THE WORLD BANK W O R L D B A N K O P E R A T I O N S E V A L U A T I O N D E P A R T M E N T The Poverty Reduction Strategy Initiative An Independent Evaluation of the World Bank's Support Through 2003 OPERATIONS EVALUATION DEPARTMENT ENHANCING DEVELOPMENT EFFECTIVENESS THROUGH EXCELLENCE AND INDEPENDENCE IN EVALUATION The Operations Evaluation Department (OED) is an independent unit within the World Bank; it reports directly to the Bank's Board of Executive Directors. OED assesses what works, and what does not; how a borrower plans to run and maintain a project; and the lasting contribution of the Bank to a country's overall development. The goals of evaluation are to learn from experience, to provide an objective basis for assessing the results of the Bank's work, and to provide accountability in the achievement of its objectives. It also improves Bank work by identifying and disseminating the lessons learned from experience and by framing recommendations drawn from evaluation findings. W O R L D B A N K O P E R A T I O N S E V A L U A T I O N D E P A R T M E N T The Poverty Reduction Strategy Initiative An Independent Evaluation of the World Bank's Support Through 2003 2004 The World Bank http://www.worldbank.org/oed Washington, D.C. © 2004 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 Manufactured in the United States of America First edition September 2004 1 2 3 4 07 06 05 04 The findings, interpretations, and conclusions expressed here 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 cannot guarantee the accuracy of the data included in this work. The boundaries, colors, denominations, and other information shown on any map in the work do not imply on the part of the World Bank any judgment of the legal status of any territory or the endorsement or acceptance of such boundaries. Rights and Permissions The material in this publication 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 courtesy of the World Bank Photo Library. ISBN 0-8213-5972-X e-ISBN 0-8213-5973-8 Library of Congress Cataloging-in-Publication Data Battaile, William G., 1964- The poverty reduction strategy initiative : an independent evaluation of the World Bank's support through 2003 / William G. Battaile Jr. p. cm. -- (Operations evaluation studies) Includes bibliographical references. ISBN 0-8213-5972-X 1. Economic assistance--Developing countries--Evaluation. 2. Economic development projects--Developing countries--Evaluation. 3. Poverty--Developing countries. 4. World Bank. 5. International Monetary Fund. I. Title. II. World Bank operations evaluation study. HC60.B318 2004 338.91'09172'4--dc22 2004051224 World Bank InfoShop Operations Evaluation Department E-mail: pic@worldbank.org Knowledge Programs and Evaluation Capacity Telephone: 202-458-5454 Development (OEDKE) Facsimilie: 202-522-1500 E-mail: eline@worldbank.org Telephone: 202-458-4497 Facsimilie: 202-522-3125 Printed on Recycled Paper Contents v Acknowledgments vii Foreword xi Acronyms and Abbreviations xiii Summary of Findings and Recommendations 1 1 The PRSP Initiative: Objectives and Relevance 1 What Is the PRSP Initiative? 4 What Is This Review Evaluating? 5 An Assessment of Relevance 9 2 Implementation of Design Features 9 Application of the Underlying Principles 21 The Value Added of Process Requirements 27 3 Preliminary Results 27 Value Added in Planning and Links to Implementation 31 Capacity Enhancement 33 Preliminary Outcomes 37 4 Alignment of the Bank's Assistance 38 Alignment of CAS Formulation 38 Alignment of the Bank's Financial Assistance 42 Alignment of the Bank's Analytical Work 45 5 Findings and Recommendations 46 PRS Process 47 PRSP Content and Analytical Underpinnings 48 Partnership and Alignment 49 Looking Forward 51 Annexes 53 A: Summary of the BWI Policy Papers on the PRS Initiative 55 B: Guidelines for Country Case Studies 59 C: PRSP Status of 81 IDA-Eligible Countries as of May 2004 63 D: Survey of National Stakeholders in the 10 Case Study Countries 75 E: Review of Joint Staff Assessments of PRSPs i i i 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 : A N I N D E P E N D E N T E VA L U AT I O N 81 F: Management Response 91 G: Chairman's Summary: Committee on Development Effectiveness (CODE) 95 Endnotes 103 References Boxes 3 1.1 Stated Objectives of the PRS Initiative 4 1.2 Key Evaluative Material for This Review 6 1.3 Some Common External Criticisms of the PRSP Initiative 7 1.4 The PRSP Initiative Means Different Things to Different People 12 2.1 Parliaments and the PRSP 14 2.2 Participation Has Largely Not Been Sustained 18 2.3 Coordinated Donor Support for PRSP Formulation in Ethiopia 19 2.4 PRSP Helps Donors Coordinate Budget Support and Streamline