34416 WORLD BANK OPERATIONS EVALUATION DEPARTMENT IMF INDEPENDENT EVALUATION OFFICE EPENDENT IND E VA LUA TION OFF IC E The Poverty Reduction Strategy Initiative Findings from 10 Country Case Studies of World Bank and IMF Support ALBANIA CAMBODIA ETHIOPIA GUINEA MAURITANIA MOZAMBIQUE NICARAGUA TAJIKISTAN TANZANIA VIETNAM 2005 http://www.worldbank.org/oed The World Bank http://www.imf.org/ieo Washington, D.C. © 2005 The International Bank for Reconstruction and Development / The World Bank 1818 H Street, NW Washington, DC 20433 Telephone 202-473-1000 Internet www.worldbank.org E-mail feedback@worldbank.org All rights reserved. 1 2 3 4 07 06 05 The findings, interpretations, and conclusions expressed herein are those of the author(s) and do not necessarily reflect the views of the Board of Executive Directors of the World Bank or the governments they represent. The World Bank does not guarantee the accuracy of the data included in this work. The boundaries, colors, denominations, and other information shown on any map in this work do not imply any judgment on the part of the World Bank concerning the legal status of any territory or the endorsement or acceptance of such boundaries. Rights and Permissions The material in this work is copyrighted. Copying and/or transmitting portions or all of this work without permission may be a violation of applicable law. The World Bank encourages dissemination of its work and will normally grant permission promptly. For permission to photocopy or reprint any part of this work, please send a request with complete information to the Copyright Clearance Center, Inc., 222 Rosewood Drive, Danvers, MA 01923, USA, telephone 978-750-8400, fax 978-750-4470, www.copyright.com. All other queries on rights and licenses, including subsidiary rights, should be addressed to the Office of the Publisher, , World Bank, 1818 H Street NW Washington, DC 20433, USA, fax 202-522-2422, e-mail pubrights@worldbank.org. Cover photo: World Bank Photo Library. ISBN 0-8213-6219-4 e-ISBN 0-8213-6220-8 Contents vii Acknowledgments ix Acronyms and Abbreviations 1 1 Introduction 1 Facing the poverty challenge 3 The scope of the present report 4 Summary findings from the 10 country case studies 9 2 Albania 9 Growth depends on private sector development 11 The PRS process—better development planning 13 World Bank support—emphasizing country ownership 15 Bank-Fund collaboration 15 Conclusion—building credibility, coordinating aid 17 3 Cambodia 17 A long road from conflict to growth 18 The PRS process—awkward timing, incomplete ownership 21 World Bank support—a need for effective local support and for better alignment with the PRSP 22 Conclusion—learning from experience 25 4 Ethiopia 25 Addressing economic fragility 27 The PRS process—good for government-donor relations 30 World Bank support—more decisionmaking in the field 31 IMF support—constructive engagement 32 Conclusion—a revived partnership 33 5 Guinea 33 Good potential to benefit from the PRS approach, but weak institutional capacity 34 The PRS experience—a “good document,� but inadequate process and limited implementation 38 IMF support—a challenging policy environment 40 Bank-Fund collaboration 40 Conclusion—form over substance iii T H E P O V E R T Y R E D U C T I O N S T R AT E G Y I N I T I AT I V E 43 6 Mauritania 43 Abundant resources, unmet needs 44 The PRS process—a centralized effort that may undermine implementation 47 World Bank support—overall alignment with the PRSP 48 Conclusion—a process in danger of reversal 51 7 Mozambique 51 Poverty reduction as a peace dividend 53 The PRS process—a common point of reference with strong government ownership 56 World Bank support—an effective local presence 57 IMF support—too low a profile with civil society 58 Bank-Fund collaboration 58 Conclusion—toward general budget support 61 8 Nicaragua 61 Poverty reduction in a polarized country 63 The PRS process—too little focus on growth 66 World Bank support—discussing options for growth 68 Bank-Fund collaboration 68 Conclusion—modest results under difficult initial conditions 71 9 Tajikistan 71 Surmounting regional strife to reduce poverty 73 The PRS process—better governance for a better business climate 76 World Bank support—a challenging shift to community-driven development 77 IMF support—challenges of shifting to a new way of doing business while continuing to foster fiscal discipline 78 Bank-Fund collaboration 79 Conclusion—implementation requires clearer priorities, greater coordination 81 10 Tanzania 81 A new resolve to fight pervasive poverty 83 The PRS process—a sharper focus, better data, and wider participation 86 World Bank support—too intrusive or not intrusive enough? 88 IMF support—gradual alignment with the PRSP and opening up of the debate on program design 89 Conclusion—a healthy process that has become embedded in domestic institutions overcomes defects in the original document 91 11 Vietnam 91 Successful poverty reduction 92 The PRS process—concurrent strategies 94 IMF support—flexibility, despite some persistent disagreements 95 Bank-Fund collaboration 95 Conclusion—merging parallel policy processes 97 Annex: Guidelines for the Country Case Studies 101 Endnotes iv CONTENTS Boxes 2 1.1 Five principles of the Poverty Reduction Strategy Initiative 3 1.2 Principles underlying the IMF’s Poverty Reduction and Growth Facility 12 2.1 Key elements of Albania’s National Strategy for Social and Economic Development 19 3.1 Cambodia’s National Poverty Reduction Strategy: broad priorities and action areas 27 4.1 The four pillars of Ethiopia’s PRSP 27 4.2 Civil society mobilization around the PRSP 29 4.3 An architecture for government–donor dialogue 36 5.1 Impact of the participatory process on the macroeconomic framework of Guinea’s PRSP 37 5.2 The PRSP dissemination strategy in Guinea 46 6.1 Key elements of Mauritania’s PRSP 54 7.1 Priority areas of Mozambique’s PRSP 75 9.1 Key areas of Tajikistan’s PRSP 93 11.1 Vietnam’s PRSP policy matrix Figures 10 2.1 Albania at a glance 10 2.2 The changing structure of Albania’s economy 14 2.3 Evolution in World Bank lending to Albania, 1992–2004, by sector 18 3.1 Cambodia at a glance 26 4.1 Ethiopia at a glance 30 4.2 Spending in poverty-oriented sectors in Ethiopia 34 5.1 Guinea at a glance 39 5.2 Official aid to Guinea, 1998–2002 39 5.3 Changes in consumer price index in Guinea, January 2002–March 2003 44 6.1 Mauritania at a glance 52 7.1 Mozambique at a glance 52 7.2 Real GDP growth and inflation in Mozambique, 1991–2002 62 8.1 Nicaragua at a glance 72 9.1 Tajikistan at a glance 73 9.2 Real GDP growth and inflation in Tajikistan, 1996–2002 73 9.3 Tajikistan’s total external debt, 1996–2002 75 9.4 External aid to Tajikistan, 1992–2001 82 10.1 Tanzania at a glance 85 10.2 Expenditures in priority sectors of Tanzania’s PRSP, 1998/99–2002/03 87 10.3 World Bank commitments by sector, FY98–00 and FY01–03 92 11.1 Vietnam at a glance 94 11.2 Poverty-reducing spending in Vietnam, 1998–2003 Tables 4 1.1 Ten country case studies of the PRS Initiative 21 3.1 PRSP targets for poverty reduction in Cambodia 66 8.1 Planned and actual PRSP expenditures in Nicaragua, 2000–2002 82 10.1 Growth and investment indicators in Tanzania, period averages, 1991–2002 92 11.1 Selected macroeconomic indicators for Vietnam, 1995–2002 v Acknowledgments This report was prepared by Bill Battaile (OED) Mozambique: the evaluation team was co-led and Steve Kayizzi Mugerwa (IEO), with inputs by Tsidi Tsikata (IEO) and Nils Boesen (OED from Steve Kennedy (editor). consultant) and included Paulo Mole (local con- sultant). The country chapters were derived from the fol- lowing 10 country case studies conducted by Nicaragua: the evaluation team was co-led by OED and IEO: Martin Kaufman (IEO) and Nils Boesen (OED consultant) and included Alex Segura (IEO) and Albania: the evaluation team was led by Basil Desiree Elizondo (local consultant). Kavalsky (OED consultant) and included Zamir Islamshah (OED), and Ylli Cabiri (local consult- Tajikistan: the evaluation team was co-led by ant). Tsidi Tsikata (IEO) and Jeffrey Balkind (OED consultant) and included Goulsara Pulatova (local Cambodia: the evaluation team was led by Basil consultant). Kavalsky (OED consultant) and included Zamir Islamshah (OED), Bill Battaile (OED), and Chan Tanzania: the evaluation team was co-led by Sophal (local consultant). Steve Kayizzi Mugerwa (IEO) and Madhur Gau- tam (OED) and included Shonar Lala (OED), Ethiopia: the evaluation team was led by Alison Rwekaza Mukandala (local consultant), and Evans (OED consultant) and included Shonar Joseph Semboja (local consultant). Lala (OED), and Mohammed Mussa (local con- sultant). Vietnam: the evaluation team was led by Jeff Chelsky (IEO) and included Soren Jensen (IEO Guinea: the evaluation team included Isabelle consultant). Mateos y Lago (IEO) and Jeff Chelsky (IEO). Mauritania: the evaluation team included Del- bert Fitchett (OED consultant), Laura Kullen- berg (OED), Zamir Islamshah (OED), and Abou Diallo (local consultant). vii ACRONYMS AND ABBREVIATIONS CAS Country Assistance Strategy CDF Comprehensive Development Framework CFAA Country Financial Accountability Assessment CSO Civil society organization ESAF Enhanced Structural Adjustment Facility EU European Union HIPC Heavily Indebted Poor Countries IDA International Development Association IEO Independent Evaluation Office IMF International Monetary Fund I-PRSP Interim Poverty Reduction Strategy Paper JSA Joint Staff Assessment MDGs Millennium Development Goals MEFP Memorandum of Economic and Financial Policy MTEF Medium-Term Expenditure Framework NGO Nongovernmental organization OED Operations Evaluation Department PARPA Action Plan for the Reduction of Absolute Poverty PER Public Expenditure Review PRGF Poverty Reduction and Growth Facility PRS Poverty Reduction Strategy PRSC Poverty Reduction Support Credit PRSP Poverty Reduction Strategy Paper PSIA Poverty and Social Impact Assessment SWAP Sector-wide approach UNDP United Nations Development Program WBI World Bank Institute ix 1 Introduction T he Independent Evaluation Office (IEO) of the International Monetary Fund (IMF) and the Operations Evaluation Department (OED) of the World Bank—independent units reporting to their respective institu- tions’ Executive Boards—recently completed parallel evaluations of the Poverty Reduction Strategy (PRS) Initiative, with particular attention to the role and effectiveness of World Bank and IMF support.1 A key source of evidence for the evaluations is expected to achieve measurable results. was a set of 10 country case studies, most Important changes in poverty-related planning completed in the second half of 2003, of processes and first-generation strategy countries that had completed a Poverty documents have already been completed in Reduction Strategy Paper (PRSP). This volume more than 40 low-income countries. contains summaries of those case studies. The Initiative’s framework for support to Readers are referred to the full case studies for low-income countries comprises two key more complete treatment of the issues raised elements: in this volume.2 This introduction first discusses the key • First, country-authored, medium-term PRSPs, features of the PRS Initiative, followed by the which are expected to draw on broad-based scope and methodology of the case studies. It consultations with key stakeholder groups and concludes with highlights of the main a thorough diagnosis of the causes of poverty messages of the case studies. and sources of growth in the country. • Second, the alignment of all external assis- Facing the poverty challenge tance to support the goals and priorities of At the end of 1999, the World Bank and the IMF . the PRSP Donor-supported programs are to launched the PRS Initiative to help low-income be derived from a country’s PRSP to ensure that countries develop and implement better strate- they are country-owned and clearly oriented to gies to fight poverty—and to make aid more achieving the joint objectives of poverty re- effective in reducing poverty. duction and economic growth. The PRS Initiative promotes a country- driven, comprehensive, and long-term process In promoting the PRS Initiative, the World that is coordinated with external partners and Bank and IMF agreed to transform the way they 1 T H E P O V E R T Y R E D U C T I O N S T R AT E G Y I N I T I AT I V E operate in low-income countries and align their The PRS Initiative complements the Heavily assistance directly with the goals articulated in Indebted Poor Countries (HIPC) Initiative, an member governments’ PRSPs (box 1.1). The international effort to reduce the external debt PRSP would thus become the centerpiece for of low-income countries to sustainable levels, by World Bank and IMF relations with low-income helping ensure that resources freed up by debt countries. The Bank and Fund also agreed to relief would be used for public sector programs enhance the complementarity of their develop- to reduce poverty. To obtain permanent debt ment instruments in support of the PRSPs of relief under the HIPC Initiative, the 38 qualifying their member governments. countries are required to produce an initial PRSP For the World Bank, the PRS Initiative means and implement it successfully for a year, as supporting country efforts for planning, reflected in an Annual Progress Report.3 implementing, and monitoring poverty Although it is too early to assess with reduction strategies through its various forms certainty the success of the new framework, it of assistance—lending programs as well as was an appropriate time to take stock of advisory and analytical activities—and using its progress to date and identify shortcomings that experience, influence, and convening power to may require corrections in the design and mobilize other international financial institu- implementation of the PRS Initiative. Therefore, tions and bilateral donors to join the Initiative. the independent evaluation offices of the two It also means supporting implementation of institutions—OED and IEO—decided to assess the PRSP through alignment of assistance from progress of the PRS Initiative and institutional the International Development Association support from the Bank and IMF. (IDA, the World Bank’s concessional lending Building on internal reviews of the PRS arm) with country priorities and with the Initiative undertaken by IMF and World Bank assistance programs of other external partners. staff, as well as by donor agencies and For the IMF, alignment meant creating a new nongovernmental organizations, OED and IEO vehicle for the provision of IMF concessional gathered information independently to assess lending. The Poverty Reduction and Growth the PRS Initiative against its objectives and to Facility (PRGF) gives a more central role to gauge the effectiveness of Bank and IMF poverty reduction and pro-poor growth consid- support for the PRS approach. The two evalua- erations in the design of IMF-supported tion units completed separate but parallel programs in low-income countries than did its reports, focused on intermediate outcomes predecessor, the Enhanced Structural Adjust- through 2003—the quality of the policy ment Facility (box 1.2). formulation process, the nature of the strategy and policy framework that has evolved, and the interaction between this framework and the Five principles support of external partners. of the Poverty The OED and IEO evaluations, discussed by Box 1.1 Reduction the Boards of the Bank and IMF in July 2004, Strategy concluded that the World Bank and IMF should Initiative continue to support the PRS process, while making changes in three main directions: • Country-driven, involving broad-based participation • Comprehensive in recognizing the multidimensional nature of • Reduce or eliminate uniform requirements, poverty and proposing a commensurate policy response curb excessive emphasis on documentation, • Based on a long-term perspective for poverty reduction and encourage customization of PRSPs to fit • Results-oriented and focused on outcomes that benefit the poor country circumstances. • Partnership-oriented, involving coordinated participation of de- • Encourage PRSPs to explore a wider range of velopment partners. policy options, including those aimed at en- hancing growth. 2 INTRODUCTION Principles underlying the IMF’s Poverty Box 1.2 Reduction and Growth Facility PRGF-supported programs have seven key features: ments of higher aid flows should be sought and built into the • Broad participation and greater ownership. The main ele- program. ments of the PRGF are drawn from the country’s PRSP. • Financing conditions are limited to key measures central to • Embedding of the PRGF in a national strategy for growth and the success of the strategy. poverty reduction so that macroeconomic and other poli- • Measures to improve public resource management and ac- cies reflect the country’s growth and poverty objectives. countability. Fiscal policies and objectives should be open • Budgets that are more pro-poor and pro-growth. Govern- to public debate. Monitoring systems should be developed ment spending should be reoriented toward activities that to improve efficient delivery of public services. benefit the poor. Tax reforms that simultaneously improve ef- • Social impact analysis of major macroeconomic adjust- ficiency and equity should be stressed. Data and monitoring ments and structural reforms. The distributional effects of sub- to track expenditures should be improved. stantial macroeconomic adjustments or structural reforms • Flexibility in fiscal targets. Macroeconomic projections should be considered and countervailing measures identi- should signal financing needs. Where warranted, commit- fied to offset temporary adverse effects on the poor. • Help define clearer partnership frameworks countries had completed full PRSPs by the end of around PRSPs, with the required accountabil- 2002), geographic balance, coverage of HIPC and ity for both countries and partners. This in- non-HIPC countries, and initial conditions in the cludes specific changes for the assistance of the country. The studies cover the 10 countries’ World Bank and the IMF. experience with the PRS Initiative up to mid-2003 and in some cases through the end of 2003. The two evaluations also made several The chief purpose of the studies was to recommendations with respect to the specific evaluate the internal processes and perform- roles of the World Bank and IMF; these are ance of the World Bank and IMF in supporting discussed in the main evaluation reports. the PRSP Initiative, not to appraise the policies of member governments. Because they cover The scope of the present report at most three years of implementation, the The OED and IEO evaluations of the PRS Initia- studies do not speculate on long-term tive were informed by 10 in-depth country case prospects for growth and poverty from the studies and surveys of stakeholders in each Initiative, although some information is country. Summaries of those case studies make presented on short-term growth outcomes. up the body of this report, which is designed to Four of the case studies were undertaken complement the main evaluation reports and jointly by OED and IEO (Mozambique, Nicaragua, provide the reader with a deeper look at individ- Tajikistan, and Tanzania), two by IEO alone ual country experiences under the Initiative. The (Guinea and Vietnam), and four by OED alone evaluations also drew on a range of other (Albania, Cambodia, Ethiopia, and Mauritania). evidence, including cross-country analyses and The teams undertaking the case studies desk reviews. Readers are referred to the main followed guidelines developed jointly by OED reports for discussions of that evidence, since and IEO at the outset of the evaluations. Those this background volume focuses on the evidence guidelines were applied for all 10 case studies that emerged directly from the case studies. (annex). All 10 case studies covered three main The case study countries were chosen to cover areas: relevance of the PRS approach, the extent a variety of country situations and stages of PRSP to which the underlying PRS principles were and PRGF implementation (table 1.1). Selection applied, and preliminary evidence on the results. criteria were: maturity of the PRSP process (all 10 In addition, the effectiveness of World Bank 3 T H E P O V E R T Y R E D U C T I O N S T R AT E G Y I N I T I AT I V E Te n c o u n t r y c a s e s t u d i e s o f t h e Ta b l e 1 . 1 PRS Initiative Year of Implementation # of PRSP Year of HDI Rankinga HIPC Case? PRSP yearsb prog reportsb new PRGF Joint IEO/OED c Mozambique 170 Yes 2001 2.3 1 Nicaragua 118 Yes 2001 2.3 2 2002 Tajikistan 112 No 2002 1.1 0 2002 Tanzania 151 Yes 2000 3.1 2 2000 IEO only Guinea 159 Yes 2002 1.4 0 2001 Vietnam 109 No 2002 1.6 1 2001 OED only Albania 92 No 2001 1.6 1 2002 c Cambodia 130 No 2002 0.9 0 Ethiopia 168 Yes 2002 1.3 1 2001 Mauritania 152 Yes 2000 2.9 2 2003 a. The Human Development Index (HDI) measures the quality of life against a range of social indicators. See United Nations Development Program, Human Development Report (New York: United Nations, 2002). b. As of end-2003. c. New arrangement approved under ESAF regime in 1999. support was assessed in the case studies in which institutional responses—not individual views— OED participated (8 of 10). Similarly, the on the PRSP and World Bank and IMF perform- effectiveness of IMF support was addressed in the ance in supporting the Initiative. case studies in which IEO participated (6 of 10). The study teams gave stakeholders the Country evaluation teams included local opportunity to comment on their preliminary consultants, recruited to provide local knowledge findings, and in most cases in-country workshops and insight, as well as to administer a survey of were held at the end of the field visits. The stakeholders. Interviews with relevant members of governments of the countries were asked to the World Bank and IMF staff were followed by in- comment on the full case study reports, which country consultations with a broad range of reflect their remarks. However, the final stakeholders in key national ministries, local judgments are the responsibility of OED and IEO. governments, bilateral and multilateral donors, local and international nongovernmental organiza- Summary findings from the 10 country tions (NGOs), representatives of the private sector case studies and media, and, wherever possible, members of parliament. Field visits outside the capital were Progress varies with individual country made, except in the case of Mauritania.4 circumstances and is constrained by Using a survey instrument jointly designed preoccupation with formal requirements and by OED and IEO, stakeholders in each country tensions regarding domestic ownership were surveyed to supplement information gained from the interviews. Respondents were • The degree of success of the PRS Initiative has drawn from various stakeholder groups varied widely across countries, depending involved in the PRS process—including govern- chiefly on the initial conditions in each coun- ment, civil society, private sector, and external try’s public sector capacity, the quality of gov- partners. The survey was designed to obtain ernment-partner relationships, and relations 4 INTRODUCTION among donors in each country. The Initiative nomic framework and choice of policies to has added the most value in countries where promote growth. Insufficient prioritization, in- government leadership and aid management adequate costing, and a tendency to avoid con- processes were already strong. It has had less troversial structural reform issues have effect in countries with weak capacity or with prevented most PRSPs from providing an ad- donor-dominated aid relationships. equate framework for making decisions on • In a few countries, PRS processes have become key policy tradeoffs. Thus, most PRSPs do not sufficiently entrenched in domestic institutions yet provide an effective strategic road map. In to be self-sustaining. In most cases, however, the some cases, progress has been made over time process has not proved able to adapt to unan- in ameliorating these shortcomings, mainly in ticipated developments—such as changes in countries that had begun embedding the PRS macroeconomic conditions—and monitoring process in domestic institutions. of implementation does not yet inform policy • Most PRSPs have not considered the full range design. In cases where the PRS approach has of policy actions required for growth and been integrated into domestic institutional poverty reduction, focusing instead on public structures, there have been important im- expenditures, especially social sector spending, provements in domestic policy processes. with less emphasis on other aspects of a • Preoccupation with formal PRSP requirements broader strategy to encourage poverty-reduc- that give countries access to resources has often ing growth, such as infrastructure and rural de- been at the expense of adaptation of the PRS velopment. The approach has so far not process to country circumstances. This was re- contributed significantly to understanding the flected in too much emphasis on producing linkages among growth, poverty incidence, documents, rather than improving underlying and macroeconomic policies at the individual policy processes. The initially ambitious timetable country level. for completion of first-round PRSPs in HIPC and • Even in the area of public expenditure, the PRGF countries reinforced the tendency to gen- practical value of many PRSPs has often been erate standardized strategies in the early PRSPs. limited, because of the still rudimentary nature The PRS approach should be implemented in a of most costing and prioritization. manner that capitalizes on existing strengths • In most cases, the IMF has had limited success and focuses efforts on missing ingredients, rather in embedding the PRGF in broader strategies than starting anew in all areas. for growth and poverty reduction—and in in- • It is not clear how conflicts between domestic ducing client countries to integrate PRGF-sup- ownership and assessments of policies by the ported programs with strategic priorities and Bank and Fund will be resolved. When such policy choices in both macroeconomic and conflicts have arisen, they have been resolved structural areas. This limited success reflects in an opaque manner, leaving countries to di- both weaknesses in the PRSPs as strategic vine what is “acceptable� to the Bank and Fund. roadmaps and ambiguities about the IMF’s role. The process of presenting a PRSP to the Boards of the Bank and the Fund has been perceived The PRS Initiative has led to increased as undermining the principle of country own- emphasis on dialogue and results ership—as “Washington signing off � on a sup- posedly country-owned strategy. • The PRS process has begun to orient stake- holders toward results, although the develop- PRSPs do not yet provide an adequate ment of country-specific indicators and analytical framework for policy choice and monitoring systems to track them was still at prioritization a preliminary stage in most countries. Available information generally was not linked to deci- • The analytical bases of many PRSPs are weak, sionmaking. most notably with regard to the macroeco- • The requirement that PRSPs must contain a 5 T H E P O V E R T Y R E D U C T I O N S T R AT E G Y I N I T I AT I V E multidimensional diagnosis of poverty has quality and usefulness of the Bank’s support highlighted the paucity and poor quality of and its relations with partners. data in many countries. In some cases this re- • The content of Bank assistance strategies for- quirement had already resulted in marked im- mulated subsequent to PRSPs generally over- provements in data quality. laps with the content of PRSPs. But since PRSP • The process of aligning sector programs under programs are broad and often not well prior- one strategic umbrella has helped policymak- itized, this overlap has not entailed major ers better appreciate the links between sector changes for the Bank. strategies and poverty reduction. Sectoral • IMF staff often interpreted the principle of coun- spending plans have been challenged to meet try ownership as precluding or at least limiting poverty reduction criteria. But, as noted, most their involvement in policy debate on the PRS PRSPs have failed to integrate sector strate- process among domestic stakeholders as well gies with the macroeconomic framework as in ensuring that key macro-relevant issues through a rigorous consideration of the sources were aired during the formulation of the PRS. of growth and other macro-micro linkages. • There are signs that the “policy space� in the • Extensive consultations conducted in most PRSP macroeconomic area has widened—in the countries during strategy formulation have sense of greater openness on the part of the brought new actors and increased transparency IMF to considering alternative country-driven to the development dialogue (but with rela- policies—at least in countries where macro- tively little change in discussions of the macro- economic stabilization is no longer a pressing economic framework and related structural issue. Program design under the PRGF has in- reforms). But civil society organizations in some corporated greater fiscal flexibility to accom- countries consider that this openness has had modate aid flows. However, there has been only limited impact on the design of PRSPs and no effective operational approach to managing on subsequent domestic policy, and that par- the tensions between “ambition� and “real- ticipatory activities have often waned once the ism� in defining medium-term external re- PRSP was formulated. Furthermore, participa- source envelopes. tory processes were typically not designed to • Donors (including the Bank and Fund) in- strengthen existing domestic institutional creased their support for national budgets and processes for policy formulation and account- streamlined the monitoring of performance, ability (through parliament, for example). In a but there is little evidence that they have few cases, institutional arrangements to sustain aligned aid programs with PRSP priorities and the process have begun to develop around the achieved greater interagency coordination. budget cycle. In general, the PRS process has not Where government-donor dialogue is weak or generated meaningful discussions—outside the donors continue to drive the agenda, PRSPs by narrow official circle—of alternative policy op- themselves have not noticeably promoted tions with respect to the macroeconomic frame- donor coordination or increased government work and structural reforms. management of external assistance. • Bank-Fund collaboration has improved due to The role of the World Bank, the IMF, and other the Initiative, with more frequent discussions donors is changing—but not as rapidly as between staff of the two institutions at the expected country level. While working smoothly to- gether in PRSP formulation and review in the • World Bank staff supported national stake- case study countries, the Bank and the Fund holders preparing PRSP documents by pro- had policy differences on the scope and se- viding analyses, commenting on drafts, and quencing of structural reforms in some coun- promoting participatory activities. Clients ob- tries that the PRSP framework alone could not served that decentralization of Bank staff and overcome.5 IMF and Bank contributions to de- authority to country offices has enhanced the veloping a better understanding of country-spe- 6 INTRODUCTION cific micro-macro linkages have also been fairly candor, and comprehensiveness of the as- limited. Although both institutions place greater sessments have been uneven, with scope for emphasis than before on poverty and social im- improvement even in “good practice� cases. pact assessment (PSIA), they have not yet de- Their main contribution has been in giving vised a way to ensure that the priorities for such feedback to the authorities on weaknesses in assessments (especially in the area of macro- , the PRSP but JSAs are virtually unknown out- economic policy) will reflect the strategic pri- side the narrow official circle and consequently orities of the PRSP and give clear indications of have had no impact on the broader policy de- what is expected from each institution. bate. They have not incorporated systematic in- • Joint Staff Assessments (JSAs) of PRSPs by the puts from development partners and, in Bank and Fund have not adequately performed practice, have played a limited role in inform- the many tasks expected of them. The clarity, ing lending decisions. 7 2 Albania T he OED evaluation team visited Albania from March 12 to 26, 2003, and met with stakeholders in Tirana, Dürres, Elbasan, Lezhe, and Shkodra. Among the sources for the analysis presented here was a survey of 97 stakeholders representing government, civil society, and international partners. This country case study provides an assessment of the PRSP process in Alba- nia as of mid-2003, and covers both PRSP formulation and implementation. Growth depends on private sector schemes led to a breakdown of public order. development Supplies of safe water and reliable power are Poverty in Albania is a multifaceted phenome- inadequate throughout the country. Roads, non. While higher in rural areas, poverty in sanitation, and telecommunications are in urban settings has risen substantially in recent disrepair. years. Using the methodology developed by Spending on health, education, agriculture, the 2002 Living Standards Measurement Survey, and local government has increased in recent 25 percent of Albanians were below the poverty years, but Albania’s share of spending for line. Basic infrastructure and public services health (2.2 percent of GDP) and education (3.2 have not kept pace with these changes. In percent of GDP) remain below the averages for particular, public order and access to health comparable lower-middle income countries. care have emerged as areas of great concern in Albania remains the poorest country in Europe. densely populated peri-urban areas. High The structure of the Albanian economy has unemployment and underemployment are changed fundamentally over the past decade considered principal causes of poverty. (figure 2.2). In 1990 Albania emerged from its The poor condition of basic infrastructure long post–Cold War stagnation and initiated a contributes to Albania’s low standard of living rapid transition program. Between 1993 and and is a major deterrent for both domestic and 1996 the economy grew by more than 9 percent foreign investment in the country. After years of per annum. Agricultural output and new service neglect of the country’s infrastructure during activities expanded rapidly, and their growth the period of central planning, conditions outweighed the decline in production and further deteriorated after 1997, when the employment in heavy industries that were a collapse of widespread pyramid investment legacy of the earlier period. 9 T H E P O V E R T Y R E D U C T I O N S T R AT E G Y I N I T I AT I V E Figure 2.1 Albania at a glance Europe & Lower- POVERTY AND SOCIAL Central middle- Albania Asia income Development diamond 2003 Population, mid-year (millions) 3.2 473 2,655 Life expectancy GNI per capita (Atlas method, US$) 1,740 2,570 1,480 GNI (Atlas method, US$ billions) 5.5 1,217 3,934 Average annual growth, 1997–03 Population (%) 0.3 0.0 0.9 Labor force (%) 0.8 0.2 1.2 GNI Gross per primary Most recent estimate (latest year available, 1997–03) capita enrollment Poverty (% of population below national poverty line) 25 .. .. Urban population (% of total population) 44 63 50 Life expectancy at birth (years) 74 69 69 Infant mortality (per 1,000 live births) 22 31 32 Child malnutrition (% of children under 5) 14 .. 11 Access to improved water source Access to an improved water source (% of population) 97 91 81 Illiteracy (% of population age 15+) 1 3 10 Gross primary enrollment (% of school-age population) 107 103 112 Albania Male 107 104 113 Lower-middle-income group Female 107 102 111 Source: World Bank. The changing structure of Albania’s Figure 2.2 economy (percent of GDP) 18% 33% 32% 3% 8% 54% 17% 11% 13% 11% 1992 2002 Agriculture Industry Construction Transport Services/Other Source: Republic of Albania, “Progress Report for Implementation 2002, Objectives and Long Term Vision of the NSSED, Priority Action Plan 2003� (Tirana, 2003) Section C, p. 3; and Re- public of Albania, “Medium-Term Expenditure Framework 2003-5� (Tirana, 2002), p. 16. The first phase of this transition came to an fundamental changes were taking place. abrupt halt in 1997 with the collapse of the Domestic uncertainty and disillusionment had financial schemes in which many Albanians had prompted many young Albanians to emigrate. invested money. The collapse and subsequent Remittances from these emigrants, as well as breakdown of public order set back the earnings from criminal activities channeled achievements of the previous five years. In through southeastern Europe, were beginning 1997 GDP fell by 7 percent (before returning to to flow into the country through informal and strongly positive growth in 1998–2001), and often illegal channels. In association with these inflation reached 42 percent. At the same time, inflows, corruption in government grew. 10 ALBANIA Since 1997 growth has been based largely on and submitted to the Bank and IMF Boards in strong private sector activity, including a boom June 2002, the PRSP was adopted in late 2002 in construction services linked in part to under the title National Strategy for Social and growing urbanization. The maintenance of Economic Development (NSSED). macroeconomic stability, with low inflation and Achieving sustainable and inclusive growth a fairly stable exchange rate, has been an is the fundamental objective of the NSSED. important contributor to the post-1997 Improved governance and growth and private recovery. sector development are its two main pillars. The current environment for private sector European integration is specifically recognized development, however, may impair further as Albania’s long-term objective. The NSSED is growth. A survey by the World Bank and the closely linked to the Medium-Term Economic European Bank for Reconstruction and Framework (MTEF), in order to orient govern- Development ranked Albania last, among 25 ment spending toward poverty reduction and transition economies, in the perceived quality to coordinate donor financing, it has five key of the business climate, with the security of goals and identifies six priority sectors for property rights and weaknesses in the legal public action (box 2.1). framework as the areas of greatest concern.1 Institutions—including the courts and police— Ownership, participation are weak, and their effectiveness is undermined Given its limited capacity, the government used by growing the growing influence of powerful local consultants to prepare the PRSP interest groups, including organized crime. document, but the Albanian authorities played an active role in the planning process. The The PRS process—better development consultants worked closely with government planning officials to ensure that the respective ministries The PRS process proved to be highly relevant would own the document. The consultations for Albania’s political process. The 1997 crisis were well planned and reached a wide led to a loss of confidence in the Albanian spectrum of civil society representatives. government. The PRS process helped fill the Discussions were open and reported in the void. As the first comprehensive development media. Throughout, the government put a strategy undertaken by the government great deal of emphasis on monitoring and through broad consultations and in collabora- evaluation within individual ministries and at tion with aid partners, it signaled a transition the central level. from humanitarian aid and stabilization to Ownership of the PRSP outside the govern- emphasis on growth and development. The ment and civil service remained uneven. In achievements of the process have been consid- general, the evaluators found significantly less erable, and the PRSP has added value to evidence of ownership outside the capital. Albania’s development planning and manage- Public participation had not been a major part ment. But much remains to be done to bring of the Albanian political process prior to the budgetary processes and external assistance PRSP, and the government had little idea of how under the PRSP umbrella. to involve civil society organizations in the The government insisted on giving equal process. Consultations with the private sector emphasis in the PRSP to issues of economic were the least satisfactory among all stakehold- growth, concerned that the process might ers. Private businesspeople and their represen- overemphasize poverty and thereby slow tative organizations came to the sessions progress toward joining the European Union seeking to discuss urgent issues affecting their (EU). In late 2000, after the Interim PRSP was own enterprises, rather than a broad poverty- presented to the Bank and Fund Boards, the reduction strategy. They became dissatisfied PRS process was renamed the Growth and with the process when their concerns were not Poverty Reduction Strategy. Drafted in 2001 addressed and they stopped participating. 