I would like today to discuss selected issues in the design and implementation of trade reforms based on our experience with adjustment programs supported by the use of Fund resources. It would be useful at the outset to examine the recent record on adoption of trade reform measures in developing countries and in economies in transition. We surveyed 78 Fund-supported programs with 59 countries approved in 1990–93 under Stand-By Arrangements, the Extended Fund Facility (EFF), the Structural Adjustment Facility (SAF), and the Enhanced Structural Adjustment Facility (ESAF).1 Measuring trade regimes, especially quantitative restrictions (QRs), across such a large group of countries is rather difficult. The assessment of QRs was based on the import and export coverage and the intensity of the respective QRs. The tariff regime was specified in terms of average tariff levels inclusive of various other duties and charges that add to the cost of imports. The tariff and QR regime was combined to provide an indication of the overall trade policy stance. Countries were ranked in terms of whether their trade regimes were relatively open, moderately restricted, or very restricted.
The survey revealed that substantial progress in trade reforms took place during 1990–93 among program countries. The trend toward reduced reliance on QRs observed in the 1980s intensified in the 1990s. The greatest progress in eliminating QRs was made by economies in transition in Eastern Europe and the Baltic countries where state involvement in trade was reduced sharply, and by program countries in the Western Hemisphere. Substantial progress was made in Africa, particularly where exchange system reforms eliminated QRs used for the allocation of foreign exchange. Progress on tariff reform was less rapid. Western Hemisphere countries made significant progress, as many of them adopted ambitious liberalization programs to restructure and open up their economies. Although many countries in Africa, Asia, and the Middle East lowered tariffs—in some cases by large magnitudes—the reductions were only in a very few cases large enough to classify their tariff regimes as “open” or “moderately restricted.” The combined (QR plus tariff) trade regime showed that at the end of 1990, 53 countries had restricted trade regimes, 5 were moderately restrictive, and only 1 had a relatively open regime. By the end of 1993, 33 countries were classified as restrictive, 22 had a moderately restrictive stance, and 4 had relatively open regimes.
The most rapid overall progress in trade reform was recorded in Eastern Europe and the Western Hemisphere. According to certain macroeconomic indicators, and without inference of causality, the survey showed that slower trade reformers as a group initially faced more difficult macroeconomic conditions (in terms of current account and fiscal imbalances, reserve levels, and dependency on trade taxes) compared with the faster reformers. The latter countries improved their fiscal and reserve positions significantly, reduced reliance on trade taxes while containing the deterioration in current account positions, and simultaneously undertook trade reform. The group of slower trade reformers experienced some improvement in their current account position, but fiscal and reserve positions did not improve, and reliance on trade taxes declined by only a small amount. The macro-economic characteristics of the slow and fast reformers suggested that fast reformers were more willing or able to undertake bolder, more comprehensive reforms.
Let me now turn to design issues. We have generally found that the overall trade reform effort is best addressed in the context of a medium-term strategy with clearly established and announced immediate and medium-term objectives. This helps avoid uncertainty regarding the direction of trade policy and provides economic agents clear signals for decisions on investment and production. Adhering to preannounced targets establishes the credibility of the reform effort, which is important for its success. Furthermore, trade reform is likely to be more successful if accompanied or preceded by complementary macroeconomic measures (e.g., fiscal, exchange rate) and structural measures (e.g., exchange system liberalization, price decontrol, privatization, measures to facilitate labor mobility).
Within the trade regime, program design has given priority to eliminating QRs early in the reform process, as these are less transparent and more restrictive than tariffs. A number of countries have considered a one-step removal infeasible and have opted for a phased reduction. In such cases, care should be taken not to bunch all of the most sensitive items toward the end of the phasing period, as this is often not credible. Targets for the phased reduction of QRs need to be precisely defined and monitorible. Fund-supported adjustment programs have often targeted a declining import-coverage ratio, or the number of product groups to be subject to QRs at each stage (the choice of an appropriate base period would be particularly important). When tariffs replace QRs, they should not exceed the tariff equivalents of the QRs and should be gradually reduced over time.
With regard to tariffs, for most developing countries and economies in transition that cannot affect the foreign currency prices of their imports, theoretical considerations call for zero tariffs across the board. In practice, however, we know that governments do not pursue the objective of pure welfare maximization, and tariffs are used for revenue, income distribution, balance of payments management, and protecting domestic industries. With the exception of the balance of payments, these objectives suggest that the theoretically optimal tariff structure should be differentiated. However, practical considerations suggest otherwise. With regard to the protection objective, the difficulties and pitfalls of targeting suggest that it is preferable to adopt the alternative approach of broader-based protection with relatively narrow differentiation among sectors. With regard to other objectives, considerations of political economy, administrative convenience, and lack of information also provide strong arguments against complex and differentiated tariff structures. In particular, if interest groups perceive that the authorities are unwilling to provide too much differentiation in protection, they may refrain from further lobbying to secure greater protection, thereby minimizing rent-seeking costs. Less complex tariff structures can be administered more easily, avoiding cumbersome paperwork and reducing the incentive to misclassify products.
These considerations have shaped the design of tariff reform in Fund-supported adjustment programs. Such reform aims for simplicity, maximum transparency, and eventually the ideal of a relatively low and uniform tariff structure. This involves moving from specific to ad valorem tariffs, amalgamating “other duties and charges” into the tariff structure, eliminating discretionary and other exemptions, reducing the number of tariff rates to a few, lowering the maximum tariff with only a few exceptions, and reducing the average tariff. In order to mitigate the antiexport bias due to high tariffs, a duty-drawback system should be implemented, whereby import duties paid on producer inputs that are physically “consumed” in export production are rebated to domestic manufacturers upon export of the goods physically embodying those imports. Unfortunately, duty-drawback schemes in many developing countries are often administratively cumbersome and inefficient.
The extent and speed of trade reform would naturally need to take into account the initial position (degree of distortions) and be tailored to individual country circumstances. Some countries have opted for gradually phased reductions of very high tariff levels and dispersion (e.g., a number of countries in Asia), while others have managed faster reform (e.g., in Eastern Europe and Latin America). The experience of successful tariff reformers shows that it is possible within a period of 2–5 years to reduce tariff bands to 3–5, to bring high tariffs (in some cases, from triple digits) initially to a maximum of 30–35 percent and then subsequently to 20 percent or less, and to reduce average statutory tariffs to about 15–20 percent initially and to about 10 percent in the subsequent stage.
Finally, I would like to talk about two particular design issues—fiscal and exchange rate—that we have repeatedly encountered in the design of Fund-supported adjustment programs.
Fiscal considerations (namely, generating and maintaining revenues) are a major determinant in limiting the magnitude of trade reforms and the speed of their implementation. Although trade taxes are not optimal instruments to raise revenues, in practice, developing countries and economies in transition rely heavily on them. A gradual approach to tariff reform has been adopted in a number of countries where trade taxes account for a high proportion of tax revenue (e.g., Bangladesh, Sri Lanka, and Zimbabwe). Since Fund-supported adjustment programs generally target the central government budget (and trade taxes tend to accrue to the central government), revenue considerations can drive tariff policy even when trade taxes account for less than 10 percent of total tax revenues (e.g., Argentina). But the concern about the revenue impact of a tariff reform may not be well founded in all cases. A well-designed tariff reform may actually improve the import tax collection rate at the same time as statutory tariff levels come down, at least in the initial stages. This is because in many cases tax collection rates are below statutory tariffs reflecting the widespread use of exemptions; and smuggling, tax evasion, misclassification, and corruption are sometimes encouraged by very high tariffs.
Even so, there have often been conflicts between short-term fiscal objectives and medium-term trade reform goals. The lesson from our experience with Fund-supported programs is that program design needs to explicitly recognize the link between fiscal and trade policy. It is essential that tariff reform be coordinated with tax reform so that revenue effects are anticipated. Tariff reforms can be effectively implemented and sustained, particularly in cases where the central government’s budget is restructured early in the reform process to decrease the relative importance of trade taxes in total revenue. Thus, a clear timetable for the development of alternative domestic (nontrade) revenue sources is often needed to ensure the success of tariff reforms.
Appropriate exchange policies are essential to ensure that trade reforms are consistent with balance of payments objectives. Exchange and trade restrictions often act as substitutes. A prerequisite, or at least a complementary measure, of effective trade liberalization is the adoption of liberal exchange systems. The effectiveness of exchange rate policies is itself enhanced when complemented by the liberalization of the trade and payments regimes. Where trade restrictions are pervasive, the equilibrium exchange rate is likely to be well above its restriction-free level. Under such circumstances, trade liberalization may lead to a deterioration in the current account, as the boost to imports, including private consumer demand, is likely to be felt more rapidly than the impact on exports. In the absence of adequate reserves, a significant reduction in trade barriers would normally need to be accompanied by a devaluation to safeguard the balance of payments. Of course, if the trade reform improved confidence so as to generate capital inflows in sufficient amounts, a devaluation may not be necessary, at least for immediate balance of payments reasons. Domestic industries, however, will need time to adjust to the new realities of a major trade reform and a devaluation may still be necessary to avert future balance of payments problems. Where the initial trade restrictions are not pervasive, or trade liberalization is confined to specific input sectors, a devaluation may not be necessary provided that the improvements in competitiveness deriving from cheaper inputs are translated into higher exports quickly enough.
