III Review of Fund Arrangements and Assessment of Overall Results

Robert Sharer, and Piritta Sorsa
Published Date:
February 1998
  • ShareShare
Show Summary Details

To assess trade policy in Fund-supported programs,9 a review was conducted of all multiyear arrangements initiated and terminated between January 1990 and June 1996.10 The review examined trade reform targeted and implemented during the program period, and noted factors that influenced trade reform. This section first describes the methodology, then assesses the restrictiveness of countries’ trade regimes at the start of the program, the targeted change in restrictiveness envisaged, and the overall restrictiveness at the end of the program. It continues with a discussion of program targets, monitoring, overall results, and ambitiousness, and factors that influenced trade reform in the programs and some implications for program design.

The review finds that (1) there was a clear need for trade liberalization in Fund-supported programs, since most countries started out with restrictive or highly restrictive systems; (2) the extent of movement toward liberal trade regimes targeted in Fund programs varied; about two-thirds of programs targeted a quantifiable reduction in restrictiveness measured by the index used in this study; (3) one-third of programs did not target a quantifiable change in trade restrictiveness, including those for countries starting with restrictive trade systems; (4) programs that targeted a reduction in restrictiveness generally succeeded in achieving their trade reform objectives, including through the appropriate use of binding monitoring (prior actions, performance criteria, benchmarks, and reviews); (5) medium-term trade reform targets although mentioned often were generally insufficiently specific and comprehensive; (6) in general, longer-term Fund involvement in countries during 1990–96 does not seem to have promoted greater trade liberalization in the programs reviewed; (7) the World Bank’s involvement was frequently cited as having influenced trade reform, and the results of Fund-Bank collaboration seem to have been effective; and (8) fiscal concerns were an important factor limiting the extent of trade reform targeted, but no clear relationship was found between the strength of trade reform targets or achievements and the countries’ fiscal objectives and circumstances.


The following subsections describe the coverage of programs as well as the method used to measure the trade policy restrictiveness of the program countries.

Selection of Review Countries

The review covers multiyear Fund arrangements of at least two years’ duration, since these are intended to target structural reform measures, including trade reform. Arrangements that came into effect after January 1, 1990 and that ended by June 30, 1996 were selected to facilitate evaluation of program outcomes. This resulted in the review covering 30 arrangements in support of 28 programs in 27 countries (Table 1).11 While the number of programs is limited, the review covers a wide variety of countries and circumstances.

Table 1.Arrangements Included in the Review of Fund-Supported Programs1(Multiyear arrangements started after January 1, 1990 and completed by June 30, 1996)
ESAFEFFSAFStand-By Arrangement
Burkina Faso3/31/933/13/91
Equatorial Guinea2/3/93
Sierra Leone24/3/92
Sri Lanka9/13/91
Middle East
Western Hemisphere
Source: IMF.

Stand-By Arrangement, Extended Fund Facility (EFF), Structural Adjustment Facility (SAF), Enhanced Structural Adjustment Facility (ESAF), and Rights Accumulation Program (RAP). Includes multiyear programs (at least two years long) that expired at the latest by the end of June 1996. The ESAF arrangements with Burundi and Rwanda, both of which began in 1991, are excluded.

Rights Accumulation Program.

Source: IMF.

Stand-By Arrangement, Extended Fund Facility (EFF), Structural Adjustment Facility (SAF), Enhanced Structural Adjustment Facility (ESAF), and Rights Accumulation Program (RAP). Includes multiyear programs (at least two years long) that expired at the latest by the end of June 1996. The ESAF arrangements with Burundi and Rwanda, both of which began in 1991, are excluded.

Rights Accumulation Program.

The review provides a snapshot of trade reform in each program and does not consider directly trade reform prior to, or following, the arrangements under review;12 however, the issue of continued Fund involvement during the 1990s in the trade reform efforts of countries is discussed. While programs for three transition economies (Hungary, Mongolia, and Poland) were included, the selection criteria led to the exclusion of the Baltics, Russia, and other former Soviet Union countries. One reason for this was that during the early 1990s Fund support for these countries took the form of shorter-term programs supported by Stand-By Arrangements or the Systemic Transformation Facility. More important, these programs were applied in highly unusual circumstances. International trade had been administered in the context of the Council for Mutual Economic Assistance (CMEA), central planning was being dismantled, and the trade instruments used in market economies were not widely in place. In light of these exceptional circumstances the experiences of these programs are not replicable for other countries. Moreover, these experiences have been discussed in detail elsewhere. Nevertheless, several of these short-term programs achieved substantial structural change, including trade liberalization. The trade reforms of these countries are summarized in Box 4.

