5 Many Borrowers; How Many Sizes? IMF Lending and Program Design
Author:
Mr. James M. Boughton
Search for other papers by Mr. James M. Boughton in
Current site
Google Scholar
PubMedClose
Language:
English
Download PDF (370.3 KB)

Abstract

This volume--the fifth in a series of histories of the International Monetary Fund--examines the 1990s, a tumultuous decade in which the IMF faced difficult challenges and took on new and expanded roles. Among these were assisting countries that had long operated under central planning to manage transitions toward market economies, helping countries in financial crisis after sudden loss of support from private financial markets, adapting surveillance to reflect the growing acceptance of international standards for economic and financial policies, helping low-income countries grow and begin to eradicate poverty while staying within its mandate as a monetary institution, and providing adequate financial assistance to members in an age of limited official resources. The IMF's successes and setbacks in facing these challenges provide valuable lessons for an uncertain future.

There is no one-size-fits-all in this business. You have to understand what is going on in each individual country; otherwise, you simply end up getting it wrong.

Stefan Ingves1

Director, IMF Monetary and Exchange Affairs Department

September 25, 1999

From the beginning of IMF lending in 1947, the staff understood that countries could have difficulty financing their balance of payments for two reasons: either because their economic policies were inadequate or because of circumstances beyond their control. with experience, though, the Fund realized a basic lesson. Even if the root cause of the problem was an external shock, such as a weak market for the country’s exports, policies would have to be adjusted to compensate unless the shock was temporary and the gap could be covered by a repayable loan. Because few shocks could be judged reliably to be “temporary,” the Fund increasingly conditioned its lending on policy adjustments. This confidence in and reliance on conditionality reached its zenith in the 1990s.

Of course, to paraphrase Leo Tolstoy, every unhappy country is unhappy in its own way. Every country that asks to borrow from the IMF faces a unique set of circumstances that requires a tailored response. The policy conditions and the mix of adjustment and financing must reflect the context. Throughout this decade, the Fund made a number of changes to its policies on conditionality, and it introduced new specialized lending windows in response to changing world conditions. This chapter reviews the overall pattern of Fund lending, policy advice, and technical assistance in the 1990s and the various changes that the Fund introduced.

Overview

When the IMF first began lending to its member countries in 1947, it made simple outright disbursements in response to requests.2 Five years later, it began the practice of offering stand-by arrangements, under which it would make a standing commitment to disburse up to a specified amount during a specified period, conditional on the country maintaining acceptable exchange rate and macroeconomic policies. Both the outright credits and the stand-by arrangements were also conditional on a finding that the borrower had a “balance of payments need” for them. Starting in 1963, the Fund established a variety of “special facilities” designed to respond to specific types of balance of payments problems, such as export crop failures or weak export markets. In 1977, it introduced the practice of lending on concessional terms to low-income countries through separate trust funds. As the range of lending practices widened, lending became an increasingly important activity for the Fund. In the 1990s, 100 countries borrowed from the IMF at least once.3

The Ebb and Flow of Lending by the IMF

The episodic nature of IMF lending is striking (Figure 5.1). The onset of a financial crisis induces a sudden spike in lending, which then tapers off quickly once the crisis is over. Before the 1990s, the major episodes were associated with the Suez crisis of 1956 (point A in Figure 5.1), the crises of confidence in the U.K. pound sterling in the 1960s (points B and C), the collapse of the official gold market in the late 1960s (point D), the “oil shock” of 1973–74 (point E), and the Latin American debt crisis that began in 1982 (point F). This episodic pattern continued in the decade covered by this History (and in the decade that followed).

Figure 5.1.
Figure 5.1.

IMF Lending, 1947–2000

(In percentage of Fund quotas)

Percent

Source: IMF financial accounts.Note: See text for meaning of labels at peaks. Figure includes all purchases except those in the reserve tranche, plus loans from administered trust funds.

For the first half of the 1990s, the Fund was actively lending to a large number of countries. The aftermath of the 1980s debt crisis, the influx of new members with acute financial shortages, and the strengthening of the Enhanced Structural Adjustment Facility (ESAF) as a vehicle for lending to low-income countries all made for heavy demand. Nonetheless, the amounts involved were not large by historical comparison.4 On average, annual lending from 1990 through 1994 was about $8.5 billion (SDR 6.1 billion), with a total of 94 borrowing countries. The Mexican peso crisis at the end of 1994 (point H in Figure 5.1) ushered in a new era of much larger lending, followed as it was by the East Asian financial crisis of 1997–98, the Russian meltdown of 1998, and the Brazilian exchange crisis of 1998–99 (point I). For the second half of the decade, the Fund had just 75 active borrowers, but it lent an annual average of $20.7 billion (SDR 14.7 billion).

The mainstay of this lending was stand-by arrangements, which accounted for more than 40 percent of the total volume (Figure 5.2 and the Appendix to this chapter). Extended arrangements (longer-term and generally larger arrangements under the terms of the Extended Fund Facility, or EFF) accounted for another 22 percent. The rest was spread across several special facilities and other lending policies, which are discussed in more detail in the next section. At times, one or another of these special facilities rose to the fore owing to its suitability for the occasion. Thus, the EFF accounted for 62 percent of the volume in 1990, with the Fund using it to help emerging-market countries obtain debt relief from commercial bank creditors. The Compensatory and Contingency Financing Facility (CCFF), amended to help countries affected by the Gulf War, accounted for 36 percent of 1991 lending. The Systemic Transformation Facility (STF) accounted for a third of the volume in 1994 as a number of new Fund members used it in the early stages of transition to market economies. And the Supplemental Reserve Facility (SRF) accounted for 44 percent of lending in 1998, being the preferred way for the international community to put enough money on the table for the countries most severely affected by the systemic financial breakdown that year.

Figure 5.2.
Figure 5.2.

IMF Lending by Facility, 1990–99

SDR billions

Source: IMF financial accounts.Note: SAF = Structural Adjustment Facility; ESAF = Enhanced SAF; PRGF = Poverty Reduction and Growth Facility.

The increased incidence of prolonged borrowing also featured in this decade. Although the Fund was intended to be a purveyor of temporary financial assistance, prolonged usage became a fact of life in the 1980s and 1990s. During those 20 years, some 70 countries became prolonged users for a substantial portion of the time.5 This phenomenon seems to contravene the spirit of the Articles of Agreement, which state in Article I that a purpose of the IMF is to “give confidence to members by making the general resources of the Fund temporarily [emphasis added] available to them under adequate safeguards.”6 In most cases, however, the Fund extended credit with an expectation on both sides that the borrower would strengthen its policies so it could repay the loan on schedule. Policy slips, adverse shocks during the life of the arrangement, and poor program design all contributed at various times to delays that required the Fund to lend repeatedly, sometimes for decades on end. In effect, this repeated lending rescheduled all or part of the original loans, conditional on the borrower agreeing to a new set of policy reforms. In almost all cases, the loans were eventually repaid in full and on the original terms. (The exceptional cases, in which arrears piled up and remained outstanding, are discussed in Chapter 16).

In all, 52 countries were prolonged users of Fund resources at some point in the 1990s, about the same number as in the 1980s (48 countries). In an important sense, though, the scope of the problem diminished in this decade. The difference between the two periods was that in the 1980s, all but 4 of the 48 countries had prolonged recourse to the Fund’s General Resources Account (GRA) through repeated stand-by or extended arrangements. In the 1990s, the phenomenon was much more concentrated in the ESAF, a trust fund better designed to accommodate prolonged borrowing and not covered by the “temporarily” restriction cited above. Only 22 countries were prolonged users of GRA resources in the 1990s (Figure 5.3).

Figure 5.3.
Figure 5.3.

Prolonged Users of Fund Resources, 1980–99

Number of countries

Source: IMF financial accounts and staff reports.Note: A prolonged user is defined here as one with five or more annual programs in the preceding 10 years and with outstanding credit of at least 100 percent of quota at the end of the year indicated. Outstanding credit excludes charges on overdue obligations to the Fund.

Within the group of prolonged users of the Fund’s general resources were instances in which the Fund recognized at the outset that it would have to engage in repeated lending before the country would be able to graduate. Perhaps the clearest example was Bulgaria, which asked to borrow from the Fund soon after it became a member in 1990. The country’s economy had been battered by decades of inefficient central planning and then by the collapse of the Soviet-based trading system (see Chapter 9). The increase in oil prices associated with the Gulf War additionally justified its borrowing from the IMF, but the most serious problems were chronic and would take several years to overcome.

A staff team headed by Anoop Singh (Advisor, European Department) made three trips to Sofia in as many months, at the end of which the team recommended a CCFF loan, to be followed almost immediately by a stand-by arrangement. In making that recommendation, Singh warned the Executive Board that the authorities were already talking about asking for a second arrangement when the time came. “Bulgaria,” he concluded, “may be expected to be a prolonged user of Fund resources in the coming years.”7 That alarmed several Executive Directors, who observed that it was inappropriate for the Fund to lend to a country that lacked the capacity to repay the Fund when the loan fell due. Yet, everyone knew that Bulgaria could not recover and become a productive part of the world economy without the financial support of the IMF. The cause was too important, and the risk was worth taking.8 The demands of the moment overwhelmed any doubts about whether the Fund was the appropriate agency to undertake that risk.

Bulgaria did become a prolonged user of Fund resources, with six stand-by arrangements and one extended arrangement in effect almost continuously through 2004. As expected, the economic reform process was long and difficult and not without hiccups. In the end, though, it succeeded. At the close of one more stand-by arrangement that the authorities successfully treated as precautionary, Bulgaria fully repaid its borrowings with a final balloon payment in April 2007.

Jordan also borrowed through a series of almost continuous arrangements for 15 years (1989–2004). Shortly after Jordan graduated from this relationship, the Independent Evaluation Office prepared a detailed assessment.9 That report concluded that the Fund had made “an important contribution” to Jordan’s recovery from terrible initial conditions. Starting from a “severe balance of payments crisis” and “deep-rooted macroeconomic and related structural problems,” Jordan made “major progress over the 15-year period of its IMF program involvement” (Independent Evaluation Office, 2005a, pp. 8, 12, and 13). The report was critical of what the Independent Evaluation Office regarded as the Fund’s excessive focus on short-term issues and lack of clarity in its analysis of the relationship between program design and longer-run goals. Greater emphasis on technical assistance aimed at building institutions and fostering domestic ownership of reforms might have led to more effective and earlier results.

The staff did not disagree with these conclusions, but it noted the relevance of “the external and political factors influencing the Fund’s decision to assist Jordan,” including “the link between Fund arrangements and Jordan’s requests for Paris Club agreements” to reschedule debts to official creditors (Independent Evaluation Office, 2005a, p. 84). Paris Club creditors would not reschedule a country’s debt-service payments in the absence of a Fund agreement, and those creditors constituted a majority on the Fund’s Executive Board. Prolonged usage, in this and other cases, may have resulted in part from external pressure to contribute to a broad international assistance effort.

The Evolution of Conditionality

Although the IMF normally lends subject to a set of conditions on the borrowing country’s economic policies, nothing in the Articles of Agreement requires it to do so. Indeed, John Maynard Keynes—the British economist who was one of the IMF’s two main “founding fathers”—wanted an institution that would lend almost automatically on demand. The Fund’s other main founder, the American Harry Dexter White, was more skeptical of the practicality of that plan, with the result that the Articles were silent on the role of policy conditionality. Eventually, conditional lending prevailed after the Fund began lending to countries that could not resolve their payments difficulties without major changes in their policies.

In some circumstances, lending without policy conditions is clearly more appropriate. If a country faces a balance of payments deficit because of a temporary decline in the value of its exports owing to a worsening of world market conditions, it may need financial assistance to weather the downturn, but it does not necessarily need to change its policies. The Compensatory Financing Facility (CFF) was created precisely to accommodate such situations. Similarly, if deteriorating world financial markets temporarily limit a country’s ability to raise funds privately, it may be appropriate for the IMF to lend quickly and without requiring a change in policies. Establishing ground rules for such lending has, however, proved to be extremely difficult, as discussed in the next section. Especially starting in the late 1970s, most IMF lending was conditioned on a specific list of economic policy adjustments by the borrowing country.10

The main story about conditionality in the 1990s is that the Fund gradually shed its reluctance to impose conditions on structural—as opposed to macroeconomic—policies. Throughout its first three decades of lending, the Fund eschewed structural policy conditions altogether except in very unusual situations. To give force to that principle, the Executive Board adopted guidelines in 1979 that included this paragraph:

9. The number and content of performance criteria may vary because of the diversity of problems and institutional arrangements of members. Performance criteria will be limited to those that are necessary to evaluate implementation of the program with a view to ensuring the achievement of its objectives. Performance criteria will normally be confined to (i) macroeconomic variables, and (ii) those necessary to implement specific provisions of the Articles or policies adopted under them. Performance criteria may relate to other variables only in exceptional cases when they are essential for the effectiveness of the member’s program because of their macroeconomic impact.11

That decision was interpreted fairly strictly in the 1980s, with structural policy conditions limited to countries with central planning or heavy government intervention in the economy; and to developing countries implementing multiyear programs supported by the EFF, the Structural Adjustment Facility (SAF), or the ESAF. For most stand-by arrangements, performance criteria—the quantitative conditions on which disbursements under the arrangement depend—were limited to ceilings on credit expansion, government deficits, and borrowing; and prohibitions on current account exchange restrictions and on arrears to external creditors. Although the breach in the prohibition began to widen in the late 1980s, the practice remained unusual. Out of 41 stand-by arrangements in effect from 1987 to 1990, only 11 contained any structural policy conditions, and those averaged fewer than two such conditions per program year (Figure 5.4).

Figure 5.4.
Figure 5.4.

Structural Conditionality in Stand-By Arrangements

Source: IMF Monitoring of Fund Arrangements (MONA) database.Note: Number of conditions is per year.

Staff, management, and Executive Directors all struggled to find the right balance on this delicate issue. If the Fund restricted its conditionality to macroeconomic variables, it ran the risk that the selection of policies—for example, what types of spending to reduce or which taxes to raise if a cut in the fiscal deficit was required—would be driven by domestic political considerations with short-run benefits but negative implications for sustainable economic growth. If the Fund imposed conditions on those structural policies, it ran the risk that it would undercut national ownership of key policy decisions and weaken the capacity of both the government and civil society to take responsibility for economic performance. Because the optimum strategy might well vary greatly across countries and over time, establishing and adhering to consistent guidelines was not easy.

The Fund did not revise its guidelines on conditionality during the 1990s, but it did revise its practices extensively. Without making a formal decision, the Fund expanded its use of four existing techniques. First, it supplemented its standard quantitative performance criteria with an increasing number of “structural performance criteria” such as ceilings on types of public sector borrowing or the extension of credit to types of borrowers. Second, it gradually added more structural “benchmarks” to stand-by and other arrangements. These benchmarks, such as progress indicators for ongoing reform programs, differed from performance criteria in that they often were not quantitative, and a failure to meet a benchmark did not automatically disqualify the borrower from the right to make the next drawing on the arrangement.12

Third, the Fund had increasing recourse to “prior actions,” which the authorities would have to complete before management would take the request for an arrangement or completion of a program review to the Executive Board for approval. The use of prior actions was first introduced as a way to get essential measures taken in a timely manner when it was impractical to make them into performance criteria, the classic example being a currency devaluation. In the 1990s, the Fund increasingly imposed a range of prior action requirements to get countries to establish track records and demonstrate the ability and the political will to implement reforms.13

Fourth, the Fund broadened the scope of program reviews. Originally, the Fund introduced reviews into multiyear arrangements so that it could delay setting performance criteria for each year until it had at least preliminary results for the previous year. By the 1990s, the initial performance criteria often were set for less than a year, so that more frequent reviews were necessary. In addition, reviews commonly covered the authorities’ success at meeting structural benchmarks as well as performance criteria.

These various extensions added flexibility and realism to Fund-supported programs, but they also removed much of the predictability of the Fund’s financial commitment. Borrowers could no longer have confidence that if they met all the performance criteria, they would be entitled to borrow the next tranche of the arrangement. They would also have to meet a critical mass of the benchmarks and satisfy the Fund that the program was qualitatively on track—judgmental decisions that the Fund would make at the time of a review, not before.

Several forces combined to pull the Fund more deeply into structural conditionality in the 1990s, in ways that paralleled the expansion of surveillance concerns detailed in the preceding chapter. First, the Fund had explicitly accepted that promoting economic growth, not just financial stability, was a major goal of its lending. As the staff put it in 1987, a “powerful argument for conditionality to include growth as a direct objective is that without such an approach medium-term viability (and the revolving character of Fund resources) may be elusive.”14 Second, the Fund was becoming increasingly sensitive to the need to avoid letting the effects of financial adjustment fall disproportionately on poor people. Reducing fiscal deficits and controlling inflation could be, and should be, beneficial to the poor, but the effects depended on the structure of the adjustments. The government might want to maintain the size of its bureaucracy, defend subsidies for money-losing state enterprises, or preserve benefits for the urban upper classes. To enhance the distributional effects of adjustment in such cases, the Fund would have to insist on structural reforms. Third, many of the Fund’s borrowers in the 1990s were making a difficult transition from central planning to market economies. Without structural reforms, that transition could not succeed.

The effect of these combined forces was that structural conditionality progressed from the exception to the norm. As shown in Figure 5.4, by the second half of the 1990s almost 90 percent of all stand-by arrangements had structural conditions, and those arrangements had an average of about 13 such conditions per program year. For extended arrangements and those supported by the ESAF, the numbers were even higher. Those facilities generally required borrowers to undertake structural reforms, and virtually all such arrangements in the late 1990s had structural conditions. On average, the number of conditions per program year was similar, regardless of the type of arrangement.