11 T H E P O V E R T Y R E D U C T I O N S T R AT E G Y I N I T I AT I V E K e y e l e m e n t s o f A l b a n i a ’s N a t i o n a l S t r a t e g y Box 2.1 for Social and Economic Development The main objectives of the NSSED are: Six priority sectors of public actions: • Annual real GDP growth of about 7 percent during 2002–2004 • Health and education • Reduction in the number of people living in poverty, in par- • Infrastructure ticular, the most affected groups • Improvement of governance • Tangible improvements in infrastructure and related services • Economic growth and increasing the access of the poor to these services • Social support • Reduction of infant and maternal mortality rates and infec- • Macroeconomic stability. tious disease incidence • Increase of elementary and secondary school enrollment rates. Comprehensiveness, long term Partnerships The lack of specific and detailed prioritization in Partnership and donor alignment have been the PRSP has been a major weakness in the weak aspects of the PRS process. Although approach. To some extent that lack reflects the Albania has received large amounts of aid multidimensional nature of poverty in the country relative to its size, and is of geopolitical and the need to work simultaneously on a wide importance to a number of key donors, donors range of fronts. But it also reflects a lack of clarity have not been willing to align their agendas about the core objectives to be achieved and with the PRSP, with some perceiving the PRSP measured. Sector programs tend to be an as a Bank-driven process. Mirroring those unordered list. In the health sector, for example, tensions, the government has been unable to clear targets are set, but they are overambitious establish a clear approach to the structure and and cover a wide range of areas. Survey respon- form of its aid coordination activities. dents agreed that the targets and plans in the PRSP The PRSP planning process provided an were not realistic. opportunity for donors to create more focused Private sector development will be key to and coordinated programs of support for Albania’s future, but the PRSP does not do an Albania. However, it closely followed the effective job of integrating the components launching of the EU’s Stabilization and Associa- needed for effective private sector development tion process in May 1999, which introduced the (notably programs in governance, judicial reform, prospect of EU membership to the Balkans and trade policy, financial sector development, tax support through the Community Assistance to administration, infrastructure provision, and Reconstruction, Development, and Stabiliza- privatization of basic services). tion program. The goal of EU accession—and The PRSP also is deficient in addressing the priorities and processes devised by the EU decentralization. The government is decentralizing to reach that objective—were a high-value in a piecemeal way, with different ministries taking competitor for the government’s attention and different approaches. Local governments have not for the scarce time of key officials. been adequately consulted, briefed, or trained to Donor coordination problems in Albania implement decentralization. The PRSP could have antedate the PRSP, with donors preparing and been used more effectively to provide a road map implementing programs independently of each for the decentralization program, which would other. As a consequence of the resulting indicate to ministries, local governments, and overlap, the sizeable donor effort in Albania is communities how the process is to unfold and how not living up to its potential. Donors are not capacity deficiencies are to be remedied. complementing each other’s efforts; legitimate 12 ALBANIA differences of view are not being resolved in a The PRSP has taken root in the overall systematic fashion; and donors are crowding budget process, which is linked to the annual into a few areas, such as governance, anticor- budget through the MTEF. Although this ruption, and crime, leaving critical infrastruc- process is still at an early stage, it has ture needs unmet. contributed to capacity enhancement in areas Different views among donors on the such as statistical development, monitoring relative roles and responsibilities of the and evaluation, and support for public debate ministries of finance and economy have long on development issues. complicated even basic planning tasks, such as Since the adoption of the PRSP, the govern- compiling a database on past and current aid ment has increased the share of pro-poor commitments. The publication of the Millen- expenditures in the budget. Most of these nium Development Goals (MDGs) early in increases have been used to pay higher Albania’s PRS process was another complicat- salaries to workers in health and education. ing factor. Given its mandate for monitoring Spending for maintenance of schools and progress on the MDGs, the United Nations hospitals also has risen, although it has been Development Program (UNDP) has tended to unevenly distributed. view the PRSP in Albania as a key element toward the achievement of the MDGs rather World Bank support—emphasizing than as a free-standing poverty reduction country ownership strategy and coordinating framework. The Bank made a substantial commitment to To a large extent, the sustainability of the Albania’s PRSP over the past three years. The PRSP will depend on the donor community. Bank played an effective and appropriate role Only if donors align their programs to support in support of consultations related to the PRSP. the PRSP will the central and local governments Although it was unduly dominant in the and communities see the advantages of formulation of the Interim PRSP, the Bank developing their own programs under the PRSP stepped back when the full PRSP preparation umbrella. began so that the government could take charge. At the same time, it kept the govern- Results orientation ment focused on the PRSP through joint work Improvements in the analytical base are a first on the Poverty Reduction Support Credit step to better prioritization. The development (PRSC), development of monitoring and of indicators and their measurement are critical evaluation capacity, and a consistent dialogue steps in improving the analytic base that makes at the technical level. informed prioritization possible. The PRS Emphasizing the need to improve data process has helped the government grasp the collection and monitoring, the Bank worked importance of monitoring and evaluation and to with Albania’s Ministry of Labor and Social make progress in establishing the structures and Affairs and the National Statistical Institute to indicators to monitor the NSSED. The selection prepare a poverty profile and a qualitative of monitorable and measurable indicators is all poverty assessment formulation of the PRSP. the more important given the weakness of the In the future, it will be important to lower still statistical data in Albania and the difficulty of further the Bank’s public profile on the PRSP using the overall statistical system to measure and to stress at every opportunity that the PRS progress. In September 2002 the prime minister process is meaningful only as a country-owned decreed that monitoring units should be process. Along with other donors, the Bank has established in each line ministry and that indica- taken steps to increase the role of parliament. tors should be established for use in policy And, as a result of PRS-related discussions, the formulation. A unit under the Council of Organization for Security and Co-operation in Ministers is now responsible for policy analysis Europe agreed to provide long-term assistance and monitoring and evaluation. to strengthen the role of the Albanian parlia- 13 T H E P O V E R T Y R E D U C T I O N S T R AT E G Y I N I T I AT I V E ment in the PRS process. Albanian MPs are to the issue has been the country’s poor become members of the Ministerial Monitoring governance record and concern about corrup- Committee and the technical committee on tion in public procurement and use of cross-sector monitoring. resources. The Bank’s lending program, distributed Albania’s PRSC was the first lending over a large number of sectors, has not operation of its kind in the Bank’s Europe and supported the development of a strategic focus Central Asia Region. It has four areas of focus: for the PRSP, although it has become more promoting growth and private sector develop- selective in the past three years, with the ment; strengthening capacity to monitor and advent of the PRSC (figure 2.3). A broad evaluate the policy agenda; improving service approach was understandable, given the huge delivery and the effectiveness of the social needs for infrastructure and the shortage of safety net; and improving core public sector alternative sources of donor support. But even functions and institutional arrangements. so, the Bank program lacked the strategic Trigger points in the PRSC that link lending to underpinning that could have made the benefit country performance are clearly related to of the whole program exceed the sum of PRSP objectives. The second phase of the PRSC projects’ individual benefits. The argument for builds on the progress made in the first phase, a multi-project strategy (as opposed to shifting while addressing shifts in priorities noted in resources to budget support) has been the the PRSP Progress Report. need for intensive work at the sectoral level World Bank management has made consis- and the difficulty of engaging specific line tent efforts to promote more effective aid ministries through budget support. Sector- coordination in Albania. The Bank has sought wide approaches (SWAPs) provide an option to work collaboratively with the U.K. Depart- for more direct involvement at the sector level ment for International Development and other while maintaining overall budget support, but bilateral donors in supporting the PRSP, and it the Bank has not opted to go this route. Part of was effective in persuading the EU that the Evolution in World Bank lending to Figure 2.3 Albania, 1992–2004, by sector Sector Rural Transport Economic Policy Social Protection Public Sector Governance 1992–01 Energy and Mining 2002–04 Private Sector Development Health, Nutrition, and Population Urban Development Water Supply and Sanitation Education Financial Social Development 0% 5% 10% 15% 20% 25% 30% 35% Share of commitments Source: World Bank. 14 ALBANIA PRSP was well aligned with the EU accession must address to consolidate its gains under the program. Through the end of 2003, however, PRS Initiative. the Bank generally had not been able to induce Progress has been made under the PRSP in donors to coordinate their efforts with the the area of monitoring. The current approach PRSP. The Bank might have done more to use combines the commitment of the government, its own program in support of donor alignment substantial support from the Bank, and training. by promoting sector-wide approaches and The process could be broader, with the Bank increased budget support. The government playing a role in promoting credible involve- also could have provided leadership in resolv- ment of NGOs and other nongovernment ing the conflict over responsibilities for aid actors with sufficient time to review documents coordination held by the ministries of finance and prepare comments. At the same time, while and the economy. public awareness was high at the time the PRSP The Bank, with the IMF, should give serious was launched, it faltered during implementa- thought to opening up procedures such as the tion. The monitoring unit in the Ministry of JSA to the participation of other partners, Finance could hold regular briefings with the including senior officials from other PRSP press to help address this issue. countries with first-hand knowledge of the A key weakness of Albania’s PRSP lies in process. interministerial coordination of multisector programs. Both the government and the Bank Bank-Fund collaboration need to give thought to handling these Coordination between the Bank and IMF at programs effectively. The PRSP structure and headquarters and in the field has been an processes, including sectoral and regional important direct benefit of the PRSP to Albania. strategies and action plans, must be sustained Most observers agree the approach taken by the through institutionalization. two institutions to balance spending to reduce Donor coordination is too important to be poverty and controlling the budget deficit has left to the donors. The Albanian government become more consistent. The same is true of should assume responsibility for the process. the balance between the need for increased tax After identifying an appropriate point of revenues and the imperative of maintaining coordination, the government should convene incentives for private sector activities. donors in a series of meetings to present the government’s strategy and priorities, followed Conclusion—building credibility, by monthly technical meetings co-chaired by coordinating aid the appropriate ministries to assess progress Albania’s PRSP has been a moderately success- and programs in the sectors. For their part, ful experience for both the government and donors must reduce the transaction costs of the Bank, and the experience offers useful their assistance to Albania through increased lessons for other countries. But there are co-financing, fewer project units, and serious issues that the Albanian government harmonized procedures. 15 3 Cambodia T he government of Cambodia began implementation of the PRSP in March 2003, shortly before the OED evaluation team visited the coun- try in April/May 2003. Thus, this chapter focuses only on the process of PRSP formulation and the genesis of the implementation phase. The team held meetings in Phnom Penh, and in Siem Reap and Prey Veng provinces. Among the sources for the analysis presented here was a survey of 73 stake- holders representing government, civil society, and international partners. A long road from conflict to growth Despite GDP growth averaging nearly 6 The effects of Cambodia’s turbulent past percent annually since 1994, Cambodia’s remain prominent in the country. Civil conflict economic fundamentals remain weak. Agricul- caused widespread poverty, shattered institu- ture is the dominant sector, accounting for 39 tions, and robbed the country of many of its percent of GDP and more than 70 percent of best-trained citizens. There is an acute shortage employment. Growth has been driven by of skilled people capable of rebuilding the expansion of the area under cultivation and, country and fighting poverty. Fifty percent of more recently, by tourism and the rapid Cambodia’s 13 million people are under 16 expansion of garment production as a years of age; the population growth rate of 2.4 consequence of Cambodia’s preferential access percent per annum imposes a heavy burden on to U.S. markets under the Multi-Fiber the country’s limited social services. Agreement. That preferential access expired at Cambodia is among Asia’s poorest the end of 2004, and a decline in garment countries. Some 30–40 percent of its people production and exports could severely depress survive on a dollar a day or less. Life expectancy domestic employment. is 54 years, infant mortality is 95 per 1,000 live Spending on health and education more births, and the rate of secondary school enroll- than doubled from 1993 to 2002, reflecting a ment is just 14 percent. Cambodia also has the commitment on the part of the government to highest prevalence of HIV infection in East Asia. address the basic service needs of the popula- Most Cambodians live in the rural areas where tion. Despite this build-up, Cambodia still lags the rate of poverty is almost double that of the behind regional averages in both sectors for urban population. domestic expenditures. 17 T H E P O V E R T Y R E D U C T I O N S T R AT E G Y I N I T I AT I V E Figure 3.1 Cambodia at a glance East POVERTY AND SOCIAL Asia & Low- Cambodia Pacific income Development diamond 2003 Population, mid-year (millions) 13.4 1,855 2,310 Life expectancy GNI per capita (Atlas method, US$) 310 1,080 450 GNI (Atlas method, US$ billions) 4.2 2,011 1,038 Average annual growth, 1997–03 Population (%) 2.9 1.0 1.9 Labor force (%) 2.5 1.1 2.3 GNI Gross per primary Most recent estimate (latest year available, 1997–03) capita enrollment Poverty (% of population below national poverty line) 36 .. .. Urban population (% of total population) 19 40 30 Life expectancy at birth (years) 54 69 58 Infant mortality (per 1,000 live births) 96 32 82 Child malnutrition (% of children under 5) 45 15 44 Access to improved water source Access to an improved water source (% of population) 30 76 75 Illiteracy (% of population age 15+) 31 10 39 Gross primary enrollment (% of school-age population) 123 111 92 Cambodia Male 130 112 99 Low-income group Female 116 111 85 Source: World Bank. The capacity in government to manage national strategies in 2000–2001. Development development efforts is fairly limited. Salaries are specialists regard these as sound strategies, low and corruption is prevalent. Rents can be with a high degree of ownership of the sought to supplement incomes from conces- ministries concerned, and consider the sions on agricultural land and forests, from the planning capacity in these ministries to be tourism and garment industries, and from the among the best in the government. substantial amounts of foreign aid flowing into the country. Political appointments that give The PRS process—awkward timing, access to rents are part of a patronage system incomplete ownership that can make it difficult to ensure that qualified The PRS Initiative is highly relevant to the individuals occupy key positions in government. poverty challenge in Cambodia. The Initiative’s Poor infrastructure limits growth and focus on government ownership and public poverty reduction by inhibiting exports of local participation also strikes an important chord, agricultural products and imports of inputs given the strongly donor-driven development such as fertilizer, while preventing poor people agenda since the post-conflict era started in the in villages from accessing health care and early 1990s. But the timing of the PRSP prepara- education. Less than a third of the population tion was awkward. Initial plans for an ambitious has access to clean drinking water and to irriga- document deadline competed with preparation tion for agricultural growth. Only 10 percent of of the legally mandated SEDP, which was all households have access to electricity. supported by the Asian Development Bank. The Cambodia issued its first post–Khmer Rouge decision to complete two parallel planning Development Plan in 1996. The Socio- processes was unrealistic given limited govern- Economic Development Program I (SEDP I) ment capacity, and PRSP formulation was delayed covered the period from 1996 to 2000. SEDP I to follow on the heels of completion of SEDP II. focused on growth and infrastructure, but not The document was completed in December explicitly on poverty reduction. After the plan 2002 (box 3.1). This early planning experience was adopted, donors provided substantial reflects preoccupation with the formal require- technical assistance to the ministries of health ment of a strategy paper and not on promoting and education. Both ministries developed the Initiative’s underlying principles, and 18 CAMBODIA C a m b o d i a ’s N a t i o n a l P o v e r t y R e d u c t i o n Box 3.1 Strategy: broad priorities and action areas The focus of the PRSP is on fiscal discipline, combined with rev- The PRSP identifies eight priority areas for poverty re- enue strengthening and structural reforms to improve the busi- duction: ness environment. • Maintaining macroeconomic stability The PRSP has three broad priorities: • Improving rural livelihoods • Expanding job opportunities • Growth through exports, tourism, and agriculture develop- • Improving capabilities ment • Strengthening institutions and improving governance • Governance and public expenditure management • Reducing vulnerability and strengthening social inclusion • Infrastructure and human development. • Promoting gender equity • Priority focus on population. Source: Kingdom of Cambodia (2002), NPRS 2003–05. underscores the importance of carefully tailoring Linking strategy and planning with resource to existing domestic processes. allocation is a key marker of country commit- ment to the Initiative. Albeit at a very early Ownership, participation stage of implementation, at the time of the There was strong ownership of the PRSP—of evaluation mission there was little evidence both the document and the process—among a that the PRSP in Cambodia was linked to the small group in the Ministry of Planning (MoP). annual budget and would affect resource This unit took leadership of the PRS process allocation. To achieve an operational strategy and liaised effectively with planning staff in would have required the MEF to play a strong several key line ministries. But senior manage- and continuing role in the process. MEF ment officials in the line ministries were not unfortunately played little part in formulation sufficiently engaged. Government ownership until the final stages, and the PRSP leaves also was undermined by inadequate involve- considerable work to be done in linking ment of the Ministry of Economy and Finance priority areas with budgeted programs. (MEF) during formulation of the PRSP. As a consequence, the PRSP was not perceived as Comprehensiveness, long term having broad ownership within the govern- Under the PRS Initiative, a national strategy is ment. This perception also reflected tension expected to provide a comprehensive, long- with the SEDP framework and the concern that term view of the challenges of poverty the Bank was imposing an unnecessary burden reduction. This is a very ambitious undertaking on the government by insisting on a separate for most low-income countries and requires that strategic process. good quality analysis and strategic thinking is in PRSP planners made a genuine effort to place before the PRSP document is prepared. consult NGOs in the preparation of the There is an impressively large body of poverty document and to incorporate their views—a analysis in Cambodia, along with demographic significant change from past planning and health surveys. Unfortunately, these surveys processes—but participation of civil society was are difficult to compare and there is still consid- still limited. The time allowed for consultation erable uncertainty about the poverty baseline. was short. In addition, drafts were made available Prioritization is a major quality issue in the only in English, and regional workshops were Cambodian PRSP, as is the case for other PRSP held in just three of Cambodia’s 24 provinces. countries. Many stakeholders interviewed were 19 T H E P O V E R T Y R E D U C T I O N S T R AT E G Y I N I T I AT I V E disappointed that the PRSP document did not authorities in promoting the PRSP principles as deliver as an operational plan—as advertised, the core of the country’s strategic planning and to be a complement to the general framework implementation process? Planning and of SEDP II. The line ministries were asked to implementation in Cambodia should take into submit programs for inclusion in the Action account pre-existing strategic planning Plan Matrix (APM) of the PRSP. Twenty-two line processes and make them compatible with the ministries and agencies provided submissions, PRSP principles. which comprised long lists of poorly costed poverty projects, on the assumption that there Partnerships would be additional funding available for these Cambodia is highly aid dependent. Improved projects. More than 80 percent of the APM was partnership among development donors and unfunded and nearly half uncosted. with the government could yield significant Subsequent to the PRSP finalization, MoP gains in aid effectiveness. Early on, donors added up the total cost of these programs and recognized the need for coordination. Prior to found that they came to more than $5 billion as the PRS process, however, most took the against a three year availability of donor funds approach of avoiding duplication rather than (in line with capacity absorption constraints) of promoting collaboration. The PRSP is promot- $1.5 billion, half of which was already commit- ing a framework for discussion of the govern- ted to ongoing projects. ment’s priorities with donors that did not exist At the end of 2003, the PRSP did not provide in previous national planning processes for a longer-term vision to guide short- and discussion. A major objective of the PRS Initia- medium-term programs in some of the key tive is to turn aid management into a govern- crosscutting strategic issues facing Cambodia— ment-driven process, but shifting to with the exception of trade—such as private government control over the allocation and sector development, the management of natural management of aid will take a long time. resources, and civil service reform. To be The PRS process focused attention on how effective, the plan must also address sectoral donors could better align their efforts to priorities and strategies. Two areas that warrant support the government’s strategy and exploit particular attention are agricultural growth and synergies among partners. As new country capacity enhancement. More than 70 percent of assistance strategies are being developed by Cambodia’s population depends on agriculture key partners in 2004–2005 (including the Asian as its primary source of income, but productivity Development Bank, U.K Department for in the sector is well below that of Cambodia’s International Development, and the World neighbors. The PRSP could be an effective Bank), it is too early to assess the nature and instrument for bringing together the relevant degree of donor alignment. The education and ministries, the donor community, local authori- health sectors in Cambodia represent good ties, and the many NGOs in the sector to design examples of effective donor coordination. In and implement a coordinated development other areas, however, donors appear to be strategy. It could provide a similar framework for focusing on their own projects with specific analyzing capacity enhancement needs— implementation mechanisms that are often not especially with regard to the judiciary, the police, well-linked to the ministry in which they are and customs and tax officials—to reduce the housed. constraints imposed by corruption on the growth of small- and medium-size enterprises. Results orientation The government’s focus on the PRSP as a The PRS process has put poverty reduction document—which reflected its perception of goals more centrally in the minds of the the incentives offered by the Bank and Fund— government, donors, and NGOs. The MoP has has led it to neglect a more important issue: formalized a process by which line ministries what forms of support could best help the rank planned programs according to their 20 CAMBODIA expected poverty impact. However, this costing and monitoring need to be greatly planning exercise has no clear link to the simplified. In the near term there is little likeli- budget process. Additional effort is needed to hood of being able to track the poverty impact improve the alignment of budgeted project of specific investments or measures, and the selection with the overall poverty impact goals value of diverting substantial resources to such (table 3.1). a goal is questionable. In Cambodia it is In order for the PRS process to be results- probably enough to track inputs on the one oriented, the monitoring and evaluation skills hand and outcomes on the other. of Cambodian officials will need to be enhanced. The instruments and processes for World Bank support—a need for implementation and monitoring of the PRSP effective local support and for better are still at an early stage. A central monitoring alignment with the PRSP unit has been established in the MoP, but to The World Bank’s support to Cambodia in date it is only weakly linked to the data collec- formulating the PRSP was substantial. Through tion units at the central and sectoral levels. In Institutional Development Fund grants, the general, the areas where implementation is Bank financed a support unit in the MoP that likely to proceed well are the education and kept the document preparation moving health sectors, where the capacity of the forward. The Bank’s close consultation in the ministries was strengthened through well- final stages of document preparation improved coordinated donor support prior to the PRSP. quality—but fostered the negative perception Implementation in Cambodia is largely among some in the donor and NGO communi- handled through individual donor-supported ties that the draft document “had gone to project implementation units, and there is no Washington.� consensus on changing this model, even The Bank could have done more to keep though it is inconsistent with longer-term donors and NGOs in the loop during the critical capacity building. phase between appearance of the first draft and Many aspects of the PRSP model are too the finalization of the PRSP. A steering group of ambitious for a country at Cambodia’s stage of donors and NGOs would have provided an institutional development. The approaches to effective way to monitor progress and to PRSP targets for poverty reduction in Ta b l e 3 . 1 Cambodia (percent) Objective and indicator Baseline year Target for 2005 Poverty headcount index 36 (1999) 31 Real GDP average annual growth rate 5.4 (2000) 6–7 (2001–2005) Infant mortality rate (per 1,000 live births) 95 (2000) 90 Maternal mortality ratio (per 1,000 live births) 4.37 (2000) 3.72 Children under 5 with moderate–severe stunting 44 38 12-year-olds completing grade 6 33 (1998) 90 Share of population with access to safe drinking water Rural 29 (1999) 40 Urban 69.5 (1999) 87 Share of population with access to sanitation facilities Rural 8.6 (1999) 20 Urban 49 (1999) 90 Source: NPRS, 2003, Table 1.1. 21 T H E P O V E R T Y R E D U C T I O N S T R AT E G Y I N I T I AT I V E provide joint advice on the best way to additional staff in 2001 that the country office proceed. began to make a significant contribution to the A major challenge for the Bank in supporting PRSP. There was also tension within the Bank preparation of the PRSP was balancing quality on responsibility for the process. It was and ownership. It worked mainly with the MoP relatively late in the process, at the urging of as a way to ensure document quality, and thereby Bank management, that the Cambodia country distanced the MEF from the process. This led to team became engaged in supporting the weak prioritization and costing in the first draft ministries’ efforts, mediating between the of the PRSP with little attempt to address cross- , ministries and the Bank’s sector staff. sectoral dimensions—thus compromising the Before the start of the PRS process, the Bank’s chief advantage of the PRSP as a costed process program, as laid out in its CAS for FY2000–03, by comparison with the broad treatment of focused on achieving a more programmatic programs in SEDP. Efforts to get the MEF back approach to financing through the application of into the process succeeded, but valuable time SWAPs. Individual donor processes and was lost. procedures would be slowly phased out as Early in the PRS process, a key issue for Bank government capacity increased, with a shift to staff was to ensure preparation of a strategy greater budget support. The CAS proposed document comparable in quality to first-round piloting SWAPs in infrastructure, health, and PRSPs from other countries. Because staff at the education. There was extensive consultation Bank and Fund believed that the SEDP did not between the Bank and the government over the provide a credible alternative to the PRS CAS and wide endorsement of the SWAP process and could not be retrofitted within the approach. The consultations also emphasized required time frame, the decision was made to the need for an additional focus beyond basic launch the PRS process in parallel with prepara- education to redeveloping the capacity and tion of the second phase of the SEDP. In leadership lost during the Khmer period. hindsight, a preferable alternative might have Although the Bank’s lending program been to merge the PRS process with the SEDP followed CAS priorities and focused on improv- and delay the preparation of the full PRSP until ing public resource and expenditure manage- the third phase of the SEDP, which could then ment and governance, lending choices be refocused to meet PRSP requirements. (primarily investment lending) did not follow Instead of pushing ahead with the PRSP in the the proposed strategy. The Bank was reluctant short term, the Bank, Fund, and the Asian to align itself with the donor-led Sector-wide Development Bank could have worked with the Investment Management (SWIM) programs in government to create the conditions required the education and health sectors. One donor for an effective PRSP in the long run. Under attributed the eventual improvement in Bank such a plan more training and technical support coordination with the health sector SWIM to could have been provided to ministries to the PRS process. In the area of rural develop- ensure a full understanding of PRSP objectives ment, recent Bank support attempts to operate and process; a public information campaign at the sector level and to work more closely could have been carried out; and work on the with partners. MTEF and development of a coherent agricul- Preparation of the next CAS provides an tural strategy could have been accelerated. important opportunity for the Bank to align its The tension between quality and ownership lending and analytic work in Cambodia with the was aggravated by the inability of the Bank’s PRS process. country office to provide effective support for preparation of the PRSP at the outset of the Conclusion—learning from experience process. The Bank office in Phnom Penh From a troubled starting point, Cambodia has opened in 1999 and had limited staff to support made progress in developing ownership of the the PRSP effort. It was not until the arrival of PRS program. There is a clear upward trend in 22 CAMBODIA the ministry-level understanding of the PRS applied in agriculture, private sector develop- approach. Dialogue between the government ment, capacity enhancement, and civil service and donors, and dialogue among donors, has reform. improved. If these gains can be consolidated, • Second, an effective cross-ministerial steering Cambodia should begin to realize the benefits group needs to be established with a secretariat of the PRS process. responsible for providing regular monitoring While it would be difficult to rate the reports. These need to be available to the Cambodia PRSP experience thus far as satisfac- media and NGOs to ensure good accounta- tory, it may well look quite different some years bility. hence, if the government, the Bank, and the • Third, the international community needs to donor community are able to absorb and make develop a coordinated approach to identifying use of the lessons of the PRS process. Almost and addressing key capacity constraints that all observers note progress in the course of limit implementation of the PRSP principles. PRSP preparation between the early stages and Given its comparative advantage, this should the finalization of the document. The MoP is be a major thrust of Bank work. succeeding in institutionalizing the preparation process, with monthly meetings and coalitions Several additional conclusions relate specifi- with seven donor subgroups. Much will depend cally to the World Bank and its staff. To be an on the success of the Progress Report process. effective partner in promoting the PRS Initia- Actions to strengthen the country office in tive’s principles, the Bank in Cambodia needs Phnom Penh should be an important step to embrace the PRSP not as a “task� being led toward that goal, ensuring that the Bank by a particular sector department, but as a reflects the principles of government different way of doing business and a shared ownership, partnership, participation, and responsibility. The use of team learning events, alignment in its own program. which can include government and partner In moving forward, the effectiveness of the representatives, can be an important means to PRSP in Cambodia as an instrument for support this. The country office should view supporting growth and poverty reduction will support for the PRSP as its principal mandate depend on the government and the interna- and be staffed accordingly. It is not yet clear to tional community taking several key steps: Bank staff in Cambodia whether the institution is giving priority to ownership or quality. • First, the coordinated approach that has been Clearer guidance is needed on how to handle effective in health and education needs to be the tradeoff. 