Events of the late 1980s and early 1990s created an unusually high demand for IMF financial resources. The strains in the international economy during the 1980s—a sharp reduction in spontaneous external financing for most developing countries and large drops in prices of many primary commodities—produced or exacerbated macroeconomic imbalance in many countries: large fiscal and external current account deficits that had been financed abroad needed to be curtailed, often just as export market conditions were deteriorating. In these circumstances, the number of countries seeking conditional financial assistance from the IMF swelled from an annual average of 19 during the 1970s to 39 during the first half of the 1980s and 44 during the second half. These numbers were pushed even higher during the early 1990s by the addition of the several new transforming economies in Eastern Europe and the former Soviet Union.1
Along with the increase in the use of IMF resources, changes in the conditionality attached to the use of IMF resources—that is, in the policies that the IMF expects a country to follow in order to avail itself of credit from the IMF—occurred during the 1980s.2 These changes resulted from several influences, two of which were particularly important. First, the macroeconomic imbalances facing countries that sought IMF support in this period—revealed in the size of balance of payments financing gaps and the level of domestic inflation—were large by historical standards, reflecting both the magnitude of external shocks and the tendency of countries to delay adjustment until crisis or near-crisis situations arose. Second, underlying distortions in the economy, such as excessive government intervention or measures that shielded domestic producers from foreign competition, impeded adjustment and often needed to be addressed for an adjustment program to be sustained.
In such circumstances, the conditionality attached to IMF support was pushed beyond the traditional emphasis on demand restraint, which was appropriate for relieving temporary or cyclical balance of payments difficulties. For many countries, especially heavily indebted ones, reliance on demand restraint alone for the needed adjustment would have required unacceptably long periods of depressed domestic demand, and supply-side policies—such as reducing the role of government in the economy and opening the economy to outside competition—aimed at resuming growth became an important part of conditionality. Complementing this shift in emphasis toward supply-side policies, programs and the conditionality they included began to be developed with more of a focus on the medium term. This reflected both the fact that many problems at issue could be addressed only through a steady, step-by-step program of structural and financial policy changes over a period of years and the recognition that the sustainability of financial policy changes needed to be a criterion for determining the appropriateness of a country’s policy program and the conditionality attached to the use of IMF resources.
In an effort to assimilate the recent experience with conditionality, both in its traditional and evolving forms, this review describes and evaluates the design of adjustment programs supported by the stand-by and extended facilities during the late-1980s and early 1990s.3 The review examines the record of countries for which upper credit tranche stand-by and extended arrangements were approved between mid-1988 and mid-1991—45 arrangements for a total of 36 countries (Chart 1 and Box 1). Its principal objective is to understand the nature and strength of the economic adjustment undertaken and the degree of sustained improvement in macroeconomic performance.
The Arrangements in Brief
IMF Financing
The types of credit arrangements (“facilities”) the IMF extends to its member countries include regular facilities (stand-by and extended arrangements), concessional facilities for low-income countries (SAF and ESAF), and various special facilities.
Regular Facilities
The two types of regular IMF arrangements have different terms and conditions, but both provide credit at market-related terms.
Stand-by arrangements typically cover periods of one to two years and focus on macroeconomic policies—such as fiscal, monetary, and exchange rate policies—aimed at overcoming balance of payments difficulties. Most performance criteria to assess policy implementation—such as budgetary and credit ceilings, reserve and extended debt targets, and avoidance of restrictions on current payments and transfers—are applied during the period of the arrangement, and purchases are made in installments.1 Repurchases are made in 3¼ to 5 years. Most programs supported by stand-by arrangements also include some policies to address structural or supply-side weakness, although because of the short duration of these arrangements, this focus is not as strong as in extended arrangements.
Under the extended Fund facility (EFF), the IMF supports medium-term programs through extended arrangements that generally run for three years (sometimes four years), and are aimed at overcoming balance of payments difficulties stemming from macroeconomic and structural problems. Typically, a program states the general objectives for the first year; policies for subsequent years are spelled out in program reviews. Performance criteria are applied, similar to those in stand-by arrangements, and repurchases are made in 4½ to 10 years.
Concessional Facilities for Low-Income Countries
The structural adjustment facility (SAF) and enhanced structural adjustment facility (ESAF)—set up in March 1986 and December 1987, respectively—offer highly concessional loans to support macroeconomic adjustment and structural reform in low-income countries. Under both facilities, the member, with the help of the IMF and the World Bank, sets out structural and financial policies for a rolling three-year period, in a policy framework paper (PFP). Within this framework, detailed yearly policy programs are formulated and are supported by SAF/ESAF arrangements, under which annual loan disbursements are made. The interest rate on loans is 0.5 percent and repayments are made in 57: to 10 years.
