The International Monetary Fund: Challenges and Contradictions

International Financial Institution Advisory Commission
United States Congress

The International Monetary Fund has grown in resources and responsibilities since it was established in 1944. It has become abundantly clear that the IMF’s “original rationale no longer fits,”1 that the world economy has changed dramatically since the fund was established, and that the increasing frequency and severity of economic crises in recent years require a rethinking of the IMF’s role in the global financial system.

The IMF has recognized this need and is introducing some reforms to the way it operates. The fund has in fact developed new missions for itself many times in response to crises or changes in the world economy. Those episodes have included the end of the system of fixed exchange rates in the early 1970s, the subsequent oil crises of that decade, the Third World debt crisis of the 1980s, the collapse of socialism, and, beginning with Mexico in 1994, emerging market financial crises.

In short, the IMF has expanded its role from providing short-term loans based on macroeconomic policy change to providing longer-term aid conditioned on structural economic reforms to providing bailout funds and becoming an international crisis manager. As the graph shows, new missions and greater lending have led to periodic increases in the fund’s resources, which donor nations have granted every time such increases have been requested.

Sources: IMF, International Financial Statistics (various issues); and IMF, Financial Organization and Operations of the IMF (Washington: IMF, 1990).

The IMF today finds itself in an awkward position. It continues to provide massive aid to countries suffering from financial crises while wishing to avoid creating moral hazard. The agency has thus proposed initiatives to “bail in” the private sector, so as to make investors bear more of the cost of bad investment decisions. Conversely, the fund has created a line of credit to provide aid to countries before crises occur, so as to prevent them. Yet preventive lines of credit are likely to increase moral hazard, while efforts to force losses on the private sector may precipitate the very crises they intend to prevent. Finally, the IMF wishes to become more transparent and improve its surveillance function. Its dual goals of preventing the outbreak of financial turmoil and maintaining relations with client countries, however, may undermine the fund’s credibility.

Many of the problems the IMF seeks to resolve would be reduced or eliminated with increased reliance on direct negotiations between lenders and borrowers in international finance and decreased reliance on IMF lending and mediation. That has become more and more evident since the early 1980s. To see why, it is useful to look at the evolution of the IMF.

Early Years

The International Monetary Fund was established at Bretton Woods in the aftermath of the Great Depression and at the end of World War II, when confidence in a liberal world economy was low. The fund’s purpose was to maintain exchange rate stability by lending to countries experiencing temporary balance of payments problems. In a world of fixed exchange rates, countries would only be allowed to alter their exchange rates if there were fundamental imbalances in their economies. In this way, the IMF would promote international stability and avert competitive devaluations.

Although world trade did increase under the Bretton Woods system, trade did not return to its 1913 level (as a share of the global economy) until the mid-1970s, after the Bretton Woods system was abandoned. Moreover, the system was not as orderly as envisioned by its founders. Sharp and sudden currency devaluations would occur. Economic historian Leland Yeager observed that “The authorities of a country desiring to change its exchange rate typically make up their own minds on the matter and then simply notify the Fund. Though this notification is phrased as a request for permission, the Fund actually faces the choice only between acquiescing or risking loss of face by seeing its authority flouted and the change made anyway.”2

The Bretton Woods system was, in fact, “unworkable from the start” and “promptly began to break down” once major European countries began lifting capital controls on their currencies in 1958.3 The breakdown occurred because countries pursued a combination of unsustainable policies-free capital flows, fixed exhange rates, and independent monetary policies. By the time President Nixon finally abandoned this system in 1971, thus ending the international system of fixed exchange rates, he was merely acknowledging economic reality. That similar scenario has been played out in the 1990s as financial crises have forced developing countries to learn that they cannot pursue both independent monetary policies and tie their currencies to the dollar in a world of free capital flows. It is somewhat ironic then, that the recent financial crises in Asia and elsewhere have elicited calls for the establishment of a “new Bretton Woods.”