In countries experiencing a real exchange rate appreciation due to surges in capital inflows, or where the nominal exchange rate serves as an anchor but inflation remains higher than programmed, pressures for protection are likely to rise. Ideally, the real appreciation should be tackled with appropriate monetary and fiscal policies and not by resort to trade restrictions.
In general, using trade instruments for nontrade objectives is a second-best solution. Trade policy needs to be geared to medium-term efficiency goals. However, there may be emergencies where better alternatives are not available and deviations for short-term fiscal and balance of payments may be needed. If trade restrictions are inevitable, they should be limited and kept strictly temporary. The phased elimination of such temporary restrictions is an important element in the design of trade reform.
Demetris Papageorgiou and Michael Michaely
This paper draws on the findings of an extensive research project that we co-authored with A. Choksi, on trade liberalization experiences.1 That research suggests criteria for successful—that is, sustained—trade liberalization policies and attempts to assess their short-term economic consequences. It assesses the potential for success of the recent wave of trade liberalization in Latin America. For Eastern Europe, the emphasis is on describing the current trade regime and outlining the necessary elements for trade reforms.
Most of this paper discusses the process of trade liberalization in Latin America, which within the last ten years has implemented profound policies to liberalize its trade regime. In this period, several countries in the region have embarked upon policies of trade liberalization unprecedented in scope and intensity in recent history. Trade liberalization was only a part of a wholesale reorientation of public policy toward a market-oriented economic environment, in the midst of macroeconomic calamity and stagnant economic growth. It was carried out by democratically elected governments with clear popular support.
In the first part, we summarize the findings of our research: What are the elements that lead to sustained trade liberalization, and what is its immediate economic impact? In the second and third parts, we discuss the evidence of the recent liberalization policies in Latin America and countries of the former Soviet Union and assess the extent to which they comply with the findings of the 1991 research. The discussion of Latin American trade liberalization is richer in empirical evidence and different in substance, than that of countries of the former Soviet Union, because the circumstances are radically different. The evidence in Latin America is confined to the unilateral trade liberalization and does not cover the several preferential trade agreements in which some Latin American countries have entered, including NAFTA, and Southern Cone Common Market (Mercosur).
Successful Liberalizations and Economic Costs
Despite countries’ fears that trade liberalization will have high economic costs, it has been found that adjustment costs incurred by successful trade liberalization are surprisingly small.2
Elements of Successful Liberalizations
The research by Papageorgiou, Choksi, and Michaely was broad in scope. It analyzed the course of liberalization in 19 countries, all of them market economies to varying degrees, during 36 distinct episodes of trade liberalization. It covered, by and large, every attempt at significant trade liberalization undertaken by developing countries from World War II until 1984, depending on the country. The results indicate that despite the complexities of the trade reform process and the differing circumstances that confronted each attempt at liberalization, the successful (sustained) programs shared common elements:
Momentum. Programs that started boldly and followed through with further measures proved more durable than ones that took a more hesitant approach.
Reduced quantitative restrictions (QRs). Programs that decisively reduced or eliminated such direct interventions as import quotas generally succeeded. Those that did not generally failed. This was one of the study’s clearest findings.
Competitive real exchange rates. Most of the successful programs began with a depreciation of the (nominal and real) exchange rate. Thereafter, no particular trend was clearly associated with success—but most programs that lasted avoided sharp fluctuations of the real rate.
Prudent macroeconomic policies. In general, the successful reformers kept their budget deficits smaller in relation to GDP than the others. In fact, reversals of trade liberalization were more commonly associated with poor macroeconomic policies than with any other factor, including the power of vested interests and short-run unemployment.
Proper sequencing of reforms. Programs tended to go wrong when capital-market liberalization preceded trade liberalization.
Political stability. Reforms, once introduced, were difficult to sustain against a background of unstable government.
One of the research’s most important findings is that strong programs at the initial stage lasted longer than weak ones. Nineteen trade liberalization episodes were classified as strong and 17 as weak.3 Five of the strong programs (26 percent) and seven of the weak ones (41 percent) were reversed. If anything, these figures understate the superiority of the strong reforms. The research defines episodes of liberalization to exclude those that collapsed within two years. If these unsuccessful cases had been included, they would have fallen overwhelmingly in the weak category. The golden rule of sustainable reform, therefore, appears to be that bold is best.
The research also found that trade liberalization programs that stay in place for six years are most likely to be sustained for many years. This could be because a six-year span normally involves a transition from one government to the next. Such transition usually presents the strongest challenge to the reforms. In addition, a new set of vested interests usually emerges from a more efficient economic environment that provides the needed political resistance to reverse the liberalization process.
Almost always, highly illiberal trade regimes use QRs extensively. Relaxing these restrictions has been the main content of sustained liberalization programs. Might other factors explain the results? For example, did the programs that included a major relaxation of QRs also cut tariffs substantially? The answer is no. Programs with radical QR reforms varied widely in their tariff reforms: sometimes tariffs were cut, other times they were raised, and in still others they were left unchanged.
The link between the depreciation of the real exchange rate at the implementation of trade liberalization and sustained reform seems to be strong in every case. Most of the trade liberalization episodes that were fully sustained experienced a real depreciation at the outset of the reforms. The remainder experienced no change in the real exchange rate. In no case did a fully sustained reform start with a currency appreciation. Most of the partially successful programs also began with a real depreciation, while those that failed were evenly divided between those with a rise or fall in the real exchange rate. The association is only less clear for the trend change in the real exchange rate during liberalization, and at its final stage.
The relevant role of macroeconomic policy in liberalization is to keep inflation low and thus maintain a real depreciation. One striking fact from the research is that only a single episode of trade reform was fully sustained after being introduced alongside an expansionary fiscal policy But the nature of fiscal and monetary policy at the beginning of trade reform does not discriminate between the partially sustained programs on the one hand and outright failures on the other.
The stance of macroeconomic policy at the end of the reform episodes seems to play a bigger role in the success or failure of reform. Sustained reform was usually accompanied by either a restrictive or neutral macroeconomic policy. If fiscal policy was expansionary and monetary policy accommodating at the end, the programs as a rule collapsed altogether or, at best, were only partially sustained. The research indicates that expansionary fiscal and monetary policies are the single most important cause of a reversal of trade reforms.
Countries have embarked on trade liberalization in a wide variety of economic circumstances. How did these differing initial economic conditions relate to the ultimate outcome of the reforms? At the risk of oversimplifying, initial conditions might be grouped under three broad headings: economically distressed, stable, and intermediate cases. The research findings indicate that liberalization programs that began under great economic strain tended to be strong, fast, and therefore relatively durable. The programs undertaken in more or less stable economic circumstances, and especially those that followed earlier successful episodes, also fared quite well. The in-between cases—reforms that began amid signs of economic deterioration but not a fullblown crisis—were the least likely to succeed.
Trade liberalization has usually been undertaken alongside other sorts of economic reform—deregulation of international capital flows, in particular. This raises the question of whether the ordering of such a package of reforms has any effect on its chances of success. The evidence from the research on this question is not as well based as the findings on trade reform proper for two main reasons. Capital market liberalization was beyond the main area of the research and few of the episodes of trade liberalization covered by it have been accompanied by capital market reforms (which were not very common until recently). The research investigated four such occurrences. This evidence is narrowly based and far from conclusive. As far as it goes, however, it supports the presumption that extensive capital market liberalization should be left until trade liberalization is well under way, but certainly should not precede trade liberalization.
Are there other factors that affect the reform not related to the design of the reform itself or to accompanying economic policies? The research investigated several such factors: economic size and income level of the countries, geographic size, and resource endowments. In this respect one factor does stand out: political stability. The main criterion was political continuity, which need not mean one continuous government over a long period, or democratic or authoritarian governments.4
Short-Term Economic Costs
The short-term economic effects of trade reform are of great concern to governments. Trade reform will succeed only if it shifts resources from inefficient uses to new tasks. The bigger the projected long-term gains, the greater the shift will need to be. And big long-term, expected gains may require strong, bold trade reforms, with a pronounced effect on reallocation of resources. But this short-term reallocation carries economic costs and political risks.
Four major concerns are normally voiced in contemplating trade reforms: balance of payments deterioration, fall in national output, rise of unemployment, and deterioration of income distribution. The research looked in detail at these concerns and found the fears of high costs misplaced. To a surprising extent, the costs of adjustments are indicated to be very small, even in the short term.
In most liberalization episodes, countries increased their foreign exchange reserves in the months immediately following the start of their program. In most cases, liberalization rescued the country from falling reserves. Indeed, a rapidly worsening external position was quite often the reason why the reforms had been implemented in the first place.
In about half of the episodes, imports did increase in the months after liberalization began. Most of these involved episodes with major relaxation of QRs. Except in a few cases, the other half showed no trend, either upward or downward; in four episodes, imports actually fell.