Performance Monitoring 22 2.5 JSAs Play Different Roles for the Bank and the Fund 22 2.6 Key Findings of IEO Review of JSAs 23 2.7 Good Practice Treatment of Alternative Growth Strategies 24 2.8 New Government in Albania Uses APR to Adjust the PRSP 25 2.9 APR Requirements Unconnected to Domestic Processes 29 3.1 The PRSP Has Raised Awareness about Exploring Synergies across Sectors 30 3.2 Good Practice: Rolling PERs Analyze Options and Problems in PEM and Detailed Blueprints for Change 32 3.3 Weak Sub-National Capacity Makes Implementing PRSPs More Difficult Figures 2 1.1 Poverty Reduction and Growth below Expectations in the 1990s 5 1.2 A Country-Level Results Chain 12 2.1 Was the PRS Process Country-Driven? 15 2.2 Opinions Diverge on Participation 17 2.3 PRSP Process Scores High on Partnership Aspects, but Much Lower on Current Quality of Donor Coordination 23 2.4 JSAs Are Weak in Some Areas 28 3.1 Is the PRSP Process Focused on Outcomes That Benefit the Poor? 31 3.2 Post-PRSP Expenditures Continue Pre-PRSP Trend 35 3.3 Quality of Policies and Institutions in PRSP and Non-PRSP Countries 36 3.4 Average Net ODA Flows per Country 39 4.1 World Bank Lending Has Shifted to PRSP Countries 40 4.2 PRSCs Have Less Sectoral Concentration Than Previous Adjustment Lending to PRSP Countries 41 4.3 Sectoral Allocation in Bank Lending and PRSPs 42 4.4 Core ESW Increased in PRSP Countries, and They Receive More Core Diagnostic ESW Than Non-PRSP Countries 43 4.5 PRSP Alignment of the Bank's Analytical Work Was Not Rated as Highly as Alignment of Financial Assistance or Strategy Tables 10 2.1 Case Study Countries 15 2.2 Top 10 Priority Sectors in PRSPs by Expenditure 34 3.1 Progress in MDGs for 12 PRSP Countries with Annual Progress Reports i v Acknowledgments T his report was prepared by a team com- ments of more than 40 Bank staff who took part prising Bill Battaile (Task Manager), Zamir in focus group sessions for the review. These dis- Islamshah, and Shonar Lala. Mariola Chros- cussions were facilitated by Matt Minahan. towska and Carlos Reyes provided research as- Peer review was provided by the Swiss Agency sistance. John Eriksson, Basil Kavalsky, and David for Development Cooperation (coordinated by Bevan provided background papers for the eval- Beatrice Ferrari), Aart Kraay, and Jerry Wolgin. In uation and, along with Alison Evans, offered addition, helpful comments were received from feedback on earlier drafts. the following OED staff: Andy Anderson, Patrick OED is grateful to the governments and na- Grasso, Fareed Hassan, Kyle Peters, and Denise tional stakeholders of the 10 case study countries Vaillancourt. for sharing their knowledge and time. Jeffrey This evaluation was conducted in parallel with Balkind, Nils Boesen, Alison Evans, Delbert Fitch- the IMF's Independent Evaluation Office (IEO). ett, Madhur Gautam, and Basil Kavalsky led the OED appreciates the collegial and professional OED case studies. Additional members of the relationship shared with the IEO team led by case study and national stakeholder survey teams David Goldsbrough. Collaboration with IEO on were Desiree Elizondo, Ylli Cabiri, Abou Diallo, the case studies involved Martin Kaufman, Steve Paulo Mole, Rwekaza Mukandala, Mohammed Kayizzi-Mugerwa, and Tsidi Tsikata. Mussa, Goulsara Pulatova, Joseph Semboja, and OED is grateful to the Swiss Agency for De- Chan Sophal. Antje Vetterlein assisted in prepar- velopment and Cooperation for financial assis- ing desk reviews for the case studies. tance to support numerous activities, including OED thanks the Executive Secretary of the the multi-stakeholder workshops, a number of United Nations Economic Commission for Africa, the case studies, and the national stakeholder sur- K.Y. Amoako, for hosting the cross-country stake- vey. OED is also grateful for Danish Consultant holder workshop to discuss preliminary case Trust Fund support for the Mozambique and study findings (Addis Ababa, January 2004). Par- Nicaragua case studies. ticipants at both the case study and the entry Annisa Cline-Thomas, Yezena Yimer, and Julia workshop (Washington, December 2002) pro- Ooro provided administrative support to the vided valuable feedback on the case studies and team. Caroline McEuen edited the report. the approach paper for the evaluation. The Review was prepared under the direction OED appreciates the time and candid com- of Victoria Elliott, Manager of OEDCM. Director-General, Operations Evaluation: Gregory K. Ingram Director, Operations Evaluation: Ajay Chhibber Manager, Corporate Evaluation and Methods: Victoria Elliott Task Manager: Bill Battaile v Foreword T he Poverty Reduction Strategy (PRS) Ini- This evaluation builds on