A technical workshop on macroeconomic analytical issues in taking the PRSP approach forward was held at IMF headquarters on April 3-4, 2003 bringing together some 40 representatives from low-income country governments, academia, civil society organizations, and multilateral institutions and donor agencies. There was also widespread interest in observing the proceedings of the workshop by Fund and Bank research and operational staff, as well as by the staff of the Executive Directors from both institutions.

Summary of Discussions

A technical workshop on macroeconomic analytical issues in taking the PRSP approach forward was held at IMF headquarters on April 3-4, 2003 bringing together some 40 representatives from low-income country governments, academia, civil society organizations, and multilateral institutions and donor agencies. There was also widespread interest in observing the proceedings of the workshop by Fund and Bank research and operational staff, as well as by the staff of the Executive Directors from both institutions.

The purpose of the workshop was to take stock of the current state of research and analysis on macroeconomic analytical issues of importance to countries' design and implementation of their national poverty reduction strategies, with a view to: (1) assessing the knowledge base and the degree to which the guidance put forward in the draft background papers was considered appropriate by countries and development partners; and (2) setting a priority international macroeconomic research agenda on low-income countries.

Drawing on three of the priority macroeconomic issues identified by the May 2002 joint Bank/Fund Review of the PRSP and the IMF Review of the PRGF, the first three sessions of the workshop examined background papers prepared by the IMF, World Bank, UNDP, government of the United Kingdom, and government of the Netherlands on macroeconomic programming and poverty reduction; macroeconomic consequences of larger aid flows; and fiscal and external debt sustainability. The final session focused on knowledge gaps, research priorities and next steps. The following is a summary of each of these sessions.

Session 1: Macroeconomic Programming and Poverty Reduction

Overview: This session focused on the need for realism in macroeconomic projections and considered possible approaches for bridging the gap between the frameworks presented in PRSPs and those in annual budgets and associated PRGF-supported programs. The question of alternative fiscal deficit paths was also discussed, including with regard to the implications for inflation, real interest rates and growth. Key questions included: How to reconcile the need for ambition with the need for realism regarding projections of growth and other macroeconomic variables? Does the proposed "two-scenario" approach adequately address this issue and, if not, is there a better approach that could achieve the same objective? If so, how can the linkages between PRSPs and budgets be strengthened?

Summary: The main issues raised during the discussion can be summarized as follows:

• Instilling Realism: There was broad-based recognition that projections of growth and other macroeconomic variables underpinning the macroeconomic frameworks of countries' poverty reduction strategies and PRGF-supported programs have been unrealistic. In the short-term, projections should be cast against the backdrop of past trends and cross-country comparisons. In the medium-term, they should be grounded in an analysis of the drivers of growth. Projections could also benefit from greater (regional) disaggregation, analysis of both the demand and the supply side, and inclusion of the labor, as well as the capital/investment, dimensions. Further research and analysis on the sources of and obstacles to growth at the country level should be a priority.

• Single versus two scenario approach: There was widespread concern expressed about the Fund staff's proposal that countries adopt a "two scenario" approach to their macroeconomic frameworks. This would not only place an additional burden on the already strained administrative capacity of low-income countries, but also call into question the operational relevance of the PRSP itself. A preferred approach would be to have a single realistic macroeconomic framework with a contingency plan for what to do in the event that assumptions regarding policy implementation, institutional reform, and external financing proved too conservative (or too optimistic). There is also a need for greater country ownership of the macroeconomic frameworks underpinning their strategies. If countries do not have the capacity to carry out the analysis themselves, development partners need to assist countries in building such capacity.

• Risk analysis: The need for risk analysis of macroeconomic projections and for greater flexibility with regard to how macroeconomic frameworks respond to exogenous shocks was raised repeatedly. Stress testing of key variables should be systematically carried out, including for deviations in key world commodity prices and shortfalls in foreign assistance. The inclusion of confidence intervals could also be usefully explored.

• Avoiding unnecessary complexity: There is a need to avoid macroeconomic models that are too complex to be used in-country. In turn, there is a need to develop relevant, user-friendly models, and to help countries build capacity to employ these.

• Poverty and social impact analysis: The process for undertaking PSIA was considered to be as important as the tools and techniques themselves. Early involvement and ownership by stakeholders is essential if PSIA is to have the desired impact on decision-making. While some called for the distributional impact of the entire macroeconomic framework (and not just of individual policies) to be carried out, it was generally recognized that this would be problematic. Knowledge and experience should therefore be built up in this regard over the coming 12-18 months.

• Deficiencies in data and data systems: There was general recognition that data systems needed to be strengthened so that countries could develop better growth strategies and conduct distributional analyses of associated policies. Greater attention needed to be placed on gaps in data systems and whether the international community's present approaches to improving these systems are paying off or are ill-designed.

• Early involvement of Fund staff: IMF staff needed to engage earlier-on in the PRSP process to assist countries in developing their macroeconomic frameworks and considering policy choices and alternatives. The Resident Representative had an important role to play in this regard and many participants believed that they should be given greater authority.