23 4 Ethiopia T he OED evaluation team visited Ethiopia in June/July 2003, and met with a broad spectrum of local stakeholders in Addis Ababa, Awassa, Shebe- dino, and Wonago. Among the sources for the analyses presented here is a survey of 83 stakeholders representing government, civil society, and in- ternational partners. This country case review provides an assessment of the progress and challenges in Ethiopia through PRSP formulation and one year of implementation, as of mid-2003. Addressing economic fragility substantial improvement over past decades. Almost half of Ethiopia’s 66 million people live Growth rates were strongly driven by agricul- in absolute poverty. Each year millions face ture, on which the Ethiopian economy is highly serious food insecurity—between 11 and 13 dependent. In 2000, agriculture comprised 43 million during the most recent drought. percent of GDP and employed 85 percent of Conflicts, such as the 1998–2000 war with the population. In 2002–2003, with drought Eritrea and commodity price shocks, exacer- and a collapse in coffee prices, real GDP growth bate chronic food insecurity caused by policy- plunged to –3.8 percent. Improvements in related failures that include overdependence rural poverty reduction stalled, while urban on subsistence farming, problems of land poverty worsened. Although some notable tenure, weak markets, poor infrastructure, and improvements were made in education and reliance on external food aid. Ethiopia relies on health during the 1990s, challenges related to development assistance for 40 percent of its the availability and quality of social services public sector spending and is likely to remain remain. For example, only one-third of the heavily dependent for many years to come. children were in primary school in 1999–2000, Ethiopia’s long-term debt sustainability is and adult literacy remains very low, especially extremely fragile. The country’s ratio of among women in rural areas. Population exports to GDP is very low, and exports are growth remains high at 2.5 percent per year. concentrated in coffee, a commodity whose Per capita spending on health is about one- price was near its 20-year low in 2003. sixth the Sub-Saharan average. Safe drinking Between 1991 and 2001, annual real GDP water is available to just 15 percent of the rural growth averaged 4.4 percent annually, a population. And HIV/AIDS is a growing threat. 25 T H E P O V E R T Y R E D U C T I O N S T R AT E G Y I N I T I AT I V E Figure 4.1 Ethiopia at a glance Sub- POVERTY AND SOCIAL Saharan Low- Ethiopia Africa income Development diamond 2003 Population, mid-year (millions) 68.6 703 2,310 Life expectancy GNI per capita (Atlas method, US$) 90 490 450 GNI (Atlas method, US$ billions) 6.4 347 1,038 Average annual growth, 1997–03 Population (%) 2.3 2.3 1.9 Labor force (%) 2.2 2.4 2.3 GNI Gross per primary Most recent estimate (latest year available, 1997–03) capita enrollment Poverty (% of population below national poverty line) 44 .. .. Urban population (% of total population) 16 36 30 Life expectancy at birth (years) 42 46 58 Infant mortality (per 1,000 live births) 114 103 82 Child malnutrition (% of children under 5) 47 .. 44 Access to improved water source Access to an improved water source (% of population) 24 58 75 Illiteracy (% of population age 15+) 59 35 39 Gross primary enrollment (% of school-age population) 62 87 92 Ethiopia Male .. 94 99 Low-income group Female .. 80 85 Source: World Bank. Ethiopia adopted a federal constitution in ment. There were three key outcomes of these 1994 with a commitment to ethnic federalism, reforms: (i) a wave of decentralization or regional devolution drawn along ethnic designed to bring government closer to the lines. In 1998, a border and trade dispute people by devolving increased authority from between Ethiopia and Eritrea degenerated into the regional level to the district level; (ii) a new full-scale war that imposed a heavy toll on the cabinet involving key appointments from economy. A peace agreement was reached in outside of the ruling party; and (iii) the 2000, but tensions remain high. Attempts to creation of five super-ministries to coordinate open the political system to competition have and renew the drive to improve government faltered over the years. The emergence of a efficiency. single-party system has increased the public’s The government’s long-term development alienation from the democratic process. strategy and the current Five-Year National Supported by Bank and Fund credits, Development Plan (2001–2005) remain centered measures were taken in the immediate post- on the Agricultural Development-Led Industrial- conflict period to support the exchange rate ization Strategy (ADLI), which dates back to management, public sector management, 1994. ADLI also provided the framework for the public expenditure management, private Interim PRSP, which focused on food security, sector development (investment and urban improved marketing systems, irrigation develop- land reform), and export competitiveness. But ment, expansion of extension services, capacity the pace of reform has lagged in part because building and human resource development in of reluctance to allow private entry into the education and health. Agricultural development financial sector. In addition, some larger will form an important part of any rural poverty companies that were to be privatized did not strategy in Ethiopia. But without a well- find buyers. developed urban system linked to national and In an attempt to rebuild confidence, in 2001 international markets, sustainable agricultural the ruling party congress reviewed the development will be very difficult to achieve. country’s development policy and implemen- Ethiopia reached the Decision Point under tation strategies and announced far-reaching the HIPC Initiative in October 2001, granting it changes in the formal structures of govern- much-needed debt relief. But like many other 26 ETHIOPIA low-income countries, Ethiopia has a low The four pillars of capacity for taking on new debt. Box 4.1 E t h i o p i a ’s P R S P The PRS process—good for government- • Agricultural development–led industrialization and food security donor relations • Governance, decentralization, and empowerment The timing of Ethiopia’s PRSP, completed in • Reform of the justice system and the civil service August 2002, was opportune from the • Capacity building standpoint of government-donor relations, which had been poor because of differences over the war with Eritrea. The PRSP gave the Civil society government an opportunity to articulate its Box 4.2 mobilization agenda to internal and external constituencies, around the PRSP while bringing together a range of policy commitments under a comprehensive and long- The PRSP provided a platform for an active engagement of civil soci- term framework (box 4.1). The government ety in the public policy process. In various ways the NGOs confronted already had a series of national strategy their capacity problems and the traditionally closed culture of the gov- documents emphasizing poverty reduction— ernment. ADLI, the Third Five-Year Development Plan, and the National Food Security Strategy. These • First, they organized themselves into an “NGO Task Force on the strategies commanded the support of the politi- PRSP� that held a series of awareness and sectoral analysis work- cal elite. The PRSP made it possible to integrate shops. They commissioned a study on NGOs’ contribution to poverty these strategies into a unified development alleviation. They monitored consultations at the district level and framework for endorsement by the wider made successive attempts to work closely with the government’s society and the external assistance community. PRSP secretariat. • Second, they formed coalitions with the donor community. Ownership, participation • Third, they engaged various stakeholders in a series of public dis- The PRSP has been well received in govern- cussions. For example, the Forum for Social Studies ran a Poverty ment circles. Ownership at the central govern- Dialogue series in which policy issues were raised and submitted ment level has been high because of strong as part of the PRS process. continuities between the existing policy agenda and the PRSP. Source: Abebe Haile-Gabriel, “The PRSP Process in Ethiopia: Second Meet- The consultative process for preparing the ing of the African Learning Group on PRSPs,� Brussels, November 2002. PRSP was broad-based, closely mirroring the federal and administrative hierarchy of the country—a major undertaking given the size of noted the benefits in terms of raised awareness the country and its population. Between about poverty issues and an increased emphasis November 2001 and March 2002, the federal on self-help at the community level. government conducted public discussions at Among domestic and international NGOs the district and regional levels. Those discus- working at the federal and state levels, views on sions were monitored by NGOs and the donor the consultative process were less sanguine. The community. Civil society organizations (CSOs) majority recognized the importance of the and the donor community also organized a process and its intrinsic value for those who series of public discussions and consultations participated, but a significant number felt that the at the regional and federal levels (box 4.2). consultations promised more than they Most respondents to an OED survey of delivered. There was little space to consider national stakeholders and the donor policy alternatives, a finding supported by survey community regarded the consultative process results. Some civil society organizations and as an important first step toward increased donors expressed disappointment that the final voice and local accountability. District officials PRSP document did not adequately reflect issues 27 T H E P O V E R T Y R E D U C T I O N S T R AT E G Y I N I T I AT I V E raised during the consultations and that opportu- vital role during PRSP formulation and nities for stakeholder engagement had declined subsequently in supporting greater alignment following the end of the consultative process. and harmonization in implementation. A public-private sector consultative forum was Comprehensiveness, long term established to address major problems identi- The PRSP presents a broad diagnosis of poverty fied by private sector partners during the and its key determinants using the results of consultative process. two Household Income Consumption and During the PRS process donors agreed on Expenditure surveys (HICES) carried out in rules of engagement, including establishment of 1995–1996 and 1999–2000, as well as a series of a Joint Donor Fund for PRSP preparation. welfare monitoring surveys conducted annually Donors felt obliged to join the DAG and agree since 1995–1996. Most of the poverty analysis on a plan for moving forward. Close collabora- details quantitative trends in income-based tion between government and donors during poverty, over time and across regions. Also the process, especially the consultation process, covered are nonincome dimensions of had the effect of building greater mutual poverty—nutritional levels (stunting and understanding across, within, and between the wasting), access to health services, literacy, and two groups (box 4.3). Related efforts to expand various dimensions of asset ownership. And cooperation around the Public Expenditure the analysis addresses key areas of vulnerabil- Review (PER) and the Country Financial ity—food security and HIV/AIDS—but in a Accountability Assessment (CFAA) also secured highly aggregated manner. a stronger basis for donor partnership. The resulting poverty profile represents a Efforts were made to improve aid effective- marked improvement over previous efforts as ness through more coordination among set out in ADLI, the Five-Year Plan, and other donors, more direct support for country strategic documents. What is missing from the budgets, fewer loan conditions, and less diagnostic document is a qualitative assess- onerous compliance procedures. Those efforts ment of peoples’ perceptions of poverty and enhanced government-donor relations in vulnerability. No substantive use is made of Ethiopia. Although donors had begun provid- qualitative findings from the PRSP consulta- ing various forms of budget support before the tions, despite obvious disparities between the PRSP, the PRS process strengthened the quantitative analysis and the perceptions approach, helping donors coordinate and expressed by ordinary people during the harmonize their efforts. consultation process to the effect that the level While progress with donor alignment and of poverty was increasing rather than declining. general budget support appears to have moved At a minimum this apparent disagreement is more quickly than in many other low-income likely to reflect discrepancies between subjec- countries, it remains difficult to assess the tive and objective data—not an uncommon extent of the realignment of donor support to result. A more serious interpretation might be the PRSP. Not all donors are equally keen on that the quantitative trends noted in the alignment, and several are adopting a wait-and- poverty profile are in some sense misleading see attitude. Limited prioritization within the because of comparability problems between PRSP is one reason given by such donors for the 1995/1996 and 1999/2000 HICES data. their caution. There is also uncertainty about how new money being made available through Partnerships various global funds, specifically for HIV/AIDS, The PRSP provided a context for a range of new will fit into the PRSP framework. partnerships between government and donors and between government and private sector Results orientation partners. The Development Assistance Group The PRSP highlights areas that need to be (DAG), composed of some 40 donors, played a strengthened if implementation of the strategy 28 ETHIOPIA An architecture for government–donor Box 4.3 dialogue Government and donors have agreed on a new dialogue archi- progress with harmonization; and (iii) provide a forum for dis- tecture for government-donor dialogue. The structure consists cussion of policy issues. of three layers: • The forum will be supported by subsidiary groups of two kinds: joint sectoral groups and groups covering processes • A high-level government-donor forum will meet quarterly at (such as public finance management). the ministerial level to encourage mutual accountability and • The architecture will be supported by a secretariat respon- better policy dialogue. More specifically, the forum will: (i) sible for horizontal and vertical communications between the oversee PRSP implementation and monitoring; (ii) oversee high-level forum and the subsidiary groups. is to be effectively monitored. Those areas The government’s prior approach to poverty include building the capacity of the central reduction had lacked a coherent framework for statistical authority and also the Welfare monitoring and measuring progress. Individual Monitoring Unit of the Ministry of Economic programs had developed monitoring and Development and Cooperation, and ensuring evaluation frameworks, and the Welfare consistency across reporting systems at differ- Monitoring Unit gathered data on key poverty ent levels. Most important of all is strengthen- trends, but none of the data had been used ing analysis and use of data at all levels of systematically to measure progress against government, specifically at the regional level specific policies and programs. In addition, and below, to reinforce the government’s there had been no clear articulation of the commitment to deeper decentralization. A key resource envelope needed to finance the requirement will be to involve nongovernment strategic aspirations of government. The PRSP stakeholders in generating, analyzing, and provided this. using the data. The PRSP contains no timetable Before the PRS process began, the govern- or plan for any of these actions. ment of Ethiopia had committed to increasing The PRSP does provide 25 indicators and spending on poverty-oriented sectors (figure targets for monitoring and evaluating the 4.2). Continued commitment to poverty- implementation of the strategy. Most relate to oriented spending is clearly a part of the PRSP the social sectors—poverty headcount, resource allocation framework, but actual education, and health. Nine performance budget figures are well below estimates, suggest- targets incorporate the main HIPC triggers ing a lack of realism and discipline in planning. required for reaching Completion Points under Increasing the poverty focus of spending the HIPC Initiative. under the PRSP will depend on the govern- Progress with monitoring and evaluation ment’s adoption of a timetable for ongoing was initially slow. Interviewees from within and reforms in public expenditure management outside government spoke of a loss of and on integrating capacities for welfare momentum linked to the emergency situation, analysis and expenditure planning from the the demands of the new block grant system, federal level down to the district level. At the and the time taken up by harmonization and local level, the challenge of aligning budget other donor-related issues. At the time of the priorities with those of the PRSP is great. evaluation mission, progress had picked up. District budgets are in flux in regions undergo- The December 2003 Annual Progress Report ing the second wave of decentralization, and sets out a comprehensive plan for a monitoring officials appear uncertain how recent changes and evaluation system, but indicators must be in the block grant system and devolution of developed, and it is unclear how the system will responsibility for service delivery are related to work in the context of decentralization. each other and to the PRSP. 29 T H E P O V E R T Y R E D U C T I O N S T R AT E G Y I N I T I AT I V E Spending in poverty-oriented sectors in Figure 4.2 Ethiopia 6.5 5.5 Percent of GDP 4.5 3.5 2.5 1.5 0.5 1991/92 1992/93 1993/94 1994/95 1995/96 1996/97 1997/98 1998/99 1999/00 2000/01 2001/02 2002/03 Agriculture & natural resource Road construction Education Health Source: Ethiopia PRSP and the Ministry of Economic Development and Cooperation. Preliminary results the levels forecast in the PRSP. The annual There is some evidence of progress as a result performance review detailed budgeted increases of the PRS process. Bank staff point to for 2003–2004, but no clear indication was given important shifts in the government’s attitude for the basis of such projections, nor was there toward policy discussions in areas previously any further prioritization of core poverty out of bounds, such as business regulation, programs. It is not clear, therefore, that the land tenure, telecoms, and financial-sector problem of unrealistic budgeting has been cured. development. At the time of the evaluation, The full cost of the programs necessary to however, those discussions had not led to achieve PRSP targets exceeded available more market-based approaches. resources envisaged in the Macroeconomic and Progress is also evident in the recent push Fiscal Framework, possibly due to optimistic toward more devolved and responsive local assumptions regarding real GDP growth in the government. The decision to push ahead PRSP (5.7 percent per annum until 2015). quickly with block grants to the districts in the Under the circumstances, a doubling and four largest regions has provided a clear indica- possible tripling of official development tion of the government’s intention to address assistance will be required to help Ethiopia critical service delivery blockages and promote meet the MDGs in poverty, education, hunger downward accountability—both central to eradication, water and sanitation, and PRSP implementation. A major challenge will HIV/malaria by 2015. Given the country’s very be providing the right incentives and overcom- limited ability to absorb further debt, increased ing capacity weaknesses at lower levels of aid flows will have to come in the form of grants. government. Although Ethiopia has a fairly sound public- World Bank support—more expenditure management system, the pace of decisionmaking in the field improvement in aligning the budget with the The Bank has played a constructive role in the PRSP has been slower than hoped. The PRS process in Ethiopia. While supporting the medium-term Macroeconomic and Fiscal government with advice and technical inputs Framework is a shadow exercise, and the full as needed, the Bank has respected the princi- procedural integration of recurrent and capital ple of country ownership. budgeting has not yet been put into effect. Bank engagement was shaped by the CAS Disbursement of budgeted funds was greater preparation process, which ran virtually in in 2002–2003 than in 2001–2002, but still below parallel with PRSP formulation. CAS consulta- 30 ETHIOPIA tions, involving a series of high-level workshops dialogue and to strengthen monitoring and during 2002–2003 on themes such as decentral- evaluation of key policies included in the PRSP. ization, private sector development, capacity The CAS program is wide-ranging, covering building, and rural development, were intended a broad spectrum of projects and programs. Its to echo issues that the government raised in the breadth reflects the scaling up of IDA assistance PRSP. It is difficult to determine if running the following the end of the war and that the PRSP CAS preparation process in parallel with the PRS itself is a wide-ranging and relatively weakly process increased costs to government. Strong prioritized document. In the medium to long synergies between the CAS and the PRSP are term, while other donors scale up their consistent with Bank guidance, but it seems programs, IDA aims to become more selective, equally important to avoid putting an extra focusing on areas that attract less support from burden on the government. other donors, such as infrastructure. In the Responses from government officials about meantime, the current CAS is ambitious. The Bank support for the PRSP were almost univer- challenge for the Bank will be to manage a sally positive. Government officials noted that broad (and high-volume) strategy and ensure relations with the Bank had improved with the sufficient complementarities for other actors in increased status of the country office and the support of the PRSP. country director based in Addis. This was echoed by members of the DAG who had IMF support—constructive engagement witnessed a significant improvement in donor- From the perspective of IMF staff, the PRS Bank relations since 2000–2001. process was a period of constructive engage- Representatives of some NGOs saw the ment. Fund staff based in Addis Ababa partici- Bank’s support for country ownership as a pated in several PRS consultations and PRSC double-edged sword. Although they agreed that missions, and the Fund was an active member a country-led, country-owned process was in DAG discussions. Washington-based Fund crucial, they were concerned that the govern- staff, though less involved on a regular basis, ment paid insufficient attention to their views did take the lead in encouraging government and demands for inclusiveness and wanted the to assess the sources of growth and to consider Bank to advocate more strongly on their behalf. alternative macroeconomic scenarios in the A more comprehensive communication strategy PRSP. The IMF country team also played a at the beginning of the PRS process could have central role in preparing the JSA in collabora- helped set clearer expectations about the tion with the Bank country team. process and the roles of government, donors, Collaboration could have been better in the and NGOs. early stages of the CFAA. Some participants The CAS for Ethiopia focuses on three believed that the IMF could have supported (or clusters of objectives—enhancing pro-poor been asked to support) the process more growth, enhancing human development directly and, as a result, taken greater ownership outcomes by improving governance, and of the findings. IMF participation in PERs was reducing vulnerability—that track closely with also an area that some donors believed needed the four pillars of the PRSP. At the heart of the to be strengthened. The IMF did commit assistance program is a series of PRSCs intended resources to the 2001 PER, including drafting a to support each of the strategic objectives of the chapter on fiscal policy, but the perception was CAS. Other new aid modalities are considered, that Fund ownership of the PER’s findings was including capital grants to local authorities for limited—with implications for the PRGF. infrastructure development and a program- A future role for the Fund in a strengthened matic approach to public sector capacity PER process is essential to future alignment of building. One of the most innovative elements the PRGF and PRSP. Some senior staff in the of the CAS is the plan to use the nonlending Ministry of Finance and Economic Develop- program as a tool for furthering national ment expressed concern about a disconnect 31 T H E P O V E R T Y R E D U C T I O N S T R AT E G Y I N I T I AT I V E between the macroeconomic framework set While the PRSP is still in its infancy, the out in the PRGF and the financing needs of the evaluation uncovered some potentially PRSP, asserting that the Fund’s programming divergent expectations about the process. The rules were not in complete agreement with the principles have been adhered to for the most financing demands of the PRSP. The Fund’s part and have provided momentum and a new view was that the assumptions in the PRSP were sense of direction for what was, until recently, too optimistic, and that a more realistic an ailing government-donor partnership. macroeconomic framework was necessary to Ensuring that the recent gains are sustained ensure that programs were prioritized and will require some changes on both sides of the targets could be met. partnership, including some fairly quick evidence that donors are truly serious about Conclusion—a revived partnership alignment and harmonization. But it appears The PRS process has been a successful experi- that the donor community is keen to support ence for both the Ethiopian government and government’s commitment to far-reaching the Bank. The government views the PRSP as an institutional and structural reforms. important opportunity to articulate its long- The impact of the PRS process on relations standing policy program to both internal and between government and civil society is mixed. external constituents. Many donors view it as There is evidence of positive change in some the first step on the road to more responsive areas, such as the creation of a formal dialogue and transparent government and a new context between government and the private sector. for aid relations. Many civil society organiza- Some observers believe that the scale of the tions and NGOs view the PRS process in a more consultations is evidence that government and radical light, hoping for a more fundamental civil society relations are moving in a positive shift in the political economy of policymaking direction. Others are less convinced and and the consideration of policy alternatives. conclude that the promise of the PRSP was Remaining aware of these differing expectations overwhelmed by deficiencies in the political and finding the right balance between respect system—poor transparency, accountability, and for country ownership and providing intellec- weak political institutions. Sustaining the PRS tual and financial leadership are likely to be process will require further institutionalization critical challenges for the World Bank. And of structures and supports, notably the sectoral, extending ownership of the PRS process regional, and district strategies and action beyond Addis Ababa will be crucial for the next plans. There is an urgent need for all involved phase of the PRSP. to link policy and practice more effectively. 32 5 Guinea T he IEO evaluation team visited Guinea in April/May 2003 and held meetings in Conakry and Mamou, one of eight regional capitals. Among the sources for the analysis presented here was a survey of 53 local stake- holders representing government, civil society, and international partners. IMF operations have not been reviewed beyond end-2002. However, develop- ments on the ground (including progress under the PRS approach) have been taken into account through end-2003. Good potential to benefit from the PRS 40 percent live below the poverty line. In recent approach, but weak institutional years Guinea has had to cope with spending capacity pressures associated with a significant influx of Guinea, with a population of some 8.5 million, refugees fleeing political conflicts in neighbor- is rich in natural resources, particularly bauxite ing Côte d’Ivoire, Liberia, and Sierra Leone. and gold. But its GDP per capita is just Defense spending has risen in response to US$550—a legacy of isolationist and central security concerns along the country’s borders. planning policies that retarded economic Revenue mobilization by the central govern- development. Since 1985 the country has been ment is poor—under 12 percent of GDP—largely embarked on a gradual process of economic because of weak tax administration, widespread opening and liberalization, with the support of corruption, and extensive use of tax exemptions. the World Bank, the IMF, and a handful of Weak and nontransparent public expenditure bilateral donors. Since the mid-1990s, Guinea management and a narrow but costly and has enjoyed an annual real GDP growth of 4 to nontransparent state-owned enterprise sector 6 percent, while inflation has averaged about have compounded the problem. The net effect 5 percent. has been to undermine the government’s ability Despite this relatively stable macroeconomic to finance programs in the priority social sectors. environment, Guinea’s social indicators are Poor transport and communications infrastruc- weak. In 2002, Guinea ranked 159 out of 173 ture, unreliable utilities, and rampant corruption countries in the Human Development Index have undermined private sector development, compiled by the UNDP. Almost two-thirds of the leaving the economy heavily dependent on a population is illiterate and more than narrow mining sector for tax and export revenue. 33 T H E P O V E R T Y R E D U C T I O N S T R AT E G Y I N I T I AT I V E Figure 5.1 Guinea at a glance Sub- POVERTY AND SOCIAL Saharan Low- Guinea Africa income Development diamond 2003 Population, mid-year (millions) 7.9 703 2,310 Life expectancy GNI per capita (Atlas method, US$) 430 490 450 GNI (Atlas method, US$ billions) 3.4 347 1,038 Average annual growth, 1997–03 Population (%) 2.2 2.3 1.9 Labor force (%) 2.1 2.4 2.3 GNI Gross per primary Most recent estimate (latest year available, 1997–03) capita enrollment Poverty (% of population below national poverty line) .. .. .. Urban population (% of total population) 35 36 30 Life expectancy at birth (years) 46 46 58 Infant mortality (per 1,000 live births) 106 103 82 Child malnutrition (% of children under 5) 33 .. 44 Access to improved water source Access to an improved water source (% of population) 48 58 75 Illiteracy (% of population age 15+) .. 35 39 Gross primary enrollment (% of school-age population) 77 87 92 Guinea Male 88 94 99 Low-income group Female 66 80 85 Source: World Bank. IDA is the country’s largest source of conces- subsequently embodied in the MTEF, which sional financing, accounting for approximately authorities began implementing in 1998. In one-quarter of net official development 1999, the application of the MTEF was assistance in 2000 and 2001. Beyond IDA and expanded to include three additional priority the IMF, the donor community is fairly narrow. sectors—justice, urban development, and Since late 2001, in light of concerns with the housing and social affairs. quality of governance, most donors have scaled Guinea was declared eligible for the HIPC back their assistance and refrain from provid- Initiative in December 1999 and reached its ing general budget support. Thus, at the time Decision Point a year later, having prepared an of the evaluation, Guinea was going through a Interim PRSP (I-PRSP) in October 2000. period of relative aid drought, which may have affected domestic stakeholders’ perceptions of The PRS experience—a “good the PRS Initiative. document,� but inadequate process and In 1995, following the release of data ranking limited implementation Guinea at the very bottom of UNDP’s Human Guinea’s full PRSP, completed in January 2002, Development Index, Guinean authorities meets the formal requirements set out in the embarked on a human development program PRSP guidelines and approaches the problem with support from the World Bank. That of poverty from a comprehensive perspective. exercise resulted in the production of a long- The strategies embedded in the document are term development strategy titled “Guinea suitably results-oriented, in the sense that the Vision 2010,� in which the government made a expected outcomes are clearly identified, and commitment to invest in human capital and the PRSP also contains elements of a long-term promote good governance. The top priorities perspective. The PRSP is built around three identified by participants in a National Consul- pillars: enhancing economic growth, promot- tative Forum conducted under the strategy ing delivery of and equitable access to basic were rural development, access to basic services, and enhancing governance and education, access to primary health care, institutional and human capacity building. development of local entrepreneurship, and These characteristics contribute to the improved governance. Those priorities were widespread perception within the donor com- 34 GUINEA munity and among informed observers that the were believed to exceed the government’s Guinean PRSP is “a good document.� 1 capacity to deliver, owing both to technical However, several qualitative shortcomings limit capacity and financing constraints. its usefulness as an operational guide to policy- making. These include the use of an outdated Ownership, participation data set for the poverty diagnosis and little Given the low literacy rate of Guinea’s popula- analysis of the determinants of poverty or the tion, the scarcity of autonomous CSOs and links between specific policies and programs their limited technical capacities, the prohibi- and poverty outcomes, both of which reflect tion of private radio and television, and the technical limitations tied to the time frame limited circulation of newspapers outside the chosen by the authorities for the formulation capital, it would have been unrealistic to expect of the PRSP. Perhaps the most important the formulation of Guinea’s first PRSP to meet shortcoming was the failure to integrate the the objectives of broad-based participation and macroeconomic framework with the rest of the broad country ownership. poverty reduction strategy. Significantly, then, both local stakeholders That failing led to other problems. First, many and international partners believed that the of those interviewed by the evaluation team in outreach effort went much further than at any Guinea—especially outside government time before, if only because it extended over a circles—found the PRSP targets and objectives longer period and involved a larger share of the to be overly optimistic (such as reaching an population. In addition, the high profile of the annual GDP growth rate of 10 percent by 2010). PRS process raised awareness throughout the Second, there was little discussion of the interac- country of the most challenging obstacles to tions between macroeconomic and sectoral poverty reduction. One example of issues that policies—in particular how sectoral policies was emphasized during this dialogue is the might contribute to reaching macroeconomic need for better governance—coupled with objectives, such as revenue and GDP growth, or fiscal decentralization and local community the impact of exchange-rate policy on export empowerment as key elements of an anti- diversification. corruption strategy. Also missing was a discussion of likely Nevertheless, the extent to which the partic- tradeoffs between competing policy objectives. ipatory process actually influenced the final Many private sector and civil society stakehold- PRSP is difficult to measure. The authorities ers, along with a number of government contend that the process validated the findings officials, expressed the view that the priority of previous participatory exercises. And close actions listed in the PRSP were insufficient to to 60 percent of survey respondents agreed address the main structural obstacles to that the final document had been modified to growth, including constraints to private sector accommodate some of their viewpoints. But a development. Fewer than half of the respon- systematic comparison of the action plans dents agreed that the document offered an envisaged for selected areas in the I-PRSP and appropriate balance between growth and in the full PRSP does not suggest that signifi- poverty alleviation. cant changes were made, although in some Finally, many stakeholders interviewed— cases details were added and priorities even those who praised the PRSP content—felt outlined. For example, the participatory that the strategy was out of touch with existing process had limited impact on macroeconomic implementation capacity. This judgment policy (box 5.1). Perhaps that is because little applied notably to the provisions related to effort was made, including by the IMF, to indicators and monitoring arrangements— highlight key policy tradeoffs in a manner that there was as yet no structure to collect and might have encouraged debate. analyze the data needed for the indicators. Although Guinea does have a participatory Other priority public actions listed in the PRSP tradition, the government does not encourage 35 T H E P O V E R T Y R E D U C T I O N S T R AT E G Y I N I T I AT I V E Impact of the participatory process on the Box 5.1 m a c r o e c o n o m i c f r a m e w o r k o f G u i n e a ’s P R S P Participation had little impact on the macroeconomic aspects of such as the reform of state-owned enterprises. Group members the PRSP. Even discussions in the thematic group dedicated to told the evaluation team that they assumed the macroeconomic macroeconomic policy were limited in scope, according to par- framework would be set in the context of separate negotia- ticipants interviewed. The group reported having active discus- tions over the PRGF. The group did not interact with the teams sions that identified, as a major priority, the importance of keeping in charge of those negotiations or with the PRGF monitoring inflation under control, given its negative impact on the poor. An- committee. other major concern was the need to increase availability of Moreover, in the absence of formal interactions between micro-credit. Unfortunately, there was no in-depth discussion of the macroeconomic thematic group and the sectoral groups, the pros and cons of policies that might achieve these objectives. there was no mechanism for integrating the plans of the latter The group did not discuss the overall macroeconomic frame- into the macroeconomic framework or for ensuring that sectoral work or structural reforms of critical macroeconomic importance, plans were conceived within a realistic budget. the free expression of views, especially critical the population (box 5.2). For example, a views, in the presence of government greater role could be given to representative representatives. Thus, a consultation process institutions such as the National Assembly and that relied predominantly on large popular the Economic and Social Council. Civil society gatherings orchestrated by senior civil servants organizations with ties to grassroots communi- and, frequently, cabinet ministers, was unlikely ties throughout the country—including to generate anything but a broad endorsement religious organizations—could have been of the policies put forward by the authorities. asked to participate in the consultations. And At the same time, the process did give partici- public debate could have been encouraged pants a forum to air their perceptions of outside the government-led PRS process—for poverty and to express their needs. And the example, in the media or the academic involvement of senior officials in the outreach community. process was viewed favorably by many of those The JSA of the full PRSP by the Bank and interviewed. Fund was not helpful in conveying the The organizations called upon to represent constraints and limitations of the participation civil society were, in some cases, far from truly process. In addition to flagging the participa- representative—a problem made worse by the tory formulation process as one of the main limited number of slots allocated to civil society strengths of the PRSP, the JSA noted that “civil in the thematic consultative groups (on society participated effectively� and that the average, 4 out of 30). The institution officially process led to “a genuinely country-owned mandated to represent the interests of civil strategy.