14 IMF Governance and Financial Crises with Systemic Importance

Ruben Lamdany, and Leonardo Martinez-Diaz
Published Date:
July 2009
  • ShareShare
Show Summary Details

Crises in systemically important countries, whose failure is likely to lead to contagion with far-reaching consequences for the international financial system, pose acute governance problems because of the stakes involved, the amount of resources that must be mobilized, and the need for rapid decision making. During crises in systemically important countries, the locus of effective decision making shifts outside of the formal organization of the IMF to forums that better reflect the international distribution of resources. Because crisis management involves high-level political decisions about exceptional access to Fund resources, and because it may involve matching official financing or coordinated pressure on financial institutions to extend private financing, G-7 Deputies (deputy ministers of finance with responsibility for international issues) usually play a central role. At the same time, the formal decision-making system (based on universal participation and representation of the membership), in which Management formulates proposals and the Executive Board approves them and exercises oversight, is necessary to provide the IMF’s legal authority.

This mixture of formal and informal governance imposes costs as well as offering benefits. Informal governance allows for flexibility and speed, but the combination of the informality of shareholder participation and the high level of secrecy that surrounds Fund decision making undermines ex post accountability. As a consequence, the IMF is unable to defend itself from external criticism by pointing to a clear historical record. The absence of effective ex post accountability also leads to three other important governance problems: (1) staff faces incentives that undermine the quality of the analysis that it presents to the Executive Board; (2) shareholders are able to exercise substantial influence over the content of conditionality that is not subject to scrutiny; and (3) Management does not resist shareholder pressures to relax the enforcement of conditionality in particular cases.

The six crisis cases reviewed for this paper1 demonstrate some important common features of informal IMF governance as well as significant variations. Each of the G-7 countries played different roles in the various cases, reflecting the different interests at play in each. In addition, the way in which informal consultations took place evolved over time, in part in response to the crises themselves, and in part because of leadership changes. These variations highlight the fact that the Fund’s informal governance takes different modalities in each crisis.

Although the substance of conditionality in programs in support of systemically-important countries tends to be delegated to the Fund, there are also variations across cases, and particular G-7 governments became more intensely involved in some countries than in others. In each of the cases reviewed, the U.S. had a dominant role within the G-7, in particular in setting structural conditionality in Indonesia and Korea, or in the case of supporting the preferred Brazilian exchange rate regime in 1998. In most cases, however, U.S. preferences did not differ significantly from those of other G-7 countries or from the strategy preferred by the Fund.

These case studies reveal governance problems in four important areas: surveillance, access to Fund resources, design of conditionality, and enforcement.

  • Surveillance. Every member country is visited, according to a regular schedule, by an IMF staff surveillance mission, after which staff prepare a report to the Executive Board. A major focus of this exercise in emerging markets is to detect early warning signs of developing crises and recommend corrective action before a crisis becomes full blown and requires drastic policy corrections. The Executive Board’s ability to identify and head off potential crises is undermined because, for systemically important crises, staff does not send clear signals about country performance. On one hand, the Board is not considered secure enough to receive confidential information. On the other, shareholders resist critical assessments of their countries, and staff is required to support the view of management in front of the Board, which constrains its ability to discuss many issues openly with the Board. Staff is also hesitant to confront member countries with critical analysis, especially when the country is in a vulnerable position that would make it react strongly. In all six cases, staff reports failed to issue early warnings that could be used to prepare for financial crises.

  • Access to Fund Resources. Ordinary IMF lending is covered by rules that limit access to Fund resources to 300 percent of a country’s quota, and a supplementary set of rules for extraordinary access. During urgent crises in systemically important countries, however, these regulations are waived, and are replaced by a very non-transparent process of informal bargaining.2 In some cases, it is very difficult to reconstruct exactly how the degree of access was finally decided. The probability that the program succeeds in stemming a crisis is generally an increasing function of the resources committed, and the amount of adjustment required in the short run is a decreasing function. The country seeking assistance requests a figure that it believes politically feasible. Political approval for access to Fund resources is secured before a program comes before the Board, but this is nevertheless the most controversial stage of Board discussions, and the one most likely to lead to abstentions. In the end, the size of financing in crises is determined by informal consultations with the leading shareholders—usually involving a conference call among the G-7 deputies—with the United States playing a central role in each case.

  • Conditionality. The Executive Board does not play a direct role in designing the conditionality in individual programs, but the fact that the Board does not formally amend conditionality does not mean that it exerts no influence. The Board exerts indirect influence over conditionality because possible objections are usually anticipated and because future programs can take account of the issues that are raised. Management is careful not to bring a program to the Board before ensuring through informal contacts that the program enjoys a comfortable majority. In practice, this means that Management has to assure itself of the support of the G-7. Informal intervention by G-7 deputies and their Executive Directors in the design of conditionality reached a high point during the Asian crisis. Although formal procedures are in place to safeguard staff autonomy, shareholders are able to exercise substantial informal influence over the content of conditionality that is not subject to scrutiny, as in the cases of Indonesia and Korea.