The Post-Bretton Woods Era

The collapse of the system of fixed exchange rates managed by the fund ended the IMF’s original mission. Lending by the fund, nevertheless, doubled from 1970 to 1975. The oil crises of that decade led the IMF to create oil facilities, or new lending mechanisms that would make it easier for countries to cope with the rising cost of oil imports. (In the event, Great Britain and Italy became the largest users of that credit line.)4 In creating the new credit window, the fund was going beyond its usual lending activities but was following a pattern it had begun in the 1960s, when it established two other credit facilities intended to help countries cope with swings in commodity prices.5 By 1975, the IMF launched the Extended Fund Facility, which would provide credit, mainly to developing countries, on more lenient terms than those available through standard IMF programs. The fund has since created a host of credit programs that provide increasingly concessional loans. Observers noted that the IMF was taking on functions more appropriate to a foreign aid agency like the World Bank than to an international monetary institution.

The outbreak of the Third World debt crisis in 1982 initiated the era of international financial rescues led by the IMF. In the post-World War II era, threats to the international financial system that were considered systemic had not yet erupted. While previous to the founding of the IMF, debt crises in poor countries involved little or no official mediation, authorities now ruled out direct negotiations between debtor countries and creditors as unrealistic. The largest U.S. money-center banks had made sovereign loans that exceeded their capital. Most banks were eager for the IMF to provide funding, as were Third World countries that wished to avoid default and gain access to easier credit.

Thus, developed countries initially responded to the possibility of a Third World debt default by treating it as liquidity problem and providing new loans directly and through multilateral agencies. As part of the deal, commercial banks were to continue lending. In the early stages, IMF loans did not necessarily require structural adjustment since authorities believed that indebted nations needed some time to get their finances in order. Under IMF programs, countries thus raised taxes and tariffs and reduced government expenditures. By 1985, it had become obvious that deep-rooted problems in developing country economies were preventing them from growing out of their debt. A new strategy was then announced by U.S Treasury Secretary James A. Baker whereby new money from the IMF and commercial banks would be based on market reforms. In exchange for that money, indebted countries were to liberalize their economies.

By 1987, it became evident that that strategy was not working. Countries did very little in the way of economic reform, though they continued to receive funding. Banks, though they continued to lend, were reducing their exposure in the region. A slow transfer of private debt to public debt was occurring in the absence of a resolution to the underlying problems that caused the debt crisis.6

IMF conditionality appeared to provide little incentive to reform. Sebastian Edwards, formerly the World Bank’s chief economist for Latin America, referred to the IMF as “participating in a big charade,” because fund programs imply that “there is a high probability that the country will attain balance-of-payment viability in the near future. For many countries this is not the case and everybody knows it.”7

By financing governments that were uninterested in serious liberalization and structural adjustment, the Fund actually delayed reforms in Latin America during the 1980s. Latin America became more indebted, private commercial banks in the United States were able to postpone recognizing losses, and the living standards of Latin Americans fell. MIT economist Rudi Dornbusch noted that “The IMF set itself up to save the system, organizing banks into a lender’s cartel and holding the debtor countries up for a classical mugging.”8 As Anna Schwartz commented in her 1988 presidential address to the Western Economics Association, “the intervention of the official players has prolonged and worsened the debt problem.”9 Peter Lindert found that as a result of IMF intervention, “Most [debtor nations] have participated in a three-party stalemate, in which official agencies, private creditors, and debtor countries agree, after repeated struggles and much uncertainty, to reschedule in a way that postpones large net resource flows.”10

The end of the 1980s and beginning of the 1990s finally did see the introduction of far-reaching market reforms, a development for which the IMF often takes credit. But that outcome resulted from economic necessity in the wake of the collapse of development planning and of other failed economic policies. By the end of the 1980s, the banks ultimately opted for a solution that some prominent figures in the financial community had recommended early on: banks set aside loan-loss reserves and wrote down the value of their loans. If banks were indeed threatened by the Third World debt situation, they could have borrowed directly from the U.S. Federal Reserve Board at a penalty. It was always questionable for the IMF to attempt replicating the role of a lender of last resort. For that reason, the Federal Reserve Bank of Minneapolis has recently noted that “the IMF is redundant to prevent worldwide financial crises,” and “should cease its lending activities altogether.”11

