Chapter 3 IMF Relationship Management in Sub-Saharan Africa
- International Monetary Fund. Independent Evaluation Office
- Published Date:
- April 2007
This chapter sets out relevant findings about Fund relationships with the authorities, donors, multi-lateral partners, and local civil society groups. It draws on face-to-face interviews and the evaluation survey.1 A key contextual issue is the changing operating environment for aid to SSA, with donors increasingly decentralizing resources and decision making to country offices with implications for how the IMF is perceived, given its more limited field presence (see Box 3.1). The evidence presented in the chapter suggests major differences of views between how IMF staff see themselves and how partners and stakeholders see them, especially in the aid arena. It thus raises questions about how the IMF acquires and processes feedback about its own performance—in view of the intrinsic value of such feedback for self-assessment, learning, and accountability and the information that such feedback may carry about changing conditions on the ground, as an input into strategy formulation and action planning.2
The evaluation team met with and surveyed representatives of ministries of finance and central banks, and also sectoral colleagues in ministries of health, education, and infrastructure and related agencies. Three emerging issues warrant highlighting: (1) the importance attached to the relationship by the authorities; (2) complaints by the ministries of finance about the Fund’s “pro-poor” orientation and the absence of countervailing complaints by the health and education ministries; and (3) the expressed interest by some interviewees in receiving more substantive content from the IMF, and in turn raising questions about the analysis underpinning the operational dialogue on the PRGF (see Box 3.2).
Feedback provided to the evaluation team in face-to-face meetings with representatives of finance ministries and central banks points to increasing openness, flexibility, and tolerance for the accommodation of aid flows by Fund missions and programs. But there also were complaints. Some interviewees criticized Fund missions for listening too little, demanding too much, and imposing their views despite the institution’s rhetoric on “ownership.” Some recalled earlier days of heated debates and difficult discussions during their countries’ stabilization periods. Others complained about mission members’ weak language skills, where relevant, and about staff turnover. One said that changes in mission chiefs were especially disruptive, sometimes triggering wholesale revisions of the program. But several interviewees said that turnover below the level of the mission chief was also a problem; it undermined rather than built capacity, by taking scarce official time to “retrain” new IMF staff all too often.3 On the positive side, the authorities volunteered praise for the work of the African Regional Technical Assistance Center (AFRITAC), whose one-on-one coaching style they very much appreciated, and for IMF Institute courses on financial programming.
Box 3.1.Location of Work
The management of IMF relationships—whether with the authorities, donors, multilateral partners, or civil society—occurs mostly in the field, under the supervision of a headquarters-based mission chief and in his/her absence, a resident representative with highly constrained resources. Current arrangements are increasingly out of step with the IMF’s bilateral and multilateral partners (including the World Bank), which have decentralized significant numbers of staff—and decision-making power—to country offices. Three observations follow, based on the evaluation team’s interviews:
The authorities interviewed by the evaluation team generally did not have problems with current arrangements. They receive priority attention—and some worry that a larger Fund presence might be misconstrued. However, there is interest in greater substantive capacity in resident missions, suggesting a skills-mix issue in some cases.
Vis-à-vis donors, the imbalance is most pressing in those countries for which general budget support has become an important donor instrument. Especially there—although in some other countries as well—donor interest in macroeconomic issues has risen, in turn increasing “demand” for IMF staff time on the ground, without an appreciable increase in “supply,” creating stresses and strains for donor and IMF staff alike, and for relationships between them. These strains color partner perceptions about the IMF’s role and effectiveness.
Vis-à-vis civil society groups, the issue is missed opportunities for exchanging information and for correcting possible miscommunications on both sides.
Box 3.2.Informing the PRGF Dialogue with the Authorities
During the evaluation team’s interviews, some country authorities called for greater focus on substantive content in the dialogue with the Fund. They were especially interested in lessons learned in other countries (within and outside SSA); analysis of and explanations for proposed wage-bill, tax-rate, and other program targets; and connections between macroeconomic policies and aid, the real economy, growth, and poverty reduction.