OED's earlier work tiative is a major effort by the World Bank to assess how well the PRS Initiative has helped and IMF to make aid more effective in re- thus far in putting the CDF principles into action, ducing poverty. Introduced in 1999, it has re- how relevant it is to countries' quest for sus- sulted in first-generation strategy documents and tainable poverty reduction, and how effectively important process changes in more than 35 low- the Bank has supported and aligned its own income countries. The Bank intends to extend the programs with the PRS approach. PRS Initiative to all low-income countries. OED conducted this evaluation in parallel The key principles underlying the PRS Initia- with an evaluation by the IMF's Independent tive--ownership, results focus, and country-led Evaluation Office (IEO) on Poverty Reduction partnership--were derived from the Compre- Strategy Papers and the Poverty Reduction and hensive Development Framework (CDF). OED's Growth Facility. OED and IEO jointly carried out multi-stakeholder evaluation of the CDF, Toward four case studies and pooled findings from their Country-Led Development, found that imple- separate case studies and analytical work. The mentation of the CDF principles requires diffi- IEO study was submitted to the Board of the IMF cult changes in the behaviors and practices of for discussion in July of 2004. both donors and recipients. Strong country lead- ership is needed to end fragmented donor pro- Main Findings grams and promote harmonization. The Bank and Fund launched the PRS Ini- The Initiative is relevant but its benefits vary tiative to underpin the Highly Indebted Poor The Initiative has improved the poverty focus of Countries (HIPC) Initiative, which was in- national strategies. The integration of sector tended to ensure that resources freed up by programs in an overall strategy has helped pol- debt relief would be directed toward poverty- icymakers to link sector strategies with poverty reducing public sector programs. OED's eval- reduction. Sector expenditure plans are being uation of the HIPC Initiative found that it challenged to meet poverty reduction criteria. provided highly indebted countries with a But progress depends on each country's public fresh start, but that countries still face chal- sector capacity, government-partner relation- lenges in finding sustainable paths for growth ships, and relations among donors. The Initiative and poverty reduction. has added the most value in countries where gov- v i i 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 : A N I N D E P E N D E N T E VA L U AT I O N ernment leadership and aid management countries. The information gathered is gener- processes were already strong. ally not linked to decisionmaking. Tensions in the Initiative's design have caused prob- The Bank has effectively supported PRSP formula- lems during implementation tion, but alignment has not entailed major changes in There is inherent tension in a Bank/IMF­driven Bank programs initiative involving conditionality that is also Bank staff supported national stakeholders meant to foster a country-driven process. This has preparing PRSP documents by providing ana- led to two problems during implementation. lytical work, commenting on drafts, and pro- First, there is no mechanism or guidance to moting participatory activities. Clients observed adapt the Initiative's processes and requirements that decentralization of staff and authority to to differing country conditions. Countries have country offices has enhanced the quality and focused more on completing documents, which usefulness of the Bank's support and its relations give them access to resources, than on improv- with partners. ing domestic processes. Second, there are no in- Since the launch of the Initiative, Bank lend- termediate indicators for what the principles ing has increased for countries with PRSPs com- are expected to achieve, which makes it difficult pleted through 2003. The content of Bank to assess progress. This lack of clarity has led to assistance strategies formulated subsequent to different and sometimes incompatible expecta- PRSPs overlaps with the content of PRSPs. But tions among stakeholders. PRSP programs are broad and not well prioritized, so this overlap has not entailed major changes Key areas need strengthening in Bank programs. · Analytical base. The PRS process has under- The Bank has conducted more economic and scored the need for more analytical work about sector work in PRSP countries than it did before how development policies and programs can the Initiative--largely in the form of diagnostic best lead to poverty reduction. First-round studies of key fiduciary