This expansion of conditionality posed unprecedented challenges for the Fund. The basic rationale for conditionality is that it can serve as a commitment mechanism for national authorities. Without conditionality, they might otherwise have difficulty persuading domestic interest groups to accept policies with overall benefits but net losses for those groups; or they might have difficulty getting electorates to support policies with long-run benefits but short-run costs.15 Conditionality is likely to be less successful when it is used to overcome the inability or unwillingness of the authorities to formulate and carry out strong policies. In practice, the difference between these two circumstances often becomes indistinct, especially when “the authorities” are not a homogeneous body.

These challenges came to a head when the Fund was called upon to help Thailand, Indonesia, and the Republic of Korea overcome a regional financial crisis in 1997 (see Chapter 11). In each of the three countries, the government in power at the time of the crisis resisted making the policy adjustments requested by the Fund as conditions for its lending. In each case, some officials, advisors, or opposition leaders were more sympathetic to the need for reform. In Indonesia, government advisors were instrumental in designing the reforms but lacked the power to get them carried out. In all three countries, domestic power struggles ensued, new governments were elected, and successful reforms were eventually implemented.

By the end of the 1990s, the need for strong domestic ownership of reforms was becoming clearer, as was the need for restraint on the part of the IMF. Structural policy conditions could be an important component of the Fund’s tool kit, but only if the conditions were designed with care and focus and were negotiated in partnership with the national authorities (see Goldstein, 2001, and references therein; and IMF, 2002). With that conclusion in mind, the Fund undertook a major overhaul of its conditionality and adopted new guidelines in 2002 aimed at focusing and streamlining its practices. At the end of that decade, it seems safe to conclude that the 1990s were the high water mark for the scope of conditionality.

New Special Facilities

If the only choices available to the IMF were to lend through a stand-by arrangement with all the usual policy and other conditions or to lend in response to shocks without negotiating policy adjustments, the second choice would be open-ended and irresistible. To reduce temptation as well as to tailor lending to country circumstances, the Fund created a web of special facilities designed for specified conditions when an ordinary stand-by arrangement would be inappropriate or insufficient. Whenever a new type of potential circumstance arose, it was necessary to establish a new facility or modify an existing one. This practice had the advantage of enabling the Fund to maintain control while responding to shocks with a measure of flexibility, but it occasionally produced results that were not needed or that were ill designed for their intended purpose.

Two new special facilities were created and used successfully by the Fund during the 1990s: the Systemic Transformation Facility (STF), which was created in 1993 as a temporary facility for assisting countries in transition, and the Supplemental Reserve Facility (SRF), which was created in 1997 as a means of providing large-scale financing to countries facing capital account crises. Two other innovations, the Contingent Credit Line (CCL) and the Y2K Facility, were created in 1999 but never used.16

The Systemic Transformation Facility

The IMF faced a major challenge at the beginning of 1993. It had taken in 22 new member countries in the preceding year. Almost all of them were undertaking a transition from state ownership and control to private enterprise, market-based pricing, and openness to international trade and finance. Most had little or no foreign exchange, little or no experience with market prices, limited administrative capacity to set and collect taxes, no national currency or experience with monetary control, and a production structure geared toward bilateral trade within a command system that no longer existed. They also faced a Catch-22—without the ability to stabilize and reform, few of them could be expected to formulate or implement a program of policies that would warrant the support of the international community on normal terms. Without that financial support, the transition would be seriously delayed and possibly aborted altogether because each country would face a collapse of output and incomes and, consequently, social unrest and backlash against reform.

At the time, the Fund’s only choices for helping these countries financially were a first credit tranche drawing or a highly risky stand-by arrangement supported by an economic reform program that would almost certainly fail. As the experience with the Russian Federation and other transition countries in 1992 had shown all too clearly, neither choice would spur reform. The first choice would provide too little financing, and the second would likely result eventually in unsustainable and unredeemable debts. A new facility was needed that would support the early stages of reform: a program aimed at stemming the output losses, stabilizing the economy after the initial price adjustments, and buying enough time for the authorities to work with the Fund and other agencies to develop a comprehensive strategy and plan.

Once the need for a new facility became apparent, the Fund responded quickly. The staff put together a general proposal in March 1993, and the Executive Board considered it in April. A preliminary discussion revealed widespread support, and on that basis the staff prepared a specific proposal—the STF—which the Board approved on April 23.17 The central element was an understanding that policy conditionality would be lighter and much less detailed than under a conventional stand-by arrangement. Each applicant would be expected to submit a letter stating its policy intentions for the next year, including steps “to put in place the basic institutions of economic management in a market-oriented system.” The letter would also spell out macroeconomic policy intentions and would commit the authorities to work toward a “comprehensive adjustment program that could be supported by a Fund [stand-by] arrangement.”18

In addition to lighter policy conditions than in stand-by arrangements, the STF also offered borrowers a longer time to repay the loans (10 years instead of 5). Access limits were relatively small: each eligible country could borrow up to 50 percent of quota in two equal tranches. This borrowing would not count toward the general access limits, but it would preclude a “first credit tranche” loan with no conditionality. Interest rates were the same as for a stand-by arrangement.

The decision establishing the STF allowed for the possibility that the Fund might simultaneously—or even earlier—approve a stand-by arrangement for a country borrowing from the new facility. In that case, the Fund’s goal in using the STF would be to front-load, top up, and extend the arrangement. This feature turned out to be important; nearly half of the countries using the STF had stand-by arrangements approved along with their first STF drawing (Table 5.1).

Table 5.1.

STF Loans and Associated Arrangements, 1993–95

article image
Source: IMF financial accounts. Note: n.a. = Not applicable; ESAF = Enhanced Structural Adjustment Facility; SBA = stand-by arrangement.

Arrangements were associated with the second STF drawing.

A large part of the rationale for developing the STF so quickly was the Fund’s eagerness to put together a strong financing package for the Kyrgyz Republic. As discussed in Chapter 9, the Kyrgyz authorities were about to launch their own currency and back it up with a stabilization program. The Fund had previously declined to approve financing until the government had a credible monetary policy ready, and now it needed to act aggressively to ensure the success of the new currency. On May 12, 1993, just three weeks after the establishment of the STF—and just two days after the authorities introduced the som as the national currency—the Fund approved a stand-by arrangement and a simultaneous STF drawing. The combination meant that the Kyrgyz Republic could draw 40 percent of quota immediately (SDR 25.8 million, equivalent to $37 million), enough to finance the first stage of an effective reform program.

The STF as a whole was scheduled to expire in 20 months, at the end of 1994. By then, the Fund expected that the initial shock of the transition process would have worn off, and most of the eligible countries would be ready for conventional stand-by arrangements. However, heavy demand for STF loans continued, and a few eligible countries took longer than expected to put basic institutional and policy structures in place.19 In June 1994, as the deadline for making STF drawings approached, Managing Director Michel Camdessus tried to persuade the Executive Board to extend the facility, increase access from 50 percent of quota to 85 percent, and allow countries to make as many as five consecutive drawings instead of the two already allowed.20 A number of Board members, led by Ewen Waterman (Australia), were skeptical about either increasing access or allowing multiple tranches, and those ideas were ultimately dropped. However, the Board eventually agreed to extend the deadline for a first STF drawing to end-April 1995, to extend the maximum gap between a first and a second drawing from one year to 18 months, and to allow second drawings to be made up to the end of 1995.21

The establishment or extension of special facilities such as the STF required the approval of Executive Directors holding at least 85 percent of the voting power.22 In April 1995, when the Board considered whether to extend the life of the STF any further, the consensus broke down. Karin Lissakers (United States) asked that the facility be extended and augmented. A majority of Directors either supported her or were willing to go along with the proposal, but others were skeptical. After some deliberation, Camdessus concluded that the facility had served its purpose well but was no longer needed. As a temporary facility, it should be allowed to expire. Several Directors then sided with the Managing Director. Hachiro Mesaki (Japan), for example, stressed the importance of strong conditionality for promoting economic reform. Prolonged reliance on the STF would weaken the effectiveness of Fund support, so the time had come “to draw the curtain and bid [the STF] a warm farewell.” Most of the Board had no strong preference on the matter. Even Krzysztof Link (Alternate, Poland), whose constituency included several STF borrowers from the Caucasus region and central Asia, noted the value of the facility but concluded that its expiration “would not create great harm to any country.” Support for an extension fell well short of the 85 percent threshold, and the motion failed.23

The last two countries to qualify for STF drawings were Uzbekistan (in January 1995) and Azerbaijan (in April). Shortly before the facility expired for good, both countries made their second drawings and simultaneously obtained their first stand-by arrangements.

In all, of the 20 countries that borrowed through the STF, 16 made two equal drawings, each one amounting to 25 percent of quota. Of the remainder, two low-income countries, Cambodia and Vietnam, shifted to concessional borrowing from the ESAF.24 As discussed in Chapter 6, Bulgaria was out of compliance with the terms of its standby arrangement in 1995. Romania’s financing needs were satisfied by its stand-by arrangement without the need for a top-up from the STF.

Eight of the 20 borrowers had stand-by arrangements approved along with the initial STF drawing, and eight others had stand-by arrangements approved along with the second drawing. Two countries—Belarus and Russia—had stand-by arrangements approved several months after the second drawing, and the remaining two—Cambodia and Vietnam—had ESAF arrangements associated with their single drawing on the STF. As shown in Table 5.1, two other low-income transition countries—Albania and Mongolia—did not draw on the STF but obtained ESAF arrangements whose amounts were later augmented to take into account the adverse effects of the transition process.

Emergency Financing and the Supplemental Reserve Facility

In 1994, as the possibility of a sudden reversal of capital inflows to developing countries emerged as an imminent threat, the Fund took a fresh look at its options for providing quick-disbursing assistance to affected countries. As detailed in Chapter 9, financial capital had been flowing freely into many emerging markets for more than four years. The longer that inflow continued, the greater the potential for imbalances, increasing the likelihood of a reversal. In response, the staff prepared a proposal for a short-term financing facility aimed at helping countries facing large and presumably temporary outflows of capital. The Fund would assess each country’s policies as usual through annual Article IV consultations. If, despite satisfactory policies, the economy got hit by outflows anyway, the Fund would be prepared to lend quickly and possibly even automatically in amounts adequate to stem the speculative pressure. In support of the proposal, the staff’s paper noted that the Czech Republic, Mexico, and Sweden had all faced such situations in the early 1990s and could have benefited from such a facility.25

Antecedents

The 1994 proposal was not the first of its type. In May 1972, as part of an initial staff proposal for systemic reform during the breakdown of the Bretton Woods par value system, Jacques J. Polak (Economic Counsellor) prepared a note for the Board on “possible Fund financing of short-term capital movements.” For some countries, such capital flows were expected to be large relative to Fund quotas, provoking the fear that existing credit facilities (in the Fund or elsewhere) would be inadequate to cover the potential imbalances. Polak raised the question of “whether reserve creation by the Fund should in some way be responsive to the needs for reserves caused by the ebbs and flows of capital movements.” If so, he asked, “should the facility in principle be available to all members … or perhaps [to] those Article VIII members that met certain additional criteria?” The first question, however, was whether the reformed system would be anchored by par values, and if not, whether fluctuations in exchange rates would absorb the pressure for large capital movements. That discussion occupied the Executive Board and the Committee of Twenty for two years, by which time everyone seems simply to have forgotten the Polak proposal.26

In 1979, the staff considered introducing “reverse stand-by arrangements.” In that scheme, if the staff team conducting Article IV discussions with a member country considered the country’s economic conditions strong and its policies sound, the Managing Director could offer a stand-by arrangement as a “seal of approval” instead of waiting for countries to develop problems and then ask for a borrowing arrangement as a last resort.27 A year later, Alexandre Kafka (Brazil) floated the idea during the Executive Board’s periodic discussion of its work program, but it gained no traction and was quietly dropped.28

Polak tried again in January 1980, when he was serving as Advisor to the Managing Director (Jacques de Larosière). In a memorandum, he proposed establishment of a “temporary intermediation facility” that would provide low-conditionality loans quickly to countries facing sudden losses of access to international bank loans. In his view, it would be a natural extension of the Oil Facilities, in view of the similarities between the effects of the oil price increases of 1973–74 and the interest rate increases of 1979–80. De Larosière forcefully rejected the idea on the grounds that Fund lending should be backed up by policy conditionality and not extended automatically.29

In January 1981, the European Department toyed with the idea of providing large-scale short-term lending without a stand-by arrangement. No special facility would be created. Instead, the Board would be asked to waive the normal conditionality requirements on a case-by-case basis for countries deemed to qualify because of their track records and current policies. This proposal was floated in the context of an exchange-market crisis in Sweden. The Fund would disburse close to 150 percent of quota up-front, with an expectation of either an early repayment or a rollover into a stand-by arrangement if necessary. De Larosière rejected that proposal on essentially the same grounds as the previous one from Polak.

Only once again in the 1980s did a proposal reach the Executive Board, and then only in the most general terms. In the immediate aftermath of the Mexican debt crisis of 1982, the U.S. authorities asked for consideration of new special procedures to deal with systemic strains. In response, the staff prepared a paper assessing “the adequacy of existing arrangements to deal with major strains in the international financial system.”

On the role of the Fund, the paper noted that “under … present policies, there are no special facilities that members can use in the event of a financial emergency arising from sudden, severe, and widespread strains in the international financial system.” It posed the question of whether a new facility might be warranted and whether SDR allocations could help finance it, but it did not elaborate a specific scheme.30

When the Executive Board met to discuss the 1982 paper, the idea became known as the “contingency fund” proposal. Directors generally agreed that it was better to prevent crises than to clean up the mess afterward, but a majority nonetheless balked at the idea of weakening conditionality. Gerhard Laske (Germany) suggested that such a scheme might be interpreted as a “safety net for the banks,” which would be counterproductive and possibly destabilizing. Summing up the meeting, de Larosière noted that “a number of Directors considered that, on the whole, it might be inappropriate for the Fund to engage in short-term bridging financing in view of the risk of impairing the effectiveness of the Fund’s adjustment programs.”31

The Emergency Financing Mechanism

In sum, all these suggestions failed to win support because they were seen to be too soft on conditionality. The 1994 proposal suffered the same result when considered by the Board. Waterman opened the discussion by suggesting it was “unlikely that a country with strong fundamentals and a track record of sound policies would be cut off from borrowing from international capital markets. … In practice, the greatest demand could come from the countries whose economic fundamentals are not sound and whose track record is not so good.” That confidence in the wisdom and rationality of capital markets and in the general need for conditionality carried the day. Ian D. Clark (Canada) expressed a liking for the idea in principle, but he seriously doubted that the Fund could outsmart the markets in assessing the quality of a country’s policies and circumstances. Several speakers also worried about the Fund’s ability to judge whether a shock was temporary or would persist.32

Camdessus cast the discussion in the best possible light (“some [Directors] were cautiously optimistic; others were skeptical but receptive”) and promised to continue working to devise a viable proposal. Three weeks later, Mexico—one of the three countries the staff had singled out as a possible beneficiary of a quick-disbursing facility—exploded in a frightening financial crisis that exposed underlying weaknesses the Fund had failed to see (Chapter 10). For the time being, that crisis ended further consideration of the proposal.

Notwithstanding the daunting challenges, the Fund still needed some means of reacting quickly and forcefully when a member country came under a speculative attack. The Mexican crisis showed that although the Fund might not be able to see a crisis coming, it had to—and could—respond quickly once the crisis erupted. To enhance that response, the heads of state and government of the Group of Seven (G7) countries, meeting in Halifax, Nova Scotia (Canada), for their annual summit in June 1995, proposed establishing an “Emergency Financing Mechanism involving a Fund arrangement with strong conditionality but with high up-front access and faster procedures to access Fund resources in crisis situations under the ‘ exceptional circumstances’ clause.”33

A staff team in the Policy Development and Review Department quickly prepared a proposal in response to the Halifax summit request, which was being pushed particularly forcefully by the United States. In the staff conception, the trigger for activating special procedures under the Emergency Financing Mechanism (EFM) would be a finding by the Fund that a country faced an exceptional crisis and was prepared to take strong policy actions to deal with it. In such “rare circumstances,” the Fund would accelerate its reaction, negotiate a program as quickly as possible, prepare only a short report for the Executive Board, and at least consider granting exceptionally large access in relation to the member’s quota. Additional financing from bilateral creditors and other multilateral lenders would be expected as a normal part of the international package of support.34

Two features made the EFM more palatable than the rejected schemes of the past. First, it did not rely on the staff’s ability to foresee a crisis or to assess the quality of a country’s policies in advance. Second, it did not rely on a distinction between external and internal causes of the crisis. As long as a country was willing to work with the Fund to deal with the problem, it could qualify for large and speedy assistance.