Because imports, by and large, either showed no trend or increased, the overall improvement in balance of payments positions must have been the result of rising exports. And so it was. Taking an average of the 31 episodes for which good trade data were available, the annual growth rate of exports for the three years preceding the liberalization was 4.2 percent, while the annual growth rate for the year of liberalization and subsequent three years increased to about 11 percent in constant U.S. dollars. What may be more surprising is that the improvement in export performance was both quick enough and big enough to bring about an immediate improvement in the balance of payments. The speed of reaction may partly be the outcome of speculative movements of (recorded) trade flows in response to largely anticipated devaluations.
How did national output fare in the short term? If anything, trade reform—especially strong reform—is associated with higher growth from the beginning. For all episodes, the average growth rate of GDP in the year preceding the reforms was 4.4 percent and the annual rate for the period including the liberalization and the three subsequent years was 5.6 percent. For the strong liberalization episodes, the corresponding rates were 3.5 percent and 5.4 percent, respectively, and for the weak episodes 5.6 percent and 5.7 percent.
On the short-term effects of liberalization on employment, the data suggest that they have been small. This evidence is based not on all episodes, as detailed data were not always available. Nevertheless, in most countries covered by the research, employment was not lowered even in broad individual sectors of the economy, such as manufacturing. This suggests that the reallocation of labor was achieved largely within sectors, causing less disruption than might have been feared.
It could be that the main concern of government in planning for trade liberalization is the short-term effects of the reforms on income distribution rather than GDP. The fear is that whatever the costs of adjustment, they will fall disproportionately on the poor. This risk would be great if the short-term costs in lost output and employment were large. But the evidence indicates that the short-term effect of liberalization on national output was strongly positive, and on employment at worst neutral.
The evidence on the precise effect of liberalization on income distribution is mixed and fragmentary. It leaves the issue open. There is no solid evidence that low-income groups gain particular benefits from liberalization (beyond sharing in the aggregate benefits). Equally, there is no support for the popular view that reform is bound to make the poor worse off.
Recent Liberalizations in Latin America
The recent wave of trade liberalization in Latin America began in 1985 with Bolivia and Mexico, followed by Uruguay in 1987, Argentina and Venezuela in 1989; and by Brazil, Ecuador, and Peru in 1990 (see Table 1). Several other countries in Latin America have also undertaken trade liberalization (Costa Rica in 1986, Guatemala in 1986, Honduras in 1990, Jamaica in 1985, Panama in 1991, Paraguay in 1989, and Trinidad and Tobago in 1985) but their reform process has been slow and uncertain. Furthermore, lack of comprehensive data for these countries makes it difficult to include their liberalization experiences in this overview of recent liberalization policies in Latin America.
Recent Episodes of Trade Liberalization in Latin America
Recent Episodes of Trade Liberalization in Latin America
| Argentina | (July 1989–April 1991) |
| Bolivia | (August 1985–February 1990) |
| Brazil | (July 1990–July 1994) |
| Colombia | (September 1989–December 1991) |
| Ecuador | (May 1990–August 1992) |
| Mexico | (July 1985–December 1989) |
| Peru | (August 1990–August 1991) |
| Uruguay | (July 1987–April 1992) |
| Venezuela | (February 1989-July 1991) |
Recent Episodes of Trade Liberalization in Latin America
| Argentina | (July 1989–April 1991) |
| Bolivia | (August 1985–February 1990) |
| Brazil | (July 1990–July 1994) |
| Colombia | (September 1989–December 1991) |
| Ecuador | (May 1990–August 1992) |
| Mexico | (July 1985–December 1989) |
| Peru | (August 1990–August 1991) |
| Uruguay | (July 1987–April 1992) |
| Venezuela | (February 1989-July 1991) |
Of the nine countries covered in this study, five (Argentina, Brazil, Colombia, Peru, and Uruguay) had attempted to liberalize their foreign trade in the 1960s and 1970s. The liberalization policy in Argentina, Brazil, and Peru collapsed altogether, in Colombia, it was partially reversed, and in Uruguay, it was sustained. For the other four (Bolivia, Ecuador, Mexico, and Venezuela), it was their first serious and persistent attempt to liberalize trade and reverse protectionist policies.
Although historical judgment may not yet be rendered, the several years that have passed since the implementation of these liberalizations do make it possible to observe a pattern of development. Could this pattern have been predicted by the inferences of the 1991 study?
A common thread that runs across most of these liberalizations is the economic circumstances under which they were implemented. All but Colombia and Venezuela faced severe macroeconomic problems. And in all but Colombia liberalization was one of several structural changes, in addition to macroeconomic stabilization.5 Indeed, trade liberalization was an integral part of the governments’ policies in Latin America for reducing their role in the economy, privatizing extensively public assets, and stabilizing their economies.
The focus, in reviewing the evidence from these trade reforms, is whether they appear to be sustainable in light of the findings of the 1991 research. We shall first examine evidence pertaining to the design of successful liberalizations, then evidence on short-term economic effects.
Design of Trade Liberalization
One of the strongest findings of the research by Papageorgiou, Choksi, and Michaely was that bold, strong trade liberalization at the initial stage of the reform was strongly associated with its success. How, then, do these nine reformers compare with those identified in that study?
Table 2 describes the policy elements that were involved. In all cases but one, major reduction in QRs were implemented along with substantial reduction in the level and dispersion of tariff rates. Table 3 provides a comparison of tariff structures before and at the end of liberalization. The contrast in these structures is remarkable. For example, Bolivia’s tariff range went from zero to 100 percent before the liberalization to only two tariff rates of 5 and 10 percent. Ecuador’s tariff range went from zero to 338 percent before the liberalization to a range from zero to 20 percent following the liberalization. Even one of the milder users of these reforms, Venezuela, went from a tariff ranging from zero to 135 percent before the liberalization to one ranging from zero to 50 percent by 1991. Equally remarkable, these reforms were implemented in a very short time—mostly less than four years, and in some instances within two years (Argentina, Bolivia, and Peru).
Policy Elements of Trade Liberalization
Indicates the direction of real effective exchange rate upon the implementation of trade liberalization.
Uruguay had eliminated all quantitative restrictions in earlier liberalization.
Policy Elements of Trade Liberalization
| Major Relaxation of Quantitative Restrictions? | Major Reduction of Tariffs? | Depreciation1 | Sustained? | |
|---|---|---|---|---|
| Argentina | Yes | Yes | No | Yes |
| Bolivia | Yes | Yes | Yes | Yes |
| Brazil | Yes | Yes | No | Yes |
| Colombia | Yes | Yes | Yes | Yes |
| Ecuador | Yes | Yes | No | Yes |
| Mexico | Yes | Yes | Yes | Yes |
| Peru | Yes | Yes | Yes | Yes |
| Uruguay | No2 | Yes | Yes | Yes |
| Venezuela | Yes | Yes | Yes | Partially |
Indicates the direction of real effective exchange rate upon the implementation of trade liberalization.
Uruguay had eliminated all quantitative restrictions in earlier liberalization.
Policy Elements of Trade Liberalization
| Major Relaxation of Quantitative Restrictions? | Major Reduction of Tariffs? | Depreciation1 | Sustained? | |
|---|---|---|---|---|
| Argentina | Yes | Yes | No | Yes |
| Bolivia | Yes | Yes | Yes | Yes |
| Brazil | Yes | Yes | No | Yes |
| Colombia | Yes | Yes | Yes | Yes |
| Ecuador | Yes | Yes | No | Yes |
| Mexico | Yes | Yes | Yes | Yes |
| Peru | Yes | Yes | Yes | Yes |
| Uruguay | No2 | Yes | Yes | Yes |
| Venezuela | Yes | Yes | Yes | Partially |
Indicates the direction of real effective exchange rate upon the implementation of trade liberalization.
Uruguay had eliminated all quantitative restrictions in earlier liberalization.