This raises the question of what inputs Fund staff use in PRGF preparation and design—and whether those inputs capture the full range of analysis and research available. This question was put to IMF staff in the evaluation survey, with the staff responses summarized in the accompanying figure. As shown, large majorities of operational staff respondents to the evaluation survey said they used the analysis carried out by the IMF’s Fiscal Affairs (FAD) and African Departments (AFR) and the World Bank. Majorities said they also used the analysis of the IMF Policy Development and Review Department (PDR), the authorities, and the IMF Monetary and Financial Systems Departments (MFD).1 Minorities said they used the analysis of other sources, including from donors, the IMF Institute (INS), AfDB, academics, and civil society. No respondents said they used the analysis of the IMF Research Department (RES) or UNDP.
These results in turn raise further questions for follow-up. First, how relevant is the Research Department’s analytic work to the macroeconomic challenges that SSA countries face? Second, how open are Fund staff to analysis and ideas that go beyond immediate operational concerns, whether generated within or out-side the Fund? Third, how does the Fund ensure that its advice is adequately informed by up-to-date research and analysis?
Proportions of IMF staff respondents who agreed/strongly agreed that they used the analytical work and experience of:
The evaluation team also met with representatives of ministries of education, health, and infrastructure in the six countries it visited. The most immediate and striking response to questions about possible influence of the IMF on their sectoral resource envelopes and access to aid was the emphasis placed on country ownership. Several sectoral interviewees even took exception to the questions, stressing that the budget was their country’s and that they and their colleagues made all the decisions. More generally, there was little blaming of the IMF for any resource shortfalls that their sectors may have encountered; interviewees said any blame belonged with their own government. Some interviewees applauded the IMF’s positive influence on the development of more realistic plans. Of course, the education and health ministries were major beneficiaries of funding from HIPC savings, which may have favorably inclined them toward the IMF. Representatives from infrastructure ministries generally made two points: (1) the need to broaden the criteria for priority expenditures to include basic infrastructure projects, a plea that was sometimes specifically supported by their colleagues in the health ministries, based on a recognition that investments in water and roads are necessary to meet the health MDGs; and (2) their desire to loosen Fund-imposed constraints on borrowing abroad to finance high-return investments in infrastructure.4 In Ghana, for example, the constraint on nonconcessional borrowing is a major issue for the authorities, who want to borrow commercially to expand infrastructure investment. Similar issues—albeit on a smaller scale and with a more distant time horizon—have arisen in Rwanda and Zambia. In all three cases, PRGF limits on commercial borrowing for infrastructure was a recurring complaint of the authorities during the evaluation team’s face-to-face country interviews.
The authorities’ responses to the evaluation survey were more positive on most questions than other surveyed groups, except for Fund staff. This is a striking and significant result, although to some extent it may reflect selection bias among survey respondents and/or reluctance to criticize Fund staff for fear of adverse consequences, despite reassurances of confidentiality. As pictured in Figure 3.1, the authorities and Fund staff generally see eye to eye on the Fund’s performance on bread-and-butter activities such as the timing of missions, the openness of the dialogue, and the use of the authorities’ analysis and experience—statistically, their responses are not significantly different from each other. However, there were two exceptions in highly relevant areas—first, on the accommodation of aid, where the authorities were significantly less positive than Fund staff on all questions, especially on infrastructure;5 second, and even more important, the difference in views on the use of additional aid scenarios in PRGF design, for which only 47 percent of the authorities agreed that these were used, compared with 88 percent of IMF staff respondents. Otherwise, the authorities and Fund staff tended to respond in broadly similar ways—and quite differently from the other groups—on other substantive questions, as noted elsewhere in this report. Almost 90 percent of the respondents from the authorities’ group were from ministries of finance and central banks, representing 25 of the 29 SSA countries under study. The remaining 10 percent were from sectoral ministries, whose survey response rates tended to be much lower.6
Figure 3.1.Survey Views on IMF Staff and Authority Interface: “Connect”
Proportions of respondents who agreed/strongly agreed that:
Three findings about the relationship between the IMF and SSA donors emerge from the evidence considered by the evaluation team. The first is the continuing high marks the donors give to the Fund’s macroeconomic assessment.7 This is true for donors with traditional project-based aid programs and donors with larger portfolios of general and sectoral budget support operations. The second finding, elaborated below, is the low marks that donors give to Fund staff proactivity in engaging with donors both one-on-one and in formal and informal meetings—largely because they see it happening less than they would like. Third—and closely related—is the stress surrounding donor-Fund relationships on the ground in the era of donor budget support—aggravated by the location-of-work issues set out in Box 3.1.
Many of the evaluation team’s face-to-face discussions with donor representatives focused on the growing importance of general budget support by donors—and reliance on the Fund’s macroeconomic analysis—and its implications for the donor-IMF relationship. Two pressure points were identified with respect to demands on resident representatives’ and mission chiefs’ time. First, the increase in budget support and budget support donors in a number of countries has raised donor interest in an ongoing dialogue with the IMF on macroeconomic issues in the context of working groups and task forces on medium-term expenditure frameworks, inter alia. Second, there are critical moments in the budget and/or donor calendar when information about the IMF macroeconomic assessment is essential. These two pressure points have occasionally erupted into major irritants for both sides; donors have become annoyed with Fund staff’s inability or unwillingness to engage with them and to harmonize with their schedules and Fund staff have become annoyed about increased donor demands on their time and schedule. Several budget support donors complained about the IMF’s inability to commit to decisions on the same time frame as them, in turn complicating the aid and budget planning cycle.
Not all donors sought greater engagement by or with the IMF. In meetings at donor headquarters, several interviewees stressed that—other than the macroeconomic assessment and sign-off—the Fund was neither expected to play a role nor taken into account in the determination of the overall aid envelope in individual country cases. That observation is reflected in donor answers to the survey question about the desirability of the Fund’s increasing its attention to additional aid scenarios going forward. All respondent groups answered positively in the 85–100 percent range, except for donors, who were in the 60 percent range.8
Surveyed donors painted a mixed picture of the IMF’s aid-related work in SSA. Their responses were less positive across the board than the authorities’, but more positive than civil society’s. In answering questions on the Fund’s proactivity on aid—where donors are clearly principals—they were less positive than in other areas (such as on the design of the PRGF) and very much less positive than Fund staff, as illustrated in Figure 3.2. Donors were negative on the question of IMF mission timing—with only about 10 percent of respondents agreeing that missions came at an appropriate time for their processes and decisions on aid, and 30 percent disagreeing.9 They were equally negative on the question of whether the IMF has increased the importance attached to additional policy and aid scenarios.
Figure 3.2.Survey Views on IMF Staff and Donor Interface: “Disconnect”
Proportions of respondents who agreed/strongly agreed that:
Donors were somewhat more positive in their survey answers on the quality of their dialogue with the IMF. A large minority (some 40 percent) of respondents characterized the discussions as full and fair exchanges of views on both sides. In addition, majorities of donor respondents acknowledged changes in the Fund over the past five years toward greater focus on poverty reduction and public expenditure management—both areas they would like enhanced attention to in the next five years. Also in other areas—such as growth, private sector development, and infrastructure—where fewer respondents saw increased attention in the past five years, majorities wanted more IMF attention over the next five years.
The evaluation team also canvassed the views of World Bank, AfDB, and UNDP staff—both through face-to-face interviews in the context of the country visits and through the evaluation survey.
World Bank staff
The complexity of the Fund-Bank staff relationship in SSA is reflected in the evaluation survey results. In some contexts, the IMF-Bank relationship is one of partners, for example, in Fund and Bank staff’s work in supporting country efforts to design and implement Poverty Reduction Strategies. While in other interactions, the Bank works more closely with donors than with the IMF. For example, in the context of the Bank’s participation in donor budget support groups through its Poverty Reduction Support Credits, the Bank behaves more like a donor, sharing donor concerns about IMF mission timing and the effect of macroeconomic policies on the absorption and spending of aid. As a result, for some survey questions, Bank staff responses are closer to those of Fund staff, while for others they are closer to donors.