The Executive Board approved the EFM in September 1995, on the understanding that it was not a formal new lending facility but only a new set of procedures for coping with extreme financial crises. To that end, the Board issued no formal decision. The procedures, which in essence just described the way the Fund had responded to the Mexican crisis, were specified in the Chairman’s Summing Up of the Board meeting.35 The rationale for this low-key approach was that many on the Board were wary of raising expectations that the Fund would just “bail out” any country that might get into major financial trouble. Marc-Antoine Autheman (France) went so far as to argue that the EFM should not be used “more than two or three times a century.” While not everyone wanted to be that strict, almost all Directors (with the exception of Karin Lissakers of the United States) agreed that its use should be limited to “truly exceptional” and thus rare circumstances.36

The agreement on the EFM allowed the Fund to approve exceptionally large loans, but it did not require a link between the emergency procedures and exceptional access. In fact, the first invocation of the procedures did not involve exceptional access. In July 1997, the Philippines was hit by contagion from the financial crisis in Thailand just as it was preparing to exit from prolonged dependence on IMF financial assistance (Chapter 12). The crisis erupted on July 2; the central bank abandoned the peg between the Philippine peso and the U.S. dollar and allowed the peso to float on July 11; a staff report was issued to Executive Directors on July 14; and the Executive Board approved an augmentation and extension of the existing extended arrangement on July 18 (see Chapter 12). Although the peso continued to depreciate for some time, the panic subsided. The Philippines was thus spared a crisis of the extent and magnitude of those that engulfed much of the region that year.

Before the end of 1997, the Fund invoked the EFM procedures three more times: for Thailand in August (21 days from the authorities’ request for assistance to the Fund’s approval of a stand-by arrangement), for Indonesia in November (23 days), and for Korea in December (8 days). In July 1998, although the Fund did not specifically invoke the EFM as justification, the Executive Board greatly augmented Russia’s extended arrangement just 10 days after the authorities asked for help. In each of these four cases, the arrangement included exceptionally large access.37 Autheman’s “two or three times a century” was already exceeded, as the world economy staggered from a series of crises that no one could have foreseen.38

The Supplemental Reserve Facility

As the Fund began providing large-scale financial support—loans that were well in excess of the normal quota-based ceilings—to countries facing massive outflows of private capital, attention turned again to the question of the way those arrangements should be structured. Just topping up and front-loading stand-by and extended arrangements on the same terms as their smaller cousins was not satisfactory. It failed to provide adequate safeguards to the Fund, and it raised the risk of moral hazard if vulnerable countries and their commercial creditors came to expect a bailout on relatively easy terms. Nonetheless, the Mexican and Asian crises showed clearly that large-scale assistance might continue to be needed.

In the wake of the 1997 crisis in Thailand, Japanese officials proposed establishing an Asian Monetary Fund, independent from the IMF, that would use regional resources and avoid the stigma many countries associated with borrowing from the IMF. As discussed in Chapter 12, that proposal did not gain much support, but it did lead to an international agreement to establish a peer-group surveillance process known as the Manila Framework. The communiqué for the meeting that set up the Manila Framework also called on the IMF to set up a mechanism to provide large-scale assistance on a more consistent and formal basis. Specifically, the finance and central bank deputies of 14 countries “urged the IMF to constructively examine the establishment of a new mechanism to provide short-term financing to augment an exceptional stand-by or extended arrangement in the light of the globalization of financial markets and the increased scale of private capital flows.”39

The staff responded to this request with a proposal quite different from the “short-term financing facility” that was not approved in 1994. The new proposal dropped the idea of trying to sort out strong and weak economies in advance, and it dropped the idea of offering low conditionality to the strong ones. It built on the concept of the EFM by noting the importance of a rapid and large-scale response, and it added two new elements: relatively high interest rates and incentives for early repayment. Assistance would be provided through a new or existing arrangement with the same policy conditions that would apply in normal circumstances.

The driving assumption behind the 1997 proposal for what was initially called the “Manila facility” was that a country hit by a sudden loss of access to private financial capital would have both short-term and longer-term financing needs. The longer-term or “underlying” need would be associated with a balance of payments deficit similar to other situations that would lead countries to ask for loans from the Fund. That sort of demand could be met in the usual way, with the amount of the arrangement linked to the country’s quota through the established access limits, as discussed in Chapter 15. The short-term need would be associated with the initial capital outflow and might be large relative to the size of the economy and thus out of proportion to the country’s quota. If the Fund-supported program worked as expected, private capital inflows would eventually resume, and this additional financing could then be repaid. The Mexican crisis of 1994–95 had shown this pattern, and the staff saw that it was reappearing in Korea and was likely to develop in Thailand and Indonesia as well.40

While the staff was formulating its response to the request from the Manila Framework, crisis erupted in Korea. The authorities asked for help, and—as noted above—the Fund responded extremely quickly by activating the EFM procedures. On December 3, Camdessus flew to Seoul to complete the negotiations. He reported back to the Executive Board via videoconference, later the same day, that the authorities agreed to shift a major portion of their borrowings to a new short-term facility once the Fund was able to enact it. With that informal understanding, on December 4—just two weeks after the Manila conference—the Executive Board approved a massive stand-by arrangement equivalent to just under 20 times Korea’s quota (Chapter 11). The next day, the staff hastily put the finishing touches on its proposal for a Supplemental Reserve Facility (SRF) and circulated it to Executive Directors.

The staff carefully designed the SRF to maximize the safeguards. Everyone involved understood fully the extent of the risk to the Fund of lending extremely large amounts to troubled countries without being able to take the time to negotiate a completely articulated program of economic reforms. The risk of creating a moral hazard if creditors believed that the Fund would always rescue them from bad lending or investing decisions also weighed on everyone’s minds. When the Executive Board met to consider the proposal in mid-December, just when doubts about the effectiveness of the Korea program peaked, most of the discussion concerned means of making the safeguard provisions even stronger. The staff duly produced a revision, which the Board unanimously approved on December 17.41

The decision creating the new facility included several provisions aimed at coping with risks and minimizing moral hazard.42 First, use of the SRF would be limited to cases posing systemic threats. Second, the borrowing country would be expected to try to persuade all creditors, “both official and private,” to maintain their exposure, and the Fund and other official creditors would assist in that effort. “All options should be considered to ensure appropriate burden sharing.” As discussed in Chapter 1, that feature effectively restored “private sector involvement” as an essential element in this type of rescue effort. Third, if that persuasion effort proved insufficient, the Fund could require the country to impose controls to stem capital outflows. Fourth, the borrower would be expected to repay the loan within two and a half years, compared with the five-year maturities for ordinary stand-by arrangements and ten years for extended arrangements. Fifth, to cover the additional risk and encourage prompt repayment, the interest rate charged on SRF loans would start at 3 percentage points above the standard rate and rise to as much as 5 points above once a loan was outstanding for two and a half years.

On December 18, 1997, one day after establishing the SRF, the Fund approved the second installment of the stand-by arrangement for Korea. From this point on, disbursements would be made under the terms of the SRF rather than the ordinary credit tranches. Korea borrowed approximately $2.1 billion on December 19 and a total of $13.5 billion (SDR 9.95 billion) through the SRF by the end of 1998. Korea repaid those drawings in installments from December 1998 through September 1999. The initial drawings on the stand-by arrangement were repaid in 2001.

The Fund activated the SRF twice more in the 1990s. In July 1998, Russia requested and received a large augmentation to its EFF arrangement plus a CCFF drawing in its last desperate attempt to stave off a default on its government debt (Chapter 7). The package totaled $11.2 billion, of which $4.8 billion was to be made available immediately. Most of the immediately available funds were provided through the CCFF ($2.9 billion) or the EFF on normal terms ($1.0 billion). The remainder ($0.9 billion) was provided under the terms of the SRF. All of it was financed by activating the General Arrangements to Borrow.43 Then in December 1998, the Fund approved a stand-by arrangement for Brazil that was primarily on SRF terms (Chapter 12). Brazil borrowed close to $9 billion through the SRF linked to that stand-by arrangement: $4 billion in December 1998 and $4.9 billion in April 1999.

The SRF was activated several more times in the next four years.44 It was terminated in 2009 as part of a general restructuring of Fund facilities.

Contingent Credit Lines

Establishing the SRF as a high-conditionality facility did not mean the IMF was giving up on the idea of low-conditionality lending. When the Russian default in August 1998 began to threaten other emerging markets around the globe, decisive and coordinated action appeared to be needed to forestall that threat. As discussed in Chapter 12, Camdessus called a special meeting of finance ministers throughout the Americas, which was held at Fund headquarters at the beginning of September. That meeting produced a general agreement that many of the Latin American emerging-market countries had greatly strengthened their economic policies and conditions in recent years but still faced a serious risk of a reversal of support from international financial markets. In a communiqué summarizing the meeting, Camdessus noted that the Fund was already providing financial assistance to many of these countries, and he was “ready to recommend the strengthening and broadening of this support, if necessary.” That was a veiled reference to the possibility that the Fund might make an advance commitment to a select group of countries judged to be vulnerable despite the implementation of sound economic policies.45

The strongest impetus for a preemptive lending facility came from U.S. President Bill Clinton. Shortly after the ministerial conference at the Fund, in a speech to the Council on Foreign Relations in New York on September 14, 1998, Clinton called for the Fund to take more direct action to ward off crises of contagion before they could destroy otherwise healthy economies:

We must develop policies so that countries can reap the benefits of free-flowing capital in a way that is safe and sustainable. We must adapt the IMF so that it can more effectively confront the new types of financial crises, minimizing their frequency, severity, and human cost. We need to consider ways to extend emergency financing when countries are battling crises of confidence due to world financial distress as distinct from their own errors in policy. We must find ways to tap the energy of global markets without sentencing the world to a cycle of continued extreme crises.46

Clinton’s speech effectively put the 1994 proposal for a preemptive short-term financing facility back on the table. When the G7 finance ministers met at the beginning of October, they “agreed to explore a strengthened capacity, based in the IMF …, to provide more effectively contingent finance to help countries pursuing sound policies to maintain stability in the face of difficult global financial conditions.” Although the tortured syntax of that sentence suggested that the ministers had argued a bit over how far to run with the idea, it was enough to convince the Interim Committee to put it on the agenda for the Fund’s work program.47

The intended beneficiary of the proposed contingency facility was Brazil. As soon as the Russian government defaulted on part of its debt in August 1998, international financial markets concluded that Brazil was the most likely market to collapse next (Chapter 12). IMF officials were pressing Brazil to tighten fiscal policies and to allow the exchange rate to depreciate more, but they had confidence that the government would take the necessary measures. Because the Clinton administration shared that view, the prospects were good that the Fund could offer large-scale preemptive financing to Brazil on a precautionary or contingent basis. The question was not whether but how to frame the offer. A conventional stand-by arrangement, even if it was announced as precautionary, would be risky, because it would inflame the political opposition in Brazil and could further ignite the already intense speculative pressure in financial markets. In the fall of 1998, however, that was the only option available to the Fund. As negotiations started in October, Stanley Fischer (First Deputy Managing Director) promised Executive Directors that if the IMF created what was then being called the Contingent Reserve Facility, Fund support for Brazil would be shifted to it.48

On October 30, when newspapers around the world were reporting that Brazil was desperately trying to reach an agreement with the IMF, the G7 raised the ante by issuing a statement in the name of the “leaders” (heads of state or government), welcoming both Brazil’s new policy commitments and the plans for new “financing arrangements to ward off destabilizing market contagion.” The G7 leaders’ statement included a specific description of what they wanted: “an enhanced IMF facility to provide a precautionary line of credit that could be drawn on if needed by countries pursuing strong IMF approved policies, accompanied as appropriate by bilateral finance, on a case by case basis, and with appropriate private sector involvement.”49

The staff visualized modifying the SRF to include the possibility of making an advance commitment to select countries for a Contingent Credit Line (CCL). A public commitment from the Fund would help countries with good policies avert a financial attack, and SRF resources would serve as a backup if a crisis hit anyway and if normal quota-based resources were too small to cope with it. The first challenge, therefore, was the same one that had doomed earlier attempts. Could the Fund reliably and convincingly separate the wheat from the chaff? The staff report proposing the facility acknowledged the problem but confined its doubts to a footnote: “In practice it would be difficult to draw such a clear line. … Nonetheless, it would be important to confine the use of a contingent credit line to members for which … policy shortcomings can confidently be judged unlikely to be the trigger for the crisis.”50 The momentum coming from the G7 was too strong for more pronounced skepticism.

Executive Directors from outside the G7 were less constrained by the political winds.51 A. Shakour Shaalan (Egypt) argued that existing facilities, notably Fund-monitored programs and precautionary stand-by arrangements, were perfectly adequate. J. Onno de Beaufort Wijnholds (Netherlands) scoffed that the CCL was “being created for a clientele which does not exist (i.e., ‘ innocent’ countries)” and would “undermine the Fund’s credibility and catalytic role.” Zhixiang Zhang (China) suggested that countries with strong policies would be unlikely to ask the Fund for assistance in advance of a crisis; in any case, the “health checks” the staff was proposing to separate strong from weak economies were still untested. Willy Kiekens (Belgium) concluded that a positive assessment by the Fund—a “credible assurance” of sound policies and strong economic conditions—would do more good for a country than a contingent promise of future loans. Even the Russian Executive Director, Aleksei V. Mozhin, acknowledged that “the countries that have been hit by the crisis did not have spotless records, to put it mildly.”52

In light of these controversies, the staff put three options up for consideration. First, the Fund could agree to monitor a member country’s policies, with an explicit commitment to consider offering a stand-by arrangement of a specified amount if the country were hit by a contagion-fueled crisis (the “Fund-monitored program” option). Second, the Fund could approve a “low-access precautionary arrangement,” with a promise to consider augmenting it with large-scale resources if needed (the “augmentation” option). Third, the Fund could approve a large-scale arrangement, with the full amount already committed and to be made available in the event of a crisis (the “commitment” option). The first two options would provide the greatest protection for the Fund but would provide relatively little assurance to the country that help would be available when needed. Only the commitment option would provide the desired assurances, but it would leave the Fund exposed to relatively high risk.53

Eventually, management and the staff decided on a watered-down version of the commitment option. The Fund would approve an arrangement with a commitment to the full amount the country might need in a crisis, but disbursement “would be subject to a review and decision by the Executive Board, taking into account the circumstances of the member, the member’s policy response to those circumstances, and the member’s track record under the program specified in consultation with the Fund.” Once it was activated, the CCL would be subject to “appropriate conditionality (performance criteria and reviews).”54

Because an applicant would have no assured access to the resources committed under a CCL, the main advantage would be the Fund’s public approval of its policies. In the final formulation, to be considered for a CCL, a member country would have to satisfy four criteria. First, the Fund would determine that the country would not need to borrow from the Fund unless it was hit by financial contagion. Second, the country would have to have a favorable assessment from the most recent Article IV consultation and a continuing positive assessment when the Board considered the request for a CCL. Third, the applicant should be “maintaining constructive relations with its private creditors” so as to limit its vulnerability and to “facilitate the appropriate involvement of the private sector” in the event of a crisis. Fourth, the country should put in place a “satisfactory economic and financial program” and submit it for approval by the Fund.55 If the Fund had any credibility with financial markets, then a country meeting these criteria should be assured of continuing support from the markets, and the CCL should serve only as a second line of defense against a widespread market meltdown.

This formulation was carefully crafted to meet two conflicting objectives: maximize the potential benefits by making large amounts of money available to stricken countries and minimize the risk by limiting access to the most-deserving countries. The discussion in the Executive Board, which took place over several days of meetings through the first four months of 1999, showed that everyone was keenly aware of the difficulty of making the new facility useful without weakening the Fund’s control over the use of its financial resources. The crisis in Brazil was already essentially resolved, with the approval of a conventional stand-by arrangement in December 1998 and a successful revision of the economic program in February 1999. Which other countries might apply for and benefit from a CCL was not clear and was not openly discussed. It was thus with an apparent sense of weariness and limited enthusiasm that Executive Directors established the CCL as a separate set of procedures within the SRF,56 just in time for the Interim Committee to welcome it at its meeting in April 1999.

From the vantage point of the IMF, the CCL was a device for certifying that certain countries had “first class” policies and no need for major adjustments. The combination of that certification and the (conditional) promise of large-scale financing if private markets pulled out was expected to play a key role in preventing crises from occurring or spreading. Fischer (2000) trumpeted it as “a watershed in the role of IMF lending” and expected “several countries” to avail themselves of it in the near future. For a time, speculation centered on Argentina as a prime candidate, but Mexico, South Africa, Korea, and other countries also were mentioned as possibilities.57 Within the Fund, Estonia—much admired for its handling of the transition—was also considered a prime candidate.

From the vantage point of countries that the IMF encouraged to apply, the CCL looked more like a way of admitting that they might be vulnerable to a speculative attack. If only one country had a CCL, its financial markets might feel as if they had targets painted on their front doors. If two countries had approved credit lines and one came under attack, the other could be dragged into the fray by association. If the Fund approved a country’s application and then later decided that a weakening of the country’s economic policies disqualified it, the termination could have devastating consequences for market access. The contingent nature of the Fund’s commitment and the fact that activation would be accompanied by policy conditions that would be specified only at that time further weakened the incentives for applying.

Although the Fund was fully aware of these difficulties when it established the facility, the terms were as favorable as the Executive Board could tolerate in 1999. They were not favorable enough. Over the next four years, no country applied for a CCL, despite a concerted effort by the Fund to inform officials of several countries that an application would be favorably received. The CCL was terminated in November 2003 without ever having been used.58

Fear of the Millennium: The Y2K Facility

Anyone coming of age in the twenty-first century will find this story hard to conjure, but in the waning years of the twentieth a great many people seriously feared that the millennium would usher in a meltdown of modern technology. The origin of this concern was ridiculously plain. When the computerization of financial records began in the 1950s, the prevailing programming technology relied on instructions encoded on 80-column cardboard punch cards. Because efficient programming required stringent economizing on the use of these columns, years were universally recorded in two-digit form, that is, without the century. Even though punch cards were phased out in the 1970s, and all of the successor programming technologies were free of the need for this streamlining of numbers, the habit persisted well into the 1990s. Consequently, as the millennium approached, a massive reprogramming effort was required if computers were to be capable of recognizing that 2000 (a year dubbed “Y2K”) was the successor to 1999 and not 99 years before it.