functions. Critical gaps PRSPs could not have been expected to fill remain in linking policies and programs to these long-standing analytical gaps, but the poverty impact. PRS process could be used more effectively to The Bank management's process for pre- begin to address them. senting a PRSP to the Board undermines own- · Growth policies. Most PRSPs to date have not ership. Stakeholders perceive this practice as considered the full range of policy actions re- "Washington signing off" on a supposedly coun- quired for growth and poverty reduction. They try-owned strategy. The analytical quality and focus largely on public expenditures and pay coverage of Joint Staff Assessments (JSAs) are more attention to spending on social pro- mixed. JSAs are not widely shared with national grams than on infrastructure, rural develop- stakeholders. These shortcomings limit the use- ment, and other areas with poverty reduction fulness of JSAs as a basis for Bank support and potential. as a feedback tool for country clients. · Alignment. External partners have supported the PRS process during formulation, but nei- Recommendations ther donors nor the Bank have defined whether The low-income countries that have completed or how they should change the content of PRSPs are those with better policies and insti- their programs to reflect PRSPs. Donors have tutions. The PRS process has yet to be intro- changed their processes by raising their par- duced in the most difficult country environments. ticipation in budget support instruments and To increase the likelihood of gains in these more streamlining performance monitoring. challenging countries--and to deepen the PRS · Results focus. The development of country- process in existing PRSP countries--the Initiative specific indicators and monitoring systems to requires reorientation along the lines noted track them is at a preliminary stage in most below. v i i i F O R E W O R D PRS Process policies and programs. Analysis should address · The Bank should emphasize improving country areas such as sources of growth, the quality of processesforplanning,implementing,andmon- strategies in other than the social sectors, and itoring public actions geared toward poverty re- the integration of macro frameworks with duction.Itshouldencouragecountriestoestablish structural and social reforms. goals and benchmarks in these areas that are ap- propriatetocountrycircumstancesandwithincur- Partnership and Alignment rentcountrycapacitytomonitoranduse.Itshould · The Bank, in concert with other donors, placelessemphasisoncompletionofdocuments. should facilitate use of the PRSP as a partner- · The Bank should provide transparent, com- ship framework and make clear the expecta- prehensive, and candid feedback to countries tions about how external partners should on their PRS processes and develop a review pro- support their PRSPs. The Bank should assist cedure more supportive of ownership and more countries in setting priorities within their effectively linked to decisions about the Bank's PRSPs and demonstrate how the content of program. Bank programs has changed in light of im- proved PRSPs. The Bank should encourage PRSP Content government-led aid management and selec- · The Bank should help countries address key tivity and integrate its own assistance with the analytical gaps about the poverty impact of efforts of other partners. Gregory K. Ingram Director-General, Operations Evaluation i x Acronyms and Abbreviations APL Adaptable Program Loan APR Annual Progress Report BWI Bretton Woods Institutions (World Bank and IMF) CAS Country Assistance Strategy CDF Comprehensive Development Framework CFAA Country Financial Accountability Assessments CPAR Country Procurement Assessment Review CPIA Country Policy and Institutional Assessment DAC Development Assistance Committee (OECD) DFID Department for International Development (U.K.) ESAF Enhanced Structural Adjustment Facility ESW Economic and Sector Work (World Bank) EU European Union HIPC Heavily Indebted Poor Countries IDA International Development Association IDB Inter-American Development Bank IEO Independent Evaluation Office (IMF) IMF International Monetary Fund JSA Joint Staff Assessment M&E Monitoring and Evaluation MDGs Millennium Development Goals MTEF Medium-Term Expenditure Framework NEPAD New Partnership for Africa's Development NGO Nongovernmental Organization ODA Official Development Assistance ODI Overseas Development Institute OECD Organisation for Economic Co-operation and Development OED Operations Evaluation Department PA Poverty Assessment PAF Performance Assessment Framework x i A C R O N Y M S A N D A B B R E V I AT I O N S PEM Public Expenditure Management