The past decade has witnessed wide-ranging economic reform in several Latin American countries. There is little resemblance between the economies of the early 1990s and their more controlled, poorer performing counterparts of the early 1980s. The new atmosphere reflects, in part, support for sound economic policymaking at the highest political levels and a change in attitudes that now permits the consideration of deregulation, privatization, and the opening up of the economy. Deregulation has spurred economic activity. The opening of the economy, with its impact of increased imports, has led to higher revenue from this source.1 Where privatization occurred, it has also led to increases in revenue. Significant efforts to reduce the fiscal deficit—even though in most instances the improvement was mainly the result of revenue-enhancing rather than expenditure-tightening efforts—and sensible credit policies have led to success in the control of often runaway inflation. In some countries, the role of the state itself has undergone careful scrutiny. Consequently, government has become less involved in running public enterprises, which began to respond more to market signals; there have been some efforts to contain the government payroll; and, in a few instances, social security has passed from public to private management.

The past decade has witnessed wide-ranging economic reform in several Latin American countries. There is little resemblance between the economies of the early 1990s and their more controlled, poorer performing counterparts of the early 1980s. The new atmosphere reflects, in part, support for sound economic policymaking at the highest political levels and a change in attitudes that now permits the consideration of deregulation, privatization, and the opening up of the economy. Deregulation has spurred economic activity. The opening of the economy, with its impact of increased imports, has led to higher revenue from this source.1 Where privatization occurred, it has also led to increases in revenue. Significant efforts to reduce the fiscal deficit—even though in most instances the improvement was mainly the result of revenue-enhancing rather than expenditure-tightening efforts—and sensible credit policies have led to success in the control of often runaway inflation. In some countries, the role of the state itself has undergone careful scrutiny. Consequently, government has become less involved in running public enterprises, which began to respond more to market signals; there have been some efforts to contain the government payroll; and, in a few instances, social security has passed from public to private management.

Tax reform comprised not only an integral part of the overall reform effort but often represented its strongest element. A majority of the Latin American countries undertook major tax reforms, even though a small but important minority do not seem to have been able to generate the critical mass of political support for significant change. The fact that tax reform represented such an important policy instrument in overall macromanagement and economic growth suggests, to no small extent, that country authorities recognized fiscal policy—and especially tax policy—as “their” policy, that is, more amenable to influence, with results that are more directly measurable than are the ramifications of other economic policies, at least in the short to medium run.2

This paper reviews Latin American countries’ recent experiences with tax policy, mainly since the mid-1980s. Based on those experiences, the paper points in the direction of further necessary reforms. In what follows, the main trends in tax policy among Latin American countries in recent years are examined. The next section compares and contrasts important elements of tax reform—or the lack of it—focusing on Argentina, Bolivia, Brazil, Chile, Colombia, and Mexico.3 The final section outlines future directions for tax reform, especially as the economies, their fiscal institutions, and, perhaps more important, the thinking on these matters mature. These relate to various issues, such as the realization that a tax system—both policy and its administration—cannot be viewed in isolation, especially as a country becomes more integrated with the rest of the world, the need to widen the universe of taxpayers as economic growth begins to result in an expanding middle class; the constraint put on fiscal management by the dearth of tax and expenditure responsibility at lower levels of government; the emergence of new and dynamic concerns, such as those of the environment, with implications for the design of a tax system; and so on. The main findings are summarized and some concluding remarks offered.

Recent Trends in Tax Reform4

At the beginning of the 1980s, most Latin American tax structures were complex and cumbersome. Often, they consisted of hundreds of taxes but yielded little revenue. Consumption and production taxes were hampered by multiple rates and led to difficulties with tax administration. The taxes were inefficient because of “cascading”; that is, they were levied not only on the value of production but also on the taxes that were paid at earlier stages of production. They tended to impair international competitiveness because they were often levied at the manufacturing rather than at the retail stage, so that, effectively, exporters also paid the tax. Income taxes were riddled with multiple exemptions and incentives, high rates, a lack of integration between personal and corporate incomes, and a lack of indexation, in general leading to low revenue productivity, inequities caused by narrow tax bases, and inefficiencies in resource allocation, including biases in the debt-equity composition of business financing. On the whole, the tax systems depended on a few highly distortive domestic taxes or on international trade taxes, including export duties, to generate revenues.

The tax reforms that followed comprised either a series of steps over a number of years, as in Colombia (an incremental approach), or what amounted to a sea change, as in Argentina (a revolutionary approach). In general, though, the common experience was one of continuous formulation of reform policies and their steady, sequenced application. Reform-minded countries attempted to simplify their tax structures, focusing, in the early stages, on income taxation but increasingly moving toward taxing production and consumption. As the economies matured and became more integrated with the rest of the world, such as the Mexican economy, their focus reverted to fine-tuning particular aspects of the income tax, which had international ramifications.

Changes in Overall Tax Structure

Among the important changes that the Latin American tax structures underwent, the following pertain to the main tax categories.

Personal Income Tax

With respect to the personal income tax, first, the band was narrowed through a reduction in the number of rates, and the rates themselves were scaled down. Between 1985–86 and 1991, the rates declined, on average, to 5–36 percent from 7–47 percent (Table 6.1). Indeed, few countries had a top marginal rate over 50 percent, while many had brought it down to the 30 percent range.

Second, there was a tendency to reduce the burden of taxation on low-income groups as indicated by an increase in the threshold level at which the personal income tax applied. Thus, on average, the threshold more than doubled from 0.8 times per capita GDP to 1.8 for Central American countries, while for South American countries it more than doubled from 0.4 times per capita GDP to 1 (Table 6.2).

Table 6.2.

Cross-Country Comparisons: Personal Income Tax Exemption and Upper Bracket

(Multiples of per capita GDP)

article image
Sources: Secondary, published Sources such as publications of tax summaries by Price Water - house, Coopers and Lybrand, International Bureau of Fiscal Documentation, the IMF, and other, similar Sources.

Averages are taken over the set of countries for which data for both 1991 and 1985 or 1986 are available.

Allowance equals 12 months’ minimum wage in zone of residence (13 months with Christmas bonus). The data provided correspond to Mexico City.

Third, as the top marginal rate came down, it began simultaneously to be applied at lower levels of income for Central American countries. An important difference between Central American and South American countries is that, in the former, the income at which the highest marginal rate applied fell by half, from over 200 times per capita GDP to over 100 times. By contrast, in South American countries, it remained almost the same, decreasing slightly from just over 13 times per capita GDP to just under 13 times.5

exempted them or taxed them at a reduced rate, only to shift back within a few years to treating them as ordinary income for tax purposes.6 This instability disappeared, however, as most countries settled on the policy of treating capital gains as ordinary income and applying the same tax rates to both.7

Third, Latin American countries reduced tax evasion by improving tax administration, such as through rapid computerization of collection procedures and processing of taxpayer information. The payment of taxes was facilitated, for example, through banks, selective auditing, and the establishment of large taxpayer unit8 but also through tax policy measures entailing estimating the extent of tax evasion and enacting presumptive taxes and minimum taxes. Measuring and taxing income at the company level is a difficult task that not only involves complex technical issues but also elicits severe and continuous threats from highly sophisticated tax “planners” looking for ways to erode the tax base. Several countries have been looking for schemes to ensure that at least a minimum contribution is obtained from company income. Some countries—Argentina, Ecuador, Mexico, and Peru—introduced a minimum contribution to the income tax based on gross assets, and many others began considering such a tax. Colombia, Costa Rica, and Uruguay legislated a similar tax based either on fixed assets or on net worth, probably owing to the greater political feasibility of this alternative even though these variants are less revenue productive (Table 6.4). Bolivia replaced its corporate income tax with only a net worth tax.

Table 6.3.

Cross-Country Comparisons: Corporate Income Tax Rates

(In percent)

article image
Source: IMF, Tax Policy Division; secondary, published Sources such as publications of tax summaries by Price Waterhouse, Coopers and Lybrand, International Bureau of Fiscal Documentation, the IMF, and other, similar Sources.
Table 6.4.

Cross-Country Comparisons: Assets-Based Taxes

(In percent)

article image
Sources: Secondary, published Sources such as publications of tax summaries by Price Water-house, Coopers and Lybrand, and International Bureau of Fiscal Documentation.

Minimum corporate income tax; can be credited against normal corporate tax. In Mexico, the income tax can be credited against the gross assets tax to avoid the foreign investors’ problem of crediting against tax liability in the home country.

The base is real estate. The tax, however, is conceived not as a property tax but as an additional corporate tax.

This tax has the form of a license to do business. The maximum tax amount is US$20,000 a year.

Although it is called a net worth tax it is, in effect, a gross assets tax because liabilities can nolonger be deducted.

Fourth, the rates of withholding taxes on foreign remittances fell, thereby bringing closer the domestic and foreign components of corporate taxation (Table 6.5). Finally, in Latin America as a group the overall structure of the corporate income tax was improved, for example, through the introduction of indexation in the face of an inflationary environment.

Table 6.5.

Withholding Taxes on Foreign Remittances

(Percent of remittances)

article image
Sources: Secondary, published sources such as publications of tax summaries by Price Waterhouse.

Simple average of figures presented.

Pertains to dividends in cash or kind, other than stock dividends. The beneficiary must be identified; otherwise, the rate is 22.5 percent. Dividends and remittances of branch profits in excess of 12 percent of registered investment are subject to a special remittance tax ranging from 15 percent to 25 percent

Services derived from agreements ruled by the Foreign Technology Law: (a) technical assisstance, technology, and engineering-27 percent (45 percent on assumed profit of 60 percent); (b) cession of right or licenses for inventions, patents, exploitation, and others-36 percent (45 percent on assumed profit of 80 percent); and (c) nonregistered agreements–45 percent (profit of 100 percent is assumed).