The Fund began internal preparations in 1997 to ensure that its own computer-based operations were Y2K-compliant. It completed that process without incident before May 1999, when its financial year 2000 began.59 In the fall of 1998, the Fund turned its attention more toward member countries, out of concern that many of them—particularly those with seriously outmoded computer systems and limited capacity to replace them—might face major breakdowns when the millennium turned. Most directly, it feared that some countries might not be able to carry out transactions with the Fund or with other bilateral or multilateral agencies because of technological breakdowns. Even domestic payments systems might break down if computers were unable to recognize transaction dates correctly. More generally, even if the technology was fixed, savers and investors might retain the fear of a breakdown. Savers could shift massively out of banks and into cash; large institutions might shift investments out of smaller banks; and low-income and other developing countries could face capital flight.

In September 1998, Camdessus asked the staff to raise Y2K issues in every consultation, regardless of whether the context was surveillance, lending, or technical assistance, to ensure that all countries were adequately prepared. During the next two years, almost every staff report contained at least one reference to the issue. To reach a wider and higher-level audience, the Fund held a seminar on the subject at the 1998 IMF/World Bank Annual Meetings in Washington. The word was thus put out that upgrading all technological systems before the end of 1999 was critically important. The remaining question was whether the Fund could do anything more directly to support the international effort.

A number of countries asked the Fund’s technology service—the Bureau of Computing Services—to provide technical assistance on computer modernization. The Fund declined to do so on the grounds that the bureau was not equipped to assist with implementation (as opposed to system design) problems.60 Instead, in September 1999, with encouragement from the U.S. authorities and a few other major shareholders, Camdessus proposed establishing a special lending facility for countries experiencing balance of payments difficulties as a result of Y2K-related problems. The Executive Board acted quickly to pass the measure in time for the Interim Committee to endorse it in its September 26 communiqué. Few objections were expressed, except that Gregory Taylor (Australia) complained that the staff had failed to explain the way in which a Y2K-related balance of payments problem would be distinguished from one arising from other causes. Without that, it was not clear how eligibility for the facility would be determined. Thomas Leddy (Deputy Director, Policy Development and Review Department) insisted that because the nature of the shocks that might arise was unknowable, it was necessary to stay flexible. Unsatisfied, Taylor abstained from approving the decision.61

The new “Year 2000 Facility” was a channel for the IMF to lend up to 50 percent of a country’s quota (or more under “exceptional circumstances”) as an outright purchase (as opposed to a series of drawings under a stand-by arrangement).62 The only policy condition would be a general test of cooperation with the Fund. As with the SRF, a borrower would have to pay a surcharge amounting to 3 percentage points above the basic interest rate on stand-by arrangements, and more if the drawing was outstanding for more than six months. Drawings under the Y2K Facility would not reduce a borrower’s access to drawings under other Fund facilities.

January 1, 2000, arrived on a Saturday, first in the western Pacific islands and New Zealand, and then it rolled westward hour by hour as the earth spun leisurely on its axis. At IMF headquarters in Washington, a small army of technical staff stayed at their desks throughout the weekend, ready for whatever calamity might arise. Financial institutions, governments, and corporations around the world conducted similar death watches. Nothing happened. To almost everyone’s amazement, three years of careful preparation in some 200 countries and countless private enterprises had squashed the “millennium bug.” Everywhere, financial records remained intact, and transactions occurred smoothly.

In the end, no country requested to use the Y2K Facility. On March 31, 2000, any possible need for it having passed, it expired as scheduled.

Use and Modification of Earlier Special Facilities

Three special lending facilities existed at the beginning of the 1990s: the Compensatory and Contingency Financing Facility, the Buffer Stock Financing Facility, and the Extended Fund Facility. Each one was controversial in its own fashion.

Compensatory and Contingency Lending

The first special lending facility established in the IMF was the Compensatory Financing Facility (CFF), created in 1963. It aimed to help developing countries that were heavily dependent on exporting primary commodities, the prices of which were highly unstable in world markets. Beginning with a $60 million loan in June 1963 to compensate Brazil for a drop in the value of coffee exports, the Fund used the CFF sporadically but in important ways for about 20 years, with large bursts of lending whenever commodity prices were generally weak. The Fund then decided, in effect, that it should usually grant compensatory credit only in conjunction with a stand-by or extended arrangement. That 1983 decision substantially tightened the standards for CFF lending, and usage declined. The expansion of the facility in 1981 to include a window to compensate for the increased cost of cereal imports provided some additional scope, as did the addition of a contingency element in 1988. The latter change resulted in the renaming of the facility as the Compensatory and Contingency Financing Facility (CCFF).63

In the 1990s, the Fund approved 42 CCFF drawings by 27 countries (Table 5.2). Almost half of those drawings (18) resulted from the temporary establishment of a window to help countries cope with the surge in oil prices after the 1991 Gulf War. After that option expired in 1992, usage dropped off sharply. The main constraint continued to be a conviction by major creditors that “in almost all cases in which members face arguably temporary [balance of payments] financing needs that are met by the CCFF, they also face pressing adjustment needs which are best met through conditional upper credit tranche support.”64 That meant that the Fund was very unlikely to approve the use of the facility for countries without a Fund-supported policy program with upper-tranche conditionality. Instead, the Fund typically approved CCFF credits to augment and accelerate disbursements under stand-by, extended, and (more rarely) SAF or ESAF arrangements.

Table 5.2.

CCFF Drawings, 1990–99

(Millions of SDRs)

article image
article image
Source: IMF financial accounts and staff reports. Note: EFF = Extended Fund Facility; ESAF = Enhanced Structural Adjustment Facility; SAF = Structural Adjustment Facility; SBA = Stand-by arrangement.

A zero in this column indicates that an External Contingency Mechanism loan was approved but not activated.

The ESAF arrangement for FYR Macedonia was inactive at the time of the CCFF drawing.

The Paucity of Stand-Alone Loans

On just five occasions in this decade, the Fund approved requests for CCFF drawings from countries that had no new or ongoing program relationship with the Fund. The first of these occasions was the most controversial because it gave rise to a highly unusual injection of politics into the considerations.

In March 1992, Israel requested a $245 million (SDR 178.6 million) loan to compensate for export shortfalls related to the Gulf War. On reviewing the request, the staff concluded that the war had substantially and temporarily depleted Israel’s export revenues and raised the cost of its oil imports. By the time of the request, those disruptions had passed. Meanwhile, since the end of 1989, Israel had been absorbing a massive influx of immigrants, including some 350,000 from the former Soviet Union. That influx was causing a sustained balance of payments problem and was making it difficult for Israel to meet its targets for foreign exchange reserves. The two problems together gave rise to a balance of payments need that warranted financial support from the Fund. Because Israel’s macroeconomic policies were considered to be sound, a CCFF drawing seemed more appropriate than a conditional stand-by arrangement.65

When the Executive Board met to consider Israel’s request, three Directors—from the Islamic Republic of Iran, Libya, and Saudi Arabia—strongly objected. Mohamed Finaish (Libya) led the attack, noting that the real motivation was to cope with immigration, not the war that was already over. To accommodate the new immigrants, Israel was expanding its settlements in the Occupied Territories of the West Bank and Gaza. Finaish, supported by Muhammad Al-Jasser (Saudi Arabia) and Mohamed Ali Hammoudi (Iran, Senior Advisor to Abbas Mirakhor), noted that these settlements were illegal under international law and contravened a resolution of the United Nations (UN) Security Council. Those three constituencies, which held a total of 9.8 percent of the voting power, voted to deny the request. Other Directors, however, decried this overt injection of political considerations into its deliberations. E.A. (Ted) Evans (Australia) questioned the appropriateness of using the CCFF to assist with a medium-term problem such as Israel’s surge in immigration, but he dissociated himself from the comments of Finaish and others and did not oppose the loan.66

The next year, the Fund approved three stand-alone CCFF drawings, for Moldova, Ghana, and South Africa. In each case, the borrowing country was undertaking difficult political transitions.

Moldova—a country of the former Soviet Union—joined the Fund in August 1992 and set about to establish basic economic institutions and develop a macroeconomic policy program in consultation with the Fund and other international agencies. Even to prepare the ground for an STF loan was going to take a full year (see Chapter 8). In the meantime, as a down payment on the support that it expected to provide over time, the Fund approved a CCFF loan for $18.5 million (SDR 13.5 million) in February 1993.

Ghana had just established a multiparty democracy, and the 1992 elections had been won by the country’s long-standing military ruler, Jerry John Rawlings. The previous government had successfully implemented a four-year program supported by the Fund through the ESAF (1988–92), and in 1993 the authorities were trying to maintain the same policy approach under difficult circumstances. The Fund and the authorities both considered Ghana ready to “graduate” from reliance on IMF financing. Accordingly, the Fund was monitoring performance through its “enhanced surveillance” procedures—the first time it had applied these procedures to an ESAF-eligible country—with the objective of encouraging domestic saving and other private sector financing. In the second half of 1992, Ghana suffered not only from the policy slips often associated with a transition to democracy, but also from a weak world market for cocoa (Ghana’s principal export crop) and a weather-induced slump in export volume. The Fund responded by approving a $65 million (SDR 47 million) CCFF loan in July 1993 and continuing its enhanced surveillance.67

As discussed in Chapter 14, in 1990 South Africa began a historic transition to democratic majority rule. In 1993, as the date for elections approached, the transitional government held discussions with the IMF about financial support. The Fund was prepared to offer a stand-by arrangement, but the authorities were convinced that accepting the Fund’s policy conditions was undesirable and could be politically suicidal. As a compromise, the authorities requested, and the Fund approved, a CCFF loan of $850 million (SDR 614.4 million) to compensate for the effects of a drought and a weak global market for gold.

After 1993, the Fund approved stand-alone CCFF drawings on only two occasions.68 Again, these special cases were marked by political transitions that made normal program implementation and support difficult.

In 1995, Rwanda was emerging from one of the most horrific internecine wars of the twentieth century. As its national unity government began to formulate a recovery plan, it entered into talks with the IMF regarding a program to be supported with ESAF loans. As an early gesture toward that end, the Fund approved a $13 million (SDR 8.9 million) CCFF loan in November 1995, justified by the temporary interruption of international trade during the war (see Chapter 14).

Algeria received the final stand-alone drawing in this period. An exporter of oil and gas, Algeria had borrowed from the Fund from 1994 to 1998 through a stand-by and then an extended arrangement, both of which were completed successfully (Chapter 14). In April 1999, the appointed government attempted to move toward democracy through multiparty elections for president, but the effort was thwarted when six of the seven candidates withdrew to protest alleged irregularities in the process. After the election, to help the economy through a troubled period and to smooth out and prolong the repayment schedule on Algeria’s outstanding obligations, the Fund approved a CCFF loan for $300 million (SDR 223.5 million). That turned out to be Algeria’s last borrowing from the Fund for at least the next decade.

Oil Imports and the 1991 Gulf War

The invasion of Kuwait by Iraq in August 1990 touched off a series of events, including the Gulf War of 1991, that disrupted the world oil market and caused a major spike in oil prices. For many of the affected countries, the Fund could respond by augmenting or accelerating stand-by or other arrangements. At the time, nearly 50 countries had active borrowing arrangements with the Fund, an unusually high number by previous standards. For other countries likely to see the cost of their imports rise, Camdessus proposed modifying the CCFF by adding a window relating specifically to oil imports. This device would give the Fund the flexibility to provide low-conditionality and fast-disbursing loans to at least 30 highly vulnerable countries that would not otherwise need to borrow.69

Although the Interim Committee seemed to endorse the idea in its September communiqué, the Fund staff was skeptical. The staff paper circulated after the Interim Committee meeting explained how an import “element” might work but declined to recommend it. Flexibility in handling requests for regular borrowing arrangements could take care of most cases; the logic of singling out one component of the balance of payments was weak; and it would be difficult to distinguish a transitory spike in prices from a shift in the trend, as required by the CCFF rules.70 That reluctance drew a rebuke from Thomas C. Dawson II (United States), who argued strongly for the proposal at a Board meeting on November 2. The Dean of the Board, Alexandre Kafka (Brazil), also considered the introduction of an oil-import window to be an “essential” component of “an effective Fund response” to the crisis. Although some Directors shared the staff’s skepticism, most were at least open to the proposal, so it was kept alive by the political momentum.71

Because of the highly technical structure of the CCFF, detailed amendments had to be drafted to maintain consistency. For the oil-import element, that meant specifying ways for the access limits and associated policy conditions to be applied in various circumstances, computation of the excess cost of imports, recomputation of the initial loan if the cost of oil turned out to be different from projections, and so forth. In just a few weeks, the staff managed to draft an amendment to the CCFF that included 45 paragraphs and subparagraphs covering all conceivable circumstances.72 The Executive Board approved this and other amendments on December 5, 1990. At the same time, acting under UN resolutions, the United States and other allied countries were preparing to go to war to free Kuwait from the Iraqi occupation. By the time the war began, the Fund stood ready to offer loans through these new procedures.

During the 18 months the oil window was open, 12 countries borrowed a total of $3.4 billion (SDR 2.5 billion) through it.73 All of them had ongoing program relationships with the Fund at the time—in most cases, stand-by arrangements (Table 5.2). None of the countries the Fund had expected to be the prime beneficiaries—those that otherwise would not need to borrow—chose to use it. The temporary window may have been a useful vehicle for the countries that did use it, to the extent that it enabled the Fund to augment its lending quickly and without renegotiating underlying programs. It also may have helped calm financial markets by demonstrating the Fund’s commitment to managing the crisis produced by the Gulf War. It would, however, be hard to argue that it was as successful as its advocates had hoped. Other aspects of the Fund’s crisis-management strategy—temporarily suspending some access limits, rephasing disbursements under existing arrangements to bring them forward, and increasing the financing available through the ESAF—were more generally applicable and more clearly beneficial.74

Cereal Imports

The “cereals window” of the CFF was established in 1981 in response to international calls for a “food facility” to help developing countries cope with unusually high costs of importing food. It was designed to provide a means of compensating countries for a broader range of commodity-related balance of payments problems and to increase the access limits for countries with multiple problems. However, the requirement that it could be activated only when a country had a net shortfall in its balance of payments limited its applicability. That is, if a country had excess costs of importing cereals such as wheat or maize at the same time it had a temporary increase in the value of its exports, it normally would not be eligible to draw on the cereals window.75

Target countries used the cereals window only lightly during the first five years of its life, but that turned out to be its heyday. From 1981 to 1986, seven countries borrowed a total of some $560 million (just over SDR 500 million) under this decision, which was about 6 percent of total CFF drawings during that period.76 Through the end of 1989, just two more countries availed themselves of it. Both of those drawings were made in conjunction with the activation of conditional lending arrangements.

In the 1990s, four countries—Algeria, Bulgaria, Moldova, and South Africa—drew on the cereals window, on six occasions (Table 5.2). Three of these drawings compensated for the effects of drought on domestic crop production. The other three compensated for other problems that required the country to increase its imports sharply and temporarily.

Bulgaria made the final drawing on the cereals window in April 1997, while the country was in the middle of a political and economic crisis, as discussed in Chapter 6. The Bulgarian economy was on the brink of hyperinflation, and a collapse of production was leading to severe food shortages. The Fund approved a stand-by arrangement, supplemented by a CCFF loan. The initial drawing on the stand-by arrangement was only SDR 23.2 million, but it was accompanied by a CCFF drawing of SDR 107.6 million. (The total drawing was the equivalent of approximately $180 million, or 28 percent of quota.) Those amounts helped to alleviate the immediate problems and gave the authorities time to establish momentum for their reform program.

Contingencies

In August 1988, the Fund established an External Contingency Mechanism (ECM) within the CFF and renamed the facility the CCFF. As explained in Boughton (2001, pp. 737–42), the mechanics were extremely complex, but the essential feature was a procedure to supplement or alter a stand-by or other Fund arrangement in the event of adverse shocks. Suppose, for example, that a stand-by arrangement assumed that the price of the borrower’s principal export commodity would remain constant throughout the life of the arrangement. Separately from the arrangement, the Fund could approve an ECM under which the Fund would offer an additional loan if export prices weakened. Other covered situations included unanticipated increases in export volumes, import prices, or interest rates on external debt. Although such contingencies could be covered by appropriate modifications to the arrangement itself—either a contingency clause in the original arrangement or a revision after the fact—the ECM was more palatable to a majority of Executive Directors because it was thought to be less likely to lead to a profusion of requests that would be hard for the Fund to contain or manage.

The popularity of the ECM was limited. Neither the Fund nor most of its borrowers found it to be helpful, simply because approving waivers or modifying programs as circumstances warranted was much easier and offered more flexibility. From the borrower’s perspective, the most troubling feature was the “symmetry provision” of the ECM, which provided that the allowable drawings under the associated arrangement could be reduced in the event of a favorable shock. From 1988 through 1992, the ECM was used in association with 11 programs, but it was activated only once (Table 5.2).77 As the staff acknowledged at the time, the procedure had “proven to be largely unworkable in practice.”78 After 1992, it was not used at all.