Tariff Structure Before and After Liberalization
Tariff Structure Before and After Liberalization
| Before Liberalization | After Liberalization | |
|---|---|---|
| Argentina | Tariff range from zero to 55 percent. Import surcharges 15 percent. Export taxes on some goods. | Three tariff rates: zero, 11, and 22 percent. Most export taxes abolished. |
| Bolivia | Complex system of tariff rates with exemption and high dispersion. Tariff range from zero to 100 percent. | Two tariff rates: 5 percent and 10 percent. |
| Brazil | Tariff range from zero to 105 percent. | Tariff range from zero to 35 percent with only nine tariff positions. |
| Colombia | Tariff range from zero to 200 percent. | Tariff range from zero to 40 percent. But 99.5 percent of tariff positions are within the range of zero to 20 percent. |
| Ecuador | Tariff range from zero to 338 percent. Prior deposits requirement surcharge. | Tariff range from zero to 20 percent. |
| Mexico | Tariff range from zero to 100 percent. Use of reference prices. | Tariff range from zero to 20 percent. |
| Peru | Tariff range from zero to 84 percent. Some export taxes. | Two tariff rates: 15 percent and 25 percent. Eighty-two percent of tariff items fall under the 15 percent rate. |
| Uruguay | Five-tier tariff structure ranging from 10 to 55 percent. Additional taxes on exports. Reference prices. | Three-tier tariff structure of 10 percent, 15 percent, and 20 percent. |
| Venezuela | Tariff range from zero to 135 percent. | Tariff range from zero to 50 percent. |
Tariff Structure Before and After Liberalization
| Before Liberalization | After Liberalization | |
|---|---|---|
| Argentina | Tariff range from zero to 55 percent. Import surcharges 15 percent. Export taxes on some goods. | Three tariff rates: zero, 11, and 22 percent. Most export taxes abolished. |
| Bolivia | Complex system of tariff rates with exemption and high dispersion. Tariff range from zero to 100 percent. | Two tariff rates: 5 percent and 10 percent. |
| Brazil | Tariff range from zero to 105 percent. | Tariff range from zero to 35 percent with only nine tariff positions. |
| Colombia | Tariff range from zero to 200 percent. | Tariff range from zero to 40 percent. But 99.5 percent of tariff positions are within the range of zero to 20 percent. |
| Ecuador | Tariff range from zero to 338 percent. Prior deposits requirement surcharge. | Tariff range from zero to 20 percent. |
| Mexico | Tariff range from zero to 100 percent. Use of reference prices. | Tariff range from zero to 20 percent. |
| Peru | Tariff range from zero to 84 percent. Some export taxes. | Two tariff rates: 15 percent and 25 percent. Eighty-two percent of tariff items fall under the 15 percent rate. |
| Uruguay | Five-tier tariff structure ranging from 10 to 55 percent. Additional taxes on exports. Reference prices. | Three-tier tariff structure of 10 percent, 15 percent, and 20 percent. |
| Venezuela | Tariff range from zero to 135 percent. | Tariff range from zero to 50 percent. |
How did these reforms treat quantitative restrictions? The 1991 study established a close association between sustained liberalization and drastic reduction in QRs at the initial stage of the reforms. Indeed, in that study strong liberalization and reduction in QRs were almost synonymous. Table 4 summarizes the initial conditions with QRs and what was left of them at the end of the liberalization. All countries, with the exception of Uruguay, which had removed all QRs in its earlier liberalization in the 1970s, had in place a complex system of QRs, such as import licensing, foreign exchange allocations, negative import lists, and outright import prohibitions. By the end of the reforms, four countries (Bolivia, Brazil, Ecuador, and Peru) had virtually no QRs or any other nontariff barrier left in their books. In Argentina, about 2 percent of local manufacturing is protected by QRs, mainly the car industry. In Ecuador, only 1 percent of tariff positions is subject to import licensing. Even in Mexico, which retained QRs covering 20 percent of local tradable production, the QRs apply primarily to the oil industry, which is an export activity. Among the eight countries that had extensive QRs before they began liberalizing, Venezuela appears to have retained the highest level of QRs; restrictions still apply to 18 percent of the import positions.
The Evaluation of Quantitative Restrictions
The Evaluation of Quantitative Restrictions
| Before Liberalization | After Liberalization | |
|---|---|---|
| Argentina | Extensive quantitive restrictions (QRs). Import licensing covering 42 percent of local production. | Quantitative restriction covering about 2 percent of manufacturing (primarily the car industry). |
| Bolivia | Extensive coverage with QRs. | No QRs. |
| Brazil | Extensive import licensing. Foreign exchange allocation controls. Reservation of Brazilian markets for local production for several products. | Most of the QRs have been eliminated. Some remain for official purchases. |
| Colombia | Import prohibition. Import licensing (together they covered 99 percent of tariff positions). | Import licensing for 1 percent of tariff positions. |
| Ecuador | Extensive import controls. Prohibition list for 9.5 percent of local production, import licensing, permit requirement. | Substantial reduction in QRs. |
| Mexico | Extensive use of import licensing covering 90 percent of tradable production. Export controls covering 25 percent of tradable production. | Import licenses reduced to about 20 percent of tradable production. Export controls reduced to 18 percent of tradable production. |
| Peru | Preferential regimes. QRs for 10 percent of tariff position. The importation of 10 percent of tariff items was prohibited. | No QRs. |
| Uruguay | No QRs. | No QRs. |
| Venezuela | The importation of goods falling under 45 percent of import positions was prohibited (11 percent), required import licensing (29 percent), or required health certification (5 percent). | The importation of goods falling under 18 percent of tariff positions was prohibited (5 percent), required prior licensing (5 percent), or required health certification (8 percent). |
The Evaluation of Quantitative Restrictions
| Before Liberalization | After Liberalization | |
|---|---|---|
| Argentina | Extensive quantitive restrictions (QRs). Import licensing covering 42 percent of local production. | Quantitative restriction covering about 2 percent of manufacturing (primarily the car industry). |
| Bolivia | Extensive coverage with QRs. | No QRs. |
| Brazil | Extensive import licensing. Foreign exchange allocation controls. Reservation of Brazilian markets for local production for several products. | Most of the QRs have been eliminated. Some remain for official purchases. |
| Colombia | Import prohibition. Import licensing (together they covered 99 percent of tariff positions). | Import licensing for 1 percent of tariff positions. |
| Ecuador | Extensive import controls. Prohibition list for 9.5 percent of local production, import licensing, permit requirement. | Substantial reduction in QRs. |
| Mexico | Extensive use of import licensing covering 90 percent of tradable production. Export controls covering 25 percent of tradable production. | Import licenses reduced to about 20 percent of tradable production. Export controls reduced to 18 percent of tradable production. |
| Peru | Preferential regimes. QRs for 10 percent of tariff position. The importation of 10 percent of tariff items was prohibited. | No QRs. |
| Uruguay | No QRs. | No QRs. |
| Venezuela | The importation of goods falling under 45 percent of import positions was prohibited (11 percent), required import licensing (29 percent), or required health certification (5 percent). | The importation of goods falling under 18 percent of tariff positions was prohibited (5 percent), required prior licensing (5 percent), or required health certification (8 percent). |
The evidence on the intensity of the reforms and of the substantial reduction in QRs would indicate, on the strength of the earlier research, that chances are good that these trade reforms will succeed. So far they have indeed been sustained, with the sole exception of the partial reversal in Venezuela.
The behavior of the real effective exchange rate (REER) at the time trade liberalization is introduced and at the end of the liberalization is presented in Table 5. The earlier study indicated a strong association between a depreciation of the REER at the introduction of liberalization and success of the liberalization. The same is true about a depreciation toward the end of a liberalization episode. The evidence from the nine episodes is mixed. At the initial stage, six countries experienced a depreciation of the REER: Bolivia, Colombia, Mexico, Peru, Uruguay, and Venezuela; in three—Bolivia, Mexico, and Venezuela—the depreciation was strong (more than 15 percent). Three countries, Argentina, Brazil, and Ecuador, saw an appreciation of their REER; it was strong in Argentina, small and temporary in Brazil, and small in Ecuador.
Behavior of Real Effective Exchange Rate During Trade Liberalization
Indicates the immediate impact on REER of the trade liberalization. The + sign indicates a sharp appreciation or depreciation change.
Describes the behavior of the REER during the entire liberalization period.
Compares the prevailing long-run level of REER before and after the introduction of trade liberalization.
Behavior of Real Effective Exchange Rate During Trade Liberalization
| When Trade Liberalization Was Introduced1 | During Trade Liberalization2 | Long-Run Comparison3 | |
|---|---|---|---|
| Argentina | Appreciation + | Appreciation/ Fluctuation | More appreciated |
| Bolivia | Depreciation + | Depreciation | More depreciated |
| Brazil | Appreciation | Depreciation/ Fluctuation | No change |
| Colombia | Depreciation | Depreciation | More depreciated |
| Ecuador | Appreciation | Appreciation | More depreciated |
| Mexico | Depreciation + | Fluctuation; no-trend | More depreciated |
| Peru | Depreciation | Appreciation | More appreciated |
| Uruguay | Depreciation | Fluctuation; no trend | No change |
| Venezuela | Depreciation + | Small appreciation | More depreciated |
Indicates the immediate impact on REER of the trade liberalization. The + sign indicates a sharp appreciation or depreciation change.
Describes the behavior of the REER during the entire liberalization period.
Compares the prevailing long-run level of REER before and after the introduction of trade liberalization.
Behavior of Real Effective Exchange Rate During Trade Liberalization
| When Trade Liberalization Was Introduced1 | During Trade Liberalization2 | Long-Run Comparison3 | |
|---|---|---|---|
| Argentina | Appreciation + | Appreciation/ Fluctuation | More appreciated |
| Bolivia | Depreciation + | Depreciation | More depreciated |
| Brazil | Appreciation | Depreciation/ Fluctuation | No change |
| Colombia | Depreciation | Depreciation | More depreciated |
| Ecuador | Appreciation | Appreciation | More depreciated |
| Mexico | Depreciation + | Fluctuation; no-trend | More depreciated |
| Peru | Depreciation | Appreciation | More appreciated |
| Uruguay | Depreciation | Fluctuation; no trend | No change |
| Venezuela | Depreciation + | Small appreciation | More depreciated |
Indicates the immediate impact on REER of the trade liberalization. The + sign indicates a sharp appreciation or depreciation change.
Describes the behavior of the REER during the entire liberalization period.
Compares the prevailing long-run level of REER before and after the introduction of trade liberalization.