PER Public Expenditure Review PFP Policy Framework Paper PIU Project Implementation Unit PREM Poverty Reduction and Economic Management Network (World Bank) PRGF Poverty Reduction and Growth Facility (IMF) PRS Poverty Reduction Strategy PRSC Poverty Reduction Support Credit PRSP Poverty Reduction Strategy Paper PSIA Poverty and Social Impact Assessment QAG Quality Assurance Group SPA Strategic Partnership with Africa UNDP United Nations Development Program UNECA United Nations Economic Commission for Africa WBI World Bank Institute WTO World Trade Organization x i i Summary of Findings and Recommendations T he Bretton Woods Institutions (BWIs: the Initiative includes a uniform requirement--com- World Bank and the International Mone- pletion of an acceptable document--it has not tary Fund) launched the Poverty Reduc- been sufficiently tailored to match the full range tion Strategy (PRS) Initiative in 1999 to improve of capabilities and public policy priorities found the planning, implementation, and monitoring in low-income countries. Most PRSPs deal pri- of public actions geared toward reducing poverty. marily with the composition of public expendi- The Initiative centers on promoting a PRS process ture and give inadequate treatment to growth in each country that is expected to be country- and other poverty-reducing actions. External driven, results-focused, long-term, comprehen- partners have supported the PRS process during sive, and partnership-oriented. Low-income formulation, but have not yet systematically borrower countries are required to complete a adapted the content of their assistance programs PRS paper, or PRSP, for continued access to World in a coordinated manner around the content of Bank and Fund concessional lending and to PRSPs. Shortcomings in the strategies them- HIPC debt relief: almost half have already done selves are partly responsible for this. so. PRSPs are also meant to serve as the frame- The PRS Initiative is an improvement over work for assistance from external partners, in- the Policy Framework Papers of the 1990s, but cluding the World Bank. This evaluation reviews remains a work in progress. It has not yet fulfilled the Initiative's progress against these aims and its full potential to enhance the poverty reduc- assesses the effectiveness of the World Bank's tion efforts in low-income countries. The World support.1 Bank should continue to support it, while mak- The Initiative's contribution so far has varied ing changes in three main directions: widely across countries, depending mainly on the initial conditions in each country's public sector · Reduce or eliminate uniform requirements capacity, government-partner relationships, and and foster better customization. relations among donors. The Initiative has added · Encourage PRSPs to explore a wider range of the most value in countries where government policy options, including those aimed at en- leadership and aid management processes were hancing growth. already strong. It has had less effect in coun- · Help define clearer partnership frameworks tries with weak public sector capacity or with around PRSPs, with accountabilities for both donor-dominated aid relationships. Because the countries and partners. x i i i 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 : A N I N D E P E N D E N T E VA L U AT I O N Tensions in the Initiative's Design Have The PRS process is expected to follow five Caused Problems during Implementation principles--country ownership, results orienta- There is an inherent tension in the design of a tion, comprehensiveness, partnership, and a BWI-driven initiative involving conditionality that long-term perspective. Countries typically re- is simultaneously meant to foster a country- port on steps they have taken to put these prin- driven process. To promote ownership, the BWI ciples into practice. But what results should a policy papers underlying the Initiative were cir- country expect from their successful imple- cumspect in important areas. This has led to mentation? Moreover, during implementation two problems during implementation. First, the process specifies no intermediate indicators there is no mechanism or guidance to adapt the or milestones for what the principles are ex- Initiative's processes and requirements to dif- pected to achieve, which makes it difficult to fering country conditions, especially the weak assess progress toward the achievement of the public sector capacity found in many low-in- Initiative's objectives. This lack of clarity has led come countries. During implementation, coun- to different and sometimes incompatible ex- tries have understandably focused on completing pectations among