Evaluation of the IMF's Role in Poverty Reduction Strategy Papers and the Poverty Reduction and Growth Facility

Chapter 5 The IMF’s Way of Doing Business

David Goldsbrough, Isabelle Mateos y Lago, Martin Kaufman, Daouda Sembene, Tsidi Tsikata, Steve Mugerwa, Alex Segura-Ubiergo, and Jeff Chelsky
Published Date:
September 2004
  • ShareShare
Show Summary Details

Key Messages

The 1999 policy papers launching the PRSP/PRGF called for far-reaching changes in the nature of the IMF’s involvement in low-income countries. These expectations were probably overambitious without much deeper organizational changes—and more administrative resources—than have occurred so far. Actual change has fallen well short of these expectations, although there are marked differences between “good” and “average” practice.

The heavy procedural requirements of the PRS process are absorbing much of the time of relatively small staff country teams at the expense of analytical and capacity-building activities.

IMF general policy advice, reflected in the PRSP Sourcebook, allowed for an appropriate broadening of the policy space. So has the country-specific internal IMF policy formulation process—at least in countries with a good policy implementation record and macroeconomic stability. There is some evidence of continued progress over time. However, staff has generally not viewed the participatory elements of the PRS process as a vehicle for exploring and resolving controversial policy issues.

IMF staff involvement has generally fallen well short of active participation in informing the broader policy dialogue. A lack of clarity about what was expected of staff generally led to a “hands-off” approach, in the interests of domestic ownership, even when key issues were not being addressed.

Despite greater sensitivity to the social implications of macroeconomic policy advice, there is not yet a systematic effort to identify and fill country-specific knowledge gaps on macro-micro linkages and PSIA and to integrate the resulting evidence into program design. The PRS process has not been used to generate systematic priorities on what the IMF itself should be delivering.

There is no systematic guidance for the role of resident representatives in the PRS process. The scope of their contribution varies widely, depending on individuals, not institutional arrangements.

The IMF’s contributions to the broader partnership framework implied by the PRS approach have been limited by a lack of clarity on what this framework means for the IMF’s way of doing business. This includes the IMF’s “signaling” role, the role of conditionality in a longer-term framework, and operational approaches to setting the medium-term external resource envelope.

The PRS/PRGF initiative called for significant changes in the IMF’s way of doing business. This chapter asks how much has changed in practice. We begin with a discussion of what the original policy papers as well as subsequent internal guidance indicated were the major expected changes in the Fund’s operating approach. We then assess: (i) the extent to which the IMF’s general policy advice to low-income countries allows for greater “policy space” and how much the IMF’s own internal policy formulation has changed in light of the new approach; (ii) how much the IMF’s approach on the ground has changed, according to the case studies; (iii) the role of resident representatives; and (iv) the contribution of IMF surveillance. We conclude by discussing some of the main obstacles to enhancing the effectiveness of the IMF’s role.

What Was Expected of the IMF Under the New Initiative?

The original policy papers envisaged substantial changes in the nature of the IMF’s involvement in PRSP countries for the new initiative to be successful. The expected changes in some critical areas are illustrated in the quotations (below) from these papers.1

  • Openness to homegrown adjustment and reform: expanded policy spaceThe staffs must be open to more flexible, home grown adjustment and reform paths, which reflect country ownership without prejudicing the goal of poverty reduction and faster growth.The Fund will need to be ready to assess new approaches and to recognize and support a healthy process of experimentation and innovation. Fund staff will be open to considering alternative adjustment paths, taking into account their impact on the poor.
  • Active participation in the domestic policy debateBank and Fund staff should be prepared to share their analyses and the key elements of their policy positions in the consultative process, even during the early stages of the policy dialogue. Missions would learn from the policy dialogue, and should be ready to consider modifying their analysis of the problems and policy options accordingly. Staff would need to play an active role in supporting the process, to ensure that domestic dialogue is well-informed.Discussions on the macroeconomic framework will become more open and iterative. Fund staff can expect to be requested by members to participate in broad-based consultations organized by the member in preparing its poverty strategy and the macroeconomic framework underpinning it. Key macroeconomic policies, including targets for growth and inflation, and the thrust of fiscal, monetary, and external policies, as well as structural policies to accelerate growth, would be subjects for public consultation.
  • Contribution to expanding knowledge on the linkages between macroeconomic policies and growth/poverty reduction[There is a] need for Bank and Fund staff to work closely together to improve our knowledge of the general link between the macro framework, growth, and poverty reduction, particularly over the medium-term so as to understand better the general issues raised.
  • Inputs into determining the external resource envelopeEarly in the process, the staffs should seek a common perspective… on… an initial view of the current resource envelope and, within this, the possible scale of expenditures for poverty reduction; [and] a discussion of the current levels and nature of external assistance, and the prospects for increased aid over the medium-term.One key component will be to mobilize external support, and here Fund staff will need to redouble efforts to identify sustained increases in resources for countries where these can be used most effectively.

The initial policy papers put little emphasis on capacity building, but as time went by it became clear that weak country capacity was a major constraint to effective preparation and implementation of PRSs. Accordingly, at the time of the 2002 review, the Board encouraged the staff to step up its efforts at capacity building in the IMF’s areas of expertise.

How these broad—and challenging—expectations were to be translated into specific operational approaches for IMF staff was not spelled out initially, although some guidance was eventually issued later (Box 5.1). There was an understandable concern not to overprescribe a common operating approach that might conflict with the emphasis on country-driven processes and the need for “learning by doing.” However, this reluctance to be overpre-scriptive translated in practice into a lack of concrete indicators on precisely what the IMF should be delivering—either initiative-wide or in individual countries.

Box 5.1.Internal Guidance to IMF Staff on the Implementation of the PRS Approach

This box addresses two related issues: (i) what internal guidance says about IMF staff’s role in the PRS approach. It should be emphasized that a large share of this guidance was developed recently (i.e., in 2003) and therefore was not available to guide staff work for most of the period under review; and (ii) what staff think their role is and whether they think the guidance is clear.

General guidance on role of staff in PRS approach

Early policy papers recognized that substantive changes in the organization of mission work would be required, but this was not translated into specific operational guidance until a relatively late stage.1 In August 2003, the Policy Development and Review Department (PDR) circulated a note to IMF mission chiefs on how to handle key challenges raised by the PRS approach. Key suggestions included: (i) engaging early on in the PRS process, with staff visits planned in line with domestic processes; (ii) encouraging and helping (through capacity-building efforts) the authorities to base their PRSP on realistic growth projections, to integrate it with the budget and medium-term expenditure framework, and to incorporate contingency spending plans; (iii) discussing with the authorities the macroeconomic implications of higher aid flows, to determine whether the negative macroeconomic consequences of higher externally financed poverty-reducing spending outweigh its benefits; and (iv) synchronizing the PRGF and PRS cycles by initiating a new three-year PRGF arrangement shortly after the completion of a new PRSP. The same memo also encouraged staff to use IMF Country Strategy Papers and Article IV consultation reports as vehicles to provide their assessment of the sources of and obstacles to growth, drawing upon the analyses of the World Bank and others.

Guidance on outreach

On the occasion of a review of the IMF’s external communications strategy in early 2003, Executive Directors, while not distinguishing between PRS countries and others, supported a more active role for resident representatives and mission chiefs in outreach. The “Guide for Staff Relations with CSOs” (IMF, 2003i) suggests to “treat public outreach as vital, but (given resource constraints) not [to] compromise other tasks.” The purposes of such relations are defined broadly and include explaining the Fund and its activities; obtaining policy inputs from nongovernmental sources; gauging forces for and against IMF-supported policies; and building national support and initiative toward IMF-backed policies. It is suggested that staff meet with CSOs early enough in the process so that the consultation is meaningful, to ensure ample opportunity for questions and comments, and to debate options. The guidelines also emphasize that relations with CSOs should be handled in ways that do not alienate or put indirect pressure on governments.

Staff views on their own role in PRS approach2

Why has there been only limited change in the IMF’s way of doing business? Respondents on average were of the view that the main factors were staff resource constraints, the demands of the review process, or doubts about the value added of the new approach. Among the factors constraining staff’s outreach efforts toward civil society were limited mission size and length and absence of suitable internal incentives. Absence of clear guidance on what was expected was generally considered a less important factor. Only about one-fifth of respondents to the survey of IMF staff agreed that the PRS/PRGF initiatives had led to significant changes in the way initial policy positions are discussed within the Fund.

Role of IMF staff in PRSP participatory process. One-half of respondents considered their role as participating only to the extent requested by the authorities. Only about two-fifth of respondents saw their role as implying active participation throughout the process, while one in ten defined their role as entirely passive.

Main changes in IMF’s way of doing business. Of the key features of the PRGF, staff thought the ones that had the most impact on their work were the requirements for more selective structural conditionality and for broad participation and greater ownership. All key features were found to have had a significant impact on staff work, PSIA excepted. Two aspects of staff work were highlighted as having undergone the most significant changes: collaboration with the World Bank (in the sense that it was improved both at headquarters and, even more, in the field) and interactions with civil society. Regarding obstacles to PSIA, funding/staffing constraints in the IMF and the World Bank and capacity constraints in member countries were emphasized.

1 Interviews with area department staff suggest that the general policy line was that staff should take a hands-off approach, out of concern not to undermine country ownership, or be seen as doing so.2 Based on survey of a group of IMF staff members with current or prior involvement in PRSP/PRGF countries. See Annex 2 for details.

Have the IMF’s Policy-Formulation Processes and Advice Changed Under the PRSP/PRGF?

We looked at two types of evidence to see what had changed in the IMF’s own processes. First, we reviewed the IMF’s general policy advice on macroeconomic issues in low-income countries to assess whether it gives greater “policy space” for the consideration of alternative policy options. Second, we examined what happened during the internal policy formulation process for all countries in our sample to see how the new approach was applied in practice.

General IMF policy advice: what has happened to the “policy space”?

As noted in Chapter 2, the policy papers establishing the PRS/PRGF approach never explicitly address the potential tensions between ownership and the BWIs’ own judgments on what policies are good for growth and poverty reduction. But a greater emphasis on ownership was not meant to imply the acceptance of any homegrown policy package. The aim was to create greater scope for country-driven alternatives and investigation of alternative strategies by expanding the policy space acceptable to BWIs, especially on issues where the evidence on links from policies to growth and poverty reduction is not well established or is likely to depend on circumstances in a particular country.2 We reviewed one particular template for the provision of general policy advice—the chapter on “Macroeconomic Issues” in the PRSP Sourcebook—to assess whether the policy space was indeed widened.3 Our overall judgment is that the Sourcebook generally achieves its objective of discussing the fiscal, monetary, and exchange rate policies that would be supportive of sustainable growth and poverty reduction in a manner that provides adequate room for different approaches, perhaps arising from the circumstances in each country (see Annex 4 for details).

However, on a few issues the discussion appears to be overprescriptive. For example, while recognizing that the pace of adjustment to external shocks will be influenced by the availability of financing, the Sourcebook essentially endorses an approach for fiscal policy to “treat every favorable shock as temporary and every adverse one as permanent.” This appears too sweeping, especially once one considers aspects of the nature of poverty that emerge from poverty assessments undertaken as part of each PRSP whereby temporary contractions can have adverse permanent effects on growth and poverty (if, for example, children are pulled out of school).

Another area where the PRSP Sourcebook is overprescriptive relates to exchange controls. It argues that relaxing controls that force assets to be held in domestic currency could give the poor access to safer assets, such as foreign currency, that can protect them from devaluations. Asserting a direct poverty-reduction dimension to such liberalization without recognizing the complexity of the policy choice involved, which depends critically on much larger sequencing issues, goes beyond the available evidence.4

Interestingly, the views of various I-NGOs on the general policy advice contained in the PRSP Source-book are much more negative than our own assessment. We discuss what seem to be the main areas of contention in Box 5.2.

Changes in internal IMF policy formulation processes between the PRGF and ESAF

We examined in detail the IMF’s internal “briefing paper process” to see whether its approach to policy formulation has adapted to the new approach. Under IMF internal procedures, the area department prepares a briefing paper prior to every staff mission to a country. Each brief presents an assessment of the country’s economic challenges, and identifies the key issues that the mission plans to take up with the authorities, along with main features of the anticipated program or policy advice. The brief is commented on by several IMF reviewing departments and is cleared by management. It is not seen by the IMF’s Board or the authorities. This internal clearance process is much more centralized than that of other IFIs and provides a rich source of material for assessing whether internal processes are moving in the directions called for by the PRS/PRGF approach.

We compared briefing papers from the ESAF and PRGF periods, along with reviewing departments’ comments on these briefs, for missions to the sample of 23 countries according to a set of uniform criteria on what we expected to see in terms of changes in approach.5 There were wide variations between “average” and “best” practice, but the following messages emerge:

  • Increased attention to protection of key social objectives. A majority of PRGF briefs considered various ways of protecting key objectives in the event of unanticipated shocks, including those of a global slowdown and reductions in aid disbursements. In particular, they discussed ways of protecting priority expenditures, even where this required some expansion of the fiscal deficit. This contrasted with earlier briefs that often indicated the potential sources of shocks but without any strategy to protect priority expenditures.
  • More limited opening of policy space in macroeconomic areas, except in countries with a strong policy implementation record and macroeconomic stability. Only about a quarter of the PRGF briefs indicated an increase in policy space in the sense of acknowledging that some issues should be addressed through homegrown options without the IMF staff pushing for a particular approach. On macroeconomic stabilization measures and structural issues, some space for the discussion of alternative options was already evident under the ESAF; on average, there appears to have been only a modest further expansion of such policy space under the PRGF, concentrated in countries with good macroeconomic performance. The most clear-cut examples of increased willingness to defer to the view of the authorities were in the areas of public finance and, to a lesser extent, financial sector reform.6 An important positive sign of change was that review departments were generally less inclined to suggest increases in conditionality as a means of resolving outstanding policy issues than had been the case under the ESAF, except for countries with a history of poor economic policy performance.
  • Broader participation was generally not viewed as a significant avenue for exploring macroeconomic policy alternatives. On the whole, PRGF briefs did not accord participatory processes under the PRS much significance as a vehicle for resolving key macroeconomic policy issues, typically not looking to such processes as a means of easing political economy constraints or generating alternative policy options. However, in the few cases where the importance of a participatory process was recognized, the briefs envisioned greater consultations by IMF staff with domestic stakeholder groups, including the private sector, and aimed to use outcomes from such discussions in the design of the program.
  • Broader call for PSIA, but still not a regular part of ex ante program design, and vagueness about responsibility for delivering the analysis. PRGF briefs in most cases pay greater attention to the idea of PSIA, in many cases highlighting where such analysis should be done. Some briefs suggested the setting up of national committees for coordinating PSIA work, to enable a more systematic scrutiny of the poverty impact of policies. The approach was premised on the assessments being under-taken outside the IMF (by governments themselves, the World Bank, or other donors), but often with considerable vagueness as to when such inputs might be delivered.
  • Heightened flexibility to accommodate external financing. The PRGF briefs indicated more explicitly than earlier briefs a willingness to accommodate higher aid inflows. The scope for a higher fiscal deficit was recognized, although few briefs linked this systematically to program design issues or discussed in detail the rationale for a particular proposed fiscal path. Generally—and partly as a legacy of the HIPC-oriented process—the briefs held to the view that the bulk of the new external financing had to be targeted to priority sectors, notably health and education, while the impact of alternative expenditure patterns was generally not explicitly considered.

Box 5.2.Views of NGOs: What Is the Nature of Their Policy Differences with the IMF?

As part of the effort to encourage broad-based inputs into the evaluation, we invited the views of a number of groups on the thrust of the macroeconomic policy advice contained in the PRSP Sourcebook. It is interesting to discuss the reaction of some I-NGOs, not to suggest that we necessarily share their judgment but because their response helps to understand the nature of their policy disagreements with the IMF:1

Concept of macroeconomic stability and degree of trade-offs with other objectives. While recognizing that large macroeconomic disequilibria can be an obstacle to poverty reduction, some critics argue that macroeconomic stability may compete, and even be in conflict, with other objectives, such as government spending to meet the MDGs, and that poverty reduction, not the macroeconomic objective, should be given priority. For example, they argue that a more flexible approach to the threshold inflation objective may be warranted, depending on the trade-offs in particular countries. We discussed this latter issue in Chapter 4 and came to the conclusion that there is no evidence that programs systematically target excessive disinflation.

Assumed supply-side responses to macroeconomic adjustment. Critics argue that the IMF’s policy advice implicitly assumes that the private sector will respond rationally to macroeconomic adjustments, which understates the potential impact of extensive market imperfections in developing countries. The speed of private sector response is one of the key “macro-micro” linkages, and the discussion in Chapter 4 suggests that program design is indeed sometimes based on assumptions not backed by strong evidence.

The efficacy of social safety nets. Critics argue that the effectiveness of such safety nets in protecting the poor from adjustment costs is questionable. In particular, they argue the introduction of such safety nets at the time of a crisis or shock rarely works well and a more effective strategy would be to develop longer-term programs in advance, along with ex ante PSIA of adjustment policies. This is a point now recognized by the IMF. Indeed, the recent IEO evaluation of the role of fiscal adjustment in IMF-supported programs recommended that, as part of its surveillance activities, the IMF invite countries that so wish to discuss how such programs might be protected in the event that extensive adjustment is required.

Greater focus is needed on what the IMF can do to improve the external environment for sound macroeconomic policies, including the problems caused by the unpredictability and volatility of aid flows and the trade environment faced by low-income countries. We take up the latter issue in the discussion below on the IMF’s surveillance role.

1 The discussion here draws in particular on comments submitted to the IEO by the EURODAD secretariat, but the submission by Oxfam (2004) makes a number of similar points.

The IMF’s Contribution on the Ground: Evidence from Case Studies

The case studies provide our main source of evidence for what has actually changed on the ground with regard to the IMF’s role. There appears to be a wide variation in country experience, but in general the involvement of IMF staff has fallen far short of the active participation in the consultative process and resulting policy dialogue suggested by the policy papers establishing the PRS/PRGF approach. As noted earlier, these papers anticipated a more active role in the form of early involvement; active staff support for well-informed domestic dialogue; open and interactive process; and staff participation in broad-based consultations. In none of the case studies did the IMF meet these expectations fully, although it made progress in a number of areas:

  • The staff in all cases engaged in some early interaction, but this was typically limited to its traditional interlocutors (the ministry of finance and central bank). The Fund also generally adapted its mission timing to the government’s own policy cycle, including that for the budget.
  • In cases where a PRGF arrangement was already in place before the PRSP,7 the macroeconomic framework was typically taken from the former with limited efforts to “open up the policy debate.” Beyond this transitional phase, there are some examples—most notably Tanzania, but also Mozambique—where the staff did engage in a more interactive dialogue on some policy aspects. It is hard to generalize from these few cases, but they do suggest that such changes in approach can add significant value in terms of better policy outcomes. (We will illustrate this in Chapter 6 when we trace through the impact of the approach on a number of specific policy issues and outcomes.)
  • There were few cases where the IMF staff had identified issues that it viewed as critical and successfully pushed for them to be on the menu of the broader policy debate around the PRS—which would seem to be a central element of ensuring a well-informed dialogue. SOE reform in Vietnam was one exception, but this was an issue where the government already had strong ownership.
  • The donor community was generally critical of the IMF’s preference to remain in the background, instead of helping to explain to stake-holders the implications of the policy measures it was recommending to governments. In a number of cases, however, there have been increased efforts in recent years to explain the rationale of Fund advice (e.g., Nicaragua, Tanzania, and Vietnam), typically led by the resident representative.
  • IMF staff (especially resident representatives) now meet a broader range civil society on a more frequent basis, although the substantiveness of these discussions varies considerably. It is most substantive in cases where the government has established an institutional framework for a broader dialogue on policy issues.

The case studies support the conclusion that there is somewhat greater openness to considering alternative policy courses and trade-offs, but only after immediate macrostabilization issues had been addressed (e.g., Nicaragua and Mozambique). In contrast, when programs went substantially off track (e.g., in Guinea), the staff’s focus was on traditional stabilization approaches that drew little from the broader PRS debate—with the important exception that significant attention was given to protecting priority expenditures.

There is also some evidence of progress over time in consideration of alternative options, although it is hard to make broad generalizations. Continued Fund involvement in the update of the macroeconomic policy framework resulted in reconsideration of elements of the strategy in some cases (e.g., Tanzania—see Box 4.2—and Mozambique). In addition, several cases show progress over time in opening up the policy space on particular issues (e.g., the approach to tax reform in Nicaragua; fiscal flexibility in Tanzania; and tariff policy in Mozambique).

Interviews with staff and evidence from the case studies suggest that, despite some efforts to broaden the investigation of linkages between macroeconomic policies and growth/poverty outcomes, there is not yet a systematic effort to integrate such evidence into macroeconomic program design. In this connection, a recent review of the IMF’s African Department reached a similar conclusion that there has been not enough emphasis on filling the gap between broad macroeconomic prescriptions from general cross-country evidence and specific country work on PRSP-relevant issues.8

IMF Contributions to the Broader Partnership Framework

The general objective of the partnership principle of the PRS approach is to help countries overcome long-standing problems of donor coordination. We address here three implications for the IMF’s role: (i) conditionality and performance monitoring; (ii) determination of the external resource envelope; and (iii) capacity building.9 In all cases, our evaluation suggests that the implications for the IMF’s way of doing business of the broader partnership framework are still insufficiently developed.

On conditionality and performance monitoring, there are two major shortcomings:

  • The potential of the JSA as a common reference among development partners has largely failed to be realized, for reasons discussed in Chapter 3: first, the clarity of the IMF’s assessment in the JSA varies considerably, weakening its role as a signal to other donors. Second, efforts to incorporate inputs from other donors also vary widely but are inadequate in most cases.
  • The “signaling” role of IMF-supported programs is often ambiguous. For example, the traditional short-term IMF conditionality may not fit well with a framework where the international community is “in for the long haul.”10 Indeed, no one—including the IMF—would want a situation where short-term interruptions in programs lead to excessive “on-off” signals for other financing flows even when all stakeholders recognize that the issues involved in the interruption are of legitimate concern. We have not been able to obtain systematic evidence on the effects of program interruptions on such financing flows, but the limited evidence available from the case studies suggests a mixed picture. In some cases, donors’ decisions about the extent and modalities of their support were only loosely tied to performance under IMF-supported programs (e.g., in Guinea, donors proved unwilling to commit to much support even when the PRGF-supported program was on track, whereas in Vietnam they provided substantial support regardless of the program status).

With regard to the determination of the external resource envelope, two factors appear to have constrained the IMF’s role in helping to mobilize financing. First, the approach to balancing “ambition” and “realism” in PRS macroeconomic frameworks is unresolved (see Chapter 4). In practice, there have been only limited efforts to integrate systematically macroeconomic frameworks and sustainability analysis (areas where the IMF has a comparative advantage) with realistic costings of other aspects of the growth and poverty reduction strategy and with assessments of absorptive capacity (areas where the IMF does not have a comparative advantage and inputs should be taken from other partners). Second, procedures for giving feedback on each country’s PRS, through the JSA, provide no mechanism for “treating different countries differently” (i.e., for signaling that some sets of policies deserve more support from the donor community). The standard language used to conclude all JSAs, as discussed in Chapter 3, mitigates against such signaling, even though the original expectation was that the JSA would play such a role.

