LENDING HOWTfELMS WWD, HSOW C E ROVED VINOD THOMAS 9274 and AJAY CHHIBBER editors A __ BANK iUtnrcIO S A WORLD BANK PUBLICATION Adjustment Lending How It Has Worked, How It Can Be Improved Vinod Thomas and Ajay Chhibber, editors The World Bank Washington, D.C. This monograph is based on the proceedings of a conference on adjustment lending held September 19, 1988, at the World Bank in Washington, D.C. The conference was organized by Stanley Fischer, vice president for development economics and chief economist, and John Holsen, director of the Country Economics Department of the World Bank. The conference was attended by Bank staff as well as participants from the International Monetary Fund, the U.S. Agency for International Development, the Inter-American Development Bank, the U.S. Fed- eral Reserve, and academic institutions. The volume editors are grate- ful to all the conference participants. Karla Cabana, Lorrie Crutchfield, and Rebecca Sugui helped organize the conference and prepare this document. The views and interpretations in this monograph are those of the authors and participants in the conference and do not necessarily represent the views and policies of the World Bank or of its executive directors or the countries they represent. Vinod Thomas is chief of the Trade Policy Division of the World Bank's Country Economics Department and Ajay Chhibber is a senior econ- omist in the Office of the Vice President for Development Economics and Chief Economist. The cover photograph, Time Machine, is by John Cleave. © 1989 The International Bank for Reconstruction and Development / THE WORLD BANK 1818 H Street, N.W. Washington, D.C. 20433, U.S.A. All rights reserved Manufactured in the United States of America First printing July 1989 ISBN 0-8213-1233-2 Contents Introduction 1 Issues in Adjustment Lending I Types of Adjustment 2 Design of Adjustment 4 Implementation of Programs 5 External Environment 7 Quality of Bank Lending 8 Coordination with Other Lenders 8 Impact of Adjustment Programs 10 Performance under Adjustment 10 Design and Implementation of Adjustment Loans 13 Conclusion 14 The Extent of Policy Reforms 15 Trade Policy 17 Fiscal Policy 18 Public Sector Management 18 Financial Sector Policy 19 Antipoverty Policy 21 Industrial Policy 21 Agricultural Policy 22 Fiscal Effects of Conflicting Policies 22 Applying the Prerequisites for Adjustment Lending 23 Overall Adjustment Program 25 Ownership of the Program 25 Realism of the Program 26 Selected Operational Implications 27 Sector Adjustment and the Macroeconomic Program 27 Conditionality and Tranche Release 28 Monitoring and Evaluating Programs 29 Conclusions and Areas for Further Study 30 Notes 32 Selected Reading 33 iii b I I Introduction The World Bank introduced structural adjustment lending in 1980 as a response to the severe balance of payments problems facing many developing countries. These problems had their origins in the second oil shock and in weaknesses of domestic policy. Although adjustment lending in most countries was expected to last from three to five years and the balance of payments was expected to reach a sustainable level in from five to seven years, adjustment lending has intensified rather than disappeared. Over the past decade the terms of trade for most developing countries have deteriorated, real interest rates have in- creased, and the debt crisis and slow growth have persisted. In addition, the sources of domestic inefficiency proved to be more difficult to overcome than previously thought. As a result, the scope of adjustment operations has widened, and the Bank has increasingly undertaken sector adjustment loans. Together with the programs of the International Monetary Fund (IMF), the World Bank's adjustment lending has supported structural adjustment. The intention has been to cushion directly the effect of external shocks and to encourage other external lenders. By linking lending to reforms in policy and institutions, the Bank has tried to help improve a given economy's response to external changes. Yet policy-based lending has given rise to much concern in the Bank and among governments and others. A recent World Bank study examined the performance of adjust- ment programs, evaluated policy reforms, and assessed the design and implementation of adjustment operations.' The results reaffirm the rationale for adjustment lending and its benefits. At the same time, they also suggest some alternatives to ensure that future programs will be more effective. This report, based on a conference held at the World Bank September 19,1988, serves as a companion volume to that evaluation. It points to some unresolved issues, outlines some of the practical implications of the evaluation's findings, and briefly recapit- ulates the evaluation. Issues in Adjustment lending Observers of adjustment lending have underscored its rewards. They have also raised a variety of issues, ranging from the very meaning of adjustment to how it is conceived and carried out. An overview of I 2 ADJUSTMENT LENDING some of the concerns will lend more meaning to the review of the impact of adjustment programs later in this monograph. Types of Adjustment Laying the foundation for reform involves marketing the policy change among local interest groups, designing monitorable pro- grams, and paying attention to implementation. Effective adjustment may need to be triggered through rapidly disbursing assistance. But it can be consolidated and deepened with greater use of policy-based investment loans, disbursed over time in step with the gradual pro- cess of institutional learning typically needed for durable policy reform. These conclusions, emphasized by Robert Picciotto at the World Bank's conference on adjustment lending, grow out of a recog- nition of the various dimensions and meanings of the word "adjust- ment." Webster's Third New International Dictionary lists several definitions. One of them is straightforward: to adjust is to "bring to a more satisfactory state." According to the Bank's 1988 evaluation, countries receiving adjustment lending (AL countries) have, on aver- age, outperformed other countries in terms of economic growth and trade balances. Exports responded particularly well to reform.2 "Ad- justment" is also defined as "an orientation to the demands of a new order," "the achievement of a truer or more effective relative posi- tion," and "the rearrangement of components after assembly to im- prove performance." These are demanding definitions. They suggest that adjustment has two phases: (1) acknowledging a new reality or one that had been denied or suppressed, and (2) reshaping and redirecting policies and institutions. In other words, adjustment be- gins with abandoning certain notions of development and adopting others; but the commitment to reform must be fortified by complex follow-up measures to ensure sustainability. It has been relatively easy to evince statements of adherence to the basic tenets of new adjustment policies. This should not be surprising: politicians are prone to promulgate new policies that are not taken seriously, to establish reforms that may be aborted, and to create institutions that are stillborn. For this reason the customary emphasis on broad statements of development policy may be misguided. The most challenging aspect of adjustment is the pain of actual practice, since adjustment can mean abandoning popular programs and adopt- ing policies fraught with social tensions. The Bank's evaluation warned against the tendency to oversell the benefits of policy pack- ages and to underinvest in the reform of institutions. Adjustment is ISSUES IN ADJUSTMENT LENDING 3 seldom easy, and the resulting disappointments have contributed to adjustment fatigue. Adjustment lending can embolden a country and cause it to set out on a difficult and politically awkward path. Yet the very process of adjustment leads to pressures that cannot be relieved immediately without external resource flows. Accordingly, the authors of the eval- uation show tolerance for tailor-made solutions, partial Bank inter- vention, and flexible sequencing of sector adjustment loans. This approach hardly typifies the ideological rigidity often ascribed to World Bank practitioners or that conveyed by two other Webster's definitions: to adjust is to "put in order" and to "reduce to a system." Evidently, the Bank's evaluation of adjustment lending is opposed to the dogma and disciplinarianism implied by such definitions. A common dilemma is the difficulty of matching the pace of reform with the typical flow of disbursements under an adjustment loan. Changes in price or trade policy are usually well supported with quickly disbursing adjustment loans. To institutionalize reform, how- ever, may require several years. This suggests that policy-based in- vestment lending may be more effective than adjustment lending in supporting institutional changes, except where push-button policy changes are required. Another problem is that those who bear the costs of reform and those who benefit from the disbursements may not be one and the same: investment lending is better suited to calibrating resource allocation to various groups. Finally, there are global arguments for moving toward policy-based sector investment and hybrid lending, that is, lending that has elements of both adjust- ment lending and investment lending: to secure a stable constituency for development assistance, World Bank Group activities should sym- bolize the assumption of mutual benefit in the lending relationship.3 Where feasible and appropriate, lending vehicles should incorporate visible benefits of cooperation to recipients in developing countries and to economic agents in high-income countries. This helps sustain a Bank role in cofinancing and in the coordination of aid, and it strengthens loyalty among donors and recipients in the adjustment enterprise. Another instructive definition describes adjustment as "a payment under an insurance policy on settlement for a loss." To be sure, developing countries have no access to insurance against malpractice by bankers or development advisers. But the evaluation appropri- ately notes that it takes two to tango