Special Facilities
A number of facilities have been created throughout the IMF’s history in response to members’ special needs. Such assistance is additional to that available under other IMF facilities, but members cannot finance the same balance of payments needs under both regular and special facilities.
Arrangements Under Review
The IMF periodically undertakes reviews of adjustment programs supported by its facilities. The present review assesses the experience of 36 countries under 45 stand-by and extended arrangements that were approved between mid-1988 and mid-1991 (that is, more than one arrangement was approved for 8 countries during the period; see Chart 1). The commitment of IMF resources was SDR 18.7 billion (about US$25 billion), over half of which was in support of six extended arrangements. Annual average access amounted to 50 percent of quota (a member’s share in the IMFs capital) under stand-by arrangements and 66 percent under extended arrangements. In eleven of the arrangements, all purchases were completed, and in five no purchases were made. In the remaining 29 arrangements, 52 percent of commitments were utilized In eight of these, only the first scheduled purchase, that available upon approval of the arrangement by the IMF Executive Board, was made.
1 Under regular IMF facilities, a member does not “borrow” from the organization; it purchases usable (hard) currencies with its own domestic currency. A member country repays the IMF by repurchasing its own currency with currencies acceptable to the IMF.
Stand-By and Extended Arrangements1
(Solid lines denote the original duration of stand-by arrangements unless otherwise noted)
1 Countries for which an arrangement was approved during July 1988-June 1991. Circles indicate the date of approval, original expiration, and extended expiration; extension periods are indicated by a broken line. Triangles indicate cancellation dates. For countries that were not eligible to make all purchases, asterisks indicate the last date on which they were eligible to purchase. EFF = extended Fund facility; SAF = structural adjustment facility: ESAF = enhanced structural adjustment facility.2 Hungary had had three previous arrangements with the IMF but is included with the other Central European countries in this review because of the common influences they faced.Stand-By and Extended Arrangements1
(Solid lines denote the original duration of stand-by arrangements unless otherwise noted)
1 Countries for which an arrangement was approved during July 1988-June 1991. Circles indicate the date of approval, original expiration, and extended expiration; extension periods are indicated by a broken line. Triangles indicate cancellation dates. For countries that were not eligible to make all purchases, asterisks indicate the last date on which they were eligible to purchase. EFF = extended Fund facility; SAF = structural adjustment facility: ESAF = enhanced structural adjustment facility.2 Hungary had had three previous arrangements with the IMF but is included with the other Central European countries in this review because of the common influences they faced.Stand-By and Extended Arrangements1
(Solid lines denote the original duration of stand-by arrangements unless otherwise noted)
1 Countries for which an arrangement was approved during July 1988-June 1991. Circles indicate the date of approval, original expiration, and extended expiration; extension periods are indicated by a broken line. Triangles indicate cancellation dates. For countries that were not eligible to make all purchases, asterisks indicate the last date on which they were eligible to purchase. EFF = extended Fund facility; SAF = structural adjustment facility: ESAF = enhanced structural adjustment facility.2 Hungary had had three previous arrangements with the IMF but is included with the other Central European countries in this review because of the common influences they faced.Methodology and Scope
This review is part of a periodic process in which the IMF reviews conditionality and examines member countries’ performance during adjustment programs. The present review, which comprises this and a companion volume of background papers.4 is the first of stand-by and extended arrangements that has been made available to the public. An unusual feature of this review is that it examines data on policies and outcomes not just from the periods during which arrangements with the IMF were in place. Rather, it follows developments in each country from the approval of its first arrangement during the period under review through 1992 (the most recent year for which data were widely available at the time of writing). For most countries, this period extends past the arrangements under review. This approach is in keeping with the increasingly medium-term orientation of programs and conditionality. It reflects the recognition that progress in stabilizing and reforming an economy is most appropriately judged against the durability of sound policies and the sustained improvement in economic performance. Developments are reported using a mix of before-after, target-actual, and control-group comparisons for individual and various groupings of countries.
The issues addressed in this review and the methodology employed represent a deliberate choice among many potentially interesting possibilities. A decision was taken to examine developments in a large group of countries, rather than to focus in detail on a few case studies. While case studies permit more depth in the analysis of policy choices and their effects, they cannot convey the breadth of problems, strategies, accomplishments, and shortcomings that a larger survey allows.