In other words, there is little in the PRS approach as presently designed and implemented that provides for effective signaling with regard to aid selectivity, let alone collective decision making on a reasonable level of aid flows to assume in formulating the medium-term macroeconomic framework for PRSPs. Consequently, the calls in the original policy documents for the IMF to play a larger catalytic role in mobilizing such financing have little practical backing in terms of operational mechanisms.11

Recent internal reviews have underlined that PRSPs should provide the basis for setting technical assistance (TA) needs and country-owned priorities for capacity building.12 However, this recognition is relatively new. The IMF policy statement on TA, adopted in April 2001, does not mention the PRSP/PRGF approach in its discussion of filters and priorities for enhancing the effectiveness of IMF TA. In practice, the record suggested by the case studies is mixed:

  • To date, the bulk of IMF TA activities—and thus the nature and scope of capacity building—has been more influenced by the demands of IMF-supported programs than long-term capacity-building needs of the countries. The planning of TA supply remains driven by functional rather than area departments, and inputs from the latter are typically not anchored in the PRS (although there are signs that this is changing as far as the African Department is concerned). The planning horizon is yearly at best, which does not mesh well with the longer-term perspective of the PRS approach.
  • Few PRSPs provide systematic guidance on capacity-building needs in macroeconomic and related areas. Most JSAs, by contrast, do offer some discussion of capacity constraints. But there is no indication that this has been used as a basis to orient the provision of IMF TA, let alone coordinate TA supply with other providers.
  • Nevertheless, the case studies do suggest that the broad areas in which the IMF is delivering TA are reasonably aligned with PRS priorities, taking account of the IMF’s comparative advantage. They also suggest that efforts to coordinate IMF TA delivery with other donors have improved, but in no case is such TA fully integrated into a common framework focused on longer-term objectives.

Role of IMF Resident Representatives

The presence of a resident representative in countries participating in the PRS/PRGF initiative can significantly enhance the IMF’s ability to understand the political and social context in which policy is made and facilitate earlier and less formal dialogue on policy issues.13 We review here what guidance is given to resident representatives, and summarize evidence on their actual role from the case studies and the staff survey.

There is no Fund-wide guidance to resident representatives on the role they are expected to play in the PRS process. Area departments prepare terms of reference (TORs) for each representative. These TORs are agreed with the host country and, in principle at least, could be a vehicle for spelling out some of the “rules of the game” for IMF involvement in the broader policy dialogue. We reviewed all TORs for resident representatives in PRS countries over the last four years and, with a few noteworthy exceptions, they provide remarkably little guidance on what was expected vis-à-vis the PRS process. The issue was not mentioned in most TORs for resident representatives outside Africa. For postings in Africa, instructions varied considerably. While some made no reference, others referred to “liaising with civil society in the context of the PRSP process.” In only a minority of countries (all in Francophone Africa) did the TORs describe explicitly a proactive role for the resident representative.

The case studies throw light on two aspects of the representatives’ role: (i) their involvement in the broader policy dialogue and participatory process on macroeconomic and related issues; and (ii) their inputs into the formulation of the IMF’s own policy formulation and program design.

The case studies show a wide variation in the role played by resident representatives in the broader policy dialogue. In Vietnam and Nicaragua, representatives were active contributors to discussions among a broad range of stakeholders, and were praised by some local stakeholders for their ability to discuss complex macroeconomic issues in a manner that was accessible to noneconomists. In Mozambique, the donor community was equally complimentary, but the civil society organizations (CSOs) considered the IMF to have been “invisible.”14 In some other countries (e.g., Guinea), resident representatives interpreted the emphasis on country ownership to imply a more “hands-off” approach, and members of the donor community and civil society expressed a desire to have more substantive interactions.

The institutional context within which each resident representative operates has a significant influence on the modalities for involvement in domestic participatory processes. In some countries, like Mozambique, Tanzania, and Vietnam, well-established forums exist within which the representative could engage with CSOs and donors. The scope in Tajikistan was more limited given the nascent state of civil society and the small number of donors on the ground. In Guinea, dialogue with CSOs was limited partly because the authorities had reservations about the extent of formal consultation between the representative and civil society.

That said, the evidence suggests that it is possible for the resident representative to engage meaning-fully with key stakeholders. This requires an ability and interest on the part of the representative to communicate openly and substantively with groups that have varying degrees of technical sophistication. It also requires the right structure of incentives. Our discussions suggest that these kinds of communication skills were not sufficiently recognized or rewarded by the IMF.

It is difficult to assess comprehensively the impact of representatives on the IMF’s internal policy formulation because it is mostly channeled through informal exchanges between the representative and IMF headquarters staff (especially the mission chief). Experience varied substantially, but the following points are worth noting: (i) In all cases, the representative reports through the headquarters-based mission chief. (ii) Internal briefing papers are always prepared in Washington; the representative is usually (but not always) asked to comment, but these inputs stay within the area department. (iii) Experienced and well-informed resident representatives can have a significant input into the Fund’s policy advice and program design, but it depends on the individuals involved. Nothing in current institutional arrangements systematizes such an input.

Role of Surveillance

Surveillance can contribute to the success of the PRS process in two ways: (i) in PRSP countries, it can foster contributions to the broader policy debate that could feed into the PRS process; and (ii) bilateral surveillance of advanced countries and multilateral surveillance can identify ways in which policies in advanced countries constrain the achievement of growth and poverty reduction objectives in low-income countries. How well has the IMF performed these roles?

Article IV surveillance in low-income countries would seem to be a natural vehicle for addressing longer-term strategic issues in the IMF’s area of responsibility as well as for providing an input into the broad public debate. To assess what role the IMF has played in practice, we analyzed staff reports and background issues papers for surveillance in PRS cases to identify what types of issues were examined in depth, including the following categories: growth and development issues (e.g., obstacles to private sector development and identification of trade-offs), sustainability and consistency between longer-term plans and macroeconomic conditions (e.g., aid envelope and path of fiscal consolidation, Dutch disease/aid dependence), transition and sequencing issues, and obstacles to growth from market access restrictions of advanced countries. In most instances, surveillance was carried out in conjunction with PRGF program activities. For the 23 countries with a full PRSP by end-2002, there were only 9 stand-alone Article IV consultations (out of over 100 consultations) in the last 4 surveillance cycles (2000–03). The case studies suggest that surveillance generally centered preponderantly around issues directly related to the program, with more limited attention to other issues that the PRS would have needed to address.15 However, surveillance in a number of the case studies did pay attention to various aspects of the problem of weak economic institutions, including regulatory and supervisory frame-works for the financial sector and budgetary processes, and governance issues.

A review of recent Article IV staff reports of major industrial countries suggests that surveillance referred to the impact of these countries’ market access and subsidy policies on low-income countries in only a very general manner. In no case was there a systematic assessment of the impact of these policies on groups of low-income countries, drawn from work done in the IMF or elsewhere. Moreover, the scope of coverage of market access issues in surveillance documents included only a partial treatment of key areas identified by multilateral surveillance, and has varied across country documents.16 Treatment of the relevant trade and subsidy issues gives relatively limited consideration to the impact on low-income countries, with no sense of magnitude of the effect of the country’s policies.

IMF multilateral surveillance (e.g., the World Economic Outlook) has clearly identified key market access issues affecting developing countries.17 The policy instruments that have the most impact on market access for low-income countries comprise tariff and nontariff barriers (mostly on textile and apparel), subsidies (principally to agriculture), and tariff escalation on various industrial goods. This has not translated, however, into analyses of the impact for particular low-income countries. For example, there do not appear to have been any attempts to link systematically the discussion of the individual PRSP country’s strategies and export projections with the constraints implied by advanced economies’ trade and subsidy policies. Such an analysis would be a useful contribution by the IMF to help expand the “partnership” principle so that the policy actions by advanced economies that would most support each country’s growth and poverty objectives would be identified as part of the PRSP process.18

Obstacles to Further Progress in Adapting the IMF’s Role

The evaluation indicates a number of obstacles to adapting the Fund’s role to meet the expectations of the new initiative:

  • The original policy documents raised highly ambitious expectations about the IMF’s role that may have been unrealistic and, in any event, would require much deeper organizational changes—and more resources—than have occurred so far. The quotations cited earlier clearly imply a radical change in the IMF’s own way of doing business, and it is easy to understand why the expectations of outside groups were so high. Discussions with IMF staff, both for the case studies and more generally, indicate that many believe that the initiative is overpromising on what can be delivered in the areas of the IMF’s primary responsibility given existing organizational constraints, which include:
    • A lack of clarity/guidance on how to operationalize a number of aspects of the new approach. These aspects included (a) the role of the IMF in the broader PRS debate;19 (b) the IMF’s contribution to determining an appropriate medium-term external resource envelope; and (c) the IMF’s broader signaling role and how in practice the trade-offs between selectivity (in financial support) and domestic ownership are to be resolved.
    • There has been only a moderate reallocation of resources to address all of the expected new tasks. The share of total IMF administrative expenses devoted to PRSP/PRGF issues was 11.5 percent in FY2003, only slightly higher than the 10.8 percent allocated to ESAF activities in FY1999.20 PRGF countries in Africa appear to have received even less staff resources than other PRGF countries: on average of 2¾ full-time equivalent staff years a year as against 3¼ staff years for non-African countries.21 These are small teams to conduct all of the tasks now expected. Not surprisingly, IMF staff surveyed for the evaluation identified internal resource constraints as one of the major factors constraining changes in the Fund’s role. Indeed, the general impression we received during interviews was that staff often feel over-whelmed by the various procedural requirements, with too little time to devote to analytical and capacity-building activities.
    • The implications of the new approach for the modalities of the IMF’s work have not been fully examined, in particular the relative weight to be given to “informing the debate” and “direct negotiation” activities and the relative roles of resident representatives and missions from headquarters.
  • Vague criteria for assessing IMF performance have resulted in mid-course assessments of the initiative that give insufficient feedback on what the IMF should be doing differently. The lack of concrete indicators on what the IMF is expected to be delivering limited the changes in the IMF’s way of doing business, in the sense that specific priorities for program and surveillance activities were rarely spelled out or linked systematically to the countries’ own priorities, as indicated in the PRS. A lack of sufficient accountability and feedback is present both at the country level and initiative-wide. It also reflects the fact that neither JSAs, nor reports on Progress in Implementation of the PRSP approach (which usually build on the former) undertake to assess systematically IMF performance.
  • Government reluctance to open the policy debate. Some country case studies indicate that a major obstacle to progress has been the government’s reluctance to open up the process to a broader domestic debate on macroeconomic policy issues—as opposed to social sector and some other sectoral areas where much greater progress has been made. More generally, a country’s political and institutional conditions may not be conducive to the change in the mode of IMF engagement envisaged by the PRS approach. Countries that have made most progress in this regard have established an institutionalized structure for seeking the views of stakeholders (e.g., Tanzania and Mozambique). The establishment and functioning of such a framework requires first and foremost government willingness and active engagement. The IMF can encourage the process, but cannot substitute for the government in initiating it.

