Trade Conditionality Under Fund-Supported Programs, 1990-2004

Trade Conditionality Under Fund-Supported Programs, 1990-2004


Trade Conditionality Under Fund-Supported Programs, 1990-2004

I. Introduction and Overview

1. Trade reform has been a key element in a number of Fund-supported programs over the years. The presence of trade-related measures stems in large part from the importance of an open trade and exchange regime to structural adjustment and growth, as well as its contribution to good governance. As with many other aspects of the Fund’s work, staff have periodically reviewed the trade policy content of Fund-supported programs (Box 1).

Recent Studies on Fund Trade Conditionality

Four trade-related reviews were conducted during 1992-2001. The 1992 study reviewed trade reforms in Fund-supported programs in the second half of the 1980s. More recent surveys and their findings are summarized below.

The 1994 review examined the extent of trade reform in Fund-supported programs in 59 countries during 1990-93. Evidence was found of significant trade reform, especially in lowering quantitative restrictions (QRs) as compared to tariff reforms. In the sample, reversals in trade reform were limited, and often reflected competitive pressures as real exchange rates appreciated, slow domestic tax reforms, or political difficulties in resisting domestic protectionist lobbies. Conflicts between short-term fiscal objectives and medium-term trade reform goals often surfaced, leading to the conclusion that early restructuring of the domestic tax base, liberalization of exchange systems, and exchange rate flexibility were important complements to sustainable trade reforms.

A subsequent study in 1997 covered multiyear Fund arrangements between 1990 and mid- 1996. Utilizing a newly developed index of overall trade policy restrictiveness and six detailed case studies, the review found that most programs targeted a marked reduction in trade restrictiveness, and the targets were generally achieved, especially when they were well specified in a medium-term context and publicly announced. Like the 1994 review, fiscal considerations exercised a significant influence on the pace of trade reform. Cooperation with the World Bank was also observed to be important in shaping trade reform, while regional trading arrangements occasionally set some of the parameters.

The 2001 review was conducted in the context of an overall examination of structural conditionality and focused on programs during 1997-99. Particular attention was paid to the restrictiveness of trade regimes at the outset of programs, whether conditionality was focused on the most restrictive aspects of countries’ trade regimes, and the incidence of trade policy conditionality across types of programs. The review uncovered only a few cases where trade policy reform was considered critical to the success of the program. Structural conditions on tariff reforms were mostly broad-based, while those on NTBs were generally sector-specific and often related to exports. The number of trade-related conditions averaged somewhat less than one measure per program year, with a slight rise during the second half of the 1990s associated with increased use of benchmarks and prior actions. Trade conditionality was applied mainly through structural benchmarks, prior actions and program reviews, with performance criteria used sparingly. Implementation rates were very good. In several programs, the member’s receptivity to trade reform may have been facilitated by the coincidence with WTO accession negotiations or Uruguay Round reforms. Notwithstanding the decline in the World Bank’s direct involvement in trade-related operations during the 1990s, the Fund continued to rely on the Bank in this area, especially in the design of sector-specific policies.

Sources: Issues and Developments in International Trade Policy, World Economic and Financial Surveys, 1992; International Trade Policies: The Uruguay Round and Beyond: Volume II., World Economic and Financial Surveys, 1994; Trade Liberalization in Fund-Supported Programs, 1997, and Trade Policy Conditionality in Fund-Supported Programs, 2001.

2. This paper is a continuation of earlier cross-country reviews of trade conditionality in Fund programs. The paper is largely descriptive, and does not seek to draw out the policy implications of the trends in trade conditionality—some of which are best addressed through detailed case studies, while others are taken up in a broader examination in the Review of the 2002 Guidelines on Conditionality (forthcoming), in the main paper covering the Fund’s work on trade (SM/05/47), and the companion paper on revenue considerations in trade reform.1 This paper examines the nature and design of Fund- supported trade policy reforms, and assesses the implementation of trade policy conditions. The current exercise covers a longer time span (15 years) than earlier reviews, and delves more deeply into the types of trade measures, and the motivation behind these measures.

3. The review offers a number of useful findings. The incidence of trade-related program commitments was heaviest in cases where it might be expected—countries that began Fund-supported programs with restrictive trade regimes, where governance was poor, or where the country in question was seeking closer regional or multilateral integration. Both the incidence and the nature of trade-related conditions changed over the years covered by the review. The steady increase in the number of conditions through the 1990s was essentially reversed in 2001-04, reflecting the streamlining of structural conditionality, the generally more liberal trade policy environment, and possibly the advent of the Doha Round of multilateral trade negotiations. The nature of trade-related measures also shifted from a focus on nontariff barriers (NTBs), to tariff reforms, and finally to customs administration.

The focus of the remainder of the paper is as follows. Section II reports on the findings of the review, and analyses the motivations for trade reform, their design, factors affecting conditionality, changes in trade conditionality over time, and the implementation of trade-related conditionality. Section III offers a summary and some conclusions.

II. Findings

5. The review examined trade conditionality in Fund-supported programs in a selected group of 34 countries2 over the period 1990 to 2004 (Box 2). The focus was on the incidence and type of trade conditionality across countries and programs, the motivations behind trade conditionality, the pattern of trade reform advocated, and the implementation rate of reforms.

Methodology of Review

The review covered the trade policy content of 34 countries’ Fund-supported programs that were initiated between January 1, 1990 and December 31, 2001. For purposes of continuity, the countries chosen were those with programs of at least two years’ duration after 1995, and were in the Fund’s 1997 review and/or the 2001 review of structural conditionality (Table 1). Three additional countries—Haiti, Kazakhstan, and Pakistan—were added, based on a judgment of the relevance to the Fund of their particular experiences in trade reform and for further regional balance. The result was an analysis of 138 Fund-supported programs, under a variety of arrangements: 58 Stand-by Arrangements, 22 arrangements under the Extended Fund Facility (EFF), 2 under Emergency Post Conflict Assistance (EPCA), and 56 arrrangements under the Structural or Enhanced Structural Adjustment/Poverty Reduction and Growth Facilities (SAF/ESAF/PRGF).1

The assessment of trade policy conditions was based on an examination of actual program documents, in particular Fund staff reports and letters of intent/memoranda of economic and financial policies. From these documents, the types of conditions were determined, initially classified as general commitments (plans for reform but not subject to specific conditionality) or “binding commitments” (performance criteria, prior actions, structural benchmarks and review clauses). For each condition, the nature of the measure was determined (comprehensive reform program, QRs, tariffs, state trading, customs administration reform, export subsidies, export processing zones, and standards). An attempt was also made to establish their principal objective (such as revenue, balance of payments adjustment, efficiency, streamlining administration, governance, or income distribution), whether they were directed at the agricultural sector, their timing (short- or medium-term), implementation record, and relation to preferential arrangements, WTO commitments, and World Bank trade reform.

1 While the number of Stand-By Arrangements and SAF/ESAF/PRGF programs were about the same, the time period covered by the latter was about three times as long. Most of the countries had an unbroken span of Fund support (of whatever type) of around 8 or 9 years.
Table 1.

Conditionality on Tariffs and NTBs1 and Trade Restrictiveness2 during Fund-Supported Programs

article image

A subset of binding trade commitments included in estimation of the Fund's trade restrictiveness index (TRI) - relating to tariffs, quotas, licensing, and state trading

According to the Fund's TRI. "More restrictive" refers to a TRI rating of 5 or higher; "less restrictive" refers to a TRI rating of 4 or lower.

A. Objectives of Trade Policy Reform

6. To ascertain the objectives of trade policy reform in Fund-supported programs, “motivations” were assigned to each trade measure found in the review of country and program documents.3 The analysis revealed that the dominant motivations for trade reform were to improve economic efficiency and to streamline trade policy administration(Figure 1)4 Often, comprehensive trade reforms were part of a package of policies aimed at improving the business environment, and included such complementary measures as decontrolling domestic prices and privatization of state enterprises.

Figure 1.
Figure 1.

Motivation / Objectives of Trade Policy Reforms 1990-2004

Citation: Policy Papers 2005, 020; 10.5089/9781498331845.007.A001

7. Improving the structure of government revenue was another important factor that appeared to motivate the inclusion of trade measures in Fund programs.5 Programs often included fiscal reform packages, incorporating measures to reduce the reliance on trade taxes, and/or to strengthen collections through better customs administration and a streamlining of tariff bands. The pace of trade tax reduction was sometimes modified to protect the balance of payments and/or ensure adequate time for other tax measures to take effect so as to protect overall government revenue. This was especially the case when complementary measures, such as exchange rate flexibility or higher nontrade taxes, were not available in the short run.6 The companion paper on Dealing with the Revenue Consequences of Trade Reform (SM/05/57, Supplement 2) observes that fiscal revenue loss has been mitigated where tariff liberalization was accompanied by a purge of customs exemptions. In this context, it is interesting to note that 41 percent of Fund-supported programs during the period that carried conditionality to liberalize tariffs also had conditionality on exemptions (either concurrently or in subsequent programs).7 A majority of these programs also had associated technical assistance in support of tax and tariff reform.

8. Improved governance was also cited as a motivation behind trade reform.8 This was most notably the case for improvements in customs administration, intended to help curb corruption, smuggling, and underinvoicing. Programs aimed at enhancing the transparency of licensing systems, or at dismantling restrictive lists, trade quotas, and licensing regimes as means to curtail rent-seeking activities. It was also recognized in program documents that the tariffication of quantitative import restrictions would help to transfer some of the rents reaped by import license holders to the government.9

B. Design of Trade Policy Reform

9. Trade policy objectives often take the form of “general commitments”— frequently describing a medium-term framework for trade policy within which the country indicates it will operate. These commitments may set the parameters for designing conditionality under the Fund arrangements, which can influence Fund financing of programs and may appear in the same or subsequent Fund-supported programs.

10. About three-fourths of the programs under review contained general commitments to trade liberalization. In a number of cases, this involved setting out comprehensive reform strategies, such as in Egypt, Indonesia, Jordan, Kazakhstan, Mexico, Mongolia, Peru, Tanzania, and Yemen.10 Most programs, however, included specific measures to reduce trade restrictions, drawing on a broad menu of instruments reflecting country circumstances. Commitments relating to tariffs (often tariff ceilings),11 NTBs and customs administration were about equally frequent over the period as a whole (Figure 2). Tariff-related measures included reducing reliance on reference prices, replacing specific tariffs with ad valorem taxes, and limiting exemptions. Other trade measures, such as lowering average rates or narrowing tariff dispersion, commitments on export taxes, and the phasing out of state trading were less prominent. Export subsidies were rarely mentioned.12

Figure 2.
Figure 2.

Nature of Program Commitments, 1990-2004

Citation: Policy Papers 2005, 020; 10.5089/9781498331845.007.A001

11. Trade measures included as trade conditionality in Fund-supported arrangements were less common than general commitments (Figure 2 above).13 Of the 138 programs surveyed, 59 (43 percent) had no conditionality on trade measures. When examined by country, however, only 7 out of 34 never had trade-related conditionality, implying that over time most countries accepted some form of binding trade-related conditions. Still, the gap between the frequency of general intentions and conditionality suggests that, while trade reforms were viewed as an important part of government economic plans, they were less often closely or critically related to the objectives of the Fund-supported program.

12. Structural benchmarks have been the preferred form of trade conditionality (Figure 3).14 Prior actions were next in terms of frequency—mainly as small, concrete components of broader, well-articulated trade reform plans.15 In fact, prior actions were the modality of choice in GRA cases. Performance criteria on trade were frequent in ESAF/PRGF-supported programs—although there were generally not many per program— but distinctly less prominent in programs supported through the GRA. In only 13 programs (out of 138) was there an explicit requirement for program reviews to assess progress in trade reform.

Figure 3.
Figure 3.

Incidence of Trade Conditionality Across Programs, 1990-2004

Citation: Policy Papers 2005, 020; 10.5089/9781498331845.007.A001

13. In keeping with Fund policy, trade-related conditionality involved measures that were well-defined and relatively easy to monitor. Over the period as a whole, tariff reform and customs reform16 were both subject to conditionality in roughly two-thirds of countries, while NTB conditionality featured in approximately half. Prior actions and performance criteria on tariffs were almost exclusively related to maximum rates, which are the easiest to verify, while commitments relating to tariff averages and dispersion were incorporated mainly as structural benchmarks and review clauses. Where other price-based measures (such as import surcharges and duty exemptions) were included, they took mainly the form of benchmarks, and occasionally of prior actions. For programs with conditionality related to customs reform, it was not unusual to specify more than one condition. In Bolivia’s programs, for example, all of the trade conditionality involved customs administration reform—five items in the 1994-98 ESAF-supported program and seven items in the 19982001 ESAF-supported program.

C. Factors Influencing Trade Conditionality

14. The incidence of trade conditionality was higher in longer term programs, and programs with low-income countries, in certain regions, and in countries with more restrictive trade regimes.

15. Consistent with the medium-term nature of trade reform, the review indicated that longer-term programs (in particular under the ESAF/PRGF), had more trade conditionality than those supported by Stand-By Arrangements—80 percent of programs under the ESAF/PRGFs carried such conditions compared with 42 percent of SBA cases. Moreover, trade conditionality was more frequent in longer term programs in low-income countries than in other countries (EFF-supported). This might be due to the greater incidence of consecutive programs for lower-income countries relative to consecutive EFFs, which could have lengthened the de facto program horizon.17

16. The set of countries reviewed also showed that the incidence of trade conditionality varied by region. In Asia and Pacific, and to a lesser extent in Western Hemisphere countries, the incidence of trade conditionality was markedly less than elsewhere (Figure 4).18 In the case of Asia and Pacific countries, of the five countries and 12 programs included, only Indonesia and Mongolia had some trade conditionality—in the form of a single measure in one program. Half of the eight Western Hemisphere countries surveyed had no trade conditionality in their programs; of the remainder, conditionality was concentrated in Guyana. Regional differences in the range of trade issues covered were minor, with the exception of sub-Saharan Africa and to some extent the Western Hemisphere (Bolivia), where customs reform dominated.

Figure 4.
Figure 4.

Regional Distribution of Trade Conditionality, 1990-2004

Citation: Policy Papers 2005, 020; 10.5089/9781498331845.007.A001

17. The review demonstrated that the initial degree of trade openness had a bearing on trade conditionality. This should, in fact, be expected given the generally positive association between openness and growth in the literature, and the fact that growth is often an objective (or the ultimate objective) in Fund-supported programs. The measure of openness used was the Fund’s trade restrictiveness index (TRI).19 Table 1 presents a summary of the relative incidence of trade-related conditionality, broken down by degree of trade restrictiveness in a given program year.20 Trade-related conditions were more frequent in countries with a “restrictive” trade regime than in countries with less restrictive regimes. Over the period 1990-2004 as a whole, in ESAF/PRGF cases, the average number of trade- related conditions per program-year was 0.36 during years wherein the trade regime was relatively restrictive, versus 0.21 in years with a less restrictive regime. In GRA cases, there is no such difference overall; however, excluding the four transition economies in this sample, all of which had a large number of trade-related program conditions (Bulgaria, Moldova, Russia, and Ukraine), the incidence of trade conditionality in program years with a restrictive regime was almost eight times higher than when the regime was less restrictive. Arguably, the objective of systemic change in the transition economies motivated far greater attention to structural reforms, including trade regimes, even though these were (formally) quite open by the mid-1990s.

18. The review also showed, as did earlier Fund surveys, that membership in regional trading arrangements (RTAs), WTO commitments, and World Bank reforms were positively correlated with trade conditionality in Fund-supported programs (Table 2). Most of the countries surveyed had been members of the WTO since its inception in 1995—several joined subsequently or are making bids for accession, which may have facilitated the ownership of trade policy reforms While the reports examined included several references to the member’s participation in the WTO, there were no cases of conditionality including the implementation of WTO obligations, or the undertaking of new or prospective obligations. Further, Fund conditions on tariffs referred to applied rates and did not require countries to change their bindings at the WTO. There were also no cases of conditionality that appeared inconsistent with WTO obligations. For example, in cases where import taxes were raised for fiscal or other reasons, WTO bound rates were not exceeded.

Table 2.

Overlap Between Trade Commitments in Fund Programs and Commitments to World Bank, WTO and Regional Trade Arrangements

article image

Related in the sense that the measure might be contained in a World Bank adjustment program or sectoral loan, or be part of commitments under a regional trading arrangement.

Refers to general commitments under programs.

Refers to binding trade commitments, such as performance criteria, prior actions, benchmarks and review clauses.

19. Regional trade commitments featured occasionally in Fund trade conditionality.21 One clear example was the structural benchmark in Guyana’s 1990-93 ESAF-supported program which stated that the government would implement the phased reduction of the common external tariff agreed by CARICOM member states.22 Many of the general commitments in Fund-supported African programs referred to the need to implement regional changes in common external tariffs and harmonize customs classifications and procedures.23

20. The World Bank was a key player with regard to trade reforms in many of the countries reviewed. Fund conditionality at times followed up on the detailed content and timetable fleshed out in separate World Bank supported programs. However, in many cases, the emphasis was on general policy intentions rather than conditionality. In Yemen’s comprehensive trade reform which began in 1996, for example, there was only one trade reform measure taken under Fund conditionality.24 Occasionally, another multilateral agency supported comprehensive trade reform—for example, Peru’s Trade Sector Loan with the Inter-American Development Bank.

D. Trade Conditionality Over Time

21. To assess the evolution of trade conditionality over time, programs in the review were divided into four sub-periods in accordance with the date they were initiated: pre-1995; 1995-97, 1998-2000, and 2001-2004. The results are summarized in Table 3.

Table 3.

Trade Conditionality Over Time

article image

Refers to performance criteria, prior actions, benchmarks and review clauses.

22. Changes over the period are striking. The rapid increase in conditionality of earlier years was reversed in the period 2001–04. The share of programs with trade conditionality increased from 55 percent of programs with at least one condition in the pre-1995 period to 70 percent of programs during 1998–2000. However, in the period 2001–04, the percentage fell to 46 percent. Trade conditionality in Stand-By Arrangement- and EFF-supported programs rose through 1997 but declined thereafter. In the period 2001-04, only one EFF- supported program (Jordan), had trade conditionality.25 Trade conditionality in SAF/ESAF/ PRGF-supported programs in general rose through 2000 but then almost halved during 2001-04. In terms of the modalities of conditionality, the frequency was similar in 2001-04 to what it had been prior to 1995, except for structural benchmarks for which it declined by around one-third. Initial data on all forms of conditionality for a broader sample of countries collected as part of the forthcoming review of the 2002 conditionality guidelines also indicates a decline in the incidence of trade conditions in the 2001-2004 period - somewhat sharper than the decrease in overall conditionality during this period.26

23. The range of issues encompassed by Fund trade conditionality also shifted during the period under review. An examination of trade conditions over the period as a whole shows customs administration reform commitments constituting 32 percent of total commitments, followed closely by tariff reform commitments (29 percent) and NTB reform (21 percent). However, when reviewed over the four sub-periods of this study, a distinct change in the relative weight of different types of trade-related conditionality emerges—in the midst of an overall nominal decline in the number of measures (Figure 5).27

Figure 5.
Figure 5.

Trade Conditionality, by Measure (Four sub-periods from 1990-2004)

Citation: Policy Papers 2005, 020; 10.5089/9781498331845.007.A001

24. Even though the bulk of conditionality in the pre-1995 period was on NTBs, such commitments declined sharply during the following periods to the extent that no such conditionality appeared in the reviewed programs during 2001-04. Conditionality related to tariff reform first increased but then declined as well, and in the period 2001-2004 constituted only about 9 percent of trade conditionality. On the other hand, the incidence of customs reform measures (as a share of the total number of trade conditions) remained relatively stable through 1997 at about 21 percent of conditions, before increasing to 42 percent during 1998-2000 and to 66 percent in the period 2001-04. Thus, during 20012004, customs reform was by far the most prominent form of trade conditionality. The shift in the relative weight of different measures as a share of conditionality raises the question of whether there might be a “life-cycle” of trade reform, with initial emphasis on NTBs giving way to tariff reductions and then to customs reform. A closer look at the progression in trade conditionality for individual countries in the sample suggests this life-cycle may be apparent in some cases, but cannot be generalized for the sample as a whole.

25. The review also revealed the emergence in recent years of trade-related issues that were relatively new to Fund-programs. Two such issues were export processing zones and import surcharges.28 Several countries have chosen to employ export processing zones to stimulate exports and encourage investment. Fund conditionality in this area has generally been motivated by concerns about governance, efficiency, and the preservation of fiscal revenues.29 In the case of surcharges, some countries that liberalized their trade regimes have opted to use temporary import taxes to raise government revenue. In these cases, Fund conditionality generally set a timeframe for their phasing out.

26. One explanation for the rise and fall in trade-related conditionality might be that the distribution of program types changed over the study period in favor of SAF/ESAF/ PRGFs. The peak of trade conditionality during 1998-2000 may thus be partly due to the fact that the review sample had twice as many new such programs in that period as it had new Stand-by Arrangements, a ratio far higher than in any other sub-period (as discussed in paragraph 15 above, SAF/ESAF/PRGF-supported programs are more likely to contain trade- related conditionality). As a result, both the earlier rise and subsequent fall in trade-related conditionality would be overstated. But as noted, the incidence of conditionality in SAF/ESAF/PRGF-supported programs also declined in 2001-04. A second significant impulse behind the rise in trade-related conditionality during the 1990s came from Fund- supported programs with European transition countries (particularly Russia, Ukraine, Moldova, and Bulgaria).

27. Another possible explanation—at least for the decline in trade-related conditionality in 2001-04—would be a general trend in members’ own policies toward more liberal trade policies. 30 As discussed previously, the incidence of trade conditionality would be expected to be lower in countries/programs with less restrictive regimes. Table 1 offers some perspectives on this question, keeping in mind that it only reports on program conditions on tariffs and NTBs (i.e., it excludes customs administration and other measures). The table shows a substantial decline in the frequency of conditions. In a total of 90 “program years” there were only 9 tariff- or NTB-related conditions in the countries surveyed during 2001-04. While there were, indeed, relatively more program years associated with less restrictive trade environments than for the period 1990-2004 as a whole, the impact on the average incidence is dominated by the decline in the number of conditions within each group, both restrictive and less restrictive.31

28. A third explanation for the recent decline in trade conditionality is the tightening of criteria for structural conditionality under the 2002 Guidelines on Conditionality.32 These guidelines, and the interim guidance issued two years earlier (in slightly weaker form), set out the requirement that the measure covered by conditionality should be of critical importance for achieving the goals of the member’s program. It is possible that the mediumterm nature of trade reforms reduced their perceived urgency (particularly for those that have no direct and positive fiscal revenue impact), and meant that they did not, in the judgment of many Fund missions, meet the criticality test.

29. Finally, the advent of the of the Doha Round of multilateral trade negotiations (launched in 2001) may also have played a role. Countries preparing for the bid-offer process of trade negotiation at the heart of the round might well have been more reluctant to engage in unilateral liberalization—holding as bargaining chips measures that they might have agreed to under a Fund-supported program.

E. Implementation of Trade Conditionality

30. Consistent with earlier studies, the track record on the implementation of Fund trade conditionality during the review period was good, and on a par with overall conditionality implementation. As shown in Table 4, 71 percent of trade-related conditions during the period, in the sample countries, were implemented on time. Excluding prior actions, this corresponds to an implementation index for trade conditionality of 1.51.33 By comparison, the implementation index for all types of conditionality, reviewed as part of the Review of the 2002 Conditionality Guidelines (SM/05/81), is estimated at 1.53. It is important to note that this comparison is made across two different sample sets (the sample used in this paper is smaller than that of the broader conditionality review), and time periods that do not perfectly match (the conditionality review examines 1995-2003).

Table 4.

Trade Conditionality Implementation

article image

The index assigns a weight of 2 to measures implemented on time, 1 to measures implemented late or rescheduled, and 0 to measures not implemented.

31. The timing of the trade reforms subject to Fund conditionality varied widely. While prior actions and performance criteria were almost invariably to be implemented within a very short timeframe, it was not unusual for the commitments under structural benchmarks and review clauses to span several years. Delays were more frequent for performance criteria and structural benchmarks (which, interestingly, had a similar performance), but more than 80 percent were ultimately implemented. In several cases, administrative or legal delays were responsible for the rescheduling of performance criteria.34 Structural benchmarks were more apt to be modified than performance criteria.35 Delays in observing trade conditions in one program were often followed up in successor programs. In some cases, conditions were included in a review clause or structural benchmark in the initial program were established as prior actions or performance criteria in subsequent programs.

32. Comparing across different types of programs, conditions under GRA arrangements were more likely to be implemented on time than those under the ESAF/PRGF—81 percent versus 65 percent. And in terms of the nature of conditions, the implementation record of tariff reform measures was significantly stronger than for NTBs and customs reforms. The difference in implementation in GRA versus PRGF programs may stem from a number of causes, including lower levels of institutional and administrative capacity. Regarding the difference in implementation across types of measures, it has been suggested that the weaker implementation of NTB and customs reform conditionality may stem from unrealistic expectations, lack of institutional capacity, and inaccurate perceptions on the part of Fund staff and country authorities of the time required to build such capacity.

III. Summary Results

33. The review offers a number of interesting findings. The range and variety of reform conditions across countries shows that country-specific circumstances played a major role in the design of trade conditionality. The principal motivations for trade-related conditions were economic efficiency, administrative streamlining and fiscal considerations. Structural benchmarks have been the preferred modality overall for trade-related conditionality, but in programs supported under the GRA, prior actions tended to be just as common. In terms of the type of trade measures covered by conditionality, customs and tariff reform and the elimination of NTBs were all broadly equally prominent. Customs reform was particularly frequent in African country programs, and—relative to other measures—during the period 2001-04. Longer-term programs, especially those supported by SAF/ESAF/ PRGFs, generally contained a higher number of trade measures. The Fund’s assessment of a country’s degree of openness appeared to have an important bearing on the extent of trade conditionality. This relationship has generally held throughout the past 14 years, and across different types of programs. Finally, the on-time implementation rate of trade-related conditionality was relatively high at 71 percent.

34. The review highlighted important changes to trade conditionality over time. The nature of Fund conditionality shifted from a focus on NTBs, to tariff reform, and finally to customs administration. In fact, during the most recent period—2001-04—there was no conditionality on NTBs and only very limited conditionality on tariffs in the sample of programs under examination. The incidence of trade conditionality increased during the first half of the 1990s. This trend was sharply reversed during 2001-04, most likely as a result of the broader policy to streamline conditionality—albeit the increasing openness of economies played a certain role, and the decline may also have been linked to the advent of the Doha Round of multilateral trade negotiations.


Dealing with the Revenue Consequences of Trade Reform (SM/05/57, Supplement 2), prepared by the Fiscal Affairs Department.


By region, the 34 countries were: (1) Africa: Benin, Burkina Faso, Cameroon, Côte d’Ivoire, Ethiopia, Mali, Mozambique, Tanzania, and Zambia; (2) Asia and Pacific: Indonesia, Korea, Mongolia, Philippines, and Thailand; (3) Europe: Albania, Bulgaria, Moldova, Russian Federation, and Ukraine; (4) Middle East and Central Asia: Egypt, Jordan, Kazakhstan, Kyrgyz Republic, Mauritania, Pakistan, and Yemen; (5) Western Hemisphere: Argentina, Bolivia, Brazil, Guyana, Haiti, Mexico, Panama, and Peru.


There is scope for overlap in motivations. A single measure on tariff reform, for example, can help to streamline administration, improve governance (by eliminating room for misvaluation), and improve revenue. A single measure might thus be assigned multiple motivations. This is reflected in Figure 1.


This was the case for both members’ statements of policy intentions and conditionality for disbursements of Fund resources under an arrangement.


The thrust of customs reform was typically to increase trade tax revenues by increasing trade volume and better controlling exemptions. Where objectives were made explicit in staff reports, such revenue considerations were important motives in 53 percent of cases over the 1990-2004 period, economic efficiency and streamlining in 58 percent, and governance in 22 percent of cases.


These observations on sequencing are taken from Trade Liberalization in Fund-Supported Programs, World Economic and Financial Surveys, 1998


If general commitments in letters of intent are included, the percentage is significantly higher.


The nine countries in the review that had trade measures motivated by governance considerations scored worse on the World Bank’s corruption index than those that did not (-0.71 compared with -0.49, on a scale of -2.5 to +2.5).


Other reasons cited for trade measures included helping the poor, addressing domestic emergencies, and fostering regional integration—examples include Zambia’s plan to remove the duty on mosquito nets to help poor segments of the population deal with the malaria situation (1999/2000 ESAF), and regional trade commitments cited by several African and Latin American countries.


For example, in Egypt’s Stand-By Arrangement-supported program that commenced in mid-1991, the government announced its intention to remove virtually all nontariff import barriers over two years, reduce tariffs in stages and reduce export quotas; such commitments to comprehensive reform also formed part of the follow-up programs supported by an EFF and another Stand-By Arrangement.


In certain cases, programs envisaged temporary increases in import taxes—e.g., in Thailand’s 1997/2000 Stand-By Arrangement-supported program, where the authorities explained the need to increase import surcharges for fiscal reasons, while committing to reverse this increase when possible.


Exceptions include Egypt’s 1991/93 program, and Korea‘s in 1997/2000.


Conditionality, which is linked with Fund disbursements, can take the form of prior actions, performance criteria, structural benchmarks, and review clauses.


Very few countries (Egypt, Jordan, and Pakistan) had programs with trade conditionality which did not include structural benchmarks.


Programs with Egypt, Guyana, Mauritania, Moldova, Russia, Ukraine, and Yemen had five or more prior actions on trade.


Includes measures related to duty drawbacks, administrative reform, exemptions, and reference pricing.


Jordan and Peru were unusual in the sample in both having three consecutive EFFs.


Comparisons of the restrictiveness of trade regimes (using the Fund’s TRI) of the countries in these two regions show them to be relatively similar and less restrictive than the reviewed countries in Middle East, North Africa, and Central Asia and sub-Saharan Africa but more restrictive than European countries.


See separate companion paper for a discussion of the qualities and shortcomings of the TRI (Review of the IMF’s Trade Restrictiveness Index, SM/05/57).


The notion of “conditionality” here relates only to those measures captured in the Fund’s trade restrictiveness index - those related to tariffs, and such non-tariff measures as licensing requirements, quotas, or state trading. The actual year of incidence (the program year in which the measure occurred) is estimated.


An area yet to be fully examined is the extent to which there is interaction between Fund conditionality on trade measures and trade commitments under regional trade arrangements. One key question would relate to whether commitments to RTAs had a significant impact on Fund conditionality, while another issue would be the extent to which participation in RTAs affected country’s implementation of Fund trade-related conditionality.


A performance criterion and structural benchmark in the successor ESAF-supported program also required observance of the CARICOM tariff reduction schedule.


In some countries’ program documentation, e.g., Mexico’s 1995-97 Stand-By Arrangement-supported program, there was awareness of the potential trade-diverting effects of RTAs and the desirability of reducing tariffs on imports from nonpartner countries.


The reforms, supported by an Economic Recovery Credit from the World Bank, entailed elimination of import licensing, replacement of import bans with tariffs and substitution of a 15-band with a 4-band tariff structure.


Of the five EFF-supported programs still active during 2001-04, only one (Jordan’s 1999-02 EFF) included a measure that actually fell within this period. The review examined programs and assigned conditionality to the year the program was initiated; Jordan’s measure therefore does not appear in Table 3.


The differences in sample countries selected, measures included, and definitions of “trade-related” conditionality make a direct comparison between the two data sets impossible. However, the sharper decline in trade conditionality relative to all forms of conditionality supports the view that other factors besides streamlining (such as independent liberalization, the Doha Round, and the possible focus of reform efforts on addressing other economic challenges) may have contributed to the recent decline in trade conditionality.


Note that Figure 5 records measures in the year the program was initiated, rather than when the measure was actually monitored/implemented.


Another recent, albeit less frequent, development is the use of antidumping measures by developing countries.


For example, structural benchmarks in Moldova’s EFF-supported program of 1996-2000 included submission of legislation to restrict free economic zones to export-oriented production and transshipment, and in Yemen’s EFF-supported program (1997-2000) the preparation of an action plan to develop the appropriate legal, regulatory and judicial framework for a free trade zone. The need to limit fiscal leakage from such free zones has been a consistent concern within the Fund.


The average trade restrictiveness rating (as defined by the Fund’s trade restrictiveness index) for each major subregion of the 34 countries surveyed declined as follows between 1995-2000 and 2001-04: Sub-Saharan Africa, from 5.7 to 3.9; Asia and Pacific, from 4.2 to 3.9; Europe, from 4.0 to 2.7; Middle East, North Africa and Central Asia, from 5.5 to 3.8; and Western Hemisphere, from 4.0 to 3.3.


There would have been 36 tariff or NTB conditions (rather than 9) in the 90 Fund-supported programs during 2001-04 if the relative frequency of programs in countries with restrictive and less restrictive trade regimes had remained constant.


Guidelines on Conditionality, Dec. No. 12864-(02/102), Sept. 25, 2002.


The implementation index was devised as part of the forthcoming review of the 2002 conditionality guidelines. The index assigns a weight of 2 to measures implemented on time, 1 to measures implemented late or rescheduled, and 0 to measures not implemented.


In Jordan’s 1999-02 EFF-supported program, implementation of a performance criterion related to the reduction in the maximum tariff was late because of a delay in the issuance of a royal decree enacting the required legal amendment. In Mauritania’s ESAF-supported program of 1995-98 the announcement of the details of the initial phase of trade reform had to await conclusion of a background study. Administrative problems also occasionally caused slippages in the timing of customs reform.


During negotiations of Fund-supported programs, the pace of trade reforms was sometimes modified to take into account fiscal considerations—the “revenue adjusted” pace of tariff reform was then somewhat slower than otherwise.

Trade Conditionality Under Fund-Supported Programs, 1990-2004
Author: International Monetary Fund