into heavy indebtedness. Exces- sively optimistic global projections, encouragement of recycling, and tolerance for slipshod policies during the romantic 1970s are part of 4 ADJUSTMENT LENDING the history of development. It behooves the Bank, then, as a coopera- tive institution, to help developing countries secure a greater margin of maneuver by playing a catalytic role. Since Bank funds are too scarce and too expensive for the enormous job at hand, they must be leveraged by burden sharing, which may be easier to secure with sector (or multisector) investment lending-or with new instruments aimed at debt reduction. Design of Adjustment One of the criticisms of adjustment programs is that they are some- times inappropriately designed (see box 1 for a summary of this and other concerns). A justification of adjustment lending (as opposed to project lending) is that it can support economywide reforms more comprehensively. Yet the design of adjustment policy in one area can conflict with goals in another area because compensatory measures have not been adequately worked out. The interactions among poli- cies and goals in adjustment lending are complex (see figure 1). For example, devaluing the exchange rate improves the external accounts but may increase the budget deficit if the government is a net pur- chaser of foreign exchange. Reducing tariff rates will increase eco- nomic efficiency but can reduce government revenue (as in Morocco or Thailand) or increase short-term unemployment. Although fiscal spending can soften the adverse effects of some of these tradeoffs, this method may be costly. Budgetary subsidies, for example, can buffer the effects of a devaluation, but a higher budget deficit can worsen the balance of payments, speed inflation, raise interest rates, and hurt the real and the financial sectors. Another concern about the design of adjustment loans is how reform affects the poor. Even if these effects are temporary, compen- satory measures to offset any negative effects on income distribution are not without merit. Promising approaches to help those directly affected by stabilization and structural adjustment (for example, civil servants and other employees laid off because of austerity measures or shifts in production) include temporary compensatory measures (such as public works programs to employ those who were laid off) and food-for-work programs to meet basic needs. Targeted food subsidy programs are also often recommended. For low-income and vulnerable groups likely to be hurt by cutbacks in social programs or by changes in relative prices, more effective and better-targeted public expenditures are usually important. Sometimes, reducing public funding for education and curative health care actually helps to ISSUES IN ADJUSTMENT LENDING 5 Box 1. Concerns of Some Observers about Adjustment Lending Design * Possible inadequacies in the technical design of adjustment programs that aim simultaneously to reduce excess domestic demand, shift resources to- ward the production of tradable goods, and reduce inefficiencies * Excessive austerity supported by adjustment lending (particularly its stabili- zation component), which frequently hurts the poorer members of society * Uniform and ideologically driven policy packages, rather than packages tailored to the particular circumstances of a country Implementation * Inadequate implementation and follow-up arrangements, which undermine the credibility and sustainability of reform * Overoptimistic expectations about the willingness or abilty of governments to persuade or compel interest groups to give up their rents in pursuit of national efficiency External Conditions * Overoptimistic assessment of global economic projections, such as prices, interest rates, and economic activity in industrial countries * Underfunding because of inadequate burden-sharing arrangements with commercial banks and official assistance agencies * Rapidly disbursed World Bank lending used to repay and bail out commercial banks Quality of Lending and the Portfolio * Open-ended adjustment lending in general, and sector adjustment lending in particular, which transfers resources to countries that are not in fact adjusting * Weak conditionality, so that the lending may not improve the recipients' economic performance and creditworthiness * Increased risk to the Bank's portfolio caused by adjustment lending Diplomacy and Coordination * Political and diplomatic difficulties that impede the Bank-government dia- logue or acceptance of the conditions imposed on the management of the economy, or both * Problems in coordinating interaction among the International Monetary Fund, the World Bank, bilateral lenders, the Paris Club, and organized creditors correct mistargeting of existing public expenditures on social pro- grams. Implementation of Programs The Bank's evaluation of adjustment lending noted that adjustment programs are sometimes inadequately implemented. Implementing price changes, for example, can be accomplished more quickly than 6 ADJUSTMENT LENDING Figure 1. How Adjustment Policies Interact Policy Reduce Promote growth Polcy Reduce and employment, Reduce goal debt alleviate poverty inflation Macroeconomic External Fiscal Finaiicial level level _ Balance of Budget deficit payments Expenditure Revenue Monev gror.th chaags T bae Interest payment, Tab policy, lnterest rate, austai suchimespl publc investment enterpriase ticciot re fore ign m pric ong inPestment, v a at capital flows / \ Sectoral Agriculture n< Inustry /\ \ Other sectors level t E Fa rm pr icing, Prcing,Humnrsuce invSestment, privZate inf rastructure export investment, institutionalizing such changes or achieving nonprice reforms. In general, follow-up and supervision of changes that take time have received less attention than has the management of push-button changes. To be effective, reforms must be followed through and sustained. Such implementation, according to Robert Picciotto, re- quires the comnmitment of the policymakers to the reform process. Policy letters are often vague and are usually negotiated with a single ISSUES IN ADTUSTMENT LENDING 7 individual or a group of individuals at or near the top of the hierarchy, and thus carry little weight. Yet adjustment involves a wide range of policymakers and decisionmakers. Their limited understanding of and commitment to the adjustment program can be a serious obstacle to effective implementation, hence the need to involve officials at all levels in the design and implementation of a reform program. Many instruments, including training, policy seminars, and technical assis- tance, exist to help carry out such a process. The appropriate sequenc- ing of these instruments is a neglected area of adjustment lending. Implementation, as noted earlier, takes time and effort. The evalu- ation noted that problems arise in matching the pace of reform with the flow of disbursements. Typically, those who bear the costs of reform may be more concentrated and vocal than those who benefit from it (for example, when an agriculture sector loan requires the withdrawal of subsidies). This can complicate the task of securing a domestic constituency for reform. Because political feasibility re- quires involving local beneficiaries, the evaluation suggested using counterpart funding for local expenditures. Also helpful would be giving greater consideration to multiyear lending that disburses slowly against a positive list of eligible imports, to hybrid adjustment lending, and to policy-based sector investment loans in general. External Environment The Bank's evaluation examined overoptimistic assumptions about the role of the external environment in adjustment programs. External funding to support reasonable and sustainable adjustment has been insufficient, and commercial banks have not shared enough of the burden. At the conference, William Cline's view was that this situation constitutes a strong case for increasing Bank lending for adjustment. Project lending takes longer to disburse. And foreign exchange carries a very high scarcity price in the 1980s as external funds dry up. The adequacy of financing is difficult to appraise, depending as it does on a society's ability to sustain measures that may have high short-term costs. Adequate financing can give a country the time and resources needed to make orderly adjustments, so long as it does not lead to the indefinite postponement of necessary reforms. Inadequate financing, by contrast, may mean that demanding adjustments are made too abruptly to be politically and economically sustainable, and it may lead to a program's collapse. External capital flows to Africa and Latin America, for example, declined drastically in the 1 980s. But some countries willing to implement comprehensive adjustment pro- grams (Colombia, Indonesia, the Republic of Korea, and Turkey, for 8 ADJUSTMENT LENDING example) have attracted commercial financing; this suggests that the availability of external financing is itself affected by a country's will- ingness to undertake an adjustment program. Quality of Bank Lending Adjustment lending is ineffective if it does not lead to sustainable improvements. In general, the benefits and repayment sources of project lending are easier to pinpoint than are those of adjustment lending (see box 2). The uncertainty about lasting improvement, mentioned earlier, heightens this risk. One fear is that adjustment lending may serve to postpone adjustments rather than promote them. It is, however, not necessarily easier in project lending than in adjustment lending to assess the true additionality of the financing or to monitor and assess the benefits, according to Marcelo Selowsky. In a world of fungibility, one is not financing the negotiated project but the marginal project in the investment list, the one that can now be financed because of the Bank's presence. Fungibility exists if the negotiated project would have been financed even in the absence of the Bank. For that reason the Bank's reviews of public sector expen- diture now assess the quality of the overall program, including expen- diture on marginal projects. But this still assumes that the loan does not affect the overall