The IMF’s move in the opposite direction probably contributed to creating the lost decade of the 1980s. As economists Lindert and Peter Morton noted in 1987, “the intervention of the Fund and the [World] Bank has impeded the striking of bilateral bargains between debtor governments and the creditor banks.”12 Shortly before joining the IMF, MIT professor Stanley Fischer expressed apparent agreement with that assessment: “I believe that the debt crisis would have been over sooner had the official agencies not been involved.” Fischer added, however, that he thought that in the absence of official intervention the adjustment crisis would have been deeper.13 But it is hard to imagine that Latin America would have suffered more had the liberal reforms that were eventually introduced in the late 1980s and early 1990s been implemented seven or eight years earlier.

Recent History

With the collapse of socialism, IMF lending again began to rise and has shot up significantly after the fall of the Mexican peso in 1994 (see graph). The fund thus took on the role of turning formerly socialist countries from central planning to the market. The creation of a new Systemic Transformation Facility in 1993, through which countries like Russia would borrow on easier terms than available under standard IMF programs, exemplified the fund’s new initiative. The Mexican peso crisis of 1994-95 initiated a new era of massive emergency lending by the IMF to countries experiencing financial crises.

In fact, there have been at least 90 severe banking crises in the developing world since 1982 and the losses in 20 of those cases have ranged between 10 and 25 percent of GDP.14 A number of flawed policies in the countries experiencing those banking problems are, of course, to blame for image
creating such financial disasters. Principal among those policies is the explicit or implicit guarantee by developing country governments that they will rescue domestic banks if they get into trouble. The result has been the creation of moral hazard at the national level. The more governments rescue unhealthy banks and their business associates, the riskier we can expect banks to behave.

The IMF has complicated the situation by adding moral hazard at the international level. When the Mexican peso fell in 1994 as a result of expansionary fiscal and monetary policies that were inconsistent with its pegged exchange rate, moral hazard was already well established. In 1995 the IMF and the U.S. Treasury decided, for the fourth time in 20 years, to rescue the Mexican government and investors from the consequences of irresponsible election year policies.15 The bailout allowed Mexico to repay in full about $25 billion worth of dollar indexed bonds. Those investors suffered no risk or losses but were able to pass the bill on to ordinary Mexicans in the form of greater debt. The redistribution of wealth from the relatively poor to the relatively well off has been a feature of subsequent bailouts. The U.S. Treasury and the fund have claimed Mexico a success, but the intervention there sent a signal to the world that if anything went wrong in emerging economies, the IMF would come to investors’ rescue.

Moral hazard helps explain the doubling of capital flows to East Asia in 1995 alone. Governments that have received emergency IMF aid since then were not discouraged from maintaining flawed policies as long as lenders kept the capital flowing-which lenders imprudently did with the knowledge that official aid would be used in case of financial troubles.

To be sure, the financial crises in Mexico, Asia, Russia and Brazil were not necessarily created by the IMF; rather, they were a product of poor domestic policies to which IMF lending contributed. (In Asia, those policies included borrowing in foreign currencies and lending in domestic currency under pegged exchange rates; extensively borrowing in the short term and lending in the long term; lack of supervision of borrowers’ balance sheets by foreign lenders; and shaky financial systems. Mexico, Russia and Brazil shared many of those misguided policies plus unsustainable fiscal deficits.)

IMF interventions in crisis countries have been justified on the grounds that fund conditionality promotes market reforms and that emergency lending helps avert systemic threats to the international economy. Yet the impact of conditionality is at best ambiguous and has clearly not worked in Russia since 1992 or Indonesia since 1997. Moreover, the record of IMF lending does not speak well either of the temporary nature of IMF loans or of IMF conditionality-70 countries have depended on IMF credit for 20 or more years. (See Appendices A and B).16 As the graph below shows, once a county receives IMFcredit, it is likely to become dependent on fund aid for most, if not all, of the following years.