Another possibility would have been to follow up on econometric studies that seek to answer the question, “Did IMF-supported programs have a statistically significant influence on macroeconomic performance?”5 At the heart of such an analysis is the comparison of actual developments in key macro-economic variables with the counterfactual—the outcome that would have occurred in the absence of IMF support. In principle, such an analysis has the attributes of separating the effect of IMF support from that of exogenous influences and initial conditions and eliminating sample selection bias in comparisons of outcomes in programs with those in a non-IMF-supported control group. It suffers, however, from two weaknesses: the arbitrariness of the counterfactual, which by its nature is unobservable; and, often, unstable results from estimating a simple macroeconomic model over a sample of fundamentally diverse economies. Moreover, attempts to quantify the “independent effect” of IMF-supported programs aim to measure the value added by IMF support but are not especially helpful in identifying ways to improve the design of such programs—an important objective of this review.6 It must be recognized, however, that without systematic controls for sample selection bias, exogenous influences, and initial conditions, the before-after comparisons reported in this review cannot be interpreted as indicating solely the effect of IMF support.
A decision was also taken not to address questions about the basic paradigm in which programs are cast. This paradigm, examined in recent studies, usually with long-term data, emphasizes the importance of financial stability, an outward, pro-trade orientation of policies, and reliance on market mechanisms rather than government dictates to determine resource allocation.7 This review focuses rather on the choice of policies and their implementation within this broad framework. These more narrowly defined policy choices still raise many critical issues because countries undertaking IMF-supported programs do not exist in a first-best world where stabilization and structural reform can be pursued at an optimal pace in a closely coordinated fashion. The design of programs reflected the frictions in a second-best world where political and administrative constraints often required short-term trade-offs between macro-economic objectives.8
The scope of this review is broad, and many interesting issues could be identified and commented upon only briefly. It is hoped that this process will motivate further study of specific issues.9 Several policy issues, however, were of particular interest, and they were studied in depth:
The composition of fiscal adjustment.
The behavior of interest rates in the face of financial liberalization and changes in fiscal financing.
The use of the exchange rate as a nominal anchor.
The role of wage controls in transition economies.
The effectiveness of policy measures for raising private savings.
These issues are the subjects of the background papers in the accompanying volume, the main conclusions of which are summarized in this volume. It must be recognized that the limited sample of countries and period under review constrained these exercises—particularly in the use of econometric techniques to examine hypotheses. Although the countries reviewed are distinctive for the severity of their external financing constraint and financial and structural weaknesses, the policy issues facing them were, in many respects, similar to those in many countries outside the sample. Because conclusions on policy design would be strengthened by including these countries, further analyses of these issues with larger samples would be useful.
Summary of Findings and Conclusions
Initial Conditions and Adjustment Strategy
Economic conditions in the 36 countries reviewed, almost without exception, were dire at the outset of their arrangements. Most were suffering the effects of long-standing and severe institutional and structural weaknesses, a lack of fiscal discipline, and weak external competitiveness, and many had been pushed to a crisis by adverse terms of trade or other exogenous developments. The associated deterioration in the external accounts, erosion of official reserves, slow growth, inadequate savings and investment, and frequently rising inflation had often persisted for years. Typically, countries had first responded to emerging difficulties by borrowing abroad without adequately addressing underlying problems. They usually had sought IMF support only after access to commercial financing had ceased and refinancing of external debt was needed. Most of the countries under review that had used IMF resources at least twice in the previous eight years (“repeat users”) continued to face extraordinary debt burdens.10
Not surprisingly in these circumstances, adjustment was protracted. Typically, not only the IMF but also the World Bank and other multilateral and bilateral agencies supported the adjustment process. On the basis of the IMF’s involvement alone, however, every arrangement was part of a process that had begun before or would extend beyond the period considered (see Chart 1). Thus, programs, whether supported through stand-by or extended arrangements, were effectively implemented over the medium term.
Although countries shared unusually severe external financing constraints, programs needed to accommodate a wide variety of initial conditions. For example, in most of the repeat users, unsustainable balance of payments positions often stemmed from external debt overhangs, while in the newer users they more frequently reflected acute current excesses of absorption over output. In many of the Western Hemisphere and Eastern European countries, excess demand pressure was manifest in high inflation, but in many African, Asian, and Middle Eastern countries, inflation was rather low. Almost all countries had major structural weaknesses, but some in the context of already open exchange and trade systems, relatively sophisticated financial markets, or small public sectors. Perhaps most important, countries’ political commitment to adjustment and reform varied widely. The Central European countries added new dimensions to this diversity: their need for and commitment to structural reform surpassed that in most other countries, and the collapse of their principal export markets and sharp deterioration in their terms of trade made their external adjustment task unusually daunting.
The overarching framework for the design of programs was a three-pronged approach involving (i) the reining in of domestic demand through fiscal and credit restraint, (ii) structural reforms to promote a supply response and improve the efficiency of resource use, and (iii) the securing of external financing to support the program (and often to clear external arrears). In principle, the extent of reliance on each element should have depended on the source of a country’s problems and its stage in the adjustment process. For the most part, countries carried out policies broadly as envisaged in programs, although there was substantial variation in the completeness of implementation. Some countries suffered reversals after the programs covered in this review. Others slipped in the implementation of policies during the period reviewed, but these slippages paved the way for strong adjustment in subsequent programs.
Economic Developments: External Adjustment Before Recovery
In most countries, macroeconomic performance improved during the arrangements and, in many, continued to do so afterward. Improvements were large in the external sector. In countries that started their arrangements in the midst of acute balance of payments crises (mainly the new users), a marked improvement occurred quickly, and in other countries with less acute difficulties underlying problems were addressed: official reserves rose to a more comfortable level; about half of the countries that had had external arrears substantially or completely cleared them; some countries (particularly among those that had initiated their adjustment efforts in the early 1980s) benefited from large increases in capital inflows—a reflection of changing conditions abroad as well as improved investor confidence in domestic policies; and current account positions converged toward projected sustainable capital inflows. There were disappointments, however, even in the external sector: particularly in sub-Saharan Africa and Central Europe, capital inflows failed to materialize as projected; several countries did not fully resolve their debt overhangs and arrears; and, in most countries, strong export performance early in the programs faded toward the end of the period reviewed.
Developments in the domestic economy were generally less impressive. On inflation, a few countries achieved dramatic reductions from very high initial rates, but many continued to experience moderately high inflation; a few even saw upward trends. For the bulk of the countries outside Central Europe, there was some strengthening of growth and on average an increase in saving rates, but the record reveals few if any countries shifting to a distinctly more rapid pace of growth backed by higher saving ratios, even among the repeat users of IMF resources. Most worrisome of all was that few countries saw any increase in overall investment ratios; nevertheless, on average, private investment rates rose as public investment rates fell, and incremental capital-output ratios fell.
Were Adjustment Strategies Flawed?
These developments raise several questions about program design that are the focus of this review. At the heart of these questions is a concern about whether the IMF’s basic three-pronged approach to adjustment and reform is flawed. More specifically, does it place too much emphasis on achieving a sustainable balance of payments, subordinating domestic objectives—for inflation, saving, investment, and growth—to that of external equilibrium? Does the one- to two-year duration of most programs prevent adequate consideration of medium-term problems and effects? Within the broad three-pronged approach to adjustment and reform, were programs adequately tailored to the specific problems of each country?
Answers to these questions, of course, vary from program to program. Ideally, they would stem from an investigation for each country of counterfactuals that could determine whether an adjustment strategy different from that supported by the IMF would have been more successful in improving all aspects of economic performance. Such exercises are beyond the scope of this paper. Rather, this paper starts from the premise that reviewing the programs in a sizable group of countries can provide insights into the priority objectives of programs, the steadiness of policy implementation, and the degree to which adjustment programs were differentiated according to individual countries’ initial positions.
Because countries typically sought IMF support only when they had a serious external financing problem (typically having reached the point where IMF involvement was needed for essential debt rescheduling), resolving this problem did command the highest priority. Nevertheless, most programs appear to have pursued this goal with sensitivity toward the ultimate goal for economic policies—to improve living standards through higher growth. Fostering sustainable growth, however, is invariably a protracted process, only yielding clear gains beyond the short term and requiring steadfast adherence to sensible policies over a long period. Most programs, therefore, aimed to secure immediate improvements in external finances while establishing conditions for greater efficiency, savings, investment, and growth over the longer term.
The three-pronged approach to adjustment and reform conforms to the weight of empirical evidence from studies on (i) the most effective ways of correcting external payments problems in the short term; and (ii) the role in promoting growth of stable external positions, manageable levels of debt, moderate inflation, and structural conditions that minimize distortions. The acceptance of this basic paradigm, however, is only the first step in designing an adjustment strategy. Within this broad paradigm, programs entail many choices—and compromises that involve acceptance of second-best policies as part of an overall package—that can affect the prospects for external stability, inflation, and the speed of the response of saving, investment, and growth. In these choices there was frequently room for improvement in the programs reviewed.
