Chapter 5 Internal Governance

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

The evaluation of experiences in the three cases studied reveal some important lessons relating to internal process issues. These involve human resource management, the role of major shareholders and the Executive Board, and relations with other international financial institutions. Many of these issues are general in nature and also arise in other cases.

Human Resource Management

Our evaluation revealed a tendency for the sharper, more candid elements of a diagnosis to be diluted in final Board papers—whether in the context of an assessment of vulnerabilities during the surveillance process or judgments of the potential risks and the probability of success in program-related documents. This problem, which has been noted in other contexts including in the recent IEO evaluation report on prolonged use of IMF resources, raises the issue of greater internal incentives to encourage frank presentations of problems. Interviews with staff members indicated a perception among some that it was difficult to make assessments on issues that were inevitably of a probabilistic nature and could not, therefore, be easily proved or disproved, especially in the short term. They feared that efforts at candor were unlikely to be supported fully within the institution if the authorities concerned were to object strongly.1

Second, APD’s staff was overstretched by the crises simultaneously occurring through the region, but the IMF’s system of internal budgetary and human resource management delayed the reallocation of resources to APD. A reallocation did eventually occur, but only once the crises were already well under way.

Third, there was a tendency to split responsibilities without clear lines of command, as manifested in the insufficient integration of APD and MAE in their country work during the crises. In particular, staff with special expertise should have been integrated more fully into the negotiating missions. The lack of full integration was most costly in the case of Indonesia. The idea of having a single MAE/area department team in crisis situations has been noted in a recent review of MAE by a Managing Director–appointed panel of outside and inside experts.2 This review has resulted in a broader reorganization of MAE, one of the aims of which is to provide a strengthened center of expertise responsible for banking crisis management and resolution issues.

Fourth, available internal knowledge was not fully used in formulating the programs, particularly in Indonesia and Korea, in part owing to the reorganization of the Asia-Pacific operations of the IMF in early 1997.3 Only a relatively small number of participants in the missions, including those assigned from outside APD, had previous experience with Indonesia or Korea. Although the problems were less pronounced in Brazil, because of the continuity maintained at the senior level, short tenure also characterized staff assignments with that country in both the surveillance and the program phases. These examples are a reflection of a broader problem with the excessive turnover of country teams within the IMF, as has previously been noted by a report of the Office of Internal Audit and Inspection as well as by the IEO’s evaluation of prolonged use of IMF resources (IEO, 2002).

While these managerial issues need to be tackled for the sake of improving performance, however, most of the weaknesses in program design and implementation identified by the evaluation did not arise primarily from human resource management problems. Thus, the evaluation team does not believe that these issues fundamentally altered the outcome of the programs.

The Role of Major Shareholders and the Executive Board

The need to respond quickly to deal with the crises required close collaboration of staff and management with the Executive Board, particularly in the cases of Indonesia and Korea where the accelerated procedures under the Emergency Financing Mechanism were invoked. Frequent informal sessions served to facilitate a flow of information, and provided Executive Directors with opportunities to voice their inputs into the program at different stages. Such close consultation was necessary for the Executive Board to fulfill its governance role in these large-access cases, in which political judgment played an even greater role than usual and speed was critical.

The major shareholders also interacted directly with management during the negotiation phase on what should be the key elements of program design and also with the authorities in the country concerned. This involvement is entirely understandable and appropriate given the exceptional size of access involved and the concern about possible systemic effects, the fact that any strategy is risky, and also the fact that bilateral support may have to be provided. In the case of Korea, the close involvement of the United States in the earlier stages probably facilitated the later U.S. decision to take a leadership role in organizing a rollover agreement among international banks. Likewise, it was the close earlier involvement of the other major shareholders that allowed them to respond promptly to that U.S. initiative by exercising moral suasion on banks based in their countries.

However, in order for the IMF to undertake its role as crisis coordinator effectively, two elements are critical. First, Executive Directors (and, through them, key shareholders and other potential sources of official financing) need to be given candid assessments of the probability of success of the proposed strategy, including frank feedback when parts of the strategy favored by some shareholders lower this probability. Second, it is important that the technical assessments of the staff and political judgments by the Executive Board not be blurred. It is legitimate and important for the Executive Board and shareholders to communicate their expectations to management and also to interact with management on what might be the contours of an acceptable program. In certain situations, shareholders concerned with an evolving crisis may wish to deal directly with the authorities, as the authorities may also wish to deal directly with them, and there were examples of such interactions in all three cases. However, any appearance of shareholders dealing directly with IMF missions in the field can be mis-interpreted.4

In the case of Indonesia, interviews with staff and internal documents indicate that there was extensive feedback from members of the Executive Board on the need to strengthen structural conditionality. This was not inconsistent with the framework envisaged by the July 1997 guidance note, which explicitly stated that the IMF “should collaborate with other multilateral institutions and donors in addressing economic governance issues” and also endorsed use of informal channels of interaction with Executive Directors to keep them “informed on a timely basis of developments in significant cases involving governance issues, including those in which third parties’ governance concerns have implications for program financing.”5 However, our assessment reveals that this feedback from the Board may have contributed to the excessive structural conditionality built into the Indonesian program. This suggests that, while greater involvement by the Board in these cases is appropriate, ways must be found to ensure that it does not lead to micromanagement of operational details.

