|
Article Excerpt
Abstract--We review the changing nature of tax policy in developing countries over the last 30 years and consider what factors determining the level and structure of tax revenues in such countries may have changed recently and how such changes may affect future developments.
INTRODUCTION
A century ago any list of the developing countries of the world would almost certainly have included three "regions of recent settlement"--Canada, Australia, and Argentina. Today, only Argentina would be included in such a list. Did different tax choices have any influence on this outcome? Arguably, they may have done so. Argentina's heavy reliance on export taxes and its failure to develop an adequate internal tax system undoubtedly played some role first in derailing its impressive economic progress in the early years of the last century and then in explaining its inability to get firmly back on the growth track in the latter half of the century.
Whether this argument is right or not, this brief story introduces two of the most important lessons that have been learned about taxation and development over the last hundred years. The more obvious lesson is simply that one must be careful not to kill the goose that lays the golden eggs--in Argentina's case, agricultural exports. A less obvious but equally important lesson is that a good internal tax system provides not only revenue but an essential element in developing a capable state. Argentina is clearly still working at this task. In contrast, Canada and Australia--responding in part to wartime exigencies--essentially managed to create such systems in the first half of the last century. (1)
Are there lessons for contemporary developing countries to be learned from such broad assertions as these? There may be, although there is far too much that is not yet understood about the historical role of tax policy and tax administration in the development of countries to explore this topic further here. (2)
In the present paper we focus less ambitiously on the last few decades where the data are better and where the subject has been more extensively worked. The main question considered is whether the setting for tax policy in developing countries is any different now than in the past, and whether or not such differences show up in how countries tax. In the next section, we begin by looking at how the level and structure of taxes has changed over time. In recent decades both technical knowledge and the degree of training have improved markedly in low--and middle--income countries around the world. If these developments have resulted in better tax decisions and those decisions have made a difference, the data should show it. We then turn to the central question discussed in this paper: are the factors now driving tax policy in the developing world different than in the past? We conclude by speculating about some factors likely to shape tax policy in low--income countries in the next decade or so and draw a few general conclusions.
TAX LEVELS AND TAX STRUCTURE
How developing countries tax themselves changes continuously. But are either tax levels or tax structures very different now than they were 30 years ago?
The Level of Taxation
Table 1 depicts the average level of the ratio of taxes (including social security taxes) to GDP for three groups of countries: industrialized, developing and transition. (3) In industrialized countries the average tax share increased from 30 to about 35 percent over this 30 year period. In developing countries, however, the tax share of output increased only slightly; indeed, since the 1980s their tax shares have been almost constant. This represents a remarkable slowdown. In an earlier study (of a more limited sample that excluded social security taxes) Chelliah (1971) found that the average tax ratio for central governments in less developed countries had increased by about 24 percent over the previous two decades (from 11.3 percent in 1953-55 to 13.8 percent in 1966-68). This general result held true for nearly every country in his sample. "Convergence" in tax levels across countries appeared to be well on its way. In reality, however, by the end of the 20" century, the tax ratio in industrialized countries was about twice that in developing countries--a much greater difference than in the 1970s. Indeed, as Table 1 shows, the tax share in the transition countries actually declined in the last decade of the century, reflecting the continuing realignment of public--private expenditure responsibilities in those countries. (4)
Almost half a century ago, Nicholas Kaldor (1963), fresh from his recent exposure to India's tax system, argued that for a country to become "developed" it needed to collect in taxes 25-30 percent of GDP. More recently, perhaps having noted that most developing countries (like India) remain well short of Kaldor's target, the UN Millennium Project (2005) was somewhat less ambitious in advising developing countries that on average they needed to mobilize only an additional four percent of GDP in tax revenue beyond their current average level of about 18 percent? However, the news is not good for those who think that a larger tax state is an essential aspect of development, for example because of the need for public investment in infrastructure: the tax to GDP ratio hardly changed in developing countries in recent decades. The average developing economy seems to have been content with (or constrained to) a level of taxes roughly equivalent to 17 percent of GDP. (6)
Most developing countries have consistently failed to meet the targets cheerfully established for them by outsiders. A few fast--growing Asian countries such as India managed to reach and even exceed the UN--prescribed four percent of GDP increase in tax ratio in the early years of this century, but it is by no means clear that these new higher levels will be sustainable. (7) As Table 1 shows, in most developing countries the tax ratio has changed surprisingly little in recent decades. The belief that some seem to hold that developing countries can increase their tax take simply through more vigorous collection efforts is particularly naive. (8) There is more to improving tax effort than simply exhorting countries to try harder.