A major objective of this review was to identify areas in which improvements could be made. At the outset, it must be recognized that many of the choices made in designing programs involve issues of which our understanding is imperfect—the optimal sequencing of structural reforms and factors that motivate private investors, to name only two; a better understanding of these issues would make for stronger programs. Program design is also frequently affected by inadequacies in data that could not be fully addressed in the short term. Yet the broad conclusion of the review is that even within our current understanding of longer-term economic relationships, the design and presentation of programs can be improved: first, by strengthening our understanding of the underlying causes of economic problems and the needs of each country; and second, by considering more explicitly the costs and implications of second-best policies that may result from political pressures or administrative constraints—for example, suboptimal composition of fiscal adjustment, inadequate structural reforms, or fiscal adjustment insufficient to provide a high degree of assurance that inflationary financing or recourse to external borrowing or arrears will not be required. Specific conclusions on these kinds of issues are summarized in the following sections.
Fiscal Adjustment: Medium-Term Frameworks Needed
For the 36 countries reviewed, fiscal adjustment strategics were tailored, in several crucial respects, to the specific nature of the imbalances.11 Fiscal adjustment, on average, was to derive about equally from revenue increases and expenditure restraint; planned expenditure cuts, however, were largest in the countries with the highest initial expenditure ratios, and planned revenue increases were largest in countries where initial revenue ratios were lowest.12 Also, targeted changes in overall deficits were greatest in the countries with the largest initial imbalances. As for outcomes, the analysis shows that fiscal adjustment was on average close to that targeted: with available data, however, and the short period considered it is impossible to account precisely for the separate roles of discretionary policy change, exogenous influences, and cyclical factors. On average, spending restraint was somewhat greater than planned, and revenue increases were slightly short of target. Proportionately, restraint of capital spending was greater on average than that of current spending.
These findings suggest that fiscal adjustment strategies were sensitive to countries’ initial conditions, and policies were largely implemented as planned. However, programs were often not presented in the context of an explicit medium-term analysis of the potential for revenue mobilization, expenditure needs, and financing. It is, therefore, not possible to judge the degree to which fiscal adjustment moved countries toward sustainable fiscal positions. Without such an explicit medium-term perspective, it is impossible to dispel any impression that programs call continually for cuts in the deficit without a clear view of the end point of the process. Programs would benefit from a greater focus on medium-term frameworks in each country—even though these would be neither precise nor unvarying standards.
Medium-term fiscal frameworks should be built around considerations of the desired path of money creation and inflation, the intertemporal dynamics governing increases in the government debt ratio, and the impact of fiscal adjustment on the contemporaneous savings-investment balance in the economy. Data requirements for such exercises are considerable and in some countries weak data preclude meaningful analysis of the medium-term fiscal outlook. Also, when possible, programs should take fuller account of situations where current decisions may adversely affect future fiscal trends or prospects for investment and growth—for example, where protecting critical capital spending is important, where capital spending now would require unmanageable future current spending, where social expenditures have borne a large share of cuts, or where revenue measures are questionable from a medium-term perspective.
Closer attention to the structure of spending and revenue raises questions about the limit to the expertise of the IMF and the legitimate scope for IMF intervention in sovereign decisions. IMF staff typically can identify egregious departures from reasonable structures of spending and revenue, but there is a real question about where on the spectrum of detailed involvement in fiscal decisions the IMF staff should be: at one end, confining their advice to questions of the level of government expenditure and the size of deficits; at the other, actively judging decisions about the structure of expenditure and revenue. In the programs reviewed, the IMF staff generally confined their involvement in fiscal structural issues to areas where expenditure and revenue policies clearly distorted relative prices and impaired the efficiency of the economy. In programs where broader issues arise, closer coordination with the World Bank, where timely analyses of public expenditures need to be carried out, is important.
Financial Programs and Other Financial Market issues: Achievements Within Limits
In broad terms, the study found that targets for money were exceeded in two thirds of the program years.13 About as often, inflation targets were missed by sizable margins. Most missed money targets reflected reasonably close adherence to ceilings on domestic credit but larger-than-targeted net foreign assets. Financial market liberalization, particularly the decontrol of interest rates and the introduction of auctions for government paper, progressed quite remarkably.14 As a result, real interest rates often rose from negative to positive levels—sometimes even to levels that could be considered excessive. In general, unsustainably high real rates were temporary and typically fell back to relatively low levels within one or two years.