Quantification of the Overall Trade Policy Stance

Measurement of trade liberalization in the countries reviewed was based on an index of overall trade policy restrictiveness, which is discussed extensively in Appendix I.13 The index consists of a 10-point scale that combines measurements of the restrictiveness of tariffs and NTBs. It measures the overall restrictiveness of a country’s trade system relative to protection levels in all Fund members. The restrictiveness of the import tariff regime was based on the average unweighted statutory tariff rate, classified into five categories, from open to restrictive. The ranges of tariff rates used for classification were based on tariff rates of all Fund members; roughly equal numbers of Fund members fall in each category. Nontariff barriers were classified into three categories based on coverage and restrictiveness, including both import and export QRs, bans, and restrictive licensing, state trading monopolies, restrictive foreign exchange practices (where they acted as binding trade restrictions), and other NTBs.14 A rating of 1 on the index represents the most open trade regime, and 10 the most restrictive. The index gives relatively more weight to NTBs since, as noted above, they generally result in larger economic distortions and less transparent policy regimes than high tariff rates.

Box 4.Trade Reform in the Baltics, Russia, and Other Countries of the Former Soviet Union

The criteria used to select arrangements included in the review resulted in the exclusion of arrangements with the Baltics, Russia, and other former Soviet Union countries, as none of these countries had Fund arrangements of at least two years’ duration that started after January 1, 1990 and ended by June 30, 1996. Nevertheless, trade liberalization was a key aspect of Fund programs in these countries and considerable liberalization was achieved. This box briefly reviews these trade liberalization achievements.1

In terms of initial conditions, the trade system had to be radically redesigned in all these countries. This was due to the role of central planning in trade, which was affected by the collapse of the CMEA and dismantling of barter trade agreements. Also, in many countries no trade system existed, because they were formerly part of a larger country. Trade reform and the accompanying move to a market-based exchange rate system were designed to provide price signals and a market-oriented incentive structure. This was aimed at restructuring domestic industry and export sectors to make the economies outward looking, globally integrated, and market based.

Fund programs, in collaboration with the World Bank, targeted the creation of simple, transparent, and tariff-based trade regimes. The Fund’s trade policy advice had three main components. One was to remove quantitative restrictions (QRs). This was linked to the elimination of state trading and state orders, ensuring that QRs were not introduced, and where necessary replacing QRs on exports with temporary export taxes. The other two components involved eliminating expeditiously export taxes and adopting an import tariff structure that was low and relatively uniform. The legacy of central planning meant that much of the initial trade reform was focused on removing export QRs.

Generally, substantial progress in trade liberalization has been made. The magnitude of liberalization has varied across countries. Most progress has been made with removing quantity-based export restrictions that were inherited from central planning. This includes dismantling of state trading and rights of strategic exporters. The incidence of QRs on imports varies, but most countries use them sparingly. Many countries have also introduced low and uniform tariffs. In some cases, progress toward implementing a low and uniform import tariff structure has been influenced by the inability to mobilize fiscal revenue from other sources. In other cases, considerations relating to accession to the World Trade Organization or regional trading arrangements, or both, have affected the pace of trade liberalization.

1 For a more detailed description, see Kirmani and others (1994a) and Leidy and Ibrahim (1996).

Changes in tariff dispersion, maximum tariffs, exemptions, export taxes, and improvements in customs administration were not explicitly incorporated into the overall index due to the complexity of aggregating additional dimensions of trade policy and the difficulties of obtaining comparable cross-country data on these variables. In many cases, however, most of these additional measures of restrictiveness moved together with average tariff levels. Thus, by including average tariffs, the index implicitly captures, at least to some extent, these other measures of restrictiveness.

A variety of sensitivity analyses were undertaken to assess whether the results of the review would be significantly affected by changes in the tariff and NTB classifications. First, an alternative import tariff classification with wider tariff ranges was specified. This left virtually unchanged the extent of movement toward liberalization targeted and achieved. Next, NTB categories were redefined by omitting export restrictions as well as foreign exchange restrictions that acted as binding trade restraints; the results were essentially unchanged in each case. While in principle the index may miss significant trade liberalization at the restrictive end, examination of the data shows this rarely occurred in the programs reviewed and did not affect the results of the study.

Aggregation of different types of trade restrictions is not without challenges. Ideally, assessment of the overall restrictiveness of the trade regime should capture the distortions in relative prices and resource allocation due to the trade regime. In practice, construction of such measures requires very substantial amounts of information and analysis. As discussed in Appendix I, in comparison with the main previous studies of trade liberalization in developing countries, the index used here has several advantages. Previous studies have (1) presented tariff and NTB measures separately without attempting to form an overall index of restrictiveness; or (2) developed very simple binary indicators of openness; or (3) conducted massive case studies to develop quantitative summary measures, such as effective rates of protection and effective exchange rates for exports and imports. This study adopts an intermediate approach that goes further than the first two categories toward providing comprehensive, objective measures of the overall stance of countries’ trade regimes, while avoiding the large resource costs of the third method. While the results of the latter may be “state of the art” in their greater precision, they too are subject to important theoretical limitations.

Review of Multiyear Fund-Supported Programs, 1990–96

The overall openness of the trade systems prior to the beginning of programs covered by the review is summarized in Figure 1.15 As can be seen, at the start of the programs, two-thirds of the countries reviewed were characterized by restrictive trade regimes, with classification of 8–10 on the index.16 Indeed, almost 40 percent of the countries reviewed had highly restrictive regimes, classified at 10 on the index.17 The preponderance of ratings in the restrictive category clearly highlights the need for trade liberalization.