Session 2: Macroeconomic Consequences of Larger Aid Flows

Overview: This session focused on the question of the macroeconomic consequences of larger aid flows, particularly with regard to competitiveness and fiscal and external sustainability. The papers under discussion maintained that, if properly managed, aid had a positive effect on growth. However, due consideration was needed regarding the possible impact of these flows on the real exchange rate and on the absorptive and administrative capacity of the country. The workshop discussion, however, broadened into the question of overall aid effectiveness and what both donors and countries should do about it. Key topics considered included the impact of aid on domestic resource mobilization efforts, the predictability and volatility of aid and its impact on growth, the use of grants versus loans for economic development, and the need for increased transparency.

Summary: The main issues raised during the discussion can be summarized as follows:

• Assessing the impact of aid: It was noted that donors tend to rally around good performers, which contributed to the problem of the impact of large aid flows on a country. Even so, there was general recognition that the relative size of increases in aid flows was not likely to be large enough to pose problems regarding Dutch Disease and that absorptive and institutional capacity constraints were the real concerns. The relative size and composition of the non-tradable goods sector also needed to be analyzed in order to determine if resource shifts resulting from an exchange rate appreciation would have a negative effect on the poor. Some argued that positive productivity gains to be had from aid flows directed at addressing structural rigidities could in fact offset the negative effect of any exchange rate appreciation on the tradable sector. Questions were also raised as to what is the appropriate time frame to measure the results and effectiveness of aid.

• Harmonization: While there has been a lot of progress on donor harmonization, problems remained. Some questioned whether there could be too much harmonization, which could lead to a common negotiated plan rather than simply common procedures. Concerns were expressed about conditionality getting out of hand, with problems in one area leading to a shutting down of donor assistance for the entire program. One participant even suggested that there was something to be gained by playing one donor off of another, although the general consensus was there were gains to be had--for countries and for donors--from greater harmonization of procedures and practices. Mention was made of a ongoing donor coordination pilot program being carried out under the auspices of the Strategic Partnership with Africa (SPA) in collaboration with the IMF, the World Bank, the European Union, and the United Kingdom Department for International Development.

• Capacity building: Emphasis was placed on the need for more "permanent" capacity building given the constant movement and turnover of country personnel. Moreover, the need to build capacity within "countries" and not simply within "governments" was stressed, particularly with regard to civil society and parliaments.

• IMF Research: Ongoing research at the IMF is looking into whether monetary policy can be used to offset any real appreciation of the exchange rate resulting from higher aid flows. Evidence to date indicates that there is an association between aid flows and real appreciation above and beyond terms of trade effects and country trends. One participant queried whether there is a risk that research conducted by the IMF would be perceived as being biased and that there was a need therefore for more independent research on this and other issues.

Session 3: Fiscal and External Debt Sustainability

Overview: This session focused on issues related to fiscal and external debt sustainability in low-income countries. Topics discussed included what indicators, time-frame, assumptions, and sensitivity tests are most useful for assessing sustainability. A preliminary draft template on fiscal and debt sustainability was put forward by the IMF for consideration and the consensus was that this template was useful, subject to certain caveats. It was broadly recognized that further work would be required, particularly with regard to the policy implications of the template.

Summary: The main issues raised during the discussion can be summarized as follows:

• Definition of sustainability: No general consensus emerged in the discussion regarding the definition of sustainability. Participants agreed that debt sustainability can be both very easy and very difficult to define. While it is easy to define it ex post as the point at which a country defaults, it is more difficult to precisely predict ex ante when this will happen. The choice of the appropriate definition of debt sustainability is therefore an issue on which there is still no consensus among researchers, and further thought and analysis is needed. As many participants pointed out, besides an "economic" definition of debt sustainability, policymakers also have to pay attention to the "political" and "social" sustainability dimensions of public and external debt. In particular, many participants raised the issue of a possible trade-off between fiscal consolidation needed to achieve debt sustainability objectives and poverty-reduction targets.

• Analyzing sustainability: Most participants agreed that the Net Present Value (NPV) of debt was a better indicator of sustainability than the nominal value, but that alternative indicators may also be useful; for example, debt service and gross financing need ratios may help capture debt-related liquidity constraints faced by governments. Moreover, several participants stressed that the definition of sustainability is highly country-specific and must be related not only to the level of debt that is deemed sustainable, but also to the path to achieve it. There was also consensus that a 10-20 year time frame is appropriate, although with the caveat that this increases the uncertainty of the projections. Many participants noted that it is important to consider domestic as well as external debt in low-income countries and welcomed the intention of the Fund and the Bank to supplement external debt sustainability with fiscal sustainability analysis in low-income countries.

• Assumptions: There was widespread agreement that more thought needs to go into the assumptions that underlie baseline scenarios. In particular, it is clear that growth is the single most important factor affecting fiscal and external debt sustainability. Thus, to have reliable analyses and better policies, one needs to have a better understanding of the determinants of growth and its links to the policies implicit in sustainability analyses. Similarly, several participants stressed the need to think carefully about what assumptions should be made about future aid flows. In particular, there was some significant support for moving more towards grants-based aid. Many participants stressed the importance of fully capturing the policy implications of such a change in the formulation of forecast scenarios and for the choice of debt sustainability indicators.