� A more candid assessment would society, the Economic and Social Council, did have noted the constraints faced and given a not have representatives in any of the initial sense of what could realistically be achieved in thematic groups and generally felt its own a country like Guinea, while recognizing what involvement in the process was unsatisfactory. had already been accomplished. Limitations on the participatory process might have been avoided, or at least mitigated, Comprehensiveness, long term by using other consultation formats, such as Officially, the PRSP became the single frame of seeking feedback on the merits of policy reference for all government policies; in options and on the resolution of tradeoffs practice, the impact of the PRSP on policymak- between intermediate objectives. The scope of ing is more elusive. the process could have been broadened by Macroeconomic conditions and financing using additional communication channels with constraints are still analyzed and discussed only 36 GUINEA Box 5.2 The PRSP dissemination strategy in Guinea The Guinean PRSP is long and written in relatively technical lan- Even members of the National Assembly and the Economic guage. Although it was made public soon after its adoption by and Social Council expressed frustration with the way their in- posting on the IMF Web site (in English only), the authorities stitution was given access to the final document (upon spe- printed only 500 copies of the document, which were dissemi- cific request, with considerable delay, and without having an nated primarily to institutional partners, domestic and foreign. opportunity to debate it in a formal session). Access to the document by the general public was severely Plans to produce a layman’s version of the PRSP, abridged limited. Outside government circles, most people interviewed and simplified, had yet to be implemented when the team vis- by the IEO in Guinea indicated that they either had not been able ited Guinea, more than a year after the adoption of the full PRSP to obtain a copy of the PRSP, or had obtained one through “in- by the government. formal� channels. within a very small circle around the minister of later, negotiations with the IMF on the finance. Budgets submitted to the National measures needed to bring the PRGF-supported Assembly since the PRSP was completed program back on track do not appear to have formally reflect PRSP priorities. But their practi- been guided by the PRSP. cal implications for resource availability at the level of line ministries and other executing Partnerships agencies must still be viewed as limited, As of April 2003, more than a year after comple- because key decisions on the allocation of tion of the PRSP, no major donor was able to spending cuts across sectors are made by the identify significant adaptations in its assistance Ministry of Finance with little consultation on program in response to the PRSP. Nor did there priorities. Nevertheless, budget allocations to appear to be much progress in the coordina- priority sectors have been relatively protected tion of strategic priorities among donors. from ad hoc cuts. Several suggested that the authorities’ failure Problems with the domestic budgetary to organize donor coordination was to blame process are mirrored in two shortcomings of for the limited effectiveness of the PRSP, noting the PRSP itself. First, the strategy did not that the authorities had not convened a multi- address how the macroeconomic program donor meeting since May 2002, although would adapt to exogenous shocks and failures government officials from various ministries of key assumptions to materialize (such as GDP had been approaching donors in an uncoordi- growth and external financing). Second, the nated manner with competing demands, using PRSP did not prioritize the public actions it the PRSP to justify specific requests. Donors outlined or establish criteria for setting priori- viewed the absence of priorities in the PRSP as ties among competing demands across and limiting its usefulness for prioritizing requests within priority sectors. There are no systematic for assistance. links between the action plans of the PRSP and the sectoral strategies and expenditure plans of Results orientation and monitoring line ministries or between the PRSP and The PRSP provides no road map for implemen- regional strategies developed (at considerable tation, nor has separate plans or institutional cost) to implement the national plan. When arrangements to implement it being developed. actual macroeconomic developments diverged Arrangements for involving civil society in significantly in mid-2001 from the path implementation of the PRSP are likewise assumed in the PRSP, the relevance of the PRSP lacking. The thematic groups that contributed was immediately challenged as the appropriate to the full PRSP have not been convened since frame of reference for policymaking. A year the PRSP was completed. 37 T H E P O V E R T Y R E D U C T I O N S T R AT E G Y I N I T I AT I V E The authorities indicated that they hoped to implement the action plans outlined in it merge the original seven thematic groups into (figure 5.2). The government’s subsequent four, each responsible for analyzing and efforts to support social spending rekindled processing the available feedback and for inflation (figure 5.3). preparing updates in their domain. Neverthe- Most stakeholders interviewed in Guinea less, 18 months after completion of the PRSP, believed that the PRSP played a useful role in no arrangements had been made to ensure the raising the profile of governance issues in the sustainability of the PRS process. If the PRS public debate. However, they also noted that process turns out to be short-lived, despite the the actual results of this higher profile had so high profile the government gave it, the far been limited. For example, although laws population may become disillusioned and and regulations to effect changes pertaining to disinclined to participate in future poverty decentralization and transparency had been reduction efforts. issued by mid-2003, actual change on the A few stakeholders—mostly in the donor ground was lagging behind, owing to adminis- community—expressed the view that the trative resistance at the central level and severe delayed takeoff of the implementation process capacity limitations at lower levels and within was largely inevitable under the circumstances. audit and inspection departments. Publication They believed that the groundwork would bear of spending allocations had begun, but journal- fruit eventually, when overall economic and ists were not permitted to verify independently political conditions were more conducive to a the data they received from the ministry. wholehearted implementation by all partners. Many survey respondents suggested that the IMF support—a challenging policy sustainability and effectiveness of the PRSP environment would hinge on the development of a reliable Since 1987 the country has drawn on conces- and transparent monitoring process. Several sional IMF resources, with one arrangement nongovernmental stakeholders expressed the under the Structural Adjustment Facility, two view that this objective would most likely be arrangements under the Enhanced Structural achieved if reporting on PRSP implementation Adjustment Facility (ESAF), and more recently and evaluation of its impact relied on existing a three-year arrangement under the PRGF. vehicles, such as the implementation reports Performance under these successive arrange- systematically attached to certain laws rather ments was generally poor. As a result, none of than on instruments prepared specially for the them could be disbursed fully, even after Bank and Fund. extending the commitment period. A new three-year PRGF arrangement was Preliminary results negotiated in early 2001. It went off-track soon Sectors identified as priorities in the PRSP have after, so that the first review could not be been receiving higher and more stable alloca- completed until July 2002, seven months tions than sectors not so identified. Non-wage behind schedule and after implementation of a current expenditure in priority sectors four-month “consolidation plan.� Slippages increased from 1.2 percent of GDP in 2001 to resumed immediately after completion of that 1.6 percent in 2002, and public investment review, primarily in the areas of public expendi- from 7.1 to 7.8 percent of GDP. Data on ture and monetary policy. After a few unsuccess- education also suggests tentative improve- ful efforts to correct the course, the program ments in school enrollment rates for girls. went off-track again in December 2002. It should be noted, however, that the A few changes have been observed in the relative drying up of external assistance since IMF’s way of doing business under the PRS the adoption of the PRSP (for reasons Initiative. Limited efforts were made early on to unrelated to the PRSP itself) has unquestion- analyze the social impact of key elements of the ably curtailed the authorities’ ability to fully program, but they faltered. However, there is 38 GUINEA Figure 5.2 Official aid to Guinea, 1998–2002 400 300 US$ millions 200 100 0 1998 1999 2000 2001 2002 Grants Net public loans HIPC and other debt relief Source: IMF. Changes in consumer price index in Guinea, Figure 5.3 January 2002–March 2003 12 10 8 Percent 6 4 2 0 Jan-02 Mar-02 May-02 Jul-02 Sep-02 Nov-02 Jan-03 Mar-03 12-month growth rate Corrected 12-month growth rate Source: IMF. evidence that more policy space is given to the IMF staff made only limited efforts to explain authorities in the negotiation of structural the rationale for the PRGF-supported program conditionality, and that meeting macroeco- to audiences beyond the authorities. nomic targets by squeezing social spending is The failure of the IMF to use the PRS process now viewed by the IMF as an unsustainable to make sure that key policy issues and strategy. Despite a poor macroeconomic tradeoffs in its areas of competence were duly environment, the IMF made efforts to adapt its aired and debated stemmed from lack of time program to protect priority spending from and staff resources, reluctance of the authori- adverse shocks and policy slippages. However, ties, and lack of clarity about what was it was unclear to Guinean stakeholders what expected of the staff in what was meant to be a tradeoffs the IMF might be willing to accept, in government-led process. part because of poor alignment between the The design of the PRGF-supported program PRGF and PRSP, in part because of the absence reflects some progress toward most of the seven of institutional forum in which to hold follow- key features that are supposed to distinguish the up discussions on necessary changes to PRGF from the ESAF, but it is still a long way macroeconomic policy, and in part because from being fully embedded in the PRSP: 39 T H E P O V E R T Y R E D U C T I O N S T R AT E G Y I N I T I AT I V E • Alignment. The PRGF was approved before Bank-Fund collaboration the PRSP was complete, and its original frame- Collaboration between the IMF and World Bank work was adopted as the macroeconomic was significantly enhanced at the working level . framework of the PRSP That framework, how- from late 1999 onward, thanks to joint missions ever, was overtaken by external shocks, in- to the country and a clear articulation of cluding the pressure of refugees from respective “lead agency� responsibilities, neighboring countries, so that by the time of leaving few areas of overlap. However, the the first review of the program there was a division of labor in reducing the burden on the clear mismatch between the macro framework government of duplicative or conflicting loan of the PRGF-supported program, which had conditions did not contribute to the achieve- , been updated, and that of the PRSP which had ment of desired reforms of state-owned- not. Consequently, the latter did not provide enterprises. Difficulties in this area were guidance on how to handle tradeoffs required exacerbated by the fact that, reflecting a lack of by the new external environment, leaving the country ownership, the PRSP did not contain a PRSP underfinanced and with unattainable forward-looking strategy. macroeconomic objectives. Regarding struc- tural policies, by contrast, the PRGF-supported Conclusion—form over substance program and the PRSP are broadly aligned. By combining a diagnosis of poverty (even if • Program design did protect priority sector based on old data) with a list of remedial actions spending and allowed for substantial flexibil- in various sectors, all cast in a long-term ity in fiscal targets, both to accommodate higher perspective, Guinea’s PRSP is a useful reference pro-poor spending and in response to shocks. tool for policymakers and donors—and is a However, that flexibility often resulted from significant improvement over past modalities of after-the-fact adaptations that, not being de- public policy formulation. But it is far from an termined on the basis of fully transparent cri- effective road map for policymaking and donor teria, gave rise to misunderstandings and coordination, for at least three reasons: exposed the Fund to charges of arbitrariness from the authorities and the donor community. • Shortcomings in the PRSP. The macroeco- • Structural conditionality was noticeably nomic framework of the PRSP lacked realism streamlined in the sense that formal program and adaptability, objectives were insufficiently conditions declined in number and became prioritized, and inevitable or foreseeable trade- more focused on IMF core competencies. Con- offs were not recognized. Nor were criteria ditionality was streamlined in areas where ei- provided to resolve them. Furthermore, the ther the authorities had demonstrated strong policy agenda contemplated in the PRSP ap- ownership of the reform agenda or where an pears to be at variance with the government’s understanding was reached with the World implementation capacity. The PRSP does not Bank to make it the lead agency in that area. offer guidance to resolve conflicts between However, this trend coincided with the pro- competing objectives or contingency plans liferation of informal conditionality, frequently against changes in the external environment. imposed in the context of interim programs ne- • Unspecified institutional arrangements for gotiated to get programs back on track, re- implementing, monitoring, and updating the flecting doubts in the Fund about the PRSP. For the most part the PRSP left these authorities’ commitment to implementing key arrangements to be defined, and they remained structural reforms and continuing governance tentative and unclear more than 18 months concerns. after the PRSP was completed. Thus the PRSP • Areas where progress has been lacking include had yet to be integrated with domestic poli- the use of poverty and social impact analysis cymaking. Without this integration, both the in program design and the realism of key un- relevance and the sustainability of the PRSP will derlying macroeconomic assumptions. remain in question. 40 GUINEA • Lower-than-expected donor support. Because the JSA was at odds with the perceptions of of the country’s disappointing macroeconomic most nongovernmental stakeholders and many performance and concerns about governance, local donor representatives in Guinea. It did donor support has been less than expected in not incorporate inputs from parties other than the PRSP. The PRSP failed to mobilize addi- the authorities and received very little dissemi- tional donor support. nation in Guinea. The impact of the PRSP has been limited to It seems likely that a better outcome was date, in part because of substantial shortfalls within reach had different approaches been in external financing, but also for lack of adopted. Such a result would have required a government initiative. Its impact on the lengthier and probably more costly formulation policymaking process is visible only in the process, greater technical support from the budgetary process, where priority sectors Bank and Fund, and an emphasis on substance have been protected to some degree, and over form. through some progress in decentralizing The JSA of Guinea’s PRSP fell short of revenues and spending authority. In other meeting several of its many purposes. Although areas, such as governance, which the PRS it offered a balanced assessment of the PRSP’s process identified as critical, impact is hard to contents and was candid in pointing out areas detect. Critically, the sustainability of the PRS that required continued attention from the process is far from ensured—two years after authorities, it was less candid in its description the adoption of the full PRSP, there were no of the participatory process and in its assess- clear institutional arrangements for its ment of implementation risks. In both cases, implementation, monitoring, or adaptation. 41 6 Mauritania T he OED evaluation team visited Mauritania in June/July 2003. Political volatility following a coup attempt in early June 2003 forced the mis- sion to limit its coverage relative to the other case studies in this vol- ume. Among the sources for the analysis presented here was a survey of 64 stakeholders representing government, civil society, and international partners. Abundant resources, unmet needs Senegal River valley. Mauritania contains consid- Mauritania’s social indicators are slightly above erable mineral resources—iron ore, copper, the averages for Sub-Saharan Africa and have cobalt, diamonds, gold, gypsum, and improved significantly in the last decade. In phosphates—but only iron ore is exploited particular, Mauritania has made substantial and commercially. In the future, revenues from rapid improvements in education. For example, hydrocarbon reserves are expected to consti- primary school enrollment for 7–12 year olds tute a major source of income for the county. rose from 49 percent in 1986 to 84 percent in Presently, fishing and mining dominate exports. 2001, and literacy has increased to 57 percent. Mauritania enjoys a 200-mile Exclusive Access to primary health care increased from 30 Economic Zone along its Atlantic coastline, and percent in 1990 to about 70 percent in 2001. its fishing grounds are some of the richest in the Between 1990 and 2000, the proportion of world. However, the fishing sector has been the population estimated to be living below the adversely affected by lack of an effective fisheries poverty line declined from 57 percent to 46 policy, sectoral mismanagement, overfishing, percent, with the decline occurring primarily in and limited technical ability to monitor and the urban areas where half of the population control the Exclusive Economic Zone. live. Roughly 70 percent Mauritania’s people are Infrastructure deficiencies are a serious Arabic-speaking Moors; the rest are members of constraint to development in Mauritania. The the southern-based ethnic groups. The popula- availability and distribution of potable water, tion is growing at an annual rate of 2.6 percent. for both direct consumption and commercial/ Mauritania suffers from a perennial food industrial uses, is inadequate. Urban waste deficit. In 2001, domestic production of cereals management systems are inadequate or was just 40 percent of total supply. The rest was nonexistent. The availability and reliability of imported or received as food aid. Efforts are electricity is limited and relatively expensive. under way to expand food production in the The extent and quality of the road network is 43 T H E P O V E R T Y R E D U C T I O N S T R AT E G Y I N I T I AT I V E Figure 6.1 Mauritania at a glance Sub- POVERTY AND SOCIAL Saharan Low- Mauritania Africa income Development diamond 2003 Population, mid-year (millions) 2.7 703 2,310 Life expectancy GNI per capita (Atlas method, US$) 430 490 450 GNI (Atlas method, US$ billions) 1.2 347 1,038 Average annual growth, 1997–03 Population (%) 2.4 2.3 1.9 Labor force (%) 2.8 2.4 2.3 GNI Gross per primary Most recent estimate (latest year available, 1997–03) capita enrollment Poverty (% of population below national poverty line) 46 .. .. Urban population (% of total population) 62 36 30 Life expectancy at birth (years) 51 46 58 Infant mortality (per 1,000 live births) 120 103 82 Child malnutrition (% of children under 5) 32 .. 44 Access to improved water source Access to an improved water source (% of population) 37 58 75 Illiteracy (% of population age 15+) 59 35 39 Gross primary enrollment (% of school-age population) 86 87 92 Mauritania Male 88 94 99 Low-income group Female 85 80 85 Source: World Bank. inadequate to move raw materials and rely increasingly on local experts in project finished goods. design, integrate project management into the Mauritania was a one-party state from national administrative system, strengthen independence in 1960 until 1978. Since then, a measures against corruption, promote private succession of military rulers has led the sector involvement in the delivery of public country. Some 24 opposition parties exist, but services, and support the government in the legislature remains overwhelmingly in the leading donor coordination. Capacity-building hands of the ruling party, the Parti Republicain efforts should be better coordinated, and civil Democratique et Social. Despite a very weak society should take a more prominent role in opposition, concerns about political stability development work. In light of these findings, persist, due to frequent conflicts between the introduction of the PRS process can be factions in the ruling camp. Presidential viewed as a highly relevant step in Mauritania. elections held in November 2003 returned the Mauritania was one of the first countries to current president for a third term, amid accusa- join the PRS Initiative and was motivated to tions of fraud from the opposition. The complete a PRSP by the prospect of debt relief Mauritanian administration is highly central- under the HIPC Initiative. A full PRSP, ized, with all major offices and agencies representing a more comprehensive approach headquartered in Nouakchott. to poverty reduction than past strategies, was Mauritania had adopted poverty reduction completed in one year, although based on as a national strategy before the introduction of outdated poverty diagnoses. More recent data the PRSP. The country’s first national poverty were used to update the strategy in two recent reduction strategy covered the period progress reports. 1994–1998. The subsequent 1998–2001 Preparation of Mauritania’s PRSP began in Programme National de Lutte contre la December 1999. Responsibility was assigned to Pauvreté was presented in March 1998. a newly established body, the Commissariat aux droits de l’homme, à la lutte contre la The PRS process—a centralized effort pauvreté et à l’insertion (the Commissariat), that may undermine implementation headed by a series of well-connected leaders A 2000 client survey administered by the World who provided a focused working environment Bank in Mauritania recommended that the Bank for an energetic and capable group of profes- 44 M A U R I TA N I A sionals. Such support at the highest level, in a poverty reduction programs to be prepared for context of scarce capacity, helped ensure that each of the 13 regions. The regional strategies the original PRSP was prepared within the are expected to include a participatory process, given time frame. The Commissariat staff a diagnosis of poverty in the region, and a interacted well with donor representatives and comprehensive approach to poverty reduction adapted to the donor “language� with which and development. the PRS process was imbued—ownership, NGO participation was organized mostly by participation, transparency, accountability, and the government, which has sought to establish so on. However, the subsequent transforma- a national umbrella group for the local NGOs. tion of the Commissariat from a planning and Participation by NGOs in 12 thematic working monitoring agency into the principal PRSP groups has been uneven, and there has been a implementation agency carries a significant lack of interest and capacity among NGOs to risk of undermining the role and capacity of the participate in macroeconomic discussions. line agencies charged with executing the According to some observers, meetings tend to poverty reduction program. be highly structured, with little opportunity to depart from the agenda developed by the Ownership, participation authorities or to question the basic approach Ownership of the PRSP by the government was to some issues. strong from the outset of the process through It is also difficult to assess the depth of civil presentation of the final PRSP document. The society participation in the PRS process. participatory process undertaken during PRSP Interviews conducted with national stakehold- preparation was greater than past attempts by ers suggest that participation is only partial, government in Mauritania, and external again often because of the limited capacity of commentators perceived the process to have their representatives. A major problem for the been successful. Stakeholders in-country, donors working to strengthen and empower however, report excessive government control these groups is that the donors often work over the choice of participants and issues through the authorities, which permits the discussed during formulation of the PRSP and government to pick and choose its favorites note that the depth of participation did not and filter aid to them on that basis. evolve during implementation. Since achieving Private sector involvement in the PRS irrevocable HIPC debt relief, policy fatigue may process has been limited. The authorities have settled in. Sustained government commit- established a consultative group to identify and ment to the PRSP therefore, is an area of concern , discuss issues, but private sector groups and in Mauritania. The ultimate test of government individuals appear to prefer to operate through commitment to the process will be whether it informal contacts with their customary continues without donor encouragement. interlocutors in the government. A national dialogue on the PRSP occurred through a variety of seminars and sessions. The Comprehensiveness, long term PRSP draft was presented to CSOs and NGOs in The long-term goal of the Mauritanian PRSP is two sessions. Additionally, four inter-regional to reduce the number of people living under poverty reduction seminars in November 2000 the poverty line to less than 17 percent by 2015. were attended by elected officials, representa- The government also aims to reduce by half the tives of NGOs, civil society, and development number of rural poor by 2015. To achieve these partners active in the 13 regions. To finalize the goals, the PRSP assumed an average growth PRSP, national sessions on poverty reduction rate of more than 6 percent in the medium- were held. term (2001–2004), an assumption not borne The PRSP was formally approved by parlia- out by events (box 6.1). ment in July 2001. The legislature directed that Mauritania’s PRSP acknowledges that poverty the PRSP be implemented through regional is a multidimensional phenomenon. In rural 45 T H E P O V E R T Y R E D U C T I O N S T R AT E G Y I N I T I AT I V E Box 6.1 Key elements of Mauritania’s PRSP The PRSP strategy contains four major themes: 2010 and to below 17 percent by 2015 and cutting rural poverty in half by 2015. Achievement of these goals was based on an as- • To enhance economic growth in order to improve the econ- sumption of 6 percent annual growth between 2001 and 2004. omy’s competitiveness and make the country less depend- The PRSP outlines five priority areas; rural development, ent on external factors as a precondition for poverty reduction. urban development, education, health, and water supply. Priority • To develop the growth potential and the productivity of the goals include reaching universal schooling by 2004, enhancing poor by promoting the sectors closest linked to the poor as the retention rate to more than 65 percent, and eliminating illit- well as the zones in which they are concentrated. eracy and gender and regional disparities in the long run. Health- • To develop human resources and access to essential infra- related goals include reducing infant and child mortality to 90 structure in order to increase the productivity of the poor and per thousand and 130 per thousand, respectively, and achieving improve their living conditions. universal access to basic health care (80 percent of the popu- • To seek institutional development based upon good gover- lation should have access to facilities within 5 km). In terms of nance and participation. water supply, the aim is to equip all villages with more than 500 inhabitants with drinking water supply systems by 2010 as well Targets for poverty reduction include reducing the number as to increase the household water connections to 85 percent of people living under the poverty line to below 27 percent by in urban areas by 2015. areas, poverty reduction is constrained by low The PRSP provided a relevant mechanism levels of agricultural production traceable to low for donor coordination, and donors initially availability of water and access to land, lack of announced their intent to align their programs financing and infrastructure, inadequate with country priorities as set out in the research and extension, poor integration of document. OED interviews suggest that livestock trade into the formal economy, and the donors’ initial enthusiasm dimmed, however, small size of the domestic market. For the urban as they came to see the PRS process as being poor, economic activities and opportunities have directed from Washington. Informants specifi- been constrained by the precariousness of land cally referred to control over the deadlines and tenure, absence of accessible financing, structure of the PRSP and to instances where inadequate skills, and lack of infrastructure and the Bank met separately with the government social services. In addition, the harshness of in Washington or Paris to draft documents Mauritania’s terrain, vulnerability to drought, before other donors had the opportunity to external debt, high birth rates, continued participate. migration to the urban areas, weaknesses in the The primary vehicle for donor coordination managerial capacity of the administration, and has been the Consultative Group meeting held low capacity of civil society organizations are also every three years, which offers the government cited as key challenges to poverty reduction. the opportunity to present, and receive support for, its development objectives. The Partnerships meetings have encouraged closer coordination In 2001 the three largest donors in Maurita- among the World Bank, United Nations, IMF, nia were the EU (42 percent), IDA (19 EU, African Development Bank, and bilateral percent), and Japan (13 percent), which donors. The quality of donor coordination is together contributed 74 percent of the $269 considered by most stakeholders to be good in million of aid disbursements that year. Total Mauritania. Progress can still be made, disbursements in 2001 continued the three- however, in crosscutting areas such as year upward trend, following an eight-year governance, decentralization, public sector low in 1998. reform, and the environment. 46 M A U R I TA N I A Although donors have been mapping their World Bank support—overall alignment programs to PRSP priorities, they have yet to with the PRSP shift from a project-based approach to general The World Bank has generally aligned its budget support, and there is considerable program with the priorities of Mauritania’s PRSP. distance between donors on standards for It began by providing economic and technical auditing and procurement. Donors have analysis to underpin the PRSP preparation expressed disappointment with the low level of process. The Bank’s May 2002 CAS for Maurita- their involvement in the formulation of PRSP nia states that the Bank’s main objective for Progress Reports. FY2003–2005 is to assist the authorities in implementing the PRSP, and the proposed Preliminary results—targets consistently operations are designed to support implemen- proved overoptimistic tation of the four strategic axes of the PRSP. Overly optimistic targets and capacity constraints The first of the four strategic axes of the have caused implementation to lag behind PRSP, accelerating private sector growth, is original plans. Estimated spending on the PRSP supported in the CAS by a Financial Support program has varied from year to year but has Reform Operation, a series of PRSCs, and a been consistently overoptimistic. For example, Partial Risk Guarantee to assist the government estimated spending on the PRSP program for in the pending privatization of its national 2002–2004 almost tripled between the PRSP and electricity company, SOMELEC. The proposed the second Progress Report. The first Progress PRSCs would strengthen public sector and Report projected very large increases in expendi- financial management, and support judicial tures in health and education, based on recently reform and capacity building in the Ministry of completed MTEFs in those sectors. Projected Finance, the Commissariat, and the Ministry of spending for education in 2002 tripled between Health. the PRSP and the first Progress Report. The second axis, growth anchored in the Actual expenditures have fallen far short of economic environment of the poor, is the projected costs. For example, allocations supported by an Urban Development Program for growth promotion were projected to and the Aftouf-Es-Saheli Water Supply Project. increase to almost 30 percent of total PRSP The Bank is supporting the government expenditures by 2004, but by 2002, actual through an MTEF for the infrastructure sector spending was still only 0.4 percent of total PRSP and by continuing work on municipal taxation. expenditures. Projections for education were The Education Sector Development Program not realistic—actual spending was equal to forms the basis for the Bank’s assistance on the about 35 percent of the target. In the second third axis, developing human resources and Progress Report, projections for the sector ensuring universal access to basic infrastruc- were revised down to about a quarter of the ture and services. This is to be accompanied by previous estimate for 2003. advisory services to measure the progress in the With the exception of health and rural health and education sectors, particularly development, the low levels of spending through support for annual rapid surveys and a emphasize the low absorptive capacity of the demographic and health survey. Mauritanian public sector for additional The proposed PRSC program, an Institu- resources. Despite analytic work by donors, tional and Capacity Building Project for Natural including the World Bank, that limitation is a Resource Management, and the Mining Sector serious risk to implementation of the PRSP Capacity Building Project support institutional program. The increased role of the Commis- development and governance, the fourth axis sariat in PRSP implementation has also raised of the PRSP. Core diagnostic work—such as a concerns, undermining the role and capacity of public expenditure review, a country financial the line agencies charged with executing the accountability assessment, and country poverty reduction program. procurement assessment report—is planned to 47 T H E P O V E R T Y R E D U C T I O N S T R AT E G Y I N I T I AT I V E support the government’s efforts in governance preparation of the first Progress Report. and to build the framework for the PRSC. Poverty and social impact analyses are planned The series of PRSCs is the centerpiece of the or are under way in public health, education, Bank’s lending strategy. The focus on PRSCs was mining, and water supply expenditures. justified by Mauritania’s strong macroeconomic The evaluation mission identified many gaps progress, an improved fiduciary framework, in the JSAs of the original PRSP and the two and the MTEFs in health and education. The Progress Reports. For example, the JSAs did not PRSCs are to take a progressively larger part of discuss key issues such as public sector the lending program, underpinned by the efficiency, expenditure levels, and implementa- analytical and diagnostic work. To cope with tion capacity. Revenues from hydrocarbon are Mauritania’s low capacity for aid absorption and expected to constitute a major source of program implementation, the shift to program- income for the country, but the use of the PRSP matic lending was to be phased in gradually, to channel hydrocarbon resources into poverty while the Bank continued with selected invest- alleviation has not yet been discussed. The JSA ment projects targeting poverty reduction, of the PRSP assesses the quality of participation institutional strengthening, and capacity in contravention of the JSA guidelines and building. A number of donors interviewed by without identifying objective, measurable the evaluation team suggested a cautious criteria for the assessment. approach to the near-term introduction of significant budget support given the govern- Conclusion—a process in danger of ment’s limited implementation capacity. reversal During the first year of the CAS cycle, the The sustainability of the PRS process in Maurita- level of Bank lending was less than foreseen. nia is threatened by several related factors. By Two activities, a Mining Sector Capacity assigning responsibility for preparing the PRSP Development Credit and a US$21 million grant to a small group of competent and well- for HIV/AIDS, were approved at the start of connected leaders, the Mauritanian authorities FY2004. The other planned FY2003 projects for were able to produce in short order a full PRSP financial sector reform and the first in the fully owned by the top echelons of government. series of PRSCs have slipped. The largest But the centralized approach to preparation— project in terms of net commitments is an coupled with government selection of partici- Urban Development Program ($70 million), pating NGOs and the weakness of civil society followed by an Education Sector Development organizations—implies that country ownership Program ($49.2 million) and an Integrated of the PRSP is limited. Implementation of the Development Program for Irrigated Agriculture PRS process is threatened both by reported ($38.1 million). “policy fatigue� within the government and the As noted, the Bank’s program of economic capacity constraints that made necessary the and sector work provided inputs into the PRS centralized approach to PRSP preparation. The process. Between 1998 and 2000, the Bank shortage of capable personnel in the Mauritan- substantially increased the size of that program ian public sector is reflected in consistent in Mauritania and introduced smaller, govern- overoptimism in estimating expenditures for ment-led, analytic products dealing with a PRSP priorities. That same lack of capacity broad range of issues. For example, Bank means that the country has a very limited ability support for the Household Survey was finalized to absorb additional external aid, despite its just after the PRSP provided updated data for pressing development needs and a generally the Progress Reports. Macroeconomic good climate of donor coordination. Most modeling advisory services and analytical work donors, skeptical of the public sector’s ability to on MTEFs in education and health contributed use aid in a manner fully consistent with the to the macroeconomic framework and , PRSP have shied away from providing assistance program planning, and the MTEFs fed into the in the form of general budget support. For 48 M A U R I TA N I A similar reasons, the World Bank may wish to future revenues from oil sales, Mauritania’s reflect on its decision to support the PRSP future may depend on a twin push to develop its through the mechanism of PRSCs. Given the public sector capacity (notably in the area of unexploited export potential of Mauritania’s public expenditure management) and to mining and fishing sectors, the underdevelop- strengthen the institutions through which that ment of agriculture in the arable southern tip of capacity can be effectively and transparently the country, and the promise of significant used to reduce poverty. 49 7 Mozambique A joint IEO/OED evaluation team visited Mozambique in April/May 2003 and held meetings in Maputo and Inhambane province. Among the sources for the analysis presented here was a survey of 57 stakehold- ers representing government, civil society, and international partners. The re- port of the evaluation team covers the experience of Mozambique with the PRSP/PRGF process through mid-2003, although it also includes analysis of the Bank’s 2003 CAS, completed in October 2003. Poverty reduction as a peace dividend for multiparty elections and a free-market Mozambique had a gross national income per economy. A UN-mediated peace agreement capita of about US$210 in 2001 and ranks near finally ended the fighting in 1992 and paved the the bottom of UNDP’s Human Development way for the country’s first general elections in Index. However, although severe poverty is still 1994, won by FRELIMO. In the second general the plight of most Mozambicans, available elections in 1999, FRELIMO again won, but the indicators of growth, life expectancy, results revealed a country deeply divided. education, and health, suggest a decrease in Real GDP growth has been strong, averaging poverty over the last decade. more than 8 percent per annum between 1993 After winning independence from Portugal in and 2002. After a sharp drop in 2000 due to 1975, Mozambique began constructing a adverse weather, growth rebounded in 2001–2002 centrally planned, state-led economy under the (figure 7.2). Underpinning the growth perform- leadership of the Liberation Front of ance has been a steady rise in domestic invest- Mozambique (FRELIMO). Basic education and ment, financed in large part by foreign savings. health services expanded substantially, but these A number of governance challenges, partly gains were soon undermined by a long-lived civil linked to the uneven distribution of growth over war and economic collapse. The changing politi- the last decade, threaten to undermine the cal situation in South Africa and pressure on the social and political stability of the country. warring factions from the international Organized crime is reportedly on the rise, with community finally led to agreement over the the judicial system unable or unwilling to take need for a domestic political resolution of the effective action. Corruption, including grand civil war. In 1990, a new constitution provided corruption involving senior officials of the ruling 51 T H E P O V E R T Y R E D U C T I O N S T R AT E G Y I N I T I AT I V E Figure 7.1 Mozambique at a glance Sub- POVERTY AND SOCIAL Saharan Low- Mozambique Africa income Development diamond 2003 Population, mid-year (millions) 18.8 703 2,310 Life expectancy GNI per capita (Atlas method, US$) 210 490 450 GNI (Atlas method, US$ billions) 3.9 347 1,038 Average annual growth, 1997–03 Population (%) 2.0 2.3 1.9 Labor force (%) 2.1 2.4 2.3 GNI Gross per primary Most recent estimate (latest year available, 1997–03) capita enrollment Poverty (% of population below national poverty line) 54 .. .. Urban population (% of total population) 36 36 30 Life expectancy at birth (years) 41 46 58 Infant mortality (per 1,000 live births) 101 103 82 Child malnutrition (% of children under 5) 24 .. 44 Access to improved water source Access to an improved water source (% of population) 57 58 75 Illiteracy (% of population age 15+) 60 35 39 Gross primary enrollment (% of school-age population) 106 87 92 Mozambique Male 107 94 99 Low-income group Female 95 80 85 Source: World Bank. Real GDP growth and inflation in Figure 7.2 Mozambique, 1991–2002 70 60 50 40 Percent 30 20 10 0 –10 –20 1991 1992 1993 1994 1995 1996 1997 1998 1999 2000 2001 2002 Real GDP Growth Average Inflation Source: IMF. party and government, is a significant problem. 240 percent in 1993 to about 65 percent in The government budget has been heavily 1999, before climbing again to nearly dependent on external aid. On average, grants 80 percent in 2002. and net external concessional borrowing were Mozambique has been implementing equivalent to 56 percent of annual total economic reform programs with the support of expenditures and net lending during the IMF and the World Bank since 1987, with the 1994–2002. Debt relief has had a substantial introduction of the Economic Rehabilitation impact on Mozambique’s debt service burden. Program. By mid-1996, the liberalization of the The ratio of external debt to GDP fell from exchange rate, prices, and interest rates was almost 52 MOZAMBIQUE complete. Subsequently, progress was also domestic policy processes, including the role achieved in the areas of privatization and restruc- of representative government and political turing of public enterprises, financial sector parties. PARPA has become the common point reform, the strengthening of tax and customs of reference on poverty reduction policies administration, the improvement of public within government and for dialogue between expenditure management, and trade liberalization. the government, other national stakeholders, Using a sector-oriented approach, and international development partners. Mozambique has a long tradition of planning in The principal objective in the PARPA is the public sector, with the core of planning reduction in the incidence of poverty from processes concentrated at the central level. 70 percent in 1997 to less than 50 percent by There has been little tradition of effective 2010. Six priority areas have been identified to involvement of civil society, and the scope for promote human development and create a cross-sectoral bottom-up planning has been favorable environment for rapid, inclusive, and limited until recently. Mozambique had a broad-based growth (box 7.1). separate Planning Commission until 1994, when The PRS approach has proven to be relevant it was merged with the Ministry of Finance to to Mozambique, and its underlying principles form the Ministry of Planning and Finance (MPF). have been applied to varying degrees in the Poverty reduction was a key objective of PARPA process. government policy in Mozambique before the advent of the PRSP process. Five strategic Ownership, participation programs were developed between 1989 and There is strong central government ownership 1999. The content of the strategies evolved of the PARPA. Its preparation was led by the over the decade. The first poverty strategy MPF, with virtually no external support— (prepared in 1990) identified three priorities: reflecting an explicit decision by the govern- employment creation, access to basic social ment. Because the draft PARPA drew heavily on services, and creation of safety nets for the existing sector plans developed using new poor. The government’s sphere of influence at participatory processes involving national that time was largely restricted to urban areas. stakeholders and international partners, and In 1995, the emphasis changed to rural areas, because annual planning cycles include consul- focusing on stimulating productive activity tations with officials from provinces and through the revitalizing of market mechanisms, districts, the government believed that the and investing in education and health, with less production of a draft PARPA for the national emphasis on establishing safety nets. consultative process already involved a partici- Only with the preparation of Mozambique’s patory process. Initially, therefore, the authori- Action Plan for the Reduction of Absolute ties resented what they saw as the externally Poverty (PARPA) developed for 2000–2004, did driven, procedural requirements of the PRS poverty reduction strategies begin to take the process. form of detailed plans with targets, time The draft PARPA went through a three- frames, and resource requirements; thus month consultation process that involved the moving poverty reduction beyond mere policy participation of business associations, labor statement to linkage with mainstream policy unions, religious bodies, nongovernmental planning and resource allocation instruments. organizations, media, central and provincial government institutions, and donors. Although The PRS process—a common point of it had its shortcomings,1 it did lead to signifi- reference with strong government cant changes in the draft, most notably, the ownership inclusion of good governance as one of the The PARPA drew heavily on existing sectoral fundamental priority areas. Unfortunately, policies and programs, but it also reinforced since its approval by the government, little has public agenda issues related to participation in been done to disseminate the document; it is 53 T H E P O V E R T Y R E D U C T I O N S T R AT E G Y I N I T I AT I V E Box 7.1 Priority areas of Mozambique’s PRSP Overall goal: To reduce the incidence of poverty from 70 percent • Basic infrastructure, with targets related to coverage and in 1997 to less than 60 percent in 2005, and to less than 50 per- quality of roads, access to electricity, and access to potable cent by 2010. water Priority and complementary areas are identified to promote • Governance, with targets related to administrative decen- human development and create a favorable environment for tralization, and improvement in justice sector services rapid, inclusive, and broad-based growth (at an average rate of • Macroeconomic and financial policies, with targets re- 8 percent per annum). lated to inflation rate, and fiscal revenue as percentage of The six priority areas: GDP. • Education, with targets related to adult literacy, rural fe- Complementary areas: Employment and business development, male literacy, and gross primary enrolment social action, housing, mines, fisheries, tourism, industry, trans- • Health, with targets related to infant and child mortality, maternal port and communications, technology, environment, and pro- mortality, birth weight, malaria mortality, and HIV prevalence tection against natural disasters. Targets are defined in relation • Agriculture and rural development, with targets related to an- to expansion of social safety net and increase in access to nual growth in agricultural production, cereals and cashew housing for low-income households. nut production, and incidence of food insecurity largely unknown in any detail outside Maputo ity, and water access rates and draws on the and provincial capitals. results of a qualitative participatory poverty Parliament and political parties were not assessment. Although the strategy provides a substantially involved in the consultation good description of regional differences of process, and PARPA was exclusively approved poverty incidence, it does not provide provin- by the Council of Ministers and not taken to the cial- and district-level targets. There is a need Assembly. for greater clarity on how national targets relate Despite the fact that the government claims to provincial and district priorities. strong ownership of the PARPA and the policies The PARPA is based on a long-term perspec- in it, many of the representatives of civil society tive, but its operational framework is appropri- and the private sector met by the evaluation ately rooted in the medium term. Government team expressed the view that the PARPA was officials emphasized to the evaluation team too strongly influenced by policies supported that the general priorities had been govern- by the IMF and the World Bank, with what they ment policy for a long time, and that the broad perceived to be excessive attention to objectives were unlikely to change even if macroeconomic stability, privatizations, and specific elements were to be modified. premature exposure of key production sectors With the support of the United Nations to foreign competition. They also saw the Development Program, a process is under way to requirement for endorsement by the Bank and define a “Vision 2025.� The government has Fund as limiting country ownership. tasked a group of individuals from government, political parties, and civil society to prepare a Comprehensiveness, long term long-term vision and development strategy for The PARPA recognizes the multidimensional Mozambique. It is too early to gauge the extent nature of poverty and adopts a comprehensive to which vision 2025 will complement the PARPA. approach to poverty reduction. It employs a wide range of indicators for measuring poverty. Partnerships In addition to income- and consumption-based The PARPA process is partnership-oriented and indicators, it also uses data on illiteracy, mortal- builds on existing mechanisms for government- 54 MOZAMBIQUE donor relations. The process was launched at a donor conditions, and making more transpar- time of broad consensus between government ent the links between government policy and donors on key priorities for poverty implementation and the disbursement of reduction. Most of the sector policies around budget support assistance. which the PARPA was built were developed by A high level forum—the Poverty Observa- the government in close cooperation with tory—brings together the government, a broad donors. Donors were consulted in the PARPA range of national stakeholders, and Mozam- preparation process, but did not impose bique’s international development partners to themselves. Both bilateral and multilateral review PARPA implementation on an annual donors offered assistance to prepare the PARPA, basis. The focus is intended to be on drawing but the MPF declined these offers—and donors lessons from past experience and developing accepted being turned down. This relatively recommendations for improving implementa- harmonious process reflected the fact that tion and monitoring. Mozambique already had a comparatively well- developed donor coordination mechanism. For Results orientation example, SWAPs—including in health and The PARPA includes about 60 “outcome� and agriculture—have developed since the mid- “intermediate� indicators at macro and sectoral 1990s. Cross-sector donor groups had existed level. In addition, it includes an operational previously but became more effective with the matrix showing principal actions to be introduction of SWAPs. undertaken as well as annual targets for several Project financing is still the dominant form indicators. The operational targets are largely of aid in Mozambique, and although a shift in taken from sectoral plans, and vary considerably the composition of actual aid disbursements in scope and precision. For example, targets has not yet materialized, peer pressure among regarding access to primary education are set at donors to move away from stand-alone projects the output-level (enrollment rates), while seems to have increased because of PARPA. The quality targets are input-oriented (curriculum donors that were already coordinating their reform, training of school directors). In health, balance of payments support before the PARPA impact-level targets on mortality rates are process, indicated to the evaluation team that included as well as targets related to service there is increased interest in the donor delivery. In some areas there are not yet specific community in moving toward general budget targets (for example, anti-corruption). support modalities and linking such support to The annually updated five-year medium- spending targets in PARPA. term budget framework (MTBF) is the princi- The willingness to consider budget support pal instrument for translating the public is a remarkable vote of confidence by donors expenditure priorities in PARPA into budgetary given that fiduciary diagnostic work by the allocations. The key challenge is to ensure World Bank pointed to high fiduciary risks implementation of public policies and plans, associated with using the government budget and to reach district and community levels in system. This may explain the intensive focus on poor areas. Capacity development and public expenditure management by the group incentives are needed to translate plans into of donors that provide budget support. This budgets, to make budgeted funds available, to group, currently made up of 11 donors, has ensure that they are spent as budgeted, to formed a macroeconomic working group, in ensure quality of spending and to enhance which the IMF and the World Bank participate monitoring and reporting. Presently the public as observers. The group is developing a joint sector is overly centralized and hierarchical, its Performance Assessment Framework. The staff poorly qualified and badly paid. Over- objectives of that framework include greater regulation and red tape flourish, and the alignment with PARPA, reducing transactions executive branch of government is not costs through increasing harmonization of effectively accountable to the legislature. 55 T H E P O V E R T Y R E D U C T I O N S T R AT E G Y I N I T I AT I V E Weaknesses in Mozambique’s public expendi- either parliament or CSOs. The focus of the ture management system were highlighted in progress report was on measuring perform- an assessment by the IMF and World Bank in ance against PARPA targets, with little discus- early 2002. sion of links between measures and outcomes, Weak institutional capacity is affecting the or of how policies might be adapted to enhance quality of monitoring. Slow progress in performance. Absence of data and weaknesses integrating the PARPA monitoring indicators in implementation capacity and in monitoring into the “normal� reporting system for the and reporting systems have hampered analysis. public expenditure system and the government Sectors and activities deemed the most budget suggests that for the foreseeable future, important for achieving PARPA’s poverty effective monitoring of PARPA will continue to reduction goals have been designated to rely on special exercises and studies tailored to receive priority in the allocation of public that specific purpose. Senior officials expenditures. Although the share of total emphasized to the evaluation team that spending accounted for by these priority areas procedures for monitoring and reporting on has increased since 1999, it fell short of the progress in implementation should be oriented PARPA targets for 2001 and 2002, including in around the strengthening of domestic the areas of education, health, and basic processes, rather than fulfilling World Bank and infrastructure. IMF procedural requirements. The PARPA process appears to have enhanced policy discussions on poverty issues Preliminary results within the government (especially between Evidence on progress toward meeting PARPA central and provincial government officials), targets and objectives is mixed, with significant and to a lesser extent, between the government progress in some areas (for example, education and nongovernment stakeholders. However, a and health), and less progress in others (for broad range of those met by the evaluation example, agriculture). But in most areas, there team stressed the need to strengthen policy is uncertainty about the extent to which analysis capacity in and outside government for progress is being made in improving quality of a more meaningful and sustainable participa- service delivery. These uncertainties reflect tory process. system weaknesses in budget execution, monitoring and reporting, and are being World Bank support—an effective local addressed as part of the reforms of the public presence expenditure management system. The Bank provided appropriate support during There is a wide range of ongoing reforms in formulation of the PARPA—without being the public sector aimed at addressing capacity intrusive. Preparation of the PARPA was strongly weaknesses related to implementation and country-driven, with the Bank providing monitoring of the PARPA. Although these comprehensive informal and formal comments, reforms are meant to address capacity as well as offering relevant analytical inputs. constraints, they appear to be taxing current Considerable work by the Bank preceded the capacity to the limit. At the same time, different PARPA process, including development of levels of government (that is, central, provin- SWAPs, a country economic memorandum, cial, and district) are putting great stock in participatory activities conducted in connection these reforms—especially aspects dealing with with the 2001 CAS, and preparation of a public decentralization of the planning and monitor- expenditure review. ing system—to improve the implementation of The end product of the formulation phase government policies and programs. was clearly country-owned, despite several The first annual PARPA progress report was differences in viewpoint between the Bank and prepared by the MPF, with inputs from other government. These included the scope of the central ministries but with no involvement of strategy, which the Bank feared was too broad 56 MOZAMBIQUE and sector-focused, and public policies that less dependent on the timing priorities of were not sufficiently connected to poverty donors and more linked to domestic needs and reduction. events. Second, the Bank can provide a balanc- Over the past five years the Bank has ing, mediating voice within the donor broadened and improved its contacts with civil community. The ability of the Bank to play this society actors, including the business role is closely linked to its ability to deliver community. That improvement cannot be timely and high quality analytical and advisory attributed to the advent of the PARPA process, inputs. It is important that the Bank ensure but benefited from it. Donors expressed a more adequate resources to provide knowledge ambivalent opinion of their relations with the services. Bank, painting a mixed picture of the quality of sectoral dialogue but acknowledging that the IMF support—too low a profile with civil Bank had been more sensitive to the views of society other donors. Reflecting the wishes of the authorities, IMF Bank lending and nonlending assistance staff did not participate directly in the prepara- since the PARPA have mostly been in line with tion of the PARPA. In particular, IMF staff did its priorities. Significant adjustments have been not participate in government-led consulta- made to the analytical work program, in line tions with stakeholders. with government priorities, including work on There was little public discussion of the decentralization, procurement, and financial macroeconomic policy content of the PARPA, accountability to advance the governance and more generally, there continues to be little objectives of the PARPA. However, activities public discussion of macroeconomic policy dropped or delayed have resulted in gaps in the issues. The IMF could help broaden discussion Bank’s support for important pieces of the of macroeconomic policies in the country by PARPA strategy, most notably rural develop- facilitating wider dissemination and discussion ment—a critical area. of the analytical work that forms the basis for A new CAS was discussed by the Bank’s its policy recommendations, including the Board in November 2003, and the planned work of technical assistance missions. assistance is broadly aligned with the PARPA. In Feedback from the authorities and donors addition, the results framework of the CAS is on the role of IMF resident representatives was linked to PARPA goals. very positive. However, civil society organiza- The Bank intends to move with other tions indicated that the IMF was invisible to donors toward providing assistance through them, reflecting its low profile outside official budget support linked directly to the PARPA, circles. The office of the IMF resident represen- through a PRSC, despite some cautionary tative should be strengthened if it is to play all signals from its own analytical work on govern- the roles expected of it. ment procedures and systems. Fiduciary risks The macroeconomic framework of the pre- may be partially offset by linking the PRSC existing PRGF-supported programs influenced directly to performance indicators related to the PARPA’s macroeconomic framework, and progress in enhancing public expenditure PRGF objectives have become broadly aligned management. to PARPA goals. The PRGF emphasizes poverty The Bank’s experience with the PARPA reduction interventions (that is, pro-poor and process has implications for the way the Bank pro-growth government budgets) and operates. First, the Bank’s qualified local measures to improve public resource manage- presence in key strategic areas is highly ment and accountability, in line with PARPA appreciated by government, national objectives on good governance. On the other stakeholders, and other donors. It is also a hand, little has been done to analyze the social prerequisite for continuous participation in impact of major macroeconomic adjustment policy dialogue processes, which are becoming and structural reform measures. 57 T H E P O V E R T Y R E D U C T I O N S T R AT E G Y I N I T I AT I V E The fiscal stance in the PRGF-supported arrangements while raising important issues program continues to reflect the goal of signifi- of alignment between government processes, cant reduction in aid dependence over a relatively on the one hand, and the requirements of short time (from very high levels associated with external development partners, including the the 2000 drought), but in reality, programs have IMF and the World Bank, on the other. Assist- become more flexible in dealing with aid inflows. ing the authorities to strengthen the analyti- Thus, program reviews did adapt to the PARPA cal content of national reporting instruments framework and the availability of higher aid than under parliamentary scrutiny would enhance originally projected. In this sense, the program the prospect for closer alignment. 2 In these did not suffer from “aid pessimism,� but it is not ways collaboration can be strengthened in possible to answer the more difficult counterfac- Mozambique and elsewhere, as the Bank and tual question of whether, if the program had Fund further support the cause of poverty targeted a higher level of external financing from reduction. the start, donors would have been “catalyzed� to provide more financing. Conclusion—toward general budget There has been significant streamlining in support structural conditionality under the PRGF- Poverty reduction as a key objective of govern- supported program. This has meant transfer- ment policy in Mozambique predates the ring responsibility for areas vacated by the IMF introduction of the PRS approach by the IMF to the World Bank. It is unclear, therefore, and the World Bank. However, the PARPA has whether the burden of aggregate conditionality become the common point of reference on on the country has been reduced. poverty reduction policies within the govern- ment (across sectors and between central and Bank-Fund collaboration provincial levels of government), and for In addition to lessons learned by the IMF and dialogue between the government, other World Bank individually, the PRSP process in national stakeholders, and international Mozambique offers useful insight into collabora- development partners. tion between the two institutions in reducing The PRS approach has proved to be relevant poverty. In the future, attention should focus on to Mozambique, and the underlying principles strengthening the processes and findings of JSAs. have been applied to varying degrees in the JSAs for Mozambique have been candid in their PARPA process. The process has been country- evaluations, pointing out strengths and areas of driven with strong government ownership; significant improvement without glossing over although the process was undoubtedly HIPC- weaknesses. At the same time, they have tended driven initially. The consultation/participation to understate the challenges to implementation process undertaken was a step forward, but its posed by capacity weaknesses. Notably, the JSAs scope was limited, with parliament playing do not seem to have contributed much to hardly any role at all, and CSO roles in enhancing partnership with other donors, who implementation and monitoring not well see the assessments as geared almost exclusively defined. PARPA is result-oriented but there is to the Executive Boards of the IMF and the World scope for further prioritization of public Bank. Some international donors would like to actions and streamlining of the targets and participate in such assessments rather than indicators for monitoring. PARPA is compre- merely provide input to Bank and Fund staff. hensive in recognizing the multidimensional Such collaborative assessments could strengthen nature of poverty and is based on a long-term donor partnerships in Mozambique and other perspective for poverty reduction. countries. There was little public discussion of the Problems associated with the production macroeconomic policy content of the PARPA. of the first PARPA Progress Report reflect Broader participation would be facilitated by the weaknesses in monitoring and reporting establishment of a government-led macroeco- 58 MOZAMBIQUE nomic working group open to representatives The PARPA is partnership-oriented and has from all stakeholder groups, to provide a forum contributed to shaping a new agenda for for macroeconomic policy discussions. Initially, international support to Mozambique, charac- its agenda could be organized around links terized by harmonized approaches by donors, between the PARPA and the budget. with increasing interest in providing assistance Integration of the PARPA into the govern- through SWAPs and budget support linked ment’s planning, budgeting, and reporting directly to the PARPA. Increased budget processes is under way but by no means support, underpinned by a common set of complete. The MTBF is the principal instru- conditions imposed by providers of such ment that links the PARPA to the government support, comes with a risk of “herd behavior� budget, but it is not yet based on detailed among donors to cut back aid in the event of costings of intervention programs. “bad news.� 59 8 Nicaragua A joint IEO/OED evaluation team visited Nicaragua in June 2003, and held meetings in Managua, Leon Norte, and Santa Rosa del Peñón. Among the sources for the analysis presented here was a survey of 83 stake- holders in government, civil society, and the donor community. The report was finalized in November 2003 and covers developments up to that point in time. Poverty reduction in a polarized country fiscal consolidation efforts. The government has Nicaragua is among the poorest countries in made significant efforts to improve macroeco- Latin America, with a gross national income per nomic stability. But its weak parliamentary capita of US$710 in 2000. Very strong growth of backing has forced it to seek support on a case- agricultural output between 1993 and 2001, by-case basis for the passage of key legislation, ascribable in part to a one-time peace dividend including reforms necessary to reach the and the availability of unoccupied land, Completion Point under the HIPC Initiative. reduced the share of the population living Nicaragua’s political instability has blocked below the poverty line from 50 to 46 percent, consensus building, limited open policy debate while extreme poverty fell from 19 to and political participation, and produced a 15 percent. But poverty remains largely a rural weak institutional framework that blurs the phenomenon, heavily concentrated in the separation of state powers, weakens the rule of eastern and northern parts of the country. law and protection of property rights, and leads Throughout the 1990s Nicaragua struggled to manifold problems of governance. Among with major macroeconomic problems, princi- parties with very different political and ideolog- pally unsustainably large fiscal deficits and debt ical positions, agreement on policies tends to levels. Repeated efforts by successive govern- occur at very general levels—if indeed it can be ments to consolidate the fiscal situation proved found at all. The need to reduce policies to unsuccessful, owing largely to recurrent their lowest common denominator to broaden spending overruns that pushed Nicaragua to the an otherwise elusive consensus has impaired verge of fiscal collapse. Hurricane Mitch in 1998 the role of consultative institutions. This high seriously damaged the country’s infrastructure degree of polarization has limited what can be and disrupted the macroeconomic situation and achieved in the fight against poverty. 61 T H E P O V E R T Y R E D U C T I O N S T R AT E G Y I N I T I AT I V E Figure 8.1 Nicaragua at a glance Latin POVERTY AND SOCIAL America Low- Nicaragua & Carib. income Development diamond 2003 Population, mid-year (millions) 5.5 534 2,310 Life expectancy GNI per capita (Atlas method, US$) 730 3,260 450 GNI (Atlas method, US$ billions) 4.0 1,741 1,038 Average annual growth, 1997–03 Population (%) 2.6 1.5 1.9 Labor force (%) 3.8 2.1 2.3 GNI Gross per primary Most recent estimate (latest year available, 1997–03) capita enrollment Poverty (% of population below national poverty line) 48 .. .. Urban population (% of total population) 57 77 30 Life expectancy at birth (years) 69 71 58 Infant mortality (per 1,000 live births) 32 28 82 Child malnutrition (% of children under 5) 10 .. 44 Access to improved water source Access to an improved water source (% of population) 77 86 75 Illiteracy (% of population age 15+) 23 11 39 Gross primary enrollment (% of school-age population) 105 129 92 Nicaragua Male 104 131 99 Low-income group Female 105 126 85 Source: World Bank. Poverty in Nicaragua is associated with high In the context of the political polarization income and consumption inequality, high rates alluded to earlier, low participation, and low of unemployment and underemployment government capacity for planning and (especially for women), land-tenure insecurity, implementation, aid donors have exerted great and poor access to infrastructure and public influence in defining Nicaragua’s development services. Fertility rates are twice the Latin priorities over the past decade. In 1993 American average; educational attainment is Nicaragua produced a medium-term develop- low; health services are inefficient and of low ment policy in close collaboration with the quality; and malnutrition is widespread. Bank and the IMF, and a three-year social Moreover, the poor are vulnerable to recurrent agenda to provide support for the poorest and natural disasters and the risk of periodic most vulnerable groups of the population and hunger, and are marginalized by insufficient increase the effectiveness and efficiency of information and lack of opportunities. Poor service delivery to the poor. women are subject to a high incidence of The launching of the Enhanced HIPC Initia- domestic violence that raises concerns about tive in late 1999 prompted the Nicaraguan their status and social cohesion. government to prepare a Progress Report on Social indicators and the coverage of social the fight against poverty that contained the services have improved in the period foundations of the later I-PRSP and the PRSP 1993–2001, but progress has been uneven. strategies and goals. This document concluded Primary enrollment accelerated and secondary that the government needed to: (i) generate enrollment improved considerably, but broad-based economic growth galvanized by efficiency at the primary level has not the modernization of the countryside; (ii) improved. However, infant and child mortality improve the human capital of all, but particu- rates decreased significantly in this period, larly of the rural poor; and (iii) try to ensure an including an impressive reduction of the infant adequate safety net so those at the bottom do mortality rate from 40 to 31 per thousand from not slip further down. These three pillars later 1998 to 2001. Maternal health has improved, became the foundation of the PRSP—with the but both prenatal care and institutional birth addition of governance at the insistence of the rates have only improved modestly. donor community. 62 NICARAGUA Two additional strategy documents, parallel making arena, the PRS approach—which to the PRS process, are worthy of mention. The requires leadership and commitment of the Plan Puebla-Panamá, launched in early 2001, government and a broad-based consultative envisaged major infrastructure investments in process to broaden country ownership and Nicaragua as part of efforts to improve generate consensus around basic policies— transportation between Mexico and Central offered a means of widening participation in American countries. Some observers found a national strategy formulation. The PRS Initia- clash between this exercise and the PRS tive was designed to have direct and important process. Second, the United Nations Develop- implications for resource allocations, prioritiza- ment Program and the Organization of tion, and explicit results and indicators— American States funded a participatory process elements that in practice had been missing to produce a “Vision of the Nation�—a strategy from previous strategies. focusing on environmental and poverty The design of the PRSP itself, with its firm issues—in 2001. links to debt relief under HIPC and continued None of the above efforts was linked to the support from the Bank and IMF, limited the budget process or to specific allocations of legitimacy of the PRS process in Nicaragua. But resources. Actual resource allocations, policy debt relief has been to some degree a catalyst developments, and legal initiatives have been for gradual building of domestic consensus on determined by other political processes. Most basic elements of a sustainable stabilization of the planning exercises have come about in program and for concentrating efforts to tackle response to donor or international events, pressing macroeconomic issues. rather than from domestically driven processes. Ownership, participation The public consultation process for the PRSP The PRS process—too little focus on was a step forward, but differences in views growth about its purpose and legitimacy have caused The Nicaraguan government embarked on the assessments to vary widely—not surprising in a PRS process because it was a precondition for country with a polarized political system. As a obtaining debt relief under the HIPC Initiative. result, ownership of the PRSP has remained This underlying motivation appears to have limited. Nevertheless, most observers agree driven many subsequent processes and that the dialogue among the government, decisions—notably the depth and scope of donors, and civil society was useful. policy debate. Once launched, the PRS process The PRSP document was prepared in a moved quickly. In just one year the government highly centralized manner by a group that submitted its Interim PRSP, reached the HIPC enjoyed support at the highest level of govern- Decision Point, and completed the PRSP. But ment. Key social ministries (education and the timing of final agreement on the PRSP was health) participated in its design, but there was decidedly poor—coming when the macroeco- limited diffusion to other strata of government. nomic situation was in crisis and elections were Moreover, there was limited dissemination of unfolding. The outgoing government had the PRSP outside Managua. seriously compromised authority. The participatory process, initially limited, On balance, the PRS approach has been was subsequently broadened as a result of relevant for Nicaragua’s national strategy pressure from stakeholders, including donors. process and its poverty reduction efforts. It The consultations centered on human capital addressed important gaps in past experiences, investments and protection of vulnerable widened participation, focused on a compre- groups, with relatively little focus on growth hensive vision of poverty, and employed a strategies and macroeconomic policies. Politi- framework emphasizing results. In a cal parties chose not to get involved in the fragmented and largely donor-driven policy- consultation process for fear of being co-opted 63 T H E P O V E R T Y R E D U C T I O N S T R AT E G Y I N I T I AT I V E and losing bargaining power in the national light on the multidimensional implications of assembly. More fundamentally, the two poverty. dominant parties perceived participation The PRSP included a comprehensive series under the PRSP as part of a political game on of targets and indicators to monitor the impact the part of the government, intended to of policies on the poor. These included: (i) weaken their influence. national objectives of poverty reduction and Civil society groups (NGO and labor union social development indicators; (ii) intermediate representatives) generally saw the PRSP consul- indicators, some measuring impact and others tations as a first, positive step, but many public service quantitative delivery; (iii) policy expressed frustration that the process had not matrix targets (some at the service output level; moved from consultation to greater participa- others at the process level); and (iv) resource tion and influence on the ultimate policy input targets, fixing, for example, expenditure decisions. At the national level, the process has levels for certain areas and programs, and had little impact on the debate on broader including the resources available through HIPC macroeconomic goals (including tradeoffs debt relief. However, while the link between involved in the choice of growth strategy), goals and intermediate indicators was clearly set which led some stakeholders to question the out in the social sectors, there were no interme- benefits of participation in consultative diate indicators for the productive area. This processes. would make it difficult to trace the effects of There has been no significant movement public policies on the growth and the impact of toward opening up the debate on the growth on the poor. macroeconomic framework and alternative With assistance from the Inter-American policy tradeoffs to a broader group of partici- Development Bank the government has pants. Many senior officials and IMF staff established SINASIP—a mechanism designed interviewed argued that the urgency of the specifically to monitor the indicators of the economic circumstances at the time the PRSP PRSP and provide information that can be used was being formulated—most notably the to take corrective actions. massive public sector deficits, the fragility of As noted in the joint Bank-Fund Joint Staff the financial sector, and the lack of government Assessment of the PRSP, the transparency of credibility stemming from poor macroeco- budget preparation, execution, and reporting nomic management and earlier governance remains inadequate, with questions surrounding problems—limited the public exploration of the classification of certain outlays as poverty- alternative macroeconomic policies. As reducing. The 2001 public expenditure review economic conditions improved and govern- found a tendency to classify all programs in some ment credibility increased, the IMF gradually rural development subsectors as important for became more receptive to government initia- poverty reduction along with virtually the entire tives on alternative policy settings and timing. social sector investment program, including many programs not especially targeted to the Comprehensiveness, long term poor. The overly broad classification of pro-poor The PRSP is comprehensive in recognizing the social spending was also underscored by some multidimensional nature of poverty and has stakeholders interviewed by the evaluation team, good quantitative and qualitative diagnoses of who linked it to the incentives created by HIPC poverty. The quantitative analysis built on living conditionality. standards measurement surveys conducted in 1993, 1998, and 1999 (post- Hurricane Mitch) Partnerships and on demographic and health surveys. A Highly aid-dependent, Nicaragua is among the qualitative poverty and exclusion study that top five recipients of official development analyzed the perceptions of the poor placed assistance in the world, but the impact of aid has vulnerability at the center of attention and shed generally been below expectations—in part 64 NICARAGUA because of severe deficiencies in the coherence concerted effort to fight poverty had been of donor-funded projects and programs, poor attempted, most observers view the PRSP as targeting, and lack of a results-oriented approach the most comprehensive attempt to date to that permits an assessment of outcomes. link poverty issues with a strategy for economic According to various stakeholders, the fine development. In a country historically charac- line between support and imposition in several terized by a great deal of economic and politi- cases had been overstepped by donors. The cal volatility, the PRSP offered a framework that value of many analytical inputs was recognized, promised some degree of policy continuity and but the perception remained that the social some broadly shared objectives. sector emphasis of the PRSP reflected However, Nicaragua’s commitment to and pressures from donors that could not be easily ownership of the PRSP and its principles have questioned. For some, the strong donor been questioned. First, the design of the PRSP backing of the PRSP was a manifestation that itself, with its firm links to debt relief under the strategy reflects a “Washington-consensus� HIPC and continued support from the Bank on how poverty reduction is best achieved. and Fund, limited the legitimacy of the PRS The new government’s desire to modify the process in Nicaragua. PRSP by introducing, in fall 2002, a new national Even if the government and the donor development strategy (ENADE) with a greater community commit to the PRSP as the guiding emphasis on directly productive investment framework for their programs and projects, caused tension between donors, who had severe challenges to the government’s expected the PRSP to bring a more durable implementation capacity remain. The public policy framework, and the government, which sector has undergone a process of moderniza- felt constrained in its ability to pursue a revised tion and restructuring, but serious problems and strategy. ENADE promised a significant modifi- capacity constraints make implementation a cation of some of the key propositions of the challenge. Implementation capacity at the PRSP; in particular, substantial modifications of municipal level, in particular, is very weak the (largely donor-funded) public investment because of the lack of qualified staff, centralized program, with less emphasis on social decisionmaking practices, and limited resources. infrastructure and more on productive Nicaragua’s poor system of public expendi- infrastructure in areas judged by the govern- ture management is a particular hindrance to ment to be of high growth potential. PRSP implementation. Some of the problems Although donors and the government agree include the fragmentation of policymaking (for that a key role of the PRSP is to align donor example, incremental budgeting processes and support, progress in reorienting ongoing the effective separation of recurrent and invest- projects and programs in key sectors of the PRSP ment budgeting), the widespread practice of is advancing very slowly because of limited earmarking budget funds, and the govern- capacity in government sector agencies and the ment’s inability to determine whether funds unwillingness or inability of donors to reconsider are being used efficiently or for the purposes projects that have already been approved. intended, which places the public sector in a Improved aid coordination will require weak position to address governance issues. A efforts by all participants to clarify their own 2002 joint study by the IMF and the World Bank policies, positions, and procedures, to cede also concludes that despite recent improve- leadership to other donors where appropriate, ments, Nicaragua still has a relatively weak and to help the government coordinate support. system of budget formulation, budget execution, and budget reporting. Results orientation The PRS process brought poverty issues to the Preliminary results fore of Nicaragua’s public policy agenda. Although limitations in budget reporting need Although this was not the first time that a to be taken into account in interpreting budget 65 T H E P O V E R T Y R E D U C T I O N S T R AT E G Y I N I T I AT I V E data, early evidence on expenditures from the The modest results shown to date must be 2002 PRSP Progress Report offers some viewed in the context of the very difficult promise. In 2002, Nicaragua spent $211 million starting conditions described above. The PRSP on PRSP outlays, just under the planned did not initially strengthen policy processes, amount of $217 million (table 8.1). As a while poor public expenditure management percentage of GDP, PRSP expenditures were has been a hindrance to implementation. actually higher than planned. Expenditures of 15.6 percent of GDP represent an increase World Bank support—discussing options relative to 1997 levels, in which expenditures for growth on PRSP outlays were estimated to be 14.7 The Bank has provided important support to percent of GDP. The composition of expendi- the PRS process, including relevant analytic tures in 2002 was biased toward social protec- inputs and advice that are well-aligned with the tion, environment, and decentralization, at the needs of Nicaragua. These included two expense of human capital investment and, in poverty assessments, a PER, and an agriculture particular, education. growth study. Ongoing analytical work includes It is still too early to assess whether expendi- a country financial accountability assessment, a tures have translated into results, but progress country procurement assessment report, and a has been made on some fronts. Data for 2002 poverty and social impact analysis. The quality were not available at the time of this evaluation, of the technical assistance and analytical advice but of the 29 intermediate targets specified for provided by the Bank is held in high regard by 2001 in the PRSP, 20 have been achieved most informants. according to the PRSP Progress Report. The The Bank’s CAS is well-aligned with the PRSP nine that are lagging are clustered in primary and closely follows the main strategic pillars, schooling and infant, child, and maternal detailing specific support to each pillar in mortality. The Progress Report notes that some recognition of the constraints to poverty targets are of particular concern because they reduction identified in the PRSP. The Bank is have been static or slightly below the base year aligning its loan program with the PRSP by of 1999. moving to programmatic lending—in the form Planned and actual PRSP expenditures in Ta b l e 8 . 1 Nicaragua, 2000–2002 Plan vs. 2000 2001 2002 actual PRSP outlays Actual Actual Revised Plan Actual (percent) In US$ millions 218.5 268.4 217 211.8 –2.4 As percentage of GDP Total 17.3 14.9 14.1 14.4 15.6 8.3 Internally financed (including HIPC) 10.2 9.2 8.4 9.2 9.2 0.0 Externally financed 7.1 5.7 5.7 5.2 6.4 23.1 Shares by PRSP pillar Economic growth 37 35 44 43 –1.9 Human capital investment 33 29 35 27 –22.1 Social protection 14 13 9 13 43.0 Governance 1 0 1 2 17.1 Crosscutting themes 15 23 11 15 43.1 Total 100 100 100 100 Source: Nicaragua PRSP and PRSP Progress Report. 66 NICARAGUA of a Programmatic Structural Adjustment tion of the macroeconomic framework, reflect- Credit and planned PRSCs. ing, inter alia, the difficult initial macro-stability The Bank is improving its behavior concern- conditions. Moreover, the macroeconomic ing the PRSP principles. According to stakehold- framework of the PRSP was quickly superseded ers, the Bank has become more open to dialogue by events, and the PRSP lacked operational with the country and is taking steps to increase guidance on policy tradeoffs. Some sensitive its responsiveness to civil society and the govern- but critical issues (resolution of the banking ment. The agriculture growth study, for example, crisis and its fiscal consequences) were not part is being done at the request of the government. of the broader policy debate, at the election of But the durability of a strategy requires both the government. good substance and sound process. The Bank The initial design of the PRGF-supported could do more to improve implementation in program did not differ significantly from earlier Nicaragua by strengthening country ownership. ESAF-supported programs. A key element of The Bank also could do more to advance the the macroeconomic framework was the path to debate on broad-based growth and the fiscal consolidation. This was in part because tradeoffs between growth and poverty-related both the government and the IMF staff consid- spending in Nicaragua. It is widely believed that ered stabilization a priority and made it the the emphasis on social-sector spending has main axis of the program. More recently, fiscal been imposed by the Bank. In reality, there is flexibility appears to have increased—for an inherent incentive and a bias toward social example, in the accommodation of potentially sector spending by government under the higher aid flows. HIPC Initiative. Structural conditions in the PRGF- The key issue is the absence of a frank supported program were narrowed to focus on discussion of growth issues in Nicaragua. The financial system and public sector issues that balance between growth and social welfare is have major macroeconomic relevance in both technically complex and fraught with Nicaragua and are highlighted in the PRSP as contentious political issues—but precisely for key priorities. But the burden of quantitative that reason it must become a more public performance criteria did not diminish. debate in an appropriate forum. The Bank The policy space for government-led initia- should promote and participate in this discus- tives and flexibility to political considerations sion, integrating the process under the PRSP in the context of the PRGF-supported program framework. The ongoing Development Policy did expand over time in response to greater Review, with its focus on sources of economic government credibility and a more stable growth, may play a positive role in this regard. macroeconomic environment. That expanded flexibility was reflected, for example, in the IMF support—beyond stabilization policy choices associated with the approval of The Fund’s policy formulation process for the 2003 budget, the passage of tax reform Nicaragua under the PRS approach was initially legislation, and the process of asset recovery not very different from previous practices. from liquidated banks. Macroeconomic stabilization efforts took up Key components of the PRGF-supported most of the attention, while growth and program—the need for fiscal consolidation and poverty issues did not receive the considera- macroeconomic stability—have been owned at tion that would have been expected under the senior levels of the government, but the new approach. IMF technical assistance, in program has lacked broader ownership. contrast, has generally supported the priorities Despite recognition of increased outreach highlighted in the PRSP, including tax reform efforts by the Fund, the formulation of and banking crisis resolution. macroeconomic policy is still regarded as not Program negotiations, rather than the PRSP, being open to a broader policy debate outside remained the driving force behind the formula- of traditional negotiations between the IMF 67 T H E P O V E R T Y R E D U C T I O N S T R AT E G Y I N I T I AT I V E staff and the authorities. To a large extent, this tion was that this broad-based, country-driven reflects the choice of the government. Opening strategy would lead to better implementation program negotiations to nongovernment of key policy elements, then the initial results actors is not realistic, but efforts by both the must be assessed as modest. The linkage of the government and the IMF to broaden the PRSP to budget processes has been an discussions of macroeconomic policy could important area of weakness, in part because of contribute to better understanding of various major problems with public expenditure policy choices and tradeoffs. So far, the PRSP management, which has hindered PRSP has had little impact in this area, although the implementation. use of technical commissions, akin to the one The assessment of modest results, however, employed to formulate tax reform proposals, should be weighed against what realistically may be worth exploring further. could have been expected from the initial PRSP cycle in a country characterized by social and Bank-Fund collaboration political fragmentation, weak institutions, and The World Bank and IMF have collaborated fragile economic conditions. Recent develop- closely since the PRSP. Resident representatives ments seem to point to a change in processes have played a key role, providing inputs to the in the right direction, including consultation respective programs and participating regularly and the broadening of policy dialogue. in meetings with civil society groups. The area The PRS process has brought the critical and of governance and legal reform, of consider- complex issues of poverty reduction out into able importance in Nicaragua, is an example the open, thereby laying the groundwork for a where collaboration appears to have been more effective approach to poverty reduction. effective, leading to a clearer delineation of Expectations of what could be achieved in the responsibilities and conditionality. short term may have been too high—and the The JSAs of the Bank and Fund were insuffi- close linkage to HIPC undoubtedly caused the ciently candid in some key areas (notably with government to press the pace, with some regard to constraints on participation in the adverse effects on the depth and comprehen- PRS process). Nevertheless, they played a siveness of the policy debate. Nevertheless, useful role in identifying country-specific risks progress in this direction is no trivial or to the implementation of the strategy, as well straightforward task and needs to be judged as areas in which further development of accordingly. specific policies or policy prioritization was The consultation process for the PRSP was a needed. However, the PRSP Progress Report step forward, but differences in views about the did not play the central role that would have purpose and legitimacy of participation been expected in the ongoing debate on the highlight the inherent difficulties of promoting growth strategy and modification of the PRSP, participation in a country like Nicaragua, with and the associated JSA should have signaled its polarized political system. As a result of this failing. those differences, broad country ownership of the PRSP has not been achieved. Nevertheless, Conclusion—modest results under most observers agree that the PRS process was difficult initial conditions a helpful attempt to foster a dialogue among Initially, the PRS process in Nicaragua did not the government, donors, and civil society. clarify or strengthen the policy processes that The experience with the introduction of the together determine whether and how poverty new development strategy (ENADE) illustrates will be reduced. If the intention of the process some of the potential tensions between differ- was to reduce the inherent complexities of ent elements of the PRS process, notably poverty reduction in Nicaragua by having a between the provision of greater policy space comprehensive strategy to which all significant to the government and other objectives, partners could subscribe, and if the expecta- including donor priorities. These tensions were 68 NICARAGUA increased in the case of Nicaragua by the poor the growth strategy (Pillar I of the PRSP). But alignment of the PRS process with the domestic even with the considerable imperfections in political cycle and the limited policy debate the process, the fact that the debate is now early in the process on the tradeoffs involved in taking place should be welcomed. 69 9 Tajikistan A joint IEO/OED evaluation team visited Tajikistan in June 2003 and held meetings in Dushanbe and Khudjand. Among the sources for the analysis presented here was a survey of 91 stakeholders from govern- ment, civil society, and the donor community. The team’s report covers the experience of Tajikistan with the PRS process through mid-2003. Surmounting regional strife to reduce survey of living standards revealed higher, and poverty growing, poverty rates among women—in part Tajikistan is one of the poorest countries in the a legacy of the civil war. Female-headed world—with per capita income of about households had less access to land, irrigation, US$200 in 2002. Despite this, indicators for and livestock. They also had lower food security literacy, education, and health and the infant and lower monthly income compared with mortality rate are better than for many other male-headed households. low-income countries—but among the lowest Tajikistan passed through a tumultuous in the former Soviet Union. decade after the collapse of the Soviet Union. A According to the headcount measure used long civil war and sharp reductions in incomes by the government’s statistical agency, about 83 and living standards left their mark on the percent of the population is considered poor, country’s economic infrastructure, while brain and 33 percent extremely poor. A poverty drain (chiefly to Russia) bled capacity from the assessment undertaken by the World Bank in public sector. Economic reforms gathered late 1999 found that although inequality was momentum in 1997 following a peace accord lower than in other countries of the region, it that marked the end of the war. Real GDP was rising, and that poverty incidence in the growth has been strong since 1998—albeit capital, Dushanbe, was much lower than from a low base—and a measure of macroeco- elsewhere. Children are the most vulnerable nomic stabilization has been achieved. demographic group—the higher the number of The peace agreement brought a degree of children in a household, the greater the likeli- stability to the country, creating a government hood of poverty. Poverty rates were also high considered to be one of the more inclusive in where three or more elderly persons resided in the region. Despite a functioning bicameral the same household or extended family. A 1999 legislature and an active opposition, however, 71 T H E P O V E R T Y R E D U C T I O N S T R AT E G Y I N I T I AT I V E Figure 9.1 Ta j i k i s t a n a t a g l a n c e Europe & POVERTY AND SOCIAL Central Low- Tajikistan Asia income Development diamond 2003 Population, mid-year (millions) 6.3 473 2,310 Life expectancy GNI per capita (Atlas method, US$) 200 2,570 450 GNI (Atlas method, US$ billions) 1.3 1,217 1,038 Average annual growth, 1997–03 Population (%) 0.8 0.0 1.9 Labor force (%) 2.1 0.2 2.3 GNI Gross per primary Most recent estimate (latest year available, 1997–03) capita enrollment Poverty (% of population below national poverty line) .. .. .. Urban population (% of total population) 25 63 30 Life expectancy at birth (years) 67 69 58 Infant mortality (per 1,000 live births) 90 31 82 Child malnutrition (% of children under 5) .. .. 44 Access to improved water source Access to an improved water source (% of population) 60 91 75 Illiteracy (% of population age 15+) 1 3 39 Gross primary enrollment (% of school-age population) 107 103 92 Tajikistan Male 109 104 99 Low-income group Female 104 102 85 Source: World Bank. regional and political rivalries remain a reality Tajikistan has achieved some success in fiscal in Tajikistan, aggravated by the pressing socioe- matters since the mid-1990s—the fiscal deficit conomic problems the country faces. The (excluding externally financed public invest- government remains highly centralized, with ment) fell from nearly 6 percent of GDP in 1996 the president having broad powers. to near balance in 2002. However, substantial Landlocked Tajikistan relies on good quasi-fiscal activities (especially in the energy regional relations for access to trade outlets. and cotton sectors) raise concerns about Periodic border closings by some of its sustainability, and the combination of low tax neighbors (Uzbekistan, Kazakhstan, and revenue and a heavy debt-service burden Russia) have been disruptive. The ouster of the (figure 9.3) limit the scope for increasing Taliban in Afghanistan raised the prospect for government spending on measures to reduce greater stability along the country’s southern poverty. border, while Tajikistan’s support for the war Agriculture remains the largest sector of the effort and its strategic location raised the economy and the major source of employment, international profile of the country and especially in rural areas, where more than 70 strengthened international commitment to percent of the population resides. Reforms in ensure its stability—as evidenced by a large the sector are incomplete, and further measures increase in external aid pledged by interna- to remove distortions and disincentives to tional donors at the Consultative Group investment are widely regarded as holding the meeting held in Dushanbe in May 2003. best prospects for significant poverty reduction Economic reform programs supported by in the country. Poor infrastructure and official the IMF, the World Bank, and other multilateral interference in pricing, procurement, and and bilateral donors gained momentum after marketing are major constraints. the end of the civil war. Since then, real GDP Privatization of small-scale enterprises has growth has been strong and inflation reduced been completed and progress has been made (figure 9.2). The economy remains highly on reform of large-scale enterprises, but several susceptible to external shocks, however, key sectors, including energy, are still dominated including those related to the terms of trade. by public enterprises. 72 TA J I K I S TA N R e a l G D P g r o w t h a n d i n f l a t i o n i n Ta j i k i s t a n , Figure 9.2 1996–2002 (annual percentage change) 12 180 10 160 8 140 Inflation (end-period) Real GDP Growth 6 120 4 100 2 80 0 60 1996 1997 1998 1999 2000 2001 2002 –2 40 –4 20 –6 0 Real GDP growth Inflation (end of period) Source: IMF. Figure 9.3 Ta j i k i s t a n ’s t o t a l e x t e r n a l d e b t , 1 9 9 6 – 2 0 0 2 1,400 140 1,200 120 In millions of U.S. dollars 1,000 100 In percent of GDP 800 80 600 60 400 40 200 20 0 0 1996 1997 1998 1999 2000 2001 2002 In millions of U.S. dollars In percent of GDP Source: IMF. The PRS process—better governance for in addressing poverty: (i) basic poverty benchmarks a better business climate had not been established; (ii) within government, Since the collapse of the Soviet Union and the associ- responsibility for policy formulation was spread ated reduction in incomes and living standards, among different ministries and departments; and poverty has been a major concern in Tajikistan. But (iii) the government’s capacity to implement a after the civil war, Tajikistan faced three impediments coherent set of policies was weak. 73 T H E P O V E R T Y R E D U C T I O N S T R AT E G Y I N I T I AT I V E Tajikistan submitted an I-PRSP in October the role of the participatory process in 2000 and a full PRSP in June 2002. The PRSP implementation and monitoring, and in future sets a target of reducing the proportion of the iterations of the strategy, has not been defined population living in poverty from 83 percent in by the government and is a source of concern 2001 to 75 percent in 2006 and 60 percent by among stakeholders. NGOs described their 2015. For that reason alone, it is clearly relevant interactions with government as having to the country’s needs and aspirations. The dwindled once the PRSP was finished. strategy for reducing poverty involves export- led, labor-intensive growth, with increasing Comprehensiveness, long term reliance on the private sector; efficient and fair The PRSP recognized the multidimensional provision of basic social services; targeted nature of poverty by defining poverty broadly, support to the poorest groups; and efficient to encompass income and nonincome aspects governance and improvement in security. of poverty (including indicators that measure A major goal of the PRSP is to create a skilled, well-being and access to social services). The adequately paid, and corruption-free civil PRSP also highlighted differences in poverty service capable of developing, implementing, incidence across geographical regions. To tackle and monitoring government policies. The civil poverty comprehensively, the PRSP proposed a service faces a challenge in attracting and retain- crosscutting, intersectoral approach, whereby, ing qualified officials with wages at their current for example, the improvement of infrastructure levels. Low wages have also contributed to the would be linked to the improvement of social institutionalization of corruption, with rent- services and vice-versa. The organization of the seeking emerging as a mechanism for sustain- PRSP around nine sector working groups under ing livelihoods. Strengthening of the judiciary is the direction of an overall steering committee also important for protecting the poor and for accentuated its comprehensive view (box 9.1). fighting corruption. However, in several areas (for example, constraints to private sector development and Ownership, participation links between the formal and informal sectors), The PRSP was prepared through a well- greater coverage and more detailed discussion organized, government-led process that could have given the PRSP more cohesion. involved broad participation—including line ministries, scholars, and representatives of Partnerships CSOs. Regional participation was a strong The PRSP is partnership-oriented in that it has feature of the process, as scores of workshops provided a point of reference for donor interac- and conferences were held across the country tions with the government. However, it has not to discuss papers prepared by sector working provided a basis for reconciling what appears groups. Other initiatives were undertaken to to be a substantial difference between govern- supplement the official regional consultations— ment preferences (top priority to infrastruc- among them a donor-funded NGO umbrella ture development) and what donors are willing group (Counterpart Consortium Tajikistan) that to finance (spending on social sectors and on compiled the views of local NGOs and safety nets). communities in an NGO poverty assessment The size and composition of external aid to report. The World Bank and the Asian Develop- Tajikistan has evolved significantly over the last ment Bank supported a study (“Voices of the decade (figure 9.4), moving from humanitarian Poor�) that provided qualitative feedback from assistance to development assistance from a the poorest segments of the population. growing number of donors. Donors were Although the PRS process opened up the involved to varying degrees in the preparation policymaking process to some extent, there is of the PRSP and believed, on the whole, that little familiarity with the final version of the their inputs were taken into account. There is document outside government. Furthermore, some evidence that donors are aligning their 74 TA J I K I S TA N Box 9.1 K e y a r e a s o f Ta j i k i s t a n ’s P R S P Macroeconomic management. Establish a stable economic en- Employment creation. Promote agricultural development, cre- vironment through appropriate fiscal, monetary, and ex- ate a favorable environment for private enterprise (includ- change-rate policies. ing in the informal sector), ensure a well-functioning labor Public administration. Create a skilled, adequately paid, and cor- market, and privatize state enterprises. ruption-free civil service capable of developing, imple- Infrastructure and communication. Provide access to reliable and menting, and monitoring government policies. affordable energy, transport, communications, and safe drink- Social protection. Provide the poor with income through direct ing water. Improve reliability of urban infrastructure. In rural cash transfers, and with increased opportunity through ac- areas, provide access to services. cess to productive assets. Environmental protection. Reduce the vulnerability of the poor Education. Ensure access to basic education for all, and raise to natural disasters through environmental protection, sus- the quality of education. tainable energy, and disaster prevention. Health care. Ensure wide access to curative care and public Tourism. Remove administrative obstacles to private develop- health services, and provide the right balance between pre- ment of the tourism sector. Promote a positive image of the ventive and curative care. country abroad. Agriculture. Continue process of restructuring and privatization of state-owned and cooperative farms. Accelerate land reforms. programs to PRSP objectives, but with no clear to the MDGs. However, as a guide to links to the PRSP policy matrix. Because the operations, it is insufficiently results-oriented, PRSP is not prioritized, it is easy for donor as the vast majority of its indicators relate to programs to claim alignment with some PRSP processes and policy measures rather than to objective or other. outcomes. All major issues are covered in some form, but public actions needed to achieve Results orientation goals are weakly prioritized, and there is a The PRSP is based on a long-term perspective disconnect between the PRSP and the govern- for poverty reduction, with key targets linked ment’s public investment program, with Figure 9.4 E x t e r n a l a i d t o Ta j i k i s t a n , 1 9 9 2 – 2 0 0 1 180 160 In millions of U.S. dollars 140 120 100 80 60 40 20 0 1992 1993 1994 1995 1996 1997 1998 1999 2000 2001 Bilateral Multilateral Source: OECD. 75 T H E P O V E R T Y R E D U C T I O N S T R AT E G Y I N I T I AT I V E neither well integrated into the budget A key part of the Bank’s assistance came as a process. By way of better orienting the govern- result of the Poverty Assessment Report of June ment budget toward poverty reduction, the 2000—a well-argued, analytic document. While PRSP highlights reform of the public expendi- this report preceded the PRSP, it formed a ture management system. Three other cornerstone on which the authorities could elements are in need of strengthening: (i) the base their proposals and measures. It also MTBF; (ii) the Public Investment Program established poverty-line benchmarks and (PIP); and (iii) the annual budget process. highlighted strategic issues to be tackled in the Institutional arrangements for monitoring the labor market, agriculture, education, health, implementation of the PRSP revolve around a social protection, and governance. The update unit that became operational only in the first of the Household Living Standards that is half of 2003. In conjunction with other under way in the State Statistics Agency agencies, the new unit is developing a more represents an important opportunity for the focused set of indicators, including for measur- Bank to provide specific advice to the officials ing intermediate outcomes. and data analysis so as to ensure that the poverty line benchmarks are brought up to Preliminary results date—and comprehensively. The authorities indicated to the evaluation In assessing sectoral differences, the evalua- mission that they have been implementing tion mission’s discussions with representatives measures specified in the PRSP policy matrix to of the PRSP working groups revealed that the the extent they can, citing lack of external Bank’s work on education, poverty analysis, financing as the major factor hindering social protection, health, energy, and agricul- implementation. All ministries contacted said ture/land reform were appreciated for their that their budgets for 2003 and the proposals direct relevance to the PRSP, whereas work in under formulation for 2004 were based on the private sector development, the financial PRSP. However, the evaluation mission formed sector, water, and the transportation sectors the view that not much has happened by way of were seen as having less influence in shaping integrating the PRSP, PIP, MTBF, and govern- the PRSP. ment budget. This is an important task because The Tajikistan CAS of February 2003, the PRSP and the PIP appear to contain differ- formulated after the PRSP, takes a medium-term ent priorities. perspective of 5–6 years—a good fit with the The government’s attempt to coordinate long-term perspective of the PRSP. Its guiding donor assistance has led, paradoxically, to a principles are in line with the PRSP pillars, an proliferation of units with responsibilities that alignment that is further reflected in the Bank’s include liaison with donors. This has resulted lending program: a mix of adjustment lending in overlapping functions and reduced account- and investment lending in the social sectors ability. (education and health), agriculture, watershed management, banking, private sector support, World Bank support—a challenging shift and municipal infrastructure. To improve to community-driven development governance, the Bank will focus on reforming Both within government and among CSOs one public sector management, strengthening the finds awareness and appreciation of the Bank’s customs and tax departments, and improving role in support of PRSP formulation. Bank