stakeholders (see figure 1). For the documents that give them access to re- example, stakeholders agree that the extensive sources. This attention to requirements has often consultations conducted in most PRSP coun- been at the expense of adaptation of the PRS tries during strategy formulation have brought process to country circumstances. The BWIs' new actors into the development dialogue and initially ambitious timetable for completion of increased transparency. On this criterion, par- first round PRSPs in HIPC and Poverty Reduction ticipation has been a success. But civil society or- and Growth Facility (PRGF) countries reinforced ganizations consider that this openness has had the tendency to generate standardized strategies only limited impact on the design of PRSPs or on in the early PRSPs. domestic policy and that participatory activities F i g u r e 1 W a s t h e P R S P r o c e s s C o u n t r y - D r i v e n ? Government Donor Civil Society Int'l NGO 1 2 3 4 5 Completely Disagree Completely Agree Note: The thick lines indicate the mean score from the survey of national stakeholders in the 10 case study countries (from 1: completely disagree to 5: completely agree) on a compos- ite of four questions on whether the PRSP process was country-driven. The box around the bands shows a 95 percent confidence interval. The mean score for the government respon- dents was significantly higher, at a 95 percent confidence level, than that of donors. The means for both these groups were significantly higher than that for international nongovernmental organizations (NGOs) and civil society. The difference in scores was not significant between the latter two groups. x i v S U M M A R Y O F F I N D I N G S A N D R E C O M M E N D AT I O N S have sometimes waned once the PRSP was for- lead to poverty reduction, and the PRS process mulated. Along these dimensions, participation has underscored the need for more analytical has not achieved all its aims. work on these issues. The PRSP formulation process has not been much help in reducing Other Implementation Experience and these knowledge gaps, because it has rarely en- Areas for Strengthening tailed robust consideration of policy options, cost-effectiveness, and tradeoffs. Most PRSPs do Improved focus on poverty. Most stakeholders agree not integrate the macroeconomic framework that the Initiative has improved the poverty with sectoral strategies through a rigorous con- focus of dialogue during formulation of national sideration of the sources of growth, the social im- strategies. The need for PRSPs to be based on a pact of macro policies, or other macro-micro sound poverty diagnosis has highlighted the linkages. These first-round PRSPs could not have poor data on the main characteristics of poverty been expected to fill these longstanding analyt- in many countries and, in some cases, has already ical gaps. But the PRS process could have been resulted in marked improvements in data qual- used more effectively to identify the key analyt- ity. The PRS process has played an important ical gaps and to develop a research agenda to ad- role in emphasizing multidimensional income dress them. and non-income poverty indictors. The process of gathering sector programs under one strate- Inadequate attention to growth policies. Most PRSPs to gic umbrella has helped policymakers better ap- date have not considered the full range of policy preciate the links between sector strategies and actions required for growth and poverty reduction. poverty reduction, and sector expenditure plans They focus largely on public expenditures. And are being challenged to meet poverty reduction within the realm of public expenditures, they pay criteria. more attention to health, education, and other so- cial programs than to the poverty reduction po- Added attention to monitoring, but results focus re- tential of spending in other areas such as mains weak. The PRS process has begun to ori- infrastructure and rural development. ent stakeholders toward a results focus, although the development of country-specific indicators External partners' alignment with the PRS process. and monitoring systems to track them is still at Most partners have accepted that the PRSP has a preliminary stage in most countries. Available the potential to be an overarching framework for information is generally not linked to decision- aid management. Many have provided well-co- making. Some countries have adopted PRSP tar- ordinated assistance to governments to formu- gets that are unrealistic in light of their initial late PRSPs. PRSPs have generally provided a conditions, resources, and recent track records. constructive framework for transparