would
otherwise be (Box 2). Corporate income taxes may also affect how
corporations structure their finances. For example, debt finance is encouraged
over equity finance when interest on debt capital is deductible and dividends
paid on equity capital are not.
Box 2 — Effective Tax Rates
Taxes are one of many factors affecting investment. To study the impact of
taxes on capital allocation, one must first quantify that impact. There are
two principal measures used to do so: the marginal effective tax rate
16
As noted earlier, not all such effects need be bad: for instance, if tobacco consumption falls,
people may live longer, healthier and more productive lives.
24
(METR) and the average effective tax rate (AETC). METR is the ratio of
the taxes imposed on an increment of investment to the total after-tax
income generated by that incremental investment. In contrast, AETC is the
ratio of the total taxes imposes on a taxpayer to its total income. “Income”
in these formulas is defined broadly to include, respectively, all the benefits
captured by the taxpayer from its marginal investment or its total
investment.
The principal differences between these two measures, and their
significance, may be illustrated with an example. Consider a large wellestablished American manufacturing firm that intends to invest abroad. The
potential investment location will be chosen from three small countries that
are similar in all relevant attributes (including the effectiveness of tax
administration) but the design features of their tax systems. Policy makers in
these three countries need answers to two different questions. How
“competitive” is their tax system compared to the other two? And how
would the fiscal position of the countries be affected should they be chosen
as the location for this foreign investment?
The technical question facing policy analysts in these countries is essentially
what tax measurement to employ, METR or AETR, in answering these
questions. The short answer is that both are relevant. In general, METR
should be used for estimating tax competitiveness, or the investment impact
of the formal tax structure, and AETR should be used to estimate the fiscal
position, or revenue impact, of the investment.
According to economic principles, a rational investor will continue to invest
until the marginal income from additional investment equals the marginal
cost of earning that income. Consequently, the firm will favour the
investment location where the marginal cost is lowest. In this example, the
firm would not consider investing in any of the countries if its marginal cost
of making the investment, determined without reference to taxes, is greater
than the expected marginal revenue. If the investment makes sense without
reference to taxes, however, then the METR comes into play. All else being
equal, the country with the lowest METR would be the most “competitive”
in attracting investment from the firm. The AETR generally would not be
relevant to the firm in making its investment choice. However, for
simplicity, this example ignores the possible impact of the American tax
system on the firm’s investment decisions. If that impact is taken into
account, then the AETC might be relevant to the firm’s investment
decisions, due to the way the American foreign tax credit system operates.
The AETR is the proper measure of the benefits to a country, in terms of
additional tax revenues, from serving as the host for the investment by a
foreign firm. A firm is taxed not only on the taxable income generated by
25
the last dollar invested but also on the income generated by every dollar
invested. For example, if the American firm earned US$1,000 in one of the
countries and the AETR in that country was 30 percent, the country would
obtain tax revenues of US$300.
Exactly how important such tax effects are is a matter of considerable
debate among analysts even in developed countries,17 but the consensus is that
they are almost certainly much more important than was thought thirty or
forty years ago. The efficiency costs of taxation are certainly a considerable
multiple of the administrative and compliance costs mentioned above. The
lowest estimates for developed countries are perhaps 20-30 percent of
revenues collected and much higher figures are not uncommon. Costs of this
magnitude mean that one Euro shifted from the private to the public sector can
cost the private economy 1.2 to 1.3 Euro plus the administrative and
compliance costs. Thus it is conceivable that the loss for which there are no
services in return is nearly one half as much as the value of revenues shifted to
the public sector. To the extent such costs are an inevitable consequence of
rational policy decisions (for example, to redistribute income through the
fiscal system), they may of course be considered acceptable. But losses of this
magnitude require a high payoff in benefits. Further, it is obviously critical to
design taxes to minimize such possible adverse consequences in poor
countries. This point is worth emphasis because, although effects such as
those mentioned above are real, they are not directly visible. The efficiency
cost of taxation arises because something does not happen: some activity did
not occur or occurred in some other form. Output that is not produced,
however, is still output – and potential welfare – lost. What you see is not
what you could get if the taxes needed to achieve public policy objectives
were better designed. Efficiency is no mean virtue anywhere, but it is surely
especially important not to waste resources by taxing inefficiently in countries
that have few resources to waste.
Interaction between the global economy and domestic resource
17
For a recent survey of the effects of taxes on saving, for example, see Bernheim (2002).
26
mobilization is perhaps more clear in terms of the distortionary effects of
taxes. Globalization, new technologies, digitization, and better transportation
all conspire to increase the distortion from taxation by giving both consumers
and producers more options for tax evasion and avoidance. Thus, a focus on
the full cost of taxation is more essential today than it has ever been.
Good tax policy thus requires minimizing unnecessary costs of
taxation. To do this, experience and analysis suggests that, to the extent
possible, three broad rules should be followed: First, tax bases should be as
broad as possible. A broad-based consumption tax, for example, will still
discourage work effort, but choices between taxable and non-taxable goods
and services will not be altered if all are taxed.18 A few items, such as
gasoline, tobacco products and alcohol, may be chosen for differentially
higher taxation for regulatory reasons or because the demand for these
products is relatively unresponsive to taxation so that at any given tax rate
efficiency costs are relatively low and revenues relatively high. Income tax
bases too should be as broad as possible, treating all incomes, no matter from
what source and whether in-kind or in cash, as uniformly as possible.
Second, from an efficiency perspective, tax rates should be as low as
possible, consistent with the need to finance the appropriate functions of
government. Of course, the broader the base, the lower the rate needed to
generate a given amount of revenue. Lower rates are worthwhile, however, in
their own right. The reason is simply because, as noted above, the efficiency
cost of taxes arises from their effect on relative prices, and the size of this
effect is directly related to the tax rate. In fact, the general rule is that the
distortionary effect of taxes increases proportionally to the square of the tax
rate, so that doubling the rate of a tax implies a fourfold increase in its
efficiency costs. From an efficiency perspective, it is thus always better to
18
In theory, in order to minimize efficiency losses different tax rates should be imposed on
each commodity, with higher rates imposed on those goods and services where the changes in
behaviour are the smallest. To do so, however, requires much more information about how
taxes alter behaviour than is available in most countries. Moreover, this approach does not
take administrative and equity concerns into account. For these reasons, in practice it seems
generally advisable to impose a uniform tax rate to the extent possible.
27
Income tax bases too
should be as broad as
possible, treating all
incomes, no matter from
what source and
whether in-kind or in
cash, as uniformly as
possible.
tax rates should be as low
as possible, consistent with
the need to finance the
appropriate functions of
government
raise revenue by imposing a single rate on a broad base rather than dividing
that base into segments and imposing differential rates on each segment. In
practice, of course, as discussed further below, this consideration needs to be
balanced against the equity argument for using a graduated rates schedule to
impose heavier tax burdens on the well-off.
Third, from an efficiency perspective, it is particularly important to
impose taxes on production with care since such taxes directly affect the
location of businesses, alter the ways in which production takes place, change
the forms in which business is conducted, and so forth.19 Taxation at the point
of production is nonetheless often essential in developing and transitional
countries for a variety of reasons. For instance, it is often simpler in countries
with limited administrative capacity to collect excise and even sales taxes at
the point of manufacture. In addition, to the extent that taxes recover the costs
of providing public services to businesses they should obviously be imposed
directly on producers. Finally, as discussed further below, taxes need to be
imposed on corporate income both to prevent tax avoidance by retaining
earnings within a corporation and to collect taxes from foreign-owned firms.
Tax Revenue
A tax system that attains other goals but does not meet revenue needs
is not viable. The need for sufficient revenue to prevent large budget deficits
is obvious. Except in circumstances when some short-term fiscal stimulus
may be considered appropriate for macroeconomic reasons, deficits generally
have such undesirable macroeconomic consequences as crowding out private
investment and causing inflation. Preventing deficits requires good control
over both the expenditure and revenue sides of government but the focus here
is only on the revenue side.
Effective, accurate estimates of likely revenues should be developed as
an essential component of the budget process. To do so, skilled analysts and
the appropriate data are needed in the Ministry of Finance, and possibly also
19
The VAT is rebated on exports, for example, to prevent the tax from operating as a
production levy in export markets.
28
in the legislature, to estimate revenues and more generally to investigate the
revenue, distributional and other consequences of tax policy changes.
Moreover, and importantly, the legislated budget must be structured each year
to operate strictly within the revenue estimate. All these points may seem –
indeed are – obvious. Nonetheless, even these initial conditions for good tax
policy are unfortunately not satisfied in a number of countries.
Within a sound budgetary framework, tax reforms should, as a rule be
undertaken to achieve long-term rather than short-term objectives. 20 Tax
structures should not normally be altered on a temporary basis to meet
anticipated current year shortfalls. Frequent changes in tax laws increase
compliance costs and, to the extent businesses make production and location
decisions on the basis of a particular tax structure, they may impose efficiency
costs as well. Of course, minor changes in the tax laws to clarify them or to
plug loopholes are often required simply to preserve the original intent of
those laws. Such minor repairs of the tax structure, like repairs of an earthen
dam, are appropriate and often necessary. What is particularly unwise is to use
annual tax changes to avoid the need for major structural tax reforms.
Unless tax revenues grow sufficiently quickly to finance desired
services over the long term, governments need either to reduce expenditures,
to raise tax rates, or to alter other structural characteristics of the system. To
provide a stable environment for economic activity, to the extent possible, a
good tax system should yield an appropriate revenue growth path that equals
the necessary trend of expenditures. Although of course the precise rate at
which revenues should grow differs from country to country in the light of the
demand for public services, a useful baseline is to expect that the growth rate
should be at least the same as the overall economic growth rate, unless the
country wants to increase (or reduce) the size of its government.
20
The immediate revenue effects of a tax policy change need not reflect its long-term effects,
owing both to transitional aspects of the new structure and to the fact that taxpayers often
change their behaviour temporarily to take advantage of higher or lower tax burdens in the
years before or after the changes.
29
Tax policies enacted in such
economically and politically
difficult circumstances have
often come up short in terms
of resolving the underlying
basic problem of inadequate
revenue elasticity.
The rate at which revenues increase over time can differ dramatically
depending on the tax structure, the quality of tax administration, and the type
of economic growth. When revenues consistently grow more slowly than the
demand for public services, tax reform is usually required with a goal of
raising revenue growth. The “income elasticity” of a tax system measures how
fast revenues grow relative to the economy.21 Tax elasticity is defined as the
percentage change in tax revenues divided by the percentage change in GDP
(or potential tax base, such as personal income). An elasticity equal to one, for
example, means that tax revenues grow at the same rate as the economy
(meaning tax revenues will remain a constant share of GDP without the need
to enact policy changes), while an elasticity greater than one indicates that tax
revenues grow more rapidly than income.
In principle, revenues should grow at the same rate as desired
expenditures (that is, the income-elasticity for revenues and expenditures
should be the same). In practice, however, many developing and transitional
countries have had great difficulty in achieving this target, leading to frequent
tax “reforms” aimed primarily at closing short-term revenue gaps. Tax
policies enacted in such economically and politically difficult circumstances
have often come up short in terms of resolving the underlying basic problem
of inadequate revenue elasticity. The likelihood is that the reforms are shortterm patches that make the tax system more complicated and less coherent.
Moreover, as noted above, frequent rate or base changes entail additional
administrative and compliance costs as well as, in many cases, efficiency
costs. Tax changes are costly, and an important consideration in changing
taxes should thus be to avoid future changes to the extent possible. Close
attention should thus be paid to the long-term implications of tax reforms on
revenue elasticity.
The overall elasticity of any tax structure is simply the average of the
Tax “elasticity” refers to revenue growth in the absence of any tax policy changes, while tax
“buoyancy” refers to growth including the effects of such changes. In principle, elasticity is a
better of measure of the growth potential of the tax structure. In practice, however, as in
chapter 2, data limitations often force analysts to rely on tax buoyancies.
21
30
elasticity of individual taxes, weighted by the percentage of total taxes raised
by the tax. The elasticity of a tax depends on the specific characteristics of its
structure. In general, the elasticity of personal income taxes reflects the
progressivity of their rate structure and, most importantly, the level of the
personal exemptions (or zero bracket) relative to average income levels.
Consumption taxes that cover more rapidly growing goods and services and
not just more slowly expanding traditional goods are more elastic, especially
if levied as a percentage of the price (like a VAT) rather than on the specific
number of units purchased (as with many excises). Property tax revenue
obviously rises more rapidly when reappraisals occur on a regular basis and
when property is fully valued, when on average it is rising in value, and so
forth.
Unsurprisingly, revenue growth generally slows during recessions and
accelerates during expansions. Revenue elasticity also tends to rise in
expansions and fall in recessions, thus exacerbating the volatility of revenue
flows. The elasticity of the corporate income tax is often particularly volatile
because in a recession corporate profits fall much more precipitously than
overall economic growth. Countries that depend heavily on taxation of natural
resources such as oil or minerals are especially vulnerable to cyclical swings,
with wide swings in prices shifting tax revenues dramatically and making the
stable provision of government services difficult. Other things being equal,
reliance on a balanced set of tax instruments rather than on a single revenue
source will reduce tax revenue volatility, just as an individual investor can
reduce the volatility of his or her investment portfolio by balancing different
investment categories.
Taxation and Fiscal Adjustment
Chapter 4 considers in detail some of the implications of globalization
for taxation in emerging economies, but one important aspect of taxation in
relation to globalization - the role of taxation in adjustment policy - is
considered briefly here, because in practice in many countries macroeconomic
concerns often tend to dominate all other factors in determining the
appropriate level and structure of taxation.
31
The existing revenue
systems of many countries
are inelastic, meaning
revenue growth is slow
relative to economic
growth and often even
slower compared with the
desired growth in public
expenditures.
As just discussed, in many ways the most important macroeconomic
lesson for fiscal design suggested by experience in recent decades is that many
developing and transitional countries need to improve the elasticity of their
tax systems, that is, to make them more responsive to changes in real and
nominal income. The existing revenue systems of many countries are
inelastic, meaning revenue growth is slow relative to economic growth and
often even slower compared with the desired growth in public expenditures.22
Continued tax reforms - reforms that all too often amount to little more than
ad hoc changes in tax structure and tax administration intended simply to
bring in more money to deal with immediate fiscal needs - are required simply
to maintain the current level of government spending (relative to the
economy). In such an environment of repeated fiscal crises, the changes made
to generate additional revenues are almost inevitably those that are most
politically expedient, with relatively little attention being paid to good tax
policy. Over time the result of this process is likely to be a tax system that
resembles a patchwork quilt and that deviates substantially from the system
that would best achieve the country’s policy goals.
Observers of fiscal changes in countries such as Colombia and Mexico,
for example, have sometimes referred to the “fiscal constant,” by which they
mean the observed fact that, despite frequent tax changes, the government’s
share of national income has on the whole remained remarkably constant for
decades. In some transition countries, such as Ukraine, despite frequent
reforms, the tax ratio (share of the economy paid in taxes) has even decreased
substantially. Such tax structure instability may not increase the tax ratio, but
it inevitably creates uncertainty and, as noted earlier, tends to reduce private
investment. Furthermore, the haphazard tax structure that tends to result from
this process generally increases economic distortions, often in ways not fully
perceived by policy-makers.
22
Chapter 2 showed that such countries on the whole had tax systems in which buoyancy was
slightly greater than one, but of course in many instances the underlying elasticity of the
system is lower and often less than unity.
32
Even countries experiencing economic growth may encounter problems
of declining tax revenues relative to GDP. Sustained growth usually requires
additional investment in infrastructure and human capital. Increased public
expenditure is often needed to maintain and sustain productivity increases.
With relatively inelastic tax systems, however, unless tax rates or bases are
increased, decreased public spending, with detrimental effects on growth (and
perhaps also on political stability) or increased budget deficits with similar
dire consequences are likely to result. Since deficits are, in the circumstances
of developing and transition countries, usually financed by monetary
expansion, leading to inflationary increases in price levels that are likely in
turn to lead to demands for more expenditures just to maintain service levels.
Since revenues have not expanded in step with nominal income increases, this
vicious cycle may continue until a country’s domestic and international
credibility vanishes with its rising indebtedness, and a serious macroeconomic
adjustment is required.
Matters may be even worse when, as is often the case, a vicious cycle is
initiated not by a burst of real economic growth but by unsustainable
monetary expansion. Even in those fortunate countries that have been able to
achieve sustained growth for some period, buoyed by commodity demand or
some other factor, a day of cyclical reckoning usually comes. For example, if
expenditures rise quickly when revenues from trade or natural resource taxes
increase, experience suggests that they are most unlikely to be contracted
equally quickly when revenues decline, thus launching the fiscal system once
again on a downward spiral. To a considerable extent, such swings are
cushioned in developed countries by the higher overall elasticity of their
revenue systems. Less fortunate countries, often subject to much stronger
swings, would also seem well advised to pay more attention in designing taxes
to the likelihood of such cyclical problems than generally seems to be the
case. (Box 3).
33
Box 3 — The Case for More Elastic Taxes
Not everyone would agree with the argument that a tax system that is more
responsive to growth and inflation is an essential ingredient of a sound tax
strategy in an emerging economy. There are three key issues with respect to
which ideas and interests may differ widely — the appropriate size of
government, the appropriate structure of expenditures, and the effects of
taxation on growth. If one thinks that the best government is that which
governs least, then tax policies that will, over time, tend to maintain or
increase the size of government obviously will not be welcome. One might
also oppose an elastic tax structure if one believes that public expenditures,
in practice, are sufficiently wasteful that cuts in the relative size of the
government are desirable, notwithstanding the theoretical benefits of
maintaining or increasing the size of the government. Finally, even if one
considers both the level and the structure of expenditures to be acceptable,
one might still oppose an elastic tax system because such a system
necessarily will tax the faster-growing sectors more heavily than the slower
growing sectors, with possible negative effects on overall growth.
Each of these concerns may be legitimate to varying degrees in different
countries at different times under different philosophies of the role of
government in society. Broad generalizations about the desirability of
improving the income-elasticity of tax systems are as suspect as most other
generalizations in this diverse world. Nonetheless, barring very extreme
cases where one or more of these three concerns dominates all other
considerations, it is not possible to find any good argument in favour of the
present predominantly inelastic nature of the tax systems of most developing
and transitional countries. Consequently, measures to improve the tax
system and increase revenue elasticity almost invariably constitute an
essential part of any adjustment package in such countries. The tax policy
question facing governments in most developing and transitional countries is
generally not whether revenues should be increased, but how they can be
increased.
The problem facing many countries is all too often not whether to
increase revenues but how to do so. Essentially, there are only three
possibilities: raise rates, expand bases, and improve administration. Simply
increasing tax rates within the existing system is the most obvious and often
most politically acceptable approach, but it is often the least desirable. Raising
rates in a fundamentally weak tax system in a country in which traditional tax
bases are declining and substantial structural change is taking place may not
34
increase revenue much. Even if it does, the country must be prepared to keep
increasing rates because the relationship between revenue growth and
economic growth is probably slowed by the rate increases. Further, the result
may be to make taxes both more inequitable and more inefficient. Inequity
increases because only those few unfortunates trapped within the tax system
will bear the burden, and (since the distortions associated with taxation
increase with the square of the tax rate) inefficiency will also go up sharply
with rate increases. Such problems are often exacerbated because it is often
politically expedient to increase taxes most on the politically weaker segments
of society or perhaps, in some cases, on the most economically mobile sector,
foreign investors.
Higher tax rates do not solve the problems arising from low revenue
elasticity. High rates may increase revenues if they can be effectively
enforced, but elasticity depends on the relative growth rate of revenues and
not on the level of revenues. The elasticity is, however, a function of the tax
structure. When higher tax rates do not elicit behavioural changes aimed at
avoiding a tax by reducing taxable activity and taxes are administered
effectively, progressive income tax rates may in principle increase elasticity.
On the other hand, if the higher rates discourage taxable activity excessively
or make it more difficult to administer the tax elasticity may fall even in this
case. Keeping tax rates lower thereby reduces the perverse effects of taxation.
Properly designed base expansions, through such measures as reducing
exemptions and increased efforts to bring non-payers into the tax net,
generally have more desirable effects, but they may also require more political
support to enact. Those who pay higher taxes when their exemptions are
eliminated are usually very vocal in their opposition to the reform, while those
who benefit because the elimination of exemptions available to others results
in a fairer, more efficient, and perhaps more elastic tax structure - and also
because their own tax rates are not raised - seldom voice their support in an
active manner. The small additional revenues that must be paid by each
taxpayer when rates are increased seem, as a rule, to give rise to fewer
politically significant objections than those raised by the more focused, and
35
Higher tax rates do not
solve the problems arising
from low revenue elasticity.
High rates may increase
revenues if they can be
effectively enforced, but
elasticity depends on the
relative growth rate of
revenues and not on the
level of revenues.
therefore larger, increases on a smaller number of taxpayers that occur when
exemptions are eliminated. But it is important, when possible, to expand the
base to include faster growing components of the economy in the tax net.
...whenever possible, it is
much better to respond to
short-term fiscal
pressures by moving
existing taxes, such as
VAT or income taxes,
towards as broad a base
as possible by reducing
base exclusions,
deductions, and
exemptions.
Raising revenues through base expansion also generally requires better
tax administration. New taxpayers must be identified and brought into the tax
net, new collection techniques must be developed and so forth. Such changes
take time, often considerable time, to implement. The goal of increasing
revenues through better enforcement is often espoused, as it was regularly in
Russia during the 1990s, but often, the government has no explicit plan to
achieve this goal. Raising new revenues through base broadening is thus not
as easy or as quick a means of generating additional revenue as simply raising
tax rates. On balance, perhaps the best response to pressures for fiscal
adjustment in the form of additional revenue is, where possible, a basebroadening reform that eliminates tax incentives or exemptions without
significantly increasing the number of taxpayers, thus simplifying rather than
complicating administration.
Short-term measures adopted in crisis conditions should of course be
consistent with long-term objectives. This is easier said than done, however,
in countries in which the only quick way to increase revenue may often be to
increase, say, the rates of taxes on trade or excise taxes on particular
commodities. Nonetheless, whenever possible, it is much better to respond to
short-term fiscal pressures by moving existing taxes, such as VAT or income
taxes, towards as broad a base as possible by reducing base exclusions,
deductions, and exemptions. Only in this way can developing and transitional
countries move towards a sounder long-term tax structure -- one that both
reduces the likelihood of fiscal crises and provides a better basis for any
discretionary rate adjustments that may be required in the future. Broader tax
bases are also of course a more effective means of meeting other goals of
taxation, such as growth and fairness.
Maximizing Growth
Consider a country that is concerned only with economic growth and
36
wishes to design a tax system that is as “pro-growth” as possible without
concern for fairness or other traditional goals of taxation. What might such a
system look like? The simple answer is that no one really knows for sure
because the impact of taxes on growth are difficult to determine in theory and
even more difficult to determine in practice. We suggest, nevertheless, that a
tax system designed only to promote growth probably would have the
following four features:
 It would avoid excessive burdens on business profits so as to avoid
discouraging entrepreneurial activity and risk taking.
 It would place major fiscal reliance in most cases on a broad-based
consumption tax and would be cautious in imposing taxes on saved
income in order to avoid discouraging savings and promoting
capital flight.
 It would impose taxes, to the extent feasible, on the inefficient,
non-monetized sectors of the economy, such as the traditional
agricultural sector and the underground economy in order to
encourage some migration of the resources utilized in those sectors
to more efficient uses.
 It would avoid imposing burdens on low-income workers that
might threaten their ability to work productively.
As a preliminary matter, we note that we are not recommending that
developing and transitional countries pursue a growth-only strategy. In our
view, a tax system should pursue all of the traditional goals of tax policy —
efficiency (growth), fairness (tax equity) and administrative economy (low
costs of collection and of compliance). When those goals conflict, we believe
that a government needs to make sensible tradeoffs. For example, we do not
believe that a country should substantially complicate its tax system for small
gains in efficiency or fairness or that it should pursue a growth-only agenda or
a fairness-only agenda and forsake other legitimate objectives. In addition, we
believe that policy makers often can avoid making tradeoffs among the
competing objectives of taxation through good tax design. When tradeoffs
must be made, we recognize that some differences of opinion are inevitable,
even among ourselves.
Before tradeoffs among competing tax policy goals can be made in an
37
intelligent manner, it is necessary to understand the nature of the tradeoffs
being proposed. For that reason, it is important to understand, within the limits
of current knowledge, what types of measures are likely to promote a growthonly agenda. It is also useful to know why some measures that might seem to
promote growth are unlikely to do so in practice.
Taxation of Business Profits. Some commentators have suggested
that a pro-growth strategy calls for the abolition of all taxes on business
profits. The basic intuition is that a tax on profits is a tax on the rewards of
risk-taking and entrepreneurial activities, that a tax on those rewards is likely
to discourage people from engaging in these activities, and that both of these
activities are essential elements of sustainable economic growth in most
countries. Subject to several important caveats, we believe this basic intuition
is unassailable. Our caveats are important, however, and lead us to conclude
that a moderate level of tax on business profits is consistent even with the
growth-only agenda.
One important caveat is that a tax on abnormally high profits, above
the amount needed to reward entrepreneurial activities and risk taking, is a
highly efficient form of taxation that should be included, to the extent feasible,
in a pro-growth strategy. Economists refer to these supra-normal profits as
economic rents. In principle, therefore, a country following a growth-only
agenda should exempt normal profits and tax abnormal profits.
In practice, the distinction between normal and abnormal profits is
difficult to make. Profits arising from monopoly practices or from extraction
of natural resources are likely to be economic rents, whereas profits earned in
highly competitive markets are likely to be normal profits. Even this broad
generalization must be tempered by the tax-planning opportunities typically
available to business enterprises. For example, a business enterprise may be
able to avoid most of the tax on the normal return on its tangible property by
financing the acquisition of that property with debt and deducting its interest
payments on that debt. Similarly, a business enterprise might reduce
substantially the taxes due from its intangible property by acquiring that
38
property in a transaction that required it to pay out most of the expected
profits as a deductible royalty. The reality is that most successful business
enterprises earn some normal profits and some abnormal profits, and the mix
depends upon the nature of the business and the particular economic
conditions under which it operates.
A second caveat is that the elimination of a tax on business profits will
advance the growth-only agenda only if the substitute tax is superior from an
efficiency perspective. As noted above, all taxes are likely to have some
inefficiencies in practice, and the higher the tax rates, the more significant
those inefficiencies become. In most developing and transitional countries, the
administrative limitations of the tax department are probably sufficient to
introduce inefficiencies into even the best designed taxes, and political
constraints may add additional inefficiencies. If a country replaces the
revenues from a profits tax by increasing the rates on an inefficient alternative
tax, it may not have advanced its growth-only agenda at all, or at least not by
very much.
A third caveat is that an exemption for business profits jeopardizes the
personal income tax, or, indeed, any type of personal tax. Providing an
exemption for business profits within the context of a personal income tax is
difficult, due to the serious practical difficulties that arise in distinguishing
business profits from other types of income. If income labelled “business
profits” is exempt from tax and other forms of income are taxable, it can be
expected that tax planners will arrange for a large portion of the personal
income tax base to fit the criteria for “business profits.” That definitional
problem might be avoided by limiting the exemption to profits earned in
corporate form. That approach, however, also creates serious problems
because taxpayers liable for tax under the personal income tax would have a
strong incentive to avoid that tax by placing their income-producing assets in
a corporation under their control.
Finally, a profits tax is probably the only feasible mechanism available
to developing and transitional countries for obtaining a reasonable share of the
39
profits earned by foreign investors. Obtaining taxes from that source will be
efficient for the host country, to the extent that the tax can be exported to
foreign investors. As discussed above, analysts are unclear on the way the
burden of the corporate tax is shifted, although most would agree that the
extent of the shifting depends at least in part on the facts and circumstances of
particular cases. Given this uncertainty, the promoters of a growth-only
strategy should be exceedingly cautious in recommending the abandonment of
a profits tax on the income earned by foreign investors. Exempting only
domestic earners and taxing foreigners, however, is not a practical option, due
to the competitive problems that such a policy would create.
Our conclusion is that a high tax on business profits is unlikely to form
part of a growth-oriented tax strategy. We do suggest, however, that a stable,
broad-based profits tax imposed at rates that are moderate by international
standards is defensible even in the context of that strategy. Many countries
impose a profits tax at rates of 30 percent or below. If the growth-only agenda
is relaxed somewhat to include other considerations, such as fairness, we
believe that the case for some form of profits tax is strong. As the survey
presented in section 2 shows, most developed and developing countries utilize
some form of profits tax.23 Not all of those taxes are likely to be consistent
with a growth-only agenda, due to their design features and rates. In our view,
the general direction of reform in these countries should be to improve the
design of their profits taxes rather than to abolish them.
The role of tax incentives for investment, for savings, for exports, for
employment, for regional development, and so on must be considered in a
discussion of business profits taxes (Shah, 1995). For the most part, such
incentives are redundant and ineffective, giving up revenue and complicating
the fiscal system without achieving their stated objectives. Even to the extent
that incentives may be effective in inducing investors to behave differently
than they would have done in response to market signals, the result is often
distorting and inefficient, diverting scarce resources into less than optimal
23
Low-income countries generate 11.2 percent of their revenues (1.77 percent of GDP) from
corporate profits taxes; moderate-income countries generate 7.8 percent (1.56 percent of
GDP) and high-income countries raise 8.4 percent (1.95 percent of GDP).
40
uses. Essentially, incentives improve economic performance only to the extent
that government officials are better able to decide the best types and means of
production than are private investors. Incentives also result in very uneven tax
burdens, with domestic companies often continuing to be subject to the full
tax burden, while other firms, particularly multinationals, are subject to much
lower effective tax rates because of the incentives.
Despite such strictures, many developing and transitional countries
continue to introduce and extend a variety of special tax incentives, partly in
competition with one another for increased capital flows from abroad.
Kyrgyzstan (along with many other countries), for example, which developed
a relatively good tax code in the 1990s, has aggressively looked to enterprise
zones, tax concessions, and accelerated depreciation with the hopes of
generating economic stimulus. Much experience, however, suggests that more
important to investment in general are such factors as a sound macroeconomic
policy and a stable governance system. Although there may sometimes be a
limited role for certain simple incentives as part of a growth-oriented fiscal
policy, as some East Asian experience suggests (Bird and Chen, 1998), in
general, incentives can never make up for the absence of such critical factors.
Should a country decide to introduce a few limited incentives for whatever
purpose, it is obviously critical that they should be well-designed, properly
implemented, and periodically evaluated if they are to do more good than
harm (Box 4). On the whole, however, a better strategy to encourage growth is
generally to impose taxes on as broad a base as possible and at the lowest
rates possible.
Box 4 — Tax Incentives
As a rule, disinterested tax analysts, and especially public finance
economists, do not favour tax incentives. The effectiveness of tax
concessions, tax holidays, and the like is difficult to assess, and even welldesigned incentives may be perverted to benefit existing special interests.
For tax incentives to have even a reasonable possibility of serving good
public policy, the country offering the incentives must have a stable
macroeconomic environment and a stable political and administrative
system. Nonetheless, many countries that do not satisfy these conditions
41
continue to have recourse to fiscal incentives for investment. To maximize
the possibilities of beneficial results and to reduce the damage that poorlydesigned and implemented incentives may cause, we suggest the following
rules for countries that feel compelled, by political or other considerations to
adopt tax incentives:
1. Keep it Simple. Complex fiscal incentives are unlikely to produce
desirable results at reasonable cost and are likely to be conducive to evasion
and corruption. Tax incentives, therefore, should be few in number and
simple in structure. Detailed, costly, and complex attempts to divert private
investment into pre-selected channels should be avoided.
2. Keep Records. No matter what incentives are created, clear records
should be kept as to who receives them, for how long, at what cost in
revenue forgone, and with what results in terms of investment, employment,
and so on. Clear procedures need to be in place to follow up the results of
any incentives granted.
3. Evaluate the Results. At regular intervals — say, annually or at most
every three or five years — a country should analyse the data to assess
whether the incentive is worthwhile and, if it is not, to eliminate it. A
country that is unwilling, or unable, to put incentives to this test probably
should not have any incentives.
On the whole,
however, a better
strategy to encourage
growth is generally to
impose taxes on as
broad a base as
possible and at the
lowest rates possible.
Taxation of Savings. A developing or transitional country that is
pursuing purely a growth-oriented strategy generally should avoid taxes that
tend to reduce national savings and should favour taxes that tend to increase
national savings. This basic policy proscription follows from the generally
accepted view that domestic savings are essential for financing the investment
needed for rapid economic growth.
What types of tax policies would encourage savings is a matter of
some debate. Economists generally agree that an income tax creates a bias in
favour of current consumption over future consumption. Because “future
consumption” is largely synonymous with savings, it would seem to follow
that a tax that did not contain that bias would be preferable on efficiency
grounds to an income tax. As a practical matter, the alternative to an income
42
tax — for those developing and transitional countries capable of administering
an income tax — is some form of consumption tax, either a VAT or an
income tax with a deduction for savings.
The argument summarized above for favouring a consumption tax
over an income tax as part of a growth-only agenda is subject to some
important caveats. One important caveat is that it focuses only on the
substitution effect of the income tax and ignores the income effect. As noted
above, a tax that causes taxpayers to shift from one use of their money to
another is generally inefficient, absent some special reason for favouring such
a shift. To the extent that an income tax causes taxpayers to substitute current
consumption for future consumption, it can fairly be said to be inefficient.
In determining the impact of an income tax on national savings,
however, it is not enough to look only at the substitution effect. Another
possible effect of an income tax is that it would cause taxpayers seeking to
achieve some definite savings goal to increase their savings. As an illustration
of this income effect, assume that a taxpayer has determined that he must save
US$800 per year to be able to finance his basic subsistence needs upon
retirement. The government now imposes a tax of 20 percent on his income.
As a result, he must increase his before-tax savings to US$1,000 in order to
achieve his savings goal.
As economists have long recognized, the substitution effect and the
income effect have opposite impacts on national savings (for those who are
net savers rather than net borrowers), and the relative importance of each is a
matter for empirical research. As a result, it is possible in theory that an
income tax would reduce, increase, or have no net effect at all on national
savings. The fact that these two effects tend to be offsetting, however, does
not mean that the inefficiency of the tax is reduced. On the contrary, a tax that
caused some taxpayers to decrease their savings and caused others to increase
their savings would be more inefficient, according to standard economic
theory, than a tax that had only one of these effects. The point is that
efficiency is measured by economists in terms of lost welfare to individuals,
43
and individuals suffer what economists call a deadweight loss whenever they
are induced to change their behaviour on account of a tax. These deadweight
losses do not necessarily indicate, however, that the tax system is having an
adverse impact on aggregate national savings.
A second caveat to the claim that consumption taxes advance a progrowth agenda is that an exemption for the savings component of income
generally requires an increase in the tax rate on the consumption component
of income if the tax is to raise the same revenue as a normal income tax.
Income tax rates can be lower if the base is very broad, a difficulty for many
countries that have failed to tax many forms of private sector earnings. All
taxes, including consumption taxes, have some inefficiencies in practice, no
matter how well they may be designed. An increase in the tax rate on
consumption will increase those inefficiencies, perhaps considerably. Whether
the increase in inefficiencies will offset the efficiency gains from forgoing the
tax on savings is an empirical issue. It depends, firstly, on the actual impact
that a tax preference for savings will have on the national savings rate and,
secondly, on the actual impact on efficiency of the higher rates on consumed
income. That latter effect is likely to depend primarily on the quality of the
consumption tax — the worse the consumption tax, the greater the
inefficiencies caused by the higher tax rates.
A third caveat is that an exclusion for savings in the context of a
normal income tax is not easy to achieve from an administrative perspective.
If the goal is to increase national savings, then the preference for saved
income must be limited to net savings, not gross savings. That is, the tax
system must prevent taxpayers from deducting their saved gross income and
also deducting the costs of earning that income from their taxable amount. In
particular, the tax system needs to prevent taxpayers from taking an interest
deduction that is properly treated as a cost of earning saved income (McIntyre,
1997). Experience in many countries that have offered various savings
incentives strongly suggests that limiting deductions to the costs of earning
consumed income is exceedingly difficult. This point is not relevant, of
course, to countries that are choosing between a VAT and a normal income
44
tax. It applies only if the choice is between a normal income tax and an
income tax with a deduction for savings.
Finally, a country pursuing a growth-only strategy has already
determined that it needs to mobilize resources for development
notwithstanding the preferences its residents may have for current
consumption over future consumption. In effect, a political decision has been
made to give some preference to future generations over the current
generation — to impose current hardships in the hope of a better life for the
children and grandchildren of the current generation. Given this policy
framework, a country may not want to put great emphasis on the welfare
losses of the current generation due to their time preferences for consumption.
The important issue for a growth-only strategy of mobilizing local capital is
how best to achieve the goal of increasing aggregate national savings. In this
respect, a lower-rate income tax may have the advantage because it can raise
the needed revenue with fewer distortions (other than its impact on time
preferences for consumption) than a higher-rate consumption tax. The tax
revenue, once collected, can then be used to finance investment by the
government or, if private investment is preferred, to increase the pool of
private investment capital by paying down some of the national debt
(assuming, realistically, that such debt exists).
For many developing and transitional countries, including most of the
least developed countries, the debate over the relative merits of income and
consumption taxes is not a significant policy issue. They do not have the
administrative capacity to operate a significant income tax, and they have
great difficulties in operating efficient consumption taxes. Their main hope of
increasing national savings is to impose taxes as best they can and devote
whatever portion of those taxes they can afford to investment. The countries
where the debate over income and consumption taxes is most relevant are
those that have already achieved a significant level of economic development.
These countries are the ones that are least likely to adopt a tax policy
focussing exclusively on economic growth.
45
Taxation of Non-Monetized Sectors. A growth-oriented tax system
in the many developing countries with a large non-monetized traditional
sector should reach as far as possible into that sector in order to encourage its
incorporation into the monetary (modern) sector. In many developing
countries, that sector is traditional agriculture. In some developing and
transitional countries, it is the underground economy.
Imposing higher taxes on traditional agriculture may be difficult
politically and administratively, and it may be inequitable as well. Such
taxation, however, is generally conducive to growth. The growth potential of
traditional agriculture is limited; especially in light of the way agriculture is
subsidized in many developed countries. In addition, the existing social
structures in most developing countries make the introduction of modern
farming methods exceedingly difficult. From the perspective of promoting
growth, therefore, a tax that draws resources out of the traditional sector is
likely to be beneficial. At a minimum, the tax system should not discourage
growth by taxing resources away from the growth sectors and leaving the
traditional sectors untouched by taxation.
Taxing the non-monetized sectors is no easy task. An income tax is not
a practical alternative for a variety of obvious reasons. A VAT can be imposed
on the non-monetized sectors and the underground economy only to the extent
that people living or working in those sectors purchase goods and services in
the monetized sectors. In some countries, especially some of the transitional
countries, those purchases are substantial, whereas in other countries they are
negligible. In this regard, the VAT offers a significant advantage because nonregistered taxpayers cannot claim credits for taxes paid at earlier stages of the
production process. Choosing not to zero rate agricultural inputs can help
ensure that the tax is imposed unless the farmers are registered VAT
taxpayers. In many cases, the only practical taxes are taxes on agricultural
land and presumptive income taxes (Box 5).
46
Box 5 — Presumptive Taxes
Presumptive taxes are taxes that “presume” a certain taxable capacity based
on objective (measurable, visible) indicators of production or consumption
such as land, electricity used, employees, machines, vehicles, and so on.
Such taxes, in many different forms, have a long history in many countries.
They seldom produce much revenue and often suffer from many defects of
design and administration. Nonetheless, despite such problems, presumptive
taxes may in principle play two vitally important roles in developing
countries.
First, such taxes are often the only levies effectively imposed on the often
large, non-modern (or at least “non-official”) sector of the economy. The
government may not have the resources or capacity to be able to assess the
income or consumption of farmers or small traders effectively, but it is more
likely to be capable of measuring land and the floor area of buildings or of
counting machines and employees, and so on. Presumptive taxes serve the
double function of both securing some revenue from this hard-to-tax sector
while at the same time encouraging people to become more productive in
order to pay the taxes (which are, if well-designed, “lump-sum” in nature,
that is, with income but no substitution effects and, hence, no efficiency
cost).
Secondly, presumptive taxes can sometimes serve as a backstop for “normal”
taxes in the formal sector. For example, Mexico imposes a minimum tax on
the gross assets of a business. If the profits reported for tax purposes by a
business exceed a certain minimum rate of return on the assets, the profits
tax is applied as usual, but if the reported profits are below the minimal
return, the business is, in effect, taxed on the presumed income generated by
its gross assets.
When presumptive taxes, rather than the regular tax structure, are imposed
on small traders in order to keep administrative and compliance costs
reasonable, care must be taken to ensure that what appears to be a
simplification does not in fact entail additional compliance costs. For
example, legislation to simplify taxes on small businesses in both Russia and
Krygyzstan once required small businesses to calculate tax liabilities using
the methodology imposed on larger corporations and also to calculate taxes
using the new small business structure. The net result was that small firms
were required to calculate taxes using two rather than one approach, thereby
raising rather than lowering compliance costs.
47
Excessive Taxation of Working Poor. Even in the most growthoriented tax system, in which those who make tax policy decisions do not care
at all about the well-being of the existing population, taxes should kept be as
low as possible on the poorest people simply because they must consume to
be productive. Just as some so-called “investment” — for example, in luxury
homes or elaborate office buildings — is not conducive to productivity, so
also some “consumption” is productive. If people do not have enough to eat, if
they are too ill to work, or if they are insufficiently protected from the
elements in terms of clothing and shelter, or if they are not sufficiently
educated to do the work at hand, they are unlikely to be as economically
productive as they would be if they were treated more humanely by their
government.
A tax system, in and of itself, cannot provide for the basic subsistence
needs of poor workers. In can raise revenue from other sources that could be
used to finance public welfare programs and public education. More
fundamentally, it often can avoid imposing taxes that make the poor even
poorer. As a result, tax equity (in the sense of not taxing the poor) and growth
(in the sense of enhancing the productivity of the labour force) are quite
compatible objectives up to some point.
A value-added tax might serve both a growth-only agenda and a
fairness agenda if it provided an exemption for certain specific items that
constituted a significant fraction of the consumption of poor people. For
example, a study of the VAT in Jamaica found that exempting only five
narrowly-defined items would cut in half the VAT burden imposed on
individuals falling within the lowest 40 percent of the income range (Bird and
Miller, 1991). Similarly, exemptions for the poor under an income tax would
serve both the growth-only agenda and a fairness agenda.
We note that serious problems arise under a VAT in limiting the
benefits of exemptions to the poor. As the Jamaica example presented above
indicates, it may be possible in some cases to mitigate the burdens of the VAT
48
on the poor by exempting a limited number of essential consumption items. In
general, however, most exemptions from a VAT are not easily targeted on the
poor, and they complicate tax administration. In providing exemptions for the
poor under a VAT, therefore, some tradeoff is inevitable between growth and
fairness, on the one hand, and administrative economy on the other hand.
Concluding Note. The discussion above of a growth-only agenda
assumed that the policy choices were being made in a stable political
environment and that the basic political structure of the country would not be
affected by the steps taken to implement that agenda. In reality, tax policy and