Major factors undermining the effectiveness of the fund are its inability to enforce loan conditions and its institutional incentive to lend. Especially when it lends to governments uninterested in reforms, the fund’s credit does little to promote policy or structural changes. In such cases, the fund suspends credit until it receives promises or evidence of policy change in the right direction. At that point, the fund releases credit again taking the pressure off of the recipient government to reform. Indeed, the fund cannot afford to watch a country reform on its own without the IMF’s involvement. That scenario has occurred numerous times in post-Soviet Russia, certainly delaying the country’s transition to the market. It has also prompted leading Russian liberals to denounce IMF aid to Moscow. In a July 1999 letter to IMF managing director Michel Camdessus, for example, former Russian Deputy Prime Minister Boris Federov warned against a new loan to Russia, which was subsequently approved: “I strongly believe that IMF money injections in 1994-1998 were detrimental to the Russian economy and interests of the Russian people. Instead of speeding up reforms, they slowed them.”17

The fund’s incentive to lend is well known both by the fund and by recipient governments, thus making the fund’s conditionality much less credible. Even in Asia, where some countries show signs of economic improvement, the record is ambiguous. As The Economist magazine recently pointed out, the recovery there reflects “the natural propensity of economies to bounce back… . What it does not reflect is fundamental, structural reform, in any country in the region.”18

Have recent events supported the case for official intervention on the grounds that IMF aid is necessary to resolve financial crises and avert contagion? Only after Brazil’s currency crisis-two months after the IMF bailout-did the government there take tentative steps to address the country’s problems. The collapse of the Russian ruble in August 1998 and subsequent debt default-which an IMF bailout did not prevent-rattled world markets and likely reduced moral hazard. Successive interest rate cuts by the Federal Reserve and other central banks then helped calm world markets, raising questions about the utility of the IMF in both preventing defaults and dealing with their global effects. Indeed, the Brazilian devaluation did not have the colossal consequences that the IMF and U.S. Treasury predicted, probably in large part due to the effects of the Russian crisis.

Market discipline has also been at work in Korea. As Jeffrey Sachs notes, the fund responded to the Korean crisis by providing a tranche of credit in late 1997, but “The Korean debacle ended only when Korea ran out of IMF money, forcing the international bank creditors to agree to roll over the debts owed by Korean banks.”19 Even so, the restructuring was “far from ideal” since the newly restructured debt was generously guaranteed by the Korean government at higher interest rates than that of the original debts.20

Morris Goldstein also questions the IMF approach in Korea. According to him, it is not “clear that the first round of rollovers that did take place … would not have happened anyway in the absence of a promise of accelerated disbursements from the official sector. The argument that creditors are too numerous and dispersed to make such discussions feasible did not seem to apply in this case. If the rescue package for South Korea were smaller … and disbursements were not accelerated, a larger amount would have had to be rescheduled.” Moreover, losses in Korea would not have made Western banks insolvent.21

Where We Are Now

Recent experience has prompted initiatives at the fund to improve its transparency and its surveillance function, do more to prevent financial crises from erupting, and find ways to “bail in” the private sector. The IMF’s role as a surveillance agency has been seriously tarnished by the Asian crisis. The fund provided no warning about the impending collapse of currencies and domestic banking systems and instead lauded the East Asian economies in public documents shortly before the outbreak of the crises. The financial community has not been comforted by the fund’s claims that the governments it was dealing with were not transparent (which would amount to an admission that the fund did not fully know what was going on) and that the agency did in fact provide warnings to officials in Thailand but kept that information confidential. That episode only highlights an inherent conflict in the fund’s role as both a credit rating agency for countries and an agency that attempts to prevent the eruption of financial turmoil. For if the IMF did detect alarming economic conditions in an emerging economy, the public release of that information would precipitate a crisis; not sounding the alarms, however, would further undermine the IMF’s credibility as a surveillance agency. As long as the IMF pretends to maintain both roles, that conflict will continue to exist.22