• Stress tests: While many participants considered the set of stress test used in the debt sustainability framework appropriate, some argued in favor of considering non-economic stress factors as well. For example, HIV-AIDS pandemics, natural disasters, and conflicts are likely to cause large short-term and permanent shocks to major economic variables. However, some participants thought that such shocks could be subsumed under the more general shocks to real GDP growth.

• Bigger questions: Beyond the draft template, bigger questions were raised by the participants. Low-income countries face major challenges in financing their PRSPs and, ultimately, making progress towards achieving the MDGs, yet there are limits to current resources. Countries also need to improve their capacity to deliver on the reforms required to make progress on these fronts. A range of potential solutions were discussed. Some argued that, given a potential low growth scenario, this may require full debt cancellation and a major increase in grant flows in order to avoid another debt trap and further debt write-offs. Fiscal and debt sustainability analysis should therefore hope to highlight the dangers of this lower than projected growth, but also suggest means for overcoming constraints. Others considered a more optimistic scenario whereby some countries could achieve higher levels of growth and borrow at sustainable levels. Questions remain as to whether HIPCs can take on more IDA resources in order to achieve growth and poverty reduction (given that their debt levels will then rise above the threshold of 150 percent debt/exports) and, for low-income countries in general, whether additional finance can be mobilized to support more rapid progress toward achieving the MDGs.

Session 4: Knowledge Gaps, Research Priorities and Next Steps

Overview: This session focused on identifying a set of priority research projects to be carried out over the medium-term, including by the IMF, the World Bank, other development partners, and academia. There was broad consensus on a set of research topics that should receive highest priority, as well as those which were considered to be of less importance. Issues regarding how best to approach the carrying out of such research were also discussed. A proposal by the IMF, World Bank, United Kingdom, and the Netherlands as to how to take this process forward was also briefly discussed, with details to be circulated to the participants in the aftermath of the workshop

Summary: The main issues raised during the discussion can be summarized as follows:

• The priority areas for further research included:

  • Sources of growth: There was broad consensus that the international community needed to get a better handle on the drivers of growth and what causes growth traps and growth reversal. There was also a recognized need to get a better understanding of what is meant by pro-poor growth and of the policies necessary to bring such growth about. The impact of HIV/AIDS on growth in those countries particularly affected by this pandemic also needs to be better understood.

  • Impact of shocks: Priority attention needs to be placed on commodity price instability and the policies required to address such instability. Risk management regarding exogenous shocks and the implications for reforms in structural and other policy areas require further reflection. Similarly, the impact of large shocks such as regime change and conflict on growth and poverty needs to be further explored.

  • Trade and investment: Obstacles to private sector development, improvements in the investment climate and the importance of trade on growth and poverty in low-income countries requires priority attention. In addition, the influence of policies in the north on trade and poverty in the south needs to be further analyzed.

  • PSIA of macro policies: Some participants believed that despite the current tool kits and existing methodologies, greater research was required on the poverty and social impact of the macroeconomic framework as a whole, in addition to that on individual policies. While this is not an easy exercise, research could begin by identifying the obstacles to such analyses.

  • Gender and the environment: Participants widely believed that gender and the environment needed to be mainstreamed in the global research effort.

• Some participants believed that further research on the macroeconomic consequences of higher aid flows and on fiscal and external debt sustainability should not be a priority for the IMF and the World Bank. Others (including the representatives from the IMF and the World Bank) believed that further research on the macroeconomic consequences of larger aid flows should be pursued, as it remained an important topic in the institutions' dialogue with member countries.

• How research is to be conducted: The focus of our research efforts needs to be on policy-based and/or applied research. The participants felt strongly that partnerships in research need to be neither all Washington-based nor all country-based. Rather, there is a need to build networks between the various partners in research. There is also a need to deemphasize the use of cross-country regressions in analyzing low-income country issues and focus more on country-specific research. The way economic research and analysis is conducted needs to be broadened, including the perspective of political economists since many of the implementation problems faced by low-income countries are political economy in nature. At the same time, research cannot be all demand driven and there is a legitimate need to respond to academic debate and allow for intellectual curiosity. Data remains a critical constraint to research and policy analysis and needs to be addressed urgently with technical assistance and support.

• Next steps: a summary of the workshop discussions will be circulated to all participants for their review and comment before being placed on the external websites of the IMF and the World Bank. An IMF, World Bank, United Kingdom, and the Netherlands joint proposal for taking the research effort on macroeconomic issues forward will also be circulated to participants with a view to inviting other stakeholders to participate in this collaborative effort. This is envisaged to be a long-term agenda for policy-oriented research, with the IMF and World Bank working closely with other stakeholders including significant input from low-income countries. A research conference is tentatively being planned for 2004 where researchers on low-income country macroeconomic issues can share findings of their works-in-progress and where the international research agenda can be visited based on recent developments and experience in implementing the PRSP approach.