What policy implications can be drawn? First, financial programs appear to have contributed to casing external constraints, especially increasing reserves. They were less successful in reducing money creation and inflation owing to the inherent limitations of credit ceilings: when inflation expectations are entrenched and nominal appreciations resisted, capital inflows can feed money growth that accommodates higher-than-targeted inflation. To help redress this limitation, programs would benefit from more explicit benchmarks for an appropriate monetary aggregate. However, in light of the difficulty of projecting and interpreting large fluctuations in the demand for money, such benchmarks should be used more as signals of possible dangers for inflation targets than as absolute limits. Then, any response to deviations from benchmarks can be guided by the causes of the deviation and by the short-term tradeoffs between building reserves, controlling money, and maintaining competitiveness.
Second, financial programs must ensure adequate fiscal adjustment to avoid excessive reliance on light credit, high real interest rates, and unsustainable capital inflows. Judging the appropriateness of the mix of financial policies—even ex post—is difficult. Typical indicators of insufficient fiscal adjustment and excessive reliance on credit restraint—persistently high real interest rates, large capital inflows, or weak private investment—can occur for a number of reasons unrelated to excessive credit restraint. Few of the countries reviewed exhibited clear evidence of excessive reliance on credit restraint. Nevertheless, further work on the development of indicators is needed.
Third, financial liberalization and shifts in fiscal financing can enlarge the market-clearing role of interest rates, but care is needed to avoid excessive levels of real interest rates and large spreads. In the programs reviewed, high real interest rates tended to subside as policy credibility grew. Nevertheless, the experience suggests that two aspects of liberalization be given greater attention: (i) ensuring that the shift from bank to nonbank budget financing is accompanied by adequate reductions in overall public sector borrowing; and (ii) improving prudential controls, bank supervision, and the competitive structure of financial markets to minimize oligopolistic behavior and reduce upward pressure on interest rates from adverse selection.
Exchange Rate Policy: To Anchor or Not to Anchor?
Countries that used the exchange rate as a nominal anchor—that is, fixed or followed a preannounced path for the nominal exchange rate—differed in distinct ways from those that managed their exchange rate more flexibly.15 Among the 15 countries that used the exchange rate as a nominal anchor, most initially had either very high or quite low inflation rates. By contrast, countries with moderately high inflation generally chose to manage their rates more flexibly. The differing approaches to exchange rate policy reflected the choice between two possible short-term roles for the exchange rate—on the one hand, anchoring inflation expectations and disciplining policies, and, on the other, maintaining or improving competitiveness. In general, the countries that chose to have an exchange rate anchor reduced inflation or maintained it at low levels. The cost was typically a loss of price competitiveness as inflation fell but not to the level in trading partners. In two countries, the exchange rate anchor was not sufficiently supported by other policies and proved unsustainable. The countries that chose more flexible exchange rate policies tended to see little decline or even increases in inflation, but on average some improvement in competitiveness.
The experience confirms the effectiveness of exchange rate anchors in reducing high or sustaining low inflation when supported by sound macro-economic discipline. This suggests that in a first-best world, where financial policies could be sufficiently restrictive, there would be clear rewards to using the exchange rate as an anchor for a period of at least a few months to one year. In fact, policies are often not strong enough, and a short-term trade-off between inflation and competitiveness arises. The choice between them is frequently dictated by the external constraint. Programs should, however, clearly consider the trade-off in designing exchange rate policy. Also, the trade-off needs to be kept under review—recognizing that an anchor may have to give way to greater flexibility in the face of adverse terms of trade movements or a protracted loss of competitiveness.
In countries using the exchange rate more flexibly, limits to the effectiveness of nominal depreciations in achieving real depreciations must be recognized. Explicit real exchange rate rules—that is, specific targets for the real exchange rate that guide nominal exchange rate policy—should be avoided, even in low-inflation settings: decisions on whether to adjust nominal rates to achieve real changes should take into account the prospects for needed support from fiscal and wage policies.
Medium-Term Programming: The Fundamental Challenge
Each of the programs reviewed was part of an adjustment process that began before and/or would extend beyond the period covered by the review. The protracted nature of adjustment is not surprising, in view of the depth of the financial and structural weaknesses in most of the countries under review. It meant, however, that the IMF was generally providing support over the medium term and that issues beyond the financial program and immediate external financing constraint needed to be addressed by the adjustment strategy.
