Report No. 25160 The Heavily ndebted Poor Countries (HIPC) Debt nitiative An OED Review February 20, 2003 Operations Evaluation Department Document of the World Bank I I Abbv io ns and Acronyms CAE Country Assistance Evaluation CAS Country Assistance Strategy CDF Comprehensive Development Framework CP Completion point CPIA Country Policy and Institutional Assessment DAC Development Assistance Committee DFID Department for International Development DP Decision point DRS Debtor Reporting System DSA Debt sustainability analysis E-HIPC Enhanced HIPC Initiative ESAF Enhanced Structural Adjustment Facility GAO General Accounting Office GDP Gross domestic product GNI Gross national income HIAL Higher Impact Adjustment Lending HIPC Heavily Indebted Poor Country IBRD International Bank for Reconstruction and Development IDA International Development Association IEO Independent Evaluation Office IF International Financial Institution IMF International Monetary Fund I-PRSP Interim Poverty Reduction Strategy Paper LIC Low-income country MIC Middle-income country NGO Non-governmental organization NIEO New International Economic Order NPV Net present value ODA Official development assistance OECD Organization of Economic Cooperation and Development OED Operations Evaluation Department O-HIPC Original HIPC Initiative PAF Poverty Action Fund PEAP Poverty Eradication and Action Plan PER Public Expenditure Review PRGF Poverty Reduction and Growth Facility PRSP Poverty Reduction Strategy Paper SSA Sub-Saharan Africa UNCTAD United Nations Conference on Trade and Development Director-General, Operations Evaluation : Mr. Gregory K. Ingram Director, Operations Evaluation Department (Acting) Mr. Nils Fostvedt Manager, Corporate Evaluation and Methods Ms. Victoria Elliott Task Manager : Mr. Madhur Gautam Foreword The creation of the Heavily Indebted Poor Countries (HIPC) Debt Initiative marked a turning point in the evolution of development finance. The HIPC Initiative has been a catalyst for far-reaching changes in the processes surrounding development assistance, reflecting the coming of age of a new authorizing environment with the active participation of civil society. It has introduced greater transparency and accountability in the sovereign debt regime and raised development cooperation to a higher plane, including between the Bank and IMF. This review by OED finds the HIPC Initiative highly relevant in addressing a key obstacle facing many poor countries. If the anticipated debt relief is delivered in full, the Initiative will succeed in substantially achieving its fundamental goal of reducing the excessive debt burden of the qualifying countries. But the legitimizing process that helped make the Initiative a reality has also expanded its objectives. The Initiative seeks to provide a "permanent" exit from debt rescheduling, to promote growth, and to release resources for social expenditures targeted at poverty reduction. Achieving these objectives will require actions by donors and the HIPC governments that are beyond the scope and means of the Initiative. Unmanageable debt is a problem that needs to be effectively dealt with, but it is also a result of economic and political factors constraining growth and poverty reduction. The HIPC Initiative is thus an important but small part of the overall development assistance framework. Having provided the HIPCs with an opportunity for a "fresh start," the international community still faces a challenge in helping these countries set out on a sustainable path for growth and poverty reduction. This requires actions by the HIPC governments to adopt sound policy frameworks and a balanced development strategy. It also requires actions by the international community to assist the countries to enhance their exports and build needed institutional capacities. A further challenge for donors, consistent with one of the Initiative's main guiding principles, is to provide adequate resources to meet the development priorities of HIPCs as well as other poor countries, and ensure that the HIPC debt relief is truly additional to other aid flows. The review makes four recommendations addressing the strategic issues facing the Initiative. The first is to clarify and communicate the purpose and objectives of the Initiative, and to ensure that its design is consistent with these objectives. The second recommendation is to make explicit the methodology and economic models underlying the debt projections used in the debt sustainability analyses, and to make the economic forecasts more realistic to assess better the prospects and risks facing individual countries. Third, the Initiative should maintain the standards for HIPCs' policy performance to ensure that the risks to achieving and maintaining the Initiative's objectives are minimized. And when flexibility is desirable, there should be a clear and transparent rationale for relaxing the criteria. Finally, the review recommends a greater focus on pro-poor growth to provide a better balance among development priorities relative to the current emphasis on social expenditures. Table of Contents Ackno w ledgm e t .................................................................................................................... v Executive Sum m ary .............................................................................................................. vii 1. Introduction ....................................................................................................................... 1 Progress and Current Status......................................................................................... Evaluation Design ........................................................................................................ 3 2. Evolution of Debt and the Debt Problem ........................................................................ 4 Evolution of the HIPC D ebt Burden ............................................................................. 4 The Consequences of an Excessive External Debt Burden........................................... 7 3. The H IPC Initiativ ........................................................................................................... 8 Background................................................................................................................... 8 The Political Economy of HIPC.................................................................................... 9 The HIPC Objectives................................................................................................... 10 The Process ................................................................................................................. 12 4. Design of the Initi tive .................................................................................................... 14 Background ................................................................................................................. 14 Consistency of the D esign with Overall Objectives.................................................... 14 D ebt Sustainability...................................................................................................... 18 The Debt Sustainability Criteria...................................................................... 18 D o D SA Projections Provide a Robust Forecast? ........................................... 22 Perform ance Criteria.................................................................................................. 24 Track Record Requirem ent ............................................................................. 24 Com pletion Point Triggers.............................................................................. 25 Creditor Participation................................................................................................. 26 Capacity Building........................................................................................................ 26 5. Im plem entation Experience ........................................................................................... 27 Pace of Implementation............................................................................................... 27 D ebt Sustainability Analysis........................................................................................ 27 Perform ance Criteria.................................................................................................. 31 Track Record................................................................................................... 31 ii Completion Point Conditions.......................................................................... 32 Creditor Participation................................................................................................. 37 Institutional Development: Capacity Building for External Debt Management......... 38 6. Likey Outcom e& - Prelimnai ry Findlngs ..................................................................... 40 Debt Sustainability ...................................................................................................... 40 Completion Point Countries............................................................................ 40 The Interim Countries ..................................................................................... 44 Prospects for Debt Sustainability .................................................................... 45 Progress Toward Performance Criteria ..................................................................... 46 Incremental Resource Trane ................................................................................... 50 Debt M anagement Capacity ........................................................................................ 55 Likely Outcomes.......................................................................................................... 56 7. indings and Recom m endatons .................................................................................... 5o Recommendations........................................................................................................ 62 Refeirenees .............................................................................................................................. 65 Annexes Annex A: HIPC Initiative: Status of Country Cases Considered under the Initiative, July 2002................................................................................................................................. 69 Annex B: The Objectives and Guiding Principles for the HIPC Initiative .............................70 Annex C: HIPC Initiative Flow Chart.....................................................................................71 Annex D : Evaluation Approach..............................................................................................72 Annex E: Change in Net Transfers to HIPC by Source..........................................................74 Annex F: How Does a Large Debt Stock Affect Economic Performance? ............................75 Annex G: Debtor Country Perspectives on the Heavily Indebted Poor Country (HIPC) Initiative ..........................................................................................................................78 Annex H: Creditor Country Perspectives on the Heavily Indebted Poor Country (HIPC) Initiative .......................................................................................................................... 82 Annex : Synthesis of M ain Findings from Case Studies ....................................................... 87 Annex J: Constraints to Growth as Identified by OED Country Assistance Evaluations.......93 Annex K: M anagement Response...........................................................................................95 Annex L: Report from the Committee on Development Effectiveness - Chairman's Summ ary ....................................................................................................................... 102 Boxes Box 1.1: Status of the 42 Eligible HIPCs (as of August 2002)................................................. 2 Box 4.1: Uganda's Influence on the Design of the HIPC Initiative........................................ 15 Box 5.1: Debtor Countries Endorse the Links between the HIPC Initiative and the PRSP.... 33 iii Box 5.2: PERs Highlight Absorptive Capacity Constraints in the Social Sectors in the HIPCs .636 Box 5.3: Increasing World Bank Involvement in Debt Management Capacity Building.......38 Box 6.1: Topping-Up and the Prospects for Debt Sustainability for Burkina Faso................46 Tables Table 2.1: External Debt as percentage of GDP (period average)...........................................4 Table 3.1: The Original and Enhanced HIPC Frameworks in a Nutshell............................... 11 Table 5.1: Comparison of DSA Export Growth Projections and Historical Performance......29 Table 6.1: Debt Sustainability Indicators for O-HIPC Graduates...........................................40 Table 6.2: Debt Sustainability Indicators for E-HIPC Graduates...........................................41 Table 6.3: Countries with Difficulties Staying on Track with Macroeconomic and Structural Program s in 200 1 ............................................................................................................47 Figures Figure 2.1. Evolution of the HIPC Debt Burden ........................................................................ 4 Figure 2.2: Net Resource Transfers to HIPCs Have Outpaced Debt Service ........................... 5 Figure 4.1: HIPC Target for Debt-to-Export Ratio Compared with Other Poor Countries .... 21 Figure 4.2: What Would Be a Good Debt-to-GDP Target?...................................................21 Figure 5.1: Structural Adjustment Compliance Fosters GDP Growth....................................30 Figure 6.1: Projected Debt-to-Export Ratios (2000--10).........................................................42 Figure 6.2: Projected Debt Service Ratios (2000-10).............................................................42 Figure 6.3: Key Factors Affecting Debt Sustainability: Status of E-HIPC Completion Point C ountries .........................................................................................................................43 Figure 6.4: Policy Performance Average Scores of the Early Entrants, the "Millennium Rush" G roup, and Late Entrants ................................................................................................48 Figure 6.5: Declining Aggregate Net Transfers to HIPCs ......................................................53 Figure 6.6: Sharply Declining Aggregate Net Transfers to All Developing Countries (excluding A sian crisis countries)...................................................................................53 Figure 6.7: HIPCs' Growing Share of Aggregate Net Resource Transfers.............................54 V Acknowledgments This report was prepared by a team led by Madhur Gautam. Members of the core team were Georgia Wallen, Shonar Lala, and Yusuf Ahmad. Victoria Elliott provided overall guidance to the team as manager of the Corporate Evaluation and Methods Group, with substantial input into the formulation of the final report. Professors Thomas M. Callaghy and Jonathan Eaton contributed key background papers and have advised the team at critical stages of the review. Magdalene Apenteng, William Battaile, Sunday Khan, Michael Lay, and Mirafe Marcos contributed to the country case studies. Additional inputs were provided by Judith Hahn, Robert Keyfitz, Carlos Reyes and P.A. Sharafudheen. Parveen Moses provided administrative support, with assistance from Annisa Cline-Thomas and Julia Ooro. William Hurlbut and Caroline McEuen provided editorial assistance. The evaluation has benefited from discussions, comments, and assistance from numerous people. Special thanks are due to former and current staff of the World Bank's HIPC unit, in particular Mark Dorfman, Francis Earwaker, Anthony Gaeta, Charlene Adam Gust, Young Chul Kim, Jacob Kolster, Vikram Nehru, and Axel van Trotsenberg, for their cooperation at all stages of the review. Inputs from the following persons are acknowledged with gratitude: Lisandro Abrego, Masood Ahmad, Oliver Buston, John Cady, Punam Chuhan, Stijn Classens, Shantayanan Devarajan, Sumana Dhar, Geske Djikstra, William Dorotinsky, Ceri Edmonds, Ibrahim Elbadawi, Mahyar Eshragh-Tabary, Gerald Flood, Indermit Gill, Romilly Greenhill, Jo Marie Griesgraber, Bernhard Gunter, Sanjeev Gupta, Maria lonata, Keith Jay, Kadima Kalonji, Nawal Kamel, Jeffrey Katz, Heinz Kaufmann, Tony Killick, Jeni Klugman, Matthew Martin, Calvin McDonald, Rob Mills, John Mitchell, Ashoka Mody, Catherine Pattillo, Ann Pettifor, Malvina Pollock, Robert Powell, Doris Ross, Alison Scott, Kathy Selvaggio, Sudhir Shetty, Veena Siddharth, Miriam Tamane, Melissa Thomas, Rachel Turner, D.C. van der Hoek, Nicholas Vaughan, Jian-Ye Wang, Howard White, Alexandre Widmer, John Williamson, and Hanspeter Wyss. OED is grateful to the Swiss Agency for Development Cooperation for financial assistance to support a number of activities, including the multi-stakeholder workshop and background research. OED also thanks the U.K. Department for International Development for sponsoring the Lilongwe and London workshops, Debt Relief International for organizing the workshops, and the finance ministers and their representatives from the 24 HIPCs who participated in these two workshops for their time and invaluable inputs. The evaluation benefited greatly from the feedback from the World Bank's executive directors and their advisers from 14 creditor countries on their perspectives on the Initiative. Thanks are due to the Governments of Cameroon, Guyana, Malawi, Uganda, and Zambia for agreeing to be case study countries for this review, and to the numerous government officials and civil society and donor representatives interviewed in these countries. The country case studies also benefited significantly from the assistance and inputs of the respective World Bank and IMF country teams. Finally, the evaluation team is grateful to the participants of the multi- stakeholder technical workshop held in Washington in September 2002 for their time and inputs. vii Executive Summary 1. The Heavily Indebted Poor Countries (HIPC) Debt Initiative marks a major innovation in the development finance regime. Relative to past efforts, it is more concerted across creditors and more comprehensive in its attempt to reduce the high external debt- including, for the first time, multilateral debt--of many of the poorest countries. It has made the processes surrounding the sovereign debt regime more open and accountable. The Initiative also marks a significant break with the past in how development aid is approached. Its design embodies lessons of experience linking aid effectiveness with the policy environment and aid coordination, conditionality with ownership, and social impacts of macroeconomic policy reforms with public expenditure prioritization. 2. Public concern with excessive debt burdens together with declining aid resources and poor performance in poverty reduction provided the impetus for debt relief. With the vocal support of advocacy nongovernmental organizations, these concerns came to be shared by pragmatic policymakers in donor governments and International Financial Institutions. The overarching poverty reduction mission of the development community became the core justification for the HIPC Initiative, with considerable political resonance in the wake of the Jubilee campaign. By 1999, the enhanced-HIPC explicitly identified debt sustainability and poverty reduction as its twin objectives. 3. The HIPC Initiative was triggered by a confluence of factors including the ascendancy of international civil society organizations and their growing influence on major creditors, and a change in the World Bank's leadership. Debtors had no explicit role in the design of the original Initiative and limited influence on the design of the enhanced Initiative, even though they are central to its implementation. 4. The specific objectives of the Initiative have evolved.' The original goal in 1996 was to remove the debt overhang as a constraint to economic growth and poverty reduction. The modifications introduced in 1999 brought not only deeper, broader, and faster debt relief, but also a more ambitious and expanded set of objectives-to provide a "permanent" exit from debt rescheduling, to promote growth, and to release resources for higher social spending targeted at poverty reduction. The need to create fiscal space for social expenditures was a critical prerequisite of broad-based support for the Initiative among donors, and has had a major impact on the objectives of the Initiative, its design, and its implementation. 1. Management comment: The context of this evolution is worth noting. The enormity and complexity of the development challenges facing the highly indebted poor countries were well recognized. Overcoming these challenges needed the joint efforts of the countries themselves together with the concerted financial and technical support of the international community. More recently, the Monterrey and Johannesburg Summits acknowledged this and helped enormously in shaping and advancing this dimension of the global agenda. Using a broad range of instruments, the Bank was already heavily engaged in supporting development in low-income countries, in partnership with other multilateral agencies, bilateral agencies, HIPC governments, and civil society. The HIPC Initiative became one element-albeit an important one-in this broad effort. It was seen as just one building block in more comprehensive development architecture, with sustainable development and poverty reduction as its ultimate goals. viii 5. If the anticipated debt relief is delivered, the Initiative will succeed in substantially reducing most HIPCs' external debt stocks and their debt service (on average) to below the levels of other poor countries. Thus, the Initiative is likely to achieve its original goal-to reduce the external debt burden of several of the poorest countries and give them a "fresh start." But to simultaneously achieve all three objectives-specifically to free up resources for increased social expenditures-the Initiative would have to transfer additional real resources to the countries. And to do this without diverting aid resources from poor but not highly indebted countries, the overall level of aid resources would have to increase. Although the Initiative's objectives are predicated on the assumption that (other things being equal) past aid levels would be maintained, so that HIPC resources would be additional, the Initiative's design cannot ensure that this will happen. The Initiative thus faces the risk of promising outcomes (related to its multiple objectives) that it cannot deliver by itself. 6. In fact, there was a sharp decline in global net resource transfers starting about the time the Initiative was created. As a result, although the HIPCs as a group are getting an increasing share of declining global aid resources relative to other poor countries, they are not receiving additional funds in absolute terms compared with what they were receiving before the creation of the Initiative (that is, until 1995). Because net transfers were depressed in the years immediately before most countries qualified for debt relief, these years are not the most appropriate benchmark for establishing additionality. It is too early to quantitatively determine whether there is a reversal in the recent declining trend. To the extent that the Initiative has succeeded in protecting the HIPCs' share of aggregate aid flows, it may be judged to be a limited success. However, it appears that the share of other poor countries has declined correspondingly. The resulting redistribution conflicts with the principle of performance-based allocation and could reduce the overall efficiency and effectiveness of aid. This outcome is a direct consequence of funding limitations, and it cannot be overcome through design improvements internal to the Initiative as currently conceived. A clear acknowledgment of the limitations imposed by past and current aid levels would facilitate the realignment of the Initiative's basic objectives with the resources actually available. 7. A key element in assessing the Initiative's likelihood of achieving its core objective of debt sustainability is the projection of debt indicators. The Initiative uses a debt-inventory methodology for assessing current debt levels that is a clear improvement over past practice and a sound basis for calculating the amount of debt relief for individual countries. The Initiative also projectsfuture debt levels to assess each country's likelihood of achieving debt sustainability. The methodological basis underlying these projections in the Debt Sustainability Analyses (DSAs) has not been made transparent, and the growth assumptions maintained in the DSAs have been overoptimistic in relation to historical growth rates. Debt indicators are influenced by how future debt levels, exports, and other macroeconomic aggregates evolve. The economic model behind these projections needs to be made explicit, and the economic forecasts that are the basis of projections should become more realistic. In particular, they need to better capture the potential effects of volatility in export earnings-a key risk factor. Improved risk analysis would provide a better assessment of each country's likelihood of meeting the Initiative's debt sustainability threshold. This would not by itself improve the prospects for debt sustainability, but it might foster a more informed debate about the policy changes needed in donor and recipient countries alike-as well as greater realism in setting objectives and funding arrangements for the Initiative. ix 8. Meeting the objective of debt sustainability will be a challenge for the HIPCs on both the debt and revenue sides. These countries may need to continue to borrow to meet their development needs, especially in the absence of adequate grants. Their main challenge is to ensure that the funds are invested productively and efficiently to promote repayment capacity. To do so requires improved public expenditure management along with intensified and sustained efforts to accelerate economic growth. On the revenue side, the fiscal base is narrow and the export base is typically concentrated in a few commodities subject to highly volatile market conditions. These countries need to address fiscal constraints and other policy obstacles to more rapid, broad-based growth and diversified exports. In turn, this may require improved trade facilitation services and better access to developed country markets. 9. A necessary condition for accelerated economic growth is adoption of a sound policy framework that will produce economic stability, effective public expenditure management, and efficient and nondistorting revenue generation. A main requirement for qualification for HIPC relief has from the start been a track record of strong policy performance. The application of this requirement was progressively reduced in the enhanced HIPC, particularly for the "millennium rush" countries that qualified in the second half of 2000. Many of these countries have yet to demonstrate an ability to put such frameworks in place, which raises concerns about the achievement of the HIPC objectives. 10. The Initiative's guidelines for increased public expenditures emphasize social sectors relative to other sectors where public expenditures can also help enhance economic growth. There are two disadvantages to this. First, the performance criteria emphasize expenditures rather than outcomes or impacts, yet increased expenditures may encounter diminishing returns in the short and medium run. The capacity of education and health ministries to manage increased budget resources efficiently is often weak, a substantial share of aid resources are already targeted at social expenditures, and the World Bank's own public expenditure reviews indicate that financing is not always the primary constraint to achieving outcomes. Absorptive capacity constraints in the targeted sectors and the need for investment to promote growth may warrant a different balance between social and other sectors, especially infrastructure and rural development. Second, the inflexibility in allocation of HIPC resources is a major concern voiced by debtor-country representatives. They note that the external strictures on their resource allocation can weaken budget discipline and domestic ownership. Debtor governments believe that these two issues are undermining the achievement of the main HIPC objectives. 11. To sum up, excessive debt creates problems, and must be dealt with effectively. Yet unmanageable debt is a symptom of deeper structural problems. While providing much- needed respite, as the Initiative appears likely to do, debt relief is not a panacea for broader economic development problems, nor is a one-time debt reduction a guarantee that the problem will not re-emerge. The biggest challenge facing the Initiative is the expectations of what it can achieve at current funding levels given policy and institutional constraints. Achieving its multiple objectives requires actions that are well beyond the scope and means of the Initiative. The achievement and sustainability of the individual objectives require actions by HIPC governments to promote exports and broad-based growth, together with human capital development, for sustained poverty reduction. x 12. The review makes four recommendations. The first is to clarify the purpose and objectives of the Initiative, ensure that its design is consistent with these objectives, and communicate both the objectives and how they are to be achieved clearly to the global community. The second recommendation is to improve the transparency of the methodology and economic models underlying the debt projections and the realism of the economic growth forecasts in the debt sustainability analyses, to guide decisionmaking through a better assessment of the prospects and risks facing individual countries. The third recommendation is to maintain the standards for policy performance. And when the established criteria are to be relaxed, provide a clear and transparent rationale to ensure that the risks to achieving and maintaining the Initiative's objectives are minimized. Finally, the review recommends that there needs to be a greater focus on pro-poor growth to provide a better balance among development priorities relative to the current emphasis on social expenditures. 1 1. Introduction 1.1 After almost two decades of repeated attempts to relieve many low-income countries of their external debt burdens, in 1996 the World Bank and the International Monetary Fund (IMF) proposed the Heavily Indebted Poor Countries (HIPC) Debt Initiative to provide comprehensive debt relief to some of the world's poorest and most heavily indebted countries. The Initiative was formally agreed to by the governments around the world in September 1996 and "enhanced" in 1999. The objectives of the Initiative have subtly evolved since its conception. In response to intense pressure to make debt relief broader, faster, and deeper, the goals of the Initiative were markedly modified in 1999. But fundamentally the Initiative still aims to reduce, within a reasonable time, the external debt burden of qualifying countries to "sustainable" levels. The underlying and plausible premise is that excessive debt is an impediment to the broader development goals of sustainable economic growth and poverty reduction. 1.2 The HIPC Initiative marks an important innovation in the development finance regime, with significant changes in the relationships of the World Bank and IMF with member countries and development partners. It represents a significant shift in the strategy to deal with the persisting debt crises and associated repayment problems facing low-income countries, acknowledging the HIPCs' problem as one of insolvency rather than illiquidity and recognizing the need for different actions than those taken in the past. It is designed as a concerted and comprehensive approach to deal with the external debt of poor countries in its entirety, and with an explicit objective of resolving it in a sustainable way. For the first time, it includes multilateral creditors and broadens the scope of performance requirements to include social criteria along with the macroeconomic and structural policy reform conditionality of traditional debt relief mechanisms. 1.3 The emergence of the original HIPC (O-HIPC) Initiative in 1996, and-even more- the enhanced HIPC (E-HIPC) Initiative of 1999, has been enormously influenced by nongovernmental organizations (NGOs) and civil society groups around the world. The Jubilee 2000 movement was particularly effective in making the enhanced Initiative a political reality in creditor capitals at a time of declining aid resources. The process of legitimizing and promoting the Initiative forged a direct link between debt relief and poverty, but it also gave rise to unrealistic expectations of what the Initiative could achieve. While every stakeholder group accepts that debt relief is only part of the solution in a broader attack on the obstacles to sustained poverty reduction, it is striking how critical many commentators are with respect to the actual or anticipated achievements of the Initiative. In effect, HIPC has become a lightning rod for broader policy disagreements regarding equitable and sustainable development and the role of aid. Progress and Current Status 1.4 Under the original framework, from a group of 41 countries a total of 29 were expected to qualify for relief. Of these, seven reached their decision points and six reached their completion points. The number of countries expected to qualify under the E-HIPC 2 expanded to 36 in 1999, at the time E-HIPC was launched (from a group of 41 countries, although eligibility was open to all countries meeting the E-HIPC eligibility criteria). 1.5 By August 2002, the number of countries eligible for debt relief increased to 42; of these, 6 have reached their completion points and are receiving (or should be receiving) the full benefit from HIPC relief (see box 1.1). Another 20 countries have reached the decision point and are benefiting from interim relief under the E-HIPC. Of the remaining 16, 4 are considered potentially sustainable after receiving relief from traditional mechanisms and 8 are conflict-affected countries facing particularly difficult challenges in coming to their decision points under the current framework. Box 1.1: Status of the 42 Eligible IHlPCs (as of August 2002) Reached completion point: 6 (Bolivia, Burkina Faso, Mauritania, Mozambique, Tanzania, Uganda). Reached decision point and receiving interim relief: 20 (Benin, Cameroon, Chad, Ethiopia, The Gambia, Ghana, Guinea, Guinea-Bissau*, Guyana, Honduras, Madagascar, Malawi, Mali, Nicaragua, Niger, Rwanda*, Sa6 Tom6 and Principe, Senegal, Sierra Leone*, Zambia). Not yet at decision point: 12 (Burundi*, Central African Republic*, Comoros, Democratic Republic of Congo*, Republic of Congo*, COte d'Ivoire, Lao P.D.R., Liberia*, Myanmar*, Somalia*, Sudan*, and Togo). Potentially sustainable without HIPC assistance: 4 (Angola*, Kenya, Vietnam, Yemen). * Conflict-affected countries. 1.6 The total amount of debt relief committed (that is, to countries past their completion point and the potential estimated relief to countries that are past their decision points, including those that qualified under the O-HIPC) is $41.52 billion in nominal debt service relief over time, equivalent to $25.1 billion in net present value (NPV) terms (Annex A). Of this amount, the cost of debt relief to the six countries that reached their completion points under the O-HIPC is $6.97 billion in nominal terms, or $3.46 billion in NPV terms.' According to current estimates, the total cost of the Initiative is expected to be about $37.2 billion in (2001 NPV terms), with the World Bank accounting for some $8.2 billion. This estimate excludes the potentially sustainable cases (Angola, Kenya, Vietnam, and Yemen) and Liberia, Somalia, Sudan, and Lao P.D.R.' it also excludes potential topping-up costs. 1.7 For the 34 countries (that is, excluding the 4 sustainable cases, and Liberia, Somalia, Sudan, and Lao P.D.R.), the share of the costs is roughly evenly divided between bilateral and multilateral creditors. The World Bank Group's share is 22.3 percent (the share of the International Development Association-IDA-is 20.3 percent of total). The share of the Paris Club creditors is highest at 38.6 percent. Non-Paris club creditors account for 8.8 percent of the costs and commercial creditors for 4.2 percent. 2. Of the seven that reached decision point under O-HIPC, C6te d'Ivoire has yet to reach its completion point. 3. Including Liberia, Somalia, and Sudan the costs in 2001 NPV terms are estimated to be $46.0 billion. 