tax politics are often intertwined. As a simple example, consider the salt tax.
From an efficiency perspective, that tax gets high marks because the demand
for salt is quite inelastic and a tax on salt is unlikely to affect consumption
choices significantly. Yet the salt tax was a contributing factor in the French
Revolution and was a rally cry in India’s quest for independence from Great
Britain. Presumably these major political upheavals resulted in some costs in
efficiency, at least from the perspective of the ruling parties.
Many developing countries have governments that have a fragile hold
on power. In such circumstances, an evaluation of the efficiency of the tax
should at least consider the potential of the tax for increasing political
instability. A tax measure that might get high marks for promoting growth
might be decidedly unattractive even under a growth-only agenda if it was
likely to destabilize the government. For example, if one region of a country is
experiencing most of the economic growth, a tax policy that tended to reduce
regional rivalries might be desirable even if that policy might not get the
highest marks under a narrow concept of efficiency. Ultimately, the efficiency
of a tax needs to be tested by reference to its contribution to the achievement
of the goals of the society and not simply on abstract economic theory.
49
From the perspective of
social and economic
inequality, what matters
in the end is the overall
impact of the budgetary
system on the
distribution of wealth
and income.
Taxation and Equity
In reality, of course, fairness or equity is always a central issue in
taxation. Indeed, from one perspective, the principal rationale for taxes in the
first place may be thought of as an attempt to secure equity. After all, strictly
speaking, national governments do not need taxes to secure money because
they print the money in the first place. The role of the tax system is instead to
take money away from the private sector in as efficient, equitable, and
administratively inexpensive way as possible. Equity, with efficiency and
administration, is thus one of the three principal factors shaping any tax
system. Of course, what any particular person exactly considers equitable or
fair may differ from the conceptions held by others. In the end, only through
its political institutions can any country define and implement its view of what
is an acceptably fair tax system.
The Report of the Secretary-General to the Preparatory Committee for
the High-level International Intergovernmental Event on Financing for
Development, dated 12-23 February 2001, set forth the following more
specific goals for developing countries in the design of their tax systems:
Countries should strive to develop progressive taxation systems
and should endeavour to ensure that the process of adopting taxes is
equitable and participatory through, inter alia, the following policies
and measures:
 Taking measures to ensure that the incidence of taxation
falls justly on different income classes and different
categories of income, such as wages, profits and rents;
 Extending the tax base to cover incomes from activities that are not
currently taxed;
 Expanding indirect taxes and making them more equitable by
targeting the growing modern service sector and socially and
environmentally undesirable activities.
This section considers in particular the first of these goals. The issue of
fairness in taxation may be approached in various ways. For example, one
may consider the specific tax burdens (from the income tax, excise taxes,
50
VAT, etc.) imposed on taxpayers who are thought to be in the same or in
different economic circumstances. Alternatively, one may instead focus on the
overall effects of taxation on income and level of well-being, looking at the
burden of the entire tax system. The policy implications of these two
approaches may be quite different. The first approach focuses on the
implications that details of individual taxes have on the distribution of tax
burdens across categories of people. Thus, one examines how exemptions,
deductions, tax coverage and other tax characteristics affect tax fairness.
Unfortunately, although proposals to reform taxes based on such analysis may
indeed improve particular measures of horizontal and vertical equity within
the limited group actually subject to the full legal burden of the tax in
question, these same features may also have implications for others that may
in some instance make the system as a whole less fair. From the perspective of
social and economic inequality, what matters in the end is the overall impact
of the budgetary system on the distribution of wealth and income. Both
expenditures and taxes should therefore be taken into account in considering
how government policy affects the distribution of income, as well as the
overall equity implications of any tax change.
Consider, for example, the case of a transitional country like the Russian
Federation, in which the big “untaxed sector is not traditional agriculture (as
in many low-income countries) but rather the relatively large shadow
economy. The existence of such a sector may have important implications for
the effects of particular tax policy. For example, to the extent the VAT
functions properly, it will to some extent serve essentially the same function
as a presumptive tax on the informal sector (since credits are only available
for firms that are registered as taxpayers and those earning income in the
shadow sector are taxed when they purchase commodities through the formal
sector). On the other hand, if many high-income recipients operate through the
shadow economy, the effects on equity of increasing the progressivity of the
personal income tax are not always obvious. Government employees and
employees of large, formal market firms may be the primary personal income
taxpayers and hence bear the brunt of such changes. In such circumstances, it
is not inconceivable that indirect taxes such as a VAT and especially certain
51
In short, it is important
in thinking about
taxation and equity to
focus not on
preconceived notions
about labels – for
example, that anything
called a personal
income tax is, by
definition, progressive,
while anything called a
VAT is, by definition,
regressive – but rather
on the reality of how
taxes work in practice.
excises in “higher-income” consumption goods, such as motor vehicles, may
be more progressive than a personal income tax that in reality falls largely on
a limited group of wage earners.24 In short, it is important in thinking about
taxation and equity to focus not on preconceived notions about labels – for
example, that anything called a personal income tax is, by definition,
progressive, while anything called a VAT is, by definition, regressive – but
rather on the reality of how taxes work in practice.
Such arguments do not, of course, mean that there is not an important
role for both corporate and personal income taxes in developing and
transitional countries. As noted in Chapter 2, such taxes are the largest tax
source for high-income countries, and the same is likely to be true in other
countries as their economies develop, more businesses and individuals
become part of the formal economy, and information and tax administration
improves. Right now, however, particularly in the least developed countries,
revenue systems rely heavily on consumption taxes and are likely to continue
to do so for some years to come.
All taxes, whether on income or consumption, may affect equity in
many and complex ways. They may treat people who are in essentially the
same economic position differently (horizontal equity). For example, taxes
may fall more heavily on those who consume alcohol than on those who
consume housing, or on those who get their income in the form of wages
rather than from farms or dividends. Taxes may also differ in their effects on
income distribution (vertical equity). They may tax the rich relatively more
(progressivity) or less (regressivity) than the poor. Some countries may wish
to favour cities; others, rural areas. Some may choose to favour rich savers in
the name of growth, others the poor, in the name of redistribution. Like most
policy instruments, tax policy can play many tunes. What is critical from an
24
A tax is considered progressive when the tax burden as a percent of income is greater for
higher income households than for lower income households, proportional if the percentage of
income paid in taxes stays constant as income rises, and regressive if the percentage paid in
taxes falls as income rises. A higher income taxpayer will normally bear a larger absolute tax
liability than a lower income taxpayer, regardless of whether the tax is progressive,
proportional or regressive, so the difference between the three concepts is how rapidly taxes
rise with income, not whether taxes rise with income.
52
equity perspective is, first, to be aware of the equity implications of tax
reforms for different groups, and, second, to ensure that the actual outcome of
such reforms is consistent with the intended outcome.
Taxation, for example, may not on its own be able to make the poor
richer, but it can certainly make them poorer. Some developed countries such
as Canada and the United States use their tax systems to provide income
support to certain low-income people. Such systems, however, require both
that the tax administration is efficient and that most people file tax returns.
Neither condition is satisfied in most developing and transitional countries. In
such countries, any fiscal attempts at poverty alleviation must therefore be
undertaken primarily on the expenditure side of the budget. Nonetheless, it is
important not to make the poor even poorer through taxes. As mentioned
earlier, it may therefore be desirable for this reason to exempt certain “basic
needs’ items from even the broadest-based consumption tax. Certainly, special
heavy taxes on items that constitute a significant consumption expenditure for
poor people should generally be avoided (Box 6).
53
Box 6 — Tobacco Taxes
An important recent international study (Chaloupa, 2000) has advocated
that all countries should consider much higher tobacco taxes than they
now have, essentially because of the externalities associated with
smoking and because some empirical evidence indicates that those most
deterred by higher prices are younger smokers. Governments around the
world, many of which already tax tobacco fairly heavily — essentially
because they can collect a lot of money from some trapped (addicted)
consumers — may seize on such arguments to justify levying still higher
taxes on those who smoke.
Unfortunately, there are three offsetting arguments to what may seem to
be an overwhelmingly attractive case, on both health and revenue
grounds, for taxing tobacco more heavily. First, tobacco taxes are
perhaps the most consistently regressive of all taxes. Lower income
people in all countries are those who are most heavily burdened by such
taxes. There is no way for most developing countries to overcome this
regressivity by offsetting transfer policies.
Second, smuggling creates a severe constraint on the tolerable height of
tobacco taxes. Again, this problem is especially serious in developing
and transitional countries where the underground economy is already
relatively large and enforcement is weak. Finally, the externality
arguments used to justify high tobacco taxation (public health costs,
smoker-caused fires, etc.) do not, in many cases, support taxes as high as
those that are already imposed and are especially unlikely to do so in
countries that do not have large publicly-funded health care services
Striking the right balance between these various considerations is not
easy in any country. It seems fair to say, however, that there is unlikely to
be an unexploited revenue bonanza for anyone in increased tobacco
taxation.
At the other end of the income distribution scale, taxation is one of the
few ways in which the wealthy may be made less wealthy, short of outright
confiscation. Although past attempts to redistribute income through taxation
have not been very effective in most developing and transitional countries
54
(Chu et al., 2001), it may nonetheless be important in political terms visibly to
tax those who gain the most from economic development – bearing in mind,
of course, the need at the same time to minimize the efficiency costs of unduly
high tax rates. Sustainable tax policy needs to be accepted as fair by those
affected. Even in the poorest countries, for example, automobiles and other
luxury products are more visible than income, so that it should be feasible to
collect progressive taxes on these items, and most people would consider it
fair to do so.25 Striking the right balance in such matters is what good tax
policy is all about.
In considering the equity of tax policy, it is important to understand the
economic incidence of taxation. The law may say, for example, that firms
must pay VAT to the government, but the general expectation is that the real
economic incidence of the VAT is on the ultimate consumer. Similarly, motor
fuel taxes are almost always collected high in the distribution chain (for
example, at import or from major distributors) but again it is generally
expected that consumers will pay the tax. On the other hand, the economic
incidence of property taxes may be borne either by the owners of land and
capital (who also bear the legal incidence) or by the users or renters of the
In considering the equity of
tax policy, it is important to
understand the economic
incidence of taxation.
property, depending upon market conditions.
Indeed, as a general rule, the economic incidence of a tax (at least in the
long run) is determined by market conditions, and not by whether the tax is
legally imposed on the buyer or the seller. It is often difficult to figure out just
who does pay certain taxes in the end (in the economically relevant sense of
having lower real incomes as the result of the imposition of the tax).
Businesses, for example, attempt to pass on (shift) taxes whenever they can to
someone else -- to consumers through higher product prices, to workers
through lower wages, to owners of land through lower land prices. The
incidence of a corporate income tax thus depends on such factors as the
25
As Hughes (1987) notes, taxing fuel correctly can be especially difficult in countries like
Indonesia in which petroleum products (in this case, kerosene) are an essential consumption
item for the poorest people. Bosnia seeks to impose differential rates on fuel for home heating
versus for other purposes (such as transportation), but this has proven to be very difficult to
enforce and is subject to significant evasion.
55
Indeed, as a general
rule, the economic
incidence of a tax (at
least in the long run) is
determined by market
conditions, and not by
whether the tax is
legally imposed on the
buyer or the seller.
openness of the overall economy in terms of the inflows and outflows of
capital investment, as well as on the extent to which capital moves between
the corporate and unincorporated sectors, the relative capital-intensity of
corporations, and the elasticity of demand for goods produced by corporations
and other businesses. None of these factors is easy to measure, and neither is
the incidence of the corporate income tax. Further, since differing economic
conditions in different countries mean that taxes that have the same legal
incidence may have a quite different economic incidence (Shah and Whalley,
1990), the specific characteristics of each country need to be taken into
account when evaluating the economic incidence of each tax.
Taking all these factors into account, what can be said about the role of
income taxes in developing and transitional countries? First, as already noted,
even if the only objective of such a country is economic growth, there is a
clear case for a tax on corporate income, if only to collect a fair share of the
revenue that multinational and large domestic businesses derive from
economic activities in that country. It is of course true that by taxing
corporations differently, and usually more heavily than other forms of
business, governments tend to discourage, at least in principle, the operation
of businesses in corporate form. Still, the fact that most major businesses all
over the world operate in corporate form, despite the prevalence of corporate
income taxes suggests that the corporation remains in general the best
available vehicle for mobilizing large amounts of capital for business
purposes. Since corporations are the engine of development in all modern
societies, a tax on corporations is in effect a tax on the modern, growing sector
of the economy.26 Countries need to tax that sector both to secure the revenue
they need to meet expanding expenditure demands and also to ensure that
those who benefit most from development pay their fair share. 27 But they must
also ensure that corporate taxes are not so high as to discourage growth. In
general, the best approach for developing or transitional countries is thus to
26
In many developing countries, the corporate sector may be quite small to begin with.
Nonetheless, corporate taxes may have an importance beyond the revenue they yield. To the
extent a corporate tax falls on foreign-based corporations, for example, it may provide a
useful window to the way business is conducted in developed countries.
27
For example, Sudan is able to collect nearly five times more revenue from its corporate
taxes than from its personal income tax.
56
impose a moderate, stable tax on all corporations.
As for personal income taxes (as distinguished from general wage
levies such as those imposed in many countries for social security purposes),
their principal objective is presumably largely to mitigate at least to a
moderate degree the inequalities in income and wealth that almost invariably
accompany growth.28 As noted earlier, the recent report of the SecretaryGeneral correctly noted the desirability of expanding the base of consumption
tax to encompass more services in part to reduce regressivity. Unfortunately,
it has often proved exceedingly difficult to include many services, especially
those consumed disproportionately by the rich, in the base of taxes such as the
VAT, so some degree of progressivity in the personal income tax may also
provide a useful offset to the likely regressive impact of the consumption
taxes on which the revenue systems in most lower and middle-income
countries depend. In such countries, for the most part personal income
taxation should likely have a “threshold” – the level of income at which the
tax begins to apply – well above average income levels.29 Moreover, to check
tax avoidance, it would likely also be wise to keep the top marginal rate of the
personal income tax fairly close to the rate of the corporate income tax.
Both corporate and personal income taxes thus have a role to play in
developing and transitional countries. In general, however, given the current
economic context in most such countries, the most important revenue source
in most transitional and developing countries is likely to be a broad-based
value-added tax, as argued earlier. For these countries, income taxes are
complementary taxes, to raise revenue from large business enterprises,
especially foreign enterprises, and perhaps to place some check on the growth
of income inequality. They cannot be expected to replace the revenue
generated by a VAT. Indeed, some developing and transitional countries
currently lack the administrative capacity to even attempt the implementation
28
Low-income countries collect 10.9 percent of revenues (17.1 percent of GDP) from a
personal income tax; moderate-income countries raise 12.2 percent (2.44 percent of GDP),
and high-income countries generate 31.3 percent (7.26 percent of GDP).
29
Sudan allows an annual threshold of USD864 per year, exceeding the earnings of many
public sector employees.
57
But they must also ensure
that corporate taxes are not
so high as to discourage
growth. In general, the best
approach for developing or
transitional countries is thus
to impose a moderate, stable
tax on all corporations
of income taxes.
More generally, it is important to pay close attention to the details of
precisely how particular tax instruments work in particular environments.
Some taxes that on the surface appear to be anti-growth and pro-redistribution
(such as personal income taxes with highly progressive nominal rate
structures) may, at times, have neither of these characteristics, whereas other
taxes, such as the VAT, that may seem regressive may, compared to the
alternatives, actually be mildly progressive (at least between lower and middle
income groups). One cannot judge the effects of a tax by its name but only by
close examination of the details of its design, by the way in which it is
actually implemented, and by the characteristics of the country concerned.
Thus, improving tax
administration is central
to the choice of tax
structures and to
improving taxation in
developing and
transitional countries.
Tax Administration
The best tax policy in the world is worth little if it cannot be
implemented effectively. Tax policy design in developing and transitional
countries must take the administrative dimension of taxation carefully into
account. What can be done may to a considerable extent determines what is
done in any country. In many developing countries, for example, there is a
large traditional agricultural sector that is not easily taxed. In most transitional
and developing countries, there is a significant informal (shadow) economy
that also is largely outside the formal tax structure. The potentially reachable
tax base thus constitutes a smaller portion of total economic activity than in
developed countries. The extent of this untaxed economy is, of course, in part
a function of tax policy itself. For example, the high social insurance tax rates
levied in transition countries such as Croatia create an incentive for a large
informal economy by discouraging employers from reporting the extent of
employment and encouraging the under-reporting of wage levels.30 The
resulting lower tax revenues often lead governments to raise tax rates, further
exacerbating incentives to evade taxes. Unfortunately, all too often when a
30
A number of agencies in addition to the tax administration, such as social security
administration and the customs administration and in some countries the financial police (for
example, in the countries of the former Yugoslavia), are often involved in revenue
administration. In Bosnia, for example, two-thirds of the revenue of the Entities (the main
governmental level) is collected by the customs administration.
58
country’s real tax base is small, so, almost by definition, is the administrative
capacity to reach it effectively. Thus, improving tax administration is central
to the choice of tax structures and to improving taxation in developing and
transitional countries.
As noted earlier, the resources used in administering and complying
with taxes - or, for that matter, evading them - imply real costs, in terms of the
ability of the economy to provide goods and services. An essential aspect of
good tax policy is to keep such costs as low as possible while at that the same
time achieving revenue, economic growth, and distributional goals as
effectively as possible. This is no small task. What needs to be done to
improve tax administration is generally obvious, but seems seldom to be acted
upon. In particular, three ingredients seem essential to effective tax
administration: the political will to implement the tax system effectively, a
clear strategy for achieving this goal and adequate resources for the task. It
helps, of course, if the tax system is well designed, appropriate for the country
in question, and relatively simple, but even the best designed tax system will
not be properly implemented unless the three conditions just mentioned are
fulfilled. Much attention is frequently and correctly paid to the resource
problem - the need to have sufficient trained officials, adequate information
technology and so on. In the absence of a sound implementation strategy,
however, even adequate resources will not ensure success. And in the absence
of sufficient political support, even the best strategy cannot be effectively
implemented.
As with tax policy, so also with tax administration: the most important
two ingredients for successful reform are clear recognition at the highest
political levels of the importance of the task and willingness to support
reform. Unfortunately, very few developing or transitional countries have so
far proved able to clear these initial hurdles with respect tax administration
reform. More frequently, urged by international agencies or simply desperate
for revenues, countries have launched frantic efforts to corral defaulters or to
reach new taxpayers without hurting politically powerful interests and without
providing the time, resources and consistent long term political support
59
In particular, three
ingredients seem essential
to effective tax
administration: the
political will to implement
the tax system effectively,
a clear strategy for
achieving this goal and
adequate resources for
the task
needed for effective tax administration. The widespread reluctance to collect
taxes efficiently and effectively without fear or favour is understandable in
countries which are often somewhat fragile politically. But there is no way to
obtain a viable long-term tax system without major changes in this respect.
If the political will exists, the techniques needed for effective tax
administration are not a secret. The tax administration must be given an
appropriate institutional form, which in some instances may mean a separate
revenue authority. It must be adequately staffed with trained officials. It
should be properly organized, which in most countries means on a functional
rather than tax-by-tax basis.31 Computerization and appropriate use of modern
information technology can help a lot, but technology alone cannot do the job
and must in any case be carefully integrated into the tax administration. New
computer systems have often developed parallel to the existing structure (in
the Philippines, for example), but little gain can be derived from a system that
does not recognize the skills and needs of tax officials.
Only well-trained people, with adequate political support, can
administer taxes effectively. Adequate provision must thus be made for
training and retraining staff as needed. The information needed for effective
administration must be collected from taxpayers, relevant third parties, and
other government agencies; it must be stored in an accessible and useful
fashion; and, most importantly, it must then be used to ensure that those who
should be on the tax rolls, are, that those who should file returns, do, that
those who should pay on time, do, and that those who do not comply are
identified, prosecuted and punished as appropriate. None of this is easy, and
little of it is simple, but it is not an impossible task. Countries such as
Singapore are models of what can and should be done, and such models
should be studied closely and, once adapted as necessary, implemented.
The first task of any tax administration is to facilitate compliance, that
is, to make sure that those who should be in the system are in the system and
31
Alternatively, organization by client groups may sometimes be sensible. But what is never
sensible is to assign specific taxpayers to specific officials for prolonged periods of time, a
practice still common in some countries.
60
that they comply with the rules.
First, taxpayers must be found. They may be required to register: if so,
registration must be made easy, and systems must be in place to identify
those who do not register voluntarily. An appropriate unique taxpayer
identification system must be established.
Secondly, where appropriate, tax liabilities must be determined. This
may be done administratively (as with most property taxes) or by some
self-assessment procedure as with most income taxes and VATs).32
Thirdly, the taxes due must be collected. In many countries, this is best
done through the banking system. It is seldom appropriate for tax
administrations themselves to handle any money directly.
Lastly, adequate service in the form of information, pamphlets, forms,
advice agencies, payment facilities, telephone and electronic filing, and
so on must be provided to taxpayers to make taxpayer compliance with
the system as easy as possible.
Underlying this approach is the view that in some sense the taxpayer is a
client - not necessarily a willing one of course! - to be served, and not a thief
to be caught. Unfortunately, the latter attitude seems to prevail in all too many
developing and transitional economies.
Of course, in no country are all taxpayers honest, and in some most
may not be, so the second important task of any tax administration is to catch
tax cheats. To do so, ideally the administration should have a good idea of the
extent and nature of the potential tax base, for example, by estimating what is
sometimes called the “tax gap.” At least a rough idea of the number and
nature of those not caught by the tax system is needed in order to work out
how best to bring them into the net. In some instances, the major problem may
be that the authorities simply do not know many potential taxpayers. In others,
32
Globalization confronts tax administrations with some special problems, as discussed in
chapter 4. For example, tax administrations must ensure that revenues and expenses are
properly calculated in determining taxable profits for the corporate income tax, and that
export credits and refunds are properly handled under the VAT. A number of transition
countries (e.g. Ukraine) have failed to provide the rebates provided by the VAT legislation for
exported goods. Taxpayers cannot be expected to respect tax laws when the government does
not.
61
it may be that many taxpayers who are in the system are substantially underreporting their tax base. In still others (and perhaps most), both problems are
important. Without some knowledge of the unreported base, and its
determinants, no administration can properly allocate its resources to
improving fiscal outcomes - whether through “sweeps” to find unregistered
taxpayers or the generally more productive, if technically much more
demanding, route of auditing.
In addition to exploring the nature of the tax gap and undertaking the
often difficult tasks involved in extending the reach of the tax system into the
informal economy to the extent feasible, close attention must also be paid to
the simple task of ensuring that those who are in the system file on time and
pay the amounts due. Immediate follow-up of non-filers and those whose
payments do not match their liabilities is an obvious but too often neglected
aspect of good tax administration. Adequate interest charges must be imposed
on late payments to ensure that non-payment of taxes does not become a
cheap source of finance. Similarly, an adequate penalty structure is needed to
ensure that those who should register do so, that those who should file do so,
and that those who under-report their tax bases are sufficiently penalized to
make the gamble of being caught too risky for most of them.
As with much good advice, most of what has just been said is not only
sensible but obvious. Nonetheless, it is worth saying again, since failure to
follow such advice is astoundingly common. Enforcing a tax system is neither
an easy nor a static task in any country. It is especially difficult in the
changing conditions of developing and transitional countries. Unless this task
is tackled with seriousness and consistency, however, even the best-designed
tax system is unlikely to produce good results.
No government can expect
taxpayers to comply
willingly if they consider
the tax structure unfair or
are unconvinced that the
revenue collected is put to
good use, but even a sound
tax structure and sound
expenditure policy can be
vitiated by capricious and
corrupt administration.
A third major task is to keep the tax administration itself honest. No
government can expect taxpayers to comply willingly if they consider the tax
structure unfair or are unconvinced that the revenue collected is put to good
use, but even a sound tax structure and sound expenditure policy can be
vitiated by capricious and corrupt administration. It took the developed
62
countries centuries to develop and implement sound tax administration
practices aimed at preventing dishonest tax officials from succumbing to
obvious temptations. Unfortunately, most developing and transitional
countries are currently attempting to sustain much larger governmental
structures on precarious fiscal bases and do not have the luxury of centuries to
solve such problems. They must do so now, if they are to survive.
Tax officials must therefore be adequately compensated, so that they do
not need to steal to live. They should be professionally trained, promoted on
the basis of merit, and judged by their adherence to the strictest standards of
legality and morality. They should have relatively little direct contact with
taxpayers and even less discretion in deciding how to treat them. These
statements are all clichés, no doubt, but they are clichés because they are true
– and, unfortunately, still all too relevant in many developing and transitional
countries.
Improved domestic resource mobilization is an essential element of the
stronger policy framework developing and transition countries need to have in
place in order to be able to benefit from the opportunities afforded by
globalization, rather than passively suffering the vicissitudes that may
otherwise be inflicted on countries with weak governance and policy
structures. Money alone will not ensure good governance; but it is necessary if
that goal is to be attained. Similarly, good tax administration is not sufficient
in itself, but it is necessary for effective and efficient domestic resource
mobilization, and hence deserves the highest priority in any serious reform
effort.
The failure to develop good tax administration and good tax policy
together has been a particular problem in some transitional countries. In
Russia, for example, there were both serious problems with the structure of
the VAT and with the lack of administrative experience and capacity. A
simple example is that under the initial VAT legislation, no tax liability was
due when loans were made from one business to another. Thus, a buyer would
claim to have made a loan that was never repaid (and was in fact payment for
63
goods) to a supplier and no tax was due. The result was a significant loss of
revenue. A more capable administration would have foiled this simple evasion
technique, but better legislation was also needed. Even if policies are good,
the way in which they are administered can yield very different outcomes than
those intended. Administration that is seen as unfair and capricious may bring
the tax system as a whole into disrepute. The initial failure of transitional
countries to develop their tax administrations when introducing new tax
structures resulted in very uneven tax imposition, lower than anticipated
revenue, and widespread tax evasion. Similarly, in some developing countries,
in practice corporate tax liabilities are often negotiated rather than calculated
as set out in the law. Bribery is sometimes so common that it is considered a
regular part of the compensation of tax officials. Such corruption undermines
confidence in the tax system, negatively affects willingness to pay taxes, and
reduces a country’s capacity to finance government expenditures.
Finally, some have argued that improved tax administration alone can
generate the revenues required to balance the budget or to meet guidelines set
by the International Monetary Fund, or to finance tax policy changes that will
narrow the tax base through concessions or lower tax rates. Although better
tax administration will indeed generally enhance tax collection, increased
revenue is not necessarily the only goal of improved administration, which is
needed also to improve fairness, for example. Moreover, since improving
administration takes time, any additional revenues may only accrue over a
number of years. Cutting tax rates or granting additional concessions in the
pious hope that improved administration will quickly make up any revenue
losses is never a good idea.
64
Tax Reform
No simple prescription for tax reform emerges from this brief and
general review of some important issues in developing adequate systems for
mobilizing domestic resources in developing and transitional countries. The
precise nature and design of any reform will obviously differ from country to
country, depending upon the relative importance of different policy goals, the
characteristics of the economy, and the existing tax structure and
administration. Whatever the country, however, tax reforms may take one of
two general forms. These types of reform differ only in degree, but the
distinction may be of some importance in understanding the political will
necessary to undertake the reforms.
First, reforms may be intended to correct design flaws in the tax
structure. For example, profits may not be properly measured for the purpose
of levying the corporate income tax, or taxable income for the personal
income tax may exclude fringe benefits. Changes may be needed to correct
such flaws or to deal with administrative difficulties. Such changes may be
very technical in nature and not very visible to many taxpayers, but may
nonetheless be important in achieving public policy goals.
Second, much more basic reforms may be attempted – for example, a
major change in the choice of taxes. For example, until recently many
countries levied turnover sales taxes, under which every transaction was
taxed. Such taxes may sometimes produce considerable revenue at relatively
low rates, but they also have some undesirable effects in economic terms.33
They affect both production and consumption choices, often in unintended
ways. They encourage firms to integrate vertically solely to reduce taxes, and
they lead to very differential burdens on final products, depending upon how
many stages of production (tax points) they pass through. For these reasons,
among others, many countries have chosen to shift from turnover taxes to
value added taxes. Countries make such basic reforms only infrequently,
however, and often only after considerable economic and sometimes political
33
General taxes on financial transactions, such as that in Brazil, have similarly undesirable
characteristics.
65
Cutting tax rates or
granting additional
concessions in the pious
hope that improved
administration will
quickly make up any
revenue losses is never a
good idea.
difficulties.34
Of course, even major reforms can sometimes be done gradually. For
example, Colombia adopted a sales tax with a value-added feature in 1966 and
then moved to a full VAT in 1974. Alternatively, reform can be done all at
once, in what has been called the “big bang” approach. Which approach to
take was much discussed during the 1990s in the transitional countries of
eastern and central Europe, which had inherited a tax system suitable only for
a tightly centrally-planned economy. The “big bang” approach was essentially
to adopt immediately a tax system similar to those used in many marketoriented, highly developed countries, notably those of the European Union.
The “gradual” approach was to develop a reform plan tailored specifically for
the country in question, and to move slowly from the existing structure to the
desired system.
A major problem with the big bang approach is that the necessary
institutional structures to operate a modern tax system effectively may not be
in place. Two obvious problems in most transitional countries, for example,
were the inadequacy of accounting systems and the weakness of tax
administration. A more gradual approach allows time for the accounting
infrastructure of modern taxation and the tax administration to be developed.
On the other hand, a gradual approach often presents significant political
problems. The long time lapse between conceptualization of the tax structure
and its final implementation gives reform opponents continuing opportunities
to prevent the entire set of reforms from ever being undertaken. Also, a
gradual approach entails continued uncertainty for taxpayers, and, as noted
above, such uncertainty raise costs and reduces the willingness to invest.
In the end, most transitional countries adopted more or less a full
package of modern tax laws such as corporate and personal income taxes and
value-added taxes. Some, such as Kazakhstan, adopted a complete new tax
code. In others, like Ukraine, such a code was developed but never
34
For a useful discussion of major reforms in a number of developing countries, see Thirsk
(1997).
66
implemented. Whether codified or not, most of the new tax laws adopted in
transitional countries were, by normal standards, relatively sound in design. In
many cases, however, political pressures that led to a proliferation of tax
concessions soon weakened these laws. In others, continuing administrative
problems significantly weakened the implementation of the new laws.
Although the jury is still out on whether a gradual or a big bang
approach to major tax reform is best, two lessons about the reform process
may perhaps be drawn from both this recent experience in the transitional
countries and more generally from experience with major reforms in
developing countries generally. The first lesson is simply that, with respect to
how to do it as well as what to do, each country faces its own unique problems
and hence in the end will have to find its own unique solutions. The second,
and more useful, lesson is that the dichotomy between doing reform “all at
once” or in gradual stages is to a considerable extent a false dichotomy. In
reality, tax reform is not an issue that can ever be resolved at once.
Circumstances and ideas change, and policies change with them, so that to
some extent, particularly in developing and transitional countries increasingly
open to world influences, tax reform is likely to prove to be a continuous
process, requiring constant vigilance and effort in order to ensure that fiscal
instruments are used as effectively and efficiently as possible to achieve
public policy objectives. A critical ingredient of a sound tax policy for any
developing or transitional country is thus to develop some systemic and
continuing institutional capacity to develop and adapt tax systems
appropriately as the world changes. As emphasized in Chapter 5, this is one
area in which improved international support and assistance may be of great
use to many countries.
67
Circumstances and
ideas change, and
policies change with
them, so that to some
extent, particularly in
developing and
transitional countries
increasingly open to
world influences, tax
reform is likely to prove
to be a continuous
process, requiring
constant vigilance and
effort in order to ensure
that fiscal instruments
are used as effectively
and efficiently as
possible to achieve
public policy objectives.
A critical ingredient of a
sound tax policy for any
developing or
transitional country is
thus to develop some
systemic and continuing
institutional capacity to
develop and adapt tax
systems appropriately as
the world changes.
Chapter 3
The Global Context
The key problem is that
globalization makes it
more difficult for all
countries, including
developing and
transitional countries, to
pursue some of the
traditional goals of tax
policy because
globalization potentially
limits some choices
available to countries.
Every country today has to design its tax policy recognizing that it is a
participant, willingly or unwillingly, in a global economy. The global
economy provides important opportunities for increasing national wealth.
However, it also makes countries vulnerable to international economic forces
that are not always benign. This chapter discusses some of the risks and
rewards of globalization, with respect to tax policy in developing and
transitional countries. The key problem is that globalization makes it more
difficult for all countries, including developing and transitional countries, to
pursue some of the traditional goals of tax policy because globalization
potentially limits some choices available to countries. The key solution is that,
at least in principle, a country can regain some or all of the sovereign power to
tax through mutual cooperation with other governments.
Globalization and National Tax Sovereignty
Taxation is one of the most important aspects of sovereignty. A
government that cannot tax cannot function. A government that is limited in
the way it can tax may be precluded from pursuing some societal goals. For
example, if the wealthy cannot be effectively taxed because of their ability to
avoid the tax through cross-border activities, a country’s ability to mitigate
inequalities in wealth and income is obviously constrained.
Globalization – the extension throughout the world of an economic
system based on free markets for goods, services, and capital – tends to
undermine a country’s power to tax. A tax is an involuntary exaction, paid
only because the taxpayer is under a legal obligation that it cannot avoid.
Cross-border trade and investment tend to increase the opportunities for
taxpayers to avoid taxes, legally and illegally. In some instances, the
increasing concentration of economic and political power in private (and
68
sometimes foreign) hands associated with globalization may also increasingly
challenge the authority of national governments.
Globalization has also brought many benefits to many countries. Over
the past half-century, the market economy has shown itself to be vastly
superior to alternative economic systems in satisfying material needs. The
market economies in developed countries are far from the laissez faire
capitalism of the nineteenth century, however. Western democracies, for
example, typically regulate mergers, protect the rights of workers to organize
into labor unions, and provide an economic safety net that protects citizens
against the hardships of an economic recession.
Some developing countries were initially slow to embrace
globalization and the market economy that drives it, fearing some of
the important social, political, environmental, and cultural changes that
typically accompany globalization or that their economies were too
weak to compete effectively in world markets. Nonetheless, recent
events have pushed many developing countries towards globalization
notwithstanding such fears. Many developing countries that embraced
globalization experienced high levels of growth in the 1990s. On the
other hand, the economic performance of those that eschewed the
market was disappointing. Moreover, the dynamism of globalization at
the end of the century was such that developing countries were being
buffeted by market forces whatever their chosen development model.
In significant measure, countries that had not freed markets to some
extent were suffering the downside of globalization without enjoying
the upside.
One of the benefits that developing countries and economies in
transition hope to obtain from participation in the global economy is an
increase in foreign investment. Various resolutions of the General Assembly
and the Economic and Social Council have asserted that foreign private
capital flows and investment can play an important complementary role in the
economic development process. The traditional benefits that a developing or
69
transitional country might obtained from direct foreign investment include the
following:
 A transfer to the country of managerial and administrative
expertise, technology, and other resources.
 The expansion of productive capacity and employment in the
country
 The establishment of export markets for goods and services
produced in the country.
 The development of local businesses to serve the needs of the
foreign investor or to exploit the markets created by the foreign
investor.
 The collection of tax revenue from the profits earned by the
foreign investor from activities in the country.
 The creation by the foreign investor of an infrastructure that
may promote further economic activities.
Other possible benefits from joining the global economy are that the
efficiency of domestic industries may improve in response to foreign
competition, competition from foreign producers may lower domestic prices,
consumers may enjoy a higher quality of goods and services, and the
corruption that has too often accompanied public monopolies may be
mitigated.
Globalization thus brings with it many potential rewards. But it also
entails risks, particularly because it imposes limits on the scope for
independent national monetary policy (as investors can more freely move
funds in and out of the country), and to some extent perhaps also on the ability
(or the willingness) of governments to prevent anti-competitive mergers, to
prevent the degradation of the environment, and to protect worker rights to
unionize. Most relevantly in the present context, of course, globalization may
impose constraints on tax policy, and particularly on tax policy intended in
part to offset the concentration of income and power that often accompanies
globalization.
70
Developing and transitional countries may feel, for example, that they
must compete for investment capital by offering more and more generous tax
incentives. However, any competitive advantage that a country gets from
offering tax concessions will be short lived. Competing countries are likely
soon to offer a comparable package of tax concessions. The result may be
harmful tax competition among countries that ultimately may work to their
collective disadvantage. In the end, the result may simply be that all countries
in effect will lose the ability to tax the income generated from foreign
investment – one of the “benefits” of such investment, as noted above -without experiencing any significant increase in investment flows.
Not all tax competition need be harmful, of course. To the extent that
international competition induces countries to impose better-designed and
more stable tax structures, to administer them more honestly and uniformly,
and to ensure that the revenues are more effectively utilized, it may be
beneficial. Still, it is undeniable that tax competition may to some extent
undermine attempts to impose redistributive taxes, because such taxes may
result in increased capital outflows. Even some consumption tax policies – for
example, high taxes on luxury goods – may be vitiated if people can easily
purchase goods outside the country free of tax or subject to lower tax rates. In
the end, the only way to deal with such undesired effects of tax competition is
through some degree of increased international cooperation on tax matters.
Such cooperation among countries on tax matters is certainly difficult, as the
history summarized in Appendix B shows, but it is also, as shown there, not
impossible, as discussed further in chapter 5.
Developing and transitional countries have been both the practitioners
and the victims of tax competition. In some cases, such countries have
attempted to attract foreign investment by offering rich tax incentives to
foreign corporations. In response, corporations have sometimes encouraged
developing and transitional countries to engage in “incentive bidding wars.”
Some developed countries have encouraged tax competition by offering taxsparing credits in their tax treaties, thereby making the incentives offered by
71
the OECD actions have
already shown that
concerted efforts by the
major countries of the
world can be effective in
curtailing the use of
offshore tax havens.
developing countries more valuable. A few developing countries have adopted
strict bank secrecy laws and other legal rules that facilitate international tax
avoidance and evasion, and hence in effect invited the use of their country as a
vehicle for international tax evasion. Recently, the OECD has begun a major
initiative at curtailing the use of such offshore tax havens for evading taxes
(OECD, 1998). This initiative has been controversial and has encountered
organized opposition from some governments as well as from those in the
business of promoting tax evasion and avoidance. While it remains unclear
whether the international community is prepared to act on a sustained basis to
control international tax evasion, the OECD actions have already shown that
concerted efforts by the major countries of the world can be effective in
curtailing the use of offshore tax havens.
Most developing and transitional countries would benefit from
effective international restraints on tax evasion, although a few such countries
may lose. For such efforts to be truly beneficial, however, there should be no
exceptions for certain types of cross-border tax evasion that are now
facilitated by policies in some developed countries. It may also be necessary
to provide some appropriate transitional assistance to permit existing “tax
haven” countries to move their economies in new, more productive directions.
Both theory and practice
suggest, however, that in
a market-driven economy
any jobs and profits lost
to imports in one business
are more than made up
for through expansion in
other businesses, and any
gains in jobs and profits
derived from an artificial
increase in exports are
lost in other areas of the
domestic economy
Trade Taxes
Freer trade is a central aspect of globalization and the market
economy, and rightly so. As the Monterrey Consensus states (p. 7):
A universal, rule-based, open, non-discriminatory and equitable
multilateral trading system, as well as meaningful trade
liberalization, can substantially stimulate development worldwide,
benefiting countries at all stages of development.
Opponents of free trade often contend that a relaxation of tariff
barriers will make local businesses more vulnerable to foreign competition,
resulting in lower domestic profits and fewer domestic jobs. The argument is
essentially that national wealth is enhanced by policies that discourage
imports and encourage exports. This mercantilist view has long been
72
discredited. It rests on the belief that jobs and profits lost to imports are
permanent losses and that export subsidies produce permanent gains to an
economy. Both theory and practice suggest, however, that in a market-driven
economy any jobs and profits lost to imports in one business are more than
made up for through expansion in other businesses, and any gains in jobs and
profits derived from an artificial increase in exports are lost in other areas of
the domestic economy. The market generally tends to direct labor and capital
to their most efficient uses, so that free trade tends to make a country as whole
richer, not poorer.
Nonetheless, a cold dash of free trade can have negative short-term
effects on a weak economy in which the necessary economic adjustments
occur slowly. Many of the developing countries exposed to free trade in recent
decades, after decades of protectionism, suffered badly in the short run. The
likelihood that free trade will ultimately work for the economic betterment of
the country may be little consolation to those suffering severe economic
hardship in the short run or to political leaders who are thrown out of office as
a result of popular discontent with the effects of their trade policies.
As the Monterrey Consensus recognized, a country may be able to
mitigate the adverse short-term effects of a relaxation of its barriers to imports
by negotiating a reciprocal relaxation by other countries in their barriers to its
exports. Many developing countries have some established export industries
that could expand their production of goods and services significantly if
foreign countries would eliminate their tariffs, quotas, and other trade barriers
to the importation of those goods and services. Unfortunately, reciprocal trade
agreements may be of limited utility to some least-developed countries that
have no established export businesses other than the export of natural
resources. For this reason, the Monterrey Consensus calls upon the developed
countries to open their markets to the least-developed countries without
requiring trade or other concessions in return.
Generally, however, experience over more than a century has shown
that purely bilateral trade arrangements are unsatisfactory, largely because
73
Generally, however,
experience over more than
a century has shown that
purely bilateral trade
arrangements are
unsatisfactory, largely
because cross-border trade
is not purely bilateral. It is
for that reason that the
Monterrey consensus calls
for an “equitable
multilateral trading
system.”
cross-border trade is not purely bilateral. It is for that reason that the
Monterrey consensus calls for an “equitable multilateral trading system.” The
basic international system in place today is the trading regime managed by the
World Trade Organization (WTO). The WTO has many virtues. Some have,
however, characterized it as having functioned to date as essentially a trading
club established by and managed largely for the benefit of developed
countries. From the perspective of developing countries, it has largely treated
their needs and interests as a secondary concern and excluded them from a
meaningful say in its decision-making processes. In the words of the Zedillo
Report (p.6):
Despite its youth, WTO is in urgent need of reform and support in
certain critical aspects. The necessary changes are unlikely to be
achieved from within. What may be needed is a bigger political
impulse, stemming from the construction of global economic
governance. In that endeavour, at least the following aspects of WTO
should be addressed:



Its decision-making system, which many developing
countries perceive, with reason, as selective and
exclusionary;
Its capacity to provide technical assistance to
developing countries so they can participate more
effectively in multilateral trade negotiations, trade
opportunities and the dispute settlement mechanism;
Attached to the latter, the evident underfunding and
understaffing of WTO.
Although the broader aspects of international trade policy are beyond the
scope of this report, it is important to emphasize that one clear effect of the
lower tariffs resulting from the WTO has been to lower tax revenues. Taxes
on trade are particularly important in the poorest countries. A change in trade
policy does not reduce the need for revenues in those countries. A critical
problem of trade liberalization in many countries has thus been how to recoup
from other sources the tax revenues lost from tariff reform.
74
In addition to the revenue lost directly from a reduction in tariffs, a
developing country may also lose revenue indirectly, at least in the short run,
depending on the impact of the cut in tariffs on domestic economic activity.
On one hand, previously protected producers may lose market share to
imports, thus reducing profits and profits taxes. On the other hand, domestic
businesses may increase their exports, with potentially offsetting effects on
revenues. Such effects are likely to be miniscule, however, compared with the
direct impact on revenues of decreased trade taxes. As a rule, the options
available in the most affected countries for increasing tax revenues are
limited, due to administrative constraints, political constraints, or both. Import
tariffs tend to be relatively easy for most developing countries to administer,
and the administrative machinery to do so is already in place In general, the
only viable solution is usually to impose a broad-based value-added tax,
encompassing both imports and domestic products, or, if such a tax already
exists, to increase its rates.
Tax Treatment of Capital Flows
The case for free trade in goods and services is strong under almost all
economic conditions. Only countries that are incapable of participating
effectively in the world economy are likely to benefit in the long run from
restrictive trade policies. The benefits of freeing capital flows are not quite so
clear, however, and the risks appear to be considerably greater. In particular,
there can be serious risks associated with unrestricted movements of portfolio
capital, as the global financial crisis of the late 1990s made clear. Developing
countries are particularly susceptible to such risks. If a country such as
Argentina loses the confidence of international money managers for some
reason, it may find itself in a financial crisis, much like a bank that faces
default as customers seek to withdraw their deposits all at the same time.
Reflecting the new awareness of the risks accompanying unrestricted
capital flows, the Monterrey Consensus (p. 6) states: “Measures that mitigate
the impact of excessive volatility of short-term capital flows are important and
75
In general, efforts at
restricting capital flows
tend to be
counterproductive
because investors who
believe they may lose
access to their money in a
crisis are likely to be
particularly nervous
whenever they fear that a
crisis is brewing.
must be considered.” The Zedillo Report is more explicit. It states:
Without these flows [of portfolio capital], governments and the local private
sector would not be able to reduce their cost of capital by tapping private
foreign savings. It is for this reason that progressively more developing
countries have been liberalizing their capital account in recent years. But this
has proved to be a mixed blessing. Although the infusion of capital in good
years was quite substantial, in all too many cases the boom years soon gave
way to a bust, marked by currency or banking crises, or both. Countries with
large foreign debts, particularly short-term debts and private sector debts
denominated in foreign currencies, proved vulnerable to crises, as herds of
investors fled in panic.
As a result of this experience, considerable attention has been devoted
to designing appropriate safeguards to ensure that the opening of capital
markets in the developing countries is beneficial. It is not easy to do so. In
general, efforts at restricting capital flows tend to be counterproductive
because investors who believe they may lose access to their money in a crisis
are likely to be particularly nervous whenever they fear that a crisis is
brewing. The tried and true method for reassuring nervous investors is to
make clear to them that their money is available whenever they want it.
Unfortunately, the volume of highly mobile capital in some countries is so
large that the pool of rescue capital needed to reassure the money managers
may exceed the amounts available even to the IMF.
What can developing countries do themselves to cope with these
problems? In addition to more appropriate macroeconomic policies, more
careful phasing in of financial liberalization, and better systems of regulating
financial markets, the Zedillo Report (p. 47) suggests that:
There may be occasions during capital surges when the introduction of
temporary capital inflow taxes proves to be part of the least-bad policy mix.
Chile, for example, avoided much of the economic turmoil that rocked
More generally, a lesson
from the financial crises
of the late 1990s is that a
country is less likely to be
vulnerable to capital
flight if its domestic tax
system is functioning
properly.
Latin America, at least in part because of its policy of requiring domestic
pension funds to invest much of their capital locally and its policy of taxing
foreigners on the income earned in Chile on short-term bonds.
More generally, a lesson from the financial crises of the late 1990s is
that a country is less likely to be vulnerable to capital flight if its domestic tax
system is functioning properly. Borrowing to pay the current expenses of
76
operating the government may be appropriate on occasion for short periods.
As Russia and Argentina learned, however, recurrent recourse to this practice
is generally a prescription for financial instability. A tax system that raises the
money needed for current expanses obviously reduces this risk of instability.
Moreover, the perceived ability of a government to raise and collect additional
taxes in a crisis situation provides additional reassurance to nervous investors
and makes it less likely that such additional taxes will be needed.
In addition, given the risks associated with portfolio investment, it
seems unwise for a developing country to attempt to stimulate such
investment by offering tax incentives to foreign lenders. At the same time,
however, there is no reason for tax policy explicitly to discourage such
investment. In general, therefore, developing countries should attempt to tax
income from portfolio investments, whether earned by foreigners or domestic
investors at reasonable tax rates. Countries that tax income from portfolio
investments invariably utilize a withholding tax mechanism. Withholding is
facilitated if it is applicable without distinction to both domestic and foreign
investors.
In contrast to portfolio investment, as noted earlier, many think that
foreign direct investment (FDI) is largely beneficial. During the 1990s, the
amount of FDI going to developing and emerging countries increased
markedly, as shown in the table below. It has declined some at the start of the
new century, presumably due primarily to an economic slowdown in many
countries, including some of the largest economies.
Increases in FDI were substantial until the financial crises of the late 1990s
and seem to have rebounded well after the recovery (see Table 1). Short-term
capital flows dropped sharply in the late 1990s, and their failure to recover
probably reflects investor caution after the substantial losses that some
investors suffered.
[insert Text Table 1]
77
Table 1: Taxes as a Share of GDP
With and Without Social Insurance Payments
With
Without
With
Social
Social
Social
Social
Insurance
Insurance
Insurance
Insurance
Payments
Payments
Payments
Payments
Africa
Without
Africa
Algeria
39.0
39.0
Kenya
23.3
23.3
Benin
12.7
12.7
Lesotho
27.8
27.8
Botswana
32.5
32.5
Liberia
n.a.
n.a.
Burkina Faso
14.1
14.1
Madagascar
11.3
11.3
Burundi
16.9
15.8
Malawi
14.2
14.2
Cameron
14.5
14.5
Mali
12.9
12.9
Cape Verde
18.7
18.7
Mauritania
15.7
15.7
Central African Rep.
6.1
6.1
Mauritius
17.9
16.9
Chad
6.5
6.5
Morocco
21.5
21.5
Comoros
10.4
10.4
Mozambique
11.4
11.2
.Congo (Dem. Rep.)
42.1
42.1
Nambia
27.6
27.6
Congo (Rep. Of)
26.3
26.3
Niger
Cote d'Ivoire
18.0
16.4
Nigeria
Djibouti
24.5
24.5
Rwanda
9.3
9.3
Equatorial Guinea
16.3
16.3
Sao Tome & Pri
13.0
13.0
Eritrea
19.5
19.5
Senegal
16.0
16.0
Ethiopia
12.9
12.9
Seychelles
35.8
28.2
Gabon
32.8
32.8
South Africa
22.3
21.8
Gambia
15.9
15.9
Sudan
57.2
57.2
Ghana
15.8
15.8
Swaziland
26.3
26.3
Guinea
10.5
10.5
Tanzania
9.6
9.6
9.3
9.3
Guinea-Bissau
7.9
7.9
15.2
15.2
With
Without
With
Social
Social
Social
Social
Insurance
Insurance
Insurance
Insurance
Payments
Payments
Payments
Payments
Europe
Without
Europe
Albania
19.3
15.6
Latvia
25.0
13.6
Austria
37.5
20.4
Lithuania
30.6
21.3
Belarus
40.9
31.9
Luxembourg
42.1
30.4
Belgium
31.0
16.5
Macedonia
32.1
21.7
Bosnia
29.7
29.7
Malta
27.2
20.8
Bulgaria
26.9
19.3
Moldova
22.1
16.4
Croatia
29.3
29.3
Netherlands
63.1
48.9
Czech Republic
31.7
17.3
Norway
34.5
25.0
78
Denmark
49.8
48.3
Poland
29.9
19.4
Estonia
35.4
23.2
Portugal
32.5
21.4
Finland
35.4
22.1
Romania
30.4
21.7
France
45.2
26.8
San Marino
Georgia
11.8
11.8
Germany
26.3
Greece
36.9
Hungary
Iceland
n.a.
n.a.
Slovak
32.1
19.7
11.0
Slovenia
39.4
25.8
25.8
Spain
33.5
21.8
35.6
22.0
Sweden
51.2
51.2
25.7
25.7
Switzerland
22.0
9.5
Ireland
30.6
26.5
Ukraine
34.3
34.3
Italy
44.2
29.4
United King.
32.9
26.9
With
Without
With
Social
Social
Social
Social
Insurance
Insurance
Insurance
Insurance
Payments
Payments
Payments
Payments
Asia
Without
Asia
Armenia
14.3
14.3
Singapore
14.2
14.2
Azerbaijan Rep.
15.5
12.9
Sri Lanka
15.0
15.0
7.0
7.0
Tajikistan
12.5
11.0
Thailand
14.4
14.1
Turkey
24.8
20.6
Turkmenistan
16.2
11.1
United Arab Em.
24.3
24.3
Bangladesh
Bhutan
6.8
6.8
Brunei
19.5
19.5
8.3
8.3
25.7
21.1
Hong Kong
9.2
9.2
Uzbekistan
29.4
29.4
India
8.1
8.1
Vietnam
14.9
14.9
Indonesia
13.3
12.9
Yemen
25.8
25.8
Iran
11.8
9.7
Israel
35.8
29.9
Japan
27.9
17.6
Oceania
Kazakhstan
20.2
16.4
Australia
23
23
Korea
17.6
15.3
Fiji
21.4
21.4
Kyrgyz Rep.
12.3
12.3
Marshall Is.
19.7
19.7
7.2
7.2
Micronesia
3.7
3.7
Macao
14.6
14.6
New Zealand
31.2
31.2
Malaysia
16.0
16.0
Palau
14.7
14.7
Mongolia
18.9
15.7
Papua N. Guinea
20.9
20.9
Myannar
2.1
2.1
Samoa
22.2
22.2
Nepal
8.7
8.7
Solomon
18.1
18.1
Pakistan
12.2
12.2
Tonga
17.6
17.6
Philippines
15.6
15.6
Vanuatu
21.5
21.5
Russian Fed.
27.9
19.7
Cambodia
Cyprus
Lao
79
With
Without
With
Social
Social
Social
Social
Insurance
Insurance
Insurance
Insurance
Payments
Payments
Payments
Payments
North America
Without
South America
Antigua & Barbuda
17.2
17.2
Argentina
17.5
17.5
Aruba
17.7
17.7
Bolivia
21.2
18.9
Bahamas
17.7
17.7
Brazil
19.6
14.5
Barbados
30.3
29.3
Chile
15.9
15.9
Belize
18.9
18.9
Colombia
17.4
17.4
Canada
20.3
15.3
Ecuador
20.1
15
Costa Rica
17.4
12.1
Guyana
27.2
27.2
Dominica
24.8
24.8
Paraguay
13
13
Dominican Rep.
15.3
14.7
Peru
12.5
12.5
El Salvador
10.3
10.3
Uruguay
16.7
16.7
Grenada
23.2
23.2
Venezuela
17.3
17.3
8.8
8.8
Guatemala
Haiti
Honduras
7
7
15.8
15.8
The surprising numbers in Table 1 are those on capital outflows. While
the estimation of these numbers is particularly difficult, it appears that
outflows from developing countries grew so dramatically throughout the
1990s that by 2000 outflows actually exceeded inflows. One depressing
explanation of this marked trend may be (as the World Bank has noted) an
increase in tax fraud, as some owners of capital resident in a developing
country export their capital and then repatriate it in order to benefit from tax
incentives supposedly reserved for foreign investment. Some such “roundtrip” investment may be registered as an inflow but is not registered as an
outflow, appearing as an error or omission instead.
Despite such problems, most developing and transitional countries
remain eager to attract more FDI. In the past, some developing countries were
quite selective in the types of FDI they sought to attract. Investors were
required to seek the approval of an investment authority, which was
disinclined to give approval unless the proposed investment fit within the
country’s development plans. A proposed investment might be rejected for a
variety of reasons, including that it threatened local control over natural
80
resources, that it was likely to promote frivolous consumption, that it was
expected to draw local capital from projects that were considered more
important. This planning approach to FDI is now out of favor, for two good
reasons. It rarely worked well, and it often bred corruption.
The current approach of most countries is instead to let market forces
determine what is or is not a good investment, subject only to limitations for
criminal or socially undesirable investments. This market approach is
inevitable if countries wish to participate fully in the global economy in which
firms, not countries, make investment decisions. The Zedillo Report suggests
that countries should treat foreign investors as well as they treat their domestic
investors, giving them what is commonly called “national treatment.” They
should not, however, be given special treatment. According to the report,
“foreign investors should not be exempted from domestic laws governing
corporate and individual behaviour, nor should the authority of domestic
courts, tribunals, and regulatory authorities over foreign investors and their
enterprises be curtailed.”
More importantly, foreign investors should be required to pay their fair share
of the tax burden.
As the Zedillo report says, the wrong way to attract FDI “is to hand out tax
concessions or erode domestic social or environmental standards in a race to
the bottom.”
Emerging Issues in Cross-Border Taxation
The discussion to this point suggests that globalization has several
general implications for tax policy in developing and transitional countries.
First, trade tax revenues clearly need to be replaced, generally by some form
of broad-based tax on domestic consumption. Second, the importance of
having in place an adequate and sustainable system to mobilize resources for
public purposes is accentuated by the increased need to deal with adjustments
in production structure resulting from trade liberalization. Third, the increased
inequality in income and wealth often accompanying globalization may
similarly accentuate the political and social need for some degree of fiscal
correction – a task, however, that may prove increasingly difficult in the
81
The current approach of
most countries is instead
to let market forces
determine what is not a
good investment.
absence of increased international cooperation to check cross-border evasion.
Fourth, there may perhaps be a (minor) role for some form of tax (or tax-like)
restriction on the volatility of short-term capital flows, and in any case
countries should impose similar withholding taxes on payments from firms to
portfolio investors, whether at home or abroad. Fifth, countries should in
general not give special tax concessions to foreign investors. And finally,
countries will increasingly need to develop policies and administrative skills
to deal with complex international investment transactions – again, a task that
may in many instances require international cooperation and support, as
discussed in chapter 5.
In addition, developing and transitional countries, like all countries,
face several new and troublesome problems in enforcing taxes in a world in
which national borders represent little barrier to commerce. Two such
problems are discussed briefly in this section – the taxation of electronic
commerce and the taxation of new financial instruments.
Taxing electronic commerce
Electronic commerce, or e-commerce, is commerce conducted over the
Internet, a loose network of computers and computer networks that spans the
globe. Such commerce gives rise to problems for both sales and income taxes,
but only the latter is discussed here. As a rule, most countries tax foreigners
on income derived from e-commerce under the same rules applicable to
income derived from other types of activities. Every country requires a foreign
taxpayer to have some minimum link to it before it imposes its income tax. In
taxing the income of a foreign person from e-commerce, a country must
therefore decide whether that person’s electronic presence is sufficient to
establish the minimum link required for taxation.
Most countries also have tax treaties that limit their ability to tax the
residents of their treaty partners. In general, business income, including
income from e-commerce, is taxable only if the foreign taxpayer entitled to
treaty benefits has a permanent establishment (PE) in the taxing country. To
constitute a PE, a taxpayer must have an office or other fixed place of
82
business in that country, and the activities of that fixed place of business must
go beyond certain exploratory or auxiliary activities that enjoy exempt status
under the typical tax treaty.
Issues arise as to whether various e-commerce activities in a country
constitute a fixed place of business for treaty purposes and, if they do, whether
they are within the scope of the exemptions for auxiliary activities. Those
issues have been addressed by the OECD with respect to the following
activities:



Engaging in e-commerce through a web site.
Using computer equipment, such as a server, to store or
provide access to electronic files or otherwise to assist a
taxpayer in conducting its business operations.
Using an agent, such as an independent service provider
(ISP), to obtain access to the Internet or to host a web
site.
The nature of the problem, and the way in which countries are
approaching it can be illustrated by considering briefly the first of these – the
web site issue. Under tax treaties based on the UN or OECD model (see
Appendix B), a foreign taxpayer having a PE in a country is taxable on that
portion of its business income that is attributable to the PE. If a taxpayer’s
web site is treated as a PE, then the sales income derived from sales made
through that web site would by attributable to the PE. If a taxpayer’s server or
other computer equipment is held to constitute a PE, then that portion of the
profits attributable to the operations of the server would be taxable.
A web site that appears on computer screens located in a country
allows a foreign enterprise to interact with customers in ways that are strongly
analogous to the interactions that typically occur through a bricks-and-mortar
office. Indeed, a fully-functioning web site is sometimes referred to as a
“virtual office.” The tax question is whether such a virtual office constitutes
an office or other fixed place of business for purposes of the PE rule in a tax
treaty.
Potential customers can log onto the remote seller's web site, select products
83
for purchase from an online catalogue, and consummate the sale online by
filling out a form and charging the purchase to their personal credit card. If
these electronic sales are viewed as made through a PE, the electronic sales
are taxable in the source country. If a virtual office is not treated as an "office
or other fixed place of business," however, the income derived from the
electronic sales will escape taxation in the source country (and may, for that
matter, escape taxation elsewhere as well).
The OECD (Commentary on Article 5 of the model treaty, as amended
in 2000) takes the position that a web site cannot constitute a PE since a PE
requires a “physical presence” in a country and a web site is intangible and not
physical.1. Along the same line, the OECD’s view is that the computer
equipment must be “fixed” at some location to constitute a PE, and that the
hosting of the web site typically would not cause the ISP to be a PE of its
principal.
With minor exceptions, the OECD approach to e-commerce is to
assign jurisdiction to tax to the residence country and to deny the source
country the right to tax. This position is inimical to the interests of developing
and transitional countries, because those countries are much more likely to be
the source country than the residence country with respect to e-commerce. In
the short term, the income taxation of e-commerce is not an issue of great
revenue significance for most developing or transitional countries (although
the situation in some of the latter may change soon with their impending
accession to the European Union). As yet, few such countries have much ecommerce. The e-commerce that exists, however, is mostly business to
business (B2B) commerce, which is far easier to tax than business to
consumer (B2C) e-commerce, partly because businesses typically keep books
of account and partly because the number of businesses is small relative to the
number of consumers. In the longer term, e-commerce is quite possibly going
to be an important revenue matter for many developing and transitional
countries. The growth of e-commerce over the past decade in some developed
1
Two OECD countries, Portugal and Spain, have registered their disagreement with this
physical presence requirement.
84
countries has been impressive and seems likely to extend to additional
countries in the near future.
Obviously, developing appropriate rules for taxing e-commerce is no
easy task. The appeal of residence taxation is that taxation at source is
difficult, due to the lack of a traditional place of business in the source state.
Residence taxation is also hard, however. Some mix of source and residence
taxation is probably the best approach from the perspective of tax
enforcement. As with taxing cross-border transactions more generally, in the
long run it will be important not only for developing and transitional countries
to develop the capacity to deal with such problems but, equally or more
important, as discussed in Chapter 5, the ability to take a full part in
developing international approaches to such universal issues.
New Financial Instruments
Beginning in the 1980s, financial institutions developed and marketed
a wide array of what are popularly referred to as "new financial instruments."
Many of these instruments were developed to manage risk and to generate
fees for their makers. Over time, however, they have come to be used for a
wide variety of additional purposes, including international tax avoidance.
Hundreds of different types of new financial instruments are now
traded regularly on various over-the-counter markets. Many others have been
tailored to achieve specific objectives of particular customers. The annual
sales of new financial instruments in the over-the-counter markets are
measured in trillions of US dollars. Estimates of the worldwide volume of
sales of all types of new financial instruments are unreliable because many of
the sales are private and unreported. It is safe to say, however, that the volume
of sales is very large and has been growing very rapidly for over a decade.
One important category of new financial instruments is the derivative
instrument. In very general terms, a derivative instrument is a contract that
grants the holder certain rights and imposes certain obligations, with those
rights and obligations determined by reference to the features of some other
85
asset (or collection of assets), such as corn, or a price index, or any other
reference item, such as average rainfall, that can be quantified and is likely to
change over time. The term "derivative" is used because typically the value of
the instrument is derived from a change over time in the value of the other
asset or the amount of the other reference item. More generally, the derivative
instrument increases in value if the rights provided under the contract become
more valuable over the term of the contract than the obligations imposed by
that contract.
An important, though relatively simple, type of derivative is the
futures contract. A futures contract is a financial instrument that is traded on
an organized futures exchange. Performance under the contract is guaranteed
by the exchange. The London International Financial Futures and Options
Exchange (LIFFE) and the Chicago Mercantile Exchange (CME) are two
prominent exchanges that manage the market for a wide variety of futures
contracts. The volume of trades on these exchanges is large and growing. For
example, more than 411 million contracts were traded on the CME in 2001,
up by over 78 percent from the CME's previous record year of 2000. The
underlying value of transactions on the CME in 2001 represented US $294
trillion. On the LIFFE, a record 216 million contracts with an underlying
value of UK £154 trillion were traded in 2001. As a point of comparison, the
gross domestic product of the United States for 2000 was around US $10
trillion.
The typical commodity futures contract entitles the holder to receive
some specified asset, such as corn or gold, on a specified future date in
exchange for a cash payment. That contract is entered into under procedures
established by the commodity futures exchange, and the exchange guarantees
performance of the contract. The person offering to deliver the specified asset
on the delivery date is characterized as the seller of the futures contract. The
person agreeing to make the cash payment in exchange for the promise of
delivery is characterized as the buyer of the futures contract. A person buying
or selling a futures contract on an exchange is required to pay a fee to the
exchange for its services. In general, transactions costs on futures exchanges
86
are relatively low (Box 7).
Box 7 — Futures Contracts
For a highly stylized example of the operation of a futures market, assume
that WCo, a gold watch producer, wishes to protect itself against a rise in the
price of gold. In May, it opens an account with a gold futures exchange and
gives its account manager instructions to acquire a contract giving it the right
to receive 100 troy ounces of gold in December of the current year at the
current market price for gold (the spot price) adjusted for any applicable
financing charges.
Another participant in the gold futures market is GCo. GCo is a producer of
gold and is concerned that the price of gold will decline before it is able to
extract the gold from its mine and bring it to market in December. It is
anxious, therefore, to enter into a contract agreeing to deliver gold in
December that would lock in at the current market price for gold. GCo opens
an account at the gold futures exchange and instructs its account manager to
arrange a contract to sell 100 troy ounces of gold for delivery in December.
Both of the account managers do as instructed, and WCo ends up with a
contract that entitles it to receive the gold and GCo ends up with a contract
entitling it to receive the agreed payment. Under these circumstances, GCo is
said to have sold a gold futures contract and WCo is said to have purchased a
gold futures contract.
Most persons obligated by a futures contracts to make delivery of the asset
covered by the contract do not actually make that delivery. Instead, they
make a cash settlement, paying an amount equal to their financial loss on
their futures contract or receiving an amount equal to their financial gain.
The settlement typically would reflect the difference between the spot price
for the asset at the time the contract was made and the market price for that
asset at the time of settlement. On the typical commodity futures exchange,
over 95 percent of participants make cash settlements of their futures
contract rather than take or make delivery of the asset covered by the
contract.
Derivatives and other new financial instruments are important
mechanisms for managing risk. Managing risk is an economically useful
activity and should not be discouraged by inappropriate tax policies.
Nonetheless, the increasingly widespread use of such instruments has
87
potentially serious tax implications. It has proven very difficult in developed
countries to develop tax rules that can adequately cope with the enormous
flexibility of forms that transactions using financial instruments can take. In
many ways, the simplest approach – though still one with many conceptual
and practical difficulties -- is a “mark-to-market” approach with consistent
timing, treating gains and losses on derivatives as realized in the year they
accrued, but this system is obviously hard to use in developing and
transitional countries with few or no organized trading exchanges. The
alternative approach of reporting gains and losses on a realization basis is
simpler in many ways, but it is also more open to abuse.
Unless prevented, astute tax
planners can use wellstructured financial
instruments to hollow out an
income tax completely – for
example, to shift ownership
rights, to convert dividends
into deductible interest
payments, to change the
source of income, to move
gains offshore, and to bring
losses onshore.
Unless prevented, astute tax planners can use well-structured financial
instruments to hollow out an income tax completely – for example, to shift
ownership rights, to convert dividends into deductible interest payments, to
change the source of income, to move gains offshore, and to bring losses
onshore. The problem is how to retain the obvious benefits in terms of better
risk management while at the same time defending the tax base. One possible
approach may perhaps be to require taxpayers to identify the purpose of
financial instruments at the time they are entered into, thus providing certainty
and the opportunity for fully hedging risks while at the same time eliminating
many forms of potential abuse (through so-called “tax straddles” such as those
listed above, which essentially turn one thing in reality into another thing on
paper). Although this approach is not without its problems (and costs), it may
perhaps offer a way in which developing and transitional countries can reap
some of the benefits of financial innovation while escaping at least the more
egregious fiscal costs.
88
Chapter 4
The Need for International Cooperation
Over the past century, international cooperation on income tax matters
has occurred primarily through the process of developing model tax
conventions (see Appendix B). Tax treaties, however, are just one of many
possible ways that countries can cooperate on tax matters. Indeed, much of the
useful work of the OECD over the past two decades has been in devising
technical solutions to international tax problems outside the treaty context. Its
work on tax treaties was its entrée into the realm of international taxation. It
has become a policy forum to discuss, agree on, and promote various
proposals relating to international taxation. Unfortunately, it is a forum that is
inevitably largely driven by the agendas of the developed countries.
Developing and transitional countries have an important interest in
promoting cooperation among the countries of the world on international tax
matters. Such cooperation provides them with an important way to offset
some of the loss of sovereign power to tax that accompanies globalization. It
may also provide them with a better chance to deal more effectively with
problems such as international tax evasion, transfer price abuses, and capital
flight. To promote cooperative action, the developing countries need the
support of an international institution that is responsive to their concerns.
The Report of the Secretary-General to the Preparatory Committee for
the High-level International Intergovernmental Event on Financing for
Development, dated 12-23 February 2001, recommended the establishment of
a global tax organization. The relevant portion of that report is set forth below:
There is a growing need to improve arrangements for
cooperation between national tax authorities. Increasing
international economic and financial interdependence is
89
constraining national capacity to set and enforce various tax
instruments. Governments are increasingly limited by
international competition in both the forms of tax and the tax
rates they can apply. Improved international cooperation
between taxing authorities would serve, inter alia, to reduce
opportunities for tax evasion and avoidance, contribute to
mitigating the capital-flow instability to which developing
countries are sometimes subject, and deploy tax incentives and
disincentives in support of public goods, such as avoiding
depletion of the global commons.
The achievement of these goals requires major improvements in
international cooperation on taxation matters. Forums exist in limited
membership organizations to treat these issues from the viewpoints of their
members, in particular OECD. In addition, taxation is addressed at the level of
experts in United Nations forums, notably the Ad Hoc Group of Experts on
International Cooperation in Tax Matters and certain expert groups on
accounting and other related matters that are convened by UNCTAD. But,
although OECD, for instance, has undertaken a number of outreach activities
with non-member countries, there is today no global intergovernmental forum
that considers tax questions on an ongoing basis or that can adequately put the
tax debate in a wider — including a developmental — context.
To fill this gap, an international organization for cooperation in
tax matters could merge the various international tax-related
efforts into a single entity. Such a broad-based international
organization could provide a global forum for the discussion of
and cooperation in tax matters, including the sharing of
national taxation experiences; the development of definitions,
standards and norms for tax policy, administration and related
matters; the identification of national tax trends and problems;
tax reporting; and the provision of technical assistance to
national tax authorities, particularly those of developing and
transition economy countries. Other less ambitious proposals
have also been put forward, including the strengthening of the
Ad Hoc Group of Experts on International Cooperation in Tax
Matters.
Subsequently, the High-level Panel on Financing for Development (the
“Zedillo Report”) recommended that the international community consider the
90
establishment of a world tax organization that would perform various
functions of importance to developing countries. According to the Zedillo
Report, such a new international tax body should “take a lead role in
restraining the tax competition designed to attract multinationals —
competition that . . . often results in the lion’s share of the benefits of [foreign
direct investment] accruing to the foreign investor.” Another long-term task
for the new body would be to develop a system of “unitary taxation of
multinationals” as a replacement for the arm’s length-separate entity approach
favored by the OECD.
In contrast, the Monterrey Consensus document, published at the
conclusion of the International Conference on Financing for Development, put
forth only the following recommendation for action:
Strengthen international tax cooperation, through enhanced
dialogue among national tax authorities and greater coordination of
the work of the concerned multinational bodies and relevant
regional organizations, giving special attention to the needs of
developing countries and countries with economies in transition.
The idea of any sort of comprehensive international tax organization (ITO)
thus seems to be off the agenda for the time being. Nonetheless, the Zedillo
Report is certainly correct in its assertion that the existing institutional
arrangements for dealing with international tax matters are unsatisfactory
from the perspective of the developing and transitional countries. So the
question seems to be, what can be done short of an ITO?
At the minimum, to be able to cope with the great fiscal pressures
facing them both domestically and in the global context, developing and
transitional countries appear to need some sort of center on international
taxation that had competent and adequate staffing and that is responsive to
their agendas. A possibly promising approach may be to build upon the
existing program already in place in the United Nations, namely the Ad Hoc
Group of Experts on International Cooperation in Tax Matters. That body has
done important work over the past two decades in developing and promoting
the UN model convention. It has also conducted a few workshops to assist in
91
At the minimum, to be
able to cope with the
great fiscal pressures
facing them both
domestically and in the
global context,
developing and
transitional countries
appear to need some
sort of center on
international taxation
that had competent and
adequate staffing and
that is responsive to
their agendas.
the training of tax officials in developing countries. It is woefully
underfunded, however, and does not have the institutional structure necessary
to project and defend its work adequately in the international context.
Significant restructuring would be needed for it to serve the needs of
developing and transitional countries for assistance on international tax
matters.
Should a revitalized UN center for international taxation be created, it
should have a small staff of full-time tax professionals, ideally drawn from the
accounting, economics, and legal professions. It should also have adequate
support staff to allow it to do its job properly and to stage conferences and
workshops, as well as an adequate travel budget for meetings of
representatives of developing countries and to permit center professionals to
represent the developing countries at conferences and other activities
sponsored by the OECD and other international organizations involved in
international tax matters. The center need not be large. A half dozen tax
professionals plus a support staff and a director should be enough. What is
important is competence, not size.
The main focus of the Ad Hoc Group has been on tax treaties. The
suggested center should expand the focus, just as the OECD has expanded its
focus over the past decade beyond the treaty area. The center should of course
cooperate with other international tax organizations whenever possible.
The OECD, the World Bank, and the International Monetary Fund are
not enemies of the developing and emerging countries. Within the limits of
their tasks and institutional structure, they have sought to advance various
goals of importance to many developing and emerging countries and have
done much to help such countries to deal with the serious fiscal problems they
confront. Nonetheless, these organizations do not necessarily always focus on
the problems of greatest concern to developing and transitional countries in
the international tax area.
As currently constituted, the Ad Hoc Group of Experts on
International Cooperation in Tax Matters has 10 members from developed
92
countries and 15 members from developing and emerging countries. The
developed countries were initially included in the Ad Hoc Group to provide
some technical assistance and to ensure that recommendations of the group
would have some realistic hope of being incorporated into tax treaties between
developed and developing countries. For purposes of drafting the initial UN
model convention, this arrangement appears to have worked quite well.
In recent years, however, the arrangement has become less
satisfactory. Many developed countries have appointed members to the Ad
Hoc Group who also serve as representatives to the OECD. Understandably,
perhaps, such representatives are sometimes inclined to promote the position
of the OECD within the Ad Hoc Group, even when that position seems
inimical to the interests of the developing countries. One way to reduce such
problems might be to give representatives from developed countries the status
of observers. Another might be to adopt some alternative system for selecting
representatives from developed countries. For example, the representative
might be appointed by the United Nations rather than by the developed
countries themselves. Such institutional matters will obviously require much
more thought, and discussion, before they can be resolved. But what seems
undeniable is that if the developing and transitional countries as a group are to
be able to mobilize domestic resources effectively in the world in which we
now live, their voice will have to be more effectively heard in international tax
matters than it now is.
93
Chapter 5
Recommendations for Good Tax Policy in
Today’s Global Economy
Globalization limits the sovereign power of countries to tax as it opens
national borders to greater flows of goods, services, and financial capital.
Recent technological advances, including digitization, high speed computers,
the Internet and others, have facilitated globalization and created an additional
set of limitations on government by lowering transaction and transportation
costs and making it more difficult to track economic, financial, and legal
behaviours. Nonetheless, the effects of globalization and technological
changes have not been to alter the basic parameters of good taxation, which
are much as they have been. These forces, however, have raised the penalties
for bad tax policy and have increased the rewards for good policy. The reason
is simple — the incentives created by taxation cause stronger and more rapid
responses than ever before, as these same forces allow taxpayers, investors,
and businesses to react to tax differentials and other incentives at lower cost
and with greater flexibility. Cooperative arrangements among countries,
including greater harmonization of taxes and greater assistance is assessment
and collection, can lessen these incentives and can limit the capacity of
taxpayers to avoid and evade taxes. Nevertheless, inducements for a small
number of countries to cheat or not participate in such cooperative
arrangements are always present and are rising with globalization and with
the introduction of new technologies for international communication.
As is abundantly clear from the discussion throughout the prior
chapters, no single tax structure can serve the needs of every country.
Globalization has not changed this basic fact of life. The appropriate tax
system for every country depends, among other factors, on its economic
structure, capacity to administer taxes, public service needs, government
94
structure and, perhaps most of all, to the weights attached to the different
goals of taxation. Thus, differences should be anticipated in the specific
elements of each nation’s tax structure. Countries also can be expected to vary
their level of taxation – data analysis provided here suggests that a lower
bound exists in the tax ratio used by countries if they are to finance the most
essential public services. Beyond this minimum, most countries have
considerable capacity to alter the level of taxation to meet differing local
demands for public services and desires for redistribution, as illustrated by the
wide variation in actual tax practices among countries at comparable levels of
development.
Countries must accommodate their differing goals and views towards
taxation within a series of guidelines and general principles that arise from
economic forces, legal constraints, globalization, and other factors if the
undesirably, sometimes perverse, consequences of taxation are to be kept to a
minimum and if government is to be adequately financed. Specific
conclusions on the best forms of taxation are summarized in the remainder of
this chapter. Further discussion on each of these points can be found in the
prior chapters.

Consumption tax bases, including the value added tax base,
should be structured as broadly as possible, while recognizing
that factors such as administrative capacity impose certain
limitations on what can be taxed. Broader tax structures limit
the distorting effects of taxation on choices (such as what and
how much to purchase and whether to work) and the incentives
on where and what to produce. Broader tax bases permit tax
rates to be held lower for any given amount of tax revenue that
is to be generated, and lower rates further lessen the perverse
incentives created by taxation. Other benefits resulting from
broader tax bases include that they normally result in a higher
revenue growth rate and usually allow for less regressive
consumption tax structures. Countries must set tax policy in
light of limitations on their ability to administer taxes on
certain segments of the economy. However, many of the
95
constraints on the use of broader-based taxes can be modified
by the government (endogenous factors). For example, the
government can engage in better fiscal planning, it can
improved the quality of its tax legislation, and it can take
constructive steps to development the tax administration.

Excise taxes frequently serve as an important source of tax
revenue for the least developed countries — a revenue source
that is not easily replaced. As with all narrow-based taxes,
those taxes create some inefficiencies and fairness problems.
The inefficiencies can be mitigated if the excise taxes are
levied on small set of commodities that are relatively
unresponsive to tax rates. Fairness is enhanced by avoiding the
imposition of substantial excise taxes on items essential to the
subsistence needs of the poor. Administrative costs can be
mitigated by imposing the excise taxes on commodities
produced or distributed by a small number of vendors. Broader
based tax regimes should become increasingly more important
as a country matures. Excise taxes continue to raise significant
revenues in developing and transitional countries even as their
development increases, although their relative importance
generally is reduced as they are able to administer broader tax
bases.