Of course, there is considerable variation across developing countries. While this is not the place to go into details, a recent analysis of the determinants of this variation in the tax ratio suggests that (a) developing countries that increased taxes did so largely in response to an increase in per capita GDP; (b) increased reliance on indirect taxes did not seem to drive the increases; (c) emphasis on social service spending tended to dampen it, while spending more for economic services did not seem to matter; and (d) there is some support for the argument that corruption and taxation are substitutes (Bahl, 2006). (9)
The Structure of Taxation (10)
In contrast to the stasis of the tax level in developing countries, as Table 2 shows, in recent decades there have been pronounced shifts in their tax structures. Unlike the 1950s and 1960s, when tax revenues from import duties were increasing at a faster rate than GDP (Chelliah, 1971), in more recent years international trade taxes declined from about 32 percent to about 25 percent of total taxes--a shift precisely offset by an increase in the share of domestic indirect taxes from about 25 percent to about 35 percent by the end of the 1990s. (11) Trade liberalization towards the end of the 20th century was obviously a driving force in tax reform in developing countries, as was the widespread adoption of the value added tax (VAT) and the continuing improvement in its administration. (12) As Bird and Zolt (2005) stress, in contrast to the experience in most developed countries depicted in Table 2, the personal income tax has continued to play at most a very limited role in developing countries. Developing countries have been hesitant to go too far in taxing labor in the formal sector, and labor in the informal sector has remained out of reach. Only limited data are available. In addition, in many countries recently there has been some decline in reliance on the company income tax (Keen and Simone, 2004) owing to rate reductions, base narrowing due to incentive polices, and declines in reported profitability. The long--term story with respect to income taxes has been the continued inability of the tax administration in less--developed countries to administer such taxes effectively. Much the same is true with respect to property taxes.
What Do These Trends Tell Us?
Has the failure to mobilize resources through taxation at the margin been growth--enhancing or growth--retarding? There is no agreement on this subject. Some argue that the failure to mobilize more resources has constrained governments from extending the quantity and improving the quality of public services delivered or unburdening themselves from heavy debt. (13) More specific to the question of economic development, lower taxes may constrain infrastructure investments to suboptimal levels and retard industrial development. Lower taxes in some cases have led to lower primary surpluses than are optimal, especially when combined with high debt burdens, a matter of much concern to organizations like the IMF (2006). (14) Others argue that lower taxes keep money in private hands where it is more likely to find its way into investment and job creation. This camp also holds that government expenditures in the developing countries are biased toward consumption and likely to crowd out private investment and, hence, hamper growth.
Empirical work on the impact of tax levels on growth in developing countries has come to no firm conclusions. Even endogenous growth models that allow for the effects of tax policy on growth do not give a consensus answer about whether higher taxes crowd out faster rates of economic growth (Tanzi and Zee, 1997; Mintz, 2003). It is difficult to separate the effect of the level of taxes from the level of expenditures and the budget balance. Not surprisingly, different model specifications produce different results.
The effect of tax structure on economic growth is an equally unresolved issue. In theory, distortions in the tax structure can impose a drag on the economy. Using computable general equilibrium models, the welfare cost of some taxes in some developing countries has been estimated to be more than 100 percent of the amount of tax raised. (15) Others point to the stimulus effects of tax rate reductions. The evidence here is also not clear. Ivanova, Keen and Klemm (2005), for example, find no evidence of a supply--side effect from Russia's rate reduction and adoption of a flat rate income tax, but Martinez--Vazquez, Rider and Wallace (2008) do find evidence of a labor supply effect.(16) As Lindert (2003) shows in historical context, the effects of taxes in particular country settings often depend on very detailed characteristics of tax design and implementation that are not easily captured in econometric models.
DETERMINANTS OF TAX POLICY CHOICES
Looking back to the 1970s, the prescriptions for good tax policy dispensed to developing countries by consultants and donors, when they did not amount simply to saying "copy my country," did not stray far from Adam Smith's maxims for a good tax (Goode, 1993). The recipients of such advice quickly adopted the rhetoric. Politicians almost everywhere swore that their proposed reform would be easier to understand and administer and would distribute any increase in tax burdens in an equitable way. Only the neutrality...
|