3 Evaluation Design 1.8 In the light of the increased interest in financing for development, as well as the impact of the recent global economic slowdown on the prospects for the sustainability of external debt of poor countries, this is an appropriate juncture to review the progress and prospects of the HIPC Initiative with a view of informing and, if necessary, strengthening its ongoing implementation. With 6 countries past their completion point, and another 20 past their decision point, increasing concerns are being voiced about, on the one hand, the pace at which the Initiative is progressing, and, on the other, the likely outcomes of the Initiative. 1.9 The guiding principles and the building blocks of the original and the enhanced H4IPC frameworks provide the basis for assessing whether the process being followed and the implementation of the Initiative are consistent with its objectives and design. The stated objectives and guiding principles endorsed by the Interim and Development Committees are given in Annex B and the flow diagram describing the HIPC process is given in Annex C. The details of the evaluation approach adopted for this review are given in Annex D. 1.10 Main Evaluative Questions: This review focuses on the design and implementation of the HLPC Initiative. The purpose is to assess the prospect of the Initiative achieving its intended immediate objectives-reducing debt to sustainable levels and creating the fiscal space for increased spending for poverty reduction-and the potential for contributing toward its underlying development goals-sustainable economic growth and poverty reduction. Using OED's objectives-based evaluation framework, this review seeks to answer the following overarching questions corresponding to the framework's principal assessment criteria: * Relevance: Is the HIPC Initiative's design adequate and appropriate to achieve its stated objectives and intended outcomes? * Efficacy and Efficiency: Based on the experience so far, is the HIPC Initiative achieving or likely to achieve its objectives, and to achieve them efficiently? * Sustainability: How resilient to risks are the expected outcomes of the HIPC Initiative? * Institutional Development: To what extent does the design of the Initiative help build country capacity to ensure that the HIPC objectives are achieved and can be sustained? 1.11 Coordination with the IMF: In preparing this review, OED has interacted not only with World Bank staff and management but also the IMF's Independent Evaluation Office (IEO) and the staffs of the IMF's Policy Development and Review Department and Fiscal Affairs Department. The key background papers were shared with IMF staff for critical comment and they participated in seminars to discuss the findings of the background papers, along with the staff of the World Bank's HIPC unit. IMF staff also participated in the technical workshop held in Washington, D.C. in September 2002. 4 2. Evolution of Deb &Ind 9he Debt piroblbRem 2.1 This chapter briefly discusses the evolution and magnitude of the debt problem and past responses to it. It then presents the main findings from a review of the literature on the likely consequences of a large debt stock. Evolution of the HPC Debt Burden 2.2 The 42 HIPCs accounted for about 14 percent of the developing world's population in 2000 but only about 5 percent of the total gross national income (GNI).4 Their share of total external debt of all developing countries, approximately 8 percent, is small relative to their share of population but large in relation to the size of their economies (table 2.1). 2.3 The large stock of Table 2J: External Debt as Percentage of GDPI debt in the HIPCs has a long (peided aveir%p) history, but it did not start out as large. Following a c7eoyV994 1985-19e9 1990 994 199&-2000 decade of good growth in HIPC 38 70 120 103 the 1960s (following Other IDA countries 21 33 38 33 independence for a large Other LMI* countries 22 30 27 26 number of the African * Lower-middle-income. states), the economic shocks Source: Global Development Finance and World Development Indicators. of the early 1970s, combined with serious economic, social, and structural constraints to Figure 2Jl: lEvollution of the EllIPC ID)ebt Balr(dlen raianbod-segowh rapid and broad-based growth, resulted in a long and persistent 1200 - ----- ----- economic decline-lasting until the Total stock (norrinal) early 1990s. Oil shocks, declining looo terms of trade, and highly volatile 800 commodity markets led to a spiral of i 00 - ~soc (nal fiscal imbalance brought about by 600 oPPG stocks (nomnAl) declining revenues, resistance to painful fiscal adjustments, and heavy o Present value reliance on borrowing to meet the 200 PPG stocks deficit. Large inflows and mounting arrears led to the rapid accumulation 0 of debt (Daseking and Powell, 1999). P F 2 .0 . A 8 Figure 2.1 shows how the nominal PPG = Public and publicly guaranteed external debt stock of debt of the 42 HIPCs has exploded relative to exports since Source: Global Development Finance, World Development about 1980. The debt burden, Indicators, and NPV calculations for Easterly 2001 a. measured in terms of the NPV of 4. The developing world is defined as all low, lower-middle- and upper-middle-income countries as defined by the World Bank, a total of 156 countries (see World Bank 2002). 5 external debt-to-exports, also increased sharply until the mid-1980s.' 2.4 The debt of low-income countries (LICs) escalated even as major Initiatives were under-way to solve the middle-income country (MIC) debt crisis of the 1980s. But the two groups have important differences (Cline 1997). For LICs, and especially Figure 2.2: Net Resource Transfers to HIPCs for the HIPCs, bilateral and multilateral Have Outpaced Debt Service creditors have been the primary sources for the past two decades. The MIC debt 16 Aggregate net transters of the 1980s arose largely from private 14- commercial sources. And while the 12. MICs experienced large negative net transfers during the crisis period, the i Official net transfer LICs have consistently had positive and 8- substantial net resource transfers (figure 6- 2.2).6 The official creditors have had a 4 long-standing commitment to ensuring Off icial debt service positive net transfers to the HIPCs, despite the latter's generally low o 1 .. . . . productivity growth, high debt-service g g M 0 0 M D 0 9 8 levels, and, in some cases, virtual Source: Global Development Finance. insolvency (World Bank 1994). 2.5 The prolonged economic decline, with annual per capita growth rates averaging -2.2 for Sub-Saharan Africa (SSA) during 1980-89, and deteriorating terms of trade led to increasing debt service problems and mounting arrears. Many LICs were unable to fully service their debts; in 1993, those then classified as severely indebted low-income countries (SELICs) paid only 40 percent of the scheduled debt service (World Bank 1994)." Under the circumstances, additional flows of credit would not have been justifiable under commercial financing practices-as evidenced by the early exit of commercial creditors (other than state- sponsored export credits). The official creditors, however, were not operating under commercial principles. Often driven by political concerns and domestic commercial considerations, they eventually committed to maintaining positive net transfers. As figure 2.2 shows, official net transfers have been twice the level of total debt service, and gross inflows have been three times (or more) the total outflows. 5. The nominal stock of debt is not a good indicator of the actual burden of debt on a country when a proportion of the debt is contracted on concessional (or below market) rates. Data on present value are from William Easterly's dataset (discussed in Easterly 2001 a), which uses GDF debt-inventories to calculate future obligations on publicly guaranteed debt, discounted using the LIBOR as the market rate. These data are supplemented with the present value estimates from GDF publications after 1998. 6. There are some exceptions to this at the individual country level, but the occurrences are few and not sustained. 7. A total of 36 countries were classified as SILICs, all of whom with the exception of Nigeria, are currently classified as HIPCs. 6 2.6 The positive net transfers have been maintained using various mechanisms: initially through rescheduled debt payments, then by substituting concessional loans and grants for nonconcessional loans. The figure in Annex E shows how the sources of net transfers to HIPCs have changed since 1980. Nonconcessional lending from the International Bank for Reconstruction and Development (IBRD) and other multilateral sources was replaced by concessional lending from the International Development Association (IDA) and other multilateral institutions. Thus, although net transfers from the IBRD turned negative in the late 1980s, overall transfers from the World Bank have remained positive. Similarly, starting in the mid-1980s, most bilateral donors reduced nonconcessional lending, progressively curtailed concessional lending, and sharply increased grants. 2.7 With a persistent debt crisis and the increasing acceptance that it reflected the insolvency of most HIPCs, actions to relieve the debt burdens of poor countries have progressively intensified.' Major bilateral lenders switched from nonconcessional rescheduling to concessional rescheduling, and finally to debt stock reduction actions through the Paris Club.' Multilaterals were more limited in their options, reflecting a combination of the relatively small proportion of the overall debt burden (in present value) owed to them, concern about maintaining their financial integrity, and their continued treatment of the HIPC debt problem as one of liquidity.o Nevertheless, the increasing recognition of the strains imposed by debt service and the emerging multilateral debt service problem led to the introduction of the Special Program of Assistance for Africa in 1987, debt buybacks through the IDA debt reduction facility, fifth dimension grants to help former IBRD borrowers to meet their interest payments, and the establishment of the Enhanced Structural Adjustment Facility (ESAF), a concessional arm for IMF assistance to LICs. 2.8 These actions, especially those to relieve debt (starting in 1988), did have an impact in checking the rise of the debt burden, as shown by the flattening of the NPV of debt-to- exports ratio in figure 2.1. Additional Paris Club actions with successively increasing relief elements and, eventually, the pick-up in growth account for the decline in the ratios of both the nominal and NPV of debt stock to exports in the mid to late 1990s. An important consequence of the multilateral actions, particularly the expansion of concessional finance, was the shift in the composition of debt outstanding and the corresponding debt service obligations. The share of official sources, particularly multilateral concessional debt, increased significantly. 2.9 The rising share of multilateral debt service posed potentially severe problems because these institutions did not offer an option of rescheduling or debt relief, and dire 8. The rapid increase in arrears and repeated rounds of Paris Club reschedulings are a strong indicator that the previous strategy was not effectively addressing the HIPCs' problems. 9. Early efforts to address the problem through nonconcessional rescheduling, and even the subsequent rescheduling on concessional terms, simply postponed reckoning with the problem of excessive debt stocks. Instead, with continued weak economic performance, these reschedulings exacerbated the debt stock buildup. 10. At end-1993, the share of multilateral creditors in total nominal debt stock was 27 percent, with the Bank accounting for half of this (both IDA and IBRD). In terms of present value, the multilateral share was 19 percent (with the Bank accounting for 37 percent of the multilateral share). Bilateral creditors' shares were 59 percent in nominal stock and 64 percent in prevent value terms. 7 consequences were associated with falling into arrears on multilateral debt. At the same time, the increased frequency of debt restructurings with the Paris Club made it apparent that the existing mechanisms were inadequate to deal with the problem. Eventually, a comprehensive solution was sought to address the insolvency of HIPCs, which actions merely designed to provide temporary relief or to check a rise in debt burdens could not resolve. The Consequences of an Excessive External Debt Burden 2.10 It is generally accepted that a large debt stock can impair development, but there is less agreement on how this impact might occur, particularly in the case of HIPCs. Annex F summarizes the main findings from a review of the research and academic literatures on this topic. The dominant theory underlying the adverse consequences of an excessive debt stock is the "debt overhang" hypothesis." This was the straightforward strategic rationale for the HIPC Initiative as originally conceived and operationalized (World Bank and IMF 1998, 1999). But as discussed in Annex F, this has not been demonstrated convincingly for the HIPCs. Low domestic investment, capital flight, and limited flows of foreign direct investment reflect a combination of factors affecting the investment climate, including the policy environment and structural and social constraints. In the continued presence of these factors, a reduction in debt stock by itself is unlikely to lead to significant increases in private investment. 2.11 A more plausible argument is that high debt service payments crowd out high-priority public expenditures. Despite the large positive net transfers, the fiscal space may be too small to simultaneously accommodate large debt service obligations and to fund the necessary infrastructure and social investments for broad-based and equitable growth. This partly reflects the inefficiency of existing aid processes (such as project finance and tied procurement), but it also reflects insufficient efforts to increase budgetary revenues, inefficient management of public expenditures, or both. To effectively reduce the fiscal strain would thus require not only debt reduction (to lower the debt service obligations), but also concurrent actions on the policy front by donors and recipients alike. 2.12 Furthermore, the positive net transfers have been maintained through a complex and inefficient restructuring and negotiation process (see Annex F). The uncertainty surrounding the process and the general inefficiency associated with high debt stocks can have a negative influence on both the level of investment and the effective use of existing capacity. There is some evidence suggesting this might be a more important channel through which an excessive debt stock can affect economic performance in the HIPCs. 11. The theory postulates that beyond an optimal point, the debt stock has a negative impact on growth by discouraging investment. 8 3. 'The MPC Hnitive 3.1 This chapter provides a brief review of the rise of the HIPC Initiative. After a brief background summary, it describes how the political pressures underlying the changing international financial governance and emerging alliances of civil society groups brought the Initiative into being. The chapter then presents the key objectives and characteristics of the HIPC process. Background 3.2 In the mid-1990s, increasing poverty and a general perception of development failure in many of the least developed countries, primarily in SSA, coincided to intensify the pressure for a strategy to deal with the low-income country debt crisis. The complex relations between heretofore relatively distinct groups of actors started to come together and influence official thinking on the issue (Callaghy 2002)." The growing influence of civil society, with NGO networks at its forceful core, transformed the international debt regime through political reach into the higher echelons of international governance, especially with the G-7 governments and the Development and Interim Committees of the Bretton Woods Institutions. The alliance's efforts led to a call by the governing body of the World Bank at its annual meetings in 1994 and 1995 to study multilateral debt and develop an effective strategy to deal with the issue. 3.3 Within the World Bank, concern was already growing about the rising debt problem for a large number of its poorest borrowers, particularly in the Africa Region. At first the voices were disjointed; they were also discounted under the presumption that the poor countries had a short-term cash-flow constraint that was readily amenable to policy reform. The major tool in this strategy was structural adjustment programs, which provided the resources to get through the short-run problems and targeted much needed policy reforms. By the early 1990s, however, it had become clear that structural adjustment programs were not working as expected. Lack of ownership of reform programs combined with governance dysfunctions, weak public expenditures management, and inadequate emphasis on infrastructure, private sector development, and agricultural productivity had hindered supply responses to macroeconomic policy adjustment-and they still do. 3.4 To address increasing concerns about LIC debt, especially the rapidly increasing multilateral debt, the World Bank established a working group to assess the magnitude of the multilateral debt problem and develop possible mechanisms to deal with it. The group sought to complement the World Bank's existing instruments (noted in chapter 2) and existing ("traditional") mechanisms for providing relief for bilateral and commercial debt. The mechanism proposed by the group, a "multilateral debt fund," was deliberately designed to be a concerted and comprehensive effort to effectively deal with the HIPC debt problem. 12. There has been a "triple helix" of relationships, as Callaghy calls them, between three sets of actors that were central to the debt relief movement from the late 1980s: the official agencies and processes of the international debt regime, the NGO networks, and an epistemic community of economists and other scholars who have played key advisory roles on both sides of the debate on how to deal with the debt problem. 9 3.5 When a draft working paper prepared by the group was leaked to the press in 1995, it had an unexpected catalytic effect. The development community quickly embraced the idea, which soon translated into the proposal for the HIPC Initiative put forth by the World Bank and the IMF in 1996. Through more uniform rules, the HIPC Initiative marked a significant advance from traditional debt relief mechanisms for eligible low-income countries. It transformed the debt regime toward more open and accountable norms, and introduced some key innovations including, for the first time, a systematic treatment of multilateral debt, the notion of debt sustainability, and the focus on poverty reduction. The evolution of the "multilateral debt fund" to the final form of the HIPC Initiative and the various pressures that have influenced its emergence and design are discussed in detail in a background paper to this review (Callaghy 2002). The Political Economy of HIPC 3.6 The emergence of HIPC is a response to the central structural dilemma of recent times-the emergence of a group of weak states and economies that have not been able to benefit easily or quickly from economic reform and globalization. This dilemma poses important difficulties for the functioning and evolution of the international political economy, and ultimately for international peace. The causes of this structural dilemma are many, complex, and very deeply rooted--developed country policies, external trade patterns and other external shocks, heavy reliance on primary commodities, weak formal economies, flagging economic reform efforts, poor investment climates, corrupt and oppressive governments, weak domestic capacities, civil conflict and war, environmental degradation, and disintegrating physical and social infrastructure. All of this is reinforced by limited access to private international capital flows, despite the implicit bargain with the international community that such access would sustain economic reform efforts. 3.7 The major evolution of the treatment of sovereign debt was the move from debt collection, to debt rescheduling, to aid and structural adjustment, to debt "sustainability," to forgiveness and poverty reduction-what one G-7 official called the "slippery slope of debt." The slippery slope was expertly greased by the strategy adopted, and very effectively implemented, by major advocacy NGOs in ratcheting up the debate and ultimately winning ground on debt relief." The resulting momentum brought about the original HIPC Initiative and the enhanced HIPC. 3.8 Stakeholders agree that debt relief is essentially a political issue. It has significant economic consequences, but to deal with the debt problem requires political commitment and resolve. This is evident in the evolution of the debt regime. Not surprisingly, these changes were brought about by a confluence of factors: 13. The Jubilee 2000 campaign was highly successful in creating a mass movement in support of debt forgiveness by mobilizing the support of prominent public figures, religious leaders, academics, and entertainers. The broad political resonance of the simple and forceful "debt relief for poverty reduction" message helped to mobilize a coalition that combined the forces of mainstream development advocates with those of radical critics of the aid process, thus overcoming opposition to debt forgiveness at a time of decline in overall development aid. 10 o Slow and uneven learning by bilateral and multilateral creditors about the existence of a group of states that were not benefiting much from structural adjustment, while greatly increasing their debt loads in the process o The growing pressure, influence, and effectiveness of a new set of actors in international economic governance-networks of NGOs that believed the existing situation for these states was unjust and untenable, and that had new ideas and proposals of their own, with a social movement to back them up o The influence of a group of economists, both inside and outside creditor institutions, including in the World Bank, who provided knowledge, advice, and technical understanding on this issue o The leadership of a group of small creditor states and eventually several G-7 members o New leadership at the World Bank o The successful efforts by negotiators of key creditor countries and donor institutions, both bilateral and multilateral, to work through the complex and often politically contentious details. Many of the ideas inherent in HIPC were proposed by southern states during the New International Economic Order (NIEO) events of the late 1970s and early 1980s, but nothing came of this intense and polarized state-to-state bargaining. One of the striking things about the rise of HIPC is precisely that the debtor states were not a major driving force behind the innovation. Rather it was made possible by NGOs that shifted the battle from the corridors of power into the domestic political arenas of the Organization for Economic Cooperation and Development (OECD) industrial democracies. The weak power position of the debtor states and the concomitant strong influence of the NGOs helps to explain how the HIPC Initiative eventually became focused almost exclusively on a particular approach to poverty reduction.' One of the main complaints by almost all debtors and many of the creditors is that this is very often at the expense of broader developmental concerns (see Annexes G and H for summaries of consultations with the two groups). The HIPC Objectives 3.9 The main features of the HIPC Initiative, for both the original and the enhanced frameworks, are given in table 3.1. A full statement of the objectives and the guiding principles of the Initiative are given in Annex B. 3.10 The original framework had a straightforward focus on the key issue that the Initiative was created to address-to reduce the debt stock. " The strategic rationale for this 14. At the time it was created, the principal objective of the HIPC Initiative was to reduce the external debt of eligible countries to a level that would be sustainable, thereby removing a major constraint on investment and growth and spurring further adjustment, in part by galvanizing private external investment. The victory by the NGOs-in establishing a direct link between debt relief and poverty reduction, by explicitly targeting debt service "savings" to social spending-was the outcome of a very successful campaign by the Jubilee 2000 movement. 15. HIPC addresses only the long-term public and publicly guaranteed disbursed and outstanding debt. 11 was that it would remove the disincentive effects on private investment of a "debt overhang" and allow progress toward the underlying development goal of economic growth and poverty reduction. The Initiative was viewed as one element of an overall strategy to achieve debt sustainability for the HIPCs. At the same time, the "savings" from reduced debt service, to the extent that they were actually realized, would generate the much-needed fiscal space in the HIPC governments' budgets to pursue economic growth and poverty reduction. Table 3.1: The Original and Enhanced HIPC Frameworks in a Nutshell Element Original Enhanced Stated objectives To bring the country's debt down to Maintains the original focus to remove the sustainable levels, subject to satisfactory debt overhang and provide a permanent policy performance exit from rescheduling, plus free-up resources for higher social spending aimed at poverty reduction to the extent that cash debt-service payments are reduced. Qualification criteria IDA-only countries (Poverty) Same. Applied retroactively to include Unsustainable level of debt after full use of countries already past decision or traditional mechanisms completion points under the original Strong record of policy performance framework. 41 countries eligible, 29 expected to qualify 41 eligible in 1999, currently 42, of which 38 are expected to qualify. Debt sustainability Guiding principle, Target overall debt Principle for change- Provide a clear exit sustainability to provide a durable exit from unsustainable debt burden to remove strategy from the rescheduling process the debt overhang and provide an appropriate cushion against exogenous shocks Indicators: Targets Target range for main indicator: Uniform application of single target: NPV debt-to-exports: 200-250% NPV debt-to-exports: 150% NPV debt-to-revenue: 280% with NPV debt-to-revenue: 250% with Export/GDP: 40%; Revenue/GDP: 20% Export/GDP: 30%; Revenue/GDP: 15% Calculation of relief Fixed at completion point, based on Fixed at decision point, using actual data on projections of debt indicator for completion NPV debt for year prior to decision point point and 3-year average for exports Time of relief delivery Completion point (CP), irrevocable Decision point: on an annual basis, interim commitment relief is bulk of anticipated post-CP relief, it is irrevocable Forward-looking Debt sustainability analysis to project profile Same assessments of key debt indicators Performance criteria Guiding principle, Action only after the debtor Principle for change: To strengthen the has shown, through a track record, the ability incentives for debtor countries to adopt to put to good use whatever relief is provided strong programs of adjustment and reform For decision point 3-year track record of macroeconomic Same plus stability and policy reform Interim or full Poverty Reduction Strategy Paper (PRSP) For completion point Further 3-year track record of macroeconomic Maintenance of macroeconomic stability stability and policy reform Completion of PRSP, plus one year PRSP implementation for E-HIPC Performance benchmarks for structural and social reforms Interim period 3 years Flexible, with the introduction of floating completion point Creditor participation Guiding principle: Comprehensive debt relief Principle for change: Same action: coordinated among all creditors Plus debt relief should be additional to involved with broad and equitable reinforce the wider tools of the international participation community to promote sustainable New external finance to be on appropriately development and poverty reduction I concessional terms 12 3.11 Over time, the objectives evolved, in subtle ways at first and then significantly with the launch of the E-HIPC in 1999. While the original goal of promoting growth by removing the debt overhang has been retained, the transformation, evident in the stated objectives and the guiding principles (see Annex B), took place in two dimensions: (a) a shift in focus of the original objective related to debt sustainability and (b) the addition of an explicit twin objective. 3.12 The objective related to debt sustainability became more ambitious: from reducing debt as part of a broader strategy to achieve long-run sustainability in 1995; to reducing debt to sustainable levels and thus providing a durable exit strategy from the rescheduling process in the original formulation of the Initiative in 1996; to providing a "robust" exit from debt reschedulings and the achievement of debt sustainability in 1998 (World Bank and IMF 1998); to a "permanent" exit from the rescheduling process and a "clear" exit from unsustainable debt in 1999 (World Bank and IMF 1999). 3.13 The objectives were expanded to specifically target the freed resources to social spending, ostensibly suggesting that debt relief would generate additional resource flows. The addition of this objective had a significant impact on the implementation of the Initiative. To strengthen the link between HIPC relief and poverty reduction, the performance criteria were broadened to include the requirement of a Poverty Reduction Strategy Paper (PRSP)." And, while the performance criteria to reach the completion point in the O-HIPC framework included both structural and social measures, it was decided in E-HIPC to "give more weight than under the [original] framework to social and poverty-related reforms" in the assessment of performance to reach completion point (World Bank and IMF 1999). The Process 3.14 The current HIPC process is described in detail in the flow diagram in Annex C. The primary qualification criteria and the core building blocks of the process, along with the modifications to the specific elements of the design from the O-HIPC to the E-HIPC framework, are given in table 3.1. Among the qualification criteria, the focus on poverty is straightforward: only countries that are eligible for concessional loans from IDA qualify for HIPC relief, excluding the IDA-IBRD blend countries. The second and third criteria are linked to two main building blocks that define debt sustainability and policy performance. The focus of this review is the three building blocks of the process: debt sustainability, performance criteria, and creditor participation. 3.15 The total amount of assistance provided by the HIPC Initiative, in the form of debt service relief over the next 30 years or so, is divided roughly evenly between the multilateral and bilateral creditors. The processes involved in accessing debt relief at the country level are more protracted for bilateral than for multilateral creditors. Although the amount of debt 16. The PRSP represents a long-term program for poverty reduction, prepared by the country authorities in consultation with all stakeholders. PRSPs mark a new kind of interaction of the Bank and the international development community with borrowers. They put into practice the principles of the new Comprehensive Development Framework adopted by the Bank, promoting country ownership and broader stakeholder participation, improve coordination among development partners, and increase the focus on results. 13 relief from the Paris Club is determined in the context of the HIPC Initiative and is not subject to subsequent negotiation, the HIPCs are obliged to meet with the Paris Club after the decision point to agree on the broad terms of the debt relief to be delivered by the participating creditors. After this agreement, the HIPCs must reach a formal agreement with individual Paris Club creditors on the modalities of debt relief. They also have to secure comparable treatment from non-Paris Club bilateral creditors and commercial creditors. The process of accessing debt relief from the major multilateral creditors is simpler, with debt relief generally taking the form of reduced debt service payments, cancellation of credits, or the provision of grants. The World Bank's HIPC unit estimates the amount of relief due from each multilateral creditor, although the delivery modalities still vary across creditors. 3.16 The financing of the debt relief is straightforward for the bilateral creditors. It is complicated for the multilateral creditors given their status as preferred creditors. To finance the multilateral share of the relief, the World Bank established the HIPC trust fund in 1996. The trust fund is structured to allow multilateral creditors to participate in the Initiative in ways that are consistent with their financial policies, and it also helps address resource constraints for certain multilateral creditors. The two main sources of financing for the trust fund have been bilateral contributions and transfers from IBRD net income and surplus; contributions from other multilaterals are also permitted. The trust fund either prepays, or purchases a portion of the debt owed to a multilateral creditor and cancels the debt, or pays the debt service as it comes due. 14 4. Design of the Ilnitiative 4.1 This chapter assesses whether the design of the Initiative is appropriate to achieve its goals. It first assesses the consistency of the Initiative's overall design with its objectives, and then reviews the three key elements of the design as discussed in table 3.1: debt sustainability, performance criteria, and creditor participation. Then it briefly touches on the lack of capacity building in the Initiative's design. The discussion draws on a background paper to this review (Eaton 2002), a review of HIPC documents, and data analysis. Background 4.2 The design embodies the emerging empirical evidence and lessons of experience on the link between aid effectiveness and policy environment, conditionality and ownership, social impacts of macroeconomic policy and public expenditure reforms, and aid coordination. While the original design was essentially developed by the staffs of the World Bank and the IMF, the broad participatory process adopted to review and then enhance the HIPC framework in 1999 was critical to the evolution of the Initiative, given the key role played by the international NGO community. The debtors had no explicit role in the O-HIPC design, and limited influence on the E-HIPC design. This is noteworthy, since the HIPC process envisages the debtor government and the civil society in poor countries firmly taking the driver's seat and owning the process. Their limited input into the design of the process may affect the Initiative's outcomes. 4.3 Nevertheless, adaptive "learning" among participants has been a strong feature of the evolution of the HIPC Initiative's design and the processes for implementation. As the first country to qualify under both the O-HIPC and E-HIPC frameworks, Uganda has been a valuable learning ground (see box 4.1). The interactive process with stakeholders, in particular the NGOs, and their suggestions and inputs have helped to formulate new ideas- from process issues, to the introduction of the participatory poverty reduction strategy concept, to technical issues such as retrofitting the sustainability thresholds. Consistency of the Design with OveraH Objectives 4.4 The HIPC Initiative has acquired three distinct objectives: (1) Debt sustainability by "[providing] a permanent exit from rescheduling"' (2) Raising long-term growth rates by "removing the debt overhang" (3) Poverty reduction by "[freeing] up resources for higher social spending.. .to the extent that cash debt-service payments are reduced."" 17. That is, the Initiative is intended to reduce debt service obligations so that the country can more easily meet with its export revenues and future transfers, eliminating the need for future reschedulings, defensive lending, and debt forgiveness. 18. A press release was more emphatic, stating, "The debt reduction operation.. .will create room for additional public expenditures on poverty reduction." 15 Box 4.1: Uganda's Influence on the Design of the HIPC Initiative As the first country to qualify for HIPC debt relief, Uganda has served as a valuable learning ground to refine the parameters for determining debt sustainability. The main debt sustainability indicators used in the Initiative are the NPV of debt-to-exports ratio and the debt service-to-exports ratio. Initial HIPC proposals used a single year's export earnings to calculate the key ratios and, consequently, to determine debt relief and debt sustainability. Ugandan authorities opposed the use of a single year's exports, arguing that it would not adequately capture the extreme volatility of Uganda's export earnings. The authorities were concerned that a decision point that coincided with unusually high international prices for coffee or other commodities, and therefore higher export receipts, could potentially lower the amount of debt relief they would receive. Underlying their concern was the rapid tapering off of the 1994/95 and 1995/96 boom in the international price of coffee. They favored a six-year export average but that was not acceptable to the World Bank and the IMF. According to the Ugandan authorities, a six-year backward-looking average (in 1995/96) would have almost doubled the amount of debt relief Uganda could receive. The World Bank and IMF eventually settled on a three-year export average as a basis for the debt-to-export calculation for all countries. -Uganda was quick to commit itself to linking debt relief to poverty reduction under O-HIPC, and it was first among low-income countries to successfully formulate a national poverty reduction strategy, long before the advent of the Poverty Reduction Strategy Paper. These successes have strongly influenced the design of E-HIPC by the World Bank and IMF. Uganda's poverty reduction strategy- known as Poverty Eradication Action Plan (PEAP)-was formulated in 1997 following an 18-month long consultation involving stakeholders that included central and local governments, NGOs, civil society, donors, and academia. The consultative process was subsequently broadened to include a diagnostic of poverty through direct consultations with the poor, carried out jointly by the government and civil society through a Participatory Poverty Assessment Project. In 1998/99, the government established the Poverty Action Fund-a virtual fund within the budget to finance expenditures identified as priority poverty-reducing areas in the PEAP-to "ring-fence" debt relief funds, including donor budget support, in support of PEAP objectives. Uganda's PEAP become the model for the Poverty Reduction Strategy Papers. * Based on information obtained from interviews with Ugandan authorities and Austrian Development Cooperation (1999). 4.5 Although each of these objectives is clear, the design of the Initiative does not ensure that all three goals can be met simultaneously. In specific contexts and given limited grant resources, the goals may instead conflict with one another. 4.6 The central issue is whether the debt relief provided by the Initiative results in an increase in net resource transfers to the HIPCs-"additionality." The impact of debt forgiveness on net resource transfers to countries must be viewed in the context of overall aid flows. For net resource transfers to these countries to increase as a result of debt relief, donors need to maintain the levels of other aid flows. " Equally important, the impact of debt 19. Net resource transfer is defined as disbursements (including loans and grants) less debt service payments (principal repayments and interest payments). Debt relief reduces the debt service obligations of a debtor, but for this to lead to an increase in net resource transfers, disbursements need to be maintained at the previous levels (for full additionality) or decline less than the reduction in debt service (for partial additionality). If both the inflows (disbursements) and outflows (debt service payments) decline by equal amounts, the net transfer into the debtor country would remain unchanged. 16 relief on the availability of other aid should be assessed both at the level of the individual HIPC and at the global level for aid recipient countries in the aggregate. 4.7 For the individual countries eligible for HIPC relief, an increase in net resource transfers is necessary to achieve the third objective-to free-up resources for higher social spending for poverty reduction. The design of the Initiative cannot ensure that this increase will in fact take place. Evidence suggests that past debt relief efforts were not additional (Birdsall and Williamson 2002). As will be seen in Chapter 6, while it is too early to assess the additionality of HIPC debt relief, the evidence so far on the net resource transfers to the HIPCs shows a declining trend. 4.8 To achieve the Initiative's first objective-a "permanent exit from rescheduling"- the increased net transfers would also need to be sustainable. That is, their terms need to be sufficiently concessional to obviate the need for future rescheduling. If additionality can be achieved only by transfers in the form of loans, the objective of a "permanent exit from rescheduling" will be more difficult to achieve. With limited grant resources, there is thus a tension between the first and third objectives: the more the Initiative achieves toward the first objective, the less it may do to accomplish the third. 4.9 The means of achieving the second objective, growth, may also create a tension with the first objective. If the intent is to eliminate the crowding-out effects of debt service, additional resources will be needed to allow growth-promoting fiscal expenditures or public investment. The "debt overhang" argument, that debt scares away private investment, remains speculative given other factors affecting the investment climate. To the extent it does induce private investment, however, debt reduction may lead to higher growth even if there is no additionality." But for this argument to hold, private investors may be attracted by creating enclaves to protect their investments, while foreign private lenders must have the confidence that, at some time in the future, they will be able to take their returns out of the country-that is, the country will start making net resource transfers to creditors. This has not happened historically. 4.10 If the primary goal of the Initiative is to facilitate an exit from debt rescheduling while not affecting overall aid or redirecting aid resources away from other poor countries and toward the HIPCs, then the lack of additionality is not a problem. If there is no increase in transfers as a result of the HIPC Initiative, then what are the gains from reducing the debt stock? The gains are what Birdsall and Williamson call "efficiency" gains. One is the elimination of complex debt negotiations, which require the extensive involvement of senior government officials, and a reduction in the uncertainty about the outcome of the negotiations, which generates economic instability. A second gain would be to allow creditors to be more selective in their lending. The Initiative involves more coordination among creditors, and this could lead to a rationalization in the use of funds for eligible countries. The third would be to foster greater ownership of the country's development program by allowing governments to pursue their own priorities by putting resources more directly in their hands. This would result, for example, by replacing project lending with general budget support and reducing economic 20. Investment incentives may improve, for example, through indirect positive economic effects from a more stable economic climate or eliminating the expectation of future crises. 17 costs associated with tied aid." Finally, a transfer-neutral reduction in aid would lessen the transactions cost associated with external financing. All of these effects have the potential to enhance growth and tax revenues.22 4.11 It should also be mentioned that if the Initiative does not entail any additionality, then the total "cost" of the program would be less than the direct costs estimated by HIPC documents. The Initiative would not incur additional costs if the overall level of financial assistance were maintained by giving up debt-service payments, but also simultaneously reducing the provision of transfers, including those for financing the debtors' debt service obligations. To the extent this occurs (as argued by Birdsall, Claessens, and Diwan 2001 and Sachs and others 1999) there are efficiency gains from ending this "shell game" or "forced lending," but it is cost-neutral to the international donor community. Nevertheless, debt forgiveness might end up reshuffling obligations among individual donor/creditors, making the discussion of appropriate "burden sharing" important. 4.12 At the global level, whether or not HIPC debt relief is additional to overall aid flows has important implications for the distributional impact of the Initiative between the highly indebted and non-highly indebted poor countries. As the Report of the High-Level Panel on Financing for Development notes, the issue of additionality is a key factor in appraising the desirability of debt relief in overall development finance, as it determines who actually pays for it (United Nations 2000). If the purpose of the Initiative were solely to "end the cycle of periodic reschedulings," without requiring additional resources, as was the main objective in the O-HIPC, then such distributional issues would not arise. Aid allocation would remain independent of the Initiative. But to the extent that the program results in additional transfers to the HIPCs, it expands the commitment of aid resources to these countries. Unless overall aid resources increase, less will be available for the non-HIPC countries. Whether and how this might be affecting the two groups of countries is discussed in Chapter 6. 4.13 To sum up, each of the three stated objectives of the Initiative is highly desirable. But to meet them simultaneously requires actions beyond the scope of the Initiative. In particular, to free up resources for higher social spending, it is necessary (though not sufficient) to transfer additional resources. The Initiative, however, has no means of ensuring this, especially with fixed or declining aid resource envelopes. Nor is it clear how the Initiative could have been designed to provide such surety. The Initiative is a limited instrument. A clear public acknowledgement of this reality and a prioritization of its principal goals and how they are to be achieved in specific contexts would help create the correct expectations of what the Initiative can realistically achieve. 21. An assessment of the net efficiency gains of moving from project to budget support, however, needs to weigh the relative inefficiency of project-finance versus the inefficiency of budget systems. Recent analyses by the Bank and IMF have identified the need for substantial improvements in the budget management systems in most of the HIPCs (see World Bank and IMF 2001a, 2002a). 22. Thus, debt reduction may lead to efficiency benefits that could also be secured through reform in aid delivery mechanisms. However, attempts at such reforms have so far not been successful. 18 Debt SustainabiHty 4.14 Debt sustainability was the core objective in the O-HIPC and remains a central objective in the E-HIPC. It is assessed through the debt sustainability analysis (DSA) conducted jointly by the World Bank, the IMF, and the country authorities. There are two key components of the DSA. First, the debt sustainability criteria are critical in determining which countries become eligible for HIPC relief and the level of relief to be provided in each case. Second, the forward-looking debt projections are the main instrument to assess the likely success of the Initiative for each country in maintaining its future debt burden at a sustainable level. The Debt Sustainabiity Criteria 4.15 The eligibility of countries to qualify for HIPC assistance and the amount of relief are based on actual and historical data. The main criterion to judge eligibility and calculate the amount of debt relief is the ratio of NPV of debt-to-exports. The NPV of debt is estimated using a detailed debt-inventory methodology, which aggregates the debt service payments, scheduled for each outstanding loan, discounted by the prevailing currency-denominated market interest rate. Although cumbersome, this accounting procedure is necessary to arrive at a robust estimate of a country's external debt burden, and its introduction is a very positive feature of the Initiative. For very open economies, a fiscal criterion is used-defined as the ratio of NPV of debt to government revenues (using the same debt-inventory methodology). The indicators have remained the same from O-HPC to E-HIPC, but the threshold levels were modified to deliver deeper relief and to cover more countries, as detailed in table 3.1. Criticisms have been leveled against three elements of the debt sustainability criteria: the choice of the primary indicator, threshold levels at which debt is judged to be unsustainable, and the exclusion of domestic debt from the analysis. 4.16 The choice of primary indicator: The main indicator is the ratio of NPV of public and publicly guaranteed long-term external debt to the average of the past three years of exports of goods and nonfactor services. Several NGOs do not think that this stock-based approach adequately captures the burden of debt service on the government." Nor do they consider exports as relevant to the governments' overall repayment capacity. They suggest using the debt-service-to-revenue ratio (Oxfam 2001). The alternative "sustainable development" approach is also linked to government revenues (proposed by EURODAD 2002).2 But to use 23. In the O-HIPC, the amount of relief was determined at the decision point based on the projected difference between the indicators and the HIPC target ratios at completion point. In the E-HIPC, the DSA projections have no bearing on the calculation of the amount of relief to be delivered. 24. Debt service will not be an accurate indicator of the debt burden facing a country if the debt is not being fully serviced. NPV of debt is more accurate: by removing the grant element embedded in concessional terms, it makes the debt stock comparable across countries and should better reflect the burden of debt service due. The available data show a constant relationship between debt service paid as a ratio of NPV of debt to the level of NPV of debt for non-HIPCs. For the HIPCs the ratio of debt service paid to NPV of debt falls sharply with the level of debt, indicating that at high levels debt is not being fully serviced. 25. The "sustainable development" approach proposes a bottom-up costing of the Millennium Development Goals that makes social and other "essential" expenditures top priority in the government budgets 19 revenues as the basis for judging eligibility, sustainability, and the amount of debt relief presents several challenges. The data on government revenues are notoriously poor. The HIPCs' low level of revenues relative to GDP partly reflects their poor economic condition, but also inadequate revenue collection efforts, as many OED evaluations have found. In many countries, a large part of the revenues is kept off budget (Thomas 2001). But most important, including revenue as a qualifying indicator would reward governments with poor revenue collection efforts, as argued by Birdsall and Williamson (2002), and provide adverse incentives to reform the revenue systems. 4.17 An alternative indicator is the debt-service-to-GDP ratio, which overcomes the problems associated with a revenue-based indicator (suggested by Birdsall and Williamson 2002). This indicator also raises some concerns about incentives for policy formulation and implementation. It is easier to measure and is less directly manipulable than government revenue, but depending on how it is measured in individual countries, it may be more difficult to compare across countries. And depending on the structure of the economy, it may not accurately reflect the country's external debt burden or repayment capacity. Nevertheless, the relative merits of using GDP versus exports need to be studied in more detail.' Different indicators have their advantages and disadvantages. From a design perspective, the less manipulable the indicator the better; thus, for practical and operational reasons, the NPV of debt-to-exports is preferable to other indicators. That said, alternative indicators convey useful information and need to be regularly monitored and reported. This is done in the HIPC progress reports. 4.18 Finally, although the E-HIPC has poverty reduction as an objective, its primary indicators relate to debt, not to poverty criteria such as health or nutrition (Sachs and others 1999). The eligibility criterion that relates to poverty is that countries must be IDA-only borrowers-that is, have a per capita income of $875 or less.' IDA-only borrowers constitute the poorest 76 of the worlds 117 developing nations.' Dagdeviren and Weeks (2001) show that that HIPCs house only 9-12 percent (5-6 percent in PPP terms) of the world's poor and argue that the aggregate impact on poverty would be limited. It is likely that different criteria could have better targeted the Initiative to maximize its impact on global poverty. But this would be the case only if the Initiative provides additional flows. To the extent that there is redistribution toward the HIPCs, the Initiative will be disadvantageous to the non-highly indebted poor (EURODAD). Being budget based, this approach raises the same concerns as the revenue based indictors discussed above (Birdsall and Williamson 2002). 26. The empirical evidence on the choice between the debt-to-exports and debt-to-GDP as a better predictor of a debt crisis is ambiguous: debt-to-exports is a better predictor if no other factors are included; debt-to-GDP is a better predictor if some economic variables (liquidity of the economy and openness) are included (Cohen 2001). An alternative way of assessing the choice of indictor is to determine which indicator, exports or GDP, better explains arrears. Using 1980-2000 annual data from Global Development Finance for the 26 HIPCs that have reached their decision points, results from a fixed effects regression of total arrears on a time trend (year), GDP and exports show that exports have a statistically significant negative impact on arrears while the impact of GDP is not significant. 27. This is the current operational cutoff for IDA, measured in GNI per capita in 2001 U.S. dollars. Bolivia and Honduras currently exceed this cutoff. 28. Developing nations defined as 63 low-income countries and 54 lower-middle-income countries. 20 countries that have regularly serviced their debt and/or have practiced more prudent debt and macroeconomic management, thus raising moral hazard issues. 4.19 In fact, the manner in which poverty reduction is targeted in the HIPC design is to earmark debt relief "savings" for poverty reduction. This has been translated to conditions that require substantial allocations of debt relief resources for social expenditures or include targets for social spending in government budgets. These conditions are a result of the demand by the advocacy NGOs, who have equated poverty reduction with social expenditures (Jubilee 2000 Coalition 1999). 4.20 The threshold levels: The O-HIPC thresholds were criticized as being too high. The lower E-HIPC thresholds, while welcomed by NGOs and other critics, are still judged by many to be too high. Moreover, critics consider them arbitrary and lacking a scientific basis. 4.21 While there is no theoretical basis for the ratios, they were based on empirical investigations of historical data (Underwood 1990, Cohen 1996). The analysis determined that the range of 200-250 percent for the NPV of debt-to-exports ratio was "about right" and the corresponding ratio for NPV of debt-to-GDP was 50 percent." These findings provided the rationale for adopting the 200-250 range as the threshold for eligibility to the O-HIPC. In the review of the HIPC Initiative, which resulted in the E-HIPC in 1999, to a large extent under pressure from civil society groups and advocates for deeper debt relief, the ratio was lowered to 150 percent-in order to provide countries a cushion to absorb exogenous shocks. 4.22 Using the same argument as before, but more formally estimating the level of debt stocks at which countries are more likely to run into debt-servicing problems, Cohen (2001) estimates thresholds of 200 percent of NPV of debt-to-exports and 50 percent of NPV of debt-to-GDP as the turning points. Other analyses are less definitive but generally supportive. Using an alternative methodology, Elbadawi and others (1997) find the turning point for the debt-induced growth Laffer curve for a sample of African countries to be quite high (97 percent of debt-to-GDP ratio). Pattillo and others (2002) estimate the turning point to be 160-170 percent from one statistical technique, but a wide range of turning points using alternative techniques. 4.23 These findings support the shift down from the 200-250 percent range but do not challenge 150 percent as the appropriate threshold. They demonstrate the difficulty in establishing a truly scientific basis for such a number. To some extent, it is inevitable that such a threshold will be arbitrary. There is no particular level that can guarantee debt sustainability-which is a function not only of debt reduction, but also of the volume, pace, and terms of new borrowings, as well as economic and export performance. These variables are influenced by exogenous and behavioral parameters. 4.24 To assess whether the 150 percent target for the NPV-to-debt ratio provides the HIPCs a reasonable chance of attaining debt sustainability, figure 4.1 compares the ratio for 29. Underwood's analysis was heuristic, assessing the likelihood of countries running into debt service problems as determined by Paris Club reschedulings. Cohen (1997) confirmed these findings using formal statistical methods. Cohen (1996) used alternative methods to arrive at similar conclusions. 21 HIPCs with that of other poor but non- Figure 4.1: HIPC Target for Debt-to-Export highly indebted countries. The comparisnisbmde witnno. Th Ratio Compared with Other Poor Countries comparison is made with non-HIPC 50 IDA countries and non-HIPC lower- 50 middle-income countries. The figure Al450 . Al HIPCs shows that, on average, other IDA 400 ---- countries have sustained debt-to- o--- - --- x 300 - export ratios of over 150 percent. The 2 Lower-middle- non-HIPC lower-middle-income 200 Income countries have sustained up to 250 = 150 percent debt ratios.3" Through the C 100 -I 1990s, both groups maintained them so at levels under 150, with IDA g 1 g a countries averaging 148 percent for 1998 and lower-middle-income Source: Global Development Finance and World countries averaging 135 percent. Thus Development Indicators. there is some historical justification for the 150 percent target. 4.25 An alternative measure of the debt burden is the NPV of debt-to-GDP ratio. As shown in figure 4.2, the average for this indicator for the non-HIPC IDA countries remained between 30 and 40 percent in the 1990s, although earlier it was lower. The ratio for the lower-middle- income countries has historically been above 50 percent and stayed around 40 percent in the 1990s. It would seem, then, that 30 to 40 percent would be a good target to bring the HIPCs to a position comparable to that of other poor countries. Based on the NPV of debt reductions offered by the E-HIPC, the average NPV of debt-to-GDP is expected to be reduced to 30 percent. Thus, although the NPV of debt-to-GDP ratio is not explicitly used as an indicator in the Initiative, the actions envisaged are expected to bring the debt burden of the IPCs, as measured in this way, down to a level Figure 4.2: What Would Be a Good comparable to other poor countries. Debt-to-GDP Target? Similarly, the indicators for the ratio of 100 debt service-to-exports, GDP, and 90 - -- - - - - - -- revenues also show that if the Initiative so succeeds in delivering the envisaged C o 70- amount of relief, the HIPCs will be at par a 60- Lower-middler Income or below the current ratios for other poor o 50. in 40 countries. (D 40 - > 30 - 0. Z 20 -- -- 4.26 While any specific number may 10 Other IDA be arbitrary, in the interest of 0 , ,, . . ,. transparency and equity of treatment it is - 0 ' C politically necessary (although not always the optimum in specific circumstances) to Source: Global Development Finance and World Development Indicators. use a simple, uniform rule. This is a 30 The sample of MICs used here does not include any of the 1980s MIC debt crisis countries. 22 lesson learned from the dissatisfaction with the "case-by-case" approach that governed the earlier debt regime, which was perceived as being too "political." More important, it should be clear that there is no current debt level that "ensures" debt sustainability with any certainty. The ratios are composites of variables subject to the future decisions and behavior of the country and its creditors, in the case of the debt stock, and unpredictable exogenous factors, in the case of exports. The completion of the Initiative cannot guarantee that a country will avoid further debt-servicing problems down the road. 4.27 Maintaining debt at sustainable levels requires prudent debt management, a key element missing in most HIPCs in the past, and a need for close and continuous monitoring of the debt inventory, including domestic debt, to ensure that the country continues to remain on a sustainable debt path. The country and its creditors should ensure that the extent and nature of future loans are consistent with HIPC targets. 4.28 The exclusion of domestic debt: Domestic debt is excluded from the analysis but some, including many debtors, have criticized this, arguing that servicing domestic debt adds as much to a government's fiscal burden as servicing external debt. But integrating domestic creditors into the process would be unwieldy (for instance, when applying the principle of equal burden-sharing) because they are likely to be a very diverse group. Moreover, the level of domestic debt is more open to manipulations. Devising mechanisms to avoid moral hazard issues would be challenging. For example, a government can easily increase its domestic debt (such as by taking over a failed bank). Finally, reducing domestic debts might wreak havoc on fragile domestic financial markets. 4.29 In view of these characteristics of domestic debt, the decision to include only external debt for the HIPC Initiative is appropriate. Yet the effective management of domestic debt can be an important prerequisite for the success of the HIPC Initiative. Therefore, the levels and characteristics of domestic debt need to be monitored and reported at the decision and completion points, and more routinely through World Bank and IMF monitoring-for instance, in the context of the Poverty Reduction and Growth Facility (PRGF) review. Further, to the extent that the external Initiatives add to fiscal burdens, care must be taken that these do not lead to excessive domestic debt. 4.30 One issue raised in the debtor feedback workshops was to allow increased flexibility in the current framework so that HIPC debt relief could be used to deal with domestic debt servicing (although not to include domestic debt in the calculation of HIPC relief). To the extent that controlling or reducing domestic debt or debt service is critical for macroeconomic stability, such actions fall under the purview of PRGF arrangements within the context of overall fiscal management. The HIPC Initiative has recently shown such flexibility. For example, in the case of Ghana a part of the HIPC debt relief is earmarked for domestic debt retirement. However, here again, care needs to be exercised to assess who holds domestic debt and what the likely distributional implications are. Do DSA Projections Provide a Robust Forecast? 4.31 The forward-looking component of the DSA projects the key indicators for debt sustainability using forecasts for key macroeconomic variables. These projections do not 23 have any impact on the eligibility criteria or the amount of relief that will go to individual countries. The projections are used to assess whether, after receiving HIPC relief, the profile of a country's debt burden is likely to remain within the prescribed range over the next 10 to 20 years, and therefore to demonstrate whether the Initiative is likely to achieve its main objective. Debt projections can be evaluated both in terms of the methodology they employ to connect endogenous variables to exogenous variables, and the assumptions they make about the exogenous variables themselves. The former is assessed here; the assumptions will be discussed in the next chapter on implementation. 4.32 This review sought to assess the models underlying the forward-looking DSA projections, notably the consistency with which they are applied and their sensitivity to selected assumptions. Key components of the debt projections for each country are the balance of payments projections, derived from the IMF's macroeconomic framework. Because the methodological basis of these projections was not made available, it is not possible to attest to the consistency between the balance of payments projections and the analysis of debt sustainability. In particular, it is unclear how the Initiative itself is incorporated into projections of imports and net resource transfers or how the debt projections are integrated with the macroeconomic framework. It would be useful for the World Bank and IMF to provide an explicit statement of how the debt projections, including their balance of payments and fiscal components, are arrived at." A review of some of the analysis in the HIPC documents and consultations with World Bank and IMF staff raise concern that accounting identities are either not observed or are forced by using certain lines as arbitrary "buffers." While it is necessary to make assumptions in exercises such as the DSA, the basis for the key assumptions needs to be documented and made explicit. 4.33 Other external reviewers have experienced similar difficulties in seeking to comprehend the basis on which DSAs are conducted. According to the U.S. General Accounting Office (GAO 2000, Appendix VI) "inconsistencies and gaps in the types of information provided in the debt sustainability analyses presented challenges in our analyses of the documents."" An independent research and advisory group, Development Finance International (2001), makes similar points: "It is important that there is consistency between the new borrowing assumptions of the DSA and the HIPCs PRGF limits and that there is an objective basis for the specified new borrowing limits of HIPCs. The debt sustainability analysis needs to be directly linked up the Poverty Reduction Strategy Papers so one can see if debt relief (both amounts and timing) coupled with new inflows is sufficient to achieve HIPCs poverty reduction plans." The reporting by the World Bank and IMF on the implementation of the HIPC Initiative has been more comprehensive and standardized in the context of the E- 31. In addition, counterfactual scenarios with different projections about exports and the terms of future lending should compare the situation with and without the HIPC Initiative. 32. The Appendix concludes, "Critical information that we derived from the debt sustainability analyses is not presented explicitly within HIPC documents. In some of the decision point documents, the actual amount of debt relief a country would receive is not explicitly presented and had to be deduced from other data provided. In addition, the amount of borrowing that each country would need after debt relief is not reported and required a complex methodology for U.S. to derive. Finally, although the documents discuss the resources that debt relief would contribute to poverty reduction activities, they do not mention that these are financial resources that each country would have to borrow" (page 110). 24 HIPC. Nonetheless, a more transparent, explicit, and consistent methodology for the debt projections that connects all the relevant components of the fiscal budget, national accounts, financial flows, and balance of payments remains necessary. Performance Crifteria 4.34 There are two sets of performance criteria a country has to meet to be eligible for debt relief under the Initiative. One is an ex ante requirement of the establishment of a strong track record of policy performance to reach the decision point. The second is a set of conditions, called the completion point triggers, that must be met to reach the completion point, when the full amount of HIPC assistance promised is delivered and made irrevocable. Track Record Requirement 4.35 There is little disagreement among researchers and policymakers that growth is essential for both debt sustainability and poverty reduction. There is also little dispute about the need for supportive economic policies, macroeconomic stability, and sound economic management to promote growth and poverty reduction, as reiterated in the Monterrey Consensus statement. The lack of sustained macroeconomic adjustment policies and structural reforms has contributed to debt buildup in the HIPCs (Brooks and others 1998). The country case studies for this review confirm that poor fiscal management and failure to adjust to evolving economic circumstances were significant contributing factors in all of the country cases reviewed (Annex I). 4.36 OED's Country Assistance Evaluations (CAEs) conducted since 1997, available for 13 HIPCs, identify significant shortcomings in macroeconomic and fiscal management and the policy framework. Annex J summarizes the constraints to growth and long-term sustainability identified by the CAEs. A common cause of rising external debt has been the combination of adverse terms of trade, weather shocks, and poor fiscal management (lax and inefficient tax administration and uncontrolled expansion of unproductive public expenditures, with consequent heavy reliance on external finance). Poor economic performance has been a result of political instability, poor macroeconomic policies, and institutional constraints, including poor governance. Infrastructure, agriculture, and private sector development have typically been neglected in the design of poverty reduction strategies, as highlighted by OED's independent review of IDA 10-12 (Gwin 2002). 4.37 OED studies of World Bank structural adjustment programs have also identified the policy environment and fiscal management as the principal causes of poor economic performance in the early years of adjustment. Where successful, adjustment operations and structural and sectoral reforms have promoted growth and have been a significant factor in reducing poverty. But the studies also found shortcomings in the programs, such as fragmented treatment of different fiscal issues and vague or imprecise conditionality. A significant shortcoming was the lack of attention to the social impact of adjustment. The challenge has been poor compliance with conditions and the lack of borrower ownership of the reform program. The Higher Impact Adjustment Lending (HIAL) operations sought to improve the design and enhance the effectiveness of adjustment lending through improved 25 country selection, and conditions that are fewer and better designed but more effective. These operations have had better outcomes than their predecessors and have been different in their attention to social concerns and poverty (OED 1999). The issues of ownership and the diverse social impacts of adjustment lending are also highlighted in reviews of ESAF (Botchwey and others 1998). 4.38 Given the importance of an appropriate policy framework and sound macroeconomic management to ensure that the relief resources are well utilized, the O-HIPC Initiative required of a total of six years of track record-three years prior to the decision point, and three years to reach completion point. In the E-HIPC this was reduced to a three-year track record before decision point (to be applied flexibly to give credit for past performance), and a floating completion point that would allow the country to undertake reforms at its own pace. The design incorporated lessons of experience on the need to streamline conditions, with fewer conditions attached to the new PRGF programs. To foster ownership, the Initiative also envisages the PRGF conditionality to emerge from the new PRSP process, with broad participation from civil society in defining and supporting the adjustment program for attaining macroeconomic stability and a sound policy framework. 4.39 Thus, the design of the current framework is consistent with the need to focus on the policy framework. The modifications to the track record requirements in the E-HIPC are also consistent with a key lesson from the middle-income country debt crisis of the 1980s: the need to focus on countries with convincing policy track records, but once that is accomplished, the process should not delay the delivery of debt relief (Cline 1997). Completion Point Triggers 4.40 To reach the completion point, each country has to comply with a set of conditions. Two generic requirements are (a) staying on track with the IMF's PRGF macroeconomic stabilization and reform program and (b) the development and implementation (for at least one year) of a participatory PRSP. In addition, the staff of the World Bank and IMF, with the government, identify key structural and social development actions or reforms which would promote progress toward sustainable development. Beyond the two generic requirements, the design allows for the triggers to be tailored to suit the particular circumstances of a country and expects them to be designed in full consultation with its government. Should this materialize, the foundation for sustainable growth and poverty reduction required by HIPC should be secured. 4.41 The added focus on poverty reduction led to the inclusion of social sector policy reform and performance criteria as explicit completion point triggers, and very often in the allocation of HIPC "savings" for social expenditures. As noted in Chapter 3, the modifications to the HIPC framework introduced in 1999 specifically shifted the focus of the performance assessment by giving more weight to social and poverty-related reforms. 