This rate was reduced from 25 percent to 15 percent starting January 1, 1993. Treaty rates in excess of 15 percent are automatically reduced.

Payments of interest (loans) to foreign financial institutions are subject to a withholding tax of 15 percent.

Withholding tax will be reduced as indicated above for dividends.

Taxes on dividends are withheld at the basic tax rate with surcharges. If the dividends are paid from undistributed profits of prior years, credit is allowed for the tax already paid on such profits by the company.

No withholding required on interest remitted or credited abroad on loans. A special tax of 0.5 percent to 2 percent on the portion of the loans payable up to two years is levied (only once) at the time loans are registered at the Central Bank of Ecuador. If the loan is due after two years, the special tax is not payable.

Interest on cash foreign-source loans brought into the country is not subject to withholding taxes.

The withholding taxes are an average of different interest and royalties rates.

Interest payments to nonresidents are exempt of Mexican income tax in the case of (a) loans to the federal government; (b) fixed-rate loans for five or more years, by duly registered financial institutions; and (c) certain securities and bank acceptances issued in foreign currency.

lnterest payments to nonresidents are exempt from Mexican income tax for (a) loans to the federal government; and (b) loans for three or more years by duly registered financial entities that promote exports by special financing; (c) these gains are taxable as interest (d) When royalties are paid for the use of patents in connection with the technical assistance required for their use under the same contract, both the licensing fee and amounts paid for the technical assistance will be subject to the lower 15 percent rate, (e) The nonresident taxpayer may elect to pay at the regular corporate tax rate on net profit if he has a resident representative and advises the customer accordingly. The latter, then, makes no withholding.

Treaty signed but not in force at time of publication.

The rate will be reduced to 10 percent five years after the treaty enters into force.

This rate applies where share ownership is at least 10 percent.

The 10 percent rate, applicable to interest on loans from banks and insurance companies, will be reduced to 4.9 percent five years after the treaty enters into force.

Under certain circumstances, exemptions are granted.

Taxable income is determined as gross rentals less depreciation computed as provided by law.

Payments for transfer of technology or for information regarding commercial, industrial, or scientific knowledge are deemed to be royalties.

Consumption Taxes

A significant focus of Latin American taxation was consumption, in particular the VAT. By the early 1980s, some ten Latin American countries had introduced a VAT. By 1994, almost all of them had implemented one (Table 6.6).

Table 6.6.

Value-Added Tax: Percentage Rates1

article image
Source: IMF.

Rates shown in bold type are so-called effective standard rates (tax exclusive) applied to goods and services not covered by other especially high or low rates. Most countries use a zero rate for a few goods, and Ireland and the United Kingdom use it extensively to ensure that substantial amounts of goods and services are free of the VAT.

Supplementary VAT rates of 8 percent and 9 percent on noncapital goods imports; through “catch-up,” these can revert to 18 percent retail.

Effective rate (legislated tax-inclusive rate is 13 percent).

On interstate transactions depending on region.

On intrastate transactions.

Venezuela was the last country to introduce a VAT in late 1993, but had removed it by March 1994.

In many instances, a rudimentary VAT had been introduced and was levied up to the manufacturing-importing stage or as a production-type VAT that disallowed credit for capital goods purchases. Many of them continued in this form through the mid-1980s. In the latter half of the 1980s and into the 1990s, Latin American countries began to reform their VATs, reducing the number of rates (Bolivia, Chile Colombia, and Mexico), expanding the base by reducing the number of previously exempted goods and covering a greater number of services (Argentina, Bolivia, Chile, Colombia, and Mexico), converting to consumption-type VATs (Argentina, Chile, Colombia, and Mexico), and improving administration.

The VAT played the most important role in improving the revenue performance of the tax system. To enhance revenue, those countries with a VAT measured evasion9 to assess the need for improvement, consequently strengthening VAT administration. Countries that experienced high increases in their tax-GDP ratios, such as Argentina and Chile, often achieved the increases with the VAT instrument.10 Most countries increased the rate of the VAT, owing, once again, to the leading role it played in generating revenue.

Commensurate with VAT reform, countries reduced the long list of excises, attempting to tax only such items as beverages, tobacco, petroleum products, electrical and electronic consumer durables, and automobiles. In the last few years, the excises on electrical goods have been rapidly disappearing, further shortening the list.

Taxes on International Trade

Finally, the late 1980s and early 1990s have been witness to fundamental reform of the taxation of international trade. All major countries have basically done away with export duties, which had been a mainstay of tax revenue a decade earlier. Customs tariff reform became common. The dispersion of tariff rates was reduced and their levels fell drastically. Tariff reform was carefully sequenced, with rates usually ending up within a band of 10–20 percent. Some countries hastened to reduce tax rates faster than originally planned (for example, Colombia), some introduced one tariff rate (such as Chile, with 11 percent), and some considered the elimination of import tariffs.

Revenue Consequences of Tax Reform

Table 6.7 presents data on the ratio of tax revenue to GDP of selected Latin American countries. It compares data from the early 1980s with data from the mid- or late 1980s and the 1990s. It divides the sample countries into three groups: high-tax (ratios of more than 20 percent of GDP); medium-tax (ratios of 15–20 percent); and low-tax (ratios of less than 15 percent). It estimates the tax revenue for the general government—rather than only for the central government—wherever possible. Total tax revenue has been broken down into different types of tax to the extent that such information was available.11 The, high-tax group includes Brazil, Chile, Costa Rica, Nicaragua, and Uruguay. The medium-tax group includes Argentina, Colombia, Ecuador, Mexico, Panama, and Venezuela.12

Table 6.7.

Cross-Country Comparisons: Tax-to-GDP Ratios

article image
article image
Source: IMF staff estimates.

Only Mexico City; does not include provincial tax revenues.

Contributions are classified as nontax revenue.

The low-tax group includes Bolivia, Guatemala, Paraguay, and Peru.13

Table 6.7 shows that most countries that carried out tax reform with the objective of improving the efficiency, equity, neutrality, and administrative feasibility (simplicity) of their tax systems also experienced a perceptible increase in their ratio of tax revenue to GDP, of between 2 percent and 4 percent of GDP, at least at the central government level, the usual target of tax reform. This outcome resulted whether, at the start of the reform, their ratio of tax revenue to GDP was low or high. It is understandable because a typical preoccupation of authorities undertaking reform was a downward slide in tax revenue, which they attributed to the tax structure that they wished to correct. Thus, even though tax reform need not necessarily be linked to a revenue increase, in effect, it invariably was.

Countries whose ratio of tax revenue to GDP is low initially might continue to have rather low ratios—compared with other countries—after implementing reforms, although their own ratio would increase.14 Thus, throughout the 1980s, and especially from the mid-1980s, most Latin American countries experienced increases in their tax revenue-GDP ratios at the same time as the tax reform was being implemented, a trend that seems to have continued into the early 1990s.15 As a result, high expectations regarding continuing increases in these ratios are unrealistic. Thus, to achieve fiscal balance, a much greater effort to curtail public expenditures may be necessary. This point will be taken up in the context of future directions for reform.

Anatomy of Tax Reform: Case Studies

The forces behind the tax reform and its main components in selected cases merit a more detailed analysis. Not all country cases can be considered here even though each would provide a unique view of the tax reform process. The countries being considered—Argentina, Bolivia, Brazil, Chile, Colombia, and Mexico—were selected with a view to discussing a variety of experiences, highlighting similarities and contrasts. Nevertheless, given the restriction of space, descriptions will focus on those aspects of the tax reforms that stand out as the most striking. In contrast, Brazil’s tax system is complex and provides a unique opportunity for examining in detail a case that demands attention.


Among all Latin American countries, Argentina’s tax reform probably took place in the most revolutionary manner. As late as 1989, the tax system was riddled with complexities. The VAT had multiple rates, its base had eroded, and it yielded less than 2 percent of GDP in tax revenue. An array of tax incentives ate into the base of the corporate income tax, and revenue from it had become negligible. Tax administration needed all-around improvements. These were just a few of the system’s shortcomings.

The 1990s have witnessed the most radical changes in Argentina’s tax structure and administration. In many countries, tax structure improvements are not necessarily accompanied by reform of the tax administration, the latter following slowly, if at all. The benefits from such an approach to reform are very often minimal. In Argentina, by contrast, the concurrent efforts made in tax policy as well as in tax administration were probably mainly responsible for the success of the reform. There was a phenomenal rise in the ratio of tax revenue to GDP. Between 1988 and 1992, it jumped from 11 percent of GDP to 16 percent of GDP.

VAT revenue jumped from less than 2 percent of GDP to nearly 6 percent of GDP. Evasion seems to have been minimized, buttressed by tax administration measures. Innovative and imaginative ideas, focused especially on VAT administration, have been tried. Businesses that fail to make timely or correct declarations are closed immediately—clausura preventiva—for three days (Durán and Gómez-Sabaini, 1994). In 1990, 700 taxpayers were penalized in this way. In 1992, the number rose to 12,000. This response had a strong impact on VAT compliance. New invoicing requirements and controls, as well as an increased VAT rate for when registered enterprises sell to nonregistered vendors, were also introduced. Improved information on VAT taxpayers, such as the number of sales to which the VAT applies, and relations between credits and debits also helped improve collection from other taxes.