investment budget of the government, that is, that no part of the financing has brought about higher consumption. This determination is very complex. Thus the benefits of project lending are often just as difficult to evaluate as are those of adjustment lending. There are three prerequisites for effective adjustment lending, which were emphasized by John A. Holsen. The first requirement is that the country have an overall adjustment program that includes stabilization and longer-term aspects. The second is that the govern- ment own the program. The third is that there be adequate financing to support a sustainable program that takes into account political and social conditions. These prerequisites have sometimes been ignored in sector adjustment loans. Coordination with Other Lenders Even if the principles just discussed are broadly accepted, their appli- cation is difficult. How are agreements on macroeconomic programs to be recorded, monitored, and assessed? How will they affect collab- oration between the Bank and the NF? The nature of adjustment lending calls for coordination among the country, the Bank, the IMF, ISSUES IN ADJUSTMENT LENDING 9 Box 2. Project Lending as Opposed to Adjustment Lending Arnold Harberger highlighted the differences between traditional project loans and adjustment loans. They have different effects on the ability to generate cash flow for repaying the loan. In general, if the beneficiary group is identifiable, it is better if the operation (be it a project loan, a structural adjustment loan, or a sector adjustment loan) is financed by that group rather than by general taxes. If the beneficiary group is not identifiable in any straightforward way-that is, if it is a broad segment of the population, such as purchasers of imported goods- then one looks to some general way of meeting the cash flow needs of the operation. Policy changes made as a result of structural adjustment loans or sector adjustment loans may yield large increases in national income and output; any such increases should produce automatic increases in taxes of a magnitude sufficient to cover the debt service needs of the program. This is the ideal situation. In this case, since those with increased income are automatically beneficiaries of the program, the beneficiaries in a sense provide the financing. When there are no direct flows to the public treasury or such flows fall seriously short of what is heeded to service a loan, then additional sources of funds should be sought. Under a sector adjustment loan, the first choice would be to raise revenue from the affected sector; only if this were not possible would the population in general be tapped as a source of revenue. Under a structural adjustment loan, or in any other case in which general revenue must be relied on, the first choice would be to eliminate defects in the revenue system (by getting rid of a significant tax loophole, for example) to add directly to revenue. In such a case, it would not matter whether the revenue was raised from the population as a whole or from a limited group. If there are no obvious defects then general revenue would have to be raised as efficiently as possible. Thinking about the source of the cash flow gives an added degree of discipline to the process of administering an adjustment program. In particular, it helps combat the debt rollover syndrome. In the best case, a country with constant or growing debt services and repays old debt with revenue from the true produc- tivity of the old projects or programs; newly incurred debt is invested in new operations, which are expected to be highly productive themselves. Rolling over or rescheduling payments on old debt normally signals that the old loans did not prove to be as productive as hoped (it may also reflect new adverse shocks from other sources). It should be realized that even in the best of circumstances, rollovers of old loans impinge on the possibility of using the same supply of funds for new and productive operations that would contribute to the growth of the country's economy. and other creditors. Considerable understanding and diplomacy are needed to ensure a nonconfrontational and effective process. The Bank's evaluation of adjustment lending recommended a com- plementary role for the World Bank and the IMF in macroeconomic policy. Views of commentators vary, with some recommending a sharper division of labor-that is, the IMF would handle macro issues and the Bank would focus on investment issues-and others calling for a greater overlap. Questions have been raised about who should 10 ADJUSTMENT LENDING do what, how differences can be resolved, how the policy dialogue should proceed, how policy understandings should be conveyed, and so on. (Some of these operational questions will be discussed in the final section of this monograph.) Impact of Adjustment Programs Compared with measuring the impact of traditional projects, measur- ing that of adjustment programs is more imprecise. The indicators to be used in evaluating the program depend on its multiple objectives. The final goal of adjustment is to improve growth and income distri- bution, with minimal effect on debt and inflation. At the same time, intermediate indicators such as the size of the fiscal deficit, the real exchange rate, and the rate of investment need to be watched closely. Another source of uncertainty in evaluating performance under an adjustment program is that the comparator is how an economy would perform in the absence of an adjustment program. Lacking a precise model, evaluators often compare countries with and without adjust- ment programs or the same country before and after adjustment. Performance under Adjustment According to Jaime de Melo, the Bank's 1988 evaluation based its assessment of the impact of adjustment lending on nine indicators: growth in gross domestic product (GDP); growth in exports; the real exchange rate; inflation; and the ratios of investment to GDP, current account balance to GDP, budget balance to GDP, external debt to exports, and debt service to exports. When the change in these indicators in the adjustment lending and nonadjustment lending (NAL) countries was compared, the conclusion was that thirty AL countries that re- ceived loans before 1985 on average have performed modestly better than the NAL countries examined. This modest improvement occurred despite a relatively less favorable environment. Moreover, the perfor- mance of twelve AL-intensive countries (recipients of three or more adjustment loans) was considerably better than that of the AL coun- tries as a whole despite the much larger external shocks faced by the AL-intensive countries (see table 1). Several considerations nonetheless raise concerns for the sus- tainability of adjustment programs. Investment ratios have declined and budget deficits are higher in highly indebted countries. Although current account deficits are shrinking, debt-service ratios and debt- export ratios have increased. It is not possible to conclude, therefore, IMPACT OF ADJUSTMENT PROGRAMS 11 Table 1. External Shocks and Performance under Adjustment Low-income Middle-income All cou1ntries countries countries Country group AL NAL AL NAL AL NAL 30 AL countries External shocks as percentage Of GDPa 4.6 3.8 5.2 1.4 5.0 2.4 Percentage of performance measures showing improvementb 53 47 55 45 54 46 12 AL-intensive countriesC External shocks as percentage of GDPa 10.3 3.8 7.0 1.4 8.4 2.4 Percentage of performance measures showing improvementb 62 38 63 37 63 37 a. Sum of terms of trade and interest rate shocks for 1981-86. b. The AL and NAL columns add to 100. c. AL-intensive countries have received three or more adjustment loans. Notes: AL = adjustment lending. NAL = nonadjustment lending. Source: World Bank data. that the AL countries have been growing out of indebtedness. Al- though the principal may be refinanced, the burden of interest seri- ously limits the capacity to import and the use of domestic savings to finance investments. Another concern is the stagnation in nutrition levels since 1980, especially among low-income groups and in Sub- Saharan countries, where per capita caloric intake remains below "minimum" requirements. Objections may be raised about the methodology used for the comparison of AL and NAL countries. Determining the success or failure of structural adjustment loans and sector adjustment loans simply by comparing economic performance of countries with and without adjustment loans can be misleading. Mohsin Khan explained that the conclusions in the Bank's evaluation should be more tenta- tive. The problem is that the conditions in AL countries are likely to differ substantially from those in NAL countries before the period of comparison, which affects the evaluation of performance. In short, the 12 ADJUSTMENT LENDING AL countries are not randomly selected. This means that the estimates for the control group are biased because differences in outcome are attributed exclusively to whether a country has adjustment loans. In fact the different starting positions of AL and NAL countries are in themselves a cause of differences in subsequent performance. One solution to this problem (the one adopted in the evaluation) is to compare changes in performance rather than absolute differences among countries. Another, more complex approach is to identify, first, the relevant reduced-form relationships that link policy instruments to policy targets and, second, the policy reaction functions, which show how the use of policy instruments changes when the state of the economy changes. These empirical approaches, however, are not easily implemented and may be subject to error. In Mohsin Khan's view, however, the error is likely to be smaller than in the approach used in the Bank's evaluation. Another approach would be to classify countries by performance and then compare AL and NAL countries within subgroups. This alter- native, however, may be impractical because there would be too few comparators in some subgroups to yield significant statistics. It is possible, though, according to Arnold Harberger, that the better average performance of the AL group in the evaluation reflects the