4.42 The World Bank and the IMF are required to monitor progress toward the triggers, but the initial design did not incorporate any safeguards against poor public expenditure management or other governance issues. The process mandated specific uses for the relief "savings" but the need to "track" expenditures and outcomes only emerged later as the 26 realization set in that debt relief essentially amounted to budget support. These issues have been recognized and actions are being taken to monitor the use of HIPC "savings" as part of tracking overall poverty reducing expenditures by governments under regular IMF and World Bank program reviews. A major constraint in doing this is the weak budget formulation, execution, and reporting systems. Multi-donor efforts are under-way to improve public expenditure management systems in HIPCs (World Bank and IMF 2002a). Crediftor Participation 4.43 The Initiative was designed to provide a comprehensive treatment of the HIPCs' debt problem through a concerted and uniform approach with the participation of all creditors: the multilateral, Paris Club and non-Paris Club bilateral, and commercial creditors. Although agreed to by governments, there is no formal or legal accord that obliges creditors to participate. The legal sanctity of the loan contracts signed between a creditor and a debtor is not supplanted by the HIPC process. In essence, the decision to participate in the HIPC Initiative is a voluntary one, and the Initiative is essentially an informal agreement among nations. Moral suasion is the main enforcement mechanism. 4.44 Although the design assumes the full participation of all creditors and emphasizes equal burden sharing, the process requires individual debtors to follow-up with the creditors, as was the case with the previous Paris Club treatments. The current design does not adequately take into account the capacity constraints faced by HIPC governments, or the inefficiency of the process in terms of the time and resource requirements to follow through with individual bilateral creditors." More important, it does not address previously known problems with the non-participation of a number of creditors, particularly the commercial and non-Paris Club bilateral creditors who have refused to participate in the past. Given that this is a persistent problem, the assumption of full participation maintained in the DSAs is also invalid. Capacity Building 4.45 Debt management is an important element in debt sustainability. It has been long recognized by the World Bank and the IMF that poor debt management, because of weak institutional capacity, has been an important factor exacerbating debt problems in many countries. But capacity building for debt management has not been an objective of the Initiative, and the design does not address any aspect of capacity development or technical assistance for debt management. Decision and completion point documents discuss debt management issues as part of their forward-looking assessments for individual countries, but the Initiative does not address any aspect of capacity development for debt management.' This remains a design shortcoming. 33. The processes for the multilateral development banks (MDB) have been streamlined, with technical problems and methodological issues addressed through semi-annual MDB meetings chaired by the World Bank. This has helped secure smoother and more timely debt relief from most of the MDBs. 34. Technical assistance has been provided by other agencies, primarily UNCTAD, through some bilateral donor-financed agencies such as Debt Relief International (DRI), and regional Initiatives. 27 5. Implementation Experience 5.1 This chapter reviews how the Initiative is being implemented. It assesses the extent to which the Initiative is being implemented efficaciously, equitably, and in accordance with its key objectives. Following a brief background on the pace of implementation, the discussion below is structured around the four key elements of the HIPC process: the debt sustainability analysis, application of the performance criteria, creditor participation, and institutional development. Pace of Implementation 5.2 The O-HIPC got off to a slow start. From 1996 to 1999, only seven countries became eligible for debt relief. The modifications in the process introduced by E-HIPC, particularly the lowering of the debt sustainability thresholds, made additional countries eligible. With continued pressure by civil society and impatience of the Development Committee at the seemingly slow pace of progress, implementation was accelerated in the latter half of 2000 to meet a self-imposed challenge by the World Bank and the IMF to bring at least 20 countries to decision point by December 2000. With extraordinary efforts on the part of the staffs of the World Bank and IMF, helped by the enhancements introduced in 1999, the HIPC Initiative exceeded its much-publicized target, with 22 countries reaching their decision points by December 27, 2000. The pace has since slowed down, with only another four countries qualifying for debt relief by August 2002. The challenge of bringing the remaining countries to their decision points, particularly those affected by conflict, remains. 5.3 The late-2000 "millennium rush" reflected the flexibility in the implementation of the Initiative, and is generally perceived to have been beneficial in bringing several countries into the process (see Annex H on creditor perspectives). At the same time, it is also perceived by many, including several creditors and NGOs, as having diluted adherence to the standards and process requirements incorporated in the design, and has raised the issue of equity of treatment. Debt Sustainability Analysis 5.4 The debt sustainability analyses (DSAs) done at decision point, and updated at completion point, do not have any impact on the amount of relief that a country receives under the current framework. They do have profound implications for assessing the likely outcomes, and hence the overall success of the Initiative. Two assumptions in the DSA influence the outcomes: the level and terms of new financing and an assessment of future economic and export performance. 5.5 The level and terms of new financing assumed in the DSAs have been criticized as being too optimistic (GAO 2000). While the interest rates for loans assumed in the DSAs are lower than they have been in the past, they are not much lower than current levels. Further, they are based on a survey of creditors and donors about their future commitments. The level and terms of new borrowings are also monitored under the IMF PRGF reviews, which place restrictions on the amount of nonconcessional borrowing and set limits on the minimum grant element for a loan to be considered concessional. 28 5.6 The economic assumptions in the DSA projections have been the main subject of criticism. The assumptions about the future growth rates of exports and GDP have been criticized as being too optimistic (GAO 2000; Martin and Alami 2001). These concerns have also been consistently raised by the debtors and creditor officials (Annexes G and H). 5.7 To assess this criticism, table 5.1 compares the export growth assumptions (for the 5 years after the decision point) maintained in the DSAs with historical performance for the 24 countries that reached their decision points by end-2001." The projections were updated in 2002, taking into account the experience of 2000 and 2001. To assess how the updated projections compare with the original decision point DSA assumptions, table 5.1 also gives the assumptions for the next five years (2003-2008) maintained in the original and the updated projections. Export earnings, and trade flows in general, are very volatile. To assess the degree of confidence in the forecast assumptions, table 5.1 also includes the upper and lower bounds of what might be considered reasonable forecasts based on historical experience." 5.8 The overall simple average of the growth rate assumed in DSAs across all 24 countries is more than twice the historical average for 1990-2000 and almost six times the average for 1980-2000. Individually, the forecast is below the historical estimate for 1990- 2000 for only four countries, but is within statistically reasonable bounds for eight (shown by the shaded cells in the table)." For the other 12 countries, the assumptions often far exceed historical rates (Chad is an exception, pending the expected increase in oil revenues). Also, the lower bound is in most cases much lower than the assumed rate, and in many cases is negative. The high probability of such poor outcomes requires appropriate sensitivity analyses to reflect the real possibilities. The DSAs, however, do not typically simulate the effects of such low growth rates. 5.9 The findings show that there was optimism in the assumptions maintained by the DSAs done at decision point. A similar analysis of the assumptions in updated DSAs (for 2003-2008) shows that although slightly better, the assumptions continue to be optimistic. In only nine cases are the assumptions in the updated DSAs within the reasonable (upper) bounds of confidence (in addition to Chad, Sa6 Tom6 and Principe is also an exception for this time period, as oil exports are expected to rise significantly by about 2006). Comparing the old and new forecasts for the same period (2003-2008) also shows that the new assumptions are slightly more optimistic. 35. The historical growth rates are obtained using econometric regressions, which have the advantage over the average year-on-year growth rates in not being as influenced by the starting and ending dates and by outliers. 36. The bounds represent the commonly used 95 percent statistical confidence interval for the estimated growth rates. 37. The detailed comparisons are restricted to the performance in the period 1990-2000. Economic performance in many countries was much better in the 1990s than in the 1980s because of more effective policy reforms and, in some cases, buoyant export markets for primary products, such as coffee, in the mid-1990s. Thus, although a longer historical period (say from 1980 to 2000, as used by GAO 2000) may be better for increased precision, this would ignore possible structural breaks marking a departure from the past. Table 5.1: Comparison of DSA Export Growth Projections and Historical Performance OSA assumptionsa Historical rformanceb Enhanced 2000-05c 2003-08 1990-2000 1980-2001 HIPC Lower Upper Lower Upper decision point Country Original Original New Estimate bound bound Estimate bound bound date Bolivia 10.52 8.24 8.18 5.18 2.88 7.48 4.68 4.22 5.13 Feb-00 Mauritania 444 7.11 1.20 -2.61 5.01 1.88 1.06 2.69 Feb-00 Uganda gE O9, ! 1.7 15.14 8.39 21.90 0.31 -2.05 2.67 Feb-00 Mozambique 18.98 6r495q 21.27 10.37 8.96 11.78 -0.61 -2.57 1.35 Apr-00 Tanzania 11.51 10.50 7.70 3.35 1.72 4.98 0.97 -0.64 2.59 Apr-00 Senegal 9.02 6.44 6.25 -0.16 -1.85 1.54 3.14 1.53 4.75 Jun-00 Benin 7.32 8.19 8.74 4.18 1.51 6.85 4.20 3.07 5.32 Jul-00 Burkina Faso 10.89 11.20 9.81 -1.20 -3.75 1.35 2.47 1.16 3.77 Jul-00 Honduras 13.72 9 11.28 9.40 13.16 2.98 1.79 4.17 Jul-00 Mali 7.35 '-5 4.93 3.01 6.84 3.31 2.12 4.49 Sep-00 Cameroon 8.04 6.74 4.89 0.63 -2.14 3.41 0.65 -0.72 2.02 Oct-00 Guyana 416- 4.49 4.76 9.55 5.99 13.12 -0.17 -2.43 2.08 Nov-00 Gambia, The 7.58 6.31 4.34 0.05 -2.25 2.35 3.73 2.77 4.68 Dec-00 Guinea 8.54 7.15 7.26 -1.79 -2.73 -0.84 4.33 3.31 5.34 Dec-00 Guinea-Bissau 1 4' 2 11.75 3.97 19.54 1.61 0.63 2.59 Dec-00 Madagascar 9 1 8.40 6.34 10.46 0.85 -0.30 2.00 Dec-00 Malawi 372 2.13 -1.60 5.86 2.74 1.42 4.07 Dec-00 Nicaragua 9.66 15.25 9.50 21.01 0.21 -2.17 2.60 Dec-00 Niger 3.41 7.61 4.86 -0.02 -2.44 2.39 -0.17 -1.28 0.94 Dec-00 Rwanda 15.06 14.79 12.24 0.31 -7.91 8.53 -0.14 -2.19 1.91 Dec-00 Sa6 Tomd and Principe 10.59 9.84 67.96 5.32 2.67 7.96 -0.51 -1.30 0.29 Dec-00 Zambia 12.73 6.56 5.36 -2.39 -5.12 0.33 0.13 -1.28 1.53 Dec-00 Chad 65.07 64.17 94.51 2.53 -1.97 7.03 4.14 2.93 5.35 May-01 Ethiopia 08i1O5 7.45 3.55 11.36 0.51 -0.81 1.82 Nov-01 Simple average 11.74 10.45 14.33 4.73 1.40 8.06 1.72 0.34 3.09 Without Chad, Sa6 Tom6 9.37 8.04 8.25 4.80 1.49 8.11 1.71 0.30 3.12 _ Note: Ghana and Sierra Leone qualified in 2002 and are not included in the sample due to lack of data at the time of analysis. a. From HIPC database, Spring 2002. b. From data provided by IMF. Reported are coefficient estimates from a regression of natural logs of exports of goods and nonfactor services on time (year) The lower and upper bounds correspond to the 95 percent confidence interval of the estimated growth rate. c. Estimates for Ethiopia and Chad are from 2002-2007. 30 5.10 The forecasts for GDP growth rates assumed in the DSA are also criticized as being overly optimistic." As noted in Chapter 4, assumptions about economic growth and low interest rates are important for the achievement of debt sustainability in the DSAs. The projections assume that if countries follow appropriate policies, the projected growth rates could be realized. Looking at the evidence, there is some merit in the argument. Recent analyses of countries classified as good, weak, and poor compliers with structural adjustment conditions show that committed structural adjustment can produce tangible results (OED 1997; World Bank 2002). Using the findings from the OED review of structural adjustment programs, formal statistical evidence shows policy compliance does matter and does result in a higher rate of growth." Other studies also show that the new structural adjustment , programs are better designed, incorporating past lessons on ownership and streamlined conditionality, and are performing relatively better than the programs of the 1980s (OED 1999; SPA 2002). Evidence from country cases studies (such as Guyana, Malawi, and Uganda) shows that significant economic reforms, undertaken in the 1990s, were associated with much improved economic performance. 5.11 The HIPC requirements of a prior policy track record and continued performance in maintaining and furthering macroeconomic and structural reforms thus take on added importance in future debt sustainability. Assuming that these policy reforms are fully implemented, as agreed to by the countries, it is reasonable to add a "policy premium" to historical growth rates for future projections. Available statistical evidence suggests the magnitude of this growth premium to be between 1 and 2 percent per capita real Figure 5.1: StructuraR Adjustment CompRance growth (Easterly 2001a; Noorbaksh and Fosters GDP Growth Paloni 2001). 4.5 5.12 The potential magnitude of the .8 "policy premium" is confirmed by S God some simple data analyses. The 1997 0 3.5- -O----- ---- - --OED study of Structural Adjustment 3 - Loans completed between 1988 and 1994 by SSA countries rated each 0 country by the level of compliance 2 Wewith the conditionality included. PCI Poo' -Figure 5.1 shows the correlation between the degree of compliance, A 1rated as good, weak, and poor, and < 0.5 growth rates for the 33 HIPCs included 0 198-85in the study. The trends in the growth 1980-85 1986-90 1991-95 1996-2000 rates for HIPCs confirm the findings from the broader studies. The figure Source: OED 1997 and World Development Indicators, shows that starting from a worse 38. On average, for the 24 countries that reached their decision points by end-2000, the assumed real GDP growth rates are more than double the average growth rates achieved in 1990-2000. 39. This finding is consistent with the external review of ESAF, which notes the impact of structural adjustment has been varied across countries. In countries where the program has been implemented, economic performance has improved (Botchwey and others 1998). 31 position, the good compliers have achieved higher growth rates than the poor compliers, both in the short and the long run. The difference in the real GDP growth rates between good compliers and poor compliers is less than 2 percent per annum. Thus, while policy-based projections are justified in assuming higher growth rates, evidence also suggests that these projections need to be reasonable. 5.13 Further, recent research findings also suggest a need for conservatism in projections. While volatility entails both down and upside risks, negative shocks tend to do more harm than the good delivered by an equivalent upside shock. That is, negative shocks in export prices tend to have a much greater impact on growth than do positive shocks (Collier, Gunning and Associates 2000; Dehn 2000; Deaton and Miller 1995). At the same time, positive shocks do not appear to have any lasting effects, which implies that windfall gains are not being properly saved and invested (Dehn 2000). These findings suggest a need to better incorporate volatility in exports in the DSA projections, perhaps through more robust sensitivity analysis. Performance Criteria 5.14 The implementation of the two types of performance criteria, the track record on policy performance (to be eligible for the decision point), and the completion point conditions (to reach the completion point) are considered in this section. Track Record 5.15 An important difference between the past debt relief efforts and the HIPC Initiative is the Initiative's application of a uniform set of rules to assure equity and fairness in the treatment of individual countries. The main requirement for both the O-HIPC and E-HIPC has been a three-year track record of sustained policy and structural reforms and macroeconomic stability. From the very beginning this condition has been applied flexibly. None of the countries qualifying for HIPC assistance under the O-HIPC had to wait for three years; they were given credit for strong policy performance in prior years. The same has been the case in the E-HIPC. 5.16 The application of a track record has been progressively reduced during the E-HIPC. All six countries that successfully completed the O-HIPC process had a strong policy track record in terms of performance under IMF ESAF/PRGF, World Bank adjustment programs, or both. Similarly, all nine countries that entered the E-HIPC process between January and July 2000 had a satisfactory or strong track record. But of the 13 countries that reached decision point between August and December 2000, only 2 had established a track record.' 5.17 Not only did these 13 more recent entrants (that is, the "millennium rush" group) have weaker economic track records, they also-in the aggregate-have a weaker development program. They have, for example, lower ratings for the outcome of World Bank assistance in OED's Country Assistance Evaluations. Their completed World Bank projects 40. Of these 13 millennium rush countries, all 11 reaching their decision point in November and December 2000 had a mixed track record. 32 have significantly lower satisfactory outcome ratings than the earlier entrants. And their Country Policy and Institutional Assessment (CPIA) scores in 2001 were lower than those of the earlier entrants (about one-half of one point lower on a six-point scale). " 5.18 Based on this evidence, the HIPC countries that reached their decision point during the "millennium rush" are less likely to achieve good development results than their predecessors, other things being equal. Completion Point Conditions 5.19 The HIPC Initiative incorporates three types of conditions for reaching completion point: o Process conditions, directly linking the HIPC process to development of a PRSP o Performance benchmarks, also termed completion point triggers o The use of HIPC resources for earmarked purposes. The implementation of each is discussed. 5.20 Process conditions. The Poverty Reduction Strategy Paper (PRSP) operationalized the E-HIPC Initiative's objective of linking debt relief resources to the promotion of poverty reduction. Support for a more participatory and holistic approach to development was emerging in its own right at the time of the HIPC review in 1999. The PRSP embodied such an approach and incorporated lessons of experience from past development efforts, emphasizing the need for ownership and participation, as well as the holistic view expounded by the Comprehensive Development Framework (CDF). By putting the government in the driver's seat and including conditionality that called for the participation of a wide cross- section of stakeholders, the PRSP aimed to improve ownership, transparency, accountability, and the focus on results. 5.21 Skeptical at first, civil society supported the PRSP as a potential step in the right direction, but many of its members were concerned that the development of a full PRSP might delay countries from reaching their decision points. An interim PRSP (I-PRSP}- essentially a statement of intent and a roadmap to carry out a full PRSP-was determined to be a compromise criterion for decision point qualification. A full PRSP, with satisfactory implementation for one full year, was retained as a requirement for completion point. 5.22 The weak capacity of many countries to carry out complex consultations and strategy formulation prompted concerns among some stakeholders that the link between the HIPC and the PRSP processes would either delay the delivery of HIPC relief or provide an incentive for countries to rush the PRSP to secure the debt relief. To address this concern and to deliver relief faster, the innovation in E-HIPC was to introduce the provision of "interim debt relief' between the decision point and the completion point, when debt relief would be provided in 41. The Country Policy and Institutional Performance (CPIA) Index is a summary measure of the policy and institutional performance of a country as assessed by the World Bank's country team. It covers a broad range of areas, including macroeconomic and structural policies, public sector management and service delivery, and policies for equity and social inclusion. The CPIA ratings are used, among other things, to allocate IDA resources. 33 full and irrevocably. The annual debt service relief during the interim period is projected to be substantial, in most cases amounting to over 80 percent of the annual debt service reduction expected after the completion point. 5.23 Most debtor officials and stakeholders acknowledge the HIPC Initiative as the main rationale and motivation for the PRSP process in HIPCs, but they note that the provision of interim debt relief largely removed the incentive to hastily complete the PRSP. OED's country case studies and feedback from debtor representatives at workshops held for this review indicate that the link between the HIPC and PRSP processes has been beneficial, as discussed in box 5.1.42 Box 5.1: Debtor Countries Endorse the Links between the HPC Initiative and the PRSP The participants at the two workshops organized to gather feedback from debtor government representatives for this review considered the PRSP to be an integral part of the HIPC Initiative and the primary instrument for achieving the HIPC objectives of debt sustainability and the promotion of poverty reduction. The majority of speakers held that the costs of preparing the PRSP, while substantial, were justifiable; most maintained that the benefits outweighed the costs. Internally, the PRSP process has created a new partnership among development partners, helping to identify national priorities, build broader ownership of the development agenda, and improve governance through greater transparency. Several of the representatives underlined the growing-and in some cases unprecedented-contribution of civil society to the PRSP. In the words of one official, "HIPC has caused civil society to organize itself in order to influence the process." By most accounts, the policy dialogue surrounding the PRSP is much improved compared with past experience with adjustment programs. Speakers noted that in the past, the discussions had been dominated by the International Financial Institutions and inordinately focused on economic growth. While a number of speakers said that the PRSP had succeeded in putting countries "in the drivers seat," most perceived the requirement that the Joint Staff Assessment be endorsed by the Boards of the World Bank and IMF as tantamount to requiring endorsement of the PRSP itself. They held that this undermined the spirit of true ownership. Some also questioned the extent to which donors were willing to accept the PRSP as the single framework for development cooperation. Several speakers pointed to the need for closer alignment between the macroeconomic framework and the PRSP and called for greater flexibility by the World Bank and IMF to achieve this. Representatives credited the HIPC Initiative with bringing poverty reduction into sharper focus and held that a major challenge for the PRSP would be to unite the growth and poverty reduction objectives. Speakers emphasized the importance of a holistic approach to development-improving human capital, nurturing the growth of the private sector, and building and maintaining infrastructure, among other objectives. 5.24 An assessment of whether the process has been rushed as a result of being linked to the HIlPC process requires an in-depth comparative assessment across HIPC and non-HIPC countries. Emerging findings from case studies for the ongoing evaluation of the CDF by OED indicate that some of the initial PRSPs appear to have been rushed. But how the link to HIPC 42. Annex E gives a summary of proceedings from debtor feedback workshops and Annex Ga synthesis of the country case studies. This finding supports the conclusions of the PRSP review by the World Bank and IMF (2002b). 34 relative to the potentially much larger volume of development aid has affected the process needs further study (since, at least in principle, PRSPs are to be the basis for future aid flows). Due consideration also needs to be given to capacity constraints and domestic political factors. These issues will be taken up in the review of the PRSP process by OED in the coming year. 5.25 Local NGOs, donors, and other stakeholders in the case study countries expressed satisfaction with the nature of consultations and the participatory process undertaken by the government. In many places, the process has been called "pathbreaking" in the way it establishes a new relationship between the government and civil society. Across countries, however, the degree of PRSP ownership varies. In some cases, there is substantial ownership by the political leadership of the country, but less so at the lower operational levels (as in Guyana). In others, political leaders or parliamentarians have not been substantively involved in the process (as in Malawi).43 The overall ownership of the PRSP is also affected by other factors such as the role of the International Financial Institution (IFI) Boards in endorsing them as the basis for international assistance (see box 5.1). To what degree and how the process is affecting the ownership of the PRSPs is another area the OED review of PRSPs will pursue. 5.26 While the consultative process has been quite satisfactory, the quality of the substantive aspects of the strategy has been a subject of debate. The strategies have been criticized by donors in some countries as being long wish lists in need of further prioritization and costing (World Bank and IMF 2002b; DFID 2002). Most strategies, as presented, are financially untenable. While growth and economic development figure prominently in the stated objectives of most strategies, the identification of priority activities is weak, as is the articulation of how the growth is to be achieved. 5.27 Among the case study countries, only Uganda had completed its PRSP at the time the studies were conducted. As noted in box 4.1, Uganda has been a leader in the process innovation on participatory poverty reduction strategy formulation and implementation. Even in Uganda, although the Poverty Eradication Action Plan (PEAP) provides a credible strategy to meet its four objectives, the government has focused much of its attention on two of the four legs of its PEAP relating to social outcomes and "quality of life" issues. There has been less focus on its export diversification and private sector development strategies. Government officials now recognize the need for a more balanced approach to development. 5.28 Performance benchmarks: HIPC decision point documents specify the conditions that the country has agreed to meet by completion point. Most O-HIPC decision point documents included a comprehensive list of key structural reforms under ESAF and on-going IDA Structural Adjustment Credits as part of the conditions for completion point. In addition, they required successful implementation of reforms or satisfactory progress under IDA structural adjustment and other credits. 5.29 The conditions under E-HIPC have been considerably reduced from an average of over 41 to 13 per country. The reduction has mostly been in economic policy and structural reform conditions. A significant development in the E-HIPC is the inclusion of governance conditions, with two-thirds of all countries having at least one governance condition. The 43. These findings are in line with the PRSP review conducted by the Bank and the IMF in early 2002. 35 most common types of structural reforms for the E-HIPC decision point countries were in public expenditure management (13 countries). Coverage of rural development has also decreased in the E-HIPC. While more than half of the O-HIPC countries had one or more conditions relating to this sector, only 8 percent of countries under E-HIPC (2 countries) had at least one rural development condition." - 5.30 The use of HIPC resources: Since the O-HIPC was not specifically aimed at poverty reduction, it did not seek to specify how the funds released from debt relief were to be used. Six of the seven O-HIPC countries had no targets or guidelines for the use of these funds. The debt relief was intended to be used for public spending in the social sectors and for supporting the cost of structural reforms. This approach is consistent with the current development research findings that money is fungible, earmarking is inefficient, and growth and reform are also essential for poverty reduction. 5.31 While the number of conditions was streamlined under E-HIPC, the conditions are more specific than in the O-HIPC with regard to sectoral allocations, targets by sector, and details about the intended use of HIPC funds. With the addition of poverty reduction in the E-HIPC, the focus in implementation has also shifted the balance between the alternative uses of HIPC funds. The emphasis on ensuring that resources from IPC "savings" are used for poverty reduction has increased, and this has been operationalized mainly by targeting the savings for social expenditures, mostly for education and health, as mentioned in Chapter 3. 5.32 All of the countries reaching their E-HIPC decision points have conditions targeting funds for specific sectors. Of the 20 that reached decision point after July 2000, all but one had a detailed description of how the funds were to be used, with 13 specifying numerical targets for sectoral allocation. Based on averages for these 13 E-HIPC countries with numerical targets for sectoral allocations, 65 percent of the E-HIPC savings were focused on the social sectors, 13 percent on rural development, 8 percent on infrastructure, 4 percent on governance, and 2 percent on structural reforms. Across the E-HIPC countries, 49 percent of the resources are to be allocated to education and health. 5.33 Country case studies reveal that at the central level, HUPC expenditures are being allocated largely as anticipated in the decision point documents. The budgetary resources for the targeted sectors, particularly education and health, have increased, although there is also some evidence of fungibility.45 However, the World Bank's own analysis of these sectors in Public Expenditure Reviews (PERs) highlights that increased financing is necessary but not sufficient to improve the quality of services and outcomes. The PERs, available for 17 of the 26 decision point HIPCs, reveal that the HIPC conditions only partially address the core constraints to improved performance in these sectors. Most PERs focus on social sectors, on health and education sectors in particular, and do not provide an assessment of the overall efficiency of public expenditures. The limited scope of the PERs does not permit an analysis of the absorptive capacity across sectors or the areas of priority for public expenditures. 44. By rural development, this report refers specifically to activities that enhance agriculture and agricultural productivity, rather than social services or poverty alleviation that may have a rural dimension. 45. Total allocations, including HIPC allocations, have gone up, but governments' budgetary contributions have either declined somewhat or have remained flat, while the overall (non-HIPC) budget has increased. 36 5.34 In Guyana, for instance, under the E-HIPC one trigger called for achieving satisfactory progress in increasing the numbers of teachers as identified in the interim PRSP matrix. But while the PER attributes the shortage of teachers mainly to a brain drain from the country, and recommends that teachers be given more voice in school management and be rewarded for achieving measurable goals, the I-PRSP called for "building two new dorms at the Teacher Training college, a review of teachers' diploma program and working with the Teacher's Union to improve the condition of teachers." Of these three, only the latter is consistent with the recommendations of the PER for reducing the brain drain. 5.35 Weak performance in social sectors results from low efficiency of expenditures, poor service quality, capacity shortfalls, low utilization, and inequitable allocations. Although many of the completion point triggers do address relevant sectoral constraints, increased budgetary allocations in most cases have been superimposed on weak institutions, which is likely to limit the effective use of HIPC resources. Some examples are given in box 5.2. Box 5.2: PERs ighligh t Absorptive Capacity Constraints in the Social Sectors in the HIPCs Low efficiency in education and health remains a major constraint to sectoral performance in almost every HIPC, although it was rarely addressed directly in completion point triggers. For example, one- third of HIPC resources in Honduras are allocated to the education sector. The PER for Honduras notes, "in spite of Honduras' relatively strong funding effort and high coverage, its outcomes in primary education are not notably superior to those of its neighbors." The PER finds that the costs of inefficiency are huge-about 36 percent of spending by the Secretariat of Public Education does not produce a student qualified at the grade level-and also recommends that the total spending level in education not be raised. Poor quality in many cases has become a cause of high dropout rates and repetitions in education, and is a factor that increased funding cannot necessarily address. In Madagascar, the poor quality of the education system (along with the inability of families to afford education) has been the main cause of the fall in enrollment rates and worsening indicators in primary education since the 1980s. While the triggers attempt to address the shortage of teachers in rural areas, they do not address their poor performance, which is the main driving force behind the low quality of education. Capacity shortfalls in the face of rising enrollments are the underlying cause of quality issues in several countries. Ethiopia has experienced this in the education sector, where the main obstacle to effective implementation is capacity shortfalls in all departments and at all levels. The PER states that "children had been encouraged to enroll in school even though there were insufficient resources or funds for operational costs with a consequent decline in quality." The triggers require an increase in enrollment rates for girls and a reduction in repetition rates, but do not address the basic capacity problems, ranging from issues such as buying textbooks, equipment, and transport, to improving management skills, training teachers, and building new classrooms. Low utilization or excess capacity has emerged as a constraint in Bolivia and Uganda. The Bolivia PER recommends that further incentives be designed for certain groups-for example, a negative price may be required to encourage the consumption of certain health services. In Uganda, quality and staffing in primary health care and education are major issues. Neither have any triggers to address these issues under the E-HIPC. 37 Creditor Participation 5.36 To deal effectively with the totality of a country's debt, one of the core principles of the HIPC Initiative is comprehensive and concerted action by all creditors, as discussed in Chapter 3. The participation of creditors has proved to be a challenge. The four major categories of creditors are the multilateral, Paris Club bilateral, non-Paris Club bilateral, and commercial creditors. 5.37 The feedback from debtors and country case studies indicates that most multilaterals are delivering on their promises. Of these, the World Bank and the IMF have been the most punctual and efficient. There have been some delays with the other multilaterals, but the anticipated relief is being delivered. With respect to the Paris Club, performance has varied across creditors. The process of reaching formal agreements with each individual creditor has proved to be taxing and expensive. While the relief will eventually be retroactive to the date of the Paris Club agreement, the delays have in some instances caused budgetary problems for HIPC governments, who had already anticipated such relief in their budgets. 5.38 The participation of non-Paris Club bilateral creditors has been very limited.' This has been the history prior to the HIPC Initiative, and was also experienced in the O-HIPC. The current framework continues to leave the government with the responsibility of reaching agreement with all creditors. Under the rules of equal burden sharing, the HIPCs cannot service the debt of any creditor fully, and many non-Paris Club creditors have either not responded to requests of the HIPCs or have simply refused. 5.39 Even more troublesome is the behavior of commercial creditors. Most have refused to participate in the process, and many have even successfully sued HIPC governments to recover outstanding debt.' In some instances the debt has been acquired by vulture funds at discount from sovereign creditors. For some countries these cases are proving to be quite burdensome, and most have been settled in favor of the commercial creditors. The legal case in favor of the HIPCs is weak: as noted in the previous chapter, the HIPC Initiative is an informal agreement, like the Paris Club, and does not have the legal authority to challenge a formal loan contract." 46. For the 26 countries that have reached their E-HIPC decision points, in 2000 about 10.4 percent and 2.5 percent of the nominal value of total external debt was owed to non-Paris Club and commercial creditors, respectively. Of the total debt relief to be delivered (in 2001 NPV terms), 11.2 percent and 2.3 percent are the shares of the non-Paris Club and commercial creditors, respectively. 47. Of the total obligations to commercial creditors of $2.2 billion, about $350 million or 16 percent is currently subject to litigation. Ten countries have been sued, some by multiple creditors, in a total of 23 different cases. These include two lawsuits by non-Paris club official bilateral creditors, Iraq and Burundi, although Burundi has recently suspended its court claim. The face value of total claims is $345.8 million. Of the total original claims of $167.4 million on which the judgments have been made in favor of the creditors, the total reward is $416 million. 48. It should be noted that the creditors with court judgments in their favor have not as yet been successful in converting them into payments by HIPCs. This may possibly discourage other creditors-especially those with speculative motives-from pursuing further litigation. 38 fInstitutional Development: Capacity Building Tor Externall Debt Management 5.40 A crucial element for future debt sustainability is prudent debt management. Poor or non-existent debt management has been an important factor in the creation of the HIPCs' debt problem, but in the past governments have not given it the attention it needs. 5.41 Outside of the HIPC framework, capacity building efforts in debt management have been increasing since the mid-1990s, and a number of international and regional organizations, along with several donor agencies, are playing close attention to this area.49 These efforts cover a wide spectrum of activities, from the provision of reference materials and computer software to hands-on training and assistance in preparing for debt renegotiations. After the initiation of the HIPC in 1996, five donor agencies collaborated to establish Debt Relief International (DRI) to build the debt management and negotiation capacity of HIPC governments. Three regional agencies devoted to capacity building in Africa have been established since the inception of the HIPC Initiative: the West African Institute of Financial and Economic Management (WAIFEM) in 1996, the Macroeconomic and Financial Management Institute of Eastern and Southern Africa (MEFMI) in 1997, and the Pble-Dette Initiative of Banque des Etats de l'Afrique Centrale in 2000. The Center for Latin American Monetary Studies (CEMLA) also established a National Debt Strategy in 1999, outlining a technical assistance program related to financial data systems. Among other activities focused on debt management, the Commonwealth Secretariat and the United Nations Conference on Trade and Development (UNCTAD) are leading providers of debt management computer software used by the HIPCs. More recently, the World Bank has stepped up its efforts to provide technical assistance for debt management as discussed in box 5.3. Box 5.3: fIncireasing WoirRd Bank Involvement in Debt Management Capacity Building There are growing efforts within the World Bank to strengthen debt management capacity in the HIPCs. The World Bank and IMF published "Guidelines for Public Debt Management" in 2001 and a companion volume will include case studies of proven debt management strategies among 18 industrialized, emerging, and developing countries. The Public Debt Management Group in the Treasury unit has broadened its mandate to include debt management capacity building and technical assistance for all World Bank member countries, including the HIPCs. This will include providing advice in the context of planning for Country Assistance Strategy (CAS) and project design. In addition, in a partnership with the Harare-based MEFMI, the group recently organized a seminar based on case studies conducted for Tanzania and Malawi. The Development Economics data group of the World Bank is also involved in providing technical assistance training to complement existing debt management software with simulations and linkages to macroeconomic models. One of the main conclusions from a recent inquiry into the debt management capacity of the HIPCs was that the World Bank and the IMF should be more proactive in identifying problems in debt management in the context of the HIPC Initiative (that is, decision and completion points) and in helping countries to access the resources needed to resolve them. 49. These organizations also include public agencies such as the International Development Research Center in Canada; the Agencie Intergouvernementale de la Francophonie; and the United Nations Institute for Training and Research. 39 5.42 In workshops and interviews conducted for this review, several HIPC government officials affirmed that the Initiative had significantly raised the profile of debt issues in their respective countries, and had in most cases helped to accelerate their government's efforts to improve debt management. While most officials in charge of debt management acknowledged technical weaknesses in managing the debt stock and vetting new loans, many also pointed to emerging or on-going efforts to strengthen debt management. Among the HIPC review case studies, both Guyana and Uganda have developed comprehensive debt management strategies and Malawi has recently instituted a process to streamline the review of new loans (World Bank and IMF 2002c). 40 6. LikeRy ODutcomes-PireRhImnary IFEndings 6.1 Given the short time since the HPC Initiative was launched, its economic and social outcomes will not be known with any certainty for some years. Based on available evidence and the progress made so far, this chapter assesses the likelihood of the Initiative achieving its objectives along the three major dimensions critical to its success: debt sustainability, improved economic performance, and the incremental resource transfer of the HIPC Initiative. The final section summarizes the tentative conclusions emerging from this analysis on the prospects for achieving the HIPC objectives. Debt Sustainability 6.2 This section discusses the current status of the countries that have reached their completion point or are in the process of doing so and assesses their prospects for achieving debt sustainability. The initial outcomes of the countries that have reached their completion point are considered first, followed by the status of "interim" countries-that is, those between their decision and completion points. It concludes by drawing the lessons emerging from the early experience of these countries on the prospects for debt sustainability. Compktion Point Contries 6.3 O-HIPC: Six countries reached completion point under the O-HIPC, receiving total debt relief of $3.1 billion in net present value terms. Table 6.1 shows that despite the requirement of an additional three years of track record of policy performance, none of the six had to wait the full three years before reaching the completion point. Table 6.1: Debt Sustainability fIndcatoirs foir O-HIIPC Graduates Debt Oustainability indctor: months debt/axports or debt/revenues from DP (percent) Country to CP Target Actuol at CP Action tken Uganda 12 202 196 Target reached Mali 24 200 197 Target reached Bolivia 12 225 218 Target reached Mozambique 15 200 254 Additional relief of $274 million Burkina Faso 33 205 279 Additional relief of $114 million Guyana 17 107 D/X 117 Not applicable 280 DIR 410 No additional relief Source: HIPC decision point and completion point documents; data from the World Bank HIPC Unit. 6.4 Uganda, Mali, and Bolivia successfully reached their targets for the O-HIPC completion point. Additional bilateral support from Japan ultimately brought Bolivia's debt- to-export ratio down to 198 percent. Uganda beat its target, through a combination of better than expected export performance, appreciation of the U.S. dollar, higher discount rates, and a reallocation of arrears from the public sector through privatization. Mozambique and 41 Burkina Faso qualified for additional relief because they had a higher debt-to-export ratio than targeted at decision point. Both experienced slower export growth than projected, lower commodity prices, and an increase in NPV of debt as result of a decline in discount rates and U.S. dollar depreciation. Burkina Faso's debt stock also increased because of previously unrecorded loans. 6.5 Guyana, which qualified under the "fiscal window," missed its target debt-to-revenue ratio by a wide margin. Lower export earnings, a decline in central government revenues, and the depreciation of the Guyanese dollar contributed to this outcome."o The O-HIPC framework did not provide additional relief for missing the fiscal criterion. Guyana thus completed the O-HIPC as the only unsuccessful case. 6.6 E-HIPC: Table 6.2 shows the current debt-to-export ratio for the six countries that had reached their completion point under the E-HIPC as of August 2002. Figures 6.1 and 6.2 show the DSA projections of the NPV of debt-to-exports ratio and the debt service ratio for the next 10 years. 6.7 Mozambique is firmly positioned to maintain its debt ratios below the E-HIPC threshold and the prospects for Tanzania to do so also appear good. Both have benefited from strong export performance and lower than anticipated new borrowing. However, Tanzania's narrow export base is a source of risk. Table 6.2: Debt Sustainability Indicators for E-HIPC Graduates Months Debt sustainability Indicator: debtlexports from DP to or debt/revenues (percent) Country CP Target Actual at CP Current Action taken Bolivia 16 150 143 131 Target reached Mozambique 17 150 127 108 Target reached Tanzania 19 150 137 143 Target reached Mauritania 28 137 D/X 174 174 Including voluntary bilateral debt 250 D/R 247 247 forgiveness, D/X fell to 149 percent and D/R to 206 percent. Burkina 21 150 150 216 Completion point ratio of 199 Faso percent brought down to 150 percent with additional relief of $129 million Uganda 3 150 150 171 Increase in debt stock from previous year not identified at completion point Source: HIPC decision point and completion point documents; data from the World Bank HIPC Unit. 6.8 The external debt outlook for Bolivia is mixed: the debt-to-export ratio has improved since completion point (with additional debt relief more than offsetting a fall in exports), but the debt service ratio has worsened as a result of new external debt on nonconcessional terms 50. Between 1996 and 1998, exports fell by $28 million, central government fell as a percentage of GDP from 4.3 percentage points, and the Guyanese dollar depreciated from $141 to $165 per U.S. dollar. 42 to cover the budget deficit. For Mauritania, while the debt-to-revenue ratio is likely to be maintained just under the target 250 percent, the debt-to-export ratio has worsened, as has the debt service ratio-a result of higher than anticipated new borrowing and lower exports.s Figuire 6.1: Pirojected Debt-to-Export Ratios (200--10) 3 --0- Bolivia S250 20 --c- Burkina Faso * 200 0- M ~fauritania -;*-Moxambique -100 950 -==Tanzania 0 a e a C=-= Uganda CmCM CM! C~ Figure 6.2: Pirojected Debt Seirvice Ratios (20@@-10) 30 - -Bolivia 25 airtns-s 25- -o3- Burkina Faso 20- 0 -=nMaurtana "15 - -44*-Mozamnbique 10- W-:>:-Tanzania 5. 0 0 , oUganda Source: HIPC unit, Spring 2002 database. 51. Bilateral debt forgiveness above and beyond the HIPC Initiative is taken into account in this calculation. Excluding bilateral debt forgiveness, the debt-to-revenue at completion point would have been 174 percent. 43 6.9 Downturns in export earnings and higher than anticipated new borrowings have significantly undermined the debt sustainability prospects for Uganda and Burkina Faso. Uganda's debt-to-export ratio (based on end-June 1999 data) was brought down to 150 percent at the time of the completion point, but deteriorated soon thereafter. By end-June 2001, the ratio had deteriorated by 43 percent (relative to the E-HIPC decision point projection of 128 percent), a little over half of which is accounted for by lower exports, and the rest by an increase in the NPV of debt through new borrowing. For Burkina Faso, the increase in the NPV of debt (49 percent) more than accounts for the decline in its debt-to- exports indicator (46 percent)-an outcome of a large public investment program. In both cases, however, the debt service ratios remain well within the target range. 6.10 This early experience with countries successfully completing the requirements for both the O-HIPC and E-HIPC highlights the main risk factors for debt sustainability: export performance and the levels and terms of new borrowing. Figure 6.3 summarizes the influence of these factors on the debt sustainability of the countries that have reached their completion point under the E-HIIPC. Quadrant 2 represents a good outcome for debt sustainability, and quadrant 3 a bad outcome. Quadrants 1 and 4 represent mixed outcomes with the net effect depending on the relative contributions of the opposing effects on debt sustainability. As discussed above, Bolivia is a mixed case. Figure 6.3: Key Factors Affecting Debt Sustainability: Status of E-HIPC Completion Point Countries Exports Low High S1 2 o -3 Tanzania Bolivia Mozambique 0 0 6.11 Export performance has been a decisive factor in almost all cases. Tanzania has managed to overcome the reduced earnings brought about by a decline in coffee prices through increased exports of gold-highlighting the critical need for export diversification. Uganda provides a similar example: although it has been hit hard by a 68 percent decline in coffee prices over the past two years, an increase in other exports has helped cushion the impact of this decline. 44 6.12 The NPV of debt is affected by discount rates and currency fluctuations, but the biggest influence is new borrowing. Lower borrowing has improved the prospects for Tanzania and Mozambique, while higher borrowing by the others has made matters worse. And although the sample is small, the findings suggest an association between deteriorating export performance and higher borrowing-doubly affecting debt sustainability. 6.13 Whether and how the HIPC Initiative has affected the level of new borrowing is not known. There is no obvious correlation between new borrowing and the level of HIPC relief (in terms of percentage of pre-HIPC debt stock reduced). The possible relationship between export performance and new borrowing, as also between new borrowing and the objectives of poverty reduction and growth (for instance, Burkina Faso's new borrowing was for infrastructure and poverty alleviation projects), has implications for the macroeconomic framework used to underpin the debt sustainability analysis. As noted in Chapter 5, since the assumptions underlying the models are not known, it is not possible to draw the implications of these findings. The Intepim Countries 6.14 Of the countries past their decision points, 11 (65 percent) show a worsening of their debt ratios." Of the full set of 20 countries, based on available data, the H[PC unit currently expects half to exceed the target for their debt sustainability ratios at their completion point." 6.15 As with the completion point countries, the dominant influence on the economic outcomes of the interim HIPCs in 2000-2001 has been export performance, which has led to a worsening of their debt profiles." Average (unweighted) export growth for HIPCs during 2000-2001 was 4.9 percent, compared with projections averaging 9.0 percent. Still, despite the short-term increase in the NPV of debt-to-export ratios, the current projections show that most countries should fall below the sustainability target over the medium to long term. 6.16 These countries, in the aggregate, have borrowed less (in nominal terms) than anticipated for both 2000 and 2001." This is partly a result of lower disbursements from multilaterals because of interruptions in the PRGF-supported macroeconomic programs in several countries. Some countries increased their debt stocks through higher than anticipated new borrowing, and in some the debt stock data have been revised to fully account for outstanding debt. 52. Updated information on the NPV of debt is not available for 8 of the 17 countries. As new data become available, the prospects for these countries might change. 53. Bank Management notes that this conclusion was based on a rough exercise to estimate the costs of topping- up. Many of the parameters used in the calculation of the NPV of debt were unchanged from the decision point. 54. With some exceptions, export growth was lower than projected for most interim HIPCs. At the same time, countries that had the same or improved debt sustainability indictors also had significantly improved export performance. 55. Benin and Honduras had higher than expected borrowing and projections were broadly in line with actual borrowing in four countries. 45 6.17 The overall pattern emerging for the interim HIPCs is similar to that of the countries that have reached completion point. Borrowing and exports are negatively correlated."6 The conclusions and lessons are also similar. Prospects for Debt Sustainability 6.18 To assess the prospects for future debt sustainability, the appropriate test is the projected trend for the key indicator. The projected paths shown in figures 6.1 and 6.2 are based on export projections and other underlying assumptions in the country DSAs. As discussed in Chapter 5, the projections appear to be optimistic for many countries. For the completion point countries, the realized export growth for 2000-2001 has been in line with or better than the assumed growth rate in the DSAs for three of the six countries, or a 50 percent success rate. These outcomes reiterate the need to better incorporate export volatility in the DSAs. 6.19 Taking the projections at face value, the long-run debt sustainability prospects are tenuous for Bolivia, Mauritania, and Tanzania, with debt-to-exports ratios close to the 150 percent threshold, and poor for Burkina Faso and Uganda, which remain at around 200 percent. The alternative indicator of debt burden, the debt service ratio, indicates that the prospects for the debt service burden are better for all but Bolivia." 6.20 The flat profile for most countries highlights the importance of stabilizing and improving export performance and closely monitoring new borrowing to ensure long-run debt sustainability. The profile, however, is not influenced by the level at which the threshold for sustainability is set: the level of the debt-to-exports threshold provides a benchmark, but the profile endures irrespective of where the benchmark is set. 6.21 A critical element for improving the prospects for debt sustainability, and for poverty reduction, is a credible strategy for growth. As the Initiative is currently being implemented, there is little emphasis on growth other than the maintenance of macroeconomic stability and the development of human capital. Other factors affecting growth, such as the investment climate, infrastructure development, and economic productivity-for example in agriculture-have received little attention in most countries. 6.22 Additional borrowing may be essential for most, if not all HIPCs, to improve the prospects for growth and poverty reduction (see Elbadawi and Gelb 2001, for an assessment of the financing needs for Africa). The need for additional financing posits a potential conflict with the goal of debt sustainability. Replacing loans with grants would help avoid this conflict, but the prospects for this happening at the scale necessary are limited for the near future." But 56. The one exception is Nicaragua, where new borrowing was overestimated at decision point. 57. The fiscal indicator, debt service-to-revenue, also shows that, with the exception of Bolivia and possibly Mauritania, all countries would reach 10 percent or lower in the next two to three years. 58. The expansion in the use of grants to assist debt-vulnerable countries in the IDA13 replenishment will to some extent help in addressing the conflict between the need for new resources and maintaining debt sustainability. The share of IDA13 financing to be delivered in the form of grants over the next three years is expected to be between 18-21 percent. Of this, a maximum of 44 percent (amounting to 8 percent of total 46 as in the case of Burkina Faso (see box 6.1), in Box 6.1: Topping-Up and the addition to monitoring the levels and terms of the Prospects for Debt Sustainability for new borrowing, it is important also to ensure that Burkina Faso new borrowing is put to productive use to enhance At completion point, Burkina Faso had the the country's repayment capacity through growth highest debt-to-export ratios of all the and for poverty reduction. completion point countries and had the longest elapsed time to complete the 0- Progress Toward Performance Criteria HIPC (34 months) and the E-HIPC (21 months). A large public investment 6.23 This section reviews countries' progress program results in a hump in the debt toward performance criteria that are critical to the stock and keeps the debt sustainability achievement of HIPC objectives. Among these is ratio persistently high throughout the next the development and implementation of a decade. Topping up of NPV $129 million reduces the debt hump in the short run, but participatory PRSP. It is too early to comment on does not appear to make a great difference the outcomes from PRSPs, as many are still under in long-term debt sustainability figures. In preparation or have only recently been completed, order to achieve and maintain debt but the issues of ownership and balance on sustainability, Burkina Faso needs to development priorities (both identified in the preserve macroeconomic stability and previous chapter) are sources of risk to achieving continue with structural reforms to HIPC objectives. Assessing the implications of diversify its economy, stabilize economic these and other issues related to the PRSPs growth and enhance exports. requires a separate in-depth review, which is currently being conducted by OED. This section reviews the progress on policy performance and the performance benchmarks included as completion point conditions and assesses their efficacy in contributing to the Initiative's objectives. 6.24 Policy performance: From the start, the HIPC Initiative has required beneficiary countries to have and maintain strong policies to promote economic growth and poverty reduction. As noted in the previous chapter, the ex ante performance criteria were relaxed to allow more countries to enter the program by the end of 2000. While many stakeholders view this flexibility as a good thing, it also implies that to reach the completion point, these countries would face a tougher challenge than countries that were already implementing the necessary policy and structural reforms. 6.25 The group of countries that completed the O-HIPC program had a strong track record of policy performance going into the program, and all but one maintained good performance to qualify for the E-HIPC. Guyana was the exception; it qualified for the E-HIPC but had a mixed record just prior to the decision point. The countries that reached the E-HIPC completion point similarly had a strong record prior to entering the program and maintained their performance to completion point." Since the E-HIPC completion point, policy IDA13 funds) can be allocated as grants to assist debt-vulnerable countries with per capita GNP of less than $360 (and NPV of debt-to-exports above 150 percent). Individually, these countries can get up to 40 percent of their IDA allocation in the form of grants. On average, with annual disbursements IDA of about $7.7 billion for the next three years, this amounts to about $600 million annually. 59. All countries met the macroeconomic performance criteria under the IMF program. 47 performance has been broadly satisfactory for four of the six countries. Burkina Faso is also largely on track, but revenue collection remains poor, and Bolivia has experienced policy slippages and is operating under a shadow program. 6.26 Several of the 20 current interim countries have weaker policy performance than the completion point countries. By mid-2002, half of them had experienced policy slippages and subsequent interruptions in their IMF program. As shown in table 6.3, of these 10 countries, 9 had reached decision point in the second half of 2000, and 8 had reached it in November or December. Most had a previous track record of variable performance. Thus it appears that the policy track record prior to decision point is a reasonable predictor of subsequent performance. Table 6.3: Countries with Difficulties Staying on Track with Macroeconomic and Structural Programs in 2001 NPV of debt-to-export at CP likely to exceed 150 Country Decision Point Mixed pre-HIPC track record noted percent? Guinea December 2000 Difficulties 10/99-3/00, but back on track by No end-September 2000. Guinea-Bissau December 2000 Satisfactory until 1998 civil war, which lasted No over a year. Guyana November 2000 Mixed track record post-1997, including No delays in IMF review. Honduras July 2000 Broadly satisfactory performance. No Malawi December 2000 Mixed performance, especially on Yes (anticipated at DP) expenditure control. Nicaragua December 2000 Slippages in PRGF occurred in early 2000. No Niger December 2000 Satisfactory until the 1999 coup. New Yes (anticipated at DP) govemment in 2000. Rwanda December 2000 Broadly satisfactory but delays in Yes (anticipated at DP) implementing agreed policies and shortfalls in meeting targets noted. Sa6 Tom6 and December 2000 Track record under PRGF and IDA No Principe programs equal 1 year as of 12/00. Senegal June 2000 Favorable track record post-1994 Yes devaluation - Source: HIPC decision point documents, World Bank and IMF 2002d * After additional bilateral support. 6.27 Another measure of countries' policy performance is the CPIA (Country Policy and Institutional Assessment) index. Some of the countries had worse policies in 2001, as measured by the CPIA, than they did prior to the launch of the HIPC. Figure 6.4 shows that the average CPIA score for the early entrants was high in 1998 (before entering the HIPC program) relative to the other HIPCs, and has since continued to improve. The score for the 13 countries that qualified during "the millennium rush" improved slightly until 2000, but slipped back to the 1998 level in 2001 (after the decision point). The four late entrants show a deteriorating trend. 48 6.28 Analysis of the four main Fcigure 6.4: Average Poicy PMerformane components of the CPIA index shows Scores of Early Enants the "Miternut Rush" that for HIPC countries the quality of Group, and Late Entrants economic management and social 4.00 policies has improved, but structural policies have slightly deteriorated. The performance on public sector a u Jul management is roughly unchanged. 8 a.so i 00 0200 These scores reflect the emphasis on . -0-- Aug-Dec macroeconomic policy and 2000 management, through the PRGF, and 3.00 i -After Dec the increasing focus on social sector 2000 policy and performance in recent years. They also indicate a need for 2.50 additional focus on structural 1998 1999 2000 2001 policies. Source: World Bank (PREM Network) data. 6.29 As indicated in the previous chapter, the "millennium rush" group faces a tougher challenge in achieving desired development results. Current evidence suggests that these countries are indeed at greater risk in their efforts to achieve the HIPC objectives. On average, despite weaker policy frameworks and lower economic performance in terms of real GDP growth (2.6 percent through the 1990s) relative to the early entrants (4.4 percent), the assumed real GDP growth rates in their DSAs were significantly more optimistic.' The actual performance in 2000- 2001 has not lived up to these expectations: the early entrants have performed better (4.8 percent) than the millennium rush group (3.8 percent), with both groups growing at a lower rate than assumed. The millennium rush group is also more likely to have deterioration in the key debt sustainability indicator, the debt-to-export ratio, as they are more likely to experience export under-performance. The new borrowing-to-GDP ratio is also higher for these countries and the fiscal deficit worse than for the early entrants. 6.30 Progress on other performance benchmarks: In addition to the requirement for strong policies, the Initiative includes other performance benchmarks that countries must meet to reach their completion points. As discussed in Chapter 5, the most common of these are targets for expenditures--often including specific conditions on the use of HIPC relief-on public sector education and health programs. HIPC progress reports regularly provide updates on the progress in social expenditures. Given the poor state of monitoring and budget systems in many of the HIPCs, significant effort is being made to improve them, particularly for public expenditure management. Even if the expenditure or other input-oriented targets are met, they may not necessarily ensure demonstrable outcomes toward the Initiative's objectives. Three issues are important here: (a) the totality of expenditures in individual sectors, (b) the tracking of expenditures beyond the central level and, more important, the tracking of outcomes, and (c) the balance among development priorities. 60. The ratio of the assumed real GDP growth rates for 2000-2005 to the historical 1990-2000 actual growth rates was 1.4 for the early entrants but 2.4 for the millennium rush group. 49 6.31 On the first issue-the totality of public expenditures-the evidence from HIPC documents and the country case studies shows a significant increase in social sector expenditures, but the extent to which this represents a change in totality of resources is not known. As mentioned in Chapter 5, a substantial share of donor assistance and government revenues is not captured in the budget. The Development Assistance Committee (DAC) data show that social sectors have received a major share of the sector allocable donor assistance (rising from about a quarter in 1990 to over one-half in 2000, with an almost corresponding decline in aid allocation for "production services"). To the extent that some of the donor assistance is being shifted from project to budget support (as in Uganda), the increases in recorded budget expenditures may not represent an actual increase in total expenditures. Also, to the extent that the increase in government budgets as a result of HIPC assistance is offset by a corresponding decline in other donor assistance, the net effect on specific sectoral expenditures may be less. 6.32 On the second issue of expenditure tracking, in principle the countries track overall public expenditures for poverty reduction. In most countries, these are essentially social expenditures, and mostly in the form of financing for health and education. The IFIs also regularly track estimates of social expenditures in HIPC progress reports. Reporting on pro- poor spending in reviews of PRGF programs also places emphasis on social expenditures (see, for instance, Gupta and others 2002). 6.33 The country case studies show that processes are in place for tracking debt relief inflows and for monitoring the use of HIPC expenditures at the central level (Annex 1). Some countries also have transparent dissemination of this information, as in Zambia, but for most the tracking of expenditures at the lower levels is still lacking."' Furthermore, although monitoring the proper use of resources is important, it is not a good indicator of outcomes. So far, there is little evidence on outcomes or results to determine what the increased expenditures are achieving. There are some exceptions to this, with Uganda as an example. Surveys are being undertaken there to track access to and quality of public services. The outcomes from Uganda show a marked improvement in service delivery, but also highlight the need for improving the efficiency of social expenditures. A similar survey for Ghana reveals substantial leakage of funds allocated to primary health and education facilities.` Surveys in some other countries have recently been conducted or are planned, but their results are not yet available. Efforts are also being made to rely on civil society groups for monitoring at the local level, as in Malawi. Early evidence from Malawi shows some impact-but less than expected. 6.34 Finally, perhaps reflecting the recent increased emphasis on social spending by donors and the IFIs, HIPC progress reports note that over half of government revenues will be earmarked for social expenditures in the coming years. Paradoxically, most recipients consider the focus of the Initiative to be excessive on social sectors, and too little on growth and "wealth creation" (Annex G). Considering that many PRSPs are yet to be completed, 61. This problem, however, is not specific to HIPC and is part of the broader set of problems with public expenditure management in most countries. 62. The Ghana Public Expenditure Tracking Survey found that only two-thirds of the funds for recurrent expenditures reach their intended destinations (primary schools and primary health care clinics). 50 these anticipated allocations are also inconsistent with the intended role of the PRSPs. Both these issues question the degree of ownership of the HIPC-related programs and activities. 6.35 The inflexibility in the use of HIPC resources for building essential infrastructure and for economic services, even when there are known absorptive capacity constraints in the social sectors, is viewed by debtor representatives as an inefficient use of scare resources. The conditions attached to HIPC are viewed as inadequately attuned to the holistic development goals of the HIPC governments or to enhancing the "post-HIPC prospects" to ensure debt sustainability and poverty reduction. Creditor country representatives were also critical of the lack of focus on growth opportunities (Annex H). While welcoming the focus on poverty reduction, a strong consensus among the debtor representatives consulted for this review was that the potential gains from their participation in the HIPC Initiative-long-term debt sustainability and "breathing space" for increasing social expenditures-could not be sustained without increased economic growth. Incremental Resource Tiransfer 6.36 This section discusses the implications of the Initiative on the overall resource transfers to the HIPCs. It first discusses the financial impact of HIPC assistance on the beneficiary countries, and then looks at the overall resource flows to these and other countries, including the likely distributional implications for other poor countries. 6.37 HIPC assistance: The HIPC progress reports provide a detailed account on the delivery of HIPC assistance to the beneficiary countries, including those that have reached their completion points and those receiving "interim" assistance (as in World Bank and IMF 2002d). They regularly report on the key indicators related to debt, including the level of the debt stock, and the debt service burden as reflected in the ratio of debt service to exports, GDP, and government revenues. Only the broad aggregates are discussed here. 6.38 Relief provided by the HIPC Initiative alone will cut the debt of the 26 decision point countries in present value by half, from $52 billion after traditional relief to $27 billion before additional bilateral assistance. When additional relief from major bilateral donors is added, the total debt is expected to be reduced by two-thirds, from $62 billion to $22 billion (in present value terms). A similar pattern of benefit emerges from the other debt indicators. Total debt service as a percentage of exports is expected to be halved, from 16.5 percent pre-HIPC to 8 percent in 2001-2005, compared with 20 percent for other developing countries.' Debt service-to-GDP is expected to be cut from 4 percent to 2 percent, and debt service-to- government revenues is expected to decline from about 24 percent to an average of around 10 percent by 2005." At the same time, social sector expenditures as a percent of GDP are expected to rise to 10 percent (about five times debt service, compared with two times before 63. The envisaged decline in debt service in 2005 (that is, with the full impact of HIPC relief) as a ratio of current (2000) exports would decline to under 15 percent for all but countries but Bolivia and Zambia, with an average of about 10 percent. For most countries, it would be at or below 10 percent. The ratio for Zambia is expected to fall after 2005, and for Bolivia some years later. 