Even the one activated case was odd, in a way that illustrates the frustrations of this restrictive and complicated facility. In February 1991, the Fund made its first loan to Bulgaria (which had just become a member a few months earlier) in the form of a CCFF loan to compensate for excess oil-import costs associated with the Gulf War. As discussed in Chapter 9, that loan was followed a few weeks later by a stand-by arrangement, which was fully drawn. In March 1992, as the staff completed the terms for a second stand-by arrangement, it determined that it had overestimated the size of the temporary oil excess and that Bulgaria had not been eligible for the initial CCFF loan at all.

The problem in the Bulgarian case was not that oil prices were too low. The problem was that oil prices had remained high, so the excess was not temporary enough. Under the Fund’s rules, that required Bulgaria to repay the original loan. Because the country’s external financing needs were huge at this early stage of the transition to a market economy, no one wanted to force such a repayment. Fortunately, the 1991 stand-by arrangement included a contingency element that committed the Fund to make a new CCFF loan in the event of a further or more sustained increase in the price of imported oil. To resolve the matter, Bulgaria repaid the original loan (SDR 60.6 million) at the same time that the Fund activated the ECM (SDR 56.9 million). The two amounts were nearly offsetting, and the new stand-by arrangement, approved in April, took care of the difference and preserved the momentum of the reform program.

The End of the CCFF

In the first quarter of 2000, the Fund made one last attempt to simplify the CCFF to make it more transparent and easier to use. The staff recommended, and the Executive Board agreed, to eliminate the ECM, which clearly was not serving its intended purpose. That move restored the name to the original Compensatory Financing Facility (CFF). The Board also accepted staff recommendations to retain the export shortfall and cereals windows but to simplify the access limits. After a lengthy debate that echoed earlier battles going back to the 1960s, the Board formalized the understanding that the CFF should be used only in association with a conditional stand-by or other Fund arrangement, except in the “rare” case in which the country’s balance of payments was otherwise satisfactory.79

Despite all of the Fund’s efforts to fine-tune the facility, no country requested to draw on it after 1999. When the Fund undertook to simplify its system of special facilities in March 2009, it eliminated the CFF altogether.

Buffer Stock Financing Facility

The second special facility created by the Fund was the Buffer Stock Financing Facility (BSFF), established in 1969. At that time, several international buffer stock agreements were in force requiring members of those agreements to purchase specified primary commodities when market prices would otherwise be cyclically depressed. For example, from 1956 to 1985, major tin exporters entered into a series of formal arrangements to maintain buffer stocks and thereby try to stabilize the world price of tin. If a country represented to the IMF that it faced difficulty financing its balance of payments owing to the financing requirements of an approved commodity agreement, the Fund could lend to it through the BSFF without the normal policy conditions of a stand-by arrangement. From 1971 to 1984, the Fund lent in this way to 18 countries participating in buffer stock arrangements for rubber, sugar, and tin.80

By the 1980s, the efficacy of buffer stock agreements for stabilizing commodity prices was being called increasingly into question. Agreements on cocoa, coffee, sugar, and tin all expired in the 1980s (see Gilbert, 1996; and Cashin, Liang, and McDermott, 2000). That left only the International Natural Rubber Agreement (INRA) as a basis for IMF lending through the BSFF. In April 1990, the Fund approved the 1987 INRA as a successor to the 1979 INRA and its eligibility for buffer stock financing. The staff indicated that they did not expect any participating country to use the BSFF, but thought it helpful to retain it as a backup in case of need.

No country requested a loan from the BSFF after 1984. When the 1987 INRA expired in 1993, no eligible price stabilization schemes remained in force. A new rubber agreement took effect in 1997, and two years later its officials asked the Fund to consider approving it. The Fund declined to do so and terminated the BSFF in February 2000.81

The Extended Fund Facility

The Fund established the EFF in 1974 as a substitute for an ordinary stand-by arrangement for developing countries needing time and assistance to implement wide-ranging structural reform programs. The EFF offered both higher access limits, initially up to 140 percent of quota rather than 75 percent, and longer repayment terms, initially eight-year maturities rather than five. To qualify, a country had to submit a Letter of Intent specifying a policy program covering at least three years and including adequate structural reforms. These terms and relationships were revised over time to make the EFF more attractive to borrowers, for example, by extending the repayment schedule to 10 years, starting in 1979. In 1979–83, the Fund approved 24 EFF arrangements totaling nearly $25 billion (SDR 21.3 billion), which equated to more than half the total value of the stand-by and extended arrangements approved in those five years. The Fund then grew wary of making longer-term and relatively large commitments, and the facility was little used for the next five years (three arrangements totaling SDR 1.2 billion).82

The renaissance of the EFF began in May 1989, when the IMF agreed to support the Brady Plan to reduce the external debts of certain heavily indebted developing countries (Chapter 9). One of the requirements for a country seeking debt reduction was to submit a program worthy of EFF financing. The Fund then would approve an EFF arrangement that included provisions for augmenting or setting aside specified amounts for debt relief once the country had negotiated a Brady deal with its commercial bank creditors. That led quickly to three large EFF arrangements—for Mexico, the Philippines, and Venezuela—in 1989, with a total commitment (including later augmentations) of more than $10 billion (SDR 8.25 billion).

The EFF remained in place throughout the 1990s and proved to be popular both with the Fund and with its middle-income borrowers.83 (Low-income countries had access to the ESAF on similar but much less expensive terms.) From May 1989 through end-1999, the Fund entered into 36 extended arrangements with 27 countries (Table 5.3). Several arrangements supported Brady deals, and several others assisted heavily indebted countries recovering from the 1980s debt crisis in other ways.84

Table 5.3.

Extended Arrangements in Effect, 1990–99

article image
Source: IMF financial accounts.

Final approved amount of the arrangement. In a number of cases, the original amount was subsequently augmented or reduced.

Although a number of EFF arrangements were large, either absolutely or in proportion to the borrower’s quota, size was not their defining characteristic. In most of the cases in which large-scale IMF lending was part of the solution to a major financial crisis, including Mexico in 1995; Thailand, Indonesia, and Korea in 1997; and Brazil in 1998, the Fund’s support took the form of stand-by arrangements on scales made possible by invoking the exceptional circumstances clause (see Chapter 15). The rationale in those cases was that the country was coping with a liquidity crisis and had a reasonable expectation of regaining market access and repaying the Fund quickly (as in fact happened). The EFF was distinguished from other arrangements by an expectation that the country faced deep-seated structural problems that would not be resolved in a year or two, even if initially triggered by a financial crisis similar to other crises. Thus, the large arrangements for Argentina in 1992 and 1998, Russia in 1996, and Indonesia in 1998 were formulated for the EFF rather than as ordinary stand-by arrangements.

As with stand-by arrangements, EFF arrangements occasionally were conceived as “precautionary,” meaning that the authorities stated their intention not to draw on the arrangement. During the 1990s, 12 EFF arrangements were treated as precautionary for at least part of the time they were in effect, but only three remained so throughout their lives.85 In 2000, the Executive Board accepted a staff recommendation to avoid the use of precautionary EFF arrangements in the future. Experience had shown that in most cases, if a country had a longer-term structural imbalance warranting an EFF commitment, it would also need to borrow to finance its balance of payments.

Of all the special lending facilities in effect for the IMF’s general resources during or before the 1990s, only the EFF survived the streamlining exercise of 2009. Special policies (without formal facilities) remained for lending in certain circumstances such as postconflict situations or the aftermath of natural disasters. The evolution of those policies in the 1990s is the subject of the next section.

Other Specialized Lending Programs

In a few instances, the IMF created special lending policies without creating a formal facility. In these cases, the Fund would apply the usual rules for lending “in the credit tranches” without a stand-by arrangement, but it would establish separate criteria for the circumstances under which it would lend. At the outset of the 1990s, the only such policy in effect was lending in response to natural disasters. In the course of the decade, the Fund approved lending for two more special situations: recovery from war or other conflicts and currency stabilization following a period of high inflation.

Natural Disaster Relief

Beginning with a loan to Egypt in 1962 that responded to disastrous crop failures, the Fund occasionally bypassed its normal negotiation of policy reforms when a natural disaster hit a member country. That policy was formalized in 1982 with the adoption of guidelines specifying the range of applicable circumstances, the amounts that the Fund would generally be willing to lend, and other terms. The Fund decided that it would quickly lend small amounts (normally 25 percent of quota) when a natural disaster would otherwise result in a “serious depletion of … external reserves.” In lieu of a negotiated policy program spelled out in a Letter of Intent, the applicant would be required only “to describe the general policies it plans to pursue, including its intention to avoid introducing or intensifying exchange and trade restrictions.” These requirements reflected those that the Fund applied to requests for drawings in the first credit tranche. Rather than a new policy, it mostly clarified long-standing practice. The major new element was an understanding that these loans would normally be limited to 25 percent of quota. In four of the nine earlier cases, the Fund had lent larger amounts.86

Under the 1982 policy, the Fund made emergency loans on eight occasions through 1989, and then use almost stopped. From 1990 through 1997, the only such loan was to Pakistan (Table 5.4). In September 1992, the Indus River flooded, inundating a swath of southern Pakistan, forcing several hundred thousand people to evacuate their homes, and disrupting economic activity throughout the country. Pakistan had completed a SAF arrangement in December 1991 and was preparing to negotiate a larger ESAF arrangement with the Fund. As soon as the floods subsided, the Fund sent a staff team to Islamabad to assess the damage and the economic effects. Soon afterward, the authorities requested emergency assistance, which the Executive Board approved on November 25.87

Table 5.4.

Emergency Loans for Natural Disaster Relief, 1990–99

article image
Source: IMF financial accounts.

A more widespread natural disaster also occurred in 1992, in the form of one of the most severe African droughts of the twentieth century. The worst effects were in southern Africa, but drought conditions persisted in the east as well, as far north as the Horn of Africa. Like Pakistan, most of the affected countries were eligible for concessional assistance. In contrast, however, most of them either had programs already in place or were well into the process of negotiating them. Rather than offering emergency assistance on General Resources Account (GRA) terms, the Fund accelerated, and in several cases augmented, its ESAF support to many of the affected countries.88

The next use of this policy came in October 1998, after a massive flood submerged three-fourths of the land area of Bangladesh. At that time, Bangladesh had not borrowed from the Fund for five years. The country’s only outstanding debts were from an ESAF arrangement concluded in 1993. Elections in 1996 had brought in a new government and a period of stability that was brightening the country’s economic prospects. In the summer of 1998, the authorities were engaged in discussions with the staff about a possible successor ESAF arrangement when the floods disrupted the economy and weakened the outlook considerably.89 On October 28, the Executive Board approved an emergency loan of $138 million, or 25 percent of quota, with an expectation that this response would soon be followed by larger support through the ESAF. However, owing in part to political uncertainty and in part to the underlying challenges arising from the country’s extreme poverty, policy implementation did not live up to the Fund’s expectations. Bangladesh’s next borrowing, in the form of a three-year Poverty Reduction and Growth Facility (PRGF) arrangement, did not occur until 2003.

Two devastating hurricanes passed through the Caribbean region in 1998. Hurricane Georges hit the Dominican Republic and Haiti on September 22 with 105-knot winds, killing more than 600 people and causing untold physical damage. In the Dominican Republic, the damage estimates ranged around 8 percent of GDP. The authorities had no other need to borrow from the Fund because foreign direct investment was covering much of the country’s external financing requirements. To help with the short-term reconstruction effort, the Fund provided an emergency loan of $56 million (25 percent of quota). Haiti, where the impact was estimated to be close to 2 percent of GDP, was out of compliance with the terms of its ESAF arrangement at the time of the hurricane. Rather than granting a waiver so it could lend on concessional terms, the Fund approved emergency assistance of $21 million (25 percent of quota) through its general account.

Hurricane Mitch was even worse. One of the two or three deadliest storms of the century, it caused heavy damage to several islands before slamming into Honduras and Nicaragua on October 27 with 155-knot winds. Before it finally weakened, it killed more than 9,000 people. Two weeks later, Camdessus visited Nicaragua, Honduras, and El Salvador to discuss and assess the devastation, which he found to be “dramatic and heartbreaking.”90 Honduras, where the Managing Director was escorted by the Catholic archbishop of Tegucigalpa, Cardinal Oscar Rodriguez Maradiaga, was especially hard hit. Whole villages had been obliterated by mudslides, and survivors were still numbly searching for family members who might lie buried in the sludge. The Fund responded with emergency loans to Honduras ($66 million) and to St. Kitts and Nevis ($2.3 million). Reflecting the magnitude of the destruction in Honduras, that loan was exceptionally large in relation to quota, at 50 percent. Assistance to Nicaragua took the form of an augmentation of the existing ESAF arrangement and an acceleration of debt relief under the Heavily Indebted Poor Countries Initiative. El Salvador had a precautionary stand-by arrangement already in place, but the authorities chose not to draw on it, nor to request emergency assistance.

The Fund made its final loan for natural disaster relief in the 1990s to Turkey. As discussed more fully in Chapter 12, Turkey was carrying out a staff-monitored program in August 1999 when it was struck by an earthquake that measured 7.4 on the Richter magnitude scale, killed more than 15,000 people, and left hundreds of thousands homeless. The Fund responded with an emergency loan of about $500 million (37.5 percent of quota) and then accelerated its negotiations on a three-year stand-by arrangement.

Postconflict Assistance

Armed conflict, whether internal civil war or between countries, inevitably interrupted normal relations with the international community, including the IMF. Severe conflicts often left countries without basic economic, financial, and legal institutions and without the administrative capacity to carry out sustainable economic policies. Postconflict recovery required coordinated efforts among aid agencies, bilateral donors, policy advisors, and creditors. When a large number of diverse situations arose in the early 1990s, the lack of planning and coordination became all too clear. In response, the 1995 G7 summit meeting in Halifax called on “the Bretton Woods institutions and the UN to establish a new coordination procedure, supported as necessary by existing resources, to facilitate a smooth transition from the emergency to the rehabilitation phase of a crisis, and to cooperate more effectively with donor countries.”91

It had indeed been a tumultuous few years. The collapse of the Soviet Union had given rise to conflicts among some of the newly independent countries, especially between Armenia and Azerbaijan, and to civil wars (notably in Georgia and Tajikistan). The 1991 Gulf War disrupted economic activity and trade across much of the Middle East. The 1994 civil war in Rwanda affected both that country and its neighbors in central Africa. The U.S. invasion of Panama in 1989, a 1991 military coup in Haiti, and a long-running civil war in El Salvador that ended in 1992 brought disruptions to the Caribbean and Central America. The restoration of more peaceful conditions in Southeast Asia led to a need for postconflict assistance to Cambodia and Vietnam. In each case, the IMF’s offers of technical assistance, policy advice, and loans had to be embedded within a broader multiagency rescue effort.

For once, the Fund rejected the central premise of the G7 request. Although it was certainly true that careful coordination was needed in each postconflict situation, the evidence did not suggest that coordination failure had been a generalized problem. In any case, it did not seem either desirable or practical “to establish a new coordination procedure.” Rather, case by case, agreement on who the lead agency should be was needed.92 For one country, it might be a bilateral donor with strong cultural and historical (usually postcolonial) ties. For another, it might be a regional development bank. Rarely if ever would logic dictate that it be the IMF. The staff concluded, and Executive Directors agreed, that the only need was for an implicit understanding that in the future, all the agencies involved in helping a postconflict country would designate one among them to take the lead, and all would coordinate their work within that framework.93

The Fund established a policy on emergency postconflict assistance (EPCA) in 1995, but that policy focused primarily on the Fund’s provision of financial assistance rather than on strategic coordination. In most cases, the borrowing country had not been ready or able to carry out a set of economic policies that the Fund could support with a stand-by arrangement, so it had devised an ad hoc proposal each time. For transition economies such as Georgia and Vietnam, the STF had been handy while it existed, but it was expiring in 1995. For Rwanda, the CCFF was appropriate, but the range of situations in which the facility could be used was limited. That left only the first credit tranche, but only if the country had not already drawn on it and was able to submit a plan for cooperating with the Fund in solving its problems. The Fund needed a policy under which it could lend quickly on an emergency basis, without regard to the restrictive rules of these facilities and other policies.

Additional problems arose when a country had outstanding payments arrears on earlier loans from the Fund. Chapter 16 details numerous conflicts that led to arrears, only some of which were resolved in the 1990s. Many of those debts would be unpayable without external help. For example, the signing of a peace agreement in Paris in October 1991 began a lengthy process by which Cambodia restored internal peace and emerged from more than 16 years of international isolation. The Fund began providing technical assistance and program monitoring soon afterward, but it could not resume financial assistance until Cambodia cleared its long-standing payments arrears. That process was completed only in October 1993. The Fund then quickly made a small conventional loan through the STF and entered into negotiations for concessional lending. The first ESAF arrangement was approved seven months later, in May 1994.

The EPCA policy the Fund established in September 1995 had two basic elements.94 First, it acknowledged that in most cases the Fund should not be the lead agency but should work in close coordination with others. Second, it extended the Fund’s policy on emergency assistance for natural disaster relief to cover postconflict emergencies, subject to certain conditions. To be eligible for EPCA, the borrowing country should have an “urgent balance of payments need.” Moreover, its administrative capacity and commitment should be adequate to plan a comprehensive recovery program, but not yet adequate to carry out a program that would warrant the Fund’s regular support through a stand-by or other conditional arrangement. For its part, the Fund should have an identifiable role “in catalyzing support from other official sources,” and that role should be part of a coordinated and comprehensive international effort to assist the country’s recovery from the effects of the conflict.