The trend of the REER during the liberalization was mixed. Argentina’s, for example, continued to appreciate and, by the end of the liberalization, had appreciated considerably more than it had on average in the three years before liberalization. Bolivia’s REER continued to depreciate with some fluctuations, and at the end of the episode it depreciated in relation to the period prior to the trade liberalization. Brazil’s REER depreciated over the liberalization with sharp fluctuations, but at the end it remained at about the same level as it had been in the three years preceding the reforms. In general, three countries saw their REER appreciate during the liberalization period (Argentina, Ecuador, and Peru); three continued to depreciate; two indicated no trend one way or another; and Mexico’s REER for the first half of the liberalization period continued to depreciate substantially but reversed its course in the second half and began to appreciate quickly.
At the end of the liberalization, Argentina and Peru had an REER that was more appreciated than the average REER for the three years prior to liberalization. Bolivia, Colombia, Ecuador, Mexico, and Venezuela had a more depreciated REER, while Brazil, and Uruguay saw no change.
This evidence suggests that the behavior of the REER in these countries did not provide effective support to the liberalization. A strong and sustained depreciation would have provided a strong impetus to the tradable sector and exports in particular and restrained the rapid growth of imports.
The earlier research indicated that expansionary fiscal and monetary policies are the single most important cause of a reversal of trade reforms. Judging from the course of the rate of inflation during the reforms, several countries have done a remarkable job in controlling fiscal and monetary policies. Argentina, Bolivia, Mexico, Peru, and Venezuela substantially reduced their inflation rate during the reforms. Ecuador and Brazil were unable to reduce it. The difference between the two, however, was that Ecuador’s inflation remained at the annual level of between 45 percent and 50 percent, while Brazil’s remained at the level of hyperinflation. Colombia experienced a minor increase while Uruguay, unable to insulate itself from Argentina’s and Brazil’s inflation, experienced a dramatic doubling of their rate of inflation (to 100 percent). Looking at inflation rates beyond the liberalization period, it appears that Argentina, Bolivia, possibly Brazil, Colombia, Mexico (until recently), Peru, and Uruguay have managed to rein in inflation, an indication that fiscal and monetary policies have been working. Venezuela has been slipping behind on this front.
Exchange rate policies radically changed with the introduction of trade reforms and the liberalization of capital markets. All economies unified their exchange rate. Argentina fixed the nominal rate while all others had a “flexible” nominal exchange rate.6 In all countries individuals and businesses are permitted to buy and sell foreign exchange.
The need for stabilization was paramount for all of these countries. So was the need for structural reforms. With the exception of Colombia, and to a lesser extent Bolivia and Ecuador, structural reforms meant considerable change in the role of the public sector, in addition to trade liberalization, financial liberalization, extensive deregulation, and privatization of public assets. Argentina, Brazil, Mexico, Peru, and Venezuela did just that. The stabilization policy involved tightening monetary policy and very high real rates of interest. The combined effect of high interest rates on the one hand and the privatization of public assets, including public utilities, on the other brought in substantial and sustained capital inflows. As capital markets were liberalized along with the liberalization of merchandise trade, capital inflows caused the REER to depreciate less, or appreciate more, than it would have done in their absence.
The initial conditions of the economy when the countries embarked on the liberalization can be described as total collapse. Table 6 summarizes the trade regime and the macroeconomic situation at the outset of the reforms. All the trade reforms began under great economic distress, demanding (and enabling) strong and fast programs. In addition to a highly protective trade regime, most of them (with the exception of Colombia and Venezuela) had an annual rate of inflation exceeding 40 percent. Four countries had inflation rates in excess of 1,000 percent (Argentina, Bolivia, Brazil, and Peru); three others had lower, but still very high, inflation rates (Ecuador, Mexico, and Uruguay). The governments had no other choice but to embark on a program of reform that was both strong and decisive and therefore credible.
Policy Environment on the Eve of Trade Liberalization
High-tariff countries are those where the top rate is in excess of 50 percent including surcharges. Tariff dispersion in all these countries was more than 50 percentage points (the difference between the lowest and highest tariff rates).
“Hyper” indicates annual rates of inflation in excess of 1,000 percent; “high” indicates annual rates of inflation above 40 percent but less than 100 percent; “moderate” indicates annual rates of inflation between 20 percent and 40 percent.
Policy Environment on the Eve of Trade Liberalization
| Trade Related | Macroeconomic | ||||
|---|---|---|---|---|---|
| QR Coverage | Tariff Rates1 | Export Policies | Foreign Exchange | Inflation2 | |
| Argentina | Extensive | High/ dispersed | Subsidies/ taxes | Multiple rates/ controls | Hyper |
| Boliva | Extensive | High/ dispersed | Subsidies/ restrictions | Controls | Hyper |
| Brazil | Extensive | High/ dispersed | Subsidies | Controls | Hyper |
| Colombia | Extensive | High/ dispersed | Restrictions/ free trade zones | Controls | Moderate |
| Ecuador | Extensive | High/ dispersed | None | Multiple rates/ controls | High |
| Mexico | Extensive | High/ dispersed | Subsidies/ restrictions | Multiple rates | High |
| Peru | Extensive | Subsidies/ restrictions | Multiple rates/ controls | Hyper | |
| Uruguay | None | None | Floating rate/ no controls | High | |
| Venezuela | Extensive | High/ dispersed | Subsidies | Multiple rates/ controls | Moderate |
High-tariff countries are those where the top rate is in excess of 50 percent including surcharges. Tariff dispersion in all these countries was more than 50 percentage points (the difference between the lowest and highest tariff rates).
“Hyper” indicates annual rates of inflation in excess of 1,000 percent; “high” indicates annual rates of inflation above 40 percent but less than 100 percent; “moderate” indicates annual rates of inflation between 20 percent and 40 percent.
Policy Environment on the Eve of Trade Liberalization
| Trade Related | Macroeconomic | ||||
|---|---|---|---|---|---|
| QR Coverage | Tariff Rates1 | Export Policies | Foreign Exchange | Inflation2 | |
| Argentina | Extensive | High/ dispersed | Subsidies/ taxes | Multiple rates/ controls | Hyper |
| Boliva | Extensive | High/ dispersed | Subsidies/ restrictions | Controls | Hyper |
| Brazil | Extensive | High/ dispersed | Subsidies | Controls | Hyper |
| Colombia | Extensive | High/ dispersed | Restrictions/ free trade zones | Controls | Moderate |
| Ecuador | Extensive | High/ dispersed | None | Multiple rates/ controls | High |
| Mexico | Extensive | High/ dispersed | Subsidies/ restrictions | Multiple rates | High |
| Peru | Extensive | Subsidies/ restrictions | Multiple rates/ controls | Hyper | |
| Uruguay | None | None | Floating rate/ no controls | High | |
| Venezuela | Extensive | High/ dispersed | Subsidies | Multiple rates/ controls | Moderate |
High-tariff countries are those where the top rate is in excess of 50 percent including surcharges. Tariff dispersion in all these countries was more than 50 percentage points (the difference between the lowest and highest tariff rates).
“Hyper” indicates annual rates of inflation in excess of 1,000 percent; “high” indicates annual rates of inflation above 40 percent but less than 100 percent; “moderate” indicates annual rates of inflation between 20 percent and 40 percent.
Concerning the sequencing of trade and capital market liberalization, all of the countries introduced trade reforms before or at the same time as capital liberalization. Bolivia, Peru, Uruguay, and Venezuela appeared to have introduced the liberalization of these two markets at the same time, while Argentina, Brazil, Colombia, and Ecuador staggered the reforms. This pattern partly agrees with the presumption that capital market liberalization should wait until trade reforms are undertaken.
An important element of these liberalization episodes is the political environment which they were implemented. With the sole exception of Venezuela, where several riots and two failed military coups d’état followed the reforms, the reforms were fully supported by the public. They were undertaken by elected governments, and several of them have withstood the task of transition from one government to the next (Brazil, Bolivia, Colombia, Mexico, and Uruguay). The test of political transition is an important predictor of the long-term success of the reforms; Venezuela is the only case that has failed to pass the test.
Short-Term Economic Costs
Two important aspects of the short-term economic effects of liberalization, balance of payments, and national output are addressed here. Tables 7 and 8 indicate the performance of exports and imports before, during, and after liberalization. The evidence from the individual countries indicates that in six countries (Argentina, Bolivia, Ecuador, Peru, Uruguay, and Venezuela) the growth rate of exports following liberalization was higher on average than before the reforms were introduced. In Colombia, the growth of exports was higher after liberalization only if the year of liberalization is included. On the import side, seven countries showed stronger import growth on average after liberalization than before (Argentina, Bolivia, Colombia, Ecuador, Mexico, Peru, and Venezuela). Brazil had higher growth rates for imports following liberalization if the year of liberalization is included, and Uruguay had lower import growth after liberalization. On average for all episodes, exports grew by 4.3 percent annually in the three years prior to liberalization, while the growth rate jumped to 7.0 percent following liberalization. Equally important, the growth rate of exports went from 2.4 percent in the year before liberalization to 7.4 percent in the year of liberalization.
Trade Performance
(Percentage change in trade flows in 1987 U.S. dollars)
T1 is the average annual change in trade flows for the three years prior to the year of trade liberalization; T2 is the average annual change in trade flows for the year of liberalization and the subsequent three years; T3 is the average annual change for the three years immediately after the year in which trade liberalization was introduced.