Figure 1.Distribution of Overall Restrictiveness, Initial Conditions1

Source: IMF staff estimates.

1“Initial conditions” correspond to Fund arrangements that came into effect as indicated in Table 2.

Reflecting the profile of overall restrictiveness, import tariffs could also be classified as generally restrictive; tariff rates were widely dispersed for many of the programs reviewed. Exemptions also add to tariff dispersion; the existence of exemptions, which are not reflected in the tariff measures included in the index, was explicitly noted in the initial conditions in half of the 28 programs.

Regarding NTBs, again reflecting the profile of overall restrictiveness, many ratings were at or near the restrictive end of the scale. In addition to the many cases of quantitative restrictions on imports and exports (in the forms of quotas, bans, or restrictive licensing), state trading monopolies, usually covering commodities described as vital to the national interest, were mentioned in 17 programs.

Trade Reform Objectives

The extent of movement toward an open trade regime targeted in the Fund-supported programs reviewed was mixed (Table 2).18 The average targeted improvement in restrictiveness was slightly less than 2 points on the 10-point scale, with a range of zero to 5 points (Figure 2). Almost one-third of the 28 programs (9 programs) did not target an improvement in openness large enough to register on the 10-point scale; 8 programs targeted an improvement of 1–2 points; and 11 programs targeted an improvement of 3 or more points.19

Table 2.Initial Overall Rating and Targeted Change in Restrictiveness Rating1
Initial Overall RatingTargeted Overall ChangeTargeted NTB ChangeTargeted Tariff ChangeTargeted Overall RatingFinal Overall Rating
Mozambique (1990)10-4-1-266
Zimbabwe (1992)10-4-1-268
Burkina Faso II (1993)10-3-1077
Hungary (1991)10-3-1076
Bangladesh (1990)10-2-1088
Comoros (1991)10-2-1088
Egypt (1991)10-2-1088
Ethiopia (1992)10-2-1088
Guyana (1990)10-20-288
Burkina Faso I (1991)100001010
Jordan (1992)100001010
Tanzania (1991)9-4-1-155
Poland (1991)9-1-1-187
Mauritania (1992)900099
Panama (1992)8-4-1-148
Mali (1992)8-3-1055
Lesotho (1991)800088
Philippines (1991)800088
Sri Lanka (1991)7-5-1-222
Nepal (1992)7-4-1-133
Zambia (1992)7-3-1044
Benin (1993)6-3-1036
Equatorial Guinea (1993)5-20-233
Argentina (1992)500054
Sierra Leone (1992)500054
Jamaica (1992)4-10-133
Mongolia (1993)300033
Peru (1993)300033
Source: IMF staff estimates.

The methodology of combining NTB coverage and average import tariffs into an overall index of trade restrictiveness is described in Appendix I. Each program’s effectiveness date is indicated in parentheses.

Source: IMF staff estimates.

The methodology of combining NTB coverage and average import tariffs into an overall index of trade restrictiveness is described in Appendix I. Each program’s effectiveness date is indicated in parentheses.

Figure 2.Targeted Change in Overall Restrictiveness

Source: IMF staff estimates.

Programs supported by Fund arrangements generally targeted at least some reduction in tariff protection, but in many cases these were modest. Slightly more than one-third of programs targeted a reduction in tariffs sufficient to change the classification of tariff protection. Regarding NTBs, slightly more than one-half of programs targeted a reduction sufficient to lower the classification of the restrictiveness of NTBs.

The sequencing of trade reform in these programs generally first targeted NTBs, which was followed or accompanied by a reduction in tariff protection. For the 18 programs in the restrictive classifications, 13 targeted movement toward more liberal regimes, and 11 of those 13 incorporated a reduction in NTB classification; only 2 of the programs with these countries targeted tariff reform alone. This indicates that Fund programs generally focused initially on removing NTBs as the most distortionary element of the trade system.

End of program Outturns

The overall openness of the trade systems at the conclusion of the programs reviewed is summarized in Figure 3. The number of countries in the restrictive category declined by one-third, from 18 to 12 and a corresponding increase occurred in the open category. The average actual movement along the 10-point scale was slightly less than 2 points. However, the extent of trade liberalization varied considerably, with a broadly even division between those with no change in overall trade restrictiveness; an improvement by 1 or 2 points; and an improvement by 3 points or more. In four cases, countries liberalized more than targeted under the program, or liberalized unilaterally if the program did not contain trade policy targets. Actual liberalization was less than targeted under the program in only three cases.

Figure 3.Initial, Targeted, and Final Distribution of Overall Trade Restrictiveness

Source: IMF staff estimates.

Reductions in overall restrictiveness were due to reductions in both tariff and NTBs.20 Ten of the 28 programs reviewed included lowering tariff protection sufficiently to lower the index classification; in 8 of these cases the targeted reduction was achieved. Moreover, although not directly reflected in the index, progress was made in reducing maximum tariffs and exemptions, hence reducing tariff dispersion. Average tariffs were reduced significantly in two cases but still remained very high, so that tariff protection remained classified as restrictive. In three cases, the envisaged reduction in the level of tariff protection was not achieved. A reduction in tariff exemptions was achieved in 7 programs.