Trade taxation should be reduced in the context of bilateral and
multilateral agreements to lower the barriers on trade.
Revenues lost from reduced trade taxes generally should be
replaced with higher indirect taxes. A shift from trade taxes to
broad-based consumption taxes allows countries to replace
differential, high-rate levies on imports with lower-rate, neutral
taxation of domestic and imported commodities. The result will
be more neutral taxation and a more open economy in which
inefficient businesses are not protected and which should
improve the capacity of the country to produce, consume, and
96
export.

Whenever feasible from an administrative perspective,
personal income taxation should be one component of nearly
every country’s tax structure. A personal income tax has the
potential for achieving some redistribution of income through
the tax system, and it has the administrative advantage of
providing for the collect of revenues at the source. The revenue
potential from income taxation is relatively low for the least
developed countries, but their capacity to collect revenues from
that source should expand as the country develops.

Corporate income taxation should be a component of the tax
system of almost every developing and transitional country, to
allow, among other things, for taxation of foreign earnings at
the source and to prevent individuals from avoiding taxes
under the personal income tax by hiding assets within a
corporation. Corporate taxes should be broadly based, evenly
levied on domestic and foreign corporations, and imposed at
modest rates (which are in line with personal income tax rates).
Tax concessions should be used very sparingly and only as part
of very targeted economic growth strategies. In particular,
countries should avoid negotiated tax concessions that are not
open and transparent.

Tax rates must be kept moderate. To raise the revenue needed
to finance public expenditures, a system with low rates needs
to have broad tax bases and a balanced mix of tax instruments.
The perverse consequences of taxes rise very rapidly with tax
rates. Incentives for evasion and avoidance also rise with tax
rates, creating significant costs on the domestic economy.

Tax structures, including the choice of tax instruments and the
specific tax bases and rates imposed, should be selected to
97
provide overall revenue growth that is consistent with a
country’s service delivery responsibilities and its need to
finance development expenditures. If properly designed, the
revenue growth will occur without the need for frequent rate
increases and other significant changes in tax legislation. Small
modifications in tax structures will always be required to offset
unintended consequences and opportunities for evasion and
avoidance. Adjustments also will be necessary from time to
time in response to an evolving economic structure. But major
policy changes because of poor revenue growth should and can
be avoided by designing the tax structure to achieve the
appropriate revenue growth

Taxation should extend into non-monetized and traditional
sectors, and specifically into the agricultural sector. Taxation of
agricultural land and presumptive taxation are among the
effective means to levy such taxes. The VAT, if properly
implemented, can also help in taxing these sectors. Failure to
tax these sectors results in tax subsidies to traditional economic
sectors and discourages movement into the modern sector of
the economy. Long-run economic growth will be advantaged
by more even taxation of traditional and modern sectors of the
economy.

A focus on good administration is imperative to achieve the
expected revenue performance, neutrality, and fairness from
the tax system. Good administration requires an appropriately
structured tax administration that is staffed with adequately
trained officials and has the resources to operate efficiently.
Perhaps most important of all, the tax administration must have
the political support needed to collect the taxes that have been
imposed in an even, transparent, and fair manner.

From an administrative and compliance perspective, many
98
taxes are most efficiently collected at the point of production.
Nonetheless, such taxes must be structured very carefully
because taxes on production have the potential to create many
additional perverse effects including on where businesses
locate, the ways production occurs and the location of
production. These effects are nearly always adverse to
economic growth.