One example in which Bank staff are closer to donors in their views than to Fund staff is with respect to PRGFs’ accommodation of aid earmarked for sectors such as education, health, and infrastructure. Figure 3.3 illustrates the results, which show a large disconnect between Fund staff thinking and that of Bank staff and donors—and the authorities, especially on infrastructure. The disconnect probably reflects different meanings attached to the word “accommodate” by Fund staff and by other survey respondents—with Fund staff meaning in line with Fund policy on the accommodation of aid, which as seen in Figure 2.3 can mean a very low spending rate out of incremental aid for countries with inflation rates above the critical 5–7 percent threshold. Whereas, to donors and Bank staff “accommodate” may mean that additional aid is simply allowed to be spent. Nevertheless, the size of the gap is worrying, especially the gap between Fund and Bank staff views given that it relates to views on the interface between core operational products of the two institutions—the PRGF in the Fund and financial support for key sectors (education, health, and infrastructure) in the Bank.
Figure 3.3.Survey Views on Accommodation of Earmarked Aid
Proportions of respondents who agreed/strongly agreed that PRGFs accommodated the use of aid earmarked for:
On the critical partnership issue, the quality of the collaboration between the two institutions received disappointing responses from surveyed Fund and Bank staff. Figure 3.4 suggests that about half of surveyed staff are positive on the collaboration on strategy issues for growth and poverty reduction. The good news is that IMF and Bank staff have similar views on these questions, and the apparent differences between their responses are not statistically significant. The bad news is that the (shared) view of the collaboration is not more positive. Ideally, effective collaboration on SSA would be a top priority for both institutions, and fully reflected in staff views about how it is actually working on the ground.
Figure 3.4.Survey Views on IMF Staff and World Bank Staff Interface: “Disappoint”
Proportions of respondents who said that collaboration was productive/very productive on:
The survey scores are even lower for Fund and Bank staff responses on the analysis of absorptive capacity for current and additional levels of aid—both overall and in health, as an example of a specific sector. In follow-on questions asking for the reasons for identified problems, most Bank staff responded that IMF staff had not asked for their inputs and most Fund staff responded that country-specific work programs lacked deliverables, time frames, and resources supporting the agreed division of labor. Either way, the bottom line is the same—the need to move beyond general understandings about lead agencies to specific agreements on deliverables with time-bound and fully-costed work programs based on specific country program needs. This conclusion is consistent with the earlier discussion of Fund-Bank collaboration on the PSIA—in the section “Poverty and social impact analysis”—where it was noted that the collaboration does not work well in areas where one institution (typically the Bank) is meant to supply the other institution (typically the IMF) with specific inputs and expertise, as it is perceived to be an unfunded mandate. Clearly, a different business model is at work in areas of more successful Fund-Bank collaboration, such as fiscal governance, where both institutions operate as principals and the challenge is to coordinate better—avoiding duplication and contradiction and achieving synergies.
African Development Bank staff
Face-to-face interviews with AfDB staff in Tunis and SSA capitals painted a picture of increasing openness by Fund missions and resident representatives in SSA countries. AfDB staff reported that they saw more (and more genuine) interaction between the Fund staff and the authorities and donors, including the AfDB. At the same time, they expressed concerns about what they saw as the Fund’s limited engagement with civil society. They said they looked forward to increased collaboration on governance and sectoral policies and strategies as the AfDB increased its investment in economic and sector work. In their survey responses, AfDB staff was positive about IMF proactivity at formal and informal aid meetings. Looking forward, almost all AfDB respondents stressed the importance of greater investment by the Fund in additional policy and aid scenarios and involvement in policy monitoring efforts conducted jointly by donors at the local level.