Program objectives for reducing NTBs were somewhat more extensive; 15 of the 28 programs reviewed included targeting a reduction in NTB classification. In 13 of these 15 cases, the targeted reduction was achieved. At the end of the arrangement period there were only 5 cases that had the most restrictive categorization of NTBs, compared with 14 at the beginning. On the other hand, of the 18 programs in the restrictive category, which by definition had the highest level of NTBs, 7 did not target NTB liberalization sufficient to lower the restrictiveness classification.

Targeting, Monitoring, and Overall Results

The review found that Fund programs were successful in achieving the targeted change in the trade regime. As shown in Figure 4, in three-fourths of the programs reviewed, the envisaged outcome was achieved, and in a further 14 percent it was exceeded (although it should be noted that these results include the one-third of programs that did not target any change in restrictiveness). This result was equally true for targeted import tariff and NTB liberalization. As most programs targeted some quantifiable change in restrictiveness, and these targets were generally achieved, the aggregate stance of trade policy in the countries reviewed improved markedly.

Figure 4.Overall Trade Liberalization, Final Outcome Versus Program Objectives1

Source: IMF staff estimates.

1 Includes cases where there was no targeted change in the restrictiveness rating.

To an extent, the success in attaining program targets reflects appropriate use of monitoring, namely, prior actions, performance criteria, structural benchmarks, and reviews. As illustrated in Figure 5, binding monitoring21 of at least one trade liberalization measure was used in 23 of the 28 programs and implemented on time in nearly three-fourths of the cases. Of the measures not subject to such monitoring, about half experienced problems with delayed implementation.

Figure 5.Monitoring Instruments Used in Fund Programs

Source: IMF staff estimates.

1Multiple instruments may have been used in a program.

Another notable conclusion of the review is that although some statement regarding medium-term trade liberalization was mentioned at the outset in two-thirds of the arrangements, it was generally insufficiently comprehensive or detailed to permit monitoring of the implementation, either in terms of specificity of targets or implementation dates.22 They were generally not supported by intermediate steps to measure progress in phasing in the targeted liberalization, nor by other non-trade measures that would support the liberalization, such as developing alternative tax revenue sources. These results suggest that trade policy implementation could have been strengthened by placing more emphasis on clearly defined, quantifiable, and monitorable medium-term policy objectives and targets at the beginning of the arrangement. The case studies suggest that this might be further strengthened by the early public announcement of the medium-term targets.

Program Ambitiousness

The following subsections set forth some general observations and factors that may be considered in making judgments about ambitiousness of trade reform and try to assess the ambitiousness of the programs reviewed.

General Principles

The objective of medium-term structural adjustment programs should be to move toward an open trade system characterized by the absence of quantitative restrictions and other NTBs (except for health and security reasons) and low, relatively uniform tariff protection, applied in a transparent and evenhanded manner. Indeed, movement from restrictive to open categories, as measured by the index, means the progressive elimination of NTBs for controlling trade, and reducing tariff protection to low levels by comparison with all Fund members. However, the speed with which this objective should be accomplished is an inherently difficult issue. In principle, such movement should be undertaken as rapidly as possible, but in practice judgements will need to be made on a case-by-case basis, taking account of the particular circumstances of the country concerned.

Clearly a major element determining the pace at which reform should be pursued would be the country’s initial conditions. Trade regimes that are more restrictive at the outset incur higher costs from trade distortions and hence need stronger trade liberalization efforts. Economic principles would seem to indicate that, in sequencing trade reform, one should first target the most distortionary aspects of the trade regime, namely, NTBs, in tandem with other reforms. In addition to reaping the benefits from reducing distortions, the tariffication of NTBs almost invariably increases revenue, as discussed below.

The experience of countries that have achieved markedly improved trade and economic growth performance through the adoption of liberal outward-looking policies is also instructive. The experience of four such “good practice” countries, Chile, Colombia, New Zealand, and Singapore, is reviewed in Box 5.23 These countries made steady and sustained progress in moving from the restrictive to open categorization over a period of about seven to ten years (Figure 6). Their experiences, and the results of the review and the case studies below, confirm that far-reaching trade reform is almost inevitably a long-term process. This highlights the need to avoid policy slippages en route,24 which would make the eventual success of reform and reaping its economic benefits an even more drawn-out process. It also underlines the fact that, for countries with Fund-supported programs, the reform efforts directed toward the goal of openness may need to be viewed as a continuous process over more than one medium-term arrangement. Both of these factors underscore the need for specified and monitorable medium-term trade targets to be included in the initial objectives of arrangements. The relatively long-term nature of successful reform reinforces the need for close Fund-Bank collaboration, as the intensity of the Fund or the Bank’s separate involvement may not be even over the full period of the trade reform effort.

Figure 6.Trade Policies in selected “Good Practice” Countries

Source: IMF staff estimates.

The need for commitment to liberalization on the part of the authorities also underscores the importance of dealing with the nontrade considerations that may constrain trade liberalization on a medium-term basis. As noted above, these include the real or perceived fiscal constraints, administrative capacity, particularly of tax administration, the short-term adjustment costs and disruption involved in any reallocation of economic resources, and the lobbying and rent seeking that are so often associated with trade-related activities.

Assessing the Programs Reviewed

Given the above considerations, it is difficult to translate the range of the trade liberalization objectives of the programs covered by the review into a general assessment of the ambitiousness of medium-term Fund-supported programs. Clearly a number of programs targeted far-reaching and extensive trade reform; the trade reform objectives of Sri Lanka’s 1991 ESAF, for example, were ambitious by any standards, as were the reforms targeted in a significant number of other programs. However, if the goal of progressive movement toward an open trade regime is accepted, then many programs may also be characterized as clearly unambitious. In particular, almost one-third of the total did not target a quantifiable reduction in trade restrictiveness, as measured by the index.25 Excluding the programs of the three countries classified as having open regimes, 28 percent of countries did not target a reduction in restrictiveness. Also, as shown in Table 3, programs in countries classified as restrictive at the outset of the program were not more ambitious in targeting trade reform than programs in countries classified as moderate. In fact, the extent of trade reform targeted was on average less for restrictive than for moderate regimes, and the proportion of programs that targeted no quantifiable trade reform was the same for both categories. This may reflect the fact that highly restrictive countries are those that have resisted liberalization more strongly.

Table 3.Ambitiousness of Programs Reviewed
Initial ClassificationNumber of ProgramsAverage Targeted Change in IndexPrograms That Targeted No Change in Overall Restrictiveness
Restrictive182.05 (28 percent)
Moderate72.42 (29 percent)
Source: IMF staff estimates.
Source: IMF staff estimates.

On balance, the findings of the review indicate that, notwithstanding the difficulties of judging the ambitiousness of trade reform efforts, a substantial number of programs targeted and achieved extensive trade reform. However, more liberalization of the trade system should probably have been targeted in a significant number of Fund-supported programs, and particularly in those countries with initially restrictive regimes.

Factors Affecting the Implementation of Trade Reform

The review identified the prevalence of a number of factors that might be expected to hinder or promote trade reform. These focus particularly on fiscal considerations as well as the balance of payments, but also include longer-term Fund involvement during 1990–96, World Bank involvement, the impact of the WTO and other international organizations, and regional trade arrangements. The impact of these factors on program design and implementation is discussed in more detail in Section IV.

Not all factors affecting reform, most significantly “ownership” and political economy constraints, could be considered in the review. Other related factors that can constrain trade reform include lack of an adequate social safety net, supply constraints for countries with undiversified economies or inadequate physical infrastructure, and concerns about the effects of trade reform on unemployment and the impact of openness on consumption patterns and income distribution. While critically important to the design, pace, and implementation of reform, the political economy of trade liberalization is beyond the scope of this study (Box 3).

Box 5.Trade Policies in Selected “Good Practice” Countries

This box provides some benchmarks for trade reform by looking at four countries known to have open or “good practice” trade policies and with a variety of experiences. The countries, Chile, Colombia, New Zealand, and Singapore, are examples of reformers with strong export growth and economic performance. Their trade regimes have been classified according to the 10-point scale from 1984 to 19961 and Figure 6 illustrates how they have opened their trade regimes during that period. Open trade regimes have been a major factor in the strong trade and growth performances of the good practice countries. During 1986–95 their annual export growth averaged 13 percent, while real GDP growth averaged 5 percent. New Zealand’s growth record is the most striking—growth shifted from about 1 percent a year to as high as 6 percent after the trade and other major reforms, although some of the increase may be due to cyclical factors.

Chile is often cited as an example of a successful trade reformer. During the 1980s, it virtually eliminated NTBs. Tariffs, which had earlier been increased in response to external shocks, were progressively reduced to a uniform rate of 11 percent.2 The share of imports facing NTBs during 1991–93 was just 0.1 percent.

When Colombia started reforms in 1984, over 80 percent of its imports required permits, while imports of some 800 products were prohibited and tariffs averaged near 60 percent. Over the next decade, NTBs were virtually eliminated and the average tariff was reduced to under 12 percent.

In Singapore, the proportion of tariff lines covered by NTBs was reduced to 1 percent in the late 1980s and to less than ½ of 1 percent in the early 1990s. Singapore’s tariffs have averaged less than 1 percent since 1980.

In 1984, New Zealand’s tariffs averaged 16 percent and most imports were subject to restrictive import licensing. By 1993, NTBs applied to less than ½ of 1 percent of imports, and tariffs had been cut to 8.5 percent, with continuing reductions programmed to reduce the average to 3.0 percent in 2000.3

1 Background sources for information on tariffs and nontariff barriers were PECC (1995); GATT (1990); GATT (1991); WTO (1996a, 1996b, and 1996c).2 Although Chile’s average rate is just above the level needed for a rating of 1 under the 10-point classification, the uniform tariff is likely much less distortive than some other tariff regimes with an average tariff several percentage points below Chile’s.3 New Zealand’s 1996/97 tariffs averaged 6.2 percent.