Developing countries should promote cooperation on
international tax matters through a variety of mechanisms,
including the establishment of international organizations that
can represent their interests in various tax discussions around
the world. Developed countries are normally well represented
by organizations such as OECD, but developing countries have
no coordinated front in many international tax discussions and
negotiations. Effective cooperation among developing and
transitional countries and with the developed countries can
extend the taxing options available to all countries.
99
Appendix A. Taxation and Decentralization
The problems of taxation that developing and transitional governments
face when joining the world economy have been faced for decades by subnational governments throughout the world. A sub-national government
typically relinquishes control over its borders to create a free trade zone within
the nation state of which it forms a part. In particular, sub-national
governments have difficulty raising tax revenue from mobile tax bases, such
as personal income and corporate income, and they even confront difficulties
imposing local sales taxes, due to opportunities that their residents often have
for making tax-free purchases outside their jurisdiction. These problems have
been made worse by the globalization of their national economy, for now the
sub-national governments are buffeted not only by internal economic
pressures but by external ones as well.
The solutions to the tax problems that sub-national governments
confront are similar in many respects to the solutions recommended in this
report for nation states. In general, the sub-national governments need to
recapture the sovereign power they gave up (or never possessed) through
cooperative measures. They also need to rationalize their tax policies and
improve their tax administrations.
Fortunately, a mechanism for achieving the needed cooperation among
sub-national governments is already in place --- namely, the national
government. In all too many cases, however, national governments have failed
to provide the framework for cooperation that is needed for sub-national
governments to obtain the revenues needed to provide local services and other
accouterments of government. In this section, we make suggestions on how
that framework should be structured.
The Trend towards Decentralization
One might expect that the problems associated with globalization would
have created pressures for centralizing power at the national government and
for reducing the importance of sub-national governments. The trend, however,
is in exactly the opposite direction. All over the world, countries are moving,
100
sometimes quite rapidly, to decentralize governmental functions. This
movement is taking place notwithstanding the real problems that globalization
creates for all governments and particularly for small ones. Given this trend
toward decentralization, it has become increasingly important that the tax
problems of sub-national governments be taken seriously by policy makers
and by tax analysts.
The role of decentralized governments in the delivery of public services
and the collection and administration of taxes differs widely around the world.
Countries such as Belarus, Mongolia, the Netherlands, and South Africa rely
very heavily on local governments, which are responsible for 40 percent or
more of combined national and sub-national expenditures. Other countries,
such as Costa Rica, Jordan, Lebanon, and Panama rely much less on local
governments, with the sub-national share of expenditures often comprising
less than 5 percent of combined total spending.
The reasons for decentralization vary widely but include efforts to
strengthen democracy, enhance government accountability, and empower
citizens. The expected benefits of decentralization also vary. Some countries
anticipate that decentralization will lead to better service delivery because the
diverse demands and needs of the population can be served more effectively
by local officials, who are better informed about what people want. Local
officials often can be held more accountable than national officials by the
local population, which has greater access to the local mayor than to members
of the national legislature. Decentralization may be seen as a means of
accommodating the varying interests of different ethnic groups by reducing
the potential number of friction points, thus helping to expand nation building.
There may also be dis-economies of delivering some services at the national
or even regional level, meaning that local service delivery can be less
expensive.
Fiscal Independence
Decentralization can only be successful when national and sub-national
governments have established a framework in which sub-national
101
governments can flourish. Perhaps the most vital single requirement is that
sub-national governments be financially sustainable. The Committee of
Experts on Public Administration noted that the “conditions for
decentralization’s success include national Governments to prompt, initiate,
monitor and guide the process of political and financial decentralization, as
well as the creation of an adequate financial decentralization, and the creation
of an adequate financial resource base for local governments to function
properly.” The national government must create an environment in which subnational governments can succeed, and the sub-national governments must
have sufficient resources to succeed.
This appendix focuses on one aspect of an effective decentralized
governmental structure — access to the necessary resources. An adequate
financial structure is an imperative if sub-national governments are to deliver
the services that people demand. Sub-national governments including both
intermediate governments, such as states, provinces and counties (regional
governments), and the lowest level of government, such as cities,
municipalities and villages (local governments). As noted above, sub-national
governments cannot command the resources they need to fulfill their
obligations without major support from the national government. In many
cases, that support should come from direct transfers of revenues collected at
the national level and from institutional arrangements, such as tax-base
sharing, that allow sub-national governments to raise tax revenues on their
own. Local governments also should utilize user fees for services or taxes that
are nearly equivalent to user fees when those approaches prove feasible.
In almost every developing and transitional country, transfers from the
national government must play an important role in the creation of an
adequate local financial base. The national government has a much greater
capacity than local governments to impose and collect many taxes, including
the VAT and income taxes. This greater capacity to tax makes it logical for
the national government to generate a significant share of tax revenues and
then to shift revenues to the government responsible for service delivery.
Transfers from the national government are also a key means of reducing
102
disparity in access to resources. The national government can help equalize
access to key services, such as primary education and infrastructure, by
transferring more revenues to those local governments that have the least
capacity to raise revenues on their own or the greatest need to deliver key
services.
Local, regional, and provincial governments, however, are not as
accountable or transparent when they use national resources to finance their
activities as when they use resources they obtain under their own taxing
powers. The local population generally has little information about the volume
of resources available to a local government when all the money is provided
in the form of transfers. As a result, the voting public has little ability to
require officials to use the money well. Local government officials often have
no control over expenditure levels when financing comes from the center. As
a result, local officials would not be able to work with their populations to
ensure that the best set of services is delivered even if they felt political
pressures to do so. If decentralization is to work well, therefore, sub-national
governments must have the ability to impose and collect significant taxes on
their own authority. If those sub-national taxes are properly designed and are
implemented effectively, they may lead not only to an increase in total
resource mobilization but also to improved utilization of those resources.
As noted above, the economic competition associated with
globalization may have a detrimental effect on the capacity of sub-national
governments to generate revenues. Collecting tax revenues is more difficult at
the local or regional level because much economic activity can flee or can be
hidden easily. For example, businesses or individuals can evade or avoid
income taxes by receiving income outside their home jurisdiction, or they can
purchase goods in another community and thereby evade or avoid local sales
taxes.
Methods of Decentralizing Control over Tax Revenue
The extent of tax decentralization can be determined in terms of the
degree of local or regional control over four factors:
103




tax revenues received;
choice of tax rates;
definition of tax bases, and
tax administration.
Control over revenues at the local or regional level obviously is necessary to
decentralize government. To achieve genuine fiscal decentralization, however,
local and regional governments must have control over some aspects of the
process of imposing and collecting taxes.
Local Control of Tax Rates. From the perspective of economic and
political accountability, the most essential elements of fiscal decentralization
are that local and regional governments should be allowed to determine the
rates of taxes for which they are politically responsible and that these taxes
should be sufficiently important in terms of the revenues generated to ensure
that the setting of rates affects expenditures in noticeable ways. Control over
the tax rate gives local and regional officials the ability to influence the level
of services provided (which is necessary if government is to be devolved), and
it makes them more accountable to their citizens, who are better able to see
the amount of revenue available to the local and regional governments.
Control over rates can come in several forms. If the sub-national
government is fully responsible for the administration of a tax, as it might be
in the case of a property tax, then it could be given full authority over the
rates. A national requirement of some minimum rates, however, might
actually enhance control at the local level, as it lessens the problems of tax
competition that otherwise would come into play. That minimum rate,
although established under national legislation, ought to be determined
through consultations with local officials. A maximum rate also might be
appropriate in some cases. Generally, competitive forces can be relied upon to
prevent unreasonably high rates. A cap on rates is useful in some cases to
prevent local jurisdictions from attempting to capture an unreasonable share of
the potential tax revenues from certain captive taxpayers. For example, a local
government should not be permitted to impose excessive taxes on the property
of an extractive industry located within its borders because the potential
104
revenues from that industry ought to be shared more broadly among other subnational governments or with the national government.
Local Control over Tax Bases. Providing for local control over tax
bases and tax administration are more problematic. Local control over tax
bases is inappropriate in many cases because administration and compliance
costs are often lower if the same tax bases exist across the country. A
consistent base means that businesses operating across the country are not
subjected to multiple tax structures. There are also lower administrative costs
if taxes are administered at the national level.
The theoretical efficiencies of national control over the definition of the
tax bases may be lost, however, through the normal operation of politics. All
else being equal, a political leader seeking support of the populace would
prefer to reduce taxes rather than to raise them, because taxes are always
unpopular with those who have to pay them. The chief political reason for
raising taxes is to be able to provide the public services that the populace
wants and expects from the government. If the political leaders at the national
level of government have the power to determine local tax revenues collected
by local governments by defining the local tax base broadly or narrowly, they
have a strong political incentive to go for the narrow tax base. By doing so,
they suffer no revenue loss at their level of government, yet they are able to
satisfy the desire of their constituents for lower taxes. This problem is one that
has arisen in many countries. The situation in Bosnia provides a good current
example. And the problem can be expected to arise whenever the detriments
and benefits of taxation are artificially divided.
The problem described above has no full solution. A compromise
solution can be found, however, through institutional arrangements that
require a national or uniform definition of the tax base and that give local
governments substantial input into the content of that tax base. Solutions of
this type have been worked out in countries as different as Australia and
Papua New Guinea. One workable technique is for the national government,
in consultation with the affected sub-national governments, to work out an
105
acceptable definition of the tax base and then to require substantial local input,
perhaps even with veto power, for any changes to be made in the tax base.
Suitable arrangements also would be needed to allow for minor adjustments in
the tax base to block the tax-avoidance schemes that are an almost inevitable
response to taxation.
Local Control over Tax Administration. Local administration of some
taxes, such as the property tax, may work well. Even with the property tax,
however, some national or regional assistance might be highly desirable. A
national office, for example, may be better equipped to value certain types of
property, especially property used for industrial purposes. It also might assist
in the drafting of a competent taxing statute. In addition, a regional office may
be better able to do cadastral surveys in the region to ensure that all property
of importance to local governments in the region has been entered onto the tax
rolls.
For income taxes, including the personal income tax and the business
profits tax, and for broad-based sales taxes, such as a VAT, local
administration is almost certainly not the best means of collecting revenues.
National collection of local and regional taxes may be better than establishing
numerous small and probably inefficient local collection agencies. Companies
operating throughout the country are better controlled by a national
administration that can monitor all of their domestic business activities. It is
easier for a national administration, operating on a large scale, to hire
specialized auditors and to develop a computerized information system.
National tax administration also makes it convenient to transfer revenue
agents from time to time in order to limit the likelihood of tax fraud.
Separating the collection of taxes from the level of government entitled
to the revenues, however, presents problems similar to the problems,
discussed above, that arise when the national government defines the tax base
and the local governments receive the revenues generated from that tax base.
In general, a national tax administration may not collect local taxes effectively
when it has no real stake in the results of the collection process. The national
106
tax administration obviously has fewer incentives to collect local taxes and to
respond to local officials than it does to collect national taxes and answer to
the legislature and to the national government. Hence the national
administration may put little effort into ensuring that local taxes are collected
effectively.
If the national administration is to collect local taxes, the legislature and
the tax authorities must introduce good incentives for effective collection. One
mechanism that would provide some limited benefits would be for the
national tax officials engaged in the collection of local taxes to be paid by the
local governments, either directly or through charges made to the local
governments for services rendered. The local governments then have the
ability to withhold payment if the services were inadequate. Because the costs
of collection are typically a small part of the total revenue collected, however,
a charge for services cannot be expected to overcome fully the political
dynamics favoring non-collection at the national level.
Another mechanism that has worked well in many countries is for both
the national and the sub-national governments to share the revenue from a
particular tax. For example, the national government might collect its own
VAT and also a VAT imposed by sub-national governments as a surcharge on
the same tax base. Similarly, the national government might allow the subnational governments to impose a surcharge on a national income tax, either
the personal income tax or the corporate profits tax. The national tax
administration would have a strong incentive to collect the sub-national tax
properly because its failure to do so would have serious effects on its own
revenues.
The surcharge system need not be restricted to national taxes. Local
governments might make arrangements for a regional government to jointly
impose a property tax, with the regional government having the primary
responsibility for valuing property and making assessments. The local
government, however, might have responsibility for collection of the tax. Both
the regional government and the local government would share in the tax
107
revenues, with each government’s share of the revenue determined in advance
and communicated to taxpayers.
The fundamental problem with shared collection arrangements is that
the local government may end up losing control not only over the collection
process but also over the rates and the definition of the tax bases. To allow for
control over the rates, the sub-national governments should be permitted,
perhaps within some specified range, to determine the amount of their
surcharge on the national tax. As suggested above, the local and regional
governments that are sharing a tax base with the national government also
should have serious input into the design of the tax base and should have some
authority to prevent revenue-losing alterations in the tax base.
A system of a national tax with a sub-national surcharge has some
superficial similarities to the tax-sharing (revenue sharing) system that is
found in almost all transitional countries and in a number of developing
countries as well. Conceptually and practically, however, the two systems are
totally different. Tax sharing is no more than a disguised intergovernmental
transfer of resources, with the recipient governments bearing no responsibility
at all for the tax rates or the amounts they receive. Under the surcharge
system, however, local and regional governments can establish the level of
their surcharge on the national tax, and the sub-national share of the tax would
be identified specifically to taxpayers when the tax is paid. In addition, the tax
would be paid over directly to the relevant sub-national government, without
the need for an appropriation of the amount by the national government.
In general, the administrative and compliance costs associated with a
surcharge system are likely to be low. Administrative costs should be low
because the same tax officials would collect the revenue from the same
taxpayers, and the amount of the tax would be computed with respect to the
same tax base. In the case of a personal income tax, the surcharge could be
remitted, often by an employer, to the national government, which then would
transmit the remittance to the sub-national government levying the surcharge.
Similarly, the vendors responsible for collecting a VAT would remit both the
108
national portion of the tax and the surcharge to the national government.
One complication that would arise in administering a surcharge is that
the national government would need to collect the information needed to
determine which sub-national government should obtain the revenue.
Assuming the rules for dispensing the revenue are fairly simple, the
administrative costs of collecting the necessary information should be modest.
In the case of a sales tax, the revenues presumably would be shared in relation
to the sales made in each sub-national jurisdiction and the rates of the
surcharge imposed by each sub-national jurisdiction. The proper rules for
distributing the revenue raised from a surcharge on the personal income tax
are not self-evident. Perhaps the least complex rule would be distribute the
revenue in relation to the residence of the taxpayer. An alternative rule would
focus on the source of the income taxed. That rule works easily enough if the
tax is essential a wage tax, but it does not work well if the base of the tax is
defined more broadly.
More complex problems would arise in apportioning a surcharge on a
corporate profits tax. The most appealing method of apportionment would be
to link apportionment to the source of the business profits. Making such a
linkage, however, is exceedingly difficult unless done though some type of
apportionment formula. Some developed countries, including Canada,
Switzerland, and the United States, use an apportionment formula to
determine the share of the corporate tax base attributable to their sub-national
governments. (McIntyre, 2003) The use of an apportionment formula is
workable, but no one should suggest that it does not entail serious
administrative and compliance costs. To avoid those costs, most developing
and transitional countries might wish to make the corporate profits tax
exclusively a national tax.
Selecting the Tax Mix
The design of a good sub-national tax system is even more country
specific than the design of a good tax administration or a good national tax
system. Taxation is a local matter, and local taxation is even more a local
109
matter. Still, some useful generalities are possible. We suggest the following
guidelines in designing sub-national taxes.

Sub-national governments, and particularly local governments,
should adopt user fees to finance the serves they provide
whenever possible. When subsidies for the poor are
appropriate, they should be provided directly through a subsidy
program rather than indirectly through a lower charge for
services.

Developing and transitional countries should make better use
of property taxes to fund local and regional services. The
property tax, which often has effects analogous to those of user
fees, is an especially useful tax for local governments.

If regional governments are to be given significant
responsibilities for providing nationally critical services, such
as education and health care, they need access to a major tax
base, such as a personal income tax or a VAT. As discussed
above, that access could be provided by allowing the regional
government to impose a surcharge on the national tax.