staff the budget process. provided useful comments on the draft PRSP, The Bank is increasing its program of analyt- helped the working groups mount workshops ical work to strengthen the PRSP implementa- and consultations with civil society, and taught tion phase. All major nonlending activities are key officials about best practices in tackling associated with key areas of the PRSP, including poverty through workshops in Moscow and the business environment and investment Budapest. climate review, public expenditure review, 76 TA J I K I S TA N living standards and measurement survey and • Improving interactions between the Bank’s re- poverty update, social sector expenditure gional office in Almaty, Kazakhstan, and the review (completed), and education sector Dushanbe office, especially on PRSP issues. review (completed). The PER is providing the prioritization that was missing in the PRSP by IMF support—challenges of shifting to a highlighting the main expenditure programs to new way of doing business while be supported in the budget. continuing to foster fiscal discipline A key element in the CAS revolves around IMF staff, especially the resident representa- community-driven development (CDD). tive, provided substantial support to the Discussion of the CAS within the Bank’s board authorities in the preparation of the I-PRSP, but highlighted the riskiness of the CDD approach, were more removed from the PRSP preparation while agreeing that it had a role to play in the process. In particular, they did not participate country. Risks outlined by Board members in the government-led consultative process, included the substantial concentration of seeing it as a domestic stakeholders’ affair. authority at the central government level, However, the PRSP acknowledged the capacity constraints in coordination between influence of the IMF on its macroeconomic the central government and the communities, framework; for example, it indicated that and limited prior Bank activity in the area. The monetary policy would be guided by the PRGF- evaluation mission’s discussions with NGOs supported program that was then under and field visits outside Dushanbe confirmed negotiation. those challenges and risks. Successful The inclusion of macroeconomic manage- implementation of the CDD approach in ment as one of the nine sector working groups Tajikistan will require managing capacity ensured discussion of macroeconomic issues constraints; close monitoring to ensure timely in the PRSP consultation process. Several of the scaling up; and close coordination with NGOs, policy recommendations in the NGO poverty local government, and other donor assistance. assessment referred to the earlier state- Most Bank actions address the basic interventionist rather than market-based problem flagged in the formulation of the approach to economic management, but PRSP: weak institutional capacity throughout others emphasized the importance of strength- Tajikistan. The World Bank Institute’s selection ened institutions and greater transparency as of Tajikistan as a focus country for its capacity- essential elements of a sound macroeconomic building programs should make it possible to framework. In practice, however, there was deliver training and distance learning to key relatively limited public debate on either officials and to have those activities geared to aspect. The issues raised demonstrate the support of PRSP objectives. Regional learning potentially useful role IMF staff could play in opportunities and seminars at the Bank’s raising broader awareness of macroeconomic Kazakhstan hub office, and at other ongoing issues and in promoting broader participation PRSP exercises such as in the Kyrgyz Republic, in policy discussions. could also play a valuable role. The IMF’s technical assistance has been Five other areas merit the Bank’s close broadly consistent with priorities and attention: objectives set out in the PRSP. In recent years, the areas covered by this assistance have • Creating stronger partnerships with other included bank restructuring, tax policy and donors on PRSP issues administration, external debt management, • Monitoring the effectiveness of Bank support monetary operations, macroeconomic statis- in PRSP implementation tics, and economic legislation. Much of the • Better sequencing of the Bank’s work on assistance has been closely coordinated with poverty and social impact assessments (PSIAs), program activities; for example, providing help notably on energy and land reform in meeting structural conditions or redressing 77 T H E P O V E R T Y R E D U C T I O N S T R AT E G Y I N I T I AT I V E capacity gaps revealed by poor implementation increased public spending—if linked to of program commitments. poverty reduction and financed in a noninfla- Overall, the internal IMF policy formulation tionary way (usually through concessional and review process has adapted to the external financing). Although the program’s PRSP/PRGF approach in some areas—such as formal conditions do not cover concessional allowing room for alternative options for borrowing, there is an understanding that meeting some program objectives. In other external borrowing to finance the PIP will be areas, however, notably those related to fiscal limited to no more than 3 percent of GDP. IMF flexibility and to PSIAs, there was little evidence staff justified this limit on grounds of debt and of movement toward the “different way of doing fiscal sustainability, as well as limits in the business� suggested by the PRS approach. country’s capacity to absorb aid. However, The PRGF-supported program is playing a some senior government officials question the crucial role in maintaining a measure of rationale for the limit, viewing it as arbitrary. In discipline in Tajikistan’s public finances, but the the view of the evaluation team, justification of authorities have gone along with important the 3 percent of GDP limit needs to be more aspects of the PRGF-supported program that firmly grounded in a quantitative sustainability they do not fully agree with, mainly because they framework and also in growth analysis. perceive that to do otherwise would reduce The total number of structural conditions development assistance to the country. Partici- under programs has fallen, and their composi- pation in the formulation of Tajikistan’s IMF- tion has changed slightly. In particular, supported programs continues to be limited to conditions have been dropped for public a relatively small circle of government and enterprise restructuring and privatization but central bank officials. Typically, the discussions remain on sectoral issues deemed to have revolve around a draft memorandum of macroeconomic relevance, particularly when economic and financial policies (MEFP) other international financial institutions— prepared by IMF staff. This approach discour- especially the World Bank—have no lending ages government ownership and runs the risk of operation in the sector. Although staff have cited perpetuating dependence on IMF staff. IMF staff government capacity constraints as a reason for and the Tajik authorities should set a timetable streamlining structural conditionality, in practice for turning over responsibility for preparing the streamlining has been interpreted to mean first draft of the MEFP to the authorities. clearer division of labor in relation to the World There has been a modest increase in govern- Bank. The aggregate burden of IMF and World ment spending on social services in recent Bank structural conditionality does not seem to years, and this trend is projected to continue have decreased. under the current PRGF-supported program. There has not yet been a PSIA on any aspect The program also provides for some compen- of Tajikistan’s IMF-supported program, sation to low-income households to mitigate although programs have recognized that the impact of tariff increases in the energy measures aimed at eliminating quasi-fiscal sector and has recently placed a floor on deficits in the energy sector will have severe government spending as part of conditionality adverse effects on the well-being of a large part under the program. With respect to growth, the of the population. PRGF emphasizes governance measures— especially to stop public officials’ interference Bank-Fund collaboration in the affairs of private enterprises—more than The staffs of the Bank and Fund have tried— fiscal measures. not always successfully—to coordinate their The feature of “appropriate fiscal flexibility� work programs. Areas in which collaboration encapsulated in the overall design of the PRGF has been very effective include support to the (see chapter 1) is intended to allow room for government in ensuring that the priority status programs to be more accommodating of of the education sector in the PRSP is translated 78 TA J I K I S TA N into appropriate budget allocations (based on outcome-based indicators are under develop- World Bank analytical work). Collaboration has ment by a new PRSP implementation unit that also been good on issues related to the agricul- became operational in the first half of 2003. ture sector and land reform, though the IMF The PRSP has provided a point of reference has now pulled back from those areas. Collabo- for donor interactions with the government, ration on energy issues appears to have but to date it has not closed the gap between improved over time. government preferences (infrastructure Areas in which collaboration has either not development) and what donors want to finance been effective or has been slow in producing (social sectors and safety nets). Coordination results include PSIAs and expenditure policy. A among donors is poor, and donor alignment promising area for more effective IMF–World with the PRSP should be improved. Bank collaboration is in the preparation of PSIAs The Bank is increasing its program of analyt- on aspects of Tajikistan’s IMF-supported program. ical work in Tajikistan to support implementa- The staffs of the IMF and the World Bank tion of the PRSP. All major Bank activities are maintain contacts with staff of other interna- associated with key focus areas of the PRSP, tional financial institutions, especially the Asian because the 2003 CAS reflects the priorities of Development Bank, a major source of conces- the PRSP. sional loans to Tajikistan. Closer collaboration The consultative process around the PRS among the three institutions would enhance process, including the forums organized for the the effectiveness of assistance being provided NGO Poverty Assessment Report, demonstrated to Tajikistan in at least three areas: (i) support that there is scope to broaden participation in for the development of a well-functioning macroeconomic policy discourse in the country— Medium-Term Expenditure Framework that and that the IMF can contribute to this process. integrates the PIP into the budget process Although the PRGF-supported program is (requiring closer coordination of the various playing a critical role in maintaining a measure of strands of the assistance that each institution is fiscal discipline in Tajikistan, substantial differ- already providing to the budget process or ences of views between the authorities and the PIP); (ii) energy sector reforms; and (iii) PSIAs. staff on aspects of external financing are undermining ownership of the program. In partic- Conclusion—implementation requires ular, the limit of 3 percent of GDP set for external clearer priorities, greater coordination borrowing to finance the authorities’ public The PRSP was prepared through a well- investment program needs to be clearly justified organized, government-led process that in the context of a medium-term framework. involved the participation of line ministries, Participation in the formulation of Tajikistan’s scholars, and representatives of CSOs. But that IMF-supported programs is limited to a openness has not been sustained since the relatively small circle of government and central PRSP was adopted, in part because the roles of bank officials and IMF staff. To foster ownership nongovernment sectors in implementation and and contribute to building capacity for undertak- monitoring are not well defined. We found little ing macroeconomic policy formulation and familiarity with the final version of the PRSP analysis, we recommend that the IMF give the outside government circles, even among those Tajik authorities responsibility for preparing the who participated in the consultation process. . first draft of the MEFP We also recommend that Although the PRSP recognized the multidi- participation in the formulation of the structural mensional nature of poverty and proposed to reform components of programs be broadened tackle it comprehensively through a crosscut- to include implementing agencies; this is ting intersectoral approach, it is insufficiently especially important in areas where IMF involve- results-oriented, lacking adequate outcome ment partly reflects the absence, or limited indicators, clear priorities, and integration with presence, of other international financial institu- the government’s PIP and budget. A set of tions (including the World Bank). 79 10 Tanzania A joint IEO/OED evaluation team made two visits to Tanzania in Febru- ary and April 2003. The latter mission held meetings in Dar es Salaam and Dodoma. Among the sources for the analysis presented here was a survey of 100 stakeholders representing government, civil society, and in- ternational partners. The report of the evaluation team covers Tanzania’s ex- perience with the PRS process up to mid-2003. Subsequent developments, including the PRS review process for the second PRSP and the Third Annual Progress Report, are not discussed in the team’s assessment. A new resolve to fight pervasive poverty donor countries as well as multilateral institu- By any measure, poverty in Tanzania is deep and tions, resulting in considerable inflows of aid. pervasive. In 2001–2002, 12.6 million Tanzani- However, the Ujamaa experiment failed to ans, 36 percent of the population, were living deliver the promised rapid social advancement. under the poverty line, with one in five living in Although its goal was a self-reliant economy, abject poverty. Inequality is high, and regional Tanzania became more dependent on donors. and urban/rural income disparities are marked. The late 1970s and early 1980s were marked Landlessness, lack of agricultural inputs, inabil- by a deepening economic and social crisis. The ity to obtain enough food, lack of access to war with Uganda at the end of the 1970s, and health and education services, dependency and the second oil shock, led to a severe resource powerlessness, and physical disability are the shortage. The collapse of the economic system defining characteristics of the poor. in the 1980s led to a decline in growth. Beginning in 1967, the government adopted Tanzania’s leadership came under increasing a blueprint for the country’s development pressure to undertake democratic reforms, strategy that involved the nationalization of both from within the country and from its most of the economy, as well as the resettle- external development partners, but change— ment of villagers in larger communities in the combining economic reforms with political countryside to improve their proximity to liberalization and decentralization—has not services. This policy stance, based on Ujamaa been easy, in part because of opposition from (or African socialism), was supported by many entrenched groups with a stake in the existing 81 T H E P O V E R T Y R E D U C T I O N S T R AT E G Y I N I T I AT I V E Figure 10.1 Ta n z a n i a a t a g l a n c e Sub- POVERTY AND SOCIAL Saharan Low- Tanzania Africa income Development diamond 2003 Population, mid-year (millions) 36.0 703 2,310 Life expectancy GNI per capita (Atlas method, US$) 290 490 450 GNI (Atlas method, US$ billions) 10.6 347 1,038 Average annual growth, 1997–03 Population (%) 2.3 2.3 1.9 Labor force (%) 2.4 2.4 2.3 GNI Gross per primary Most recent estimate (latest year available, 1997–03) capita enrollment Poverty (% of population below national poverty line) 35 .. .. Urban population (% of total population) 36 36 30 Life expectancy at birth (years) 47 46 58 Infant mortality (per 1,000 live births) 85 103 82 Child malnutrition (% of children under 5) 44 .. 44 Access to improved water source Access to an improved water source (% of population) 56 58 75 Illiteracy (% of population age 15+) 21 35 39 Gross primary enrollment (% of school-age population) 105 87 92 Tanzania Male 109 94 99 Low-income group Female 102 80 85 Source: World Bank. system. The reluctance of the country’s leaders otherwise dominated by serious domestic to embrace change led to a cutback of external disturbances. The most important budgetary aid in 1994. challenge has been the government’s poor After a reform-oriented government was capacity to generate revenue. The main causes elected in October 1995 in the country’s first are tax evasion, discretionary and statutory tax multiparty elections, economic performance exemptions, a very limited tax base, and weak improved, with increased growth, lower tax administration. inflation, and the return of aid donors (table But social indicators paint a stark picture. In 10.1). Since 1990, agricultural output has grown 2002, life expectancy at birth was only 43 years, at an average annual rate of 3.7 percent. Mining while a third of children under five suffered has accounted for a large part of the recent from malnutrition and more than 40 percent expansion as investment by multinationals began from stunting. Infant mortality rates exceeded to generate higher output. In contrast, manufac- 100 for every 1,000 live births. In recent years, turing output has been below expectations. this situation has been exacerbated by In combining economic and political HIV/AIDS, now the leading social concern, with reforms with peaceful conditions at home, serious demographic and economic implica- Tanzania has distinguished itself in a region tions. It is estimated that 10 percent of the Growth and investment indicators in Ta b l e 1 0 . 1 Ta n z a n i a , p e r i o d a v e r a g e s , 1 9 9 1 – 2 0 0 2 (percent) Economic indicator 1991–1995 1996–1999 2000–2002 Real GDP growth 1.8 3.8 5.5 Real GDP per capita growth –1.0 1.1 2.7 Inflation, average 27.4 14.0 5.3 External public debt (percent of GDP) 131.4 98.5 75.7 Source: Tanzanian authorities and IMF. 82 TA N Z A N I A population was infected in 2000. An important tion point for participation in the HIPC Initia- exception to the above numbers is adult tive. Officials at the Ministry of Finance initially literacy. With 76 percent of the adults classified were frustrated by the lack of clarity of what as literate, Tanzania is more literate than many was expected from the PRSP. A country-led and of its neighbors—a product of high primary comprehensive poverty reduction strategy, set school enrollment rates in the 1970s, which in the context of the long-term Development have not, however, been sustained. Net enroll- Vision 2025, already existed in the form of the ment rates for primary education increased NPES. A 1994 report prepared by a joint donor marginally in the 1990s, but still were lower group had initiated a fruitful dialogue with the than in the early 1980s. Since 2000, rates have country’s external partners, and the TAS— increased substantially—to 65.5 percent—as a aimed at coordinating donor assistance—was result of the cancellation of user fees, but the under formulation. So doubt about the added quality of education has declined markedly value of the PRSP is understandable. Neverthe- during the period. less, three years of implementation suggest In the second half of the 1990s, the govern- that the PRSP process has indeed added value ment introduced national strategies for in some important areas, although a process development and poverty reduction and that built more explicitly on existing domestic changed the process by which it made budget initiatives would have been preferable. and policy decisions. Published in 1998, Development Vision 2025 foresaw Tanzania Ownership, participation becoming a middle-income country by 2025. A key contribution of the PRS process has been The National Poverty Eradication Strategy the opening up of a dialogue between the (NPES, 1997) set goals for economic growth, government and civil society. Although inputs income, primary education, access to water and from civil society had little influence on the sanitation, mortality and health, employment, initial policy content of the PRSP, civil society is and other areas, but it failed to specify priori- now more actively engaged in national policy- ties among its many planned interventions, making through channels such as participation making it inappropriate as an action plan. The in the PER, which has broadened to become Tanzania Assistance Strategy (TAS, 2000) was the main avenue for public policy debate and the government’s attempt to restore local strategy formulation in Tanzania. The institu- ownership of development programs after the tionalization of the policy dialogue, through deterioration of relations with donors in the the PER process, where the World Bank played early 1990s. By 1998—before the PRSP process a critical supporting role, has been crucial— was launched—the PER process led by the although it is not possible to attribute this World Bank had become the main avenue for development to the PRS approach per se. The policy debate and strategy formulation among PER process provided a forum for follow-up of domestic stakeholders and donors. the policy debate, very important given initial gaps in the poverty reduction strategy. It is The PRS process—a sharper focus, broadly participatory, government-led, and better data, and wider participation encourages domestic and external partnership. The PRSP process picked up positive and It also facilitates a multisectoral and results- negative features of earlier strategy exercises. oriented approach to the PRSP through links to Some of the eventual shortcomings of the PRSP, the budget and assessments of the poverty including lack of concrete operational focus of expenditure. guidelines and costing of interventions, were Ownership of the PRSP is strongest at the spillovers from earlier efforts. level of top leadership and in the Ministry of The initial formulation of the PRSP took Finance, followed by priority sector ministries. place within an extremely compressed Ownership is much lower in nonpriority central timetable, under pressure to reach the comple- ministries. Among civil society and local 83 T H E P O V E R T Y R E D U C T I O N S T R AT E G Y I N I T I AT I V E government, ownership has been a function of over issues of quality. There is a need to include each organization’s involvement in the PRS targets from—and gradually expand process, and groups based in the capital are resource commitment to—the secondary and more favorable in their view of the process than tertiary sectors to enable Tanzanians to face the those outside. Parliament has been largely challenges of globalization. outside the process, and its ownership is low. The government is planning to revisit the One area in which stakeholder involvement PRS in an effort to strike a better balance has been limited is macroeconomic policy— between income and nonincome aspects of where the government, donors, the IMF, and poverty. According to the Second PRSP the World Bank tend to dominate. While this is Progress Report (April 2003), next steps will partly because of the weak technical capacity of include more attention to crosscutting issues, civil society groups, it also reflects limited including the environment, gender, HIV/AIDS, success by the IMF in framing key issues in a and employment, as well as the implications of manner digestible to a wider audience and in the restructuring of local government. The ensuring that all major issues were included in report notes that a measure of policy that debate. IMF staff have been reluctant to coherence has returned to agriculture with the intervene too actively in what was meant to be completion of a sector strategy. a government-led domestic debate. The financial sustainability of the PRSP in the Some NGO representatives believed that the long term is an additional concern. Invest- World Bank’s role in the overall PRS process ments in primary education and health care, was too influential—sentiments stemming in without significant accompanying efforts to part from the position of the PRSP as a boost domestic resources through income condition for access to HIPC relief. But it was generation and savings, are not tenable in the the World Bank’s insistence on a greater role long run. The government has recognized that for the NGOs that made possible a level of it must broaden its own revenue base in order consultation that, despite its shortcomings, to ensure the sustainability of programs in was wider than experienced before, and which education. opened the door for the broader consultative process that followed. Partnerships From the beginning of the PRSP process, Comprehensiveness, long term external partners were consulted regularly and The narrow focus of PRSP implementation on at key points in the process leading to the PRSP. human capabilities and survival raises concerns This participation was, on the whole, at a about whether the priorities of the PRSP are healthy arm’s length so as not to interfere with consistent with the objectives of the long-term the indigenous development of the PRS. Development Vision 2025—achieving a high During the implementation phase, the process quality of life for all Tanzanians, a society has continued to engage development partners characterized by good governance and the rule in shaping national policies and strategies. of law, and an economy that is competitive, External partners view the PRSP as the guiding diversified, semi-industrialized, and macroeco- framework for development assistance and nomically stable. have attempted to streamline their processes An example of a less than comprehensive accordingly. focus of the PRSP is in the education sector, External partners are also substantively which absorbs the bulk of PRS funding. PRSP involved through various forums, but particu- targets focus exclusively on primary education, larly through working groups. They continue which has also seen the highest budget to wield substantial influence on strategy and increases in recent years. The implementation policy formulation at the sectoral level, which of the Primary Education Development Program eventually feeds through to the national level. (PEDP) has favored access and enrollment rates Discussions are often dominated by external 84 TA N Z A N I A partner representatives and specialists. Such Results orientation domination, and related complaints about The PRSP process is clearly results-oriented external participants setting meeting agendas and focused on outcomes that benefit the or writing consultants’ terms of reference poor. It has enhanced the prioritization and the indicate the continuing influence of external mobilization of financial resources in the partners on domestic policymaking, raising priority sectors (figure 10.2). The share of concerns about the ownership of strategies and priority-sector expenditure increased from an programs, even in high-priority sectors. average of just over half of recurrent expendi- An important outcome of donor efforts at aid tures in FY99 and FY00, the two years prior to harmonization and policy alignment was the the PRSP, to 68 percent of recurrent expendi- establishment of the Poverty Reduction Budget ture in FY02. Improvements in strategic Support (PRBS) group in early 2000. Consisting resource allocation through the MTEF process of 11 like-minded donors, the PRBS group partic- have been supported, and indeed facilitated, ipates in the PER process, holding regular by the participatory PER process, active in meetings with the government, and is seen as an Tanzania since FY98. However, spending within important player in the poverty reduction sectors is not well targeted to the poor, strategy. Support from the PRBS group is directly although the PER uses PRSP targets as the channeled to the budget, with loan conditions benchmark for assessing overall performance based on a commonly negotiated performance in the sector. assessment framework that focuses on The PRSP process has substantively macroeconomic management, public expendi- enhanced national processes for poverty ture management, and governance. Generally, monitoring, and a master plan now guides all external partners are keen to participate in these monitoring activities, but a major gap remains harmonized processes. between PRSP monitoring and actual policy Through the PER process, the government formulation and performance. Information has been able to establish a database on donor across the results chain (inputs, outputs, commitments to the country over the medium outcomes, and impact) has not been integrated term, enabling it to design its MTEF. by policymakers in a meaningful way. Capacity Two issues of concern on partnerships are constraints and the political economy of the low levels of involvement of the business changing policies and budget allocations imply community and the high transaction costs for that the penetration of the process and princi- the government of numerous donor missions, ples at the lower levels, arguably more critical some of which are still not well coordinated. levels for achieving results, is slow. Expenditures in priority sectors of Figure 10.2 Ta n z a n i a ’ s P R S P, 1 9 9 8 / 9 9 – 2 0 0 2 / 0 3 1,200 In percent of total expenditures 60% Billions of Tz shillings 1,000 50% 800 40% 600 30% 400 20% 200 10% 0 0% 98/99 99/00 00/01 01/02 02/03 Billions of Tz shillings % of total expenditures Source: World Bank. 85 T H E P O V E R T Y R E D U C T I O N S T R AT E G Y I N I T I AT I V E In retrospect, the issue of poor domestic marginal impact on the overall disease burden capacity, individual and institutional, has not and other health indicators. been adequately considered, given the broad agenda of reform—both within the process of World Bank support—too intrusive or not implementing the PRSP and more broadly as intrusive enough? part of ongoing government reforms. Establish- During the short formulation period of the PRSP , ing a realistic mechanism for implementation the World Bank played a facilitative role, provid- at the subnational levels of the multiple initia- ing technical advice to the government and tives being implemented in fiscal and adminis- identifying key issues that influenced the sectoral trative decentralization, civil service reform, contents of the strategy and helped promote a and public sector reform remains the biggest more participatory process. Stakeholders were challenge that the government faces in deliver- critical of the Bank’s “intrusive� influence in the ing the results promised by the PRSP. Lack of initial stages of the formulation of the strategy capacity, particularly acute at the subnational but acknowledged that this tendency to level, has led to serious delays and may intervene was tempered by the country office. ultimately lead to inadequate outcomes. Overall, the principal drafters of the PRSP described the formulation as a government-led Preliminary results process, with the Bank and Fund not influencing In the initial phase of its implementation, the the substance of the final strategy. PRSP focused on improving service delivery, Civil society members, in contrast, criticized particularly in education and health, by increas- the Bank for not exerting more influence on ing social sector expenditures. The income- the government during formulation to increase oriented elements of a broader growth and participation. But they were cognizant of the poverty reduction strategy were delayed by the dilemma faced by the Bank in urging the late completion of sector-specific strategies, government to expand the consultative notably for agriculture and rural development. process while respecting the country-led This also reflects the legacy of the initial focus principle. Further, the Bank, like many civil on social expenditures of the HIPC process. society groups, was reluctant to trade progress The completion of sector strategies and action toward the Completion Point under the HIPC plans in later years and the more elaborate Initiative for a more extensive consultative PRSP progress reports have helped to expand process. During implementation, the Bank has on and balance the implementation of the promoted the participation of civil society strategy. within the national PRSP processes—with Although it is too early to give a definitive encouraging results. Civil society participation evaluation of the impact of the PRS approach, has been a feature of the PER process and is at the time of the evaluation mission observers being further strengthened year by year. reported a noticeable improvement in service An area where the Bank could have been delivery, including at the village level. The better prepared for the PRSP was in assisting Second PRSP Progress Report records a sharp Tanzania to conduct a household budget increase in gross enrollment at the primary survey before 2000. During implementation of school level, from 77.6 percent in 1990 to the PRSP, Bank staff have provided substantial 100.4 percent in 2002, although concerns technical advice in analyzing the results of the about the low quality of education persist. In latest household budget survey and in the health sector, the supply of drugs has establishing a national poverty monitoring greatly improved, including at local health system. The Bank’s technical expertise and units, which is attributed to an innovative cost- sectoral inputs in education and health and sharing approach drawing on the views of local through the PER process have supported the communities. However, the high prevalence of first and second PRSP Annual Progress HIV/AIDS and malaria has implied a rather Reports—those contributions have been 86 TA N Z A N I A greatly appreciated. The PER process, in partic- development, improved social infrastructure, ular, has been well recognized for its support of public sector reform, and institution building— the PRSP and its principles, and the Bank has is in line with the priorities of the PRSP. New played a leading role in promoting and expand- commitments across the three-year CAS period ing the PER from an external technical assess- reflect a similar picture of alignment: projects ment to a country-led participatory process that fall into PRSP priority sectors comprised 75 involving a wide range of stakeholders. percent of new commitments. Compared with In other areas of the Bank’s analytical support, the previous three-year period, Bank commit- some key shortcomings came to light in the ments have broadened from public sector formulation of the PRSP—and some of these governance and private sector development remain. There are continuing gaps in knowledge into priority areas such as education, social of the linkages between micro- and macroeco- protection, and water supply (figure 10.3). nomic policy, specifically in relation to: (i) the The recently introduced PRSC signals a social and poverty impacts of macroeconomic further shift toward alignment with country and structural policies; (ii) the sources of processes. The PRSC is a programmatic lending growth; and (iii) the poverty impact of alterna- instrument specifically designed to support the tive policy and strategy options. At the sectoral implementation of the PRSP. It supports level, the depth and breadth of the Bank’s analyt- government efforts in three areas: (i) reforms ical work was uneven prior to the formulation of in agricultural and rural development, (ii) the PRSP. Work in agriculture, for example, institutionalizing Tanzania’s monitoring and should have started much sooner, and the Bank evaluation systems, and (iii) strengthening could have been more assertive in promoting the public sector performance and efficiency and growth agenda in the PER and MTEF. effectiveness in the use of resources. It The Bank’s lending portfolio in the last three contains fewer conditions than previous policy- years has been well aligned with the PRSP. The based lending instruments. focus of the current CAS—private sector and The Bank has continued to promote the infrastructure development, sustainable rural partnership principle. Its proactive role in the World Bank commitments by sector, Figure 10.3 FY98–00 and FY01–03 Water Supply and Sanitation Social Protection Rural Sector Private Sector Development Public Sector Governance Health, Nutrition, and Population Financial Sector Economic Policy Environment Energy and Mining Education 0% 10% 20% 30% 40% 50% 60% 70% Share of commitments FY01–03 FY98–00 Source: World Bank. 87 T H E P O V E R T Y R E D U C T I O N S T R AT E G Y I N I T I AT I V E participatory PER process, the alignment of the was expected to remain relatively unchanged at PRSC within the framework of the multidonor about 16 percent of GDP. Grants as a percent- PRBS group, and its engagement with the age of GDP were also projected to decline by a Development Assistance Committee of the full percentage point during the program. The OECD are important steps that have enhanced implication was that increasing priority the collaboration of donors in Tanzania. expenditure would require reduction in Complaints remain about the large numbers of nonpriority spending relative to GDP. At that missions that regularly arrive from the Bank. At stage, the higher resource inflows expected the same time, stakeholders have acknowl- from the HIPC Initiative and grants had not yet edged that complaints at the sectoral level are been sufficiently factored into the projections. strongly correlated with capacity problems The original PRGF thus projected a fairly within ministries, an area that Bank and donor restrictive fiscal stance. technical assistance has been unsuccessful in Over time, the consultation processes under addressing. the PRSP had a significant impact on the content of macroeconomic policies—and the IMF support—gradual alignment with the PRGF-supported framework did show consid- PRSP and opening up of the debate on erable flexibility. In this sense, the PRGF program design eventually became aligned, although The IMF’s program in Tanzania, supported by incompletely, with the key objectives of the the PRGF, complemented the overall strategy PRSP—even though the alignment did not for growth and poverty reduction in three occur through preparation of the original PRSP. respects. First, public expenditures were In assessing how the PRSP process affected the increasingly prioritized, with emphasis on scope of policy choices available to the govern- accountability and ability of local level recipi- ment and how the PRGF adapted to these ents to provide feedback to the center. Second, policy choices, we examined in depth four the program encouraged further reforms in key policy issues in the macroeconomic area. They areas of economic activity such as taxation, land included two cases, the PRGF programming reform, and the financial sector, notably framework and cash budgeting, where a microfinance, with a view to expanding private broader debate in the PER working groups led sector activities, not only in the modern sector to a more flexible fiscal stance, and increased but also in the informal sector and the rural priority expenditure better insulated from areas where most of the poor derive their liveli- temporary budget shocks; one case, trade hood. Third, with respect to growth, the reform, where the political nature of the program focused on the consolidation of problem and the lack of a domestic debate on macroeconomic stability and creation of the lines of the PER evaded easy resolution; and incentives for investment. a last case, fiscal federalism, which, though The medium-term objective of the PRGF was considered a priority by the government and to create conditions favorable for higher with clear relevance for the PRSP, was growth, with a particular focus on poverty approached very cautiously by IMF staff, in light reduction. Macroeconomic objectives included of the fairly limited local capacities and the increasing growth from 5.2 to 6 percent IMF’s own experience in other countries. annually by the end of the program and Lack of capacities within government led to reducing inflation from 7 percent to rates delays in policy formulation and implementa- comparable with those among Tanzania’s main tion, although the process of interaction and trading partners—about 4 percent. Monetary debate finally led to the convergence of views expansion was to be kept at about 10 percent on key macroeconomic issues. Concerns by per year. Surprisingly, for a program focusing some key government officials and donors that on poverty reduction and the expansion of the initial macroeconomic framework did not service delivery, total government expenditure sufficiently accommodate increased aid flows, 88 TA N Z A N I A helped trigger a broader debate on aspects of However, three years after its launch, policy- program design. That debate benefited from makers have been able to flesh out the strategy contributions from local and foreign academic and improve poverty monitoring mechanisms. consultants. Discussions in the PER macro The process has provided an opportunity to working group and workshops organized by mobilize and focus resources in the priority the government, informed by these technical sectors; enhance the monitoring of poverty contributions, did help move the process reduction targets; and encourage broad-based forward and contributed to some significant participation of civil society in national adaptations in the macroeconomic framework processes—all previously lacking in national underlying the program design. In this sense, strategy formulation. the substantive revisions made in the original Ownership is ultimately judged by commit- PRGF program design reflected an effective ment to the strategy, which has been high, contribution of the institutional framework set especially among the top leadership. Among up for consultations between stakeholders and civil society and local government, ownership the use of outside expertise to strengthen has been mixed—with Dar-based groups more capacity. However, this participatory process favorable to the process than those outside, was incomplete, as it largely involved govern- and the priority ministries more committed to ment, the Bank and the Fund, and bilateral the PRSP than others. An important limitation donors—with civil society stakeholders largely is that parliament and the private sector have on the sidelines. been only marginally involved in the process. The IMF’s technical assistance is focused on To extend the political legitimacy of the PRSP the key areas of expenditure management, tax process and ensure sustainability, it will be administration, and fiscal decentralization, and necessary to bring both groups more formally therefore appears to have been aligned with into the process. PRSP objectives. However, government officials With time, the PRS process has become more are demanding more say in the use and mode participatory and inclusive. Early conflicts over of delivery of the IMF’s technical assistance, civil society participation reflected ambiguity which they still see as too closely linked to about what participation was meant to program targets rather than capacity building. achieve—consultation versus participation in The PRS process has had impacts on internal policy decisions. Civil society involvement was IMF procedures, with the timing of IMF more limited in macroeconomic policy issues missions becoming better aligned to the than in other areas. While partly an issue of budget cycle of the government. Still, with a technical capacity, that limitation also reflected limited presence in the field, the IMF has yet to ambiguity about the IMF’s role in the broader meet demands for broader interactions with policy issues, including its role in framing those civil society groups and to devote more time to issues in a manner digestible to a wider explaining its macroeconomic policies. audience and ensuring that all major policy The IMF’s ability to gauge the impacts of issues were included in the domestic debate. these interventions on poverty has been However, the institutionalization of the policy limited by the lack of poverty and social impact dialogue, through the PER process, where the assessments. World Bank played a critical initiating role, was crucial. It provided a forum for follow-up of the Conclusion—a healthy process that has policy debate that was very important given become embedded in domestic initial gaps in the strategy. The PER process institutions overcomes defects in the today is broadly participatory, government-led, original document and encourages domestic and external partner- Tanzania is a good illustration that it is the PRS ship while focusing the PRSP on results through process, not the document itself, that matters. links to the budget and assessments of the The initial PRSP was weak in many respects. poverty focus of expenditure. 