Balance of Payments and Fiscal Considerations

Meaningful and sustainable trade reform needs to be developed as part of a coherent package of structural and macroeconomic policy reforms.26 Fiscal and balance of payments positions are central to the macroeconomic framework, and appropriate exchange rate policies are an integral part of Fund-supported programs. Balance of payments concerns may influence the extent and pace of trade liberalization, particularly if appropriate exchange rate policies are not adopted and if existing restrictions are thought to be needed to manage the balance of payments position to preserve macroeconomic stability.27 In the review, balance of payments difficulties were cited explicitly only once, in connection with the introduction of a temporary measure that increased the restrictiveness of the trade regime. There was no evidence that exchange rate considerations directly influenced the extent of trade reform commitments or implementation.

Fiscal policy concerns28 are frequently cited as a reason for delaying trade reform. In practice the fiscal impact of liberalizing the trade system could be positive, negative, or neutral, depending on the reforms introduced and the specific circumstances of the country concerned. It should be noted that in the context of Fund programs the ratio of trade taxes to GDP may increase during trade reform, because of factors such as exchange rate changes, increases in the import-to-GDP ratio (including the impact of increased donor financing), and improvements in tax and customs administration. Although lowering the average tariff in a system with already low or moderate tariff rates will, other things being equal, generally reduce revenue,29 at least in the short term, the fiscal impact of most other aspects of trade reform is likely to be positive or ambiguous (Table 4). Tariffication of QRs and other NTBs on imports or exports will improve government finances, possibly significantly, as will the elimination of tariff exemptions (excluding export duty drawback schemes) and trade-related subsidies.30 Reductions in very high tariffs may also increase fiscal revenue, if the effect of expanded imports outweighs the reduced tax rate.31 Simplification of the tariff system through a move to more uniform rates or a reduction in the number of rates may also tend to increase fiscal revenue through increased transparency and simplification of tax administration. Even lowering export taxes may, in some circumstances, have a positive impact on revenue, depending on the extent and speed with which it expands total trade and reduces illegal activity such as smuggling.

Table 4.Fiscal Impact of Trade Reform Elements
Trade ReformExpected Fiscal Impact
Reduce moderate or low average tariffNegative
Reduce high average tariffAmbiguous
Lower maximum tariffAmbiguous
Reduce tariff dispersionAmbiguous/positive
Replace quantitative restrictions with tariffs and liberalize other nontariff barriers on tradePositive
Eliminate tariff exemptionsPositive
Eliminate trade-related subsidiesPositive
Eliminate state trading monopoliesAmbiguous/positive
Eliminate export taxesAmbiguous/negative
Sources: IMF and World Bank staff estimates.
Sources: IMF and World Bank staff estimates.

The net fiscal impact of trade reform will depend crucially upon the mix of elements in the trade reform package, as well as the countries’ initial circumstances. Clearly the above argues for the early focus of trade reform to be broad based and to include the likely positive revenue elements. Papageorgiou, Michaely, and Choksi (1991), World Bank (1992), and Thomas, Nash, and others (1991) each examined individual country cases from the 1970s and 1980s and found that trade reform may well be associated with an increase in revenue from import taxes when elimination or tariffication of nontariff barriers is an important focus of the trade reform.

Even if a first-round revenue loss from trade reform is anticipated, a preferred policy response would be to increase revenues from other less distorting, broader-based taxes,32 such as value-added taxes (VATs) applied equally to imports and domestic production.33 Optimal tax theory starts with the objective of raising revenue with the least economic distortion.34 A tariff is an inefficient tax that tends to raise prices on both imports and their domestic substitutes while only collecting revenue on imports. The distortion to the economy from taxing both imports and domestic substitutes at equivalent tax rates is generally less than that of taxing imports alone, and the revenue yield is greater from taxing both.35

Tax administration concerns also favor tariff reform in spite of the low cost of administering trade taxes relative to the revenue; this is true both because many elements of tariff reform simplify tax administration by reducing evasion or eliminating exemptions and also because the broader-based alternative tax sources generally increase collected revenue, with less economic distortion. Many of the prescriptions for trade reforms, particularly low, relatively uniform tariffs with few exemptions and an absence of NTBs, tend to make revenue collection more efficient through simplification and transparency. Fund tax policy advice has generally advocated simple tax rate structures with low rates and broad-based application, recognizing that overly complicated tax systems tend to generate inefficiencies, high compliance costs, tax evasion, and rent seeking.