A tax on gross assets was introduced as the basis for a minimum contribution by enterprises. This measure was intended to counter the erosion of the income tax base that had been caused by the widespread prevalence of tax incentives, indexation only on the side of accumulated liabilities and losses, and other factors. The income tax had basically degenerated to a withholding tax on interest. The introduction of the minimum tax, effectively eliminating the possibility of using accumulated losses to reduce tax liability, had a positive impact on income tax revenue, which rose from 1 percent of GDP in 1989 to almost 2.5 percent in 1992. Once revenue from the corporate income tax stabilized, the minimum tax was abolished in 1994. In terms of tax administration, criminal proceedings against enterprises that abused tax incentives—which had been found to be quite widespread—were begun. A large taxpayer unit was introduced to monitor and control more closely 1,000 or so taxpayers initially; it quickly expanded to include 150,000 (progressively smaller) taxpayers.16

International trade tax reform also helped improve efficiency. Quantitative restrictions on imports were eliminated in 1991, import duties were sharply reduced, and a minimum import duty was introduced. Export duties that had been used in earlier years to generate revenue were also abolished. Dynamic, younger personnel were hired for the customs administration. Deterrent measures were initiated, including a rolling publication naming importers whose declarations had been modified by the customs administration.

The total number of taxes was drastically reduced. Small taxes, such as stamp taxes, and particularly distortionary taxes, such as one on bank drafts, were abolished. The disappearance of small taxes made the tax system more transparent and easier to administer. Only a few taxes were necessary to generate much of the revenue.

In addition to the administrative measures implemented with respect to particular taxes, personnel reform in the tax administration was undertaken. Corruption and absenteeism, which had vitiated earlier attempts at genuine reform, were reduced by making it easier to fire and hire employees. To conclude, the change that occurred in Argentina was possible only because reform had the support of the highest authorities, who, with a concomitant basic change in philosophy, communicated to the general populace that tax evasion would no longer be tolerated.


Bolivia was one of those Latin American countries that had become dependent on petroleum as the main source of government revenue. Faced with falling international petroleum prices and a heavy external debt burden, it was further burdened by a complicated tax structure that led to the collapse of tax revenue in a hyperinflationary environment. The resultant rise in the fiscal deficit inevitably fed hyperinflation and made it impossible to achieve macroeconomic stabilization. Thus, in the context of an imperative need to correct overall economic policies, Bolivia embarked on fundamental tax reform in 1986 (see Harberger, 1988).

The tax reform could be said to have emphasized a drastic simplification of the tax structure, commensurate with the reality of tax administration standards. The focus of the tax structure became the VAT, with available resources concentrated on its administration. The VAT law was simple yet covered most goods and services, with few exemptions. Other taxes were given a minor role. The corporate income tax was abolished, and a 3 percent tax on the net worth of enterprises was put in place.17 The primary objective of the personal income tax—taxing all sources of individual income in an equitable fashion—became less important. Instead, it was given the role of acting as a “complementary regime” for the control of the VAT. Thus, the tax would be retained—at the same rate as the VAT—from different sources of income, such as wages, interest, and dividends, and the VAT paid on purchases could be used as a credit against such withholdings. Selected excises continued to be levied. The total number of taxes was drastically reduced so that tax administration could be better targeted. The result of the reform was that tax revenue from sources other than petroleum jumped from less than 4 percent of GDP in 1985 to over 8 percent in 1988. That ratio increased steadily to over 12 percent by 1993.

Nevertheless, revenue pressures at the time of the tax reform made it necessary to consolidate some of the earlier distortionary production taxes into a 1 percent tax on all transactions. Despite its cascading nature, its revenue generation capacity—yielding almost 1 percent of GDP for every percentage point in tax rate—subsequently led to its being increased to 2 percent.

Other difficulties with the tax system also seem to have crept in. First, little revenue is collected from the personal income tax, and most of that revenue is derived from foreign remittances. A secondary market for VAT receipts has developed, so that VAT receipts bought in this market are used for claiming credit for almost all of the personal income tax withheld from other sources of income. Second, some small nuisance taxes, such as a double tax on airline tickets purchased by Bolivians, have been introduced. Third, special regimes for preferred sectors, such as transportation—irrespective of the size of the business—ensure them a lower tax liability. With an effective 15 percent tax rate, the VAT currently yields less than 6 percent of GDP (and less than 5 percent if the petroleum sector is excluded). Finally, there remains the question of whether a country with an overall ratio of nonpetroleum tax revenue to GDP that is not high in a cross-country comparison can continue to avoid a regular income tax, especially if, as is typical, it is unable to withstand expenditure pressures from both the central as well as the lower levels of government.18

To conclude, the overall success of the Bolivian tax reform cannot be doubted. Simplicity was achieved. By crossing different taxes, the authorities introduced imaginative forms of control into a much reduced tax structure. The revenue objective was achieved. However, this success does not suggest that the time is not ripe for a revamping of the tax system with the objectives of removing remaining impurities and further developing the overall tax structure.


The Brazilian tax system continues to be very complex for a number of reasons. Brazil has made numerous attempts to modify its structure in recent years to achieve multiple objectives. Many special incentives are granted through the tax system by various government ministries at all tiers of government. These result in a wide range of effective rates of taxation that could be very different from the nominal rates. Despite the complexity of the tax structure, or perhaps because of it, its resultant burden is not low by standards of middle-income economies (23–24 percent in terms of GDP in recent years) and also reflects the generally high nominal rates of taxation.

Brazil depends on the taxation of production and consumption for almost half its tax revenue. Some of these taxes operate on the credit principle, while some are cascading taxes with negative effects on the efficiency of production. Indeed, the various Brazilian variants would translate into the equivalent of a 25 percent pure consumption VAT. The second largest generator of revenues is payroll taxes, which yield about one-fourth of total tax revenue. Income taxes yield somewhat less than one-fifth of total tax revenue. The personal income tax base has declined through legislative changes, while incentives have tended to erode the corporate income tax base. Because the economy has been relatively closed, import duties have not yielded much revenue. The share of international trade taxes in total tax revenues is less than 2 percent. Rural property is, in general, difficult to tax; urban property and services are lightly taxed by municipalities except in certain larger cities such as Sao Paulo, where reform has taken place. Thus, the amount of revenue derived from property taxes is negligible.

The VAT in Brazil is limited to certain sectors of activity. It uses a great variety of tax rates and exemptions. It does not allow credit for the purchase of capital goods in general but accords preferential treatment to selected investment goods and other inputs. Thus, it is neither a consumption-type nor a production-type VAT. Both the IPI (a multiplerate federal tax on manufactured products) and the ICMS (a state-level VAT) systems have become complex, and their economic and distributional effects are difficult to gauge. Professional services are taxed on their turnover through the ISS (a tax at the municipal level), with a negligible collection of revenue.

In general, the personal income tax remains an area of neglect in many Latin American tax systems. That of Brazil is no exception. Tax laws allow generous deductions from the tax base, suggesting that the personal income tax has a dormant revenue-raising potential. The exemption level in U.S. dollar terms is also one of the highest among medium- and high-income countries. Law No. 8383 of 1991 increased the low marginal tax rate from 10 percent to 15 percent. However, the exemption level was raised to 2.4 times per capita GDP—one of the highest in Latin America; 19 also, some deductions from the tax base were increased. This combination must have resulted in some revenue loss.

Five different taxes have been levied on corporate profits in recent years: an ordinary corporate income tax as well as a surcharge yielding an average effective tax rate of 30 percent, plus a social contribution tax, a withholding tax, and a state tax, which result in a high cumulative tax rate.20 The overall rate, which is high by Latin American standards, must be computed by taking all the taxes into account in the appropriate order. The rate for the first $150,000 a year in profits is about 40 percent and about 50 percent for amounts that exceed this figure.

The administrative cost of dealing with five taxes on corporate profits plus several other taxes and surcharges could be assumed to be quite high for taxpayers. All taxes have their complexities, are calculated on different bases, and are paid at different times. Many medium-size firms have a special department just to administer tax assessments and payments.

Argentina, Mexico, and Peru have implemented a minimum business tax based on gross assets. Those schemes usually impose a benchmark, a minimum contribution to the corporate tax, that is levied on either gross income or gross assets. Other Latin American countries are considering a similar approach. Brazil has not yet attempted to legislate a minimum corporate tax based on gross assets.

In Brazil, many sectors are restricted from foreign investment. There are limits on the acquisition of rural property (no more than 50 defined units of rural land), participation in banks (30 percent of voting shares), the petrochemical industry (one-third of shares), and transportation (limited participation in water transportation). Foreign ownership of newspapers, radio, and television is prohibited. The computer market is basically reserved for Brazilian companies. Road transportation of freight is restricted to Brazilian individuals or to corporations whose voting capital is at least four-fifths Brazilian. Without government permission, no enterprise can be established in the 150 kilometer-wide border zone unless it can prove that at least 51 percent of its capital belongs to Brazilians and that two-thirds of the workers and the management are Brazilian.

The current rate of taxation of financial intermediaries’ income is over 60 percent. This high rate may be viewed in the context of excess inflationary profits generated under high inflation. However, with a decline in inflation, the tax on most financial transactions would play a negative role. Even under high inflation, the use of monetary policy, such as raising the reserve ratio on demand deposits, is preferable. Brazil considered introducing a tax on bank debits. Such a tax may in general be expected to burden financial intermediation and could damage the financial system.

Brazil’s various social security contributions represent about one-tenth of GDP and about one-fourth of total revenue of the general government. The flow of resources among the different agencies and programs is directed through a multitude of funds, which add to the complication of a system characterized by fractioning and partial overlapping of responsibilities. In summary, the system has become increasingly complex, raising the standard compounded corporation tax rate significantly. Both the top marginal rates of the payroll tax structure and the average tax rate on workers’ incomes are high by international standards. Also, the very base of the tax—payrolls—tends to be strongly procyclical, thus contributing to the creation of financial problems for the social security system during economic slowdowns and downturns.