64. The debt service envisaged by 2005 (that is, with the full impact of HIPC) would be equivalent to 14.5 percent of the current (2000) revenues. 51 HIPC), or over half of the average government's revenues (up from a third before HIPC). Total savings to HIPC governments are expected to be about $41.5 billion over the next 30 years or so. Annually, HIPC debt relief amounts to about $1.3 billion. 6.39 The direct financial impact of HIPC assistance is already being felt, not only in the countries that have reached their completion point, but also in those that have reached their decision point. The impact, however, varies across countries, depending on their particular circumstances. The interim relief to the countries past their decision points is broadly comparable to the full debt relief expected after the completion point for all but three countries (Chad, The Gambia, Niger). Some have experienced a curtailment of interim relief from the IMF (or have yet to receive interim relief) because of policy slippages. IDA has provided interim relief to all of the HIPCs that have passed their decision point, even in the context of an interruption in the IMF program. 6.40 These estimates of the financial impact are based on the assumption of full participation of all creditors. Although the majority of creditors are participating or have given assurances of participation, even for the completion point countries the delivery of assistance is not 100 percent of the amount promised by the HIPC Initiative; assurances range from 80 percent to 96 percent. For countries in the interim period, the range of assurances given is for similar magnitudes of the expected relief. Finally, the share of the non-cooperating commercial and non-Paris Club bilateral creditors in the total debt is small, but an important concern of debtors is that the potential reward from litigation by these creditors to recover past loans could result in much larger and immediate payouts." 6.41 Ov'erall resource transfers: As noted in Chapter 4, the Initiative will increase net resource transfers to the HIPC countries only if other aid flows (including grants and - concessional loans) do not fall dollar-for-dollar with the reduction in debt-service obligations resulting from current debt relief. This additionality of resources made available by debt relief is critical to simultaneously achieve all of the Initiative's multiple objectives. Without additionality, it is not apparent that the fiscal space needed for the mandated social and other expenditures would be created. And, as also noted in Chapter 4, additionality is assumed by the Initiative, but there is no mechanism by which it can ensure that this indeed takes place. 6.42 It is generally difficult to determine whether HIPC has generated incremental resource transfers, because there is no rigorous way to know what would have happened in the absence of debt relief. One option is to use the projected flows used for the HIPC DSAs. These projections are based on a survey of the donor and lender community to assess their anticipated transfers. This provides one basis for establishing a counterfactual, and is discussed below. But it is unclear how these projections by the donor/lenders have taken the Initiative into account. Three analytical possibilities arise: The donors/lenders provide the same gross future amounts irrespective of the Initiative. In this case the reduction in debt-service obligations brought about by the Initiative is fully additional. 65. As noted in footnote 48, however, so far the litigating creditors with court judgments in their favor have not been successful in converting the judgments into payments by the HIPCs. 52 o Donors/lenders do not anticipate the Initiative and decide to cut back future gross transfers to maintain net resource transfers. In this case there would be no additionality, but a comparison of the flows immediately before and after the agreement to deliver HIPC relief would reveal a change in gross flows with no change in net transfers. o Donors/lenders already anticipated the impact of the Initiative early on by cutting back on their gross transfers by their anticipated obligations under the Initiative. In this case the before-after comparisons of net transfers may indicate additionality, but net transfers would already have fallen previously. The net result would be little or no additionality, and could even be a decline in overall net transfers. 6.43 The analytical difficulty in assessing additionality is exacerbated by the complexity and poor quality of the data on international aid flows (Birdsall and Williamson 2002; Renard and Cassimon 2001). Recognizing that the full impact of the HIPC Initiative in terms of additionality will only become clear in the coming years, an attempt is made below to identify any discernible trends in the recent aggregate aid flows." The two sources of these data are the DAC reporting system, which gives the donors' perspective on aid flows, and the Debtor Reporting System (DRS) of the World Bank, which reflects the debtors' perspective. Both sources show similar trends in net resource transfers to the HIPCs.' As discussed in Chapter 4, it is important to consider the changes in net resource transfers both at the global level-because of distributional implications-and at the level of the individual HIPCs. 6.44 At the global level, net resource transfers to the HIPCs rose during the first half of the 1990s, but then sharply turned downwards as shown in figure 6.5. The net transfers averaged about $14.2 billion from 1990 to 1995 and about $11.9 billion from 1996 to 2000. A similar but sharper decline is observed in figure 6.6 for all developing countries." The figures also show some stabilization after 1997, but a decline again in 2000. 6.45 Some further analysis shows that between 1990 and 1995, net transfers to all developing countries grew at a rate of 2.3 percent per year, but declined at a rate of -13 percent from 1995 onward. For HIPCs the corresponding growth rates are 0.5 percent before 1995 and -3.7 percent after 1995. 6.46 It should be noted that the impact of interim debt relief has yet to be fully captured in the DRS or the DAC data. However, even if the entire $1.3 billion to be delivered in annual debt service relief (noted above) to the 26 HIPCs were to be included, it would not offset the aggregate decline in net transfers of $14 billion, including the over $2 billion decline to all HIPCs combined. 66. A majority of the countries that have qualified for HIPC debt relief reached their decision points in 2000, whereas the data on financial flows are currently available only until 2000. 67. Some of the similarity in trends in the DAC and the DRS data is a result of both systems using the same source of data for multilateral flows (directly from the multilateral institutions). 68. The group of developing countries includes low- and lower-middle income Part I countries as defined by DAC (that is, excluding countries in transition), but excludes the three East Asian financial crisis countries in the group to remove large movements in 1997-2000. 53 6.47 Out of these shrinking transfers, the share of the 26 Figure 6.5: Declining Aggregate Net Transfers to HIPCs countries that have reached 17 . their decision points sharply increased after 1998, as shown 1990-95 trend in figure 6.7.' The share of 15- the remaining HIPCs has also 14 increased since 1998, but not 3 .2 13- as sharply. Thus there appears g to have been redistribution 12 from non-HIPCs to HIPCs 11 . since 1998. This shift has 1995-2000 trend important implications for the 10- principle of selectivity and 9- effectiveness of aid. 8 Considering that the HlPCs 1990 1991 1992 1993 1994 1995 1996 1997 1998 1999 2000 have weaker policy Source: OECD, DAC database. frameworks, as judged by the group-average CPIA index, than Figure 6.6: Sharply Declining Aggregate Net other poor countries (non-HIPC Transfers to All Developing Countries (excluding IDA countries as well as Asian crisis countries) non-HIPC lower-middle- so income countries), this 45 redistribution conflicts with 40 - 990-5 trnd the principle of performance- based allocation and is likely 35- to reduce the overall C so efficiency of development ai0 -0 25- aid.o as : 20 6.48 These findings 15 1995-2000 trend suggest that although the HIPCs as a group are getting 10 an increasing share of 5 - declining global aid 1990 1991 1992 1993 1994 1995 1996 1997 1998 1999 2000 resources, they are not Source: OECD, DAC database. receiving additional funds in absolute terms relative to what they were receiving before the Initiative was created (that is, before 1996). While the reasons behind the sharp decline since 1995 are not known, because the resource transfers were depressed in the years immediately preceding most countries' 69. The shares were roughly constant, with a flat trend, between 1990 and 1998. 70. The average 1998 CPIA score was 3.10 for HIPCs, 3.30 for non-HIPC IDA countries, and 3.42 for non- HIPC lower-middle-income countries. The corresponding averages for 2000 are 3.07, 3.25, and 3.54, respectively. 54 Figure 6.7: HIfCs' Growing Share of Aggregate Net Resouirce Transfers 100%- Other 90% developing 80%- countries 70%- Remaining 60%. HIPCS 50%- 40% - 30%- ~Decision point HIPCs 20%- 10%_ 0%_1 11 1990 1991 1992 1993 1994 1995 1998 1997 1998 1999 2000 Source: OECD, DAC database. qualification for debt relief, these years are not the appropriate benchmark for establishing additionality. 6.49 At the country level, the data from the spring 2002 HIPC database corroborate the decline of net transfers in 2000 shown above. Although the full impact of the HIPC Initiative may not be best measured by looking at the data for the year immediately before and after the delivery of relief, because of the bulk of the countries reached their E-HIPC decision point within about a year of each other it may be possible to corroborate the decline observed in the aggregate data from those at the country level. As noted above, the projected transfers from the decision point documents, based on a survey of donors, provide a benchmark for assessing the subsequent actual transfers in 2000. A decline in actual net transfers in 2000 is consistent for decision point countries as well as countries that have reached completion point: net transfers were about 19 percent lower than anticipated for the decision point countries taken together, and 18 percent lower for the completion point countries. The DAC data also shows a decline of about 14 percent for all 26 HIPCs. 6.50 The change in transfers, however, varies by country. In Uganda and Cameroon, HIPC relief resources have been additional, as donors have continued to provide assistance in the light of their sustained policy and reform efforts. Two cases, by contrast, show the opposite trend. In Malawi, even prior to deterioration in the country's policy performance in 2001, non-ODA and, to some extent, ODA loans had declined as some donors adopted a no-loans policy toward HIPCs. While this is good for debt sustainability, it also reduces the overall transfers of aid, as the decline in loans has not been fully replaced by grants. The recent withdrawal of budget support by most creditors because of concerns with governance and accountability has exacerbated the decline in development financing to Malawi. Given that donors have funded a substantial part of the Malawian budget in recent years, this may reflect increasing selectivity as the donors feel less pressure to continue providing funding (for example, to meet debt service obligations, among other things). While, again, this may be 55 viewed as a positive outcome by some from a development effectiveness perspective, it also implies reduced additionality. 6.51 The second case is Zambia, where balance of payments support declined considerably after the decision point, reflecting donors' concerns about governance. Beyond these specific country cases, additionality varies across HIPCs. The latest HIPC implementation report (July 2002) notes that inflows to the decision point HIPCs have increased in 2001 over the average of 1998-2000, but that the situation varies by country. The full impact of HIPC in terms of additionality will only become clear in the coming years. Debt Management Capacity 6.52 As noted in chapter 5, a number of donors and organizations are devoted to improving debt management in the HIPCs through technical assistance and training. While most agencies note that their programs are having a positive impact, the extent to which most groups monitor or are able to measure outcomes varies considerably but is generally limited. Beyond qualitative accounts, evidence of measurable improvements in the debt management capacity of the HIPCs is scarce. 6.53 A recent self-assessment survey on debt management in 33 IPC countries sponsored by the World Bank and IMF found that the countries that have already passed completion point appear to have made more progress in debt management capacity than the other HIPCs (World Bank and IF 2002c)." The need to strengthen institutional capacity (both the legal framework and the organization and staffing of the debt management function) remains the core challenge, which is also dependent on political support for making necessary changes. The World BanklRMF survey of debt management officials also found that transparency and accountability related to new borrowing were weak across countries. The capacity constraints were identified in three main areas: (i) institutional and political constraints, (ii) human resource constraints; and (iii) coordination between debt and macroeconomic policies. Even in cases where staffing levels are adequate (60 percent), a vast majority of the respondents (75 percent) judged staff skills to be only rudimentary. 6.54 The survey found that there is an ample supply of training and capacity building activities to the HIPCs, but the lack of coordination among agencies providing the services is reducing the overall efficiency of the efforts." OED consultations with regional providers of debt management training indicates that the rationality of capacity building for debt management (that is, ensuring that needed services are being offered by the organizations best able to supply them) has suffered from the lack of information-sharing across external training providers. Going forward, there is a need for more attention to debt-management capacity building as part of the HIPC process, as well as more coordinated and targeted technical assistance across providers. 71. The report underscores that these countries were among the better performing HIPCs and that findings are based on a self-assessment. 72. See paras. 5 40-5.42 and box 5.3 on the various capacity building efforts currently under way. 56 Likely Outcomes 6.55 Achieving debt sustainability remains a challenge for the HIPCs-both because they are likely to continue to need to borrow to meet their developmental challenges and because their export base is concentrated and their earnings too volatile. In this context, replacing grants with loans would further the debt sustainability objective, and avoid potential conflicts with the other objectives that require additional resources. But at present the prospects for this happening on the scale necessary in the near future are limited. The debate on the indicators and levels does not change the underlying prospects for debt sustainability: it merely changes the benchmark against which debt may be considered sustainable. For example, even if the target debt-to-exports ratio were reduced to 100 percent, export volatility or the need for new borrowing would remain, and the sustainability profile would look the same. Relative to the new definition of "sustainability," most of these countries would still not have achieved debt sustainability. Lowering the threshold will deliver more debt relief (although not necessarily more resources) and may be a goal in itself, but that is independent of the achievement of sustainability. 6.56 While the export projections need to be made more realistic, the pressing issue remains increasing and stabilizing export earnings through diversified exports. This in turn may require improved access to developed country markets. Most countries are likely to continue to need new borrowing to meet their development needs, and hence incur new debt. Their main challenge is to ensure that the funds are invested productively and efficiently to promote repayment capacity. To do so requires actions to improve public expenditure and debt management, improve the investment climate, and make the necessary investments aimed at accelerating and sustaining overall economic growth. These central issues need to be given more attention. 6.57 A necessary condition for economic growth is the adoption of a sound policy framework. Many of the HIPCs have not demonstrated this track record, which raises concerns about the achievement of the HIPC objectives-not only of growth and poverty reduction, but also of debt sustainability, which requires revenue generation. 6.58 The performance criteria are heavily weighted in favor of the social sectors, and more specifically on social expenditures, with little focus on outcomes or impacts. The current balance between growth-enhancing public expenditures and social expenditures is viewed by the debtor governments as undermining the achievement of the main HIPC objectives. 6.59 A necessary (though not sufficient) requirement to simultaneously achieve the Initiative's multiple objectives-specifically to free up resources for increased social sector spending-is an increase in net resource transfers, by at least the amount of the promised HIPC relief. The recent trends in net transfers, however, show a marked decline in flows to the HIPCs, as well as to all poor countries in the aggregate. Moreover, the data show a marked shift in the distribution of shrinking aid resources away from other poor countries, and in favor of the HIPCs. 6.60 In light of the limited instruments at Initiative's disposal, and recognizing that it is an important but only a small part of the overall development framework, it is important to 57 clarify: (a) what are the principal objectives of the Initiative, (b) what are the specific means to achieving those objectives, and (c) how can those means be ensured? This would help in better assessing the Initiative's performance relative to its goals and the expectations they help to create. 58 7. Findinsand Reconendaion 7.1 This chapter summarizes the main findings from this review in relation to the main evaluative questions laid out in Chapter 1: o Relevance: Is the HIPC Initiative's design adequate and appropriate to achieve its stated objectives and intended outcomes? o Efficacy and efficiency: Based on the experience so far, is the HIPC Initiative achieving or likely to achieve its objectives and achieve them efficiently? o Sustainability: How resilient to risks are the expected outcomes of the HIPC Initiative? o Institutional development: To what extent does the design of the Initiative help build country capacity to ensure that the HIPC objectives are achieved and can be sustained? 7.2 Relevance: The core purpose of the Initiative, to reduce the high debt levels of HIPCs, is highly relevant from political economy perspectives as well as from economic and aid effectiveness perspectives. This is evidenced to by the wide support for HIPC in all major international forums, from G-7 meetings to the Monterrey consensus. But the HIPC Initiative has acquired multiple objectives, while the instruments at its disposal have remained the same. The objective of promoting growth by removing the debt overhang has been maintained, the expectations of what it can deliver on debt sustainability have increased, and creating fiscal space for increased social expenditures has been added as an explicit objective. The latter two expansions in objectives are discussed in turn. 7.3 (a) Debt sustainability: The objective related to debt sustainability became more ambitious-a result of political pressures-fueling expectations of what the Initiative can realistically achieve. The main objective of the original HIPC Initiative was to reduce debt to a sustainable level as part of a strategy to achieve debt sustainability. Over time, this objective was transformed to the stated E-HIPC objective of assuring a "permanent" exit from debt rescheduling. 7.4 The notion of "debt sustainability" has been contentious, with controversy about how to measure it and how to "ensure" it. The review concludes that while not perfect, the main indicator used in the Initiative, the NPV of debt-to-export ratio, is operationally preferable to alternative indicators for practical reasons. The current threshold level also appears to be reasonable in comparison with the debt levels of non-highly indebted poor countries. The level of the threshold determines the amount of debt relief a country may be eligible for, but achieving the threshold does not by itself ensure long-term debt sustainability. The latter is assessed through debt sustainability analysis, the robustness of which has not yet been convincingly demonstrated. While the use of the debt-inventory methodology for assessing the current level of debt is a positive feature of the Initiative, the methodological basis underlying the projections offuture levels of debt remain unclear. The lack of transparency of the economic models behind these projections and the overoptimistic growth assumptions have made debt sustainability analyses ambiguous, giving rise to the impression that the process is politically manipulable. The threshold levels are also partly a function of the level of resources available to the Initiative. 59 7.5 A one-time debt reduction is not sufficient to guarantee that countries will be able to avoid future debt problems. The prospects for debt sustainability depend on a number of factors affecting the country's repayment capacity, especially the levels and terms of new borrowings, the productive use of additional resources to generate revenues and promote growth, and export stabilization and diversification. These factors cannot be directly addressed by the Initiative. It would have been more realistic for the Initiative to set the more modest objective of reducing debt to a level that provides countries with a reasonable chance of sustaining their external debts. Even this more modest objective would require full creditor participation to deliver the promised level of relief, and prudent debt management. The Initiative assumes that all creditors will participate, but cannot assure this. And the design of the Initiative does not address the critical issue of capacity building for debt management, assuming instead that efforts outside of the Initiative will fill the necessary gaps. 7.6 Finally, a critical element of strategy to achieve debt sustainability, and for poverty reduction, is a credible strategy for broad-based growth. Here the Initiative's design, specifically the link to the PRSP process, is consistent with lessons of experience on the need for a sound policy framework and the development of a comprehensive poverty reduction strategy that is locally owned. If fully implemented, the PRSP process might foster the rise of stronger civil societies and a have a positive impact on inclusion, transparency, and accountability processes. Whether this will facilitate economic reform efforts that are sufficiently strong to make a major dent in the structural dilemma and promote growth remains to be seen. As the Initiative is being implemented now, there is little emphasis on growth-related activities beyond the establishment of a sound macroeconomic framework and investment in human capital. Factors such as trade access, investment climate, and infrastructure development are critical to promote growth in HIPCs, but have received little attention so far. 7.7 (b) Freeing up resources for poverty reduction. The objectives were expanded in E- HIPC to explicitly target the resources released by debt relief toward higher social expenditures aimed at poverty reduction. The achievement of this objective rests on the key assumption that the debt relief provided will be additional to other aid transfers. This is necessary to "free-up" resources for increased poverty-reducing social spending. The design of the Initiative, however, has no means to ensure that this will in fact happen--debt forgiveness by itself does not guarantee additionality. Without additional resources, it is unclear how the fiscal space is to be created, and in the absence of additionality what the implied tradeoffs are among priority actions for poverty reduction. 7.8 Past debt relief efforts have not been additional. The available data show a substantial decline in net resource transfers after 1995, coinciding with the build-up of the momentum for HIPC debt relief. There also appears to be a marked increase in the share of HIPCs in the overall aid flows. There is not enough evidence yet to definitively determine the full impact of HIPC debt relief, or whether the recent declining trend in net resource transfers has been reversed. 7.9 Further, with limited aid resources, the poverty reduction and debt sustainability objectives may conflict with each other. The decline in loans is good for debt sustainability but it may reduce the availability of overall resources for poverty reduction and growth. 60 Replacing loans with grants would help avoid this conflict between the need for increased real resources and debt sustainability, but the prospects for this happening in the near future are limited. 7.10 Thus, while the Initiative is relevant to the circumstances of the HIPCs, its design is not consistent with the stated objectives. The objectives are over-ambitious and it is not clear how the limitations can be overcome by design improvements. The design would have been more appropriate for a more modest objective---of delivering debt relief to some of the poorest countries. 7.11 Efficacy and Efficiency: If the anticipated debt relief is delivered, the Initiative will succeed in substantially reducing the HIlPCs' external debt stocks, and the level of debt service that the HIPCs are currently paying will fall to levels comparable to, or lower than, what other poor countries are paying. Thus, the Initiative is likely to achieve its original goal of reducing the external debt of the HIPCs and providing them with a "fresh start." 7.12 Notionally reduced debt service provides HIPC governments with the fiscal space to pursue their development priorities, but the actual impact on overall development expenditures, as well as on the budget, will become clear only with time and more detailed data on budgets and overall resource transfers. As noted above, the data for recent years show a substantial decline in net resource transfers to the HIPCs in absolute terms. Further, the distribution of resources in favor of HIPCs conflicts with the principle of performance- based allocation and could reduce the overall efficiency and effectiveness of aid. 7.13 Debt reduction by itself will not necessarily ensure debt sustainability, which depends crucially on growth and export volatility. More realistic growth forecasts and better risk analysis of the projected debt burdens in the debt sustainability analyses would provide a better assessment of each country's likelihood of meeting the Initiative's debt sustainability threshold and would help promote a more informed debate on the policy changes needed in donor and recipient countries. It would also help bring more clarity and realism in the setting of the Initiative's objectives and the needed funding arrangements. 7.14 The Initiative requires countries to have a track record of sustained policy and structural reforms. This requirement has been applied flexibly to bring more countries into the program. But the progressive relaxation of the requirement for the "millennium rush" countries that qualified in late 2000 raises the risk of not achieving the HIPC objectives as these countries face a tougher challenge in meeting the necessary conditions to reach their completion points. And, the majority of the "millennium" rush countries have indeed experienced policy slippages after reaching their decision points. While it is too early to assess the impact of this flexibility on outcomes, other things being equal, these countries are less likely to achieve good development results. 7.15 The findings on the PRSP process-a key performance criterion-are that the participatory process has been satisfactory, but the degree of ownership varies across countries. The growth elements of the strategy are weak. This finding is consistent with the reviews of PRSPs by the World Bank and external partners, which indicate that although growth is universally accorded high priority in the PRSPs, a common weakness is the lack of 61 focus on how the anticipated levels of growth are to be realized and on the prioritization of actions necessary to achieve the key objectives of the strategy. 7.16 The Initiative also set performance benchmarks for public expenditures with an emphasis on social sectors, particularly for health and education. This is evident both in the conditions attached to the use of HIPC resources and on input- and expenditure-related targets for public expenditures as completion point triggers. In principle, the beneficiary countries track overall government spending for poverty reduction. In practice, most countries focus on social expenditures, mostly in the form of health and education expenditures. The IPC progress reports regularly track social expenditures and the PRGF reviews also place emphasis on pro-poor spending. 7.17 Reflecting this emphasis, HIPC progress reports indicate that over half of government revenues will be earmarked for social expenditures in the coming years. Most recipient countries consider this imbalanced and inconsistent with their focus on broader development goals and with the intended role of PRSPs to set priorities. 7.18 This emphasis on social expenditures also has other disadvantages. The performance criteria largely focus on inputs and expenditures rather than intermediate or final outcomes and impacts. The impact of these inputs is likely to be limited by absorptive capacity constraints, the fact that substantial aid resources are already targeted at social sectors, and the uniform application of conditions across countries-even where funding may not be the core constraint to achieving social sector outcomes. The inefficient use of resources in the targeted sectors and the lack of focus on other growth-enhancing and poverty-reducing expenditures are also likely to limit the achievement of the HIPC objectives. 7.19 Sustainability: The biggest risk facing the Initiative is that of unrealistic expectations about what the Initiative can achieve, arising from the potential lack of additionality and overly ambitious interpretations of debt sustainability. Other important sources of risk are the lack of full creditor participation, which could reduce the delivery of anticipated relief; weak capacity for debt management in virtually all HIPCs; and the perceived lack of clarity surrounding the forward-looking aspects of the debt sustainability analyses. 7.20 Long-term debt sustainability requires the development and institutionalization of a credible growth strategy to generate levels of income, job creation, and revenues necessary to attain fiscal sustainability and repayment capacity. A key element of this strategy is stabilizing export earnings by diversifying the export base and gaining market access, with worsening terms of trade and an inequitable international trade regime remaining a significant source of risk. To meet their development challenges, these countries may continue to need to borrow, and the key challenge is to ensure that all resources are used productively and efficiently. The recent progress on increased availability of grants will help in financing development in a sustainable manner, but the amounts of grant aid are still limited and far short of the financing needs of the HIPCs. Beyond official finance, the HIPCs need to create a hospitable environment for private investment through continued and substantial policy and structural reforms, including providing the necessary infrastructure and other services. 62 7.21 Institutional development: The creation of the HIPC Initiative was a major innovation in development finance. It has made the processes in the sovereign debt regime more open and accountable and spurred development cooperation. It has particularly encouraged much better collaboration and coordination between the World Bank and the IMF, both at the operational level and in providing technical advice to governments. While not incorporated into the original design, the most significant institutional development impact of the HIPC Initiative may well be improvements in the HIPCs' public expenditure management systems. 7.22 The HIPC Initiative has been a catalyst for the PRSP process, which holds considerable promise in helping countries improve governance, transparency, and accountability, while promoting ownership of country poverty reduction strategies. The Initiative falls short in capacity building for debt management, which was not addressed in its design. A number of efforts are ongoing, but they remain uncoordinated and inefficient in providing the needed assistance to HIPC governments. Recommendations Recommendation 1: Clarify the objectives and ensure the consistency of the design of the Initiative with the main objectives. The purpose of the Initiative needs to be clarified. Debt relief is a limited instrument; a clear public acknowledgement of this reality and a prioritization of its principal goals and how they are to be achieved in specific contexts would help create correct expectations of what the Initiative can realistically achieve. Specifically, the goal of debt sustainability needs to be clearly distinguished from the goal of poverty alleviation. Both goals are worthy, but require different responses. Increased spending for poverty reduction requires increasing the resources delivered to poor countries (in conjunction with improved policies and the efficiency of resource use). Debt sustainability requires redesigning the way the resources are delivered and, more important, addressing the underlying structural dilemma facing these countries. Management should take steps to focus the Initiative and identify its appropriate role in the overall development framework. Its objectives should be clearly stated, along with a clear articulation of how they are to be achieved and what the roles of different players are in achieving those objectives. Given the publicity surrounding the Initiative, any adjustment of the objectives and design needs to be explained clearly to the global community. Recommendation 2: Improve the transparency of the methodology and economic models underlying the debt projections and the realism of economic growth forecasts in the debt sustainability analyses. The debt projections component of the debt sustainability analyses (DSAs) should be based on a simple, transparent methodological framework. The current lack of clarity surrounding the projections adds to the controversy on the achievement of HIPC objectives and is a significant but unnecessary source of reputational risk for the World Bank. A better analysis of the prospects and risks would also allow more informed decisions on how to modify the current framework to maximize the Initiative's development effectiveness. 63 Recommendation 3: Maintain standards for policy performance. The critical importance of policy performance for debt sustainability, growth, and poverty reduction is widely recognized. For effective implementation, flexibility is often necessary, but there needs to be a clear and transparent rationale for relaxing the criteria or standards for policy performance. Without such a justification, it is critical that standards established for the Initiative be maintained-not only in the interest of fairness and equity, but also to improve the likelihood of achieving the Initiative's objectives and for the durability of results obtained. Recommendation 4: Performance criteria need to increase the focus on pro-poor growth. Growth is critical for both debt sustainability and poverty reduction, but at present the Initiative places a heavy emphasis on social expenditures as the primary means of poverty reduction. The Initiative's performance criteria should be better balanced between growth- enhancing and social expenditure priorities, and tailored to the individual country circumstances. The World Bank should make better use of the knowledge already available and fill the knowledge gaps through additional diagnostic work, particularly on the overall efficiency of public expenditures, identifying sources of growth, and developing appropriate sectoral strategies as the basis of appropriate benchmarks. 