89 T H E P O V E R T Y R E D U C T I O N S T R AT E G Y I N I T I AT I V E On the whole, the PRS process has been However, both central and local government results-oriented and focused on outcomes that structures lacked adequate technical skills to benefit the poor. It has enhanced the mobiliza- implement and follow up on key aspects of the tion of resources and the focus on priority process. Lack of capacity led to serious delays sectors—with PERs as major inputs. During the and still may produce inadequate outcomes. process the shares of priority sectors in expendi- The PRSP process has fostered greater coordi- tures have risen steadily, although PER analyses nation among the World Bank and IMF teams have shown that spending within priority working on Tanzania, helping to dispel the sectors needs to be better targeted to the poor. uncertainty among Tanzanian officials over what The process has substantively enhanced was expected from a PRSP. But important gaps national processes for poverty monitoring, and remain—notably the setting of priorities for a poverty monitoring and master plan now timely delivery of PSIAs. The PRS process should guides all monitoring activities. However, the be more explicit about what the Bank and Fund feedback between monitoring of results and will deliver. For example, a clearer delineation, as policy actions requires improvement. part of the PRSP framework, of what analytical The success of the PRS process was premised inputs should be delivered, when, and by which on effective management and coordination at institution, would be consistent with the results- the subnational level, particularly in rural areas. based emphasis of the PRS approach. 90 11 Vietnam T he IEO evaluation team conducted interviews with stakeholders in Hanoi and Ho Chi Minh City in June/July 2003. Among the sources for the analysis presented here was a survey of 78 stakeholders in gov- ernment, civil society, and the donor community. The evaluation reviews de- velopments up to December 2003. Successful poverty reduction expenditures, including poverty reducing Vietnam was one of the first Asian countries to expenditures. produce a PRSP. Its tradition of basing public Per capita income rose steadily from US$272 policy on medium-term socioeconomic in 1996 to US$428 in 2002. The 1997/98 planning and its reputation for consensus- Vietnam Living Standards Survey estimated that based decisionmaking (albeit within the 90 percent of poor households lived in rural framework of a single-party system) made it an areas. With growth strongly concentrated in interesting case for evaluation. Vietnam is major centers such as Ho Chi Minh City, this widely considered a development “success has led to perceptions of a widening gap story,� having reduced poverty levels in the between rich and poor within the country. past decade more rapidly than virtually any Literacy rates—of more than 90 percent in other low-income country. 1998—are high compared with other low- The Vietnamese economy experienced income countries. strong growth over the last decade, averaging Vietnam is one of the largest aid recipients just under 6 percent annually since 1996 (table in the world; Japan is its largest bilateral donor, 11.1). During the Asian crisis, growth slowed to providing more than half of total official aid. 3.5 percent, but recovered thereafter. Inflation Vietnam was one of the pilot countries for the has been in the single digits since 1996, falling CDF and was a case study for OED’s 2002 to negligible levels in 1999 and 2001. Central evaluation of the CDF.1 It also was one of the government fiscal deficits have been modest six countries chosen in 1998 as a pilot for for the most part, averaging less than 1 percent enhanced World Bank–IMF coordination. of GDP per year between 1996 and 1999. They The authorities have demonstrated clear increased to more than 3 percent in 2000 and commitment to development and poverty 2001, reflecting increases in government reduction and, with the benefit of solid growth, 91 T H E P O V E R T Y R E D U C T I O N S T R AT E G Y I N I T I AT I V E Figure 11.1 Vietnam at a glance East POVERTY AND SOCIAL Asia & Low- Vietnam Pacific income Development diamond 2003 Population, mid-year (millions) 81.3 1,855 2,310 Life expectancy GNI per capita (Atlas method, US$) 480 1,080 450 GNI (Atlas method, US$ billions) 38.8 2,011 1,038 Average annual growth, 1997–03 Population (%) 1.2 1.0 1.9 Labor force (%) 1.6 1.1 2.3 GNI Gross per primary Most recent estimate (latest year available, 1997–03) capita enrollment Poverty (% of population below national poverty line) 29 .. .. Urban population (% of total population) 26 40 30 Life expectancy at birth (years) 70 69 58 Infant mortality (per 1,000 live births) 20 32 82 Child malnutrition (% of children under 5) 30 15 44 Access to improved water source Access to an improved water source (% of population) 77 76 75 Illiteracy (% of population age 15+) 4.6 10 39 Gross primary enrollment (% of school-age population) 115 111 92 Vietnam Male 117 112 99 Low-income group Female 113 111 85 Source: World Bank. Selected macroeconomic indicators for Ta b l e 1 1 . 1 Vietnam, 1995–2002 (percentage of GDP) 1995 1996 1997 1998 1999 2000 2001 2002 Real GDP growth 9.5 9.3 8.2 3.5 4.2 5.5 5.0 5.8 Inflation (end of period) annual percent change 12.9 4.5 3.6 8.6 0.1 –0.5 0.7 4.0 Total expenditure 23.8 24.9 24.8 22.2 20.6 23.9 25.6 24.8 General government education and health expenditures — — — 5.0 4.9 5.2 5.8 — Total poverty reducing expenditures a — — — — 5.1 5.4 6.0 6.1 a. IMF estimate based on (i) current expenditure on education, training, health, and family planning and (ii) capital expenditure on education, training, health and poverty-related projects in the agriculture, transportation, electricity, and water sectors. Source: IMF. have exercised strong leadership in setting the phase in the country’s policy cycle. The reform agenda. For the most part, the authori- planning phase led to the production of the ties, the donor community, and the IMF have current 10-Year Socioeconomic Development agreed on broad reform objectives—for Strategy and 5-Year Socioeconomic Develop- example, poverty reduction and private sector ment Plan (SEDP). The I-PRSP was presented development—but there have been disagree- to the Executive Boards of the Bank and the ments on how, and how quickly, these IMF in March 2001, along with a request for a objectives should be achieved, resulting in three-year PRGF-supported program and the lengthy delays in reaching agreement on an first World Bank–supported PRSC. The full IMF-supported program. Comprehensive Poverty Reduction and Growth Strategy (CPRGS, as the PRSP is known in The PRS process—concurrent strategies Vietnam) was finalized just over a year later. At The government of Vietnam initiated the the same time, the second review under the preparation of an Interim PRSP in mid-2000, PRGF-supported program was also completed. which coincided with an intense planning Its principal elements are outlined in box 11.1. 92 VIETNAM Ownership, participation Vietnam’s PRSP Vietnam’s CPRGS is widely considered to be Box 11.1 policy matrix domestically owned. The authorities maintained a firm hand throughout the drafting process, Promote economic growth, raise income, and create opportunities much of which was done in Vietnamese, with for the poor only selected drafts translated into English. A legal environment for fair and competitive business By contrast, the participatory process used A stable macroeconomic environment in the formulation of the CPRGS, was not fully Public administration reform country driven, with international partners (the Develop major sectors and areas to ensure sustainable growth, and World Bank and international nongovernmental for poverty reduction organizations) initially assuming a leading role Policies and measures to develop sectors and create opportunities in the organization and financing of public for the poor consultations. Government engagement with Balanced development, gender equity, and advancement for women the participatory process intensified during the Social safety net to support the poor and vulnerable formulation stage, but the true extent of the Monitor and evaluate authorities’ ownership of, and commitment to, Improved macroeconomic statistics a participatory approach to policymaking will Poverty monitoring only be tested once they take the driver’s seat in future participatory consultations. To ensure that the process is viable over the long term, several shortcomings should be expenditures and poverty reduction. This addressed. For example, discussions of the difference led one major donor to encourage macroeconomic framework and many of the the authorities to add a new chapter to the associated structural reforms underpinning the CPRGS on large-scale infrastructure. strategy were notably absent from the partici- patory process, reflecting, among other things, Comprehensiveness, long term a lack of interest and expertise in discussing the The CPRGS addresses poverty as an issue macroeconomic framework on the part of the defined not only by levels of income, but also organizers of the participatory exercises, the by demographics, ethnicity, gender, and access absence of a systematic effort on the part of the to social services. The CPRGS also acknowl- IMF to encourage public discussion of these edges that the rapid economic growth has led issues, and the perceived lack of pressing to a widening of income distribution and the macroeconomic problems that would warrant emergence of urban poverty as a major the allocation of scarce time during participa- challenge in parts of the country. tory workshops. The CPRGS is seen as results-oriented and, The relationship of the CPRGS to the on balance, has made a positive contribution to existing indigenous planning processes and the policymaking process in Vietnam. It is documents, notably the 10-Year Strategy and generally seen to have helped improve the 5-Year Development Plan, also remains interministerial coordination which, along with unclear, with the authorities looking to the its timetable for policy implementation during socioeconomic development plan as the 2003–2005, has promoted transparency. While embodiment of the country’s development it has led to improvements in the mandate and strategy, whereas donors and the international capacity of data producing entities such as the financial institutions view the CPRGS as playing General Statistics Office, the CPRGS’s monitor- this role. The two strategies are largely consis- ing and evaluation framework—containing 136 tent, but their emphasis differs: the socioeco- indicators—appears to have been excessively nomic development plan places more ambitious and would have benefited from emphasis on industrialization and moderniza- streamlining and priority setting to better tion, while the CPRGS emphasizes social reflect the country’s administrative capacity. 93 T H E P O V E R T Y R E D U C T I O N S T R AT E G Y I N I T I AT I V E However, prioritization among these indicators authorities’ priorities in the CPRGS initially with regard to development objectives and masked differences between PRGF program measures continues to be a challenge for the commitments and the authorities’ desire for a government. substantial and continuing role for the public sector in the economy. Partnerships At the level of the macroeconomic framework, The CPRGS process has brought some increase there were persistent differences of view between in donor coordination, notably through staff and government authorities on GDP formation of a Like-Minded Donors Group, measurement and macroeconomic prospects. some of whose members are involved in the co- This resulted in two separate medium-term financing of PRSCs. Scope exists to improve forecasts, with the authorities’ forecast underpin- coordination—some major donors are still ning the CPRGS and that of IMF staff underpin- reluctant to adapt their programs to align with ning the PRGF-supported program. However, the strategy. The breadth and generality of the this difference of view was transparently reflected CPRGS, moreover, have made it relatively easy in IMF staff reports and the CPRGS. to justify most pre-existing donor activities and Despite such differences, and perhaps therefore have not constrained donors from reflecting the authorities’ good track record in pursuing their own preexisting priorities. conducting fiscal policy, significant fiscal flexibil- ity was integrated into the PRGF-supported IMF support—flexibility, despite some program. The program accommodated higher- persistent disagreements than-expected levels of spending in priority Performance has been mixed with respect to social sectors, partly by avoiding overly conser- adapting IMF-supported programs to reflect vative projections of official development both PRSP principles and the key features assistance (figure 11.2). Particularly notable was under the PRGF. Initially, alignment of policies the absence from the PRGF-supported program in the PRGF-supported program with those of of quantitative conditionality (such as perform- the CPRGS was only partial, although it ance criteria or benchmarks) on major fiscal improved over time. This was particularly true variables. with respect to reform of state-owned During the years of the PRGF-supported enterprises (SOEs), where vagueness about the program, poverty reducing spending increased Poverty-reducing spending in Vietnam, Figure 11.2 1998–2003 Total Poverty Reducing Expenditure... …and as a Share of Total Expenditure 7.0 30 6.5 Percent of GDP 6.0 25 Percent 5.5 5.0 20 4.5 4.0 15 1999 2000 2001 2002 2003 1999 2000 2001 2002 2003 (Budget) (Budget) Source: IMF. 94 VIETNAM as a share both of GDP and of total expendi- reform. This is an area of overlapping interest ture. However, it is not clear that this trend can between the Bank and the Fund, but one for be attributed to the PRGF-supported program which the Bank has the lead responsibility for (or the CPRGS), since it began well before the the design and monitoring of conditionality. adoption of the program (or the strategy). Reflecting different mandates and institutional The government made some effort to assess priorities, the Bank’s objective of promoting the expected impact of proposed program private-sector development through large measures on the poor and most vulnerable. Of numbers of equitizations of small- and particular note were efforts to establish and medium-sized SOEs—while in line with the finance a social safety net for workers displaced government’s priorities—did not address the by SOE reform. It also made a preliminary Fund’s medium-term fiscal concerns, which attempt to assess the social and poverty impact were focused mainly on the largest SOEs. In the of trade liberalization, but the analysis was of event, the IMF attempted to address its longer- only limited value in informing policy design. term structural concerns through short-term Beyond these examples, PSIAs did not figure macroeconomic conditionality (a limit on prominently as an input to PRGF-supported credit growth), a tactic that was neither program design, partly reflecting the IMF adequate nor successful. The rather general staff ’s sense that proposed reforms would not language on SOE reform that was contained in have a major negative impact on the poor. the CPRGS did not help to clarify these issues. Looking forward, an agenda for PSIA work has This experience shows that, in formulating been endorsed jointly with World Bank staff, conditionality in areas of overlapping Bank- including with respect to trade and tax reforms. Fund interest, it should not be assumed that Bank and Fund objectives are identical—nor is Bank-Fund collaboration it always appropriate that they should be so. The JSA of the CPRGS does not appear to have Therefore, where differences in approach or played a significant role in the CPRGS process— priority exist, they need to be discussed more few bilateral donors or officials were aware of its explicitly. This should go beyond a simple contents. If JSAs are to serve their intended description of the division of labor in staff purposes (one of which is to encourage coordi- reports to include substantive discussion of the nation between the Bank and the Fund), they consistency of Bank and Fund objectives. must provide a clear and frank assessment of a Even though Vietnam was known for strong country’s PRSP. In Vietnam, two omissions from ownership of its reform program, the Fund’s the JSA undermined its usefulness. There was design of the PRGF did not take advantage of no explicit discussion of the implications of the the opportunity to streamline conditionality. ongoing disagreement with the authorities over Instead, the PRGF required numerous prior parts of the macroeconomic frameworks conditions and increased the number of contained in the PRGF-supported program and structural conditions. the CPRGS. Second, the JSA failed to provide an accurate description of the relative roles of Conclusion—merging parallel policy external agencies and the authorities in the processes participatory exercise. Vietnam is an atypical low-income country Problems related to Bank-Fund coordination because of the extent of domestic ownership for also were evident with respect to program the reform process, the breadth and level of conditions. The clear division of labor between donor support, the strong tradition of consen- the Bank and the Fund on structural condition- sus-based decisionmaking, and high literacy ality did not overcome shortcomings in the rates. It raises important questions about the coherence of the two institutions’ approach on value added of an IMF-supported program in a some issues. This was particularly evident with low-income country experiencing macroeco- respect to state-owned enterprise (SOE) nomic stability and no pressing need for external 95 T H E P O V E R T Y R E D U C T I O N S T R AT E G Y I N I T I AT I V E financing from the IMF. While program financ- “doing something right.� This should raise the ing was clearly appreciated in Vietnam, it is not comfort level of the international community obvious that a PRGF-supported program was the with a poverty reduction strategy that is truly most appropriate mechanism to support sound home-grown and country-driven. macroeconomic performance. Most donors With respect to the impact of SOE reform on appear to have valued primarily the “signalling� displaced workers, the PSIA undertaken before role that the program provided, but discussions adoption of the PRGF-supported program and with bilateral donor representatives suggested PRSC I proved useful. However, a medium- that the absence of such a program would not term agenda for PSIA was articulated only two have significantly affected their aid decisions. years after the design and approval of the While these considerations limit the extent PRGF-supported program. Where resources to which the lessons of Vietnam can be extrap- are not readily available within the Bank or the olated to other country situations, a number of Fund to undertake the PSIA in a timely manner, lessons do emerge. One concerns the relation- staff should be encouraged to approach other ship between the PRSP process and indigenous stakeholders to meet PSIA needs. planning processes. Vietnam remains commit- Finally, the importance of broad-based ted to using the SEDP as the paramount ownership and participation implies greater mechanism to articulate its policy agenda over emphasis on building long-term relationships the medium term. Comments from donors and between the Fund and domestic stakeholders. the Bank and Fund, in contrast, suggest that IMF’s resident representatives are well placed the CPRGS, not SEDP, should fill this role. to develop such relationships—and they have Paralleling this difference of opinion, the SEDP done so in Vietnam, with good results. Staff at is more widely known within Vietnamese IMF headquarters should take full advantage of society, and the CPRGS is better known among this achievement by systematically integrating international stakeholders. the resident representative into program The merging of these two processes could negotiations and surveillance work. be an important part of any effort to clarify More broadly, if the participatory expecta- policy priorities and guide budgetary alloca- tions of the PRS Initiative are to be met, a tions. The resulting product could be more sustained effort may be needed over the long clearly linked to the internal budgetary process term to build domestic capacity for meaningful and resource allocations. The significant participation in discussions of the macroeco- progress made over the last decade in reducing nomic framework. Such a strategy would need poverty in Vietnam suggests that the authori- to promote understanding of the relevance of ties are—as one donor representative said— the macro-economy to poverty reduction. 96 ANNEX: GUIDELINES FOR THE COUNTRY CASE STUDIES The 10 case studies summarized in this volume criteria set out above: Mozambique, Nicaragua, were prepared under the following guidelines, Tajikistan, and Tanzania (joint OED-IEO); which were developed jointly by OED and IEO Albania, Cambodia, Ethiopia, and Mauritania in advance of the country work. (OED); and Guinea and Vietnam (IEO). Approach Details on Methodology Given the country-specific nature of the PRS The following guidelines define the basic process, case study analysis will be used as a parameters for the case studies, and can be key source of evaluative material. The method- further tailored to reflect country characteris- ology will combine two types of case studies tics. The guidelines have been informed by identified in the evaluation literature, “program experience from lead missions in Albania implementation� (primarily explanatory) and (OED) and Tanzania (joint OED-IEO). “illustrative� (mainly descriptive). 1 The case A local consultant will be sought to assist studies will be “program implementation� case in each of the case studies. The local consultant studies as they intend to investigate the will be expected to contribute to the team in implementation of the PRS process in each line with their level of expertise and country relative to its underpinnings. They will knowledge. This is expected to yield a range of also be illustrative in nature as they will add in- local consultant assistance across the case depth examples to other sources of evaluative studies, from assisting in scheduling meetings material including external evaluations, and providing local knowledge, to preparing thematic studies, cross-country analysis, and background papers and briefs. Local consult- survey work. ants will also be used to coordinate implemen- To capture the diversity of country experi- tation of the country stakeholder survey (see ences with the PRSP Initiative, OED plans to below). conduct eight case studies for the evaluation, Key informant interviews will be held in four of which will be conducted jointly with IEO. Washington D.C. and in the field. Interviews will IEO is planning to conduct another two case be conducted with relevant World Bank and IMF studies independently from OED. In total, staff who have been associated with the country therefore, OED and IEO will perform an in- team and/or the PRS process. This will be depth investigation of 10 of the 23 countries followed by in-country consultations to be with completed PRSPs through December 2002. conducted for a period of at least two weeks. A range of criteria were considered in select- These consultations are the primary input for the ing the PRSP countries for case studies: fieldwork and will be targeted at a broad range of geographic balance; HIPC and non-HIPC; range stakeholders including; key ministries (central of country conditions and PRSP experience; and line), key government units involved in the and both mature and recent programs. Country PRS process, local governments, bilateral and selection also took into account the country multilateral donors, local and international coverage of previous or ongoing work. 2 The NGOs, private sector representatives, media, following countries were selected based on the and, wherever possible, Parliamentarians. 97 T H E P O V E R T Y R E D U C T I O N S T R AT E G Y I N I T I AT I V E Interview modalities will vary and may include documents and databases, in-depth interviews group consultations. Finally, an exit workshop of key informants both in Washington D.C. and (see below) will be held, which will target in the country, and an in-country workshop. stakeholders consulted during the mission. Relevant documents may include official In-country field visits outside of capital will government data and reports, Bank and Fund be conducted in each case country, unless analytical and advisory work, related policy precluded by country conditions. Advance field documents, staff reports to the Boards of the visits were conducted in both Albania and IMF and World Bank, internal documentation, Tanzania and provided considerable value- prior OED and IEO evaluations, external added. The modalities of field visits will vary evaluations and materials from research from country to country. institutes and civil society. An informal exit workshop will be held in An extensive analysis of the data will be each case country at the conclusion of the carried out using the technique of triangulation mission. The purpose of the workshop is to in which the reliability of findings will be seek feedback on the emerging findings of the developed through multiple data sources and case study from a representative group of findings will be validated through agreement stakeholders, including stakeholders involved across multiple types of data. An example would in the field interviews. be assessing process changes in public expendi- A stakeholder survey jointly designed by ture management in a country. The country’s OED and IEO is being conducted in each Public Expenditure Review may yield an initial country to provide input for both the OED and finding which could then be verified with IEO studies. The survey will supplement multiple data sources (for example, interviews information gained from key informant with Bank staff in the regional Public Expendi- interviews and will target each category of ture thematic group, officials in the Ministry of stakeholder involved in the PRS process. Finance, and civil society). Economic data on Coverage will include the central government budget execution could be cross-checked ministries (Planning and Economy/Finance) against budget laws and qualitative data on and Central Bank, line ministries, donors, disbursements and procurement, that is, data NGOs, and representatives of the private sector would be validated across multiple types of and other civil society groups (for example, data. Likewise, to assess the realism of trade unions, labor unions, academia, media, macroeconomic projections made by the IMF, and faith organizations). The goal of the survey one may assess the extent of deviations from is to obtain institutional responses (not individ- recent trends, compare projections with ual views) on the PRSP and Bank and IMF outcomes, and cross-check that against views performance in supporting the process. As a expressed at the time by the authorities, review result, the criteria for respondents include good departments, and the World Bank. knowledge of the PRS process and familiarity with their respective institution’s positions on Phases and Key Deliverables specific issues. A core set of stakeholders has Phase 1—Desk Review: A desk review will be been identified, which each country survey conducted in advance of field work in each of must target, including key ministries involved in the case countries. The desk review will analyze the PRS process, but outside of these parame- and summarize main findings from available ters, the specific coverage and number of research materials and collate the necessary responses is likely to vary from country to data. A desk brief will be prepared following a country. Local consultants will be used to standard outline to provide an initial basis of administer surveys in each case country further information for the country study. enhancing the independence of the exercise. Phase 2—Washington-Based Interviews Case studies will use multiple methods for and Mission Planning: Phase 2 will build on data collection, including reviews of relevant the desk review through interviews with Bank 98 ANNEX: GUIDELINES FOR THE COUNTRY CASE STUDIES and Fund staff. Key persons for interviewing in may present their findings to key internal the field will be identified, evaluative questions stakeholders at a World Bank or IMF brown bag will be tailored around the emerging evaluative lunch. The draft final report will be shared with hypotheses, and arrangements for the the authorities for comment. workshop and field interviews will be finalized. Phase 3—Country Visit: The field work Staffing of Joint Case Study Teams will investigate the key evaluative questions in- For joint case studies, teams will consist of at least country and seek to address open issues. one core evaluation team member from each Structured interviews with key informants and evaluation unit, along with at least one senior in- an in-country workshop will form the basis for country national consultant. Either OED or IEO consultations. As noted above, an in-country will be assigned overall responsibility for coordi- stakeholder survey will be applied after field nating the case study. The team will also be work has been completed. supported by junior consultants to be assigned Phase 4—Final Washington-Based by each unit as deemed necessary. The OED and Analysis: In Phase 4, the case study team will IEO teams will nominate a case study coordina- agree on the main findings and supportive tor. His/her responsibilities include overseeing evidence, submit them to a new round of coordinating interviews, arranging the in-country triangulation with staff from the Fund and the workshop, overseeing the production of inputs, Bank, and draft a report. The case study teams and production of the final report. 99 ENDNOTES Chapter 1 risk to implementation, but without implying that it 1. See “The Poverty Reduction Strategy Initiative: put the credibility of the PRSP in question or sug- An Independent Evaluation of World Bank Support gesting remedial actions. Through 2003� (OED 2004). “Report on the Evalua- tion of Poverty Reduction Strategy Papers (PRSPs) Chapter 7 and the Poverty Reduction and Growth Facility 1. In discussions with the evaluation team, repre- (PRGF),� document number EB/EVC/04/3, June 1, sentatives of CSOs characterized the process as in- 2004. These documents are available in PDF format adequate, since it excluded the poor, and did not at the World Bank and IMF Web sites: allow sufficient time for preparation (for example, http://www.worldbank.org/oed/ prsp/ and http://www. materials were distributed late, and there was insuf- imf.org/External/NP/ieo/2004/ prspprgf/eng/index.htm ficient time for CSOs to consult members at grassroots 2. The full country case studies are available at levels). The quality of consultations was also ques- http://www.worldbank.org/oed/prsp tioned by respondents to the OED-IEO survey of 3. A larger group of more than 80 low-income PRSP stakeholders. The majority of respondents dis- countries is required to complete a PRSP to receive agreed that their group had been adequately con- continued concessional lending from the World Bank sulted during formulation of the PRSP. and IMF. 2. The process for preparing subsequent Annual 4. A coup attempt in Mauritania during early June Progress Reports has moved significantly in this di- 2003 and subsequent political uncertainties obliged rection, with the primary emphasis on a domestic re- the mission to limit its coverage relative to the other porting instrument that is also used by the Bank and case studies in this volume. the Fund. 5. Information on developments in aggregate Bank- Fund conditionality, discussed in IEO’s main evalua- Chapter 11 tion report, is too limited to permit conclusions on 1. See Toward Country-led Development—A Multi- trends in aggregate conditionality. Partner Evaluation of the Comprehensive Develop- ment Framework: Findings from Six Country Case Chapter 2 Studies. (OED: Washington, D.C., 2003). Available on- 1. See Transition—The First Ten Years: Analysis line at http://www.worldbank.org/oed/ and Lessons for Eastern Europe and the Former So- viet Union. World Bank, 2002, p. 61. Annex 1. See Lois-Ellin Datta, “Case Study Evaluations,� Chapter 5 GAO/PEMD-91-10.1.9, United States General Ac- 1. The most recent JSA by the World Bank and the counting Office, November 1990. Also Linda G. Morra IMF reached the standard conclusion that Guinea’s and Amy C. Friedlander, “Case Study Evaluations,� PRSP “presents a credible poverty reduction strategy Operations Evaluation Department, World Bank. and provides a sound basis for Fund and IDA con- 2. For example, the extensive country studies by cessional assistance.� The shortcomings discussed in the Multi-Donor Evaluation of the CDF covering coun- this section were unevenly addressed in the JSA. The tries with completed PRSPs (Uganda, Bolivia, Viet- JSA did convey that weak institutional capacity was a nam, and Burkina Faso) contain a significant amount 101 T H E P O V E R T Y R E D U C T I O N S T R AT E G Y I N I T I AT I V E of information on the PRSP process in these countries. See Toward Country-led Development—A Multi-Part- ner Evaluation of the Comprehensive Development Framework: Findings from Six Country Case Studies. (OED: Washington, D.C., 2003). 102 ENDNOTES 103 T H E P O V E R T Y R E D U C T I O N S T R AT E G Y I N I T I AT I V E 104 ENDNOTES 105