Imports as a share of GDP, at less than 5 percent, are quite small. The government recently undertook tariff reform, announcing a stepwise reduction of the average unweighted tariff at an accelerated pace, to slightly over 14 percent, while reducing the maximum tariff to 40 percent. However, unlike the norm in a growing number of countries, including Argentina and Mexico, Brazil has no minimum import duty. A minimum duty can be justified on several grounds: it is efficient in that it provides uniform effective protection, it minimizes import misclassifications, and it serves as a tax handle as imports rise. The authorities recently also curtailed explicit export taxes although a few are still levied at a low rate, for example, on hides and skins and unprocessed minerals. There is also implicit export taxation in that credit for two federal turnover taxes earmarked for social expenditures is not allowed to exporters. Except for those on windfall profits, direct and indirect taxes on exports result in a fall in the international competitiveness of Brazil’s exports. There is therefore no role for export taxes in a modern taxation environment.

The scope of fiscal federalism in Brazil is unique among countries with similar levels of income. As may perhaps be expected in a large country, rules of tax assignment do not always conform to commonly accepted economic criteria. Revenue sharing among tiers of government involves complicated formulas for various combinations of tiers (for example, union-state, union-municipality, and state-municipality). Different consumption taxes are assigned to different tiers of government; their revenues are used for general budgetary expenditures, for earmarked expenditures—mainly for social security benefits—and for revenue sharing among the different tiers. Income taxes are put to general as well as specific use and are also shared among government tiers. Payroll taxes are used entirely for social security.21

Recent trends toward greater fiscal decentralization have eroded the union’s control over the use of total tax revenue. The stabilization capacity of federal finances is thus weakened. The high proportion of shared tax revenue has also tended to limit the union’s efforts to improve the yield of shared income and encouraged instead the use of alternative, unshared, but often distortionary sources of federal revenue. At the same time, the wide scope of unconditional vertical tax transfers and limited decentralization of expenditure responsibilities could have led local governments to underutilize their own revenue bases, such as user charges and property taxes. Current fiscal federalism arrangements have also generated tax competition. The states’ operation of the ICMS will gradually erode the tax base through special treatments and incentives.

In general, local jurisdictions can enhance their tax revenue efforts by reducing the scope of unconditional vertical tax transfers, decentralizing expenditure responsibilities, setting, at a national level, minimum property tax rates and valuation methods, and providing technical assistance on tax administration matters to the smaller municipalities. To increase the efficiency and yield of property taxes, tax schedules should be as simple and flat as possible, and the tax base should be automatically linked to an inflation index. Provided that standards of property taxation are defined nationally, the administration and proceeds of the rural property tax could be viewed as a local government jurisdiction.

To conclude, it seems clear that Brazil is at an ideal juncture to undertake fundamental tax reform. Given the country’s continental dimensions, however, the scope of the effort has to be grand, not only touching upon the structure of the tax system itself, but also carefully determining tax assignment rules within its overall fiscal federal structure.


Over the years, Chile has used the tax system for specific objectives. Major tax policy changes in 1984 reflected an ongoing economic strategy from the previous decade toward greater emphasis on private sector activity. The objectives of the changes seemed to be focused on encouraging savings of individuals and enterprises. Average tax rates were effectively reduced through integration of the corporate and personal income taxes as well as through simplifying the income tax, while income tax incentives were provided for stock market investments. Although many of the changes concentrated on income taxes, the VAT base was expanded and some small taxes abolished.

The common view of income tax incentives is that they are not very effective for generating savings and investment. Recent, selected studies on Chile arrive at a similar conclusion that the incentives seemed to have generated some short-run gains in savings and economic activity, although an actual reduction in the tax rates may have been more beneficial in the longer run (Toro, 1994). Also, the incentives gave rise to some important loopholes that particularly benefited the banking sector, which was able to skim off interest differentials through its lending operations to incentive-seeking industry. In sum, the household sector, which historically saved little, did not change its savings behavior, while the enterprise sector increased reinvestment through retained profits, more as a result of increased economic stability than as a result of tax changes. It is noteworthy that between 1986 and 1989, the economy grew by more than 7 percent a year.

In 1990, another set of tax policy changes followed, with the objective of improving equity, reflecting the Chilean authorities’ desire to consolidate the benefits of economic growth by increasing social expenditures on health, education, and housing. Their specific goal, which was to increase revenue by US$600 million, resulted in all-around increases in the rates of major taxes, including an increase in the VAT rate to 18 percent from 16 percent. Recognizing that these changes might lead to increased evasion, the authorities also stepped up administrative measures, focusing mainly on intensified auditing. The result in terms of revenue was impressive. From less than 15 percent of GDP in 1989, domestic tax revenue increased to over 18 percent in 1993. In terms of the VAT, a rate of 18 percent produces 9 percent of GDP in tax revenue. Thus, each 1 percentage point of tax rate now generates about 0.5 percent of GDP—one of the highest yield rates in the world. It is also worth mentioning that international trade taxes were reformed, with export taxes being abolished and a single import tariff being imposed at 11 percent.

To conclude, the real rate of growth of the Chilean economy between 1990 and 1992 continued to be remarkably high—nearly 6.5 percent annually. First, the increased tax rates, the higher social expenditures, and the fortification of tax administrative measures did not seem to curb economic growth. Rather, the overall economic environment seems to have been responsible for supporting economic growth.22 Second, Chilean tax policy cannot be said to have conformed at all times to the general philosophy of simplification, base expansion, and rate reduction. Sometimes, it reflected other economic strategies and resulted in an income tax system with high tax rates and differentiated taxation and with preferred sectors and individuals identified within the system. To what extent those strategies were helped by manipulating the tax system remains open to question. The system’s success in revenue generation depended in large measure on a broadly based VAT and an efficient tax administration supported by a small, efficient staff.


Tax reform in Colombia has been a continuous effort over a number of decades. Many special commissions have been set up to examine broad or specific aspects of the tax system. Colombia’s highly qualified tax professionals and the authorities’ open welcome of cross-fertilization of ideas attracted the attention of international tax experts, and many commission reports were submitted to the Colombian government.23 Important reform-oriented measures were taken in 1983, 1986, 1990, and 1992–93, which gradually improved the country’s tax structure.

The objectives of the 1983 reform (like those of the 1984 reform), were to encourage economic activity through the reduction of double taxation and tax incentives. Among the measures actually taken were wide-ranging improvements in various taxes that simply attempted to correct some of the structural deficiencies of the prevailing tax system. The personal income tax structure underwent changes, such as a reduction in tax rates to reduce “fiscal drag.“At the same time, an increase in the personal exemption level reduced the number of taxpayers and facilitated tax administration. Other changes included monetary correction for financial incomes of individuals and presumptive taxation for commerce and financial intermediation. The tax rate for limited companies was reduced (multiple corporate income tax rates prevailed), and double taxation was eliminated, although various incentives were granted. The sales tax was converted to an income-type VAT, and the tax base was extended to include several previously exempted services.

The 1986 Colombian reform (like the 1990 Chilean reform) attempted to achieve greater neutrality and equity, especially through restructuring the income tax, as well as to improve tax administration. Individual income tax rates were further reduced, 90 percent of salary earners began to pay tax through retentions only, corporate income tax rates were unified, and attempts were made to correct any continuing biases in the debt-equity ratios of companies caused by the tax system. On the administration side, the reform simplified tax declaration forms, introduced tax payments through banks, thereby releasing many tax officials for other functions, and created large taxpayer units.

The 1990 tax policy changes were carried out in the context of economic restructuring and modernization. They focused on promoting savings, improving the capital market through, among other measures, the encouragement of repatriation of capital, and opening up the economy through customs tariff reform. Actual measures included a reduction in the corporate income tax rate, but the authorities also saw fit to introduce special provisions with the objectives mentioned above. Thus, they introduced a much lower tax rate for income from repatriated capital, exempted income tax from stock market incomes, and halved the withholding tax rate on repatriated income from foreign capital. The VAT rate was unified at 12 percent and its base was further expanded. The import surcharge applicable to all imports was reduced to 13 percent from 16 percent (eventually to be reduced to 8 percent in 1994), while the average tariff was reduced to 7 percent from 16 percent. On the administration side, a national directorate for taxes was created, with a view to consolidating tax evasion strategies. This action was complemented by a further, drastic reduction in the number of taxpayers who needed to file declarations.

Tax policy changes continued in 1992–93 with the objective of lowering the fiscal deficit in the face of inflation, which hovered stubbornly near 30 percent. Tax rates tended to be increased. An income tax surcharge of 5 percent was introduced to fund national security expenditures. Public commercial and industrial enterprises, mixed enterprises, public funds, and financial cooperatives became taxable. However, the withholding tax on foreign remittances was reduced, with the intention of reducing it gradually to 7 percent by 1996. The rate of the VAT, which was converted to a consumption-type VAT to reduce cascading, was increased to 14 percent for five years and its base was broadened. Import restrictions were virtually eliminated. On the administration side, domestic taxes and customs were consolidated into one directorate to improve cross control and facilitate anti-evasion measures.

In summary, the Colombian tax policy changes can be characterized as incremental, combining particular short-run goals, such as the encouragement of savings, investment, and capital inflow, and medium-term goals consisting of corrective features to remove distortions, improve equity, and simplify the tax system.24 Colombia experienced no revolutionary change—such as those that occurred in Argentina or even Bolivia—its style resembling more the Chilean-style reform, which was based on a wide scope of tax policy mixes. Nevertheless, Colombia’s success with the VAT has been somewhat less complete than that of Argentina or Chile, and, in its dependence on income taxes, it can be compared with Mexico. As in all the other countries being examined, the tax policy changes progressively increased the ratio of tax revenue to GDP over the years under consideration.


Mexico is another country in which tax reform is neither a recent nor a onetime phenomenon but can be characterized as an ongoing process. The VAT and fiscal federal arrangements were developed in the early 1980s, while further modernization of the tax system—responding to changing economic conditions—and even sophistication—based on theoretical or conceptual merit—have been achieved in the 1990s. Just as in Bolivia in the mid-1980s, Mexico in 1986–87 faced rapid inflation, high real interest rates, and a heavy debt-service ratio together with a fiscal deficit-GDP ratio of nearly 16 percent. Among other policy improvements, a fiscal strategy was engineered to achieve macroeconomic stabilization through improved revenue productivity as well as to induce economic growth through eliminating inefficiencies in the tax system. The result is that the Mexican tax structure and its administration are markedly different today from the way they were a decade ago.