The quotations are taken from IMF (1999a); World Bank and IMF (1999c); and the Concluding Remarks by the Chairman of the IMF’s Executive Board, IMF (1999b).


Bird (2004) makes a similar point.


Available at http://www/worldbank.org/poverty/strategies/sourcetoc.htm. Like the rest of the Sourcebook, the chapter appropriately emphasizes that it is not intended to be prescriptive and is only to be used selectively as an informational resource. A broader overview of recent IMF macroeconomic research on low-income countries is given in IMF (2003b). The purpose of our review was not to provide a detailed judgment on the merits of specific policy advice per se, but rather to ask two questions about the degree of “policy space” for each policy issue: (i) how detailed are the prescriptions in the Sourcebook; and (ii) how strong is the broader evidence supporting these prescriptions.


The IMF’s own work on the effects of financial integration and liberalization suggests much less definitive conclusions. See IMF (2002e, Chapter 3).


See Tables A4.1 and A4.2 of Annex 4 for details. For each criteria, the processes were rated on a four-point scale in terms of degree of consistency with the PRS/PRGF principles. It should be stressed that this assessment focused on the contents of the briefing papers, not on final outcomes of negotiations. There is no suggestion that ESAF briefing could or should have been “consistent” with approaches that were only developed later, but the comparison gives an indication of what changes have taken place over time.


Examples include upward revisions of the fiscal deficit in Bolivia and Mauritania to bolster social expenditure; formulation of an alternative to VAT on agricultural produce in the Kyrgyz Republic; and the modification to pace and sequencing of financial sector reforms in a number of African countries to allow for the evolution of a domestic consensus on the issue.


Often a “converted” ESAF.


See “Report of the Task Force on the Organization and Management of the African Department” (IMF, 2004c). The report proposed and management accepted a departmental reorganization, including enhanced staff resources, that would give greater prominence to such policy issues.


Other issues with regard to donor coordination are addressed in the OED report.


In a few countries (including Tanzania and Mozambique), groups of bilateral donors providing general budgetary support have agreed on common PAFs to streamline and coordinate conditionality. The World Bank has begun to join such groups. The PAFs sometimes draw upon IMF conditionality or program targets, but there have been no efforts toward broader integration or streamlining.


The IMF has made some progress in developing an analytical framework for debt sustainability to help the country (and its aid partners) avoid borrowing paths—even at concessional rates—that involve a significant risk of entering into new debt problems. See, for example, IMF (2004b).


See IMF (2004d). This issue will be reviewed in more depth in the context of the ongoing IEO evaluation of IMF TA. The discussion here draws upon interviews with TA coordinators in various IMF departments.


IMF resident representatives are present in most countries engaged in the PRSP and PRGF. Of the 23 countries in our sample, only Yemen and The Gambia did not have resident representatives at any time since 1999.


See Nordic Countries (2003) for a donor assessment.


The most notable exception is Vietnam, which did not have a program for over four years. In a number of other cases, surveillance documents have included sections on issues relevant for longer-term growth (e.g., on competitiveness and exports in Guinea, and on agriculture and trade issues in Tajikistan), although such analysis receives only limited attention in the staff appraisals.


See, for example, World Bank and IMF (2003a).


See also World Bank and IMF (2002e). More recently, the Global Monitoring Report (World Bank and IMF, 2004a) has provided a more systematic discussion of the obstacles to reaching the MDGs originating in the policies of developed countries.


This does not imply that the IMF staff itself has to do the detailed empirical work to make such assessments. Moreover, the impact of a general liberalization of market access would likely vary across low-income countries, since some would be more adversely affected by the loss of existing preferential access.


Although IMF staff surveyed for the evaluation identified absence of clear guidelines as of only moderate significance in explaining why there has not been more change in the IMF’s way of doing business, the case studies and interviews with staff indicate that this is because the general assumption has been one of a “hands-off” approach to the broader policy debate.


Data are taken from the IMF’s internal Budget Reporting System. These costs vary from year to year, depending on the intensity of PRGF operations and HIPC-related activities. However, a similar result was obtained for FY2002.


Figures are from the “Report of the Task Force on the Organization and Management of the African Department” (IMF, 2004c). They include staff from nonarea departments (e.g., PDR) regularly assigned on a part-time basis to missions, but do not include resident representatives.

    Other Resources Citing This Publication