All countries pursued structural reforms—typically fastest and first in exchange, trade, financial, and, in Central Europe, price liberalization. Reform was slower in the areas of taxes, public spending, and, particularly, public enterprises and labor markets. There was no evident pattern in the sequencing of reforms, which appears to have been more a function of political opportunities and administrative capabilities than optimal sequencing considerations. In some cases this may have led to adverse effects from inappropriate sequencing. In general, however, the analytical basis for comparing different states of a second-best world—that is, different sequences for removing existing distortions—is weak, and insisting on specific patterns of sequencing may often have slowed or even obstructed structural reform as politically difficult actions were delayed. Thus, except when there are clear risks in a specific pattern of reform, it may be wisest to move as rapidly as possible with reforms when they are technically and politically feasible, and to work to channel any resulting tensions into a consensus for further change. It is important, however, to recognize and explore the costs of sequences of reforms that are particularly risky; for example, financial market liberalization should not take place before adequate prudential controls and bank supervision are in place.
Although widespread, labor market rigidities were addressed to only a limited degree. In many countries, real wages exhibited a lack of responsiveness in the organized sector to labor market conditions. Backward-looking wage-indexing arrangements (de facto or explicit) existed in many countries, especially those with high and intermediate inflation rates. Modifying them was essential to lowering inflation, reducing real wages to market-clearing levels, and girding economies against disruptions from possible real shocks. In most countries, notwithstanding recognition of the problem, a political consensus for changing indexation practices could not be secured. Poor data and the political sensitivity of wage issues hampered an evaluation of the degree to which labor market practices and wage-setting behavior constrained job creation, investment, and growth. Greater efforts are needed to collect wage and employment data and, where possible, apply analytical techniques common in industrial countries for judging whether real wages are excessive.
On savings, the analysis generally bears out the conclusions of other studies that the channels for policies to influence saving rates directly are limited.16 It shows that changes in government savings were the most direct channel of influence on aggregate saving rates, although on average about half of any change was offset by opposite movements in private savings. Other principal, systematic determinants of private savings—actual and expected growth rates and the terms of trade—were beyond the direct control of policies. The more policy-controlled influences—changes in real interest rates and exchange rates, financial reform, and pension reform—proved to have weak and unpredictable effects over the program period. Moreover, the very success of programs in attracting capital inflows and raising expectations about future growth may stimulate private consumption in the short run and depress private saving ratios.
Explaining the sluggishness of investment ratios during programs will ultimately require rigorous work with longer time series. For now, several perspectives on the poor investment response suggest themselves. First, it seems unlikely that deficient final demand was the main culprit: on average, real consumption and export volume grew during arrangements. Excess capacity, especially where output had fallen prior to the arrangements, however, may not have been fully reabsorbed, so that an investment response awaited the meeting of capacity constraints. Also, the efficiency of investment may have improved owing to better structural and financial policies. Second, deficient savings, domestic or foreign, may have been a constraint on investment, although some countries that had large inflows of capital saw more of a response in consumption than investment. Significantly higher external financing on generous terms could have been channeled into higher investment but possibly only through the public sector. Third, on average, the ratio of government investment to GDP fell. How these developments may have affected longer term trends in overall investment is unclear: in some countries, worthwhile infrastructure projects were probably a casualty, but in others, especially countries with very high initial public investment rates, wasteful projects were probably cut. Only a case-by-case analysis of changes in the composition of government investment could produce a well-informed judgment on the effect of restraint.
Generally, the experience bears out the findings of broader samples of developing countries over longer periods:17 countries need to establish a track record of sound policies before private investment strengthens significantly, and this takes time. Beyond the need for policy discipline, many of the modifications of program design discussed above could help to hasten the investment response putting fiscal adjustment in a medium-term framework: for example, reducing inflation more effectively, more publicly, and attending to rigidities in labor markets, especially wage determination.
The modest response of growth raises many of the same issues as the weak investment response. In fact, the two are related: any shift to distinctly faster growth in the countries reviewed would have required increases in investment rates, although clearly ample scope existed for improving the efficiency of resource use. Growth did rise in most of the new users of IMF resources even as generally aggressive fiscal retrenchment and other stabilization measures were implemented,18 Apparently, this resulted from improvements in the efficiency of resource use, often as economies rebounded from unusually weak or negative growth preceding the program. Nevertheless, more attention needs to be focused on why countries found it difficult to shift to distinctly faster growth tracks (for example, growth rates in the repeat users of IMF resources averaged about 2 percent). Further study of this issue would be closely related to that of the influence of program design on investment.