Nevertheless, there are some fiscal arguments for gradualism in trade reform. Trade taxes may be an important source of fiscal revenue, particularly in low-income countries (see Stotsky, 1995). At some point in the process of tariff reduction, further reduction always lowers revenue, at least in the short run. To the extent trade taxes account for a significant part of fiscal revenue and total revenue is low, fiscal considerations may be an important factor in the design and pace of trade reform, particularly regarding lowering average tariffs. A shift to a broader-based tax, such as a VAT, will reduce reliance on trade taxes, but generally requires careful preparation. Further, a country may have difficulty collecting taxes because of an overly complex tax system, lack of effective sanctions for tax offenders, lack of trained personnel, tax avoidance and evasion, and generally weak administrative capacity (Faria and Yücelik, 1995). In these cases, technical assistance to improve the efficiency of tax administration and increase enforcement should probably precede implementation of substantial trade tax reforms that are likely to reduce fiscal revenues.

In addition, to the extent that adjustment programs tend to be introduced in countries with relatively weak and undiversified fiscal systems and large fiscal imbalances, mobilizing alternative revenue sources and diversifying tax sources away from trade taxes may be a longer-term project, which also argues for technical assistance at an early stage. In these cases, however, there is still no reason to delay implementation of trade reform elements that increase revenue or are revenue neutral. Indeed, the difficulties in collecting more efficient taxes strengthen the case for proceeding rapidly with the revenue-increasing elements of trade reform, particularly the tariffication of QRs and other NTBs and curtailing exemptions.

With regard to the impact of fiscal considerations in undertaking the trade reforms for the countries covered by this review, the fiscal conditions likely to present the most difficulties would include relatively heavy dependence on trade taxes, relatively low revenues in relation to gross domestic product (GDP), and a weak overall fiscal position. On this basis, the review included four indicators to assess initial fiscal conditions: (1) the ratio of trade taxes to total taxes; (2) the ratio of trade tax revenue to GDP; (3) the ratio of tax revenue to GDP; and (4) the targeted change in the ratio of budget deficit to GDP.

The results for these indicators are summarized in Figure 7. As can be seen, a direct relationship between the initial fiscal conditions and the ambitiousness of trade reform targeted, including for the ratio of trade taxes to total taxes, is not evident for the programs reviewed. Only five cases were found in which the initial fiscal conditions may have been expected to act as a constraint to trade reform. In those cases, countries lacked a VAT and at least two of the three conditions of high dependency on trade taxes—low revenue effort and large targeted reduction in the budget deficit—were satisfied.36 Despite these adverse initial conditions, three of these countries targeted and implemented trade reform of at least 3 points on the 10-point scale, and the other two countries targeted and implemented trade reform of 2 points. The evidence available in this review, albeit limited, indicates no direct relationship between fiscal conditions and trade reform objectives and that trade reform should not necessarily be ruled out because of a backdrop of difficult fiscal circumstances. Moreover, while it is beyond the scope of this study to assess the availability of alternative revenue sources and other fiscal actions, the review indicated that exemptions were noted as a problem in at least half of the cases, state trading monopolies in more than half the cases, and as indicated above, NTBs were prevalent throughout. Moreover, the evidence from the case studies, while also limited, suggests that more attention could have been given to fiscal policies that would support trade reform.

Figure 7.Initial Fiscal Indicators for the Programs Reviewed

Source: IMF staff estimates.

Nevertheless, the problems posed by introducing trade reform should not be minimized, particularly in cases where that reform is perceived as likely to have negative first-round fiscal implications. Even if alternative revenue sources or expenditure restraints appear to be readily available, action in these areas is likely to pose its own political and economic challenges. Moreover, the consideration of fiscal alternatives will be in the context of program design, which usually involves fiscal pressures in many areas of both revenue and expenditures, including outlays in support of structural reforms other than trade policy. This underlines the importance of introducing broad-based trade reform programs that include, at an early stage, elements such as the tariffication of QRs and other NTBs, reduced exemptions and trade-related subsidies as well as early technical assistance to broaden the tax base and improve tax administration.

Longer-Term Fund Involvement During 1990–96

As noted above, the review provides a snapshot of trade reform in Fund-supported programs rather than a comprehensive analysis in the context of Fund relations with countries over a longer period. Such longer-term involvement in fact characterizes the great majority of the arrangements covered in the review. This raises issues about the degree of liberalization included in any single program, the sequencing and timing of trade reform with other program measures in a longer-term context, and the need for continued steady progress in trade reform. Figure 8 shows all arrangements during the 1990–96 period for each country included in the study. As can be seen, for only two countries (the Comoros and Zimbabwe) was the program reviewed the only one supported by the Fund during the period. Twenty-one programs were followed by a successor arrangement,37 and 20 were preceded by an arrangement. In 15 cases, there was both a preceding and successor arrangement.

Figure 8.Arrangements Included in the Review1

Source: IMF, Transactions of the Fund Yearbook.

1Programs shaded solid blue were included in the review portion of the study. See text for selection criteria and List of Abbreviations. Where overlaps occur, the period of the later arrangement is shown.

