Chapter 5: Management of IMF Research
- International Monetary Fund. Independent Evaluation Office
- Published Date:
- March 2013
65. This chapter examines the organization of research activities at the IMF, looking in particular at resource allocation; prioritization and coordination; collaboration across IMF units and with authorities; the review process for research documents; incentive structure; and dissemination.21
A. Resources for Research
66. Research activities accounted for about 10 percent of the IMF’s gross administrative budget and about 8 percent of staff time over the evaluation period.22 Area departments together accounted for nearly a third of the time devoted by the IMF to research, RES for about a fifth, MCM and FAD each accounted for just over 10 percent, and INS accounted for about 5 percent of the overall research time.
67. In RES, where about 55 percent of staff time was devoted to research, nearly 80 percent of staff reported that they found enough time for the research needed for their day-to-day work, and half had enough time to conduct “self-initiated” research. In other departments, a much smaller share of staff time was devoted to research—for example, about 10 percent in INS, MCM, and FAD—and only 40 percent of staff reported that they were given enough time for the research needed for their work (and only 15 percent had enough time for “self-initiated” research). This is consistent with staff’s view that in these departments research is considered a residual activity.
B. Prioritization and Coordination
68. There was limited coordination or prioritization across departments (see Box 4), despite the wide-spread view among staff that priorities for research should be set across departments, if not IMF-wide. Attempts have been made in the IMF to introduce a coordinating mechanism, for instance through the Committee on Research Priorities as recommended in the 1999 Mishkin Report, but none of these has endured. On the other hand, a majority of staff responding to the survey (and nearly 90 percent of the respondents in RES) reported that research priorities were discussed within their division “somewhat frequently” or “very frequently.”
C. Collaboration on Research Projects
69. Collaboration on research projects across units within the IMF was infrequent and resulted mostly from informal personal contacts and individual preferences. Occasionally RES or SPR would coordinate interdepartmental research, or an area department sought to have regional units coordinate a study on a cross-country issue. But almost two-thirds of staff reported that there was little incentive to collaborate across the institution. In particular, there was little collaboration between RES and MCM—which may explain, at least in part, why macro-financial linkages did not receive more attention.
Box 4.Views of Current and Former Directors of Research
This box reports the views of four current and former directors of the IMF Research Department.
Main goal of IMF research. Interviewees held the view that the main goal of IMF research was to aid the IMF in surveillance and give credibility to IMF advice to member countries. They thought research needed to be applied and closely linked to the operational work of the Fund. A strong research program was needed to attract and maintain quality researchers. One former director said that IMF research was an important public good.
The setting of research priorities. There was limited coordination on setting research priorities across the institution. Within the Research Department priorities were set in a variety of ways including open-ended brainstorming, top-down decisions, interests of individual researchers, and reaction to current economic developments. In other departments, research priorities were derived from operational needs or determined by Management.
The role of the Research Director as Economic Counsellor to the Managing Director and leader of IMF research. Views varied on how easy it was to perform the dual role of Economic Counsellor and Director of the Research Department. At least two directors said that they faced tension between these roles. One stated that his focus had been mainly on providing advice to the IMF and member countries, and not on research per se. On the other hand, there was a unanimous view that it was not feasible for the Economic Counsellor to coordinate, let alone manage, research across the Fund, because of other demands on the Counsellor’s time. Also, it was not clear how the Counsellor could perform this function in the current organizational structure.
The technical quality and the review of working papers. WPs were commended as a good vehicle to transmit the staff’s research to the public, even though there was considerable variability in their technical quality. Some noted that the lower quality papers tended to come from area departments, where there was less time to conduct research. One director stated that it is important to recognize the long gestation period needed for research to be produced and then again the lag between its production and its impact. Another stated that he did not read WPs and had no interest in vetting those papers. Most acknowledged that there was little screening of WPs and felt that a reexamination of the review process was timely, but expressed concern that attempts to set quality standards might lead to censorship. One director thought that establishing a standard quality assurance process, such as an early seminar, would allow vetting without censorship.
Quality and review of selected issues papers. Views on SIPs varied widely. One director said that he read SIPs on countries he was visiting and found them informative. This director also noted that SIPs have a very different objective than WPs and hence their review process should differ from that for WPs. Another director said that the quality of SIPs was very mixed, partly because they were prepared under tremendous time pressure. Other directors were less familiar with SIPs, with one admitting to never having read one. One called for re-energizing SIPs, by having them address big issues so that research was used to back the recommendations the Fund was providing.
Diversity in research. The general observation was that IMF research reflected a range of views and approaches. One director, however, noted that this was limited by the lack of diversity in staff’s educational backgrounds. Directors also noted that “sometimes Management has its views and this influences research and policy advice.”
70. Collaboration with researchers from outside the IMF was very common among RES economists (64 percent reported collaborating at least “somewhat frequently” with outsiders). The Annual IMF Research Conference is an example of collaboration led by RES with researchers from academia and from other international organizations. Such collaboration was very infrequent in other departments (less than a quarter of economists had ever worked on a research project with non-IMF researchers).23
D. Review of Research Documents
71. There was no systematic IMF-wide process to review research products. The review process varied across departments and research products. The WEO and GFSR were subjected to a formal and structured interdepartmental review process. REOs were reviewed through a structured process in each area department, but these processes varied across departments. There was no systematic and uniform process for reviewing WPs and SIPs. In most departments, division chiefs reviewed WPs and mission chiefs were responsible for reviewing SIPs. Departmental seminars were used in some, but not all, departments as part of the review process and to gather comments.
E. Incentives and Performance Review
72. Research was a criterion for assessing staff performance in most departments, but there was no uniform approach and the weight that it was given often depended on the individual supervisor. Survey results indicate that staff thought that too little weight was given to research and in particular to its quality. In general, performance reviews focused on the number of publications rather than their quality, except in RES.
73. Dissemination is critical to ensure impact from research. The dissemination of the WEO, the GFSR, and a few other key documents was very effective and its messages seemed to have reached the intended audiences. On the other hand, many staff and authorities noted that it was hard to know what had been published recently and that they had difficulties distilling the policy implications of new research. This was in part due to the large number of publications and the lack of a simple search engine. In particular, they noted that there was no searchable database of SIPs.
74. During the evaluation period, the IMF moved towards Internet-based dissemination, reducing the production and distribution of hard copies. At the same time, it started a free email service to inform registered users about the issuance of WPs and some other products. The shift to electronic dissemination facilitated access in most member countries and by staff, but authorities in 40 percent of ECF-eligible countries indicated that it had diminished their usage of IMF publications, partly because of connectivity problems.
For more details see Background Document III: “Management of IMF Research” (www.ieo-imf.org).
These estimates include the full range of research activities considered research by this evaluation, a more comprehensive definition than the one used in IMF budget documents. The evaluation examined the resources dedicated to research by reviewing budgets and reports on the use of staff time. Changes in the IMF’s budget and time reporting systems during the evaluation period complicated the analysis of time spent on research. An additional complication was that time reporting practices varied across individuals and departments, and that departments did not explicitly allocate resources for many research activities that were considered intermediate outputs. Calculations of time spent on research include all time reported as spent on the WEO, GFSR, and REOs, without differentiating between research and other related activities.
To improve transparency and facilitate collaboration, both inside and outside the institution, the IMF could consider establishing an online repository for nonproprietary data, thus allowing replication and validation of results by member country officials and other stakeholders.