  • Enforcement. The Executive Board formally approves all disbursements of Fund resources, and when performance criteria are not met by the program’s review date, the disbursement is withheld unless the Board decides to issue a waiver or modify the conditions. However, it is a frequent occurrence that major shareholders use their informal influence to urge Management to propose a waiver. Since the default outcome is that the program will be suspended when a performance criterion is not met, shareholders’ informal influence has the effect of relaxing the enforcement of conditionality. This weakens the IMF’s credibility as an arbiter of sound policies and dilutes market discipline.3 Weak enforcement of conditionality was pervasive in the cases of Russia and Argentina, and laid the groundwork for the crises that occurred in 1998 and 2001, respectively.

Findings and Conclusions

Informal practices have arisen to reconcile the need for decisive action and confidentiality with the existing formal institutional framework. As a practical constraint, during systemic crises management, decision making needs to adjust to give a dominant role to those who are able to promise substantial infusions of official financing and who have access to the private financial institutions most likely to help resolve crises. A key symptom of the cost to IMF credibility is the fact that IMF lending to countries with systemic importance fails to generate catalytic effects that mobilize private capital flows, but is instead often complemented by “concerted” lending by private institutions in creditor countries.

Although the Fund has assimilated numerous lessons from the experiences of the six crises reviewed for this paper (most of which have been articulated by outside observers—the emphasis on better data standards, more transparency, streamlined conditionality, an exceptional access framework, contingent credit lines, and proposals for new frameworks for dealing with sovereign debt), it has not come to grips with the fundamental governance problems that make the Executive Board an ineffective locus for surveillance. An important part of the problem is that the secrecy that surrounds IMF decision making makes it difficult to hold the institution or particular individuals inside or outside the Fund accountable for the roles they played leading to the crises as well as in crisis management. Ex post accountability is particularly important in crisis cases because ordinary procedures have to be accelerated and informal procedures come to the fore, and this is a blind spot of IMF governance.

In response to external criticism, the Fund has in recent years greatly increased the number of documents that it makes public, but information on the details of its decision making is not disclosed to the outside world or distributed within the organization, and this undermines its ability to learn from its own experience. Furthermore, formal mechanisms for expost accountability are weak, and this is an area where informal mechanisms cannot help. Stronger ex post accountability would reinforce incentives for staff to provide candid analyses, and would improve the Fund’s ability to resist pressures to engage in risky lending or to modify programs in ways that serve the interests of only a subset of the membership. The experience of other organizations, such as central banks, indicates that publicity is the most effective protection for institutional autonomy. Therefore, the paper concludes with the following recommendation:

Formal changes in IMF procedures will have limited effectiveness in improving the management of systemic crises. Lasting improvements in institutional governance in general, and particularly in regards to crisis management require improvements in transparency. Documents that are no longer operational should be reclassified for public use and internal memos and correspondence should be transferred to the archives.


This Executive Summary is derived from an IEO Background Paper that makes extensive use of highly classified material and cannot therefore be published given the conditions under which access to that material was provided. In particular, the author had access to all of the background material used to prepare earlier related IEO reports (IEO 2003, 2004). The author wishes to thank former and current IMF officials for their cooperation in providing interviews and access to documents, the IEO for its support and valuable input from its staff, and Borislava Mircheva and Roxana Pedraglio for valuable research assistance. In addition, Leonardo Martinez-Diaz, Thomas Bernes, Nils Bjorksten, Mariano Cortes, Iqbal Zaidi, and Borislava Mircheva provided valuable comments. Any errors are the author’s.

Mexico (1995), Indonesia (1997), Korea (1997), Russia (1998), Brazil (1998), and Argentina (2001).

In response to this experience, in 2002 the Fund introduced the Extraordinary Access Framework to attempt to reconcile the formal rules with the informal practice.

A quantitative study of 26 post-Communist countries demonstrate that countries that received substantial amounts of U.S. foreign aid were subject to much shorter program suspensions when their programs went off track. They received waivers or their conditions were modified so that they could quickly get back into good standing. As a result of the weak incentives that they faced, their economic policies were more inflationary, and they failed to implement conditions and went off track more frequently. Aid from other OECD countries had no effect (Stone, 2002). A study of 53 African countries reveals a similar pattern with respect to U.S. foreign aid, and also indicates that members of the CFA Franc zone are treated in the same way, particularly if they vote with France in the UN General Assembly, and that members of the Commonwealth receive similar treatment (Stone, 2004).

    Other Resources Citing This Publication