This new policy was not a fundamental change for the Fund. It did not greatly expand or modify the Fund’s role or its ability to lend when necessary. The policy mainly simplified its procedures in these cases and made them more transparent. Specifically, it enabled the Fund to lend a relatively small amount (usually 25 percent of quota) quickly and without the additional justification required by the CCFF. It also established the principle that the Fund should be involved in coordinated reconstruction efforts if at all possible, notwithstanding the objections of those who might seek to preserve the “monetary character” of the IMF by limiting its financial role to situations involving countries capable of carrying out a fully fledged adjustment program. Finally, it expressed agreement that the Fund would seek bilateral subsidies to cover the interest on GRA credits for ESAF-eligible countries availing themselves of early Fund support.

Executive Directors found extending the emergency financing procedures controversial, for two reasons. First, a number of them felt that the existing policies were perfectly adequate and could be applied flexibly without adopting a new policy. More seriously, several Directors argued that postconflict assistance was primarily the province of aid agencies and multilateral development banks and that the proposed role for the Fund threatened the institution’s monetary character. No formal vote was taken, but in the course of the meeting, seven Directors, with 26 percent of the voting power on the Board, expressed opposition. Hence, the Summing Up referred to the policy being adopted by “Directors in their majority.”95

Bosnia and Herzegovina made the first loan application to be considered under the new policy while still in the final stages of acceding to membership in the IMF (see Chapter 2). When EPCA was established, North Atlantic Treaty Organization (NATO) military forces were conducting an air campaign against Bosnian Serbs in an effort to force settlement of an ethnic war that had been raging for more than three years. Soon afterward, the UN arranged to fly representatives of several multilateral institutions into Sarajevo to begin advising Bosnian officials on reconstructing the economy once peace was restored. Remarkably, the situation was not yet postconflict. The UN plane had to “corkscrew” into the airport to minimize the risk of being shot down. As recounted in gripping and harrowing detail by the journalist Sebastian Mallaby, conditions on the ground were both dangerous (sniper fire was part of daily life) and lacking in basic comforts (Mallaby, 2004, Chapter 5).

The IMF was represented in this group by Scott B. Brown (Assistant to the Director of the European Department), a veteran of several earlier postconflict situations who came primarily to provide technical assistance on budget preparations and other matters.96 While he was in Sarajevo on this first trip, multilateral peace negotiations began in Dayton, Ohio (United States), and the major powers soon asked the Fund to shift into high gear to normalize relations.

Brown returned to Bosnia and Herzegovina in mid-November with a full staff mission team while the Dayton talks were still in progress. In less than a month, the membership process and the settlement of arrears were complete.97 On December 20—the same day the UN turned its peacekeeping operations in Bosnia and Herzegovina over to NATO—the Executive Board approved its first financial assistance to the newly independent and peaceful country—an EPCA loan for $45 million (Table 5.5). Although the World Bank and other agencies were also providing technical assistance by this time, the EPCA loan was the first financial assistance to Bosnia and Herzegovina from any external institution.

Table 5.5.

Emergency Postconflict Assistance, 1995–99

article image
Source: IMF financial accounts. Note: ESAF = Enhanced Structural Adjustment Facility; n.a. = Not applicable; PRGF = Poverty Reduction and Growth Facility; SBA = Stand-by arrangement.

After the loan to Bosnia and Herzegovina, most EPCA loans were to African countries: Rwanda in 1997, the Republic of Congo and Sierra Leone in 1998, and Guinea-Bissau and Sierra Leone in 1999.

As noted above (p. 220), Rwanda had borrowed through the CCFF in 1995. Two years later, the government was still struggling to reconstruct its economic institutions, but it had accomplished enough to warrant a Fund-supported policy program. As negotiations for an ESAF arrangement continued, the Fund provided interim financing through two EPCA loans. (An ESAF arrangement was approved in June 1998; see Chapter 14.)

The Republic of Congo (Brazzaville) borrowed from the ESAF in 1996, but policy stability collapsed after a civil war erupted in 1997. When fighting stopped in 1998, the Fund offered an EPCA loan because a resumption of the ESAF arrangement was not yet possible. The war soon resumed, but when it ended in 1999, the country held its first multiparty elections. The Fund provided a second EPCA loan in 2000 and recommenced concessional lending with a PRGF arrangement in 2004.

From 1994 through 1997, Sierra Leone borrowed through the ESAF, even though successive governments were still fighting off a brutal insurrection (Chapter 16). The restoration to power of a civilian government in 1998 created the conditions for EPCA lending, followed by a PRGF arrangement in 2001.

Guinea-Bissau established a democracy in 1994 and joined the CFA franc zone three years later. All too soon, the people’s hopes for stability were dashed when an insurrection erupted and eventually overthrew the elected government. Once new elections were scheduled in 1999, foreign assistance resumed, including an EPCA loan from the IMF in September.

Aside from these loans to African countries, EPCA loans in the 1990s were confined to transition countries. After the first such loan, to Bosnia and Herzegovina, the Fund lent under this policy to Albania and Tajikistan. In Albania, a massive collapse of the financial system in 1997 led to civil unrest that the government was incapable of controlling. A UN peacekeeping force restored order, and the Fund helped the reconstruction effort with an EPCA loan followed by a resumption of ESAF lending (see Chapter 6). In Tajikistan, a June 1997 peace accord ended a long-running civil conflict and enabled postconflict and other support from the IMF (Chapter 8).

As it happened, all the postconflict borrowers in this period except Bosnia and Herzegovina were eligible for concessional financing. In each case, the EPCA loan served as a bridge to larger and longer-term ESAF or PRGF arrangements once the authorities were able to prepare comprehensive reform programs (see Table 5.5). The policy adopted by the Fund in 1995 envisaged that donor countries would subsidize the interest cost of EPCA loans, thus reducing the borrower’s debt service to a level commensurate with the ESAF. In practice, that idea was too difficult to apply, and it never materialized. Instead, in most cases the Fund agreed to top up the subsequent arrangements to enable the borrower to repay the EPCA loan. In effect, this practice rescheduled the loans into the ESAF and later into the PRGF.

In 1998 and in subsequent reviews, the Fund considered various proposals to strengthen the provision of assistance to postconflict countries. The general conclusion, however, was that the policy was working well as long as it was applied flexibly and quickly in each applicable case. The policy continued with little change through the next decade.98

Support for Currency Stabilization Funds

When a democratically elected government took office in Poland in 1989, it decided to try to establish a market economy quickly—in a “big bang,” as the expression went—and to stabilize conditions by anchoring expectations with a pegged exchange rate. Because Poland started with very low foreign exchange reserves, a group of 13 industrial countries agreed to establish a currency stabilization fund (CSF) by putting $1 billion at the authorities’ disposal on an as-needed basis. The fund was to be administered by those countries’ Executive Directors at the IMF, and its activation was to be conditioned on the successful implementation of a Fund stand-by arrangement. The experiment was a clear success. The authorities pegged the zloty to the U.S. dollar, and the existence of the CSF as a secondary line of defense against a speculative attack gave the policy the necessary credibility. Poland did not have to draw on the fund, and it was allowed to expire at the end of 1992.99

The success of the Poland CSF inspired the Russian authorities to ask for similar support when they were joining the IMF in 1992. As explained in Chapter 7, the G7 agreed in principle to fund a $6 billion CSF for the ruble, but it never came to fruition. Subsequently, the subject occasionally arose in talks between Russia and the Fund, especially after the authorities pegged the ruble to the dollar in July 1995. Meanwhile, Ukraine also asked the Fund to sponsor or help arrange for a CSF in support of its “radical economic reform” program, launched in October 1994.100

The IMF responded to these requests by establishing a new lending policy.101 Basically, the Fund would approve a stand-by or extended arrangement with sufficiently large access to accommodate a separate pool of funds within it, set aside from the rest of the arrangement. This pool (the CSF) could be drawn upon only to finance short-term intervention in exchange markets. To qualify, the country needed a credible set of financial policies designed to maintain a fixed exchange rate as an anchor for expectations. To apply this policy sensibly, the Fund would conclude that a currency peg was viable and appropriate but might not be credible to financial markets, perhaps owing to past weaknesses in policy implementation.

The staff first proposed introducing a CSF policy in December 1994, but quite a few Executive Directors reacted skeptically. The Canadian Director, Ian Clark, opened the discussion by noting that the policy could be perceived as a general and inappropriate preference by the Fund for fixed exchange rates and exchange rate-based stabilization policies. Jung-Ho Kang (Alternate, Korea) worried that it could be used to support weak programs. If policies were strong enough to warrant financial support from the IMF, why would they lack credibility to such an extent that a CSF was warranted? Marc-Antoine Autheman (France) regarded the Fund’s existing policies to be sufficiently flexible. Alexandre Kafka (Brazil) suggested that a CSF would have to be quite large to be of any real use. Vicente J. Fernández (Alternate, Spain) thought that the whole idea of exchange rate-based stabilization was bad policy. Several other Directors, led by Karin Lissakers (United States), were more sympathetic and encouraging.102

Seven months later, the staff came back with a revised and more specific proposal, to create a new policy but not a new facility. No country would be offered an IMF-financed CSF except as one element of a stand-by or extended arrangement. That would ensure that any CSF would be firmly linked to upper credit tranche policy conditionality, but this new proposal had the disadvantage of restricting the size of a fund to the quota-based access ceilings for the regular facilities. To ensure even more control, the proposal gave the Executive Board the right to refuse activation of a CSF if, after the initial approval of the arrangement, it deemed policies to have weakened. Most of those who expressed doubts during the first discussion remained skeptical, but a majority favored moving ahead. Accordingly, the staff prepared a final version, and the Board approved it in September 1995.103

Once the policy was adopted, its restrictive features proved to be overwhelming. The low level of potential access, the limited circumstances under which it could be used, and—perhaps most important—the retention by the Executive Board of discretion in approving activation of CSFs, meant that the policy was of little use to countries contemplating the introduction of an exchange rate anchor. No country applied for a CSF, and the Fund abolished the policy in April 2000.104

Technical Assistance, Training, and Resident Missions

The IMF formally began providing technical assistance to its member countries in 1964, when it created the Fiscal Affairs Department, the Central Banking Service, and the IMF Institute. Each of these units worked according to a sort of template, helping countries establish treasury systems and central banks and providing training on macroeconomics and finance. Over time, those activities were supplemented by technical assistance from the Legal Department and the Bureau of Statistics, each in its own area of expertise. In the late 1980s, the Fund found itself providing more than 100 staff years of assistance annually—mostly free of charge—to many developing countries. Although this activity was small scale when compared with that of some other multilateral agencies, it accounted for a significant portion of the Fund’s administrative budget.105

In the course of the 1990s, this activity more than doubled in volume, to more than 300 staff years.106 One obvious reason for this burst of activity was the influx of new members, especially those in transition to market economies. Many of those countries regarded the Fund’s technical assistance as more valuable than its lending. The countries emerging from the former Soviet Union lacked most of the basic building blocks of a functioning economy. Monetary policy had been conducted in Moscow; revenues had been generated from state-owned enterprises rather than through taxes; international trade had been conducted on barter or bilateral settlement terms, with centralized decisions on production and the direction of trade; and bankruptcy procedures and market regulation did not exist. Officials and staff needed training and guidance on overcoming these and other daunting initial conditions.107

A second reason for the increase was that the Fund shifted from the template approach to one driven more by the needs of individual countries. That is, much of the increased technical assistance was provided to help countries undertake specific reforms, often in the context of Fund-supported programs. This assistance in policy design and implementation supplemented the traditional emphasis on capacity building and institutional development. Because it required knowledge of and experience in the country or region, the Fund frequently hired outside experts for this purpose—especially for work in the transition countries and new members—rather than using career staff.

In several instances in the 1990s, the Fund provided technical assistance to non-member countries. This practice started with the Soviet Union, through the Special Association agreement signed in October 1991 (Chapter 2). For the year or so after the union was dissolved and membership negotiations proceeded, the Fund—along with numerous other agencies—provided as much assistance as it could to get institution building and other reforms started.

In November 1994, staff from the Monetary and Exchange Affairs Department (MAE), in cooperation with officials from Israel and Jordan, began providing technical assistance to the Palestinian authorities in the West Bank and Gaza Strip to help them establish a Palestinian Monetary Authority. That work, initially directed by Arne B. Peterson (Advisor, MAE) focused on developing the payments system, regulating and supervising commercial banks, and managing the foreign exchange market. Experts from the Fund’s Legal and Middle Eastern Departments, the Bank of England, and the Bank of Finland also participated.108

In 1997–98, as described in Chapter 2, the Fund provided assistance to North Korea. That effort aimed to initialize a process that could lead eventually to membership, but in the end it came to nothing.

In October 1999, the UN succeeded in quelling a violent attempt by Indonesian military forces to suppress the independence movement in East Timor. As the UN prepared to establish an interim civil administration, it asked the IMF and the World Bank to provide technical assistance on establishing the rudiments of a national economy in East Timor. Immediately afterward, a Fund staff mission headed by Luis M. Valdivieso (Advisor, Asia and Pacific Department) went to East Timor to advise officials on restoring stability in the payments system, establishing a viable system of public finance, and rebuilding international trade.109 That effort continued and broadened until the country became a member of the IMF (as the Democratic Republic of Timor-Leste) in 2002.

The biggest recipient of Fund technical assistance in the 1990s was Namibia, which joined the IMF in 1990. Namibia, the last African country to join the Fund, was emerging from nearly a quarter-century of guerilla warfare and had just become a fully independent country. The Fund responded with a massive injection of staff and expert assistance, totaling more than 47 staff years by the end of the decade (Table 5.6). Russia was the second largest recipient of these services, and the next two on the list were contiguous neighbors: Angola to the north of Namibia, and Ukraine on Russia’s western flank. Overall, the IMF’s technical assistance during this decade focused primarily on Africa, as it had in the 1980s.

Table 5.6.

Top Recipients of Fund Technical Assistance by Region, FY1990–99

(Staff years)

article image
Source: Staff calculations. Note: BEAC = Bank of Central African States.

As technical assistance began to absorb an increasing portion of the administrative budget—as much as 15 percent by the late 1990s—some Executive Directors began to worry that the effort lacked a clear purpose and focus.110 Responding to those concerns was difficult, because assessing the benefits of technical assistance is inherently subjective and is complicated by the long gestation period before the benefits may be fully apparent. The Fund commissioned a few evaluations of aspects of its assistance programs during the 1990s. In 1999, it finally decided to undertake a comprehensive internal review. That review concluded that technical assistance was generally working well in that it was much in demand and appreciated by recipients. The report made several common-sense recommendations for improvements: develop an overall framework, integrate technical assistance with surveillance and program work, focus on those areas in which the Fund has a clear comparative advantage, and regularly reassess what works and what does not.111 That report induced the Fund to develop, for the first time, a policy framework for the provision of technical assistance, which it published in 2001.112 Four years later, the Independent Evaluation Office conducted a further review (Independent Evaluation Office, 2005b), which found that the Fund was continuing to make progress but with somewhat uneven results.

The Fund also expanded its training programs for country officials in the 1990s, notably by helping establish five regional training institutes through joint ventures. This process began with a November 1991 proposal by the Fund for a regional training center for officials from transition countries. After extensive discussions among multilateral organizations and officials from the transition countries and from western Europe, the Joint Vienna Institute (JVI) was established in September 1992 as a temporary training center.113 Locating this joint venture in Vienna was felicitous, in part because of the city’s convenient location for officials from countries of the former Soviet Union and the transition economies of Central and Eastern Europe, but also because Austria had always maintained good relations with many of those countries. Andrew J. Beith, who had been serving as director of the IMF’s Offices in Europe, was named as the JVI’s first director.

At the outset, the JVI was intended to be a temporary venture to provide training to officials of transition countries, only until more general educational and training facilities could take over the task. During its first four years, through the end of 1997, 8,300 participants from 33 countries completed training courses at the JVI. By then, it was clear that the demand for such training was going to extend much longer than the original termination date of August 1999.114 All the sponsors agreed to extend the life of the JVI to 2004, and in 2003 they further amended the agreement to convert the JVI into a permanent institute.

In 1993, the IMF set up the Pacific Financial Technical Assistance Centre (PFTAC) in Suva, Fiji, as a joint venture with the UN Development Program (UNDP). The PFTAC was designed to provide training to officials from 15 Pacific island countries. Additional financing was provided by the aid agencies of Australia, Japan, and New Zealand; the Asian Development Bank; and the South Pacific Forum. The IMF-Singapore Regional Training Institute was established in 1998 as a joint venture with the host country. The IMF-AMF Regional Training Program in Abu Dhabi, United Arab Emirates, began offering courses in 1999 as a joint venture with the Arab Monetary Fund. The Joint Africa Institute was also established in 1999, as a joint venture with the World Bank and the African Development Bank. Each of these institutes accepted students from among the officials of countries in the regions in which they were located.