The Relations with Other International Financial Institutions

In its role as crisis coordinator, the IMF supplemented its own resources with additional financing from other IFIs, including the World Bank, the ADB, and the IDB, and also drew upon the analyses of these institutions in specific areas of their expertise. The relationship was not always smooth, however, and public disagreements sometimes erupted, developments that could not have been supportive of the efforts to restore confidence.

Very little difficulty arose in this respect in Brazil, where both the World Bank and the IDB worked almost exclusively in the social sector. In Asia, the working relationship with the World Bank and the ADB was more difficult, as all three institutions worked in the financial sector and their areas of responsibility necessarily overlapped. While good working relationships eventually developed as the areas of responsibility became more clearly defined over time, much depended on the personalities of the mission members. The lack of an effective mechanism to resolve differences of view led the ADB to suspend temporarily its collaborative relationship with the IMF in Indonesia in late January 1998 because of a disagreement over the establishment of the IBRA.

This experience suggests that when future arrangements call for similar collaborative efforts with regional development banks, it is important that the terms of reference for their engagement in IMF-supported programs be agreed at the very outset, so that there is a clear understanding of the demarcation of responsibilities. Staff from these IFIs should be given access to all relevant information that is at the disposal of the IMF and be invited to comment on the content of the program in areas where these institutions have particular expertise and are expected to provide financing.6 A procedure should also be established to resolve any difference of views, so that all relevant IFIs can speak with one voice on matters of substantive policy.

In the case of IMF–World Bank collaboration, there were significant frictions in the case of Indonesia. The IMF initially obtained information from the World Bank as inputs into structural conditionality, without having the Bank staff’s direct involvement in the drafting and negotiation of the program documents. Given its preference for more direct involvement, the January 1998 program ensured that the World Bank, and especially its Indonesia-based staff, was actively involved in formulating the detailed structural conditionality. In the future, it will be necessary to have a clearer understanding on the role of the World Bank in the structural component of an IMF-supported crisis-management program. The managements of the IMF and World Bank have already acted to put in place strengthened procedures.7

Despite the active involvement of World Bank staff in the IMF-supported programs in Asia, there was public criticism of the IMF strategy (especially on fiscal and monetary policy) from the Chief Economist of the World Bank, which attracted considerable attention. It is relevant to ask whether these criticisms were appropriately considered within the IMF. The IMF and the World Bank had earlier agreed, in the so-called Concordat on Fund-Bank Collaboration, on a general procedure to resolve differences of view between the two institutions on economic issues. The evaluation team has not been able to uncover any evidence of dissenting opinions from the World Bank surfacing formally through the procedures established under the Concordat. It is possible that this may be because differences of view on strategy did not follow a simple IMF–World Bank divide.8 It is difficult for the evaluation team to draw any general conclusion except to say that the established collaborative procedures clearly broke down at one of their major tests, with significant adverse consequences.


Several staff members referred to previous occasions (not involving any of the three country cases under study here) where, in their view, staff had made candid assessments but had not been supported by the Executive Board when the country concerned objected. While the IEO makes no judgment on the validity of such assertions, the perception that there is insufficient backing for candor clearly does matter. These issues have also surfaced in previous evaluations of surveillance, including the Whittome Report and the Crow Report.

“Review of the Monetary and Exchange Affairs Department,” November 2002. This review also flagged some more general concerns about the role of MAE in supporting area departments in tackling financial crisis situations and resolving problems in distressed banking systems. Issues raised, which go beyond the three country cases evaluated here, included: (1) MAE tended to move too slowly in reaching a firm position on policies that were needed to address urgent problems; and (2) there were problems with the consistency of advice between different crisis countries. See also “Report of the Task Force on the Review of the Monetary and Exchange Affairs Department,” December 2002.

The Central Asia Department and the South Asia and Pacific Departments were merged to form what is now APD, effective January 1, 1997.

The country annexes provide some examples where such interaction did take place and had some adverse effects.

“The Role of the Fund in Governance Issues,” EBS/97/125, July 1997.

For example, the Indonesian case study notes complaints from ADB staff that it was not sufficiently informed and consulted about the evolution of the strategy in areas where it was involved. Some IMF staff suggested that this reflected confidentiality concerns as well as the fast-moving nature of the negotiations, which created time pressures that led to incomplete communication among the IFIs.

See “Strengthening IMF–World Bank Collaboration on Country Programs and Conditionality,” SM/01/219, August 2001; “Strengthening IMF–World Bank Collaboration on Country Programs and Conditionality—Progress Report,” SM/02/271, August 2002; and “Staff Guideline Note on Operationalizing Fund-Bank Collaboration in Country Programs and Conditionality,” April 2002.

In this context, the World Bank’s Operations Evaluation Department provides its own analysis of the Bank’s crisis response in Indonesia, showing that there were differences between the assessment of the Office of the Chief Economist and that of the Bank’s regional staff (World Bank, 1999b).

    Other Resources Citing This Publication