On the basis of this longer-term Fund involvement, some might argue that broad macroeconomic imbalances should be vigorously addressed at the outset, with important structural reforms, including trade reforms, addressed extensively in subsequent programs. This does not appear to have been the case in the countries reviewed. First, the initial degree of restrictiveness of the trade regimes of countries that had completed earlier Fund arrangements was similar to that of the entire sample of programs reviewed. Thus, on the basis of the information available in this study, prior arrangements during 1990–96 do not appear to have promoted a start to the reform process. Second, trade liberalization objectives for countries that had already completed Fund arrangements were about the same as in other programs, that is, completion of an earlier arrangement did not lead to trade reform being addressed more extensively in the subsequent arrangement. In contrast, it may be noted that the good practice countries moved steadily from the restrictive to open category, with reform efforts sustained throughout the entire adjustment period. These findings are tentative, as only a short period was surveyed. Nevertheless, they underscore the need for steady and sustained progress to reach the objective of a liberal trade regime.

World Bank Involvement, the WTO, and Other International Organizations

The World Bank’s involvement was cited in three-fourths of the programs reviewed as a factor that influenced trade reform. In two-thirds of the cases, the Bank’s involvement took the form of loans with trade conditionality, and in the remaining cases it took the form of policy advice, combined with technical assistance. The available information provided limited insight into the modalities of the Fund-Bank collaboration and whether it had any impact on the degree of program ambitiousness. However, in the review and also in the case studies in Section IV, there was no indication of differences of view. Further, as noted above, the implementation record of agreed objectives is strong. This suggests that the results of Fund and Bank collaboration in trade policy have been effective in achieving the results targeted but leave open the question as to whether joint involvement increases the ambitiousness of the reform effort.

Trade liberalization by Fund members that is linked to the WTO results from either negotiations on WTO accession or implementation of multilateral commitments. The accession process can take several years and the liberalization linked to accession may be phased over a similar period.38 In the programs reviewed, the WTO was cited as affecting the degree of trade liberalization in three cases; in two, implementation of certain trade measures was part of the countries’ liberalization obligations under WTO agreements; and in the other, trade liberalization was linked to WTO accession. Hence, the effect of WTO obligations appears to have been limited for the programs reviewed. Interaction with other international organizations was limited to two instances where regional development banks were mentioned as having a modest influence on trade liberalization.

Regional Trading Arrangements

Regional trading arrangements (RTAs) are numerous and can play an important role in a country’s trade policy stance (Box 6). IMF staff have, with few exceptions, taken up trade reform issues with the individual RTA members rather than on a regional basis, even in the case of customs unions.

In practice, membership in an RTA has served either to promote or inhibit the trade liberalization of individual countries, depending on whether a country’s trade liberalization objectives are more or less ambitious than those of its RTA partners and whether the RTA constrains unilateral action by members setting common external tariffs and other trade policies, as is the case for customs unions. If the objectives of the RTA are progressive in relation to the initial position of its members, it may promote more rapid liberalization than might otherwise occur (as appears to have been so with Guyana and Jamaica in this review). In other cases, such membership may have acted as a constraint to trade liberalization. Although cited specifically in only five of the country cases reviewed, the influence of RTAs may be an increasingly important factor in trade policy because of the growing extent and coverage of RTA membership;39 during 1990–97, so far 62 RTAs were notified to the WTO, compared with only 13 during the 1980s.

Box 6.Regional Trading Arrangements

Regional trading arrangements (RTAs) and other regional trade initiatives are numerous and can play an important role in the trade policy stance of Fund members. The Fund has stressed that nondiscriminatory trade liberalization on a most-favored-nation (MFN) basis is the first-best policy with the ultimate objective being global free trade. RTAs should be implemented in a manner that promotes rather than impedes this objective. The following principles are more likely to result in RTAs that serve as building, rather than stumbling blocks, to multilateral liberalization: (1) the RTA should include all sectors; (2) there should be a short and concrete schedule for full liberalization; (3) substantial external liberalization should precede or accompany preferential liberalization to avoid trade diversion; (4) rules of origin should be transparent, liberal, and uniform; (5) customs unions should strive to adopt low common external tariffs; (6) use of antidumping and countervailing duty actions should be subject to strict disciplines; (7) the economic benefits of RTAs can be enhanced by deeper forms of integration (e.g., liberalization of trade in services); and (8) rules of accession should be liberal.

The WTO rules governing RTAs take account of some of these principles. They specify that substantially all trade in goods between members should be covered by an RTA; external trade barriers to third parties should not be raised; trade barriers within RTA partners should be eliminated within a reasonable period of time; and RTAs should be notified to the WTO. Nearly all WTO members have reported the WTO that they are parties to at least one RTA.

The impact of RTAs on trade policy varies considerably. While trade liberalization on an MFN basis is an important complement to preferential liberalization between RTA members, the modalities for achieving such liberalization pose challenges in the case of customs unions; these challenges are not present in free trade agreements. Customs unions include common external tariffs and other common trade policies that constrain the trade policies of member countries. In this case, the effect of the RTA will depend, inter alia, on whether a country’s trade liberalization objectives are more or less ambitious than those of its RTA partners. If the ambitiousness of the RTA’s trade liberalization is greater or smaller than that of the individual country, it will quicken or slow, respectively, trade reform of that member country. In contrast, free trade agreements usually include no restrictions on external MFN liberalization.

    Other Resources Citing This Publication