Finally, the Fund’s Resident Representative offices served as a way for management and staff in Washington to stay abreast of developments in its borrowing countries, for the staff to provide ongoing technical assistance and policy advice to the authorities, and for the Resident Representative to explain the Fund and its functions to officials and to civil society in those countries. The practice of setting up offices in member countries began in March 1956, when the Fund assigned a staff member to work in Paraguay for six months to assist the authorities in implementing policies to control inflation following devaluation of the guarani and liberalization of the foreign exchange market.115 Two more offices opened later that year, in Bolivia and Haiti. In 1965, when seven overseas offices were in place, the IMF formally established a policy providing for Resident Representatives of the Fund, as distinguished from resident or technical advisors with more limited responsibilities. The program gradually grew, and by the end of the 1980s the Fund had more than 30 Resident Representative (or “resrep” in the common parlance of 19th Street) offices around the world. Most postings were made to help oversee implementation of Fund-supported programs, but the policy also allowed for a post to be established when the Fund had some other form of “intensive involvement,” such as a staff-monitored program or ongoing negotiations.116

The early 1990s witnessed a sharp rise in the number of Resident Representative posts, from 32 in April 1990 to 69 four years later (Figure 5.5). This growth was unrelated to the number of active stand-by and other arrangements, which were almost flat during that period. Instead, the influx of new member countries with immense and diverse needs for technical assistance and policy advice drove the growth. Many of the new members were not yet prepared to implement fully articulated policy programs, and an important role of the Resident Representative was to assist in that preparation and to coordinate a regular flow of technical advisors and staff missions from headquarters. In several cases, starting with Poland in 1990 and then including China, India, and Russia, the Fund’s local offices were staffed by more than one Resident Representative.117 After 1994, the number of overseas offices remained roughly constant.

Figure 5.5.
Figure 5.5.

Resident Representative Posts, 1990–99

Source: Annual Reports.

Appendix: Credits and Loans, 1990–99

(Millions of SDRs, except as noted)

article image
article image
article image
article image
Source: IMF financial accounts. Note: n.a. = Not applicable.

Excludes emergency assistance.

References

  • Blustein, Paul, 2001, The Chastening: Inside the Crisis that Rocked the Global Financial System and Humbled the IMF (New York: PublicAffairs).

    • Search Google Scholar
    • Export Citation

  • Boughton, James M., 2001, Silent Revolution: The International Monetary Fund 1979–1989 (Washington: International Monetary Fund).

  • Boughton, James M., 2005, “Who’s in Charge? Ownership and Conditionality in IMF-Supported Programs,” in Globalization and the Nation State: The Impact of the IMF and the World Bank, ed. by Gustav Ranis, James Raymond Vreeland, and Stephen Kosack (Abingdon, Oxfordshire: Routledge), pp. 19–35.

    • Search Google Scholar
    • Export Citation

  • Boughton, James M., and Alex Mourmouras, 2004, “Whose Programme Is It? Policy Ownership with Conditional Lending,” in The IMF and Its Critics: Reform of Global Financial Architecture, ed. by David Vines and Christopher L. Gilbert (Cambridge: Cambridge University Press), pp. 225–53.

    • Search Google Scholar
    • Export Citation

  • Cashin, Paul, Hong Liang, and C. John McDermott, 2000, “How Persistent Are Shocks to World Commodity Prices?” Staff Papers, International Monetary Fund, Vol. 47, No. 2, pp. 177–217.

    • Search Google Scholar
    • Export Citation

  • de Vries, Margaret, 1985, The International Monetary Fund 1972–1978: Cooperation on Trial (Washington: International Monetary Fund).

  • Fischer, Stanley, 2000, “Strengthening Crisis Prevention: The Role of Contingent Credit Lines,” speech at the Bank of Mexico, November 15. Accessed at http://www.imf.org/external/np/speeches/2000/111500.htm.

    • Search Google Scholar
    • Export Citation

  • Fischer, Stanley, 2001, “International Economic Policy under the Clinton Administration,” comments at a conference at Harvard University, June 27. Accessed at http://www.imf.org/external/np/speeches/2001/062701.htm.

    • Search Google Scholar
    • Export Citation

  • Gilbert, Christopher L., 1996, “International Commodity Agreements: An Obituary Notice,” World Development, Vol. 24, No. 1, pp. 1–19.

    • Search Google Scholar
    • Export Citation

  • Goldstein, Morris, 2001, “IMF Structural Conditionality: How Much Is Too Much?” Working Paper No. 01-4 (Washington: Peterson Institute for International Economics). Accessed at http://www.piie.com/publications/wp/01-4.pdf.

    • Search Google Scholar
    • Export Citation

  • Independent Evaluation Office of the IMF (IEO), 2002, Evaluation of Prolonged Use of Fund Resources (Washington: International Monetary Fund).

    • Search Google Scholar
    • Export Citation

  • Independent Evaluation Office of the IMF (IEO), 2005a, IMF Support to Jordan, 1989–2004 (Washington: International Monetary Fund). Accessed at http://www.ieo-imf.org/eval/complete/pdf/12062005/report.pdf.

    • Search Google Scholar
    • Export Citation

  • Independent Evaluation Office of the IMF (IEO), 2005b, IMF Technical Assistance (Washington: International Monetary Fund).

  • International Monetary Fund (IMF), 2002, “Guidance on the Design and Implementation of IMF Conditionality: Preliminary Considerations,” Policy Development and Review Department (Washington: International Monetary Fund). Accessed at http://www.imf.org/external/np/pdr/cond/2002/eng/guid/053102.pdf.

    • Search Google Scholar
    • Export Citation

  • Ivanova, Anna, Wolfgang Mayer, Alex Mourmouras, and George Anayiotos, 2003, “What Determines the Implementation of Fund-Supported Programs?” in IMF-Supported Programs: Recent Research, ed. by Ashoka Mody and Alessandro Rebucci (Washington: International Monetary Fund), pp. 160–86.

    • Search Google Scholar
    • Export Citation

  • Mallaby, Sebastian, 2004, The World’s Banker: A Story of Failed States, Financial Crises, and the Wealth and Poverty of Nations (New York: Penguin Press).

    • Search Google Scholar
    • Export Citation

  • Phillips, Michael M., 1999, “IMF Aid Plan Contains All but a Client,” Wall Street Journal, August 13, p. 15.

  • Potter, Barry H., and Jack Diamond, 2000, Setting Up Treasuries in the Baltics, Russia, and Other Countries of the Former Soviet Union, IMF Occasional Paper No. 198 (Washington: International Monetary Fund).

    • Search Google Scholar
    • Export Citation

  • Thomas, Alun, and Uma Ramakrishnan, 2006, “The Incidence and Effectiveness of Prior Actions in IMF-Supported Programs,” IMF Working Paper 06/213 (Washington: International Monetary Fund).

    • Search Google Scholar
    • Export Citation
1

Opening remarks, press briefing on “Report on Financial Sector Crisis and Restructuring: Lessons from Asia” (September 25, 1999); accessed at http://www.imf.org/external/np/tr/1999/tr990925.htm.

2

In technical terms, these disbursements are “purchases” of currencies by the member and are not loan disbursements. For more on terminology, see the Preface.

3

The peak in IMF lending in proportion to world trade came in the 1960s, primarily because of large lending to the United Kingdom. The peak in the percentage of member countries that borrowed from the Fund came in the 1970s, owing to the ready availability and low conditionality of the Oil Facilities. The peak in the number of borrowing countries came in the 1990s.

4

The small spike in 1991 (point G in Figure 5.1) is attributable partly to the Gulf War and a fiscal crisis in India, both of which raised the numerator, and partly to a delay in increasing quotas, which kept the denominator low.

5

Any definition of prolonged borrowing is somewhat arbitrary. The definition used here is the same as that used in Boughton (2001), pp. 618–19. A prolonged user at any given date has outstanding obligations to the Fund at least equal to its quota and has had at least five annual Fund-supported programs in the preceding 10 years. In one dimension, this definition is broader than the one used by the Fund’s Independent Evaluation Office (Independent Evaluation Office, 2002), in that the IEO required a country to have seven annual programs over 10 years. It is, however, more restrictive in another dimension, specifying as it does a threshold for outstanding obligations. For this purpose, outstanding obligations exclude overdue interest charges and penalties, which normally are capitalized and included in the total. The definition of a Fund-supported program excludes programs supported only by purchases in the first credit tranche, the Compensatory Financing Facility or Compensatory and Contingency Financing Facility, the Buffer Stock Financing Facility, or emergency disaster relief.

6

The word “temporarily” was added as part of the first amendment, in 1969, but the change was regarded as a clarification and not a change in policy.

7

“Bulgaria—Use of Fund Resources—Compensatory and Contingency Financing Facility—Request for Financing for Fluctuations in the Cost of Oil Imports,” EBS/91/22, Suppl. 1 (February 14, 1991), p. 16. Also see Singh’s remarks at the Executive Board meeting, minutes of EBM/91/27 (February 25, 1991), pp. 18–19.

8

Minutes of EBM/91/27 (February 25, 1991) and EBM/91/37 (March 15, 1991). The CCFF loan was approved at the first meeting; the stand-by arrangement at the second.

9

Also see the discussion of Jordan in Chapter 16, p. 854.

10

For the history of conditionality through 1989, see Boughton (2001),Chapter 13, and references therein.

11

Executive Board Decision No. 6056-(79/38), adopted March 2, 1979; reproduced in Boughton (2001), pp. 629–31.

12

Failure to satisfy a performance criterion would preclude the Fund from approving a drawing unless the Executive Board explicitly granted a waiver. As a general rule, the Board granted waivers if the deviation was minor or temporary or if the authorities were taking sufficient corrective actions.

13

For a review, see Thomas and Ramakrishnan (2006).

14

“External Adjustment, Financing and Growth—Issues in Conditionality,” EBS/87/40 (February 25, 1987), p. 19.

15

For analyses of the relationships linking conditionality to ownership and program effectiveness, see Boughton (2005); Boughton and Mourmouras (2004); and Ivanova and others (2003).

16

This review covers only the facilities for borrowing from the Fund’s General Resources Account. The separate trust funds for lending on concessional terms are covered in Chapter 13.

17

Systemic Transformation Facility. Decision No. 10348-(93/61) STF, adopted April 23, 1993. This document may be accessed at http://www.imf.org/external/pubs/ft/history/2011/index.htm.

18

Paragraph 3(a)(i) of Systemic Transformation Facility. Decision No. 10348-(93/61) STF, adopted April 23, 1993. This document may be accessed at http://www.imf.org/external/pubs/ft/history/2011/index.htm.

19

The original proposal—”A Facility to Help Members Respond to Systemic Disruptions in Economies in Transition,” EBS/93/56 (April 1, 1993)—envisaged about 15 countries borrowing up to SDR 3.5 billion. The final tally was SDR 4.1 billion by 20 countries.

20

Minutes of EBM/94/54 (June 17, 1994), pp. 4–5. Also see “The Role of the Fund in Financing the Economies in Transition—Access and Cofinancing Trust Accounts,” EBS/94/121 (June 3, 1994).

21

For Waterman’s and others’ responses, see minutes of EBM/94/54 (June 17, 1954). For the amendments to the STF, see the footnotes to “Systemic Transformation Facility, Decision No. 10348-(93/61) STF, adopted April 23, 1993,” accessible at http://www.imf.org/external/pubs/ft/history/2011/index.htm. In April 1995, the Board declined to approve any further extensions; minutes of EBM/95/41 (April 19, 1995).

22

Article XXX of the Fund’s Articles of Agreement permits the Fund to exclude purchases under special facilities from the calculation of the member’s reserve tranche position. To qualify, such facilities must be approved by an 85 percent majority.

23

Lissakers made her request at EBM/95/26 (March 20, 1995), p. 76. The substantive discussion took place at EBM/95/41 (April 19, 1995). The quotations from Link and Mesaki are on pp. 39 and 42, respectively, of the minutes for the latter meeting.

24

Vietnam’s one drawing, at 5 percent, was the only one not equal to 25 percent of quota.

25

“Short-Term Financing Facility,” EBS/94/193 (September 26, 1994).

26

For a discussion, see de Vries (1985). The proposal is reproduced therein in Volume III, p. 16, and discussed in Volume I, pp. 123–34. The reference to “Article VIII members” is to countries that have accepted the obligation not to impose capital controls. The Committee of Twenty was the predecessor of the Interim Committee.

27

Memorandum from E. Walter Robichek (Director, Western Hemisphere Department) to C. David Finch (Director, Exchange and Trade Relations Department), “‘Reverse Stand-by Arrangement’,” December 14, 1979; IMF archives, Historian’s files. Robichek evidently envisaged Brazil as a prime example of a qualifying country.

28

Minutes of EBM/80/161 (November 5, 1980), p. 12.

29

On this and the next paragraph, see Boughton (2001), pp. 560–63 and 747–48.

30

“The Adequacy of Existing Arrangements to Deal with Major Strains in the International Financial System,” EBS/82/194 (October 22, 1982).

31

Minutes of EBM/82/150 (November 19, 1982), p. 13 (Laske); and EBM/82/151 (November 19, 1982), p. 29 (de Larosière).

32

Minutes of EBM/94/104 (November 30, 1994), pp. 4–5 (Waterman), 12–13 (Clark), and 71 (Camdessus, in the next paragraph).

33

“The Halifax Summit Review of the International Financial Institutions: Background Document,” June 16, 1995, Section 5; accessed at http://www.g8.utoronto.ca/summit/1995halifax/financial1/index.html. The document was issued in the IMF as EBS/95/86 (June 20, 1995).

34

“An Emergency Financing Mechanism,” SM/95/216 (August 22, 1995).

35

Emergency Financing Mechanism. Summing Up by the Chairman, EBM/95/85 (September 12, 1995). This document may be accessed at http://www.imf.org/external/pubs/ft/history/2011/index.htm.

36

Minutes of EBM/95/85 (September 12, 1995). Autheman’s remark is on p. 45; the Chairman’s Summing Up establishing the procedure is on pp. 53–57; and Lissakers’ objection to “truly exceptional” is on p. 57.

37

The three Asian cases are discussed in detail in Chapter 11. Russia is discussed in Chapter 7. On the exceptional access, see Chapter 15, especially Table 15.3.

38

In the following decade, the IMF invoked the EFM eight more times: for Turkey in 2001, Georgia in 2008, and—in response to the crises of the “Great Recession”—for Armenia, Hungary, Iceland, Latvia, Pakistan, and Ukraine in 2009.

39

“A New Framework for Enhanced Asian Regional Cooperation to Promote Financial Stability,” Meeting of Asian Finance and Central Bank Deputies: Agreed Summary of Discussions, paragraph 7, Manila, Philippines, November 18–19, 1997; accessed at http://www.mof.go.jp/english//if/if000a.htm.

40

“Supplemental Reserve Facility,” EBS/97/225 (December 5, 1997), pp. 7–11.

41

Minutes of EBM/97/121 (December 15, 1997) and EBM/97/123 (December 17, 1997).

42

Supplemental Reserve Facility. Decision No. 11627-(97/123) SRF, adopted December 17, 1997. This document may be accessed at http://www.imf.org/external/pubs/ft/history/2011/index.htm.

43

Activation of the General Arrangements to Borrow is discussed further in Chapter 15. Following the debt default in August 1998, Russia did not make any further drawings on the EFF arrangement or the SRF.

44

Brazil drew on the SRF again from 2001 to 2003. The other users were Turkey (2000–02), Argentina (2001), and Uruguay (2002).

45

See Fischer (2000) and “Communiqué of Meeting of Economic Policy Makers in the Western Hemisphere Region,” press release PR/98/37 (September 3, 1998); accessed at http://www.imf.org/external/np/sec/pr/1998/pr9837.htm.

46

“Global Economy,” speech at the Council on Foreign Relations, September 14, 1998; accessed at http://www.cfr.org/publication/9349/global_economy.html?breadcrumb=%2Fpublication%2Fpublication_list%3Ftype%3Dtranscript%26page%3D66. Administration officials had discussed this proposal with Camdessus while the speech was in preparation, and he had encouraged them to include it. See the transcript of his press conference on October 8, 1998, at http://www.imf.org/external/np/tr/1998/TR981008.HTM.

47

“Statement by the G7 Finance Ministers and Central Bank Governors, Washington DC, October 3, 1998,” paragraph 3; accessed at http://www.g7.utoronto.ca/finance/fm100398.htm. The Interim Committee Communiqué (October 4, 1998) simply reproduced the G7 sentence word for word.

48

Minutes of EBM/98/108 (October 26, 1998), p. 5.

49

“G7 Leaders Statement on the World Economy” (October 30, 1998); accessed at http://www.g7.utoronto.ca/finance/g7_103098.html. The statement was issued by the U.K. Chancellor of the Exchequer, Gordon Brown, who was then chairing the G7 finance ministers.

50

“Review of the Supplementary Reserve Facility and Preliminary Consideration of a Contingent Credit Line,” EBS/98/214 (December 9, 1998), p. 7n.

51

G7 member countries were not uniformly enthusiastic about the proposal. As Blustein (2001, pp. 331–36) reported, Germany was particularly concerned about the moral hazard implications. After the October 1998 meetings, however, German officials focused their interventions on an effort to tighten the standards for using the facility, not preventing its adoption; see in particular the statement by Bernd Esdar at EBM/99/6 (January 13, 1999), pp. 28–31.

52

Minutes of EBM/99/6 (January 13, 1999), pp. 6 (Shaalan), 11 (Wijnholds), 15 (Zhang), 35 (Kiekens), and 40 (Mozhin).

53

“Review of the Supplementary Reserve Facility and Preliminary Consideration of a Contingent Credit Line,” EBS/98/214 (December 9, 1998), pp. 7–16.

54

“Further Considerations Toward a Contingent Credit Line,” SM/99/54 (February 24, 1999), p. 14.

55

Summing Up by the Chairman, minutes of EBM/99/48 (April 23, 1999), pp. 32–37.

56

On April 23, 1999, the Executive Board approved an extension of the Supplemental Reserve Facility to provide for contingent credit lines. The SRF decision was amended to add a second section. The amended document may be accessed at http://www.imf.org/external/pubs/ft/history/2011/index1.htm.