(S) indicates strong liberalization, and (M) moderate liberalization.
Data for these countries for these years are not available on a comparable basis.
Trade Performance
(Percentage change in trade flows in 1987 U.S. dollars)
| Three Years Prior to Liberalization | Year of Trade Liberalization | Three years After Liberalization | |||||||||
|---|---|---|---|---|---|---|---|---|---|---|---|
| T11 | -3 | -2 | -1 | +1 | +2 | +3 | T21 | T3 | |||
| Argentina (S)2 | Exports | -3.3 | -15.0 | -12.1 | 17.1 | 26.4 | -4.0 | 0.6 | 2.0 | 6.2 | -0.5 |
| Imports | 4.4 | 22.6 | 11.8 | -21.2 | -10.1 | -18.2 | 69.9 | 75.8 | 39.3 | 55.8 | |
| Bolivia (S) | Exports | -3.3 | -1.1 | -6.1 | -2.8 | -10.0 | 17.5 | 3.8 | 12.6 | 6.0 | 11.3 |
| Imports | -18.0 | -44.0 | 10.7 | -20.7 | 65.7 | -11.8 | 8.5 | -30.8 | 7.9 | -11.4 | |
| Brazil (S) | Exports | 11.5 | 14.5 | 17.3 | 2.8 | -7.7 | 4.9 | 12.9 | …3 | 3.4 | 8.9 |
| Imports | 4.6 | -8.0 | -7.1 | 28.8 | 13.6 | 7.4 | -1.2 | … | 6.6 | 3.1 | |
| Colombia (M) | Exports | 17.5 | 31.5 | 24.5 | -4.3 | 39.9 | 13.4 | 15.6 | 13.2 | 20.5 | 14.1 |
| Imports | -0.3 | -13.3 | -1.5 | 13.9 | -2.6 | 16.4 | -7.1 | 50.3 | 14.2 | 19.9 | |
| Ecuador (S) | Exports | 1.0 | -17.3 | 29.0 | -8.8 | -1.8 | 20.0 | 0.2 | … | 6.1 | 10.1 |
| Imports | -4.2 | -4.6 | -15.5 | 7.4 | -10.3 | 31.1 | 4.6 | … | 8.5 | 17.8 | |
| Mexico (S) | Exports | 15.8 | 12.7 | 25.7 | 9.1 | -5.8 | 4.0 | -12.9 | -10.5 | -6.3 | -6.5 |
| Imports | -9.2 | -37.9 | -26.4 | 36.8 | 13.8 | 15.5 | -23.3 | 1.4 | 1.8 | -2.1 | |
| Peru (S) | Exports | 3.5 | 5.6 | 4.6 | 0.4 | 25.2 | 17.6 | 5.2 | … | 16.0 | 11.4 |
| Imports | -7.0 | 21.2 | -5.0 | -37.1 | 22.6 | 9.6 | 26.1 | … | 19.4 | 17.8 | |
| Uruguay (S) | Exports | 3.2 | -10.7 | -2.7 | 22.9 | -5.9 | 10.0 | 7.3 | 7.9 | 4.8 | 8.4 |
| Imports | 13.0 | 10.8 | -5.7 | 33.9 | 11.6 | -1.9 | 1.8 | 4.0 | 1.9 | 1.3 | |
| Venezuela (S) | Exports | -7.0 | -6.5 | 0.3 | -14.7 | 6.0 | 15.4 | 8.3 | -4.9 | 6.2 | 6.3 |
| Imports | 3.1 | -7.5 | 4.1 | 12.8 | -22.8 | 5.8 | 54.9 | 24.3 | 15.5 | 28.3 | |
| Total | Exports | 4.3 | 1.5 | 9.0 | 2.4 | 7.4 | 11.0 | 4.6 | 3.4 | 7.0 | 7.1 |
| Imports | -1.5 | -6.7 | -3.8 | 6.1 | 7.9 | 6.0 | 19.4 | 27.8 | 13.0 | 14.5 | |
T1 is the average annual change in trade flows for the three years prior to the year of trade liberalization; T2 is the average annual change in trade flows for the year of liberalization and the subsequent three years; T3 is the average annual change for the three years immediately after the year in which trade liberalization was introduced.
(S) indicates strong liberalization, and (M) moderate liberalization.
Data for these countries for these years are not available on a comparable basis.
Trade Performance
(Percentage change in trade flows in 1987 U.S. dollars)
| Three Years Prior to Liberalization | Year of Trade Liberalization | Three years After Liberalization | |||||||||
|---|---|---|---|---|---|---|---|---|---|---|---|
| T11 | -3 | -2 | -1 | +1 | +2 | +3 | T21 | T3 | |||
| Argentina (S)2 | Exports | -3.3 | -15.0 | -12.1 | 17.1 | 26.4 | -4.0 | 0.6 | 2.0 | 6.2 | -0.5 |
| Imports | 4.4 | 22.6 | 11.8 | -21.2 | -10.1 | -18.2 | 69.9 | 75.8 | 39.3 | 55.8 | |
| Bolivia (S) | Exports | -3.3 | -1.1 | -6.1 | -2.8 | -10.0 | 17.5 | 3.8 | 12.6 | 6.0 | 11.3 |
| Imports | -18.0 | -44.0 | 10.7 | -20.7 | 65.7 | -11.8 | 8.5 | -30.8 | 7.9 | -11.4 | |
| Brazil (S) | Exports | 11.5 | 14.5 | 17.3 | 2.8 | -7.7 | 4.9 | 12.9 | …3 | 3.4 | 8.9 |
| Imports | 4.6 | -8.0 | -7.1 | 28.8 | 13.6 | 7.4 | -1.2 | … | 6.6 | 3.1 | |
| Colombia (M) | Exports | 17.5 | 31.5 | 24.5 | -4.3 | 39.9 | 13.4 | 15.6 | 13.2 | 20.5 | 14.1 |
| Imports | -0.3 | -13.3 | -1.5 | 13.9 | -2.6 | 16.4 | -7.1 | 50.3 | 14.2 | 19.9 | |
| Ecuador (S) | Exports | 1.0 | -17.3 | 29.0 | -8.8 | -1.8 | 20.0 | 0.2 | … | 6.1 | 10.1 |
| Imports | -4.2 | -4.6 | -15.5 | 7.4 | -10.3 | 31.1 | 4.6 | … | 8.5 | 17.8 | |
| Mexico (S) | Exports | 15.8 | 12.7 | 25.7 | 9.1 | -5.8 | 4.0 | -12.9 | -10.5 | -6.3 | -6.5 |
| Imports | -9.2 | -37.9 | -26.4 | 36.8 | 13.8 | 15.5 | -23.3 | 1.4 | 1.8 | -2.1 | |
| Peru (S) | Exports | 3.5 | 5.6 | 4.6 | 0.4 | 25.2 | 17.6 | 5.2 | … | 16.0 | 11.4 |
| Imports | -7.0 | 21.2 | -5.0 | -37.1 | 22.6 | 9.6 | 26.1 | … | 19.4 | 17.8 | |
| Uruguay (S) | Exports | 3.2 | -10.7 | -2.7 | 22.9 | -5.9 | 10.0 | 7.3 | 7.9 | 4.8 | 8.4 |
| Imports | 13.0 | 10.8 | -5.7 | 33.9 | 11.6 | -1.9 | 1.8 | 4.0 | 1.9 | 1.3 | |
| Venezuela (S) | Exports | -7.0 | -6.5 | 0.3 | -14.7 | 6.0 | 15.4 | 8.3 | -4.9 | 6.2 | 6.3 |
| Imports | 3.1 | -7.5 | 4.1 | 12.8 | -22.8 | 5.8 | 54.9 | 24.3 | 15.5 | 28.3 | |
| Total | Exports | 4.3 | 1.5 | 9.0 | 2.4 | 7.4 | 11.0 | 4.6 | 3.4 | 7.0 | 7.1 |
| Imports | -1.5 | -6.7 | -3.8 | 6.1 | 7.9 | 6.0 | 19.4 | 27.8 | 13.0 | 14.5 | |
T1 is the average annual change in trade flows for the three years prior to the year of trade liberalization; T2 is the average annual change in trade flows for the year of liberalization and the subsequent three years; T3 is the average annual change for the three years immediately after the year in which trade liberalization was introduced.
(S) indicates strong liberalization, and (M) moderate liberalization.
Data for these countries for these years are not available on a comparable basis.
Performance of Trade Flows Before and After Liberalization1
(Percentage change weighted by 1987 U.S. dollars)
The trade flows of Brazil, Ecuador, and Peru for the third postliberalization year are excluded because comparative data are not available.
T1 is the annual average growth rate for the three years prior to liberalization. T2 is the annual average growth rate for the three years following liberalization and the year of liberalization.
Performance of Trade Flows Before and After Liberalization1
(Percentage change weighted by 1987 U.S. dollars)
| Three Years Prior to Liberalization | Year of Trade Liberalization | Two Years After Liberalization | ||||||
|---|---|---|---|---|---|---|---|---|
| T12 | 3 | 2 | 1 | 1 | 2 | T2 | ||
| Exports | 7.4 | 5.9 | 13.5 | 2.7 | 1.2 | 6.5 | 3.7 | 3.8 |
| Imports | -4.9 | -21.3 | -10.1 | 16.7 | 4.8 | -1.4 | 5.7 | 3.0 |
The trade flows of Brazil, Ecuador, and Peru for the third postliberalization year are excluded because comparative data are not available.