65 References References Austrian Development Cooperation. 1999. Uganda's External Debt: Management, Restructuring, and the HIPC Initiatives. A Case Study. Birdsall, Nancy, and John Williamson. 2002. Delivering on Debt Relief: From IMF Gold to a New Aid Architecture. Washington, DC Center for Global Development and the Institute for International Economics. Birdsall, Nancy, Stijn Claessens and Ishac Diwan. 2001. "Will HIPC Matter?: The Debt Game and Donor Behavior in Africa." Presented at WIDER Development Conference on Debt Relief, August 17-18, 2001, Helsinki, Finland. Botchwey, Kwesi, Paul Collier, Jan Willem Gunning and Koichi Hamada. 1998. "External Evaluation of the ESAF: Report by a Group of Independent Experts." Washington, DC: IMF. Brooks, Ray, Mariano Cortes, Francesca Fornasari, Benoit Ketchekmen, Ydahlia Metzgen. 1998. "External Debt Histories of Ten Low-Income Developing Countries: Lessons from Their Experience." IMF Working Paper WP/98/72, Washington, D.C. Callaghy, Thomas, 2002 Innovations in the Sovereign Debt Regime. Background paper. OED, Washington, D.C. Claessens, Stijn, Enrica Detragiache, Ravi Kanbur and Peter Wickham. 1997. "Analytical Aspects of the Debt Problems of Heavily Indebted Poor Countries." In External Finance for Low-Income Countries. Washington, DC: IMF. Cline, William R. 1997. "Debt Relief for HIPCs: Lessons from the Debt Crisis of the 1980s." In External Finance for Low-Income Countries. Washington, DC: IMF. Cohen, Daniel. 1996. "The Sustainability of African Debt." World Bank Policy Research Paper 1621. Washington DC: World Bank. . 1997. "Growth and External Debt: A New Perspective on the African and Latin American Tragedies," Centre for Economic Policy Research Discussion Paper No. 1753. . 2001. "The HIPC Initiative: True and False Promises." International Finance 4 (3): 363-80. Also published by the OECD Development Centre in 2000 as Technical Paper No. 166. Available at http://www.oecd.org/dev/publication/tpla.htm. Collier, Paul, J.W. Gunning and Associates. 2000. Trade Shocks in Developing Countries. Oxford, U.K.; Oxford University Press. Dagdeviren, Hulya and John Weeks. 2001. "How Much Poverty Could HIPC Reduce?" Presented at WIDER Development Conference on Debt Relief, August 17-18, 2001, Helsinki, Finland. Daseking, Christina, and Robert Powell. 1999. "From Toronto Terms to the HIPC Initiative: A Brief History of Debt Relief for Low-Income Countries." IMF Working Paper WP/99/142, Washington, D.C. Deaton, Angus, and R. Miller. 1995. "International Commodity Prices, Macroeconomics Performance, and Politics in Sub-Saharan Africa." In Princeton Studies in International Finance. No. 29. Princeton, N.J.: Princeton University. 66 References Dehn, Jan. 2000. "Commodity Price Uncertainty in Developing Countries." World Bank Policy Research Working Paper No. 2426. Washington, DC: World Bank. Devarajan, Shantayanan, Andrew Sunil Rajkumar, and Vinaya Swaroop 1999. "What Does Aid To Africa Finance?" Development Research Group, The World Bank, Washington, D.C. Development Finance International. 2001. "Debt Relief and Debt Humps." Washington, D.C. Photocopy. DFID (U.K. Department for International Development). 2002. DFID Views on the PRSP Process. Available at http://www.worldbank.org/poverty/strategies/review/dfidl.pdf Dijkstra, Geske, and Niels Hermes. 2001. "The Uncertainty of Debt Service Payments and Economic Growth of HIPCs: Is There a Case for Debt Relief?" Paper presented at a World Institute for Development Economics Research Conference on Debt, Helsinki, August. Easterly, William. 2001a. "Growth Implosions, Debt Explosions and My Aunt Marilyn: Do Growth Slowdowns Cause Public Debt Crises?" The World Bank, Washington, D.C. Policy Research Working Paper 2531. -2001b. The Elusive Quest for Growth: Economists' Adventures and Misadventures in the Tropics. Cambridge, MA: MIT Press. Eaton, Jonathan. 2002. The HIPC Initiative: Multiple Goals, Additionality, Eligibility and Debt Sustainability. Background paper. OED, Washington, D.C. Elbadawi, Ibrahim, Benno J. Ndulu, and Njuguna Ndung'u. 1997. "Debt Overhang and Economic Growth in Sub-Saharan Africa." In External Finance for Low-Income Countries. Washington, DC: IMF. Elbadawi, Ibrahim and Alan Gelb. 2001. "Financing Africa's Development: Toward a Business Plan?" Paper for AERC Policy Seminar, Dar Es Salaam, February 2001. EURODAD (European Network on Debt and Development). 2002. Putting Sustainable Development First. Brussels. GAO (U.S. Government Accounting Office). 2000. Debt Relief Initiative for Poor Countries Faces Challenges. Washington, D.C.: GAO. Gupta, S., M. Plant, B. Clements, T. Dorsey, E. Baldacci, G. Inchauste, S. Tareq, and N. Thacker. 2002. Is the PRGF Living up to Expectations? An Assessment of Program Design. International Monetary Fund Occasional Paper 216, Washington, D.C. Gwin, Catherine. 2002. IDA's Partnership for Poverty Reduction. An Independent Evaluation of Fiscal Years 1994-2000. Operations Evaluation Department, The World Bank, Washington, D.C. Hansen, Henrik. 2001. "The Impact of Aid and External Debt on Growth and Investment: Insights from a Cross-Country Regression Analysis." Paper presented at the WIDER conference on debt relief, Helsinki, August 17-18, 2001. Jubilee 2000 Coalition. 1999. "Unfinished Business-The World's Leaders and the Millennium Debt Challenge." http://www.jugilee2000gk.org/ 67 References Killick, Tony. 1995. "Solving the Multilateral Debt Problem: Reconciling Relief with Acceptability." A Report Prepared for the Commonwealth Secretariat. Overseas Development Institute, London. Krugman, Paul. 1988. "Financing vs. Forgiving a Debt Overhang." Journal of Development Economics, 29: 253-68. Martin, Matthew, with Randa Alami. 2001. "Long-Term Debt Sustainability for HIPCs: How to Respond to Shocks." Debt Relief International, London. Photocopy. Mistry, P.S. 1991. African Debt Revisited: Procrastination or Progress" The Hague, FONDAD. Noorbaksh, F., and A. Paloni. 2001. "Structural Adjustment and Growth in Sub-Saharan Africa: The Importance of Complying with Conditionality." Economic Development and Cultural Change.49 (3) 479-509. OED (Operations Evaluation Department). 1997. Adjustment Lending in Sub-Saharan Africa: An Update. The World Bank, Washington, D.C. . 1999. Higher Impact Adjustment Lending. Volumes 1 and 2. The World Bank, Washington, D.C. Oxfam International. 2001. "Debt Relief: Still Failing the Poor," Washington, D.C. Pattillo, Catherine, Helene Poirson and Luca Ricci. 2002. "External Debt and Growth." IMF Working Paper WP/02/69. An earlier version of the paper was presented at WIDER Development Conference on Debt Relief, August 17-18, 2001, Helsinki, Finland. International Monetary Fund Seminar Series (International) No. 2002-56: 1-47. Renard, Robrecht, and Danny Cassimon. 2001. "On the Pitfalls of Measuring Aid." WIDER discussion paper 69, UNU Wider, Helsinki. Sachs, Jeffrey, 1989. "The Debt Overhang of Developing Countries," in Guillermo A., Calvo and others eds., Debt Stabilization and Development: Essays in Memory of Carlos Dias Alejandro, Basil Blackwell, Oxford, U.K. Sachs, Jeffrey, Kwesi Botchwey, Maciej Cuchra, and Sara Sievers. 1999. "Implementing Debt Relief for the HIPCs." Center for International Policy, Paper No. 2, Harvard University, Cambridge, MA. SPA (Strategic Partnership with Africa). 2002. "SPA 5: 2000-2002. Toward New Aid Relationships to Reduce Poverty," Africa Region, the World Bank, Washington, D.C. Thomas, Melissa A. 2001. "Getting Debt Relief Right," Foreign Affairs 80 (5). Underwood, John. 1990. "The Sustainability of International Debt." Photocopy, The World Bank. United Nations. 2000. Report of the High-Level Panel on Financing for Development (Zedillo Report). http://www.un.org/reports/financing/(June). World Bank. 1994. World Debt Tables 1994-95. Volume I. Washington, D.C. . 2002. Global Development Finance: Financing the Poorest Countries. Washington, D.C. 68 ReTferences World Bank and IMF. 1996. A Program for Action to Resolve the Debt Problems of the Heavily Indebted Poor Countries-Report of the Managing Director of the IMF and the President of the World Bank to the Interim and Development Committees. Report No. DC/96-17. _ . 1998. The Initiative for Heavily Indebted Poor Countries: Review and Outlook. Report No. DC/98-15, Washington, D.C.: The World Bank. Available at http://www.worldbank.org/hipc/hipc-review/hipc-review-and-outlook.pdf ___ . 1999. Modifications of the Heavily Indebted Poor Countries (HIPC) Initiative. Available at http://www.worldbank.org/hipc/hipc-review/ModificationsDC99- 25E.pdf __ . 2001a. Tracking of Poverty-Reducing Public Spending in Heavily Indebted Poor Countries (HIPCs). Available at http://www.worldbank.org/hipclhipc- review/tracking.pdf . 2001b. Guidelines for Debt Public Management. Washington, D.C. . 2002a. Actions to Strengthen the Tracking of Poverty-Reducing Public Spending in Heavily Indebted Poor Countries (HIPCs). Available at http://www.worldbank.org/hipc/hipc-review/tracking.pdf __ . 2002b. Review of the Poverty Reduction Strategy Paper (PRSP) Approach: Main Findings. Available at http://www.worldbank.org/poverty/strategies/review/fmdings.pdf __ . 2002c. External Debt Management in Heavily Indebted Poor Countries. Available at http://www.worldbank.org/hipc/Debt_Management.0302.pdf _ . 2002d. "Heavily Indebted Poor Countries (HIPC) - Status of Implementation" September 29, 2002. Available at http://www.worldbank.org/hipc/ __ . 2002e. "The Enhanced HIPC Initiative and the Achievement of Long-Term Debt Sustainability. " Washington, D.C. Available at http://www.worldbank.org/hipc/Long-Term.pdf. 69 Annex A Annex A: HIPC Initiative: Status of Country Cases Considered under the Initiative, July 2002 Target Assistance Levels 11 Estimated total NPV of debt-to- (in US $ mhons, present value) Percentage noninal debt Decision Conpletion Export revenue World Reduction service relief Country point point (percent) Total Bilateral Multilateral IMF Bank In NPV of debt (in US $ millions) Completion point reached under enhanced framework Bobvia 1.302 425 876 84 194 2,060 Onginal framework Sep. 97 Sep 98 225 448 157 291 29 54 14 760 Enhanced framework Feb 00 June 01 150 854 268 585 55 140 30 1,300 Burkina Faso 553 83 469 58 231 930 Onginal famework Sep. 97 Jul 00 205 229 32 196 22 91 27 400 Enhanced framework Jul 00 Apr 02 150 195 35 161 22 79 30 300 Topping up Apr 02 150 250 129 16 112 14 61 230 Mauntania Feb. 00 Jun 02 137 622 261 361 47 100 50 1,100 Mozambique 2,023 1,270 753 143 443 4,300 Onginal framework Apr. 98 Jun 99 200 1,717 1,076 641 125 381 63 3,700 Enhanced framework Apr 00 Sep. 01 150 306 194 112 18 62 27 600 Tanzania APr. 00 Nov 01 150 2,026 1,006 1,020 120 695 54 3,000 Uganda 1,003 183 820 160 517 1,950 Onginalframewoir Apr 97 Apr 98 202 347 73 274 69 160 20 650 Enhanced framework Feb 00 May 00 150 656 110 546 91 357 37 1,300 Decision point reached under enhanced framework Benin Jul 00 Floating 150 265 77 189 24 84 31 460 Cameroon Oct 00 Floating 150 1,260 874 324 37 179 27 2,000 Chad May 01 Floating 150 170 35 134 18 68 30 260 Ethiopia Nov 01 Floating 150 1,275 482 763 34 463 47 1.930 Gambia. The Dec 00 Floating 150 67 17 49 2 22 27 90 Ghana Feb 02 Floating 69 250 2,186 1,084 1,102 112 781 56 3,700 Guinea Dec 00 Foating 150 545 215 328 31 152 32 800 Guinea-Bissau Dec 00 Floating 150 416 212 204 12 93 85 790 Guyana 585 220 365 74 68 1,030 Onginal framework Dec. 97 May 99 107 280 256 91 165 35 27 24 440 Enhanced framework Nov 00 Floating 150 250 329 129 200 40 41 40 590 Honduras Jul 00 Floating 110 250 556 215 340 30 98 18 900 Madagascar Dec 00 Floating 150 814 457 357 22 252 40 1,500 Malawi Dec 00 Floating 150 643 163 480 30 331 44 1,000 Mail 523 162 361 58 182 870 Original framework Sep 98 Sep 00 200 121 37 84 14 44 9 220 Enhanced framework Sep 00 Floating 150 401 124 277 44 138 28 650 Nicaragua Dec 00 Floating 150 3,267 2.145 1,123 82 189 72 4.500 Niger Dec 00 Floating 150 521 211 309 28 170 54 900 Rwanda Dec 00 Floating 150 452 56 397 44 228 71 800 Sao Tomd and Pnncipe Dec 00 Floating 150 97 29 68 24 83 200 Senegal Jun 00 FloatIng 133 250 488 193 259 45 124 19 850 Sierra Leone Mar 02 Floating 150 600 268 332 123 122 80 950 Zambia Dec 00 Floating 150 2,499 1,168 1,331 602 493 63 3.850 Decision point reached under original framework CWte d'voire Mar 98 141 280 345 163 182 23 91 6 3/ 800 Total assistance providedfcommitted 25,102 11,675 13,296 2,043 6,394 41,520 Preliminary HIPC document lasued * Congo, Den. Rep of 150 5,773 3.556 2,217 400 700 79 9,800 Source World Bank HIPC Unit Available at httpilwww wordbank.orgiNpc/progress-to- The main original objective of HIPC was to bring each country to a position of lasting debt sustainabilty by the time it reached the completion point Poverty reduction was added as a major objective with the advent of the enhanced HIPC (E- HfPC) in 1999. Most countries believed that achieving both goals seems unlikely. Quite generalized agreement existed that HIPC alone would not provide overall debt sustainability for this group of countries as a whole, although it might make some individual countries sustainable, at least for a while. Less consensus existed, however, on what this means for future development strategy for these countries. Nonetheless, all of the creditors now believe that HIPC was the right thing to do despite worries about "slippery slope" expansion of this form of debt relief to other countries, the risks of moral hazard, and the quality of implementation. Several countries noted that debt relief would leave the larger problems of these countries untouched, resulting in "wasted" debt relief. > There was almost unanimous agreement that expectations about what HIPC can and will achieve are way out of line with what is likely to happen. This poses real dangers for the initiative and for the HffPC countries themselves. The expectations problem exists at both the international level and the domestic levels within HIPC countries. The credibility of HIPC and that of the World Bank and the IMF were seen to be at stake. One creditor linked much of this problem to the very use of the term "debt sustainability," preferring something along the lines of "robust exit" from debt burdens or "dynamic sustainability" to the existing term. This creditor believes that all of the public stakeholders involved in HIPC's emergence in 1996 realized at the time that HIPC would not solve the debt problem, but that this got lost in the process of legitimizing, selling, and defending 6. Unless otherwise indicated, HIPC will be used here to refer to the overall process, including the evolution from the original HIPC in 1996 to the enhanced HIPC (E-HIPC) in 1999. One of the 14 creditor countries interviewed is not a Paris Club creditor. 83 Annex H the Initiative. This country sees HIPC as but one milestone on the road to real debt sustainability, believing that something "beyond HIPC" must emerge, given the limitations it sees in the existing framework. Debt sustainability, presuming that one could actually determine what it entails, is simply too high a bar in practice, if not in rhetoric. > There was general concern that insufficient attention is being paid to growth and how to get it About half of the countries pointed to a real tension in the Initiative between the goals of debt sustainability and poverty reduction. A majority of the creditors agreed that there is at least some tension between debt sustainability-and hence the need for growth-and the new emphasis on poverty reduction. A major creditor noted quite strongly that a concerted focus on growth is still missing, and yet the Boards of the World Bank and the IMF go along as each new HIPC case comes before them because the process is already well under way and out of a desire to be fair to all HIPC countries. This was seen by one major creditor as constituting a sort of herd-like mentality that ignored the central issue of growth. Another major creditor noted that growth is being taken into account, if one takes a "very broad" view of growth. Smaller creditors and a couple of G-7 supporters of HIPC, however, were less concerned about the problem of growth and where it is to come from, seeing a focus on this issue as a tactic to water down HIPC, prevent the reinforcement of it, or weaken its focus on poverty reduction. > The PRSP process was generally seen as a major positive innovation; several creditors pointed out, however, that it did not emerge as a direct result of HIPC, but rather from a general shift in the development discourse that became institutionalized first in HIPC. Most creditors believed that the key change was an emphasis on poverty reduction combined with broad consultations on how to achieve it. This innovation has now led to the extension of the PRSP process to other countries beyond the HIPCs. Many creditors were worried, however, about some key aspects of the process, including the level of actual HIPC/PRSP ownership, the quality of consultations, and the eventual implementation of the strategies. A couple of countries felt that the Initiative, and especially the PRSP process, would help the better-developed HIPCs that have good leadership, but would have significantly less of an impact on others. Implementation and Process issues > A majority of creditors believed that the process of determining debt sustainability needed some rethinking. Several countries went further to say that the methodology remains pretty murky and its analytic base weak. Most creditors agreed, for example, that the growth projections built into DSAs have been generally too optimistic, even before taking various shocks into account. A few creditors mentioned that much of the work on debt sustainability in the early days of the original HIPC had been quite political in nature, with the desire of keeping the scope and cost of the Initiative as limited as possible, and were concerned that the concept was not based on solid analytic footing. One creditor speculated that perhaps the IMF and the World Bank staff were trying to Annex H 84 keep within the parameters established by the major creditors, resulting in overly optimistic DSA projections to keep costs in check. > Most creditors agreed that additionality is a problem, while admitting that it would be hard to measure accurately, much less achieve. Many countries agreed that it was very unclear whether HIPC debt relief is actually additional assistance from official sources as a whole. Individually, only a couple of the creditors could assert with some certainty that their contributions to the HIPC relief efforts were additional to their official development assistance (ODA). Beyond ODA, one country saw HIPC as a negative "seal of approval" that stigmatized these countries in regard to the private markets, making additionality significantly harder to achieve. Another creditor, however, saw it as a positive "seal of approval" once a country reached the completion point, while admitting that this "signaling" effect would take time to have a real positive impact in the private markets. > Tracking was generally seen to be a major problem, both in terms of getting the most from debt relief and being able to defend HfIPC as it exists now by being able to show its concrete results. Most creditors believed, however, that the tracking problem is but one symptom of much larger capacity issues in the HNlPCs. HIPC was widely seen as being responsible for putting the issue of debtor government capacity and public expenditure and debt management even more starkly on the table. One country pointed out that the limited capacity of some debtor countries was even keeping them from a proper evaluation of the benefits and costs of participation in HIPC. Tracking of expenditures is seen as a step in the right direction, but it is still in its infancy. In addition, it still does not begin to assess the actual impact of poverty reduction spending, nor any relationship it might have to growth or debt sustainability. > About half of the countries eipressed concern about what they consider to be the inadequate level of funding for HfIPC, especially given the likely demands of topping-up, and pressures for more of it'- The smaller creditor countries seemed more worried about this issue than others and also felt that they are shouldering a larger share of the burden than other creditors, especially for the Trust Fund. A couple of countries expressed the belief that early HIPC decisions by World Bank and IMF staff had been unduly affected by cost issues, the amount of relief to be given squeezed into expected available resources, thereby slowing up the process and distorting the way the Initiative developed. They believed that the focus was more on cost than on the goal of debt sustainability. > Most creditors agreed that flexibility in the implementation of HIPC is a good thing and has been managed relatively well given the complexity of the process and the quite varied conditions of the H1PC countries. But some concerns were raised. Two major creditors believed, however, that too much flexibility exists, seeing a speed versus quality problem, while one small creditor felt that HIPC's complexity had increased at the expense of its manageability. A couple of creditors believed that more clear and consistent rules 85 Annex H are needed on topping-up and other forms of flexibility. On a related matter, several creditors mentioned that the December 2000 "big push" to get more than 20 countries to the decision point was a necessary political move, but one that set in play procedural problems of quality and equal treatment. On a related point, one creditor felt that there is a tendency to treat politically more Important HIPCs better than others. Most agreed that quality will become even more of an issue as the remaining countries reach their decision point. In short, the cases get tougher and tougher as you go down the list. A couple of creditors were quite concerned about whether the quality of the process would balance out by the time the completion point is reached. One G-7 creditor thought that current levels of conditionality and completion point triggers are too stringent, wanting only "reasonable" ones. > Most countries agreed that HIPC had significantly and positively improved coordination and collaboration between the IMF and the World Bank, and with the debt relief and aid processes more generally. Several countries pointed favorably to the existence and performance of the World Bank and IF HIPC units, while indicating that there is still plenty of room for improvement. > About two-thirds of the countries noted that HIPC had reoriented their ODA activities by focusing them on the PRSP process, and not just for the IPC countries. > Most creditors agreed that the lack of comparable treatment by non-Paris Club countries and private creditors is a real problem because debt sustainability cannot be reached without it. Implications for the future > Most countries acknowledged that HIPC is but one, quite limited, part of the overall development picture for these countries. In short, it is far from being a panacea. Yet this line of argument led to two different views on policy directions for HIPC, one more restrictive, the other more expansive: [1] the more restrictive view, held by more than half of the countries, was that other elements of the overall development picture would take up where HIPC leaves off, thereby eliminating the need to alter HIPC in an effort to achieve more debt relief, and [2] the more expansive, less common view, was that the HIPC process ought to be pushed as far as possible in order to achieve the maximum amount of debt relief, while the other elements of the larger development context are addressed simultaneously in a more aggressive way. The existence of these two viewpoints reflects less unity of opinion among the group of G-7 and other creditors than existed at the time E-HIPC was introduced. > Almost all creditor countries agreed that there should not be a HIPC 3, regardless of existing and future "beyond HIPC" problems and pressures. Despite this view, however, it was generally recognized that debt sustainability is not likely to be reached for an important number of countries, and thus pressure will exist for significant Annex H 86 topping-up and going "beyond HIPC" in some way. One G-7 country said that while it does not favor a HIPC 3 now, it might be willing to discuss one at some future date. Related to this, most countries felt that HIPC should not be extended in time or number of countries covered. A couple, however, did believe that new countries might be added, as has already happened, if the new countries meet existing eligibility criteria. A couple of countries mentioned the possibility of including some of the members of the Commonwealth of Independent States. At the same time, reflecting the views of some middle-income countries, one country expressed concern that other highly indebted countries might be ignored if only existing criteria were used-other IDA or IDA blend countries with real debt problems, for example. > A corresponding, generaly recognized, worry had to do with creating new debt burdens through IFI, donor, and private Rending, despite the existence of borrowing ceilings in PRGFs, with the possible result of creating yet another debt problem down the road. The new emphasis on poverty reduction, and the high level of expectations linked to it, leads to pressure to find new resources to sustain poverty reduction spending beyond the use of HIPC debt relief. Borrowing is one major possibility. Related to this is the issue of the level of concessionality of any new lending and the weight between PRGF, IDA, and grant (whether IDA or other) resource flows. Several creditors worried about continued, and even increased, HIPC dependency on bilateral and multilateral public resources. One major creditor has stopped extending ODA loans to any country that joins the HIPC process, while putting more emphasis on grants. Several countries worried about the impact of HIPC on the future level of IDA resources available to non-HIPC countries. 87 Annex I Annex I: Synthesis of Main Findings from Case Studies 1. As part of this OED review, five country cases studies were conducted to assess the HIPC implementation experience-Cameroon, Guyana, Malawi, Uganda, and Zambia. This synthesis gives the main findings from the five country cases. It puts together the key similarities/contrasts to draw some stylized facts. The focus is on four aspects of country experience with HIPC relief: how the countries became highly indebted; the quality at entry of the HIPC programs; implementation experience; and prospects for debt sustainability. 2. Since this is a small sample, the findings do not provide a robust basis for generalization but do provide valuable insight into the shared experience across this range of countries. The group includes two countries (Uganda and Guyana) that reached the completion point under the O-HIPC and one (Uganda) under the E-IPC; GNP per capita levels range from US$190 to US$760; and their economies are dependent on a range of dominant exports from oil to mining to agriculture. How these countries became highly indebted 3. There are strong similarities among the case study countries regarding the genesis of their debt problems, with the notable exception of Uganda given its unique post-conflict context. Most countries became severely indebted in response to terms of trade shocks and a subsequent decline in revenues, but continued maintenance of over-extended public sectors. Their economies were particularly sensitive to export commodity price fluctuations, with adverse weather conditions also playing an important role in the agriculture-dominated economies. The main problem is the high concentration of export earnings in one or a few natural resource or agricultural commodities (table 1.1). Table 1.1: Debt Levels at Decision Point and Export Concentration Percent of total Percent share Country Decision point NPV of debt Main export product exports of top 3 exports Cameroon 205 percent of exports Oil 27 47 Guyana 469 percent of revenues Sugar 25 61 Malawi 269 percent of exports Tobacco 61 75 Uganda 294 percent of exports Coffee 56 63 Zambia 486 percent of exports Copper 48 60 All 24 HIPCs 39 60 Source: World Bank and IMF 2002e and decision point documents. 4. Terms of trade shocks in the 1980s and 1990s reduced government revenues from exports. In response, the countries looked to external sources to finance their high public sector spending rather than undertaking fiscal adjustments. External debt stocks typically rose sharply in a short period of time, accompanied in some cases by poor economic performance, even in light of satisfactory adherence to adjustment programs. The growing debt service and limited foreign exchange earnings added demand for foreign borrowing. In the case of Zambia, debt service obligations were a major contributor to balance of payments shortfall over the 1990s. Annex 11 88 Extensive balance of payments support, including over US$lbillion in adjustment lending from the World Bank over the decade, was necessary to keep the country from default with the IFIs. 5. Shift in debt composition. Four of the five countries studied experienced a significant shift in debt stock composition toward the multilateral institutions. Uganda has the highest concentration of its debt to multilaterals in the group, with nearly 90 percent; this is up from 44 percent in 1986. Cameroon, in contrast, stands out as the only country with a declining share of its debt to multilaterals over the last decade. Net transfers from multilaterals were negative for the majority of the 1990s. Quality at entry 6. The qualification process. There are two main criteria for a country to qualify for HIPC debt relief: (a) its debt burden is above the sustainability target after the full application of traditional debt relief mechanisms and (b) an established track record of policy performance (normally three years). All case study countries met the NPV of debt-to-exports criteria, with the exception of Guyana, which qualified on fiscal grounds. The record on prior policy performance across the five cases is uneven. Cameroon, Guyana, and Uganda (for O-HIPC) had an unambiguously satisfactory three-year track record. Guyana, Malawi, and Zambia (for E-HIPC) had episodes of policy slippages before qualifying for HIPC relief. In terms of policy performance after the decision point, performance has kept roughly at the same pace as before. Cameroon and Uganda have continued their good policy performance. Zambia showed strong performance in the year immediately prior to its decision point and has maintained good performance under two subsequent PRGF reviews. Guyana and Malawi, however, have experienced mixed performance since their E-HIPC decision points, which resulted in interruptions of their respective IMF programs. An emergency IMF program of assistance was approved for Malawi in September 2002, and the Guyana PRGF was belatedly approved in September 2002 after almost a year-long interruption. 7. In some cases, social sector strategies have been enhanced as a result of the HIPC Initiative. Developing formal strategies in the health and education sectors was required in Cameroon as a prior action in order to reach the decision point. These strategies will help to coordinate the expected significant increases in expenditures in these sectors. For all five case studies, the preparation of interim PRSPs (I-PRSP) was a requirement for decision point. The case study experiences with this process point to significant gains from initiating a participatory dialogue between government and civil society, and, to a somewhat lesser extent, directly with the poor. In all countries, civil society assessment of the participatory process was very positive. 8. The substantive quality of the four full poverty reduction strategies that have been completed so far has been varied. The most common criticisms from the joint staff assessments of the World Bank and IMF were the need for greater realism in ability to implement, greater attention to prioritization of activities, and costing of the proposed programs. All staff assessments for I-PRSPs in these five countries reported to their boards that they provided adequate foundations for moving forward with the HIPC decision points. 89 Annex I 9. The amount of IIIPC relief that each country qualifies for is calculated on the basis of the established ratios and using actual data, and debt sustainability analyses (DSAs) are conducted to demonstrate the countries' ability to repay existing and projected new debt in a sustainable manner. However, the realism of the DSA assumptions made at the decision point, given recent and past experience, was almost uniformly raised as a concern by those consulted for all five case studies. The April 2002 update of projections by the HIPC Unit confirms significantly lower growth and export performance in many countries, and revises the projections for many HIPCs, including those in the sample. Uganda currently has very high debt indicators as a result of the fall in coffee prices and the subsequent sharply lower export figures. The export shortfall for 2000-2001 in Uganda was 27 percent, and for Zambia it was 16-18 percent. 10. Completion point triggers. The floating completion points for all HIPCs are linked to trigger conditions agreed at the decision point. With the exception of Uganda, which did not have any triggers for E-HIPC, in the other cases the triggers cover both social sector objectives and structural reforms. For Cameroon, Guyana, and Zambia the triggers cover the key reforms, which, if adopted, should substantially improve the prospects for economic growth and, ultimately, repayment capacity. In Malawi, the study concludes that many of the key structural reforms necessary to attain HIPC objectives will have to wait until after the completion point. In most cases the triggers included existing conditions from ongoing programs. For Cameroon, Uganda, and Zambia, triggers supported the structural adjustment reforms that were part of ongoing PRGF and World Bank adjustment programs. In Guyana, the dialogue for determining HIPC completion triggers provided a framework to push discussions forward on macroeconomic and structural reforms, especially solidifying and heightening attention to civil service reform that complemented preparation of an Inter-American Development Bank adjustment loan. In Malawi, while not explicitly included as HIPC completion point conditions, the key structural reforms were covered in a World Bank structural adjustment credit that was approved at roughly the same time as the HIPC decision point. Implementation Experience 11. Interim debt relief received has on balance been near the decision point projections for these countries. Multilaterals have mostly delivered as agreed, but frequent delays have been experienced with reaching formal' agreements with the Paris Club members. In the case of Zambia, the Paris Club framework was not signed until 16 months after the decision point, which delayed bilateral agreements with members. Protracted bilateral negotiations with Russia resulted in a debt repayment schedule much higher than planned. More significant problems were noted in some countries regarding nonParis Club and commercial creditor relief. Separate negotiations for these creditors are ongoing in most of the five countries, although the amounts are generally relatively small. In some isolated cases (e.g., in Uganda and Zambia), lawsuits have been filed against the government for repayment of debt by commercial creditors or private entities that have purchased sovereign debt on secondary markets. 12. HIPC-related expenditures in most countries have not been fully spent, with the exception of Uganda. Delays in setting up ring-fencing arrangements and line ministry capacity constraints have contributory factors in some cases. In Cameroon, no spending had occurred 18 months after the decision point, though over US$100 million in debt service relief had been Annex f1 90 accumulated. Delays center around developing agreed selection and approval procedures in the HIPC Consultative Committee, made up of government, civil society, and donor representatives. In addition, original project proposals needed strengthening. In Guyana, the "implementation rate" for using HIPC resources is estimated at 75 percent, with capacity constraints of line ministries for effectively absorbing the additional resources and late finalization of the 2001 budget cited as key causes for the delay. 