The tax system had become a mixture of ad hoc arrangements, as manifested, for example, in piecemeal statutes for monetary correction. Consequently, distortionary investment decisions abounded, raising the debt-capital ratio. Additional deductions for losses in real asset values were permitted, which, in combination with a nominal interest deduction for enterprises, led to a depletion of real revenues. In the tax reform that followed, Mexico’s experience seems to have occurred in stages: tax policy reform and the innovative administration reform that followed later. This experience presents an interesting contrast to that of Argentina, where, as described above, those two facets of reform proceeded together.25 Further, in Mexico, subsequent reductions in consumption tax rates accompanied the implementation of a social contract among government, workers, and entrepreneurs.

During the first period (1986–88), Mexico’s focus was on recovering the real value of tax revenues and reducing the distortionary elements of the tax system through appropriate policy measures. Advance monthly payments for the income tax, the VAT, and excises were required for the first time. Indexation was introduced: only real interest could be taxed or deducted, depreciation was adjusted for inflation, inventories were immediately expensed, capital contributions by stockholders were adjusted, and enterprises were allowed to index previous years’ losses.

The changes were made with some caution, however. Because enterprises were heavily indebted, an immediate monetary correction would have led to a sudden rise in their tax base. Such a move could have caused liquidity problems. As a result, a transition period of five years was introduced for the full correction to be achieved. First, the corporate tax rate was effectively reduced to 35 percent from 42 percent over the period. Second, a gradual increase was made toward the full use of the new expanded tax base.

Interestingly, some cash-flow properties were also built into the corporate income tax without inciting reaction from foreign governments. Thus, with the objective of recuperating private investment, enterprises were allowed to deduct fully—without authorization—the present value of the depreciation in new investments in fixed assets (at a discount rate of 7.5 percent), while a whole class of distortionary tax incentives was abolished.

The top personal income tax rate was reduced to 50 percent from 55 percent, and the standard deduction of one minimum wage was converted to a tax credit of 10 percent of that wage. The authorities, however, continued to experiment with the VAT, retaining more than one rate, which reflected distributional and developmental objectives.

The emphasis during the second period of tax reform (1989–92), following a change in government, was on containing evasion mainly through control measures. The authorities held the view that tax rates should not be increased so that economic growth would not be hampered, while the stabilization goal would remain intact. Furthermore, as a result of the previously implemented monetary and other corrections, the new corporate tax base turned out to be considerably smaller than originally expected, and the private sector preferred to adopt the new tax base in full immediately. This was granted in 1989, increasing the need for compensatory revenue measures that would be focused on base expansion through improved control and administrative mechanisms.

The first such measure was a 2 percent gross assets tax that would act as a minimum contribution to the corporate income tax, immediately yielding almost ½ of 1 percent of GDP in revenue in 1989 (similar to that of Argentina).26 The second was a rapid expansion in the number of taxpayers, which increased from 13 million to 18 million between 1988 and 1992. The number of taxpaying enterprises increased more than 100 percent. The third measure was a reorientation of special regimes. These had originally been created to facilitate the administration of small taxpayers but had become enclaves for privileged sectors—agriculture, livestock, fisheries, transportation, and publishing—which, irrespective of the size of operations, had become immune from any revenue contribution. These regimes were simply replaced by an innovative cash-flow tax that simplified tax calculations but had a higher revenue potential. Also, there were few international repercussions of the cash-flow tax given the sectors that were affected.

Administrative improvements included redefining several types of tax-related behavior—such as delays of more than six months in filing taxes or committing acts with the purpose of obtaining unjustified fiscal benefits—as tax fraud and, thus, criminal acts. As a result, the number of taxrelated criminal cases jumped from 3 during 1980–88 to 300 during 1989–93. The tax administration was also thoroughly revamped, with increased integration of various functional areas, such as collection and audit, rapid automation, and random selection for audits and customs checks.

Nevertheless, some important tax policy changes took place in this period. First, the top individual income tax rate was reduced to match the corporate income tax rate of 35 percent, while fringe benefits were taxed at higher rates than was cash income. Second, the integration provisions for the corporate and individual income taxes were simplified to reduce the burden at the individual level. Third, the corporate tax was brought closer to the cash-flow tax by allowing immediate deductions for new investments outside Mexico City, Guadalajara, and Monterrey and immediate expensing of inventories. Whether some of these complicating features, the last one in particular, necessarily constitute reform is debatable.27 Nonetheless, sincere attempts were being made to simplify tax calculations for many taxpayers and accounting for tax purposes in general.

Interestingly, in Mexico—in contrast with Argentina—much of the focus of tax reform was on the income tax rather than on the consumption tax. When the two countries are compared, this focus is also observed in the differences in the relative shares of income and consumption taxes in total tax revenue. Thus, beginning in 1989, the VAT on selected goods and services, such as nonalcoholic beverages, telephone services, and insurance premiums, was eliminated, and the general VAT rate was reduced to 10 percent from 15 percent, while the preferential rate of 5 percent on food and medicine was reduced to zero.28 Excises on automobiles, alcoholic beverages, and tobacco products were slashed, and the mineral extraction tax was eliminated. It is not surprising, therefore, that the revenue productivity of the Mexican VAT has not been high; it hovered at less than 4 percent of GDP until 1991 when the VAT rate was still 15 percent (it was reduced to 10 percent in November 1991) and fell to less than 3 percent of GDP in 1992 with the decrease in the tax rate. In contrast, the revenue from income taxes increased steadily from just over 4 percent of GDP in 1987 to nearly 5.5 percent of GDP in 1992.

In conclusion, the selected country experiences demonstrate a certain common philosophy of relying on tax reform to achieve such objectives as simplicity, neutrality, and overall efficiency in the tax structure. Nevertheless, there were perceptible differences among countries in the use of particular tax instruments to achieve, for example, greater savings and investment. Differences also appeared as economies matured at different rates. Those that developed faster became more concerned with equity and international compatibility and with modernizing the administration of their tax systems.

Future Issues and Directions for Reform

Although Latin American countries have significantly reformed their tax structures, important tasks remain to be addressed. These tasks take various forms and emerge from different directions. First, their tax structures retain certain elements that are reminiscent of archaic, distortionary systems and should be corrected. Second, as the Latin American economies grow and become more open and integrated into the world economy, their tax systems will be expected to incorporate more modern features. Third, much of the fiscal reform that has already taken place reflects a much greater effort on the tax than on the expenditure side, a problem that becomes exacerbated as new tax assignment rules tend to allocate greater resources to lower levels of government. Therefore, there seems to be a strong need for the reform of fiscal federalism rules.

Remaining Distortionary Elements

In every tax system, even after major reform has taken place, some distortionary elements are likely to remain or to creep in because of successful lobbying or populist promises made by politicians or to meet the authorities’ urgent revenue needs. Latin America is no exception. The technician’s role under those circumstances is to work steadfastly toward eliminating the distortions. The purpose of this section is not to list all such elements in Latin American tax systems but rather to provide some examples of how otherwise neutral tax systems may become subject to inefficiencies.

To generate revenue in the short run, Argentina introduced a tax on financial transactions—essentially on checks. The tax had an initial positive revenue impact but soon encouraged financial transactions to go underground, for example, through the multiple use (endorsement) of a check. The tax was eventually repealed. The news of the initial success of the tax in yielding revenue, however, had a significant demonstration effect in neighboring countries. A strong movement for a unique tax on checks—to replace the tax system—received credibility, albeit temporarily, in Brazil. Quite a few other countries continue to consider the tax quite seriously. Needless to say, such a tax has the potential to inflict lasting damage on the domestic banking system. For example, if many agents switched to foreign accounts, it would become convenient to open an account abroad and to transact with foreign checks. The tax could also reduce the overall attractiveness of financial markets and hamper tax auditing by driving transactions underground.

Despite the almost universal implementation of the VAT by Latin American countries, some country authorities burden the system with additional distortionary taxes on production or transactions. Bolivia’s 2 percent cascading tax on all transactions, for example, is highly revenue productive—as might be expected because there is no credit mechanism—but distortionary. Other countries may levy excises, including on small household appliances, that are difficult to administer. A few countries levy other nuisance taxes for fear of losing even the smallest amount of revenue.

These types of tax seem to thrive, owing, sometimes, to campaign promises not to increase the VAT rate or even to repeal the VAT. Indeed, many statements have recently been made against the VAT. Interestingly, opposition seems not to lie with the VAT itself because, in the countries where the statements are generally made, the VAT bases have typically been eroded to accommodate various interest groups, while the tax rate of the VAT might have had to be sequentially increased to compensate for the base erosion. It is obvious that the VAT should not be eliminated but rather that its base should be extended. Those countries in which the VAT base is broad and the rate is judged to be at a ceiling may do better to explore the feasibility of expanding the role of income taxes instead of introducing additional distortionary taxes on production.


At least those countries of Latin America that can be said to be modernizing very rapidly will need, in the not too distant future, to give due consideration to the tax systems of industrial countries. Three areas that need attention are the taxation and administration of incomes, property, and agriculture.

Many Latin American countries, at the early reform stage, quite rightly simplified both their tax structure and tax administration, focusing mainly on the VAT. As their tax systems and economies mature, however, many of them continue to generate insignificant revenue from income taxes. The time may have come for many of them to consider reforming, rather than eradicating, income taxes. It is not necessary to link an income tax with a global approach, based on compulsory declarations, right from the start. A beginning could be made with a simple, even schedular, structure, using withholding and retention mechanisms. Such an approach should facilitate administration of the tax and make it more productive.