good performance of a few countries. One has to distinguish between the impact of an adjustment pro- gram undertaken by a country and the impact of Bank operations associated with that program. To evaluate the contribution of Bank operations, it is important to identify and assess the mechanisms by which these operations have contributed to the success of adjustment programs, either by increasing the probability that they will be under- taken and sustained or by increasing the benefits that derive from them. An adjustment program usually involves up-front costs to many groups in society; the benefits usually take time to emerge. A Bank operation may distribute these benefits better and redistribute costs not only over time but also across socioeconomic groups. For example, the operation may allow the financing of targeted food programs while food prices are decontrolled or devaluations are taking place, thus making the adjustment program more acceptable. Bank operations may help to bring about more efficient fiscal adjust- ments that do not introduce new distortions, along with institutional changes that make the adjustment sustainable. IMPACT OF ADJUSTMENT PROCRAMS 13 Design and Implementation of Adjustment Loans William McCleary pointed out that the Bank's 1988 evaluation of adjustment lending contained several findings on the policy under- standings that prevail under adjustment lending. Almost all adjust- ment programs have stabilization elements designed to bring domestic demand more in line with available resources. Differences in the structural adjustment elements of programs-that is, in the policies to alter incentives, improve efficiency, or strengthen institu- tions-are more pronounced, largely because countries differ most in these areas. Conditionality is a normal part of adjustment lending; compliance in various policy areas has varied significantly because of differences in technical complexity, the sensitivity to policies, and institutional capabilities. The evaluation's analysis of more than fifty adjustment loans in fifteen countries included these findings: (1) about 60 percent of the conditions were fully met under structural adjustment loans and sector adjustment loans, a proportion that increases modestly over time after loan disbursement is completed (table 2); (2) implementa- tion varied significantly across policy areas; (3) for four or five critical conditions in each loan (so called because of their expected substantial impact on structural adjustment), implementation was better than for other conditions; and (4) problems in implementation arose mainly with respect to tranching, the commitment of government, the quality of Bank advice, and the external economic environment. In addition, implementation performance of sector adjustment loans is generally weaker than that of structural adjustment loans. It is only partly speculation to say that this is because (1) sector adjust- ment loans, compared with structural adjustment loans, often have had less well-defined medium-term programs, so that intentions for policy changes and the government's commitment to the program have been unclear; (2) a smaller percentage of sector adjustment loans have been tranched, and nontranche conditions have been looser; and (3) the Bank seems to have taken the macroeconomics under sector adjustment loans less seriously than it has that under structural adjustment loans, in some cases approving projects without requir- ing an adequate macroeconomic policy framework and often pro- viding less supervision. There is concern about the implications of the evaluation's findings that compliance with conditionality had been satisfactory, whereas the performance evaluation showed only modest improvement for AL countries over the NAL group. Three possible explanations are that the 14 ADJUSTMENT LENDING Table 2. Implementation of Conditionality (percentage of conditions) During the loan period Present situation Conditions Conditions fully (1) plus fully (2) plus implemented substantial implemented substantial Policy (1) progress (2) progress Exchange rate 70.0 90.0 62.5 87.5 Trade policy 54.9 84.2 63.4 89.3 Fiscal policy 53.2 78.3 69.8 95.3 Budget and public expenditure 68.0 78.0 71.7 84.8 Public enterprise reform 61.3 86.7 70.0 90.0 Financial sector policy 71.4 85.7 73.5 89.8 Industrial policy 53.3 93.3 42.9 85.7 Energy policy 79.2 83.3 83.3 88.9 Agricultural policy 57.1 81.6 58.1 83.7 Average implementation All conditions 60.3 83.7 67.5 89.0 Structural adjustment loans 68.3 84.1 73.5 92.4 Sector adjustment loans 50.9 83.2 60.0 84.9 Sub-Saharan Africa 52.4 84.6 62.2 86.7 Highly indebted countries 66.9 88.6 73.2 91.4 Other countries 52.8 79.7 56.0 84.0 Source: World Bank data. AL countries implemented the policy package but were subject to larger external shocks, which hampered performance; the supply response to the policy changes was overestimated, particularly in the short run; and the policy package was inadequate. The implementa- tion record indicates compliance but measures less satisfactorily how well the conditions were implemented. Conclusion Vinod Dubey and Stanley Fischer pointed out that, according to available evidence, adjustment lending has been moderately success- ful. Because definitive conclusions are not possible, however, the claim is a matter of judgment. Perhaps the most convincing evidence is that the AL-intensive countries have performed significantly better THE EXTENT OF POLICY REFORMS :15 than average, despite their larger external shocks. (Even here, though, an alternative explanation is that the Bank has been willing to main- tain its lending only to the countries that were doing well.) The evaluation concluded that the Bank's adjustment lending should continue, albeit with improvements. But even had it con- cluded that adjustment lending had not succeeded, it would probably have recommended continuing the program, although with signifi- cant changes. Given the economic difficulties, including the debt crisis, facing developing countries, the need for adjustment continues, as does the need to moderate its impact on the populations of devel- oping countries. Lending in support of needed changes in developing countries contributes to development and thus it should have the effect of improving the quality of the Bank's portfolio. The Extent of Policy Reforms Countries suffering similar external shocks may perform quite differently under adjustment programs. Such differences highlight how the policies and institutional structure of an economy influence performance. These differences among countries in the application of policy also provide qualitative support to the aggregative analysis of performance considered earlier. Progress in reforms has varied across policy areas (see table 3). The trade balance of most adjusting coun- tries has improved substantially. In some cases, this improvement was attributable mostly to the lack of import financing; in others, it re- flected an orderly adjustment of internal demand and the transfer of resources to the export sector. The exchange rate in several countries has depreciated significantly in real terms. And the response in vol- ume of exports has been strong; exports of manufactured goods grew 11 percent a year on average during 1982-87 in a sample of sixteen AL countries for which detailed studies were prepared. In general, exter- nal sector policies and price policies have been reformed. Progress has been less visible in institutional and nonprice areas and in aspects of price policies that require institutional changes. According to the evaluation, five elements contributed to sustain- ing reforms. First, some of the better performers (for example, Colom- bia, Korea, Thailand, and Turkey) followed relatively sound policies over the long term and adjusted quickly to shocks. Second, programs have been easier to support when the negative effects on growth, employment, and the poor were short-lived. Third, the speed of the supply response may determine sustainability. A strong export 16 ADJUSTMENT LENDING Table 3. Assessment of Reforms Category Extent of reform Assessment Overall reforms External adjustment Policies have mattered a rapid, internal adjust- great deal, but con- ment often unsus- straints include con- tained, institutional flicts in design (as reforms slow. between fiscal and trade policy). Reforms to specific policies Trade policy Progress in exchange Strong response in ex- rate flexibility and ex- port volume, but non- port incentives. price factors need greater emphasis, as in East Asia. Quantitative restrictions Greater liberalization replaced by tariffs constrained by stabili- and tariff reform, but zation, internal oppo- reduction in protec- sition. Export tion has been slower. response assists in im- port liberalization (as in Korea, Turkey). Fiscal policy Initial deficit reduction Expenditure cuts could not sustained. not be maintained. Some success in ratio- More careful scrutiny of nalization of public current expenditure investment (as in needed. Pakistan, Turkey). Tax reform has had lim- Short-run revenue con- ited success and not cerns have domi- been comprehensive; nated. Indonesia, Mexico, and Turkey are excep- tions. Public sector Institutional reforms Introduced in almost management and divestiture slow. every adjustment pro- gram. Some reduction in Reduction effected losses by public enter- more by price in- prises. creases than by im- provements in efficiency. (continued on next page) THE EXTENT OF POLICY REFORMS 27 Table 3 (continuted) Financial sedor policy Reforms initiated in Reforms often pro- only a few cases, but tracted and require receiving greater at- sound macro frame- tention. work, as shown by ex- amples of Chile and Turkey Antipoverty policy Receiving increasing at- More work needed on tention in design of how adjustment af- both policies and sup- fects poor. plemental programs. Agricultural policy Improvements in price Especially in Sub- policy visible. Saharan Africa. Reform of parastatals Supply response to slower. price constrained by institutional factors. Industrial policy Focus on trade policy Reforms specific to in- and its impact on dustrial sector have industry, not received suffi- cient attention. Source: World Bank data. response helped the continuation of reforms in Turkey; in contrast, policy reversals in Zambia resulted partly from lags in export growth. Fourth, the financing and sustainability of adjustment programs are mutually reinforcing. The adequacy of adjustment programs has helped to mobilize external resources in Chile, Colombia, and Turkey, though neither Chile nor Colombia has ready access to outside fi- nance. Fifth, a key determinant of sustainability is the program's appropriateness for the particular country. The government's com- mitment to the program and support for its implementation are crucial. Trade Policy There has been considerable progress in achieving reforms of ex- change rate policy and in improving incentives for exports, according to Vinod Thomas. The real exchange rate in 1986-87 had on average depreciated by about 35 percent compared with the average 1980-81 level in the fifteen AL countries reviewed in detail. Volumes of exports, especially manufactured exports, grew substantially, though values increased much less because export prices were declining. Chile, Korea, and Turkey have achieved dramatic increases in exports, but most other countries must do more to boost exports with price and 18 ADJUSTMENT LENDING nonprice instruments. Several countries have replaced quantitative restrictions with tariffs and reformed tariff structures. Lowering import protection has sparked greater resistance than has a reduction of the direct bias against exports. The twelve principal borrowers for adjustment typically discriminated less against exports than did other countries from the same regions. Liberalization of import regulations is easier if exports are doing well. It would benefit each country to establish and announce a strategy that includes replacing quantitative restrictions on imports with tariff protection, reducing tariff dispersion, and gradually lowering protection on im- ports. Fiscal Policy Ajay Chhibber pointed out that progress in fiscal reform has been slower than in external accounts. Initial reductions in fiscal deficits are brought about primarily by cuts in expenditure (particularly investment expenditure). In many countries, these cuts could not be maintained. In others, external shocks had been underestimated, and interest payments rose sharply As a result, a widening of the differ- ence between public sector and current account deficits put increasing pressure on the private sector to generate a net surplus, leading to cuts in private investment and growth, often accompanied by higher inflation (figures 2 and 3). Some success has been achieved in rationalizing public investment with the help of the Bank's investment reviews. But there has been much less progress in reforming taxes and reallocating current expen- ditures. Given the urgency of reducing fiscal deficits, tax policy initia- tives have been dominated by near-term revenue measures in recent years. Exceptions are Indonesia, Mexico, and Turkey, which have undertaken thorough tax reforms. Reforms to relieve the financial pressure on the private sector and to generate increased domestic savings and investments through growth-oriented fiscal policies have generally been insufficient. Public Sector Management Institutional reforms of the public sector have been promoted by almost every adjustment program. Creation of new enterprises and growth in employment in the public sector have slowed. The losses of public enterprises have probably been reduced, but the improve- ment owes more to price increases by monopolies than to increased efficiency. Improved use of resources in utility companies has proved to be an elusive goal; the financial losses of utilities remain high. Few THE EXTENT OF POLICY REFORMS 19 Figure 2. Current Account and Public Sector Deficits as a Percentage of GDP, 1980-87 8 >\iX Public sector 8 @ \ deficit 7 6 / 5 4 Current account deficit 3 1 e 1980 82 83 84 85 86 87 Note: Unweighted average for thirteen AL-intensive countries (Chile, Colombia, C6te d'lvoire, Jamaica, Korea, Malawi, Mexico, Morocco, Pakistan, Philippines, Thailand, Turkey, and Zambia). Source. World Bank data. of the reforms attempted have had much impact on planning, policy analysis, or debt management in the public sector. After initial delays, several countries, including Jamaica, Niger, Panama, and Togo, have made progress in divesting public enterprises; others, including Ghana, Senegal, and Turkey, have not. It is too early to evaluate how these enterprises have performed since they were privatized. Progress has been made, but public sector reform has a long way to go. Financial Sector Policy In the few cases in which financial reforms have been attempted-as in Argentina, Chile, and Turkey-an appropriate macroeconomic policy framework was critical for success. Financial reforms tend to 20 ADJUSTMENT LENDING Figure 3. Public Sector Deficit as a Percentage of GDP in Selected Groups of Countries, 1980-87 11 10\ 15 N AL countries 9~ -@ - - % 25 AL countries 1_ 7 LO, . _ 6 444,~ 5 5 13 AL-i-ntensive countis~ mJ 4 1 f I 1980 82 83 84 85 86 87 Notes: Data are unweighted averages. AL = adjustment lending. NAL = nonadjustment lending. Source: World Bank data. be protracted; the organized financial system in many developing countries is technically insolvent because of poor management, insuf- ficient supervision of bank lending, and incorrect lending signals during financial repression. Much remains to be done in improving regulation and in restructuring. Because, as Alan Gelb noted, it is too early to judge how successful adjustment loans have been in reforming the financial sector, such loans are treated only briefly in the Bank's 1988 evaluation. Specific policies have been quite well implemented; the real problem is that this sector is very sensitive to the macroeconomic situation. Financial sector operations can take the form of rapidly disbursing loans, usu- ally provided to countries with serious balance of payments problems and fiscal imbalances. Where the macroeconomic situation was under control or improving when the loan was made, as in Chile, the record THE EXTENT OF POLICY REFORMS 2 2 shows that these programs have succeeded. But this degree of control over macroeconomic stability is the exception rather than the rule in most countries. Perhaps in such situations it would be better to concentrate on adjustment lending in other sectors, less vulnerable to macroeconomic repercussions. Antipoverty Policy The effects of adjustment on poverty require special attention to the design of policy and to devising supplemental programs targeted to the poor. These effects are not merely a short-term concern: adjust- ment programs have taken longer and proved much more arduous than originally anticipated. Labor-intensive activities and public works can help those directly affected by adjustment. Two examples are Bolivia's experiment with the Emergency Social Fund and Ghana's Program of Actions to Miti- gate the Social Cost of Adjustment. For low-income and vulnerable groups likely to be hurt by cutbacks in social programs or by changes in relative prices, more effective and better-targeted public expendi- ture is usually needed. Correcting mistargeted public expenditure on social programs sometimes helps, for example by redirecting public funding for education and health care, as in Brazil and Cote d'Ivoire. Social service bureaucracies can be made more efficient to achieve savings, and targeted programs can sometimes replace general subsi- dies, as in Mexico and Morocco. Shifts in public spending may en- counter political opposition, as with the reduction of food subsidies in the Dominican Republic, Egypt, and Zambia. Chile and Pakistan have had more success in this area. Industrial Policy Issues raised by adjustment in the industrial sector vary widely, depending to a large extent on the level of industrialization. In coun- tries such as Chile, India, Mexico, and Turkey, an existing, complex industrial system must be restructured to make it more competitive internationally. In Africa, adjustment must take into account the unavailability of skilled labor, inputs, and technology. Except in Asia, adjustment operations for industry have put more emphasis on trade and the exchange rate than on industrial policies. Policies that bear specifically on industry have dealt with restructuring, pricing, invest- ment incentives, and redirecting public investment from the manu- facturing sector to encourage private investment. A common source of problems has been the desire to quicken the pace of industrialization to allow a developing country to catch up 22 ADJUSTMENT LENDING with the rest of the world. The objective is laudable but its implemen- tation (mostly during the 1960s and 1970s) often required large invest- ments in heavy industries and in the manufacture of chemicals, which led to excess capacity in some countries and to public enterprises dependent on heavy protection and subsidized credit. For some coun- tries, the strategy sometimes worked because the right industries were selected; for others, the industries were viable only with contin- ued subsidies or protection or by other means of support. In those cases, adjustment in industry meant reducing losses, protection, and regulation to develop a more efficient and sustainable industrial structure. Agricultural Policy The Bank's evaluation looked at three agricultural issues: price policy and incentive schemes, institutional reforms, and investment and expenditure policy. The review of price policy found that the Bank has concentrated on increasing real prices to farmers, particularly in Sub-Saharan Africa. The evidence shows that, especially since 1984, real farm prices have risen in Sub-Saharan Africa. Progress in institu- tional reform has been much slower, although there are cases of institutional changes. But in the majority of cases, reforms are not institutionalized and changes are not automatic. Price changes, for example, have had to be renegotiated year after year. On investment and expenditure policy, the evaluation stressed the link between basic supply bottlenecks, especially in the public sector, and the lower than expected supply response in the agricultural sector. Farmers are unable to respond to higher farm prices if they face constraints in the areas of transport and irrigation, among others. Some of these issues are addressed in the Bank's review of expendi- ture policy but much more work needs to be done in this area. Fiscal Effects of Conflicting Policies The fiscal tradeoffs involved in carrying out reforms cut across nu- merous areas. Raising farm prices, for example, has implications for expenditure; similarly, when tariffs are lowered, it may be hard to replace the revenue forgone. Attention to resource mobilization, how- ever, is not sufficient. As public resources increase, the need to cut public investment becomes less pressing. The question of how much of the fiscal deficit under adjustment comes from excessive expendi- ture and how much of it is necessary to the adjustment program is central to the choice between stabilization and structural adjustment. APPLYING THE PREREQUISITES FOR ADJIJSTMENT LENDING 23 In the view of Armeane Choksi and Marcelo Selowsky, it is impor- tant to achieve adjustment in the fiscal accounts in order to benefit fully from sectoral reforms. This is particularly true when fiscal defi- cits are large and involve significant borrowing from the central bank and from domestic capital markets. Large and volatile fiscal deficits involve inflation, high real interest rates, and general uncertainty. The deficit has to be addressed first: sectoral reforms designed to correct relative prices will have important benefits only if the climate of uncertainty and inflation is reduced. Reforms in the trade and finan- cial sectors have little payoff if very high inflation causes prices to convey inadequate information about incentives in the economy and if domestic crowding out and overall uncertainty stunt private invest- ment. Countries that have solved the fiscal imbalance early-such as Chile and Uruguay-have benefited more from sectoral reforms and from improvements in world trade opportunities. This experience has important implications for Bank operations and for reevaluating past strategies. The Bank's evaluation shows that loan conditions with regard to trade policy in highly indebted coun- tries seem to have been more important than conditions with regard to budget, public expenditure, and public enterprises. In many of these countries, however, public enterprises have been the main source of public sector expansion and deficit. Moreover, many of the conditions placed on these enterprises have dealt more with internal efficiency (internal management, auditing, and accounting) than with a deeper reevaluation of the role of such enterprises and their relation with central governments. There is little point in improving internal efficiency if managers know the central government will finance any ex post deficit. Applying the Prerequisites for Adjustment Lending As mentioned before, the Bank's 1988 evaluation cited three prereq- uisites for adjustment lending: the country must have an overall adjustment program, including stabilization and structural adjust- ment; the government must own the program; and the program must be politically and financially realistic. The prerequisites are demand- ing, as elaborated by Stanley Fischer and John A. Holsen. There is considerable agreement on the need for an adequate overall adjust- ment program. In addition, most successful adjustment has occurred in countries where adjustment programs were genuinely understood and supported by the government, not supported solely by the Bank 24 ADJUSTMENT LENDING or other external agencies. The third prerequisite reflects a judgment that adjustment programs in some countries have failed because the country lacked sufficient funding to finance investment. The evalua- tion does not specify how exactly to determine a country's commit- ment to a program, leaving this judgment to management. Nor does it specify how to tell precisely whether a financing plan is adequate. The Bank's practice has been to set a target for growth and then to assess the adequacy of external finance by its consistency with the target. Listing prerequisites is one thing; enforcing is another. For the first prerequisite, the Bank's evaluation suggests that for low-income countries receiving support from the iMFs Special Adjustment Facility, Policy Framework Papers (PFPs) could be the vehicle for reaching agreement on an adequate macroeconomic and structural policy framework. To this end, it recommends trying to make the PFP a more genuinely tripartite document, with responsibility and commitment shared among the World Bank, the IMF, and the host government. When there is no PFP, the evaluation suggests other ways of reaching agreement on overall policies. Of course, only determined efforts by all parties, evaluated by the Bank's management and executive direc- tors, will ensure that the policies described in such documents are realistic and are carried out. Although agreeing to the three prerequisites does not guarantee success, it strengthens the Bank's negotiating position and its mecha- nisms for internal control. The Bank's management and executive directors will have to decide whether the prerequisites are satisfied. Whether the Bank will decline to lend when these prerequisites are not met remains to be assessed. A complicated situation arises when a country does not appear to have an adequate external financing plan: the temptation then is to lend to improve the country's situation. But this increases the risks to the Bank's portfolio. The evaluation recommends increased efforts to find other lenders to share the risk, and suggests that the Bank lend into arrears (to other lenders) if necessary and if, in its judgment, a country is making a serious effort at adjustment. The conclusions of the Bank's evaluation are logical, but its recom- mendations should not be interpreted too rigidly, especially in the early stages of an adjustment program, according to Wilfried Thalwitz. The Bank often initiates adjustment operations amid con- siderable uncertainty about the outcome of the program. Very often, the Bank must begin operations in a sector in which the government is willing to make changes, but when it has not yet established a full APPLYING THE PREREQUISITES FOR ADJUSTMENT LENDING 25 and comprehensive macroeconomic framework or agreement. For example, Ghana's adjustment program might not have come about had strict, economywide conditions been employed at the outset. The dilemma, then, is whether to go ahead in the expectation that positive changes in one sector will lead to changes in other sectors or to wait for a comprehensive economywide agreement. Decisions must often be made under conditions of great uncertainty; if the Bank were to follow the evaluation's recommendations too rigidly, it might deprive certain countries of an important opportunity. Overall Adjustment Program In deciding which actions to take in an adjustment program, the Bank must take three basic steps, emphasized by Russell Cheetham. The first is to develop more stringent guidelines for operations managers concerning the macroeconomic framework and related sectoral com- ponents. The second is to strike a balance between the LvFs require- ments for stabilization and the goal of long-term adjustment. The new organization of the Bank's country departments makes it easier to develop and integrate macroeconomic and sectoral elements of an adjustment program. The main vehicle for articulating the program should be a Bank Country Strategy Paper (csP). The third step is to reorganize the preparatory work done by country departments on adjustment assistance to reduce or eliminate fragmentation in eco- nomic and sectoral work, preparation for lending, and reports on project completion. These activities should be integrated and staff allocated with appropriate emphasis on preparing and implementing the program and evaluating results. At present, relatively too much staff time is allocated to preparation and not enough to implementa- tion and evaluation. Ownership of the Program The political realism of the requirement that a government own the adjustment program can be questioned. Opposition to reform pro- grams usually exists both inside and outside the government. The question is how to strengthen the hand of reformers. The government's public commitment to the main elements of adjustment programs may not always improve the likelihood of success. One must usually judge the probability of success in the face of various obstacles rather than attempt to elicit complete commitment at the outset from all parts of the government. Of course, commitment must come from relevant parts of the government. There is no point in getting agreement on, for example, trade-related aspects of industrial 26 ADJUSTMENT LENDING and agricultural policy unless the minister of trade is in agreement and willing to implement the proposals. Very often, the ownership question requires a case-by-case judgment. Realism of the Program There are two principal factors that affect the realism of an adjustment program. The first is the magnitude and distribution of the costs of adjustment, which are typically steepest in the beginning. The second is the magnitude and distribution of the benefits of adjustment, which typically accrue later. The criterion of realism should be applied more systematically in preparing and evaluating adjustment operations. In particular, lenders must have sufficient knowledge of the incidence and magnitude of adjustment costs and benefits. Although the costs are usually better grasped, the appropriate speed of adjustment in incentives and the resulting resource flows and the size of supply responses are poorly understood. It is often difficult to judge whether the government is committed to the program, in Steve O'Brien's view. Building an analytical base for adjustment lending requires time; this is a necessary precondition for persuading decisionmakers of the validity of the Bank's recommendations. Establishing an analytical base is crucial. Yet pol- icy planners are under pressure to move quickly