The IMF has also established a new facility this year to provide bailout funds before countries experience financial difficulties in an effort to prevent them. The Contingent Credit Lines (CCL) program is intended to serve as “a precautionary line of defense”23 that would stave off creditor panics in countries with fundamentally sound economies. But the fund has not shown good judgment in determining what countries would benefit from preventive bailouts. The two times the fund has provided such aid-Russia (July 1998) and Brazil (November 1998)-the bailouts failed to prevent currency devaluations and financial crises. Such funds merely became gifts to speculators and financial institutions, leaving both countries in greater debt.

The CCL, moreover, is based on the dubious assumption that countries with sound economic fundamentals are subject to contagion. It is difficult to find a country, however, that has succumbed to crisis that did not also already have severe domestic economic problems. Every country that has experienced financial turmoil has had fundamentally flawed economic policies. Systemic crisis has not spread to countries with sound economic fundamentals. Had the CCL existed since 1997, it is not clear which country it would have saved. The CCL is likely to expand moral hazard without providing any tangible benefits.

Finally, the IMF has reacted to criticisms that its lending has created moral hazard by exploring ways to make private-sector lenders bear more of a burden when resolving financial crises. Naturally, private sector creditors object to any moves by the IMF that would encourage debtor countries not to meet their debt obligations in full and on time. The fund practice of reducing debt by lending to countries that have not yet completed debt renegotiations with their private sector creditors would essentially force losses on private lenders who have so far benefited from bailouts. Yet any such initiatives carry a high risk. Official efforts to bail in the private sector may precipitate the financial turmoil they were designed to prevent since lenders would have an incentive to pull out of a country whenever they sensed that international authorities will force losses on them.24

More Flexibility Is Needed

There is a better approach to reduce moral hazard and prevent crises from occurring that avoids the conflicts the IMF now faces. When crises occur, direct negotiations between creditors and debtors as an alternative to IMF lending and mediation would lead to an automatic “bailing in” of the private sector. Indeed, private creditors would already be bailed in and have every incentive to renegotiate debts, perhaps including a certain amount of debt forgiveness, as would developing country debtors have every incentive to reach a reasonable agreement with creditors as quickly as possible. Recent and historical experience shows that direct creditor-debtor bargains have often overcome problems of coordination, information and insurance to both deal with financial turmoil when it erupts and to help prevent it.25 In the 19th and 20th centuries, bondholder committees, banking clubs and banking syndicates have formed to provide those so-called public goods upon which much IMF lending is now justified.

As University of Chicago economist Randall Kroszner has noted, in the new international financial architecture, less is more. Less bailout capacity and less inflexibility on the part of creditors can produce a more stable global financial system.26 Less fund involvement would also produce more market innovations that are currently precluded by official interventions. Standby lines of credit could be provided by banks to countries in the case that they be negatively affected by outside shocks. Since banks would not provide such a service to all countries, the very provision of such loans would increase investor confidence and (unlike the case of IMF lending) serve as useful signal to the markets as to which countries can be expected to have more sound economic fundamentals in place.

The market for credit risk insurance and restructuring insurance could also develop to meet the needs of market participants’ diverse tastes for risk. The severity of financial turmoil would thus be reduced because not all investors would behave the same way in times of crisis.27 In the case of sovereign bonds, contracts could include clauses concerning majority voting or workout procedures in the event of default. Again, official lending has made that outcome less probable. Even so, since the 1980s, there have been several cases of successful voluntary sovereign bond restructurings that have overcome problems requiring the unanimous consent of bondholders.28

Relying more on debtors and creditors to resolve and prevent financial crises would reduce moral hazard, increase the quality of surveillance, and lead to innovations that would reduce the severity of crises. Private creditors would be responsible for bearing the full consequences of their investment decisions, a system of real conditionality and real reform would evolve, and the parties involved would have no incentive to stall the process of debt resolution. Taking an approach based on direct creditor-debtor bargains requires that countries be allowed to default. The very possibility of default would, in many cases, help lead to debt renegotiations, thus improving the position of both debtors and creditors. As Koszner puts it, “it may indeed be better to forgive than to receive. Asking for less can result in receiving more while also making the distressed country better off.”29 Default itself need not be traumatic. Indeed, history shows that defaults have repeatedly occurred, but they have usually been partial rather than complete and lending often resumes soon after the defaults. In short, relying on direct creditor-debtor negotiations would eliminate many of the conflicts now facing the IMF.


The world economy has changed dramatically since the creation of the IMF. The fund long ago lost its original mission but has taken on new functions since the collapse of the Bretton Woods system. The rise of IMF lending and crisis mediation since the early 1980s reflects, in large part, the development of a dysfunctional relationship between lenders and borrowers in international finance. Repairing that relationship requires that moral hazard be reduced and that crisis prevention and management be more effective. The IMF’s new initiatives to deal with crises, however, are likely to be ineffective. An approach based on greater reliance on two-party negotiations holds more promise in stabilizing the international financial system than the current approach, in which the IMF too often becomes a burdensome third party.

Appendix A

Use of IMF Credit by Eligible Countries, 1947-99

Country First Year Used Number of Years Used Percentage of Years Used after Year of First Use
Algeria 1989 11 100.00
Benin 1978 22


Burkina Faso

1978 22



1968 29



1974 26


Central African Republic

1974 26



1970 30


Congo, Dem. Republic

1972 28


Congo, Rep. Of

1977 23


Equatorial Guinea

1980 20



1949 22



1978 20



1977 23



1962 35



1969 31



1979 21


Ivory Coast

1974 26



1974 26



1977 23



1963 33



1974 26



1975 25



1964 35



1976 24



1969 15



1968 24



1987 13



1983 22



1966 26


Sao Tome and Principe

1989 11



1975 25


Sierra Leone

1967 29



1964 26


South Africa

1976 15



1958 39



1979 12



1974 26



1976 24



1964 25



1971 29



1971 29



1981 18


Use of IMF Credit by Eligible Countries, 1947-99 (continued)



1964 13



1972 28



1972 28



1981 10



1974 12



1949 44



1956 29



1975 25



1976 7



1991 9



1967 24



1976 24



1965 35


Papua New Guinea

1976 24



1955 39


South Korea

1974 17


Salomon Islands

1981 10


Sri Lanka

1961 39



1976 18



1977 23


Western Samoa

1975 17




1992 8



1994 6



1995 5



1993 7


Bosnia and Herzegovina

1995 5



1991 9



1993 7



1974 11


Czech Republic

1993 1



1990 3



1992 8



1994 6



1982 16



1993 7


Kyrgyz Republic

1993 7



1992 8



1992 8


Macedonia, FYR

1993 7



1993 7



1990 5



1973 25



1992 8


Slovak Republic

1993 7



1993 4



1953 40




1995 5



1949 41


Use of IMF Credit by Eligible Countries, 1947-99 (continued)

Middle East


1957 40



1955 6



1967 2



1957 18



1971 22



1960 16


Yemen Arab Republic

1983 5


Yemen People’s Dem. Rep.

1974 16


Republic of Yemen

1990 5


Western Hemisphere


1957 33



1977 19



1983 8



1959 35



1951 33



1957 38



1954 16


Costa Rica

1961 30



1979 20


Dominican Republic

1960 36



1957 29


El Salvador

1956 23







1962 18



1971 27



1958 39



1957 33



1973 27



1976 22



1957 33



1968 28



1956 5



1958 31


St. Lucia

1980 6


St. Vincent

1980 6


Trinidad and Tobago