In almost all Latin American countries, property taxation has been virtually ignored except, perhaps, in certain large cities, particularly those in fiscally federal states. Typically, the rate structure is unrealistic, with high tax rates, while the base remains uncorrected for inflation sometimes for decades. There seems to be little reason not to introduce in Latin America the model used by various local governments in the United States, which relies on published sale values to determine the assessed values of different properties rather than on a completed cadastral survey. The model that has been used to assess properties and the administration of the property tax in Sao Paulo also merits examination for broader applicability.

Essentially, the same argument may be made for agricultural taxation. Any modern tax system must include at least a simplified mechanism for taxing agriculture appropriately. Some Latin American countries are at the economic threshold at which they should seriously consider such a mechanism.

A number of European countries—Finland, Netherlands, Norway, and Sweden—have already implemented a carbon tax for environmental considerations, and other countries are seriously considering one. The tax variants are on sulfur emissions, sulfur dioxide, chlorofluorocarbons, heavy metals, chemical fertilizers, pesticides and herbicides, phosphates, and methane, as well as the “classical” taxes on mineral oil, natural gas, and coal. There are proposals for taxes on specific forms of industrial and agricultural waste and waste produced by the construction industry, as well as taxes on tropical wood, plastic bags, aluminum cans and foil, batteries, light bulbs, and water consumption. Most member countries of the Organization for Economic Cooperation and Development have expressed their willingness to reduce carbon dioxide emissions, as manifested by their participation in the 1988 World Climate Change Conference in Toronto and the 1992 United Nations Environment Conference in Rio de Janeiro.

The rationale for a carbon tax is as follows. First, scientists have linked carbon emissions to a global warming phenomenon. Second, studies have correlated carbon emissions and energy prices inversely. An appropriately designed carbon tax should therefore help curb global warming. It is time that Latin American countries legislate the means to reduce emissions. In this context, it is surprising that, as their economies have been opened and import tariffs reduced, including those on automobiles, many authorities have continued to reduce domestic excises on automobiles. The introduction of a carbon tax would correct this anomaly and help modernize the tax structure.

Tax Assignment and Fiscal Federalism

The Latin American experience with fiscal federalism, for example, in Argentina and Brazil, 29 reveals that, over time, the lack of reform, unfinished reform, or reform that is countered by policies precipitated by short-term exigencies may increase the burden on higher-level governments. This outcome may, in turn, lead to mounting difficulties for the federal government in its macromanagement functions. Sometimes even the onset of reform may increase the burden on the federal government, as explained below.

Important taxes in terms of revenue productivity—such as income and consumption taxes—are usually subject to revenue sharing. Typically, the federal government has sole control over the revenue from international trade taxes. With tariff reform, as average tariff rates have fallen and the tax bases have diminished, the federal government has had to look elsewhere for a steady source of revenue.

As revenue shares from shared taxes have tended to increase for lower-level governments, federal governments have devised unshared taxes that are often inefficient, such as the one mentioned on financial transactions. These taxes can be temporary at best and are unable to alleviate the revenue problem at the federal level in the long run.

As the revenue performance of shared federal taxes has improved, lower-level governments have had windfall revenue gains and therefore have little incentive to improve their own revenue performance. Thus, it may be necessary to require a minimum revenue effort by lower-level governments before federal funds are dispensed to them. The challenge for the federal government would be to adhere to such requirements, once introduced.

A certain volatility caused by a lack of specific rules has characterized the distribution of tax revenue. However, distributional changes have usually had the twofold effect on lower-level government shares of increasing the strain for macromanagement and causing their revenue efforts to stagnate. The lack of rules has also plagued certain categories of expenditure.

Tax assignment and revenue-sharing reform take place faster than expenditure assignment reform. This has essentially implied that taxes get assigned to lower levels (and revenue shares of lower levels increase) faster than the rate at which larger expenditures are assigned to lower-level governments. Indeed, intergovernmental negotiations for expenditure assignment tend to leave large responsibilities to the federal government, generally in excess of what its revenue share would imply.

The common experience has been for lower-level governments to be able to finance their expenditure through provincial banks virtually without constraint. Thus, reform may be carried out at the level of revenue sharing and tax assignment. Nevertheless, unless financing reform stipulates a ceiling on lower-tier financing, fiscal federal reform cannot be complete.

To conclude, with greater decentralization of tax revenue, greater decentralization must follow for expenditure responsibilities. Furthermore, given a tax and expenditure assignment mix, guidelines are needed for deficit financing rules if serious macromanagement is not to be jeopardized. While local governments have to be provided some budgetary discretion so that the efficiency gains of decentralization can be obtained, an argument may be made for local governments to earn that discretion by relying on their own tax efforts. To the extent that they are dependent on federal financing, general equilibrium and dynamic stability of the macroeconomy must eventually entail a limit on free financing of local government deficits. Thus, first, lower-level governments should be required to report their finances regularly to the next higher level, and, second, their borrowing capacity must be curbed.

Summary and Conclusions

Much of the prereform tax policy in Latin America over the past decade reflected the need for basic economic reforms that were initiated in the face of hyperinflation, stagnating or negative economic growth, expanding balance of payments deficits, and difficulty with pursuing sufficiently tight monetary policy. The concomitant growth in fiscal deficits resulted in correction, especially on the tax side, because governments were unable to reduce expenditures quickly owing to high interest costs or wage bills that were politically difficult to curtail. In that event, authorities often found that tax policy was a relatively easy instrument to wield and used it to the maximum extent possible.

Many countries initiated, and followed through with, important tax policy reform. While steadfastly keeping the revenue goal in mind, as manifested in a steadily increasing tax revenue-GDP ratio to the extent of 2–4 percent of GDP over a few years, countries that undertook reform typically revealed a number of strong trends: personal income tax rates and their dispersion went down, the tax began to be applied at higher incomes in terms of GDP, fiscal drag was corrected, corporate income tax rates were unified and decreased, withholding rates on foreign remittances were lowered, capital gains began to be treated as ordinary income, it became more common to consider or even require a minimum contribution to the corporate income tax as an anti-evasion measure, the VAT structure was improved and its base universally expanded, excises became more selective, export taxes were generally abolished, and the customs tariff structure and dispersion were scaled back.

On the tax administration side, imaginative measures were taken in many countries although, in others, improvements lagged behind. Closing businesses and treating more evasion-related behavior as criminal acts became important deterrent measures. Personnel reform and structural consolidation of tax and customs administration made it possible to induct officials who took bold steps, supported at the highest political levels, toward curbing evasion. As a result, the general population began to believe that taxes—recently rationalized and often lowered—had to be paid to avoid high penalties.

Despite the common trends, there was considerable variation in individual country experiences in conducting particular aspects of tax policy. While in some countries, neutrality of the tax system was a stated objective, in others, income taxes were used at particular times with the express purpose of encouraging savings and investment.30 Colombia and Mexico, for example, relied considerably more on income taxes than did Argentina or Bolivia, which focused on the VAT, while Chile emphasized both areas. The revenue productivity of the VAT in Chile is among the highest in the world today. Bolivia simplified its tax structure more than any other country and, in a relative sense, perhaps experienced the most dramatic revenue increase.

As Latin American economies became more integrated with the rest of the world, their tax systems responded to varying degrees, for example, with respect to the withholding tax on foreign remittances. In addition, with foreign investors in mind, Argentina and Mexico attempted to increase the sophistication of their income tax structures by incorporating international issues, such as transfer pricing, a foreign tax credit, and so on.

A key reason for country differences has been the fact that a critical mass in favor of tax reform has been slow in forming in Brazil and Venezuela. Thus, these countries have been slow to respond to pressure to carry out fundamental tax reform based solely on the evidence from neighboring countries. Clearly, the dynamics of political consensus buttressed by support from the highest political levels are essential for the implementation of tax reform.

The reforming countries did not always succeed in eliminating completely inefficient or inadequate features of their tax systems. Some that remain are the inadequate taxation of property and agriculture. Moreover, some inefficient elements have been introduced into the tax structure with the sole purpose of generating revenue. For example, Argentina introduced a tax on financial transactions, and Bolivia and Brazil considered one; Bolivia continues to implement a distortionary and cascading turnover tax at a 2 percent rate that covers all manufactured goods; and small taxes, including on business net worth, continue despite their failure to yield much revenue. Complex fiscal federal relations in large countries like Brazil tend to hamper the implementation of efficient tax assignment principles, for example, in the case of the VAT.

In conclusion, the directions in which Latin American countries should look to conduct their tax policies are clear. One important area is that of fiscal federal relations and the determination of rules for tax assignment to the various levels of government. Decentralization must also proceed for both taxes and expenditures if the objective of a low fiscal deficit is to be pursued.

Second, as economies mature, they must consider to what extent the administered tax system resembles the legislated tax structure. For example, to simplify tax administration, potential taxpayers often have to be ignored or eliminated from the functioning ambit of certain taxes, such as the VAT or income taxes. It is important, therefore, to consider to what extent the universe of taxpayers can be increased over time to include more and more so-called minor taxpayers. Third, and relatedly, while the role of “large” taxpayer units in improving auditing and revenue performance should not be minimized, the rate at which the coverage of these units can be expanded—if not merged with the general taxpayer population—should certainly be considered an important criterion for the measurement of the maturity of a tax system. Fourth, and in the same vein, the rate at which the use of tax amnesties declines may be taken as an indicator of the maturity of the tax administration.

Fifth, the two areas that have been largely overlooked—property and agriculture—cannot go untaxed for too long if the tax structures of the more advanced of the Latin American countries are to become fully comparable with those of industrial countries. Sixth, it would seem feasible for Latin American countries to give serious consideration to selected new concepts in modern taxation, such as eco-taxes (those levied for ecological purposes, such as the carbon tax). These would comprise some of the important elements of future tax reform that may be envisaged as the Latin American countries approach the twenty-first century.