13. Addifionality of IPC implies that the net resource transfers from external sources are not reduced as a result of the debt relief. This is difficult to assess definitively, largely because of the lack of a "clean" counter-factual against which to benchmark changes in resource transfers since the decision point. Uganda is a clear case of full additionality, with large and sustained inflows of external aid. But donor support has declined relative to historical levels in Malawi and Zambia, largely because of governance concerns and for political reasons. However, more analysis is needed to determine the share of these declines that are related to the provision of HIPC relief, as opposed to other factors (such as governance issues or perceived election irregularities). It is possible that debt relief, by reducing the strain put by debt service on government budget, and hence the need to provide financial support by donors, is now permitting greater selectivity by donors. In this case, while it may be desirable to be selective for good reasons, additionality is not being achieved. 14. Fungibility of relief resources has to do with the level of total resources made available at the level of sectors targeted by HIPC. It is possible that additional resources directed at certain sectors through conditionality attached to HIPC relief may substitute for some (or all) of the government's own resources. The Malawi case study found that althoqgh the overall resources available to key sectors (e.g., health, education, and agriculture) had increased, governments' own resources had declined since the decision point relative to the pre-HIPC allocations. Government contributions to the Poverty Action Fund (PAF) in Uganda were also found to be flat, while the overall size of the budget has increased. Donors contributions of the PAF have been substantial in recent years. 15. Monitoring mechanisms. All five countries have processes for tracking of debt relief inflows, as well as adequate monitoring of HIPC-related expenditures at the central budget level. Some have transparent dissemination of this information, such as Zambia's practice of quarterly publication of HIPC account inflows and expenditures in local newspapers. However, there is relatively poor tracking of actual cash expenditures at the lower levels. This reflects more widespread public expenditure management problems, not specific to HIPC. Monitoring the effectiveness of HIPC-related expenditures was found only in Uganda, and some efforts are under way in Malawi. Performance indicators for priority social sectors in Uganda (education, health, water and sanitation) have been established and are regularly monitored. These indicators track effectiveness of all PAF resources, of which HIPC relief is the largest source. In Malawi, the Budget and Finance Committee of Parliament with the Malawi Economic Justice Network track outcomes in the education, health, and agriculture sectors. 16. Fulfllment of or progress toward the completion triggers has been mixed across the five countries. While Uganda successfully completed O-HIPC triggers on time, Guyana's O-HIPC completion point was delayed from December 1998 to May 1999 because of policy slippages. 91 Annex I Cameroon and Zambia are progressing as scheduled (that is, a notional three-year interim period, although completion points are in effect floating). They are likely to hit their expected E-HIPC completion dates but both face risks (port reform in Cameroon and the mining crisis in Zambia). Guyana and Malawi are currently showing mixed progress, due partly to recent weaknesses in macroeconomic management. 17. PRSP progress. Four of the five case countries have completed their first PRSPs. Cameroon is expected to complete its PRSP within the coming months. As noted for the I-PRSPs prepared for HIPC qualification, most field consultations with civil society confirmed notable advances in the dialogue with government under PRSP preparation. One-tenth of the population participated in some form under the Guyana PRSP consultations. A successful feature of the consultations noted in Malawi, Uganda, and especially Zambia was the interaction of civil society with government policymakers through working group arrangements. These groups allowed close consultations over an extended period of time, and in most cases resulted in tangible outputs for incorporation into the overall strategy. As noted in most joint staff assessments for the completed PRSPs, further work is needed in operationalizing the strategies, in terms of prioritizing intrasectoral activities, costing, and integration into the budget. The exception is Uganda, which was already developing a Poverty Eradication Action Plan in 1995, prior to HIPC. This framework has been strengthened over time, and is fully linked to the budget and sector plans. Prospects 18. All countries have a reduced fiscal Table 1.2: Current Debt Ratios burden from lower debt service, and a distinct increase in social and poverty- reducing expenditures. However, prospects NPV of debt External debt service for long-term debt sustainability, under Country to exports (percent) to exports (percent) current HIPC debt relief arrangements, are Cameroon 197 11 mixed across the five case study countries. Guyana 136 9 At present, none of the countries has Malawi 194 12 achieved the HIPC target of 150 percent Uganda 254a 13 NPV of debt-to-exports or 250 percent NPV Zambia 441 15 of debt-to-government revenues for Guyana 'The latest estimate for the NPV of debt-to-exports for Uganda is (table 1.2).1 Based on recent data, only 209 percent, based on an updated DSA using end-June 2002 data. Cameroon is projected to become sustainable Source: World Bank and IMF 20029. within the next three to five years. Guyana is expected to be just over the 280 percent fiscal threshold. Malawi, Uganda and Zambia are not expected to reach sustainable debt levels by the end of the decade. Progress on addressing structural causes or risks to repayment capacity have so far been modest. 7. Bank Management noted that the original HIPC long-term debt sustainability paper explains that these numbers do not reflect the full extent of HEPC debt relief that would become available after the completion point. Figures cited [in table 1.2] are higher than the 150 percent HIPC threshold because four of these countries have not yet reached completion point. Annex 11 92 19. Key risk factors to repayment capacity noted in the case studies include continued export concentration and the link to commodity price fluctuations, continued weaknesses in public expenditure management, lack of private sector activity, and the accumulation of new debt. Little progress has been noted in the area of export diversification, with the exception of Zambia, which showed good performance from its non-mining exports in 2001. Work is reported to be under way in all countries to strengthen diversification strategies. The most pressing case is in Cameroon, where the stock of oil reserves is expected to decline markedly over the coming years. The importance of export diversification has been underscored by the deterioration of debt indicators in many HIPCs given the further recent declines in commodity prices and the global economic slowdown (World Bank and IMF 2002d). 20. The terms and volume of new lending are naturally a concern for all HIPCs. While HIPC debt relief can greatly reduce the debt stock as of the decision point, there are no guarantees against slippage back to original levels of indebtedness. For the group of five case study countries, the issue is particularly relevant for Cameroon, Uganda, and, in the near term, for Zambia. According to the latest HIPC unit projections, Uganda is expected to borrow on average well over US$300 million annually through the end of this decade. Cameroon and Zambia will borrow on average over US$200 million per year. The terms of the large magnitude of expected future borrowing will have direct effects on these countries' debt sustainability. 93 Annex J Annex J: Constraints to Growth as Identified by OED Country Assistance Evaluations Burkina Bolivia Faso Cameroon Cote d'Ivoire Ethiopia Ghana Malawla Political Instability/civil war 4 4 4 4 Natural calamities Macroeconomic Imbalances 1 Low/negative growth 2.Inflation 3. Fiscal deficit 4. Current account deficit External shocks 1. Falling commodity prices 2. Adverse terms of trade 3. Falling exports 4. Misalignment of currency 5. Reliance on primary commodities 6. Declining extemal remittances/foreign aid Constraints to private sector development b institutional and capacity constraints 1.Poor economic management 2. Poor govemance 3. Corruption 4. Weak/ineffective civil service 5 Weak legal/judicial system 6 Weak regulatory system 7. Population growth 8. Higher poverty 9 Adverse social conditions a. Refugee influx from Mozambique b. Poor environment for business Anniex J 94 Annem J: Constraints to Growth as Idendtfed by OED Country Assistance EvaRuations (confinued) Mozambique Tanzania Togo Uganda Zambia Yemen Political instability/civil war Natural calamities Macroeconomic imbalances 1. Low/negative growth 2. Inflation 3. Fiscal deficit 4. Current account deficit External shocks 1. Falling commodity prices 2. Adverse terms of trade 3. Falling exports 4. Misalignment of currency 5. Reliance on primary commodities 6. Declining external remittances/ declining debt Conotrainto to privato zector dzvelopmentb InstitutIonal and capacity conatrainto 1.Poor economic management 2. Poor governance 3. Corruption 4. Weak/ineffective cvil service 5. Weak legal/judicial system 6. Regulatory system 7. Population growth 8. Higher poverty 9. Adverse social conditions a. Refugee influx from Mozambique. b. Poor environment for business. 95 Annex K Annex K. Management Response INTRODUCTION 1. The Heavily Indebted Poor Countries Debt Initiative (HIPC Initiative) launched in 1996 and enhanced in 1999 represented an advance by the international community in addressing the chronic external debt burden of the world's poorest countries. It brought together donors, creditors, NGOs, and academics around a core development issue, and led to improvements in the way debt relief is provided. The Initiative spurred the international community to mobilize and commit a large amount of money and attention to reducing the debt overhang in the poorest countries-and thereby began to eliminate a huge obstacle to growth and development. For the Bank, it represented one of the initial efforts to engage in serious dialogue with civil society and encourage active participation of beneficiaries beyond the project level. This dialogue resulted in progressively better design of the Initiative and improved transparency in its implementation. 2. Progress in Implementing the HIPC Initiative. Since its inception six years ago, the HIPC Initiative has succeeded in bringing substantial debt relief to 26 of 42 potentially eligible countries. In parallel, efforts are under way to put in place longer-term poverty reduction strategies-crafted and implemented by the HIPCs using a broad participatory process-helping to steer savings from reduced debt service payments to poverty reducing programs. Partly as a result of its early successes, expectations on what the HIPC Initiative can achieve are high, and there are frequent calls for further accelerating, broadening and deepening the debt relief provided through the Initiative. 3. The OED Review. Management welcomes this timely and comprehensive review of the HIPC Initiative undertaken by the Operations Evaluation Department, as it helps take stock of what has been learned in the implementation of the HIPC Initiative to help resolve its shortcomings and better achieve its goals. This Management Response discusses the OED report's main findings and presents views on key issues that are fundamental to the success of the HIPC Initiative. Nevertheless, it is important to remember that all of the 26 countries receiving debt relief under the Initiative so far reached their decision point only in 2000 or thereafter, so any conclusions regarding the program's accomplishments or shortfalls can only be tentative at this early stage in the program's implementation. 4. Joint Program. The HIPC Initiative is a joint Bank-Fund program in which the staffs of the two institutions work closely together. Nonetheless, these comments reflect solely the views of World Bank Management and do not in any way purport to represent nor implicate the Fund. MAIN FINDINGS AND RECOMMENDATIONS 5. This Management Response focuses on the OED Review's main findings and recommendations. Also attached are detailed responses in the Management Action Record matrix. Annex K 96 6. OED Recommendations. The OED Review's four recommendations are to: "(i) clarify the purpose and objectives of the Initiative, ensure that its design is consistent with these objectives, and both the objectives and how they are to be achieved be clearly communicated to the global community; (ii) improve the transparency of the methodology and economic models underlying the debt projections and the realism of economic growth forecasts in the debt sustainability analyses; (iii) maintain the standards for policy performance, and when the established criteria are to be relaxed that there be a clear and transparent rationale to ensure that the risks to achieving and maintaining the Initiative's objectives are minimized; and (iv) focus more on pro-poor growth and provide a better balance among development priorities relative to the current emphasis on social expenditures."' 7. Objectives of the HIPC Initiative. The OED Review paper rightly points out that the objectives of the HIPC Initiative have grown in ambition over the years. The original HIPC framework aimed to bring about a comprehensive debt reduction "to achieve a sustainable debt situation" in the eligible countries.9 The consultative HIPC Review launched in early 1999 with the debtor countries and the wider donor, NGO and academic communities, recommended that debt relief operations be better linked to the overall development and poverty reduction agenda. Participants in that review also felt that decisions regarding the use of debt service savings should involve civil society and the poor themselves. Poverty Reduction Strategies, formulated and implemented through a wide participatory process concurrent with the HIPC debt relief process, were identified as a viable means to channel debt savings toward uses considered to be high priority. 8. Debt Relief and Development. When the HIPC Initiative was enhanced with lower thresholds and faster provision of debt relief, it was made clear that there is at best an indirect relationship between debt relief and poverty reduction. Debt relief was never presented as a panacea that would overcome the towering challenge of long-term development and poverty reduction in HIPCs. On the contrary, it was always seen as a program that, by reducing the external debt stock in HIPCs, would ultimately contribute to the broader goal of assisting these countries to accelerate growth and reduce poverty. Since the inception of the HIPC Initiative, and particularly after the 1999 enhancement, the Bank consciously emphasized the comprehensive development framework in its country assistance strategies, lining up numerous programs, including the HIPC Initiative, in a consistent and comprehensive attack on poverty. Naturally, the HIPC program-just one building block in a more comprehensive development architecture- aligned itself with the objectives and goals of this larger program, but was never meant to supplant it. 9. Additionality of Resources. We fully support OED's emphasis that additionality is an important underlying principle of the HIPC Initiative, but contend that it should be "The Heavily Indebted Poor Countries (HIPC) Debt Initiative: An OED Review", Operations Evaluation Department, CODE2002-0089, World Bank, December 30, 2002. 9 "The HIPC Debt Initiative-Elaboration of Key Features and Possible Procedural Steps," IDA/SecM96-927, August 26, 1996, p. 1. 97 Annex K assessed on a country-by-country basis. For example, lower grants and/or credits may be appropriate in cases where serious policy slippages or even reversals lead to slower growth and a deteriorating macroeconomic situation. Nevertheless, staff's preliminary review of the data indicates that, much as one would expect, loan commitments and grants to HIPCs in aggregate climbed in 1999 and 2000, just about the time the HIPC Initiative was taking off. And the World Bank clearly held up its end of the bargain, because there is clear evidence that the combination of net flows and debt relief from IDA increased sharply at about the same time. 10. Positive Net Transfers and Debt Sustainability. We agree with OED that net transfers of external resources to HIPCs must be adequate to support development programs needed to reach the Millennium Development Goals. But this should not come at the expense of debt sustainability. It is imperative, therefore, that the flow of grants increase significantly to these countries, and in this context we welcome the grant component incorporated in IDA-13. At the same time, however, it needs to be acknowledged that grants from bilateral and other sources will need to rise further still to eliminate the tension between these two potentially competing objectives (namely, attaining MDGs and debt sustainability). 11. Transparency in projections. The OED Review recommends greater transparency in the methodology and economic models underlying the debt projections and greater realism in the economic growth forecasts in the debt sustainability analyses. It finds that: (i) the macroeconomic projections embodied in the debt sustainability analyses (DSAs) tend to be overly optimistic; (ii) models underpinning these optimistic projections are not easily understood or substantiated; and (iii) the sustainable debt outlook based on optimistic macroeconomic projections undermines "a better assessment of the prospects and risks" facing HIPCs. The World Bank, in collaboration with country governments and the IMF, is already taking steps to make sure that greater realism is introduced into long-term economic projections that underpin debt sustainability calculations. At the same time, it needs to be recognized that macroeconomic projections for HIPCs are inherently difficult due to unreliable data, internal conflict, the effect of unpredictable weather patterns on agrarian economies dependent on one or two key commodity exports, and volatility in commodity prices. To deal with these issues, staff are already required to develop stress tests to assess risks to their baseline scenarios and prepare alternative scenarios that reflect the vulnerabilities and uncertainties these countries face."o As far as greater transparency in methodology and economic models underlying the debt projections are concerned, the Fund and the Bank apply to IUPCs the same disclosure and reporting standards they apply to all other countries. 12. Impact of Projections. The OED Review recognizes that HIPCs are not penalized if projections by Fund and Bank staffs turn out to be optimistic. First, the calculation of debt reduction for any HIPC is not based on projections but on actual external public debt and the 'average export value over the previous three years. And 10 "Heavily Indebted Poor Countries (HIPC) Initiative - Status of Implementation," DC2002-0020, September 21, 2002, pp. 31-32; "The Enhanced HIPC Initiative and the Achievement of Long-Term External Debt Sustainability," http://www.worldbank.org/hic/Long-Term pdf, Apn1 15, 2002, p. 21. Annex K 98 second, if the actual NPV debt-to-export ratio at completion point is over the projected ratio on account of factors clearly beyond the control of the country (such as an unexpected decline in commodity prices), then the HIPC framework allows the country to receive additional debt relief and reduce its NPV debt-to-exports ratio to 150 percent. While over-optimistic projections may not penalize HIPCs directly, the OED Review correctly points out that they may have the indirect effect of encouraging HIPCs (after decision or completion point) to borrow at above sustainable levels. Management is alert to this issue and requires country assistance strategies to carefully assess the macroeconomic prospects of all low-income countries, including HIPCs, and emphasizes sensitivity analyses and preparation of low case lending scenarios. 13. Standards of Policy Performance. The OED Review paper notes that sustained policy reforms are crucial for countries to maintain debt sustainability and achieve economic growth and poverty reduction. Therefore, it emphasizes the importance of maintaining clear standards of policy performance for the HIPC Initiative, especially in the instance of so-called "Millennium-rush countries," for which OED perceived a tendency to lower standards. The original HIPC framework required that HIPCs demonstrate a track record of reform for three years to qualify for debt relief at decision point, followed by a further period of reforms, to actually benefit from irrevocable debt relief at completion point. A counterpoint was posed by many critics participating in the 1999 HIPC Review who argued that a strict track record requirement was a hurdle intentionally raised by the Bank/Fund and the creditor countries to delay debt relief. Without attempting to judge the merit of this contention, efforts were made under the enhanced framework to assist eligible HIPCs in reaching decision point more quickly, as a result of which the requirement of the three-year track record was relaxed. This was appropriate, because the quicker an eligible HIPC can reach decision point, the earlier the international community can engage with the country to assist with reform and development and the greater the likelihood of success in reaching completion point and benefiting from irrevocable debt reduction. At the same time, strong policy performance standards continued to be maintained in the design of completion point triggers. Indeed, in line with changing practice in selecting conditions for adjustment loans, the Management is also trying to reduce the number of completion point triggers, and ensure that those that are included are of strategic importance to the success of the program. 14. Emphasis on Social Expenditure. The operational nature of HIPC documents requires that they explain how savings from debt relief will be channeled into poverty reduction expenditures. A large part of the additional expenditures are concentrated in social sectors although, in some HIPC documents, they also include physical infrastructure and economic services such as water supply, rural electrification, housing, and agricultural services. These expenditures, when taken together with the programs of policy and structural reform that are also articulated in HIPC documents, constitute a potentially powerful strategy for pro-poor growth. In all cases, completion point triggers include a requirement that a PRGF program is on track, reflecting the importance of macroeconomic stability, an essential ingredient for growth. Completion point triggers also include structural reforms designed to adjust incentives toward greater efficiency, private sector development, and improved governance, further strengthening the focus on 99 Annex K growth even though they may not call for additional public expenditures. And in the future, PRSPs and public expenditure reviews will provide a better empirical assessment of where incremental resources should be allocated. 15. Attachment. Attached to this Management Response are detailed responses in the Management Action Record matrix. Annex K 100 ATTACHMENT: MANAGEMENT ACTION RECORD Major OED Management Resgonse Recommendations 1. Clarify the purpose and 1. Management will strengthen the communication of objectives of the Initiative, ensure the objective of the HIPC Initiative and clarify that it is that its design is consistent with to reduce debt stocks, and so contribute to broader these objectives, and both the efforts aimed at accelerating growth and reducing objectives and how they are to be poverty. achieved be clearly communicated to the global community When the HIPC Initiative was enhanced, debt relief in itself was envisioned only as a necessary, but never as a sufficient, condition to overcome the challenges of long- term development and poverty reduction in HIPCs. The Initiative deliberately aligned itself in each HIPC with the goals of the country program that often contained many elements of a broader development architecture. By reducing the debt stock, the Initiative was expected to contribute to a more comprehensive development effort, but not supplant it In its communications strategy, Bank Management will work with relevant stakeholders to ensure that the objectives of the HIPC Initiative are clarified and communicated consistently. 2. Improve the transparency of the 2. All debt sustainability analyses of HIPCs are based methodology and economic models on accompanying PRGF macroeconomic projections underlying the debt projections and where the key assumptions for debt sustainability are the realism of economic growth presented. The methodology and economic models forecasts in debt sustainability underlying these projections are subject to the same analyses. disclosure and reporting standards as for all other countries. Management agrees with OED's recommendation to improve the realism of growth projections and elaborated these points in the HIPC progress report of September 2002. To deal with the inherent volatility, risks and weak underlying data, staffs of the Fund and the Bank are being asked to improve the realism of long-term projections and apply stress tests that assess risks to the baseline scenarios. 101 Annex K Major OED Management Response Recommendations 3. Maintain the standards for 3. Management believes that the earlier HIPCs reach policy performance, and when the decision point, the greater the engagement of the established criteria are to be relaxed international community in assisting with reform and there should be a clear and development efforts and the greater the likelihood of transparent rationale to ensure that reaching completion point. At the same time, the risks to achieving and Management will seek to include fewer-and the most maintaining the Initiative's strategic-triggers in completion documents to ensure objectives are minimized. that while standards of policy performance are maintained, they are not unnecessarily onerous. It should be noted that floating completion point conditionality has allowed HIPCs to tailor the pace of change in accordance with their capacity and the constraints they face. 4. Focus more on pro-poor growth 4. Management will continue to assist HIPCs in assuring and provide a better balance among that savings from debt relief are allocated to development priorities relative to expenditures for pro-poor growth (including social the current emphasis on social development priorities where appropriate) consistent expenditures with the PRSP framework. An essential element of the HIPC initiative is that completion point triggers include measures allocating savings from HIPC relief to finance expenditures supporting poverty reduction and growth in line with priorities expressed in PRSPs. This will be kept under careful review so that allocations to individual sectors or programs do not exceed absorptive capacity constraints. Through the Poverty Reduction Department, the HIPC Unit, and the regional country teams, Management will continue to ask that future Decision Point cases follow this approach. Accordingly, it will work to determine Completion Point triggers for pro-poor growth on a case-by-case basis in accordance with the PRSP or I- PRSP. Annex L 102 Annemx L: Repoirt firom the Committee on Development Efefetiveness - Chzirmaln's summairy 1. On January 29, 2003 the Committee on Development Effectiveness (CODE) met to discuss the findings and recommendations of the OED report Heavily Indebted Poor Countries (HIPC) Debt Initiative: An OED Review (CODE2002-0089) and draft Management Response (CODE2003-0002). 2. Main Findings. The HIPC Debt Initiative was formally agreed to by governments around the world in September 1996 and "enhanced" in 1999. Its original objective was to reduce, within a reasonable time horizon, the external debt burden of qualifying countries to "sustainable" levels. While the link to growth has been maintained, the enhanced framework added increased social expenditures for poverty reduction as an explicit objective. The report concludes that the HIPC Initiative has been a catalyst for far reaching changes in the processes surrounding development assistance and that the Initiative has been highly relevant in addressing a key obstacle facing many poor countries. The Initiative is likely to achieve its original fundamental goal of reducing the external debt of the HIPCs. While its design has been adequate and appropriate to achieve its original stated objective of reducing the "debt overhang", the report concludes that this design is no longer consistent with the expanded set of objectives of the enhanced HIPC, notably to simultaneously provide a "permanent exit" from rescheduling, promote growth, and to free up resources for increased social expenditures. Management is in broad agreement with OED's recommendations but does not concur with parts of OED's analysis, notably the conclusion that the design, as such, is inconsistent with the objectives of the enhanced HIPC Initiative. The Management Response underlines that by reducing debt- stocks, the HIPC Initiative was always meant to contribute toward a broader, more comprehensive development architecture but not supplant it. 3. Conclusions and Next Steps. CODE commended OED for an excellent report. Members supported the thrust of recommendations of the report and were overall satisfied with the Management Response. Key messages arising out of the committee's discussion were that: (a) debt relief was not a substitute for a broader, growth oriented, development program and that HIPC needs to be seen as one of several instruments to support poverty reduction; (b) additionality was an important part of the HIPC framework but should not override performance- based allocation of resources; and (c) that realism of debt sustainability analysis (DSA) and a clear external communication supporting wider public understanding of the report's findings were also important. CODE recommended that the OED report be disclosed. The report findings will inform, inter alia, management's annual update on HIPC planned for the Development Committee in September 2003. 4. ffssues. Some of the main issues raised by the committee were the following: o Design and Objectives of the HIPC Initiatives. The committee's discussion of the design and objectives of the Initiative largely centered around the disconnect between what the Initiative could realistically achieve and the overly ambitious expectations held by some stakeholders. The committee agreed that it would be important to clarify the HIPC 103 Annex L Initiative's objectives but advised that it was the role of Governors and Governments, not management nor a panel of experts, as the objectives were established through a political process and agreed to by Governors. CODE supported the need for management to ensure that Bank communications were clear and consistent, and did not further raise expectations. Management agreed it would address this issue in its annual update on HIPC. * The committee agreed with the OED report's finding that the HIPC Initiative had objectives that could not be met through design improvements alone. Members discussed varying interpretations of what a "permanent exit" from debt rescheduling was intended to mean and suggested caution in automatically linking debt relief to debt sustainability and poverty reduction. The committee supported management's view that HIPC is one of several instruments to support poverty reduction and underlined the need to link HIPC to a broader development program that addressed the underlying structural problems. The committee believed the evaluation could have taken greater account of poverty reduction strategy papers (PRSPs) as the framework to address long-term development problems in these countries. OED noted that it was in the planning stages of a separate and in-depth review of the PRSP process. * Management noted that while the HIPC is now clearly linked to the PRSP process, and by extension, to supporting countries in achieving their growth and poverty reduction goals, throughout its evolution the Initiative has retained its central objective of reducing the external debt burden of beneficiary countries. Management also noted that the growth agenda is central to the PRSP approach in many countries where the PRSP is advanced. Several speakers confirmed that the HIPC review was an important learning process and underlined the importance of continued frank and open consultations with all stakeholders. * Social Sector Spending. The committee was in agreement with the Review's finding that the early focus on the use of HIPC debt-service savings had focused too narrowly on social expenditures. The committee felt that a broader growth and development framework was needed to accommodate other factors that also affect growth -- such as the investment climate, infrastructure development, and economic productivity -- was preferable. Some members noted that the lack of emphasis on economic growth was symptomatic of PRSPs not adequately addressing growth. They found that this was an area that required further attention and supported management's proposal to give greater consideration to economic growth when designing completion point triggers. Management agreed that the support for poverty reduction under the HIPC Initiative will be consistent with the broad policy framework expressed in PRSPs and noted that high priority public expenditures, together with programs of policy and structural reform constitute a potentially powerful strategy for pro-poor growth. * Resources. The committee discussed the implications of reallocation of funds from non- HIPC poor countries to HIPC countries. Some chairs stressed that the intention of the additionality principle did not exempt countries from the performance-based allocation Annex L 104 framework. The Management Response noted that strong policy performance standards were integral to the development of lending triggers. Some chairs pointed to a potential conflict between the country policy and institutional assessment (CPIA) allocations and debt relief. The committee noted the apparent overall decline in resource transfers to HIPC countries through 2000, and while recognizing it was early in the HIPC process, underlined that if resources were to be freed-up for poverty reduction it would be important to: (i) assess the public expenditure priorities expressed in the PRSP to gain a better understanding of how to best use HIPC resources; (ii) ensure aid flows and savings from HIPC debt relief support those priorities and reach the poor; and (iii) ensure longer term measures are in place to address potential country absorptive capacity. Some members underlined the need to identify new sources of concessional financing, including grants. The committee also noted that since trends in aid flows are influenced strongly by policy and economic performance, they should be assessed on a country-by- country basis rather than in aggregate. o Debt Sustainability Analysis. The committee supported OED's recommendation on the need for more transparency in the methodology for debt sustainability analysis (DSA) and for more realistic growth forecasts. Some members, however, noted that OED could have gone deeper into some of the factors which contributed to the debt difficulties so that appropriate lessons could be drawn. The committee also supported the idea of monitoring post-HIPC performance and stressed that debt sustainability would require giving more attention to the investment climate, volatility of trade exports, access to markets, and domestic debt. The need to address possible conflicts between debt sustainability and increased Bank lending were also cited. It would also be important to ensure that resources are productively used so that sufficient capacity to service the debt is created. Improvement of debt management capacity of HIPCs would also be key. Management informed the committee that the Bank, in collaboration with country governments and the IMF, was already taking steps to ensure that greater realism was introduced into long-term economic projections and also noted that the Fund and Bank apply to HIPC the same disclosure standards they apply to other countries and programs. o Management of Expectations. The committee agreed that some stakeholders had overly high expectations of the enhanced HIPC and were concerned that the Bank, as one of the principal caretakers of this initiative, could be vulnerable to criticism on the final outcomes, even though key decisions and actions needed to realize the broader outcomes were beyond the Initiative's purview. The committee underlined the need for management to put into place an effective internal and external communication strategy regarding the Initiative and what it could and could not do, to minimize the risk of misinterpretation of the Review's findings. Emphasis should be on what is required to achieve the objectives rather than the realism of the expectations. OED is planning dissemination activities for its report. Management confirmed that it would undertake a proactive communication strategy.