T1 is the annual average growth rate for the three years prior to liberalization. T2 is the annual average growth rate for the three years following liberalization and the year of liberalization.
Performance of Trade Flows Before and After Liberalization1
(Percentage change weighted by 1987 U.S. dollars)
| Three Years Prior to Liberalization | Year of Trade Liberalization | Two Years After Liberalization | ||||||
|---|---|---|---|---|---|---|---|---|
| T12 | 3 | 2 | 1 | 1 | 2 | T2 | ||
| Exports | 7.4 | 5.9 | 13.5 | 2.7 | 1.2 | 6.5 | 3.7 | 3.8 |
| Imports | -4.9 | -21.3 | -10.1 | 16.7 | 4.8 | -1.4 | 5.7 | 3.0 |
The trade flows of Brazil, Ecuador, and Peru for the third postliberalization year are excluded because comparative data are not available.
T1 is the annual average growth rate for the three years prior to liberalization. T2 is the annual average growth rate for the three years following liberalization and the year of liberalization.
The change in the performance of imports after liberalization was even more dramatic. It went from -1.5 percent before liberalization to growth rates ranging from 13 percent to about 14.5 percent annually after liberalization.
Looking at trade flows in a different way (Table 8), we see the result is different when trade performance is weighted by each country’s exports and imports. Actual annual average exports after liberalization increased by 3.8 percent, including the year of liberalization, while prior to liberalization they grew annually at a rate of 7.4 percent. The corresponding figures for imports are 3.0 percent and -4.9 percent, respectively.
Obviously, imports grew more rapidly than exports, although this has not yet generated trade deficits. The rapid increase of imports is explained mostly by the extent of the liberalization, the dismal performance of imports (and production) prior to liberalization, and the presence of capital inflows.
The evidence also indicates that the liberalization period coincides with an acceleration of the growth rate of the GDP, if the year of liberalization is not included. For the nine countries combined, the average GDP declines during the year trade reforms were introduced (–0.9 percent). Six countries (Argentina, Bolivia, Brazil, Mexico, Peru, and Venezuela) had negative GDP rates of growth in the year trade reforms were introduced. Undoubtedly, the decline in GDP is associated with the strong stabilization packages that these six countries introduced at the same time as trade reforms. The growth rate of GDP of all nine countries accelerates in each subsequent year following the year of liberalization.
Obviously, then, liberalization is not associated with lower economic growth. If anything, these trade reforms, being strong and fast, are associated with higher growth, supporting the findings of the 1991 study. Unfortunately, the lack of data precludes a more systematic study of the liberalization on employment and income distribution. Even if relevant data on employment and income distribution were available, it would have been difficult to sort out their relation to liberalization from the effects of the stabilization policies and other structural reforms.
Compatibility with the Earlier Experience
Based on the experience of liberalization episodes of the postwar era, what may we conclude about the sustainability of the recent reforms in Latin America? Could the Latin American countries have designed the liberalization of their trade regimes substantially better? Would a different approach to trade liberalization have been preferable, in the sense that it would increase its chance to be sustained and reduce the economic costs of transition?
Recognizing that the implementation of trade reforms varied widely from one country to another, the evidence reviewed here, nevertheless, points to some general observations common to all nine liberalizations. It strongly suggests that the design of the trade reforms themselves was good, consistent with what earlier inferences have indicated as the attributes of sustainable liberalization. The only uncertainty for their future success lies outside the confines of trade liberalization, in the interplay of stabilization and the ensuing appreciation of the REER.
All these reforms were strong, with immediate implementation, imparting credibility and determination. From the outset, they decisively reduced quantitative restrictions, which the 1991 study had found to be the most important predictor of success. All eight countries with QRs sharply reduced or eliminated them.
On the real effective exchange rate the evidence is less clear, but nevertheless, six of the nine liberalizations began with depreciation of the REER (Bolivia, Colombia, Mexico, Peru, Uruguay, and Venezuela); three of these countries experienced strong depreciations (Bolivia, Mexico, and Venezuela). At the end of liberalization, five countries had their REER depreciated in comparison with the start of liberalization (Bolivia, Colombia, Ecuador, Mexico, and Venezuela), two saw it appreciate (Argentina and Peru), and the other two (Brazil and Uruguay) saw no change.
The REER is not a policy instrument. Among other factors, it reflects overall trends in macroeconomic policy. Several countries have been able in the long term to either reduce inflation substantially (Argentina, Bolivia, Mexico, and Peru) or keep it under control, at modest levels (Ecuador, Colombia, and Uruguay). The success of these countries in stabilizing their economies may well prove to be the critical element for the success not only of the trade reforms but of the entire package of economic restructuring as well.
Failure or success in getting inflation under control permanently will affect the viability of the financial institutions and political stability. The combined influence of stabilization and structural reforms has placed significant stress on the banking system. Argentina, Brazil, Mexico, Peru, and Venezuela have experienced different degrees of difficulties with their banking system. Only time can tell whether they will be able to manage this challenge effectively.
Failure to control macroeconomic policy, and therefore inflation, may affect the will and determination of the governments to maintain their current policies. So far, there has been no major opposition to trade reforms, except in Venezuela. In all other eight countries, trade reform has been supported by the public and most of the political parties. In this respect, it is encouraging that all nine countries have agreed to participate in the multilateral trade agreement of the Uruguay Round and have indicated strong interest in participating in a free trade agreement in the Western Hemisphere.
The 1991 research found that trade reforms that remain in place for six years or more are almost certain to be sustained indefinitely. On that basis, Bolivia, Mexico, and Uruguay seem to be on a safe path. Not only have their liberalizations lasted more than six years, but they have withstood the test of governmental change as well. For Argentina, Brazil, Colombia, Ecuador, and Peru, the prospects are encouraging. Argentina and Peru will soon join Brazil and Colombia, in having new governments and the chance to test the resilience of the trade reforms. Venezuela is the only country in which the liberalization remains uncertain. Not surprisingly, Venezuela is a case of liberalization that has not started with a situation of economic collapse. Hence, public support for the policy was shaky from the inception of liberalization. The policy did not go as far for the rest of the countries,7 and the new government appears to be much less committed to the trade reforms than its predecessor.
In conclusion, the recent reformers in Latin America have met fully four of the conditions found necessary for a sustained trade liberalization by the 1991 research: strong initial steps toward liberalization, substantial reduction in QRs, initial economic conditions that required strong reforms, and political stability in the sense that there was continuous broad political support for these reforms. Some of the liberalizations have also passed the test of time and have survived political transition.
The 1991 study underlined the importance of depreciation of the REER at the start of the liberalization and at the end of it. We saw that six countries began their liberalization with depreciation, and two ended it with appreciation of the REER. The 1991 research found in particular that no country’s liberalization survived while the REER was appreciating at the end of the reform. The REER tended to appreciate beyond the period of liberalization in all countries except Bolivia. The major factors for this sustained appreciation appear to be the immediate opening of capital markets, the structural reforms, and the stabilization efforts of these countries. In the 1991 research, only 4 liberalization episodes, out of 36 examined, experienced the liberalization of capital markets at the same time or soon after trade reforms were introduced. There were few observations to base a thorough assessment of the effects of substantial inflows on the tradable sector, and the impact of substantial reductions of such flows on foreign reserves and macroeconomic policy. Nevertheless, the available evidence indicated a negative effect of capital inflows on the tradable sector and, when these inflows were substantially reduced or reversed, a negative impact on foreign reserves and on the fate of liberalization.
The recent events in Mexico would probably shed some light, in the near future, about the probability of repetition of this pattern. If similar developments in capital flows were to occur in these countries, it would be difficult to tell whether the governments will manage to withstand their consequences, as Chile succeeded in doing in the early 1980s, or succumb to the pressures, as Argentina did at the same time, and witness a collapse of the reforms. It is difficult to foretell the outcome partly because the nature of inflows is substantially different now than in the 1970s, involving to a large extent private capital and direct foreign investments, and partly because the current political climate is undoubtedly in favor of trade liberalization, unlike in the earlier times.
Speculating further, it could be the case that these trade reforms were so strong that by the time large capital inflows were recorded the economic restructuring was well under way, so that it did not have an adverse effect on the tradable sector. Partial evidence consistent with this argument is that the growth rate of exports and of GDP increased by the same order of magnitude in the postliberalization period.
Furthermore, the collective experience of these countries with the prevailing policies prior to these reforms, which inflicted severe economic costs, may have persuaded policymakers and the people at large that deregulation and trade liberalization is the way to go, notwithstanding the probable transition costs and events that in the past would have led to a severe setback.