Miguel Rodríguez

Mr. Shome’s paper offers an excellent analytical summary of the progress made over the past decade in tax reforms in the context of structural adjustment programs in Latin America. It also describes with great lucidity the future challenges facing governments with respect to taxation.

As we know, efforts aimed at restructuring the public sector and finding a solution to structural fiscal instability have become the cornerstone of the structural reforms being successfully implemented by several Latin American countries. The sharp reduction in inflation in Argentina, Bolivia, Chile, and Mexico is the result of the extraordinary progress these countries have made in restructuring their public sectors and in eliminating or substantially reducing their fiscal deficits. Mr. Shome points out how important tax reform has been in improving these countries’ fiscal situation and alerts us to the fact that the delay in implementing appropriate tax reforms in countries such as Brazil and Venezuela explains, in large measure, the persistent fiscal and macroeconomic instability in their economies.

Venezuela’s case perhaps best illustrates the importance of the role of fiscal adjustment and tax reform in economic reform. After the successful launch of its economic reform program in 1989, when trade liberalization, early efforts to restructure the public sector, rationalization of exchange and monetary policies, and debt renegotiation fueled sharply accentuated growth and reduced inflation until 1992, the Venezuelan economy plunged into a severe recession. The recession was accompanied by a resurgence of inflation and the onset of a crisis in the financial sector, which quickly led to capital flight. At the root of this grave macroeconomic instability lies a sizable structural fiscal deficit (more than 10 percent of GDP) owing to the failure to implement an appropriate tax reform to diversify the government’s non-oil revenue base. In a country with a high potential for noninflationary growth, the absence of a solid tax base has led to a marked fiscal imbalance, which is hurting the economy and threatening to cancel out all of the gains achieved as a result of the liberalization effort.

Mr. Shome’s paper stresses the major progress achieved in tax matters in the region and the positive effects on the adjustment process. First, he reviews the progress achieved in this past decade with the establishment of more rational and efficient tax systems, which are decreasingly distortionary and in which collections have increased significantly. In some countries, collections increased dramatically owing to a more efficient tax environment, laying the groundwork for further progress. The basis of the reforms (which, of course, took on different forms in different countries) was the simplification of the personal and corporate income tax structure, the introduction of a minimum tax on corporate income (such as the tax on assets), the introduction or restructuring of consumption taxes (basically the value-added tax), and customs tariff reforms. However, the countries that have made the most progress in developing taxes and that have achieved relatively high ratios of tax revenue to GDP are those that were able to undertake major changes in their tax codes. These countries clearly defined tax violations and penalties and have relied on dynamic finance ministers, who have strictly enforced the new tax laws as is done in industrial countries.

But the most significant observations in Mr. Shome’s paper pertain to the agenda for the future. First, it is necessary to build upon the efficiency gains in tax administration and legislation by eliminating the sources of distortion that still exist in most countries. In extreme cases, such as Venezuela’s, the authorities intend to abolish the VAT—which was approved only last year—and to introduce a highly distortionary tax on bank debit balances that will foster financial disintermediation, in connection with a thoroughly political economic policy that is likely to delay even further the tax reform and the resulting stabilization of the economy. Fortunately, most countries are past this phase, and the general trend seems to be toward greater progress in eliminating these distortions and promoting substantially efficient tax systems.

Second, progress in the major decentralization process occurring in several countries must be accompanied by an appropriate delegation of spending authority and sharing of tax responsibilities among the various regions and government authorities. All of these efforts would lead to greater efficiency in fiscal administration and keep the fiscal deficit under control.

Last, besides suggestions for modernizing taxation in areas such as the environment, I believe that the proposals to develop the property tax scheme and to increase income tax collections are the most important ones on the tax agenda that Mr. Shome suggests for our countries. Through an institutional effort that stresses improved administration and the use of legal mechanisms to ensure that these taxes are collected, it is possible to achieve elsewhere in Latin America results similar to those that Argentina and Chile obtained with the VAT or that Mexico achieved with the income tax.

Nevertheless, a major effort with respect to the property tax and the personal income tax would help our countries adopt a progressive tax structure, with positive results on redistribution. If, at the same time, Latin American countries also adopted a progressive social spending program, long-term productivity would increase. One of the reasons income is so unevenly distributed in Latin America is the governments’ extremely regressive fiscal policy, which has played a role in maintaining, or, rather, in aggravating the appalling distribution of wealth inherited from our feudal colonial past. Traditionally, government action in the fiscal area has been based on distortionary, regressive tax policies—with the wealthiest segments of the population paying insignificant amounts—and on a public spending policy that, associated with a populist discourse, has led to large subsidies and the transfer of income and wealth to the most privileged sectors in our countries. One of the fundamental challenges Latin America faces in restructuring its public sector—building upon the rationalization achieved in recent years—is to promote a progressive tax policy on both revenue and expenditure, thereby helping to set a pattern of economic growth through a more equitable distribution of income and wealth.

To that end, our countries will have to pattern themselves increasingly after the industrial countries, with respect to government action in the fiscal realm. On the revenue side, in addition to sustaining substantial collections from the consumption tax, revenue must also be obtained effectively from income and property taxes, with the wealthy bearing a significantly higher burden. On the expenditure side, we should invest intensively and efficiently in education and health and establish the infrastructure essential to development. In a population that is generally healthy and well trained for productive purposes, such investment would reduce the sharp disparities in relative wages that are found in Latin America and that account for 70 percent of the regressive income distribution in our countries. This is precisely what the industrial countries have done: they have invested heavily in human capital using the resources obtained from progressive taxation, endowing their populations with the ability to engage in productive work. These steps, in turn, have had a decisive impact on the level of real wages and relative remuneration, and thus on income distribution. Similar trends can be observed in the successful, newly industrializing economies in Asia, which have used progressive tax policies, encouragement of savings, and major educational efforts to establish, in their growth and development process, much more equitable distribution patterns than those found in Latin America.

These remarks on the government’s recommended fiscal actions lead me to comment on a proposition reiterated by Mr. Shome in his paper, namely, the need to persist in efforts to reduce public spending as a part of long-term fiscal adjustment. I disagree slightly, perhaps because of my familiarity with Venezuela’s experience, but also because of my conversations with many Latin American ministers and officials and my observation from time to time of the Latin American expenditure figures and how they compare internationally. Although I fully agree that public expenditure should be rationalized and every effort made to eliminate unproductive expenditure, I believe that it is going to be very difficult for many countries to reduce public expenditure further. On the contrary, when we achieve greater efficiency in the allocation and execution of expenditure, that is, when we are spending the appropriate amounts on education, health, infrastructure investment, social programs, and environmental protection, I suspect that we will have to increase public spending in comparison with current Latin American practice, including in relatively more developed countries. For this reason, despite the great progress that several countries have already achieved, we must step up our tax efforts in the future so as to be able to provide noninflationary financing for forward-looking public spending programs that are consistent with our foremost development objectives.

Last, I believe that such a comprehensive, high-quality work as Mr. Shome’s should have devoted a portion of the analysis to social security, an important component of the reform that Latin America must undertake, and to its possible repercussions in fiscal matters, promotion of savings, and macroeconomics in the region’s countries.


Luis Viana

The paper by Mr. Shome is an interesting survey of recent changes in tax policy in Latin America. It is a rich paper that stimulates discussion about tax issues. The main results seem to show that the reforms that have taken place in most of the countries were steps in the correct direction. From the paper, it can be concluded that tax reforms improved tax neutrality at the same time that tax revenues increased and the inequality of the tax system was dampened.

Based on cross-country evidence on tax rates and country case studies, the paper concluded that tax neutrality was achieved through

  • a lower dispersion of tax rates,

  • a reduction in corporate and income tax rates,

  • an increase in the tax base, mainly as a consequence of tight controls and enforcement of penalties on tax evasion and the elimination of tax exemptions, and

  • the replacement of cascading turnover taxes by a value-added tax.

After reading the paper carefully, I am left with the impression that perhaps the paper was too ambitious in tackling so many issues at the same time. Even though it is mainly a survey, many times I had difficulty evaluating the main conclusions about tax neutrality, equality, and other aspects of tax administration. These three objectives are not necessarily achieved without any cost. What is missing from Mr. Shome’s paper is an explicit consideration of the trade-offs among the objectives, which should be put forward to balance the discussion.

For example, we can conclude that per capita taxes may be the most neutral tax or, at the same time, the most regressive. We can also infer that increasing the tax base on corporate taxes will increase tax neutrality, but that, simultaneously, neutrality can be achieved only at a high administrative cost. Furthermore, it might be argued that the trade-offs among neutrality, equality, and revenue-enhancing objectives must be taken into account if we are to evaluate the evidence correctly.

The second gap in Mr. Shome’s paper is the lack of a discussion of how the objectives can be achieved independently. The criteria for evaluating tax neutrality are based mainly on uniform tax structures and a broad tax base. In this paper, little attention has been paid to the effects of the overall tax structure. First, social security taxes have not been considered, and, second, the inflation tax and implicit taxes raised through public enterprises have not been taken into account. Some conclusions for inflation taxes and implicit taxes on public enterprises can be inferred insofar as stabilization and privatization policies were pursued during the period under review. However, analyzing tax neutrality without discussing social security and inflation taxes in Latin America may be misleading.

In addition, the impact of the tax structure on the structure of industry is underestimated. The tax system may influence this structure in at least two ways: first, by discriminating among sectors, and, second, by discriminating among the sizes of firms within a sector.