Improving Data
The quality of the data for many of the countries reviewed is poor. Frequent and large revisions, even up to five years after programs, and the often inconsistent definitions across countries severely constrain program design and limit the conclusions that can be drawn about adjustment policies and their effects. A particularly serious problem for program design is the wide variation in the coverage of fiscal accounts—from central government only in many countries to the full public sector in a few—especially when financial flows between levels of government are significant. Moreover, the full extent of quasi-fiscal activity is frequently unknown. National accounts data that are based on poor survey techniques and are also often massively revised make it difficult to design programs that are responsive to developments in investment and growth and, indeed, even call into question the very reliability of the record on investment, savings, and growth.
The chapters that follow review the overall economic policies and outcomes in the 36 countries for which upper credit tranche stand-by and extended arrangements were approved between mid-1988 and mid-1991. In structure, they follow rather closely the sequence of steps taken by staff teams working with governments to formulate an adjustment program and, ex post, to understand developments during the program period. Chapter 2 begins with a summary of conditions in the countries prior to the programs under review, then briefly portrays the external environment during the program period. Chapter 3 describes the adjustment strategies followed, focusing particularly on whether and how differences in the needs and structures of individual countries were accommodated within the IMF’s overarching approach to adjustment. Policy outturns are then reviewed in Chapter 4 through both target-actual and before-after comparisons. Chapter 5 concludes the analysis with a brief report of macroeconomic developments during the period from the first program year under review through 1992, the most recent year for which data were generally available at the time of writing. Finally, for the information of readers. Chapter 6 summarizes the IMF Executive Board’s consideration of these staff studies on conditionality.
Much of the increase in demand for IMF resources was met after 1986 through the newly established structural and enhanced structural adjustment facilities (SAF and ESAF respectively). Through these facilities, the traditional terms of IMF credit were adapted to the limited debt-servicing capacity and the need of prolonged adjustment of many low-income countries. Thus, use of IMF facilities providing conditional credit at market-related terms—stand-by arrangements for short-term credits and extended arrangements for longer-term credits-rose sharply through the early 1980s but subsequently fell back to the annual average of the 1970s.
Polak (1991) has a wide-ranging discussion of the influences on and the changing nature of conditionality during the 1980s. That paper also provides an explanation of many concepts and terms related to conditionality.
Reviews of adjustment programs supported by the SAF and ESAF are conducted separately owing to differences in the nature of the problems facing the countries involved and the timeframe of the adjustment. See Schadler, Rozwadowski, Tiwari, and Robinson (1993) for the most recent review of the ESAF programs.
IMF Conditionalily: Experience Under Sland-By and Extended Arrangements, Part II. Background Papers, IMF Occasional Paper No. 129, see Schadler (1995). Hereinafter cited as Background Papers.
For example, comparisons of actual developments with the counterfactual could show no effect of IMF support in two quite different situations—when the IMF supported weak policies but policies would have been weak anyway, and when the IMF supported strong policies but policies would have been strong anynized, way. Identifying how program design could be improved requires distinguishing between these two situations.
There have been many recent studies that offer support for this paradigm. Notable among them are Edwards (1989), Fischer (1993), Khan and Knight (1981,1985), and World Bank (1993).
Polak (1991) highlights the broad acceptance of the paradigm in countries with programs and discusses constraints program design.
The findings of this review on the response of investment and output to Fund-supported policy changes have already set in motion further study within the IMF.
Of the 36 countries reviewed, 15 were repeat users, 6 were using IMF resources for the second time in recent years, and 15 were using IMF resources for the first time in recent years. Countries in this last group were effectively just beginning their adjustment processes, while for the first two groups, the programs under review represented a continuation of previous adjustment efforts.
The broad strategies for fiscal adjustment and their implementation are explored in Bennett, Carkovic, and Dicks-Mireaux (1995) in Background Papers.
These correlations reflect two influences: a reversion to mean tendency following a year when spending was unusually high or revenues unusually low; and policy changes geared toward correcting a longer-standing source of a fiscal imbalance.
Financial market liberalization and the accompanying developments in those markets are assessed in Bennett (1995) in Background Papers.
The differences in overall adjustment strategies and macroeconomic performance between these two groups are examined in Mecagni (1995) in Background Papers.
See Savastano (1995) in Background Papers for an examination of private saving behavior in detail.
For example, Serven and Solimano (1994).
The same finding emerged in the review of programs supported by the enhanced structural adjustment facility: in gen eral, the more vigorous the adjustment of financial policies, the stronger the growth of output. An important question, however, is the degree to which both policies and outcomes were influenced by common exogenous factors, such as an improvement in the terms of trade. See Schadler, Rozwadowski, Tiwari, and Robinson (1993).