Trade Regime in the Former Soviet Union
This section will be quite different from its predecessors. It would be inappropriate to expect the pattern of trade liberalization in Russia, the Baltic countries, and other countries of the former Soviet Union, which was part of an overall change of the economic regime, to share much with the liberalizations in Latin America. What we will do here is present the basic elements of the emerging trade regime and pose the question of what would be the essential ingredients of further liberalization. The former Soviet Union (this applies, to a large extent, also to the former eastern block outside the Soviet Union—the members of the former Council for Mutual Economic Assistance (CMEA)) has gone through a dramatic shift in its conduct of international transactions, just as it has in other facets of economic life. Whether to refer to this change as “trade liberalization” is an open question. If liberalization is to be judged by the prevalence of instruments of government intervention in trade as they are practiced in a market economy, then the former Soviet Union has undergone a process contrary to liberalization. If, on the other hand, liberalization is understood—as we think it should be—as a process that brings relative prices closer to what they would be without government interference, and in which prices replace non-price instruments of interference, then a substantial liberalization has indeed taken place—its degree varying from one republic to another.
Prior to the disintegration of the Soviet Union and the start of the process of change toward market economies—from the end of 1990 through early 1991—“international” trade of the present independent states was, in fact, predominantly domestic trade: About 80 percent, on average, of total trade among the states and between them and the outside world consisted of interrepublic trade. Of the other 20 percent, a high proportion consisted of trade with other CMEA countries, which shared some important attributes with interrepublic trade. This trade had none of the attributes of foreign trade. It required no means of payments other than the local currency; it was financed through the same banking system; and it involved no government barriers to trade—being, in this sense, completely “free.” But, like any other (i.e., intrarepublic) trade in the Soviet Union, interrepublic trade was not guided by any market mechanism, nor did it follow the guidelines of comparative advantage. Location of economic activity, as well as trade relations of each enterprise—whether within the state or outside it—was fully determined by commands of the planning machinery. Any relationship between these production and trade patterns and what a market would have established was, then, purely coincidental.8
The shift (still partial, mostly) to a market mechanism has led to at least some approximation of what the pattern of trade would be in a free market. But state intervention is still heavy and strongly discriminates against tradables, probably even more than in the high-protection market economies prior to their liberalization. Unlike in the market economies (Argentina prior to 1990 may have been a partial exception), the instruments of intervention in trade of the Baltics, Russia and other countries of the former Soviet Union are related to exports rather than to imports. Most of the latter are formally free of either tariffs or nontariff barriers; whereas exports are subject to heavy restrictions in almost all of the states, both through export taxes (which may take the form of forced sale of the foreign exchange proceeds at a low exchange rate) and through the prevalent requirement for licensing.9
The discouragement of exports derives from a variety of sources—most of them unique to the Baltics, Russia, and other countries of the former Soviet Union, either because of the starting attributes of the regime or because common forces have been influencing these countries. To start with, a major problem facing producers, both before and after the disintegration of the Union, has been the scarcity of inputs: it is of prime importance for an enterprise to ensure the availability of its essential inputs. This leads to the restriction of exports (of machinery and, mostly, intermediate inputs—although to some extent also of certain consumer goods) and to an increase in their availability at home. True, overall availability would have increased with the purchase of imports provided for by exports, but this is beyond the concern of the agencies involved.
A contributing factor has been the frequent maintenance of domestic low prices for exportable goods. In such a system, combined often with a high real depreciation—as will be discussed shortly—it becomes highly profitable to buy locally and resell in the export market. The licensing is then intended either to prevent the exports—for the reasons mentioned before—or to assign the (often huge) rent involved in exports to the parties that the government deems worthy. With these price differentials, it should be noted that the prevention of exports may indeed often be welfare enhancing (the benefit of exportables in domestic use exceeding the price received from export sales); but short, of course, of changing the price system altogether, export taxes rather than licensing should on all counts be the preferred instrument.
Raising tax revenues is, indeed, another common motivation for export restrictions. This is done both through a straightforward tax or, often, through the enforcement of the sale of (part of or all) export proceeds to the government at an exchange rate substantially below a “market,” or “free” level of the rate (say, the rate at which free imports are financed). The motivation for the latter instrument—the forced sale—is often somewhat different: the government does not believe that export proceeds would otherwise be offered in the foreign exchange market, and be accessible to the government (which demands it as an importer and payer of foreign debt). In other words, assuming that the government cannot in effect buy sufficient foreign exchange at the free rate, the saving perceived by the government is much higher than the observed difference between the free rate and the “penalty” rate paid to exporters.
Until quite recently, when all the countries of the former Soviet Union (except the Baltics) were part of a ruble area, a particular motivation existed to discourage interrepublic exports. By exporting to another state, the country in question would be paid in rubles, but it could produce rubles itself (though not cash rubles, Russia excluded) with no real costs, so that the export would amount to a complete waste. It is only when Russia started constraining this possibility that the motivation disappeared for the peripheral republics; but it still exists today for Russia itself.
As noted, a substantial real depreciation has been a common feature; the Baltics have been the major exception, as was Russia for most of 1993–94. This depreciation results partly from the trade regime itself, with its heavy restrictions on exports (thus leading to a vicious circle—a higher real devaluation strengthening still further the motivation for export restriction). But other factors have been crucial. These are, first, a heavy asset demand for foreign exchange, liquid foreign assets being almost the only available instrument to hedge against inflation. In most countries of the former Soviet Union, a dramatic deterioration of the terms of trade has been another source. This resulted predominantly from the change of the oil price in interrepublic trade: the price increased from a small fraction of the world price level to something close to that level—a change that benefited primarily Russia, at the expense of the other states.
The level of the foreign exchange rate has been a major suppressant of imports (this is the way export restriction turns into import restriction). Since import barriers were absent at the start of the process, and no demand for added protection (to that effected through the exchange rate) was forthcoming, imports have mostly remained free of specific barriers. Not surprisingly, the exception to this rule has been the Baltic countries, primarily Estonia and Latvia, in which real appreciation has taken place. In these two countries, a gradually increasing demand for protection against imports has been provoked. It has led, indeed, to both tariff and nontariff barriers, primarily against imports of agricultural produce. By now, these barriers have become a source of major distortions in these economies. In this sense, they have moved away from trade liberalization.
What would constitute “trade liberalization” in the countries of the former Soviet Union? Once more, such liberalization is of little meaning if not given within a more general context. That is, trade liberalization makes sense only as part of an overall change in the price system and the application of the price mechanism. Otherwise, it is neither feasible nor even necessarily desirable.
Assuming that prices in general are predominately determined by a market mechanism, and are close to their equilibrium levels, trade liberalization would consist predominantly of the following measures:
(1) Removing barriers on exports—barriers imposed now through licensing, taxes, compulsory sale of foreign exchange proceeds and (indirectly) compulsory sales of exportable goods to prescribed enterprises.
(2) Establishing foreign exchange markets. Nominal trade liberalization would create a system far removed from that of a “free trade” system in the absence of well—functioning foreign exchange markets (whether within a fixed-rate or a floating-rate regime), in which most of the export proceeds are sold and most of the financing needed for imports is acquired.
(3) The third main trade liberalization measure is negative: refraining from imposing barriers on imports. The absence of such barriers at present is not of much significance, given the high (universal) protection provided by the high real exchange rate in most of the countries of the former Soviet Union. If, on the other hand, imports remain free when the exchange rate assumes some “normal” level and the protection granted through it disappears, this would indeed be a true measure of maintaining a free trade system.
Finally, the circumstances that seem necessary for introducing liberalization and implementing its major components, would also be those that are likely to grant sustainability to trade liberalization. First and foremost, changes in the trade regime must be part of a comprehensive economic liberalization: without an overall freeing of prices, so that they approach equilibrium market levels, trade liberalization will neither progress nor endure (nor, again, should it necessarily be desirable). An overall stabilization would be essential just as in normal market economies, so would be a well-functioning market for foreign exchange, which would imply, in turn, a reasonably functioning financial system. But freeing foreign capital markets, on the other hand, does not seem a priori to be essential for sustaining trade liberalization (experience does not have much to tell us yet about this issue).
See Chapter II in IMF, International Trade Policies, Vol. II, The Uruguay Round and Beyond, Background Papers, World Economic and Financial Surveys (Washington, 1994).
See Michael Michaely, Demetris Papageorgiou, and Armeane Choksi, Liberalizing Foreign Trade: Lessons of Experience in the Developing World, Vol, 7 (Oxford, England; and Cambridge, Massachusetts: Basil Blackwell, 1991), Hereinafter referred to as “the 1991 study.”
Note: Throughout this paper the phrase “the former Soviet Union” is used to mean the Baltic states, Russia, and other countries of the former Soviet Union.
This section draws liberally from Demetris Papageorgiou, Armeane M. Choksi, and Michael Michaely, “Liberalizing Foreign Trade in Developing Countries: The Lessons of Experience” (Washington: World Bank, 1990).
For details of classification, see Michaely, Papageorgiou, and Choksi, Liberalizing Foreign Trade.
The designation of a country as politically stable or unstable refers only to the period following the introduction of trade liberalization.
For example, deregulation of internal market, privatization, and liberalization of the financial system.
Flexible in the sense that it fluctuated; central bank involvement was evident in each country with flexible rates.
But it was nonetheless strong and swift.
Aside, of course, from a few important instances related to the physical availability of natural resources: oil, for instance, was, and is, inevitably produced (but not necessarily refined) where it is found in the ground.
This form of discrimination against tradables has, indeed, been an apt representation of Abba Lemer’s famous “symmetry theorem.”