Both of these effects appear to be extremely important for private sector development in Uruguay.1 These were the main results of a survey of two hundred export firms that attempted to identify barriers to private development. Information was stratified by firm size and exporting sector. The results of the survey showed that the tax burden was one of the most important obstacles for the private sector, in particular for large firms. Small firms not only enjoy tax advantages through the tax system, but are also less likely to have their taxes audited. Social security tax evasion was estimated at 3 percent of GDP, while evasion of the value-added tax was estimated at 1.8 percent of GDP.

The above findings also underscore the importance of presenting information other than marginal tax rates to evaluate tax systems. For example, information on average tax rates to compare with marginal tax rates might have been useful. Such a comparison would have revealed a truer picture of the tax base. Average tax rates can be estimated from tax collections and potential tax bases. Comparisons by sector and by the size of the firms within a sector would have been useful for evaluating tax policy.

Finally, the paper raises two issues concerning tax inequality on which I would like to elaborate. Both of them are related to tax incidence considerations. First, no explicit comment is made in the paper about the implications of capital taxation, both property and corporate taxes. The common view is that capital taxation improves equality; in my view, it does the opposite. For a small open economy with a given rate of return on capital, taxes, in the long run, are paid by low-skill workers when investment falls. Capital taxes not only depress growth, but at the same time fall on low-income groups.

Second, the paper by Mr. Shome implicitly favors income taxes over consumption taxes as a tool for improving income equality. This view presumes that savings are never going to be spent or ignores the existence of a life income profile. I believe income distribution policies should be addressed with focalized expenditure policies rather than with tax policy.


  • Aguirre, Carlos, and Parthasarathi Shome, “The Mexican Value-Added Tax (VAT): Methodology for Calculating the Base,” National Tax Journal, Vol. 41 (December 1988), pp. 543 –54.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Bird, Richard M., “Tax Reform in Latin America: A Review of Some Recent Experiences,” Latin American Research Review, Vol. 27 (1992), pp. 7 –36.

    • Search Google Scholar
    • Export Citation
  • Bird, Richard M., and Milka Casanegra de Jantscher, Improving Tax Administration in DevelopingCountries (Washington: International Monetary Fund, 1992).

    • Search Google Scholar
    • Export Citation
  • Durán, Viviana, and Juan C. Gómez-Sabaini, “Lecciones sobre Reformas Fiscales en Argentina: 1990-1993,” paper presented at the sixth Comisión Económica para América Latina y el Caribe (CEPAL) seminar, Santiago, Chile, January 1994.

    • Search Google Scholar
    • Export Citation
  • Fundación de Investigaciones Económicas Latinoamericanas, Hacia una NuevaOrganizacion del Federalismo Fiscal en la Argentina (Buenos Aires, Argentina: FIEL, 1993).

    • Search Google Scholar
    • Export Citation
  • Harberger, Arnold C., “Introduction,” in Bolivia 1952-1986, by Jeffrey Sachs and Juan A. Morales (San Francisco: International Center for Economic Growth, 1988).

    • Search Google Scholar
    • Export Citation
  • Larrañaga, Osvaldo, “Casos de Exito de la Política Fiscal en Chile: 1980-1993,” paper presented at the sixth CEPAL seminar, Santiago, Chile, January 1994.

    • Search Google Scholar
    • Export Citation
  • McLees, John A., “Fine Timing the Mexico Assets Tax,” Tax Notes International, Vol. 3 (February 1991), pp. 117 –20.

  • McLure, Charles E. Jr., “Analysis and Reform of the Colombian Tax System,” Working Paper in Economics No. E-88-15 (Stanford, California: Stanford University, Hoover Institution, 1988).

    • Search Google Scholar
    • Export Citation
  • Musgrave, Richard A., and Malcolm Gillis, Fiscal Reform for Colombia; Final Reportand Staff Papers of the Colombian Commission on Tax Reform (Cambridge: Law School of Harvard University, 1971).

    • Search Google Scholar
    • Export Citation
  • Organization for Economic Cooperation and Development, “Mexico’s Recent Tax Policy,” a note by Mexico, Committee on Fiscal Affairs, Meeting of June 30-July 1, 1993 (Paris: OECD, 1993).

    • Search Google Scholar
    • Export Citation
  • Prud’homme, Remy, State and Local Finance and Public Policy in Brazil, OEIL Report for World Bank, University of Paris XII (Paris, June 1989).

    • Search Google Scholar
    • Export Citation
  • Shah, Anwar, “The New Fiscal Federalism in Brazil,” Discussion Paper No. 124 (Washington: World Bank, 1991).

  • Shome, Parthasarathi, “Fiscal Federalism—Revenue, Expenditure, and Macro Management,” inaugural address at the sixth CEPAL seminar, Santiago, Chile, January 1994.

    • Search Google Scholar
    • Export Citation
  • Shome, Parthasarathi, “Trends and Future Directions in Tax Policy Reform: A Latin American Perspective,” Bulletin for International Fiscal Documentation, Vol. 46 (September 1992), pp. 452 –66.

    • Search Google Scholar
    • Export Citation
  • Shome, Parthasarathi, ed. (1995a), Tax Policy Handbook (Washington: International Monetary Fund).

  • Shome, Parthasarathi, ed. (1995b), Comprehensive Tax Reforms: The Colombian Experience, Occasional Paper No. 123 (Washington: International Monetary Fund).

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Shome, Parthasarathi, and Christian Schutte, “Cash-Flow Tax,” Staff Papers, International Monetary Fund, Vol. 40 (September 1993), pp. 638 –62.

  • Tanzi, Vito, Tax Notes Today (July 27, 1993).

  • Tanzi, Vito, and Parthasarathi Shome, “The Role of Taxation in the Development of East Asian Economies,” in Taxation and Development in East Asian Countries, ed. by Anne O. Krueger and Takatoshi Ito (Chicago: Chicago University Press, 1992), pp. 31 –61.

    • Search Google Scholar
    • Export Citation
  • Tanzi, Vito, and Parthasarathi Shome, “A Primer on Tax Evasion,” Staff Papers, International Monetary Fund, Vol. 40 (December 1993), pp. 807 –28.

  • Toro, Juan, “Aspectos Exitosos y Lecciones de Reformas Tributarias en Chile,” paper presented at the sixth CEPAL seminar, Santiago, Chile, January 1994.

    • Search Google Scholar
    • Export Citation
  • Torres, Fabio Sánchez, and Catalina Gutiérrez Sourdis, “Casos de Exito en Reformas Fiscales: Colombia, 1980-1992” (January 1994).


This has been the case, in particular, for revenue from the value-added tax (VAT) on imports.


This experience has been common also among Asian countries. On the role of tax policy in the development strategies of these countries, see Tanzi and Shome (1992).


For selected country studies, see also Bird (1992).


For a description of recent trends in tax reform and IMF recommendations in tax policy, see Shome (1995a).


The difference in the upper income bracket in terms of per capita GDP in the two country groups should represent, to an important extent, the different income levels of the two groups of countries. But it also probably reflects measurement problems in hyperinflationary environments. However, there is considerable variation even within the two groups.


The instability in the treatment of capital gains was apparent in a similar table that compared information between 1980 and 1990 in Shome (1992). There, other taxes were also similarly compared. While the trends between 1980 and 1990 seem, in general, to continue for the period beginning the mid-1980s to the early 1990s (which is the focus of this paper), some important differences emerge, for example, stability in the treatment of capital gains.


Bolivia abolished its corporate income tax in 1986 and did not tax capital gains either.


For a full view of tax administration in developing countries, see Bird and Casanegra (1993).


A beginning was made with the case of the Mexican VAT (Aguirre and Shome, 1988). Many other countries followed with similar estimation procedures.


A few countries, such as Mexico, successfully used the income tax instrument to increase revenues.


Nevertheless, the information should be used only for purposes of broad comparisons, given the multiple sources used and the recency of some of the data.


It is clear from the cross-country data that those countries that depend heavily on revenue from petroleum have not had to push significantly in the direction of raising the taxation of the nonpetroleum sector. Examples are Ecuador and Venezuela, which are in the medium-tax group because they rely on petroleum as the main source of tax revenue.


These classifications are affected to the extent that the ratios may fall within one group in one period and another group in another period.


An exception is Bolivia, which moved recently from the low-tax group to the mediumtax group, or Costa Rica, which moved from the medium-tax group to the high-tax group.


An exception would be Guatemala, which, in the absence of reform, has experienced a lowering in its tax revenue-GDP ratio. Peru has recently introduced reform measures, with an observable recovery in the ratio since its collapse during the heterodox period.


For coverage of various factors that lead to tax evasion, methods to measure it, and means to control it, refer to Tanzi and Shome (1993).


The income tax was abolished also in Uruguay as part of a tax reform that focused on the VAT as the leading tax.


Tanzi (1993) has recently argued that the time may have come in Latin America to renew efforts to legislate and administer a properly structured income tax.


In general for Latin America, targeting about 20 percent of the population for the personal income tax net may be administratively feasible. This would reduce the exemption level to 1.6 times per capita GDP in the case of Brazil.


The state tax has recently been repealed.


For a detailed description of the system, see Shah (1991).


This argument, which may be seen in the same vein as that of Toro (1994) for the 1984 reform, is made by Larrañaga (1994) for the reform of 1990.


For further elaboration on the ramifications of tax reform, see Torres and Gutiérrez Sourdis(1994).


For an elaboration on the operation and method of calculating the liability of this tax, see McLees (1991).


For the pros and cons of a cash-flow tax, see Shome and Schutte (1993).


However, exempted purchases made in worker unions’ stores and the lower tax rate applicable in the border regions—zona frontera— were eliminated, and interest on credit card and financial consumer loans was included in the tax base.


Such a policy did not always turn out to be successful, however.


Gustavo Michelin and L. Viana Martorell, “Desarrollo de la Actividad Privada en el Marco del Proceso de Regulacion” (Bogota: Centro de Estudios de la Realidad Economica y Social, 1992).