United Nations Development Programme staff
United Nations Development Programme (UNDP) resident representatives interviewed by the evaluation team also commented favorably on what they saw as recent changes in the IMF approach. They said that IMF staff now consult more broadly with stakeholders and are more willing to adapt the macroeconomic policy stance to accommodate needed social expenditures. Going forward, they highlighted the importance of a more collaborative IMF strategy to help SSA countries achieve the MDGs.10 Despite the relatively favorable survey responses from its staff, UNDP case studies and research criticize IMF activities and impact. Its Ghana case study argued that the Fund’s fixation on fighting inflation crowded out attention to economic growth, employment creation, and poverty reduction.11 Its Zambia case study focused on what it saw as excessive involvement by the Fund and other international financial institutions, including on the use of aid, which it said stifled domestic initiative and constrained social service provision and in turn poverty reduction and growth.12
Evaluation survey responses and interviews during country visits point to limited and ineffective IMF engagement with country-based members of civil society. This translates into missed opportunities for dialogue on key issues, including “on the links between the macroeconomic framework and growth and poverty reduction outcomes in the context of work on PRGF-supported programs,” as the Board agreed staff needed to more actively explain to a broad audience, including civil society.13
The very clear message from civil society survey responses—and from the evaluations team’s face-to-face meetings with civil society groups during the country visits—was that Fund staff are generally unknown and unavailable to civil society in SSA. This contrasts with Fund staff opinion. As Figure 3.5 illustrates, about 80 percent of IMF staff respondents report progress in their engagement with civil society over the past five years, whereas only 20 percent of civil society respondents see such progress. Going forward, majorities of all respondent groups—including the authorities and IMF staff—agreed that greater outreach efforts were important.
Figure 3.5.Survey Views on IMF Staff and Local Civil Society Interface: “Major Disconnect”
Proportions of respondents who agreed/strongly agreed that IMF has been doing:
The evaluation team’s face-to-face interviews with civil society representatives reinforced the finding of limited interaction with IMF staff. They pointed to even more limited agreement on assumptions about how IMF-supported policies affect the use of aid and poverty reduction and the MDGs. In Mozambique, for example, local civil society organizations complained about the design of PRGF program adjusters, which they said blocked Mozambique’s use of aid, and wage bill ceilings, and complained of limited dialogue with Fund staff. As it turned out, those program elements had been recently modified, but the civil society representatives had not learned about the changes, despite efforts on their part to find out. Interviews with resident mission staff suggest that while Fund policies encourage outreach and communications with civil society, they have received little actual support for such work. Yet this is the area where the dialogue is most difficult—where differences of views between civil society and government policies make the dialogue especially sensitive. There are also time and other resource constraints to be considered, as suggested by Figure 3.6, which shows mission chiefs’ and resident representatives’ views on the time they have available to meet with the authorities, donors, and civil society. Moreover, the evaluation interviews revealed that in some instances it is the authorities who prefer that the IMF have a low profile in discussions with civil society.
Figure 3.6.With Whom Do Staff Spend Their Time?
Proportions of staff respondents who agreed/strongly agreed that Fund missions and resident representatives had ample time to meet with:
Key inputs include (1) meetings with SSA ministers of finance, central bank governors, and their staff during the 2006 Spring Meetings in Washington; (2) interviews in Accra, Dar es Salaam, Kigali, Lusaka, Maputo, and Ouagadougou; in Addis Ababa and Tunis; and in donor capitals; and (3) responses to the evaluation survey from the authorities, donors, local civil society representatives, and staff of the African Development Bank (AfDB), IMF, United Nations Development Programme (UNDP), and World Bank. See Annex 5 for survey details.
Empirical analysis carried out by the evaluation team suggests that similar mission turnover rates characterize all program countries.
See Figure 3.3 below.
Given the small number of returns from this group, it is not possible to differentiate their responses from those of the central bank and ministry of finance representatives in a statistically significant way. This said, their responses tended to be broadly in line with those of their finance ministry and central bank colleagues—albeit somewhat more positive on IMF work on the MDGs and less positive on IMF work on aid.
See also the survey in IMF (2005m).
The remaining responses were either “neutral” or “don’t know.”
In their survey responses, UNDP staff were closely aligned with those of civil society, although the low number of responses to most questions means that they are not statistically significant.
See IMF (2002b).