Chapter 1 Introduction

Kevin Barnes, Ali Mansoor, Benjamin Cohen, and Shinji Takagi
Published Date:
September 2003
  • ShareShare
Show Summary Details

The decade of the 1990s saw a succession of currency crises in emerging market economies, against the background of the increasing integration of these economies with global capital markets. These crises were preceded by large private capital inflows and triggered by sudden shifts in market sentiment, which led to massive capital flow reversals. They are often described as capital account crises to distinguish them from the more conventional crises which have their origins mainly in the current account. The IMF was called in to help in several cases, and its role has been the subject of much study and comment. Contrary to the expectation that IMF support would serve to certify the effectiveness of an adjustment program and help achieve a smooth adjustment, many of the IMF-supported programs failed to achieve their initially stated objectives. Capital outflows continued, leading to severe exchange rate depreciation and, in some cases, an exceptionally large contraction in output. Not surprisingly, the IMF was widely criticized both for its failure to anticipate vulnerabilities through surveillance and for the subsequent failure to restore market confidence quickly.

This evaluation seeks to throw light on the role of the IMF in three capital account crises, in Indonesia (1997–98), Korea (1997–98), and Brazil (1998–99). In undertaking this evaluation, we recognize that we are entering into grounds that are unusually well-trodden. These crises have been extensively studied by numerous outside observers and also by IMF staff. A number of lessons have been learned and many corrective steps have been taken in the form of revised IMF policies and procedures, as well as broader initiatives related to the international financial architecture. Nevertheless, it is appropriate that the Independent Evaluation Office (IEO) should revisit these cases in order to provide an independent assessment. In keeping with the IEO’s terms of reference, the principal focus of the evaluation is to draw lessons for the IMF in its future operational work. It will also contribute to transparency by evaluating the internal processes by which important decisions were made.

Three aspects of the evaluation that limit the scope of its conclusions must clearly be stated at the outset:

  • (i) Any evaluation necessarily benefits from hindsight. This can be useful in drawing lessons for the future but, in evaluating the past and especially determining accountability, it must be kept in mind that much of what we know now may not have been known to those who had to make the relevant decisions. It is important to distinguish cases in which critical information was not available from those in which the wrong conclusions were drawn from the available information. In the former case, the evaluation should highlight gaps in data availability which need to be corrected. In the latter, it may suggest a need to reexamine and improve analytical approaches and assumptions.

  • (ii) To be meaningful, evaluation of an IMF-supported program must imply comparison with an alternative set of policies that may have produced better results. However, it is extremely difficult to establish rigorously such a counterfactual. This is especially so in areas where there is lack of consensus in academic and policymaking communities. We indicate areas where this appears to be the case, and the learning process in such cases must proceed on the basis of best judgment.

  • (iii) The behavior of an economy is always subject to uncertainty, but the uncertainties are much greater in crises. A program cannot be judged to represent mistaken decisions ex ante just because it failed to restore confidence as envisaged. The relevant criteria for judging such decisions ex post are: (1) was there a reasonable ex ante assessment of the probabilities, with the information available at the time; (2) could more useful information have been obtained if different procedures had been used; and (3) could different policies have enhanced the probability of success. These problems are especially difficult to handle if the crisis involves the possibility of multiple equilibria where it is difficult to predict the circumstances under which one or the other equilibrium can come into being.

The evaluation makes extensive use of primary information made available to the IEO. This includes staff reports for Article IV consultations,1 briefing papers and back-to-office reports for staff missions and visits, internal memoranda exchanged among staff or between staff and management, minutes of Executive Board meetings, comments by management and review departments on briefing papers, and policy papers prepared by staff for the Board.2 The IEO, however, is not given automatic access to documents that are purely internal to management or that cover management’s exchanges with national authorities, except when such documents were shared with staff.3 Inevitably many policy decisions during the crises were made by management in close consultation with its major shareholder governments and the records available to us do not cover these consultations. Our judgments on certain policy matters are therefore based on limited information.

The evaluation team has extensively interviewed those involved in decision making in the IMF (including former IMF staff and management) as well as some current and former officials of member countries. Statements made in the text about positions or views of IMF staff and management are based on the evidence from internal documents and interviews. The team has also interacted with a number of individuals who have expressed views on the IMF’s role in these cases. The list of those interviewed by the evaluation team appears in Appendix 2.

The report comprises two parts. The main report presents a summary of our major findings on the role of the IMF in the precrisis surveillance phase and the crisis resolution phase in each country and our recommendations. The annexes contain three detailed country case studies that form the basis for our judgments in the main report.

The main report is organized as follows. Chapter 2 presents a brief overview of the IMF’s involvement in Indonesia, Korea, and Brazil. The subsequent three chapters summarize major findings from the country case studies. Chapter 3 presents our assessment of precrisis surveillance. Chapter 4 discusses our assessment of the IMF experience in seven central areas of program design and implementation, that is (1) the macroeconomic framework and projections, (2) fiscal policy, (3) monetary policy, (4) official financing and private sector involvement, (5) bank closure and restructuring, (6) structural conditionality, and (7) communications strategy to enhance ownership and credibility. Chapter 5 addresses internal governance issues within the IMF. Finally, Chapter 6 presents conclusions and recommendations.


Under Article IV of the Articles of Agreement, the IMF holds consultations, usually every year, with each of its member countries on the country’s economic policies and potential vulnerabilities. This “surveillance” function of the IMF is conceptually distinct from its role in providing financial support for adjustment programs.

Some of these Board policy papers have been published, including on the IMF’s website. These papers are cited in footnotes except when they are also available in print form, in which case they are listed in the bibliography.

Management refers to the group of senior IMF officials consisting of the Managing Director, the First Deputy Managing Director, and two Deputy Managing Directors.

    Other Resources Citing This Publication