57

In the course of a press conference explaining the establishment of the CCL, Jack Boorman (Director, Policy Development and Review Department) suggested that it was “no surprise … that Argentina’s name would come up as a possible candidate,” because it met the specified criteria for eligibility; see “IMF Press Briefing, ‘ Transparency, Standards, and CCL’” (April 26, 1999); accessed at http://www.imf.org/external/np/tr/1999/tr990426.htm. Speculation about Mexico, South Africa, and Korea, based on interviews with unnamed IMF officials, was mentioned in Phillips (1999), p. 15.

58

For a summary review of the various efforts made to modify the CCL to make it more attractive, see “Signaling by the Fund—A Historical Review” (July 16, 2004), pp. 28–33; accessed at http://wwwimf.org/external/np/pdr/signal/2004/071604.htm. After the termination of the CCL, the Fund continued to refine its thinking on the subject. That led to the establishment in 2009 of the Flexible Credit Line, which was used almost immediately by Mexico and then by others.

59

For details, see “Addressing Year 2000 Issues in the Fund,” FO/DIS/98/85 (September 11, 1998).

60

“Review of Fund Technical Assistance,” EBAP/99/59, Suppl. 1 (May 17, 1999), p. 152.

61

Minutes of EBM/99/108 (September 23, 1999), pp. 44–55. Many on the staff were even more skeptical than Taylor about the wisdom of establishing a special facility for this limited purpose. Jack Boorman—who, as Director of the Policy Development and Review Department, normally would have been the staff representative at the Executive Board meeting—later recalled that he objected so strongly that he sent Leddy instead.

62

Year 2000 Facility. Decision No. 12058-(99/110) Y2KF, adopted September 24, 1999. This document may be accessed at http://www.imf.org/external/pubs/ft/history/2011/index.htm.

63

For the history of the CFF through the 1980s, see Boughton (2001), pp. 723–42, and references therein.

64

Statement by Mark Sobel (Senior Advisor to the U.S. Executive Director) at EBM/99/56 (May 26, 1999).

65

“Israel—Use of Fund Resources—Request for Purchase under the Compensatory and Contingency Financing Facility,” EBS/92/41 (March 9, 1992). At the time of the request, Israel had no outstanding financial obligations to the Fund.

66

Minutes of EBM/92/37 (March 27, 1992).

67

The program the IMF agreed to monitor is set out in “Ghana—Staff Report for the 1991 Article IV Consultation,” EBS/92/34 (February 28, 1992). The background to the CCFF loan is explained in “Ghana—Staff Report for the 1993 Article IV Consultation and Request for Purchase under the Compensatory and Contingency Financing Facility,” EBS/93/95 (June 16, 1993). The Fund’s relations with Ghana before the 1990s are discussed in Boughton (2001), pp. 673–79. Enhanced surveillance is also explained there, pp. 429–36. Ghana resumed borrowing on a more regular basis, again by taking out ESAF and then Poverty Reduction and Growth Facility loans, in 1995.

68

The August 1999 loan to the former Yugoslav Republic of Macedonia could also be considered to be in this category. The country had an EFF arrangement in effect, but it was not able to draw on it owing to policy slips. The authorities were discussing a replacement arrangement with the staff at the time of the loan.

69

“The Managing Director’s Statement on Considerations on the Role of the Fund in Current Circumstances,” BUFF/90/182 (September 17, 1990). The estimate of 30 or more countries was made implicitly in “The Response of the Fund in the Wake of Recent Developments in the Middle East,” EBS/90/179 (October 16, 1990), pp. 23–24. The staff computed that 65 countries might benefit from an oil-import window, but close to half of those could be assisted alternatively through modifications to existing lending arrangements.

70

Although the CCFF was designed to assist countries facing specific types of payments problems, the calculations took into account positive changes in other items that might offset the problem. The closest precedents for the oil-import window were the Oil Facilities that the Fund established in response to the first oil price shock of the 1970s. Those facilities were in effect in 1974–76.

71

“The Response of the Fund in the Wake of Recent Developments in the Middle East,” EBS/90/179 (October 16, 1990), p. 24; minutes of EBM/90/155 (November 2, 1990), pp. 7–11 (Dawson) and 13 (Kafka); and minutes of EBM/90/156 (November 2, 1990), p. 34 (concluding remarks by Camdessus). On September 24, the Interim Committee “agreed that the Fund should respond on an expedited basis … through use and, as appropriate, adaptation of its existing facilities, including” the CCFF (Communiqué, paragraph 5; Annual Report 1991, p. 121). On Kafka’s role as Dean of the Executive Board, see Chapter 17.

72

Oil Import Window in the CCFF. On December 5, 1990, the Fund amended the Compensatory and Contingency Financing Facility in several ways, including by adding a section establishing a temporary procedure for compensating countries for an increase in the cost of importing oil. This section was titled “Section V. Compensatory Financing of Fluctuations in the Cost of Oil Imports.” This document, as amended, may be accessed at http://www.imf.org/external/pubs/ft/history/2011/index.htm.

73

The decision establishing the oil window specified December 31, 1991, as the terminal date for a country to submit a request (paragraph 41) but permitted drawings to be made until June 30, 1992 (paragraph 60). The first drawing was made by Bulgaria, on February 28, 1991; the last was by Romania, on June 15, 1992.

74

The Fund announced the full package of measures in a press release on November 15, 1990 (PR/90/57).

75

On the origins and early use of the cereals element, see Boughton (2001), pp. 730–33 and 748–53. The rules limited its use to SITC categories 041 through 047, which included wheat, maize, rice, barley, sorghum, and millet, plus flour from those grains.

76

“Review of the Decision on Compensatory Financing of Fluctuations in the Cost of Cereal Imports,” SM/87/86 (April 8, 1987), p. 5 and Table 2.

77

Table 5.2 begins with 1990 and thus excludes the first two cases: Trinidad and Tobago in January 1989 and the Philippines in May 1989, neither of which was activated. The account in Boughton (2001) includes a typographical error on p. 741. A sentence in the middle of the page should read as follows: “During the first 12 months of CCFF operations, the Fund negotiated and approved 29 arrangements, only 2 [not 12] of which included contingency provisions.”

78

For an analysis of the inflexibilities and other shortcomings of the ECM, see Section II of “Compensatory and Contingency Financing Facility (CCFF)—Review of Facility,” EBS/92/201 (December 4, 1992). The quotation is from p. iii. Table 1 of that paper includes a detailed account of the 11 cases in which the mechanism was used, including the two that predate the period covered here.

79

See “Review of the Compensatory and Contingency Financing Facility and Buffer Stock Financing Facility—Preliminary Considerations,” EBS/99/222 (December 9, 1999); and minutes of EBM/00/5 (January 14, 2000) and EBM/00/27 (March 16, 2000).

80

This history is described in more detail in Boughton (2001), pp. 742–44.

81

“Review of the Compensatory and Contingency Financing Facility and Buffer Stock Financing Facility—Preliminary Considerations,” EBS/99/222 (December 9, 1999), pp. 20–22.

82

The origins of the EFF are covered in de Vries (1985), pp. 361–83. For the history of the facility through 1989, see Boughton (2001), pp. 705–23.

83

The only major change in the facility during this period, other than changes in access limits and charges that were implemented as part of a more general revision of lending policies, was a simplification of procedures for supporting debt- and debt-service-reduction operations under the Brady Plan. The original decision, made in May 1989, segmented the amounts by which the Fund could augment arrangements according to the nature of the operations (debt buybacks, collateralization with Brady bonds, and so forth). Those segmentation rules were eliminated in January 1994; see “Modalities of Fund Support for Debt and Debt-Service Reduction,” EBS/93/190 (November 30, 1993); minutes of EBM/94/1 (January 7, 1994); and “Summing Up by the Acting Chairman—Modalities of Fund Support for Debt and Debt-Service Reduction,” BUFf/94/2(January 10, 1994).

84

See the related discussion in Chapter 9 and particularly the list in Table 9.1. Nearly half of the extended arrangements approved during this period—all three in 1989 and then 16 out of 36 in the 1990s—were for the heavily indebted countries discussed there.

85

The three unambiguously successful precautionary EFF arrangements were for Colombia, Kazakhstan, and Peru, all of which were in effect from 1999 to 2002. In addition, Egypt had a precautionary arrangement in effect in 1993–96, but the program was off track during the latter part of the arrangement and no drawings would have been allowed. Argentina had a precautionary arrangement that began in 1998 and was scheduled to run through February 2001. The authorities did not draw on it, but it was canceled in March 2000 and replaced by a three-year stand-by arrangement. That arrangement was announced as precautionary, but the authorities drew on it beginning in December 2000. The other seven precautionary extended arrangements that were in effect during the 1990s—for Kazakhstan (1996–99), Mexico (1989–93), Peru (1993–96 and 1996–99), the Philippines (1994–98), Tunisia (1988–92), and the Republic of Yemen (1997–2001)—were each drawn upon at least once.

86

The history of the Funds’ emergency disaster relief through 1989 is covered in Boughton (2001), pp. 744–47. The 1982 policy is reproduced therein on pp. 753–54.

87

See “Pakistan—Staff Report for the 1992 Article IV Consultation and Use of Fund Resources—Emergency Assistance,” EBS/92/174 (November 4, 1992); and “Pakistan—Emergency Assistance,” press release PR/92/83 (November 25, 1992).

88

See “Statement by the Managing Director on the Drought in Southern Africa,” EBD/92/89 (April 15, 1992); and “Statement by the Managing Director on the Drought in Southern Africa—An Update,” EBD/92/212 (September 18, 1992).

89

See “Bangladesh—Staff Report for the 1998 Article IV Consultation,” SM/98/232 (September 22, 1998); and “Bangladesh—Use of Fund Resources—Request for Emergency Assistance,” EBS/98/175 (October 21, 1998).

90

See his report to the Executive Board at EBM/98/123 (December 7, 1998), pp. 3–8.

91

Halifax Summit Communiqué (June 16, 1995); accessed at http://www.g7.utoronto.ca/summit/1995halifax/Communiqué/index.html#development.

92

The only new mechanism seriously considered was to have each country prepare a policy framework paper, similar to the documents that underpinned Fund and World Bank lending to low-income countries generally. Even if such documents would have been helpful, the idea was impractical for the emergency situations most likely to arise. See the minutes of EBM/95/82 (September 6, 1995), pp. 103–37. Also see the discussion of “tripartite documents” for countries of the former Soviet Union, which the Fund eventually rejected in 1992; Chapter 3, p. 84.

93

See “Fund Involvement in Post-Conflict Countries,” EBS/95/141 (August 16, 1995). The staff identified only one previous instance of serious coordination failure: El Salvador at the end of the civil war in 1992. In the Fund staff’s view, the World Bank got ahead of other agencies by providing financial aid before the government was ready to implement an appropriate framework for using it; “Fund Involvement in Post-Conflict Countries,” EBS/95/141 (August 16, 1995), pp. 6, 10, and 11.

94

On September 6, 1995, the Executive Board agreed to establish a policy on providing emergency financial assistance to countries emerging from internal or external conflicts. This Emergency Post-Conflict Assistance (EPCA) policy was set forth in the Chairman’s Summing Up of EBM/95/82. This document may be accessed at http://www.imf.org/external/pubs/ft/history/2011/index.htm.

95

Minutes of EBM/95/82 (September 6, 1995), pp. 103–37. The Directors who spoke against adopting the EPCA were Hachiro Mesaki (Japan), Oleh Havrylyshyn (Alternate to the Director for the Netherlands), Jarle Bergo (Norway), Abdulrahman Al-Tuwaijri (Alternate, Saudi Arabia), Yacoob Yusef Mohammed (Alternate, Bahrain), Aleksei V. Mozhin (Alternate, Russia), and Daniel Kaeser (Switzerland).

96

In the course of a 22-year career at the IMF, Scott Brown frequently found himself among the first international civil servants being sent to postconflict countries where conditions were still unsettled and possibly quite dangerous. In August 2003, eight years after this assignment, Brown was seriously injured in a terrorist bombing in Baghdad that killed Sergio Vieira de Mello (chief UN envoy to Iraq) and 21 other people.

97

The membership process for Bosnia and Herzegovina, which could not be completed until arrears were settled, is covered in Chapter 2. The settlement of arrears is covered in Chapter 16.

98

The first comprehensive review was made in 1998; see “Issues Note on Providing Additional Assistance to Post-Conflict Countries,” EBS/98/155 (September 1, 1998); and minutes of EBM/98/102 (September 22, 1998), pp. 3–25. In 2000, the Fund established a “special policy” on EPCA so that these loans would not count toward the standard access limits.

99

See Annex I of “Fund Policies with Regard to Currency Stabilization Funds—Preliminary Considerations,” EBS/94/230 (December 2, 1994).

100

The program is described in Chapter 8. On the request for a CSF, see “Ukraine—Staff Report for the 1994 Article IV Consultation and Request for a Purchase under the Systemic Transformation Facility,” EBS/94/203 (October 19, 1994), p. 10.

101

On September 13, 1995, the Executive Board established a policy on the use of Fund resources to finance currency stabilization funds (CSFs). This policy was set forth in the Summing Up of the discussion at EBM/95/86 and in an attachment with specific guidelines for CSF support. This document may be accessed at http://www.imf.org/external/pubs/ft/history/2011/index.htm.

102

Minutes of EBM/94/109 (December 14, 1994). Some years later, Stanley Fischer (2001) acknowledged that the “impetus” for this proposal and for several other revisions to lending policies in the 1990s came from the U.S. Treasury.

103

“Fund Policies with Regard to Currency Stabilization Funds—Further Considerations,” EBS/95/109 (June 30, 1995); and minutes of EBM/95/68 (July 19, 1995) and EBM/95/86 (September 13, 1995).

104

See “Review of Fund Facilities—Preliminary Considerations,” EBS/00/37 (March 2, 2000), accessed at http://www.imf.org/external/np/pdr/fac/2000/faciliti.pdf; and the Summing Up of the discussion at EBM/00/27 (March 16, 2000), accessed at http://www.imf.org/external/np/pdr/fac/2000/sum.htm.

105

For the history of technical assistance through 1989, see Boughton (2001), pp. 1017–18, and references therein.

106

Comparisons over time are only suggestive because the Fund’s definitions for reporting the provision of technical assistance were not consistent. Data for the 1990s are more comprehensive than those for the 1980s.

107

For a review of the Fund’s assistance to transition countries in setting up treasuries in the 1990s, see Potter and Diamond (2000).

108

For the initial technical assistance report, see “West Bank and Gaza Strip: Issues Related to the Establishment of a Palestinian Monetary Authority,” December 1994 (unnumbered).

109

See minutes of EBM/99/118 (October 22, 1999); “East Timor—Informal Board Meeting,” FO/DIS/99/165 (November 23, 1999); and “East Timor—Establishing the Foundations of Sound Macro-economic Management,” SM/00/178 (July 20, 2000).

110

Because most recipients were low-income countries, or nearly so, the IMF was reluctant to charge for the service. In March 1990, the Fund established an administered account to receive annual grants Japan was offering to finance assistance for some of the neediest countries. That enabled the Fund to expand its technical assistance program despite the substantial squeeze on the budget at that time.

111

“Review of Fund Technical Assistance,” EBAP/99/59 and Suppl. 1 (both dated May 17, 1999).

112

“Policy Statement on IMF Technical Assistance,” April 1, 2001; accessed at http://www.imf.org/external/pubs/ft/psta/index.htm.

113

The JVI was initially sponsored by the IMF, the Bank for International Settlements (BIS), the European Bank for Reconstruction and Development (EBRD), the Organization for Economic Cooperation and Development (OECD), the World Bank, and the Austrian central bank and finance ministry. The World Trade Organization (WTO) joined in 1998. See “A Brief History of the JVI” at http://www.jvi.org/index.php?id=168.

114

“The Joint Vienna Institute—Recent Developments and Extension of Mandate,” EBAP/98/12 (February 2, 1998); and minutes of EBM/98/21 (March 2, 1998), pp. 3–20.

115

See minutes of EBM/56/12 (February 27, 1956).

116

For the history of the Resident Representative program, see “Review of the Resident Representative Program,” EBAP/94/69 (September 2, 1994), pp. 1–3; and “Review of the Resident Representative Program—Selected Issues and Statistical Annex,” EBS/97/137, Suppl. 1 (September 17, 1997), pp 5–9.

117

Traditionally, the staffing usually included only one Resident Representative, assisted by a small number of locally recruited personnel. The Moscow office was exceptional in the 1990s in that it included as many as four Resident Representatives; see “Review of the Resident Representative Program,” EBAP/94/69 (September 2, 1994), p. 9. The length of an assignment for a Resident Representative in the 1990s was either two or three years, with an average length of 2.3 years; “Review of the Resident Representative Program—Selected Issues and Statistical Annex,” EBS/97/137, Suppl. 1 (September 17, 1997), p. 11.

  • Collapse
  • Expand
Cover Tearing Down Walls
Tearing Down Walls
The International Monetary Fund 1990-1999
Author:
Mr. James M. Boughton
ISBN:
9781616350840
Publication Date:
05 Mar 2012
Publisher:
International Monetary Fund
DOI:
https://doi.org/10.5089/9781616350840.071
Pages:
1036
Figure 5.1.

IMF Lending, 1947–2000

(In percentage of Fund quotas)

Percent
Figure 5.2.

IMF Lending by Facility, 1990–99

SDR billions
Figure 5.3.

Prolonged Users of Fund Resources, 1980–99

Number of countries
Figure 5.4.

Structural Conditionality in Stand-By Arrangements
Figure 5.5.

Resident Representative Posts, 1990–99