Different degrees of decentralization are appropriate for
different taxes (Box 7).
Service Fees. In almost all countries, local governments can and should
make greater use of user charges. Indeed, user fees should be imposed
wherever possible to finance services. They can normally be imposed
effectively when the recipients of services are easily identifiable and receipt of
the service can be separately identified for individuals or households. Obvious
examples include utility services, which should be fully priced, and any
subsidies desired for equity reasons made explicit as public expenditures.
Services such as utilities should be considered for privatization as well.
Municipal transit and some health services are other obvious examples where
user fees can play an important role.
Property Taxes. Local property taxes also should provide a significant
revenue source. In many cases, a simple tax on real property (immoveable
110
property) can be seen as the economic equivalent of a user fee because the
value of real property often is enhanced by the services that it finances. For
example, police and fire protection, repair of local roads, and expenditures for
local public schools can enhance property values in rough proportion to the
local taxes paid. The relationship of a property tax to a service fee is
obviously quite imprecise even in the best of cases, but the relationship is
often sufficiently close to mitigate problems of tax competition.
As noted above, national or regional governments may play an
important role in setting up a standard property tax law and in undertaking the
technical tasks of valuing property and training staff. The collection and
enforcement of the property tax, however, can be accomplished at the local
level. Assuming that the revenues are to remain at the local level, the tax rate
definitely should be determined at the local level, albeit within some specified
range that could be established by either the national or the regional
government. Few developing or transitional countries currently give local
governments sufficient discretion or responsibility with respect to property
taxes. For this reason and perhaps for others, these taxes are substantially
under-utilized in most of these countries (Box 8).
111
Box 8 — Assignment of Taxes to Levels of
Government
Tax type
Tax
Tax Base
Tax Rate
Administration
Revenue
Ownership
1. Custom
National
National
2. Excises
National or regional National or regional National or regional National or regional
3. Wage and
National, regional
Personal Income
or local
National
National
National
National, regional
National, regional
or local
or local
4. Corporate Income National
National
National
National or shared
5. VAT
National
National
National*
National or shared
6. Sales Tax
Regional or local
Regional or local
Regional or local
Regional and/or local
7. Propert tax
Local
National or regional Local
Local
8. Real Estate
Local
Local
Local
Local
9. Motor Vehicles
Local
Local
Local
Local
10. Social security
National
National
National
National
11. User Fees
Service providing
Service providing
Service providing
Service providing
Social Services
agencies
agencies
agencies
agencies
12. User Fees
Service supply
Regulatory
Service providing
Service providing
Privatized Utilities
companies
frameworks
companies subject
companies
Transfers
Contributions
to regulatory
oversight
*Regional governments may be given the authority to apply a surcharge to the national tax
within a specified band of rates
112
Degrees of Decentralization. The extent to which a tax is
decentralization depends in large part on the inherent characteristics of the tax
(Box 9). Property taxes and motor vehicle taxes may be decentralized almost
completely because they can be administered effectively at the local level and
they do not create major problems of tax competition. As noted above,
however, some nationally imposed minimum rate to limit tax competition and
a maximum rate to prevent local exploitation are desirable. Also desirable is
the provision of regional or national technical assistance in setting up the taxes
and in dealing with certain technical matters, such as valuation of property.
Box 9 — Property Taxes
Property tax revenues are low in most developing and transitional
economies. The coverage of the tax is not comprehensive, assessments
are low, and collection rates are typically low. Nominal tax rates are also
low, and governments usually find that rate increases in this very visible
tax are difficult to sell politically. Simply raising the legal tax rate would
in any case usually burden only the few from whom taxes are actually
collected. Increased nominal rates are likely to be acceptable only if
accompanied by improvements in tax administration such as more
comprehensive coverage, better assessments, more frequent assessment
re-valuations, and enforced penalties for late payment. In general,
revenues are higher if the property tax is based on the value of land and
buildings (instead of just on land), if there are few exemptions and no
favourable treatment of particular property classes, and if the scope for
local tax competition are limited.
Despite its problems, the property tax remains the predominant option for
raising revenues at the local government level. The potential yield of
land and property taxes is unlikely to be huge, revenues from this source
will not be very elastic, and the administrative costs typically are
substantial. Nonetheless, all local taxes are difficult to impose. An
expanded local property tax is both a logical and a desirable objective for
many countries, particularly those in which local governments are
expected to play an increasing role in allocating public sector resources.
113
Some taxes, particularly the corporate profits tax, cannot be
administered at the local level and can be administered at the regional level
only with substantial assistance from the national government. As noted
above, a regional corporate profits tax is possible as a surcharge on a national
tax. Even in that case, however, difficult problems of apportioning the revenue
from the tax would arise. With few exceptions, the preferred route in
developing and transitional countries is to make the corporate profits tax an
exclusive national tax. If the revenue of that tax needs to be shared to make
the sub-national governments financially sustainable, then it should be done
through some tax-sharing or grant arrangement rather than through the
devolution of the taxing power to the sub-national level.
Sales taxes, payroll taxes, and even personal income taxes can be
decentralized, at least in part, through the surcharge mechanism. The
surcharge mechanism is likely to work well when applied by a regional
government but less well at the local level. At the local level, the problem of
how to allocate the revenue from the tax becomes difficult, especially in the
case of the personal income, which presumably would be apportioned on a
residence basis. A local surcharge on a sales tax probably is workable as long
as the local surcharge rate is low. If the rate is significant, problems of tax
competition are likely to be significant. All of the above remarks need to be
taken as broad generalities that may not be applicable in some countries. For
example, a large country with a small number of regional governments is
likely to have far less problems of tax competition under a sales tax than a
small country with many regional governments.
The degree to which a national government should share its taxing
powers with sub-national government depends significantly on the
responsibilities of the regional and local governments for providing significant
services to the populace. If the provision of services is decentralized
substantially, then the taxing power also should be decentralized substantially.
To decentralize the power to tax, however, it is often necessary for the
national government to do more than simply assign particular tax bases to
local or regional governments. As we have argued above, it also is necessary
114
for the national government to fashion a fiscal framework that allows the
regional and local governments to cope effectively with tax competition.
115
Appendix B - International Cooperation on Income Tax Matters
The strong movement towards globalization at the end of the 19th
century led many countries to seek ways of cooperating to avoid impediments
to cross-border trade and investment and to minimize opportunities for tax
evasion. A primary mechanism for cooperation was the bilateral income tax
treaty.
The earliest income tax treaties were adopted by Prussia around the
start of the twentieth century. Its first treaty was with Austria, signed in 1899.
Soon thereafter, it entered treaties with Bavaria, Saxony, and Baden, followed
by treaties with Luxembourg and the city of Bale. Other continental
governments soon followed the Prussian lead. Hungary and Austria concluded
a treaty in 1909.
The development of tax treaties was largely stalled by the hostilities
that engulfed much of the world in 1914. With the dramatic increase in tax
rates that followed the First World War, many countries, especially in
continental Europe, became interested in dealing with double taxation
problems through the treaty mechanism. Germany took the lead after the war,
signing a treaty with Czechoslovakia in 1921 and with Austria in 1922. In
1922, Italy convened a conference in Rome of the states emerging from the
dissolution of the Austro-Hungarian Empire (Austria, Hungary, Poland,
Romania, and the Kingdom of the Serbs, Croats, and Slovenes). The goal was
to forge a multilateral agreement among the represented countries. The
practical result was a bilateral tax treaty between Italy and Austria.
116
Developments up to 1946
Beginning in the early 1920s, the League of Nations began to play a
leadership role in promoting international cooperation on direct taxes. It
published the first internationally important model tax convention in 1928.
That model was the culmination of work that had begun in 1920. The
convention recognized the importance of some sharing of taxing power
between the source country and the residence country. It did not attempt to
limit the methods the Contracting States might use to prevent double taxation.
Instead, it offered alternative means of achieving a reasonable sharing of tax
revenues from cross-border activities.
The League of Nations model convention of 1928 was prepared by a
group of government experts on double taxation and tax evasion. The initial
group, formed in 1922, had seven members, all officials from European
countries (Belgium, Czechoslovakia, France, Great Britain, Italy, the
Netherlands, and Switzerland). In 1925, the committee was expanded to
include two more European representatives (Germany and Poland) and three
additional representatives from outside Europe (Argentina, Japan, and
Venezuela). Beginning in 1927, the United States, which was not a member of
the League, participated on an informal basis. The expanded committee
produced a report and draft convention in 1927, which was circulated to
League members and some nonmember States. In 1928, the Secretary General
of the League convened a meeting of government experts in Geneva,
Switzerland, to consider the draft convention. That group of experts included
the eleven who had drafted the report and members from 16 additional
countries (Austria, Bulgaria, China, Danzig, Denmark, Estonia, Greece,
Hungary, Irish Free State, Latvia, Norway, Romania, South Africa, Spain,
Sweden, and the Union of Soviet Socialist Republics). Influence of the
international business community appears to have been modest, although the
International Chamber of Commerce played a consultative role from the early
1920s onward. The drafters of the convention relied in part on a 1923 League
of Nations report on double taxation prepared by a group of four economists
from Great Britain (Sir Josiah Stamp), Italy (Professor Einaudi), the
Netherlands (Professor Bruins), and the United States (Professor Seligman).
117
The League’s first model convention clearly embraced the twin goals
of eliminating double taxation and also preventing undertaxation. The Report
(p. 26) states:
From the very outset, the Meeting of Government experts
realized the necessity of dealing with the questions of tax
evasion and double taxation in co-ordination with each other. It
is highly desirable that States should come to an agreement
with a view to ensuring that a taxpayer shall not be taxed on
the same income by a number of different countries, and it
seems equally desirable that such international co-operation
should prevent certain income from escaping taxation
altogether. The most elementary and undisputed principles of
fiscal justice, therefore, require that the experts should devise a
scheme whereby all incomes would be taxed once, and once
only. (Emphasis added.)
The League’s group of government experts proposed three alternative
ways of taxing income from cross-border activities. All three plans used the
concept of a “permanent establishment” (PE) that had appeared in the early
Prussian treaties. Under the three plans, the income of an industrial,
commercial or agricultural “undertaking” would be taxable in the countries
where its permanent establishments are situated. (Model 1a, Art. 5; Model 1b,
Art. 2B; Model 1c, Art. 3.) The models differ primarily in their treatment of
income from moveable capital. The first model generally reserves the right to
tax in the source country, whereas the other two models generally assign the
right to tax to the residence country.
To prevent double taxation of business profits when an “undertaking”
has a PE in more than one country, the three models included in the 1928
convention provide that each treaty partner would tax the income “produced
in its territory.” The term “undertaking” was undefined; it apparently has the
same meaning as the term “enterprise,” which is an undefined term found in
modern tax treaties. In general, an enterprise is a group of corporations or
other legal entities under common control and engaged in a common business.
The allocation of income among taxing jurisdictions under the 1928
118
convention was to be accomplished through consultation between the
“competent administrations of the two Contracting State.” Some form of
formulary apportionment was an acceptable method for allocating the income
between the countries where the PEs were located. The important point was
that all of the income of an undertaking would be allocated to some PE. No
exemption or other tax relief for an item of income would be given in one
Contracting State without giving the other Contracting State the concomitant
right to tax that item of income.
The drafters of the 1928 convention were favorably disposed towards
the resolution of some tax conflicts by ceding full jurisdiction to tax to the
state of residence. On a unilateral basis, some countries, including the United
States, had resolved the difficult problems of source taxation of shipping
companies by ceding source jurisdiction to the residence country on a
reciprocal basis. This approach was unworkable in many countries because
they did not attempt to tax the worldwide income of their residents. In
addition, it was not an attractive option for countries that would lose revenue
from the reciprocal arrangement, due to the lack of foreign activities of their
residents. The model convention was written so that it would be compatible
with this development without being dependent on it.
The 1928 convention addressed only a small subset of the international
tax issues of importance to the taxation of cross-border activities. Recognizing
the need for additional work, the League of Nations established in 1929 the
Fiscal Committee. The Fiscal Committee engaged in activities before, during,
and after World War II. It ceased operations after the war when the League
was supplanted by the United Nations.
During the 1930s, much of the work of the League’s Fiscal Committee
was in developing rules and standards for allocating the income of industrial
and commercial enterprises operating in more than one country. In 1933, it
prepared a draft multilateral convention on the allocation of profits. That
model convention was revised in 1935 but never formally adopted. The basic
approach in that convention was to treat a permanent establishment as if it
were a separate enterprise and to determine its income based on what has
119
come to be called the arm’s length standard. The goal was to allow countries
to determine the income of a PE based on the local books of account, without
reference to the books of the entire firm. Apportionment by formula was
reserved as a last resort. This approach was a marked change from the 1928
model convention, which provided that all of the income of an “undertaking”
be allocated among the countries where the enterprise or undertaking has a
PE.
Work on a model convention continued during the war years. In 1940
and again in 1943, the League’s Fiscal Committee convened a regional
conference in Mexico City to revise the 1928 convention. The meeting
included representatives from countries of the Western Hemisphere, both
developed and developing (Argentina, Bolivia, Canada, Chile, Colombia,
Ecuador, Mexico, Peru, the United States, Uruguay, and Venezuela). The
regional conference produced a draft model convention, popularly known as
the Mexico Draft.
In general, the Mexico Draft took the position that the primary
jurisdiction to tax income should be assigned to the source country — the
position favored by the developing countries. In this respect, the Mexico Draft
was a major refinement of the League’s 1928 model convention. The basic
structure of the 1928 convention, however, remained unchanged. The Mexico
Draft clarified an ambiguity in the 1928 convention by providing that all of
the business income derived in a country would be taxable there unless the
activities were “isolated or occasional” and the enterprise earning the income
did not have a PE in that country. A protocol to the convention provided rules
for attributing income to a PE. These rules, in general, followed the Fiscal
Committee’s 1935 draft convention for the allocation of profits.
In 1946, the League’s Fiscal Affairs Committee, with a full
complement of members, met in London, England, to review the Mexico
draft. A new draft, popularly referred to as the London Draft, was produced.
The London draft was similar in some respects to the Mexico Draft in the
treatment of business profits. It differed from the Mexico Draft, however, in
120
requiring that an enterprise resident in one Contracting State have a PE in the
other Contracting State for that latter State to tax the enterprise on any portion
of its business profits. The London Draft also imposed significant limitations
on the taxation of investment income in the source country, contrary to the
position favored by the developing counties and adopted in the Mexico Draft.
For example, the Mexico Draft provided that royalties paid for the right to use
a patent or secret process would be taxable only in the State where the right is
exploited. In contrast, the London Draft reserved the right to tax such royalties
to the country of residence.
1946 to Present
The United Nations came into being on October 24, 1945, although the
League of Nations continued for some short period thereafter. Various
governments concerned with tax treaty issues made efforts to revitalize the
work of the League of Nations on tax conventions. Although the United
Nations took an active interest in that effort, leadership in developing a model
treaty was seized by the Organisation for European Economic Cooperation
(OEEC). The OEEC was created in 1948 in conjunction with the Marshall
Plan for revitalizing the economies of Europe. It has 16 members, all
European countries outside the Soviet block. In 1956, the OEEC created its
Fiscal Committee and entrusted it with the task of working out a draft model
bilateral tax convention. The OEEC became the Organisation for Economic
Cooperation and Development (OECD) in 1960. Membership of the OECD
was extended to the major industrial countries of the world. It now has 30
members. No developing countries are members of the OECD.
The OECD published its draft model treaty in 1963. That convention
drew heavily from the League’s London Draft, but included additional
features that favored capital exporting countries over capital importing
countries. In form, the 1963 draft recognized the importance of sharing of
revenue between the source and residence countries. Its practical effect,
however, was to erect substantial barriers to the exercise of source jurisdiction
and to afford multinational firms with major opportunities for tax minimizing
strategies through its emphasis on legal form over economic substance. The
121
1963 draft was revised slightly over the next decade and a half. It was
published in final form in 1977.
The OECD model was extremely effective in eliminating most of the
common types of double taxation. It was not very effective, however, in
eliminating international tax avoidance and evasion. Indeed, several core
features of the OECD model have promoted international undertaxation. The
OECD model provides for an exchange of information on tax matters between
the Contracting States. The goal of the exchange-of-information article is to
curtail tax avoidance and evasion. Unfortunately, the article allows countries
to decline to provide information if they have adopted bank-secrecy laws or
otherwise are disinclined to exchange information for policy reasons. As a
result, the exchange of information clause has not been effective in most
cases. However, it has been effective for a few countries, such as the United
States and Canada, that have a long tradition of cooperation on tax matters.
One feature of the OECD model that promotes tax avoidance is the
treatment of affiliated entities. Under the model, a member of a multi-national
enterprise (MNE) group will not have a PE in a country solely because it
conducts its business in that country through a related company. This rule has
been read broadly to mean that the affiliate relationship is irrelevant in
determining whether the affiliate is the PE of the MNE group (or some
member thereof). As a result, the affiliate must be an agent of the group and
must habitually make contracts on behalf of the group to be a PE of the group.
Tax planners have no difficulty in ensuring that an affiliate is never the PE of
the group unless they have some reason for wanting it to be a PE.
The affiliate rule permits what tax specialists refer to as “entity
isolation.” In general, the MNE group assigns to a separately incorporated
affiliate all of the activities and property that the MNE group needs to conduct
business in a particular country and that would constitute a PE. To the extent
possible, assets and activities that are likely to generate substantial profits are
not assigned to the affiliate. For example, valuable intangible property, such
as patents and secret processes, would not be assigned to the affiliate. The
122
intangible property is kept in an offshore affiliate, and the taxable entity is
required to pay a substantial royalty to the offshore affiliate for the use of the
intangible property in the source state. As a result of such techniques, the
MNE group typically is able to restrict the profits taxable in the source
country to some modest return on its physical assets located in that country.
Three additional features of the OECD model, in addition to the PE
rule, support entity isolation. First, the OECD model generally requires the
source state to allow a deduction for royalty payments to a foreign MNE
unless the tax authorities can show that the payments constitute some type of
sham transaction or are unreasonable in amount. Second, the OECD model
convention prohibits the source state from imposing a withholding tax on the
royalty payments. Third, the OECD model has no effective mechanism for
preventing treaty shopping — the use of a treaty by a resident of some third
country that is not a party to the treaty. Through treaty shopping, an MNE can
move income ostensibly allocated for taxation in the residence country to a tax
haven country, with the result that the income of the MNE is not taxed either
in the source state or the residence state. These three rules, in combination,
permit extensive use of what tax specialists refer to as earnings stripping.
The problems with the OECD model described above have been well
understood by tax experts for several decades. Nothing has been done to deal
with these problems for two reasons. First, the OECD typically operates by
consensus, and some members of the OECD have been prepared to prevent
agreement on any model that would undercut their ability to attract capital,
including capital from those avoiding tax in other countries. Second, the
OECD has been a victim of its own success. An important goal of a model
convention is to promote uniformity. The OECD model has been embraced by
most developed countries, with relatively minor adaptations, and, with some
more significant modifications, by much of the rest of the world. To change
the model is to upset the uniformity that the model has achieved. Given the
very long gestation period for implementing changes in a country’s tax
treaties, a change in the model would result in non-uniformity for a decade or
more. As a result, the OECD has attempted, with some limited success, to deal
123
with defects in its model through its Commentary. Changes in the
Commentary have been effective in addressing certain types of issues, but
core issues like entity isolation cannot be addressed effectively through the
Commentary.
The widespread success of the OECD model in the 1970s provoked a
reaction from developing countries. Those countries, being outside of the
OECD, were excluded from effective participation in the design of the model.
Under the League of Nations, the developed countries were the dominant
force in designing a model convention. Only in the Mexico draft were the
interests of the developing countries given high prominence. Still, the
developing countries were represented in the process of approving model
conventions. With the capture of the model treaty process by the OECD, the
participation by developing countries ended. They were disenfranchised at a
time when the number of developing countries was increasing markedly, due
in large part to the collapse of colonialism in Africa and Asia after World War
II.
The developing countries responded to the success of the OECD
model convention by developing their own model convention under the
auspices of the United Nations. The United Nations’ involvement in the treaty
process originated with a resolution of the Economic and Social Council,
dated August 4, 1967. That resolution requested the Secretary-General to set
up an ad hoc working group of tax experts and tax administrators who would
represent the interests of both developed and developing countries and would
represent different regions and tax systems. The ad hoc group would be asked
to formulate guidelines and techniques for use in tax treaties between
developed and developing countries. In 1968 the Secretary-General responded
to the resolution by setting up the Ad Hoc Group of Experts on Tax Treaties
between Developed and Developing Countries. The group was composed of
tax officials and experts from twenty countries. In principle, these officials
were to act in their individual capacity and not as representatives of their
government.
124
In 1979, the United Nations published a Manual for the Negotiation of
Bilateral Tax Treaties between Developed and Developing Countries. The
manual had been prepared by the Group of Experts and the UN Secretariat.
The following year, the United Nations published the United Nations Model
Double Taxation Convention between Developed and Developing Countries.
That publication included a model convention and Commentary, both of
which had been prepared and approved by the Group of Experts. The UN
model was based in substantial part on the OECD model, but it departed from
the latter model on some key points. In particular, it modified the definition of
a PE to allow some additional taxation of business income by the source
country, and it provided that any reduction in a country’s statutory
withholding rates would by done through bilateral negotiations. No specific
target withholding rates were established in the model. The obvious
expectation was that treaties based on the UN model would include a positive
withholding rate on royalties and that the withholding rates on dividends and
interest generally would exceed the rates recommended in the OECD model.
In 1980, the membership of Ad Hoc Group, renamed the Ad Hoc
Group of Experts on International Cooperation in Tax Matters, was increased
from 20 to 25. The membership currently includes experts and tax
administrators from 10 developed countries and 15 developing and
transitional countries. The countries currently represented are: Argentina,
Brazil, Burkina Faso, China, Côte d’Ivoire, Egypt, Finland, France, Germany,
Ghana, India, Indonesia, Israel, Jamaica, Japan Mexico, Morocco, The
Netherlands, Nigeria, Pakistan, Russian Federation, Spain, Switzerland,
United Kingdom, and the United States. A representative of the Palestine
Authority is invited to attend the meetings of the Group of Experts.
The UN model has been effective in influencing the content of
bilateral tax treaties between developed and developing countries. In
particular, virtually all of the treaties between developed and developing
countries provide for a positive withholding rate on royalty income, and the
modifications proposed in the PE article by the UN model have been adopted
widely. The core of the UN model, however, is taken from the OECD model.
125
As a result, the UN model permits companies to engage in entity isolation to
minimize their exposure to taxation in the source country. The UN model does
have a somewhat broader exchange-of-information article. Unfortunately,
developing countries have had little success in obtaining information from the
developed countries through that article.
Both the OECD model and the UN model were not modified during
the 1980s to respond to the increased globalization of the world economy. The
OECD finally began making some amendments to its model in the 1990s, and
now publishes its model in loose-leaf form. The United Nations published a
new model in 2001 — the first revision in two decades. These revisions made
many useful changes, mostly of a technical nature. The revised models,
however, have continued the emphasis on formalisms. For example, both
model conventions continue to permit entity isolation by treating the
corporations comprising a multinational enterprise as if they were independent
entities, notwithstanding their economic integration in most cases.
Although the OECD and UN models have addressed the set of
international tax issues identified by the League of Nations in the 1920s, they
have not addressed many important international tax issues that have emerged
in the 1980s and thereafter. For example, neither model addresses crossborder mergers, aggressive tax avoidance schemes, treaty shopping, hybrid
entities, overlapping controlled-foreign-corporation regimes, and global
trading of financial instruments.
126
Over the past several years, the OECD has attempted to involve
developing countries in its various activities relating to tax treaties and
international taxation in general. It has invited representatives of developing
countries as observers on some of its meetings, particularly with respect to ecommerce. It has also invited a selected group of developing countries to
register objections to positions taken by the OECD in the Commentary to its
model convention. This initiative seems unlikely to result in the developing
countries influencing the core decisions made by the OECD on tax treaty
matters or on any other matters of importance to the OECD member states.
127
References
Alm, James and Hugo López-Castaño (2002) “Payroll Taxes in Colombia,”
Unpublished paper prepared for Misión de Finanzas Públicas, Bogotá,
December.
Bernheim, Douglas B. (2002). “Taxation and Saving,” in Alan J. Auerbach.
and Martin
Feldstein (eds.), Handbook of Public Economics, Volume 3
(Amsterdam:, North-Holland).
Bird, Richard M. and Barbara Miller (1989) “The Incidence of Indirect
Taxation on Low-Income Households in Jamaica,” Economic
Development and Cultural Change, 27: 393-409.
Bird, Richard M. and Duanjie Chen (1998) “The Fiscal Framework for
Business in Asia,” in Wendy Dobson, ed., Fiscal Frameworks and
Financial Systems in East Asia, Hong Kong Bank of Canada Papers on
Asia, vol. 4 (Toronto: Institute for International Business, Faculty of
Management, University of Toronto).
Burgess, Robin and Nicholas Stern (1993) “Taxation and Development,”
Journal of Economic Literature, 31: 762-830.
Chattopadhyay, Sumen and Arindam Das-Gupta (2002) “The Compliance
Cost of the Personal Income Tax and Its Determinants,” National
Institute of Public Finance and Policy, New Delhi.
Chu, Ke-young, Hamid Davoodi, and Sanjeev Gupta (2000) “Income
Distribution and Tax and Government Spending Policies in
Developing Countries,” IMF Working Paper /00/62, Washington.
Ebrill, Liam, Michael Keen, Jean-Paul Bodin, and Victoria Summers (2001).
The
Modern VAT (Washington: International Monetary Fund)
Hughes, Gordon (1987) “The Incidence of Fuel Taxes: A Comparative Study
of Three Countries,” in David Newbery and Nicholas Stern, eds., The
Theory of Taxation for Developing Countries (Published for the World
Bank by Oxford University Press).
Jha, Prabhat and Frank Chaloupka (2000) Tobacco Control in Developing
Countries (Oxford University Press for World Bank and World Health
Organisation).
League of Nations (1928) “Double Taxation and Tax Evasion: Report
128
Presented by the General Meeting of Government Experts on double
Taxation and Evasion,” II. Economic and Financial 1928.II.49.
(Geneva: League of Nations).
League of Nations, Fiscal Committee (1946) “London and Mexico Model Tax
Conventions: Commentary and Text,” C.88.M.88.1946.II.A. (Geneva:
League of Nations).
Mann, Arthur J. (2002) “Estimating the Administrative Costs of Taxation: A
Methodology with Application to the Case of Guatemala,” DevTech
Systems, Arlington VA, August.
Martinez-Vazquez, Jorge and Robert M. McNabb (2000) “The Tax Reform
Experiment in Transitional Countries,” National Tax Journal, 53: 27398.
McIntyre, Michael J. (1997) “Tracing Rules and the Deduction for Interest
Payments: A Justification for Tracing and a Critique of Recent U.S.
Tracing Rules,” Chapter 17 of Taxation Towards 2000, John G. Head
& Richard Krever, Eds. (Sydney: Austrailian Tax Research
Foundation).
McIntyre, Michael J. (2003) “The Use of Combined Reporting by Nation
States,” prepared for Conference at Harvard Law School, “Taxation of
Business Profits in the 21st Century,” Oct. 24—26, 2002
(forthcoming).
OECD Committee on Fiscal Affairs (1998) Harmful Tax Competition: An
Emerging Global Issue (Paris: OECD).
OECD Committee on Fiscal Affairs (1992) Model Tax Convention on Income
and on Capital, updated looseleaf (Paris: OECD).
Organisation for Economic Co-operation and Development (1994) Taxation of
New Financial Instruments (Paris: OECD).
Sandford, Cedric, ed. (1995) Tax Compliance Costs: Measurement and
Policy (Bath, UK: Fiscal Publications).
Shah, Anwar and John Whalley (1990)”Tax Incidence Analysis of Developing
Countries: An Alternative View,” World Bank Economic Review, 5:
533-52.
Tanzi, Vito (1987) “Quantitative Characteristics of the Tax Systems of
Developing
129
Countries,” in David Newbery and Nicholas Stern, eds., The Theory of
Taxation for Developing Countries (Published for the World Bank by
Oxford University Press).
Thirsk, Wayne (1997) Tax Reform in Developing Countries (Washington:
World Bank).
United Nations (2001) United Nations Model Double Taxation Convention
between Developed and Developing Countries (New York: United
Nations), ST/ESA/PAD/SER.E/21.
United Nations (2001) High-level International Intergovernmental
Consideration
of Financing for Development (“Zedillo Report”) (New York: United
Nations), A/55/1000.
United Nations (2000) Millennium Declaration (New York: United Nations),
A/RES/55/2.
United Nations (2000) Report of the Secretary-General to the Preparatory
Committee for the High-level International Intergovernmental Event
on Financing for Development (New York: United Nations),
A/AC.257/12.
United Nations (2002) Report of the International Conference on Financing
for Development, (Monterrey Consensus”) (New York: United
Nations), A/CONF.198/11.
130