when the economy is in trouble or the government needs resources. In addition, the Bank's evaluation alludes to internal Bank pressures to lend. A case- by-case approach is necessary, but good decisions require time, a solid analytical foundation for policy prescriptions, and the requisite gov- ernment commitment, since adjustment is a course of action that takes years to complete. The same dilemma appears in ensuring adequate financing for adjustment programs. Planners tend to underestimate the resource requirements of adjusting countries because they underestimate im- port needs and overestimate export prospects and the potential for improving efficiency and reducing the incremental ratio of capital to output. Even when a program lacks country-specific commitments from donors, Bank staff often wish to implement the program in the expectation that additional funding will materialize. In a similar situation, the staff of the IMF would not usually pursue the program. This difference can create problems of coordination, as on documen- tation of joint programs. The best route for the country-the Bank's catalyst approach, which carries the possibility of inadequate fund- ing, or the IMF's more conservative approach-is not clear. SELECTED OPERATIONAL IMPLICATIONS 27 Selected Operational Implications The Bank's 1988 evaluation of adjustment lending has implications for the operation of adjustment programs, including the areas of addressing macroeconomic imbalances, establishing conditions for tranche release, and monitoring and evaluating programs. Sector Adjustment and the Macroeconomic Program The evaluation discussed whether all rapidly disbursing loans, in- cluding sector adjustment loans, should be required to contribute directly to reducing macroeconomic imbalances. The existence of such imbalances is a necessary condition for rapidly disbursing loans. If loans that do not contribute to reducing macroeconomic imbalances can be made, as for alleviation of poverty, the Bank in effect makes it possible for countries with balance of payments problems to receive loans that are not available to other countries. Nothing is easier than to generate a balance of payments deficit, so it is not desirable that this incentive should exist. Since adjustment loans rapidly increase the Bank's exposure, it is important that the Bank decline to lend when a loan would not reduce the need for further adjustment spending. Still, some argue against the view that adjustment loans should be re- stricted to those that reduce macroeconomic imbalances: all adjust- ment loans, whether or not they reduce such imbalances, help achieve desirable reforms in borrowing countries. The Bank's evaluation con- cludes that, under a comprehensive adjustment plan, the Bank would not be restricted to making adjustment loans that contribute directly to reducing macroeconomic imbalances. Bank management, how- ever, has asked for an assessment of such loans over the eighteen months following the evaluation. An adequate macroeconomic framework should accompany a sec- tor adjustment loan, but some argue against the idea that such loans should be limited to programs that directly correct macroeconomic imbalances. In particular, there is concern that the proposed guide- lines would exclude policy loans for social service sectors, especially in countries with no balance of payments problems. Conversely, there might be even tighter constraints on the Bank's adjustment operations (see box 2). Most developing countries are either in a fiscal bind or on the edge of one. Good programs can fail when little thought is given to cash flow. Once the programs are under way, debt service payments loom as a new and difficult problem. Even worse, a country team might get slipshod about the evaluation of the 28 ADJUSTMENT LENDING program itself if it forgets to consider sources for repayment. The team might see it as just what the doctor ordered to solve an immediate balance of payments need, for example, and not look seriously into the program's intrinsic productivity. Conditionality and Tranche Release The Bank's 1988 evaluation recommends that each adjustment loan explicitly list fewer-but key-conditions for tranche release. By in- cluding many conditions, the Bank avoids having to specify to poli- cymakers which ones are necessary, according to Stanley Fischer. This leaves open the possibility of releasing a second tranche if a govern- ment has met relatively unimportant conditions. The Bank has been moving in the direction of listing desirable (but not essential) condi- tions in the description of the loan and including only key conditions in the loan agreement itself. This seems to be the right way to focus the attention of the Bank and the country on what matters in each loan. A typical adjustment loan has a limited number of key conditions, say, four or five. These conditions are often stressed by government officials and Bank staff because they would likely influence stabiliza- tion or adjustment in the short run (over two or three years). The percentage of key conditions met was higher than that of nonkey conditions-68 percent compared with 56 percent. This shows that important conditions are better met and also confirms that adding numerous conditions to these loans is likely to generate negligible benefits. Furthermore, tranche release conditions are better met (71 percent for first tranche, 62 percent for second and third tranches) than conditions not tied to tranche release (43 percent). Still, there are examples of a borrower egregiously failing to meet tranche release conditions adequately and the Bank releasing the tranche anyway- after a respectable delay. The foregoing arguments provide the underpinning for several recommendations. Performance can be improved; a 60 percent imple- mentation rate is not spectacular and may, in any case, be inflated. Fewer conditions, more tightly worded, would ensure better imple- mentation. Large numbers of conditions make the Bank's job of monitoring implementation more difficult. Tightening conditions for second tranche release would also be helpful; the government should be expected to meet the conditions fully or with an equivalent pro- gram unless the conditions are no longer relevant. More reliance on prior conditions (actions already taken) would help, but the prospects for single-tranched loans are not clear, since the executive directors do not have a consensus on the merits of such an approach. SELECTED OPERATIONAL IMPLICATIONS 29 Few argue against putting a limit on the number of conditions. Whether the conditions are consistent and implementable is more important. Most agree that the Bank should avoid large numbers of conditions, especially because most countries would not be able to meet them. Monitoring and Evaluating Programs About 60 percent of the conditions in evaluated adjustment loans were met (see table 2). This proportion rises to 84 percent when the category "plus substantial progress" is added. Variation in implemen- tation across policy areas is pronounced. Performance is best when price changes or financial covenants (such as those regarding ex- change rates, agricultural and energy pricing, and rate of return or financial coverage) are involved and in cases in which political sensi- tivities are not strong (for example, some areas of budget policy and restructuring of public expenditure). Performance is weakest when institutional reforms (such as tax reforms, significant changes in export financing, and restructuring of public enterprises) and politi- cally sensitive measures (such as the removal of subsidies for fertiliz- ers or irrigation water, or the divestiture of public enterprises) are called for. Implementation of the conditions of sector adjustment loans is poorer than that of structural adjustment loans (51 percent of conditions met under the former, 68 percent under the latter). Since policy changes are not necessarily permanent, it is important to determine whether the intent of the original changes is preserved after disbursement of the loan is completed. By examining the postloan situation in several policy areas about which conditions had been set during the loan period, it is apparent that (1) there has been no substantial slippage in implementation of conditions in any im- portant policy area (one exception might be budget deficit reduction; few Bank loans, however, contain explicit conditions in this area, since the IMF usually handles it); (2) on average, 68 percent of conditions are being met, compared with 60 percent during the loan period; and (3) areas in which, because of institutional constraints, progress was slow * during the loan period have continued to improve. The Bank, in Russell Cheetham's view, should evaluate the results of past operations before proceeding to new ones. Bank operations in general and the Bank's 1988 evaluation itself give insufficient weight to evaluation. More systematic work is needed, especially when a series of adjustment operations must be evaluated in close succession. But what should be evaluated and how? Progress made in meeting original targets or evidence that policies were put in place even if there 30 ADJUSTMENT LENDING is not yet a supply response? Or preliminary assessment of costs and benefits of the adjustment program, by using supply response, for example? More research is needed in this area. There is some innovation in the methods of monitoring reforms under adjustment programs, particularly when the programs involve changes in the incentive and pricing systems. Changes in legislation are usually monitored. But there are no means of assessing situations in which these legislative changes have been replaced by hidden costs that affect the incentive system and thus reallocate resources. It is usually recommended, for example, that quantitative restrictions be replaced by tariffs. But if it now takes importers of a certain commod- ity six months instead of two to get the foreign exchange from the central bank, the restrictions under the new system may be worse than before. Once in a while the producers given such incentives should be sampled to judge whether the incentives represent an improve- ment and to see if hidden costs have emerged. Conclusions and Areas for Further Study Those who are familiar with the Bank's 1988 evaluation agree with the view that adjustment lending is a necessary and valuable instru- ment in the Bank kit for encouraging dialogue about lending and policy. The questions are when and how to use it, and how to improve it. This consensus contrasts with the view that the Bank is better off confining itself to traditional project lending. Concern about the volume of the Bank's adjustment lending also comes from those who view adjustment lending as risky to the Bank's portfolio. To limit adjustment operations to 25 percent of total Bank lending seems a reasonable and acceptable compromise. The evaluation's three prerequisites for making an adjustment loan are widely endorsed. Any disagreement centers on the interpretation of these prerequisites: if they are applied too rigidly, the Bank might miss the opportunity of supporting policy changes that sometimes develop in hesitant steps before the benefits of the program are revealed. Subjective judgments, therefore, are often necessary. Five areas of broad agreement about the Bank's adjustment lending can be highlighted. First, the three prerequisites have broad support as principles to be followed before adjustment lending is considered. But further, these prerequisites are worth applying to the actual practice of adjustment lending. CONCLUSIONS AND AREAS FOR FURTHER STUDY 31 Second, adjustment lending so far has been reasonably successful, given some reservations about methods of evaluation. Much more emphasis on monitoring and evaluation will be needed if the effec- tiveness of such lending is to be judged. Most agree on the need to develop an explicit macroeconomic framework to analyze adjustment programs (especially sector adjustment loans) and the effects of changes in budgetary, monetary, and exchange rate policies; opinion varies on how much quantitative analysis is desirable. Third, social costs of adjustment deserve greater attention. Indica- tors should be developed to demonstrate more readily the impact of adjustment programs on the poorest groups in society Government programs that clearly benefit the poor should be protected during adjustment. Fourth, adjustment programs are likely to remain complex in for- mulation and implementation, but in the area of lending conditions, greater focus, simplicity, and realism are desirable. This calls for more emphasis on establishing and following priorities and on supervising the implementation of reforms. Fifth, in view of the risks of adjustment lending and the limited capabilities for design and implementation, some limits on lending levels are desirable. Lending limits would also ensure an adequate balance between adjustment and investment lending. In the case of loans by the International Bank for Reconstruction and Development, the 25 percent overall limit recommended for purposes of portfolio management is reasonable. In a number of areas, consensus within the Bank would assist future adjustment lending. For example, in establishing conditions for an adjustment loan, how much credit is to be given to a government for actions carried out before the loan was applied for? And how can the Bank strengthen the interactions of the external financial commu- nity and governments in the formulation and implementation of policy? Much more analysis of several issues would be useful. In designing policy, particularly fiscal policy, under an adjustment program, Bank staff would benefit from understanding how sustainable various reforms are. Another fruitful area for analysis is the global consistency of adjustment policies, since, as a rule, a number of countries are expected to adopt similar policies, for example in agriculture, which is heavily protected in developed countries. These and other issues should be the subject of continuing assessment. 32 ADJUSTMENT LENDING Notes 1. The main findings of this study are published in Adjustment Lending: An Evaluation of Ten Years of Experience (Washington, D.C.: World Bank, 1988), hereafter referred to as the Bank's 1988 evaluation of adjustment lending, or by a shortened form of this phrase. 2. Since this is one of the resilient findings of the evaluation, a stronger case might have been made to continue to focus on trade reform in the years ahead. 3. The term "World Bank Group" refers to the International Bank for Reconstruction and Development, the International Development Association, the International Finance Corporation, and the Multilat- eral Investment Guarantee Agency. SELECTED READING 33 Selected Reading Balassa, Bela. 1982. "Structural Adjustment Policies in Developing Countries." World Development 10 (January): 23-28. Balassa, Bela, and F. Desmond McCarthy. 1985. "Adjustment Policies in Developing Countries, 1979-1983: An Update." World Bank Staff Working Paper 675. Washington, D.C. Berg, Elliot, and A. Batchelder. 1984. Structural Adjustment Lending: A Critical Analysis. Alexandria, Va.: Elliot Berg Associates. Bradford, C. I., and W. H. Branson, eds. 1987. Trade and Structural Change in Pacific Asia. Chicago: University of Chicago Press. Corbo, Vittorio, and Jaime de Melo. 1987. "Lessons from the Southern Cone Policy Reforms." World Bank Research Observer 2, no. 2: 111-42. Corbo, Vittorio, Morris Goldstein, and Mohsin Khan, eds. 1987. Growth-Oriented Adjustment Programs. Washington, D.C.: Interna- tional Monetary Fund and World Bank. Corden, W.-M. 1987. Protection and Liberalization: A Review of Analytical Issues. IMF Occasional Paper 54. Washington, D.C.: International Monetary Fund. Cornia, G. A., R. Jolly, and F. Stewart. 1987. Adjustment with a Human Face. Vol. 1. Oxford: Clarendon Press. Demery, Lionel, and Tony Addison. 1987. Alleviation of Poverty under Structural Adjustment. Washington, D.C.: World Bank. Donovan, Donald. 1982. "Macroeconomic Performance and Adjust- ment under Fund-Supported Programs: The Experience of the Seven- * ties." IMF Staff Papers 29, no. 2 (June): 171-203. Dombusch, Rudiger. 1987. Debt Problems and the World Macro Economy. * NBER Working Paper 2379. Cambridge, Mass.: National Bureau of Economic Research. Edwards, Sebastian. 1985. The Order of Liberalization of the Balance of Payments: Should the Current Account Be Opened Up First? World Bank Staff Working Paper 710. Washington, D.C. Fischer, Stanley. 1986. "Issues in Medium-Term Macroeconomic Ad- justment." World Bank Research Observer 1, no. 2 (July): 163-82. 34 ADJUSTMENT LENDLNJG Frenkel, Jacob. 1982. "The Order of Economic Liberalization: Discus- sion." In K. Brunner and A. H. Meltzer, eds. Economic Policy in a World of Change. Amsterdam: North-Holland. Genberg, H., and A. Swoboda. 1987. "The Medium-Term Relationship between Performance Indicators and Policy: A Cross Section Ap- proach." EPD Discussion Paper. World Bank, Economic Policy De- partment, Washington, D.C. International Monetary Fund. 1987. "Theoretical Aspects of the De- sign of Fund-Supported Adjustment Programs." Research Depart- ment Occasional Paper 55. Washington, D.C. Khan, A. R. 1986. "Policy Conditionality of Bank Lending: Some Issues." World Bank, Country Policy Department, Washington, D.C. Khan, M. S. 1987. "Macroeconomic Adjustment in Developing Coun- tries: A Policy Perspective." World Bank Research Observer 2, no. 1 (January): 23-42. . 1988. "The Macroeconomic Effects of Fund-Supported Ad- justment Programs: An Empirical Assessment." International Mone- tary Fund, Washington, D.C. Killick, Tony, ed. 1984. The IMF and Stabilization: Developing Country Experiences. London: Heinemann. Krugman, P. R. 1987. Strategic Trade Policy and the New International Economics. Cambridge, Mass.: MIT Press. Michaely, Michael, Armeane Choksi, and Demetrios Papageorgiou. Forthcoming. Liberalizing Foreign Trade: Lessons of Experience in the Developing Countries. Oxford: Blackwell. Mussa, Michael. 1987. "Macroeconomic Policy and Trade Liberaliza- tion: Some Guidelines." World Bank Research Observer 2, no. 1 (Janu- ary): 61-78. Nelson, Joan. 1984. "The Political Economy of Stabilization: Commit- ment, Capacity and Public Responses." World Development 12, no. 10 (October): 983-1006. Pinstrup-Anderson, Per. Forthcoming. "The Impact of Macroeco- nomic Adjustment on Nutrition." In S. Commander, ed. Structural Adjustment in Agriculture: Theory and Practice. London: Curry. SELECTED READING 35 Please, Stanley. 1984. "Structural Adjustment, IMF Conditionality and the World Bank." In Stanley Please, ed., The Hobbled Giant: Essays on the World Bank. Boulder, Colo.: Westview. Sachs, Jeffrey, ed. 1986. "Conditionality and the Debt Crisis: Some Thoughts for the World Bank." World Bank, Development Research Department, Washington, D.C. Sen, A. K. 1983. "Development: Which Way Now?" Economic Journal 93, no. 372:745-62. Stem, Ernest. 1983. "World Bank Financing of Structural Adjust- ment." In John Williamson, ed. IMF Conditionality. Washington, D.C.: Institute for International Economics. Streeten, Paul. Forthcoming. "Structural Adjustment: A Survey of the Issues and Options." In S. Commander, ed. Structural Adjustment in Agriculture: Theory and Practice. London: Curry. Thomas, Vinod. 1988. "Issues in Adjustment Lending." PPR Working Paper 2. World Bank, Policy, Planning, and Research Department, Washington, D.C. World Bank. 1988. Adjustment Lending: An Evaluation of Ten Years of Experience. Policy and Research Paper 1. Washington, D.C. . 1987. "Industrial Adjustment in Sub-Saharan Africa." World Bank, Economic Development Institute, Washington, D.C. .1986. Recovery in the Developing World: The London Symposium on the World Bank's Role. Washington, D.C.: World Bank. U The World Bank Headquarters 1818 H Street, N.W. Washington, D.C. 20433, U.S.A. Telephone: (202) 477-1234 Facsimile: (202) 477-6391 Telex: RCA 248423 WORLDBK WUI 64145 WORLDBANK Cable address: INTBAFRAD WASHINGTONDC European Office 66, avenue d'Iena 75116 Paris, France Telephone: (1) 40.69.30.00 Facsimile: (1) 47.20.19.66 Telex: 842-620628 Tokyo Office Kokusai Building 1-1, Marunouchi 3-chome Chiyoda-ku, Tokyo 100, Japan Telephone: (3) 214-5001 Facsimile: (3) 214-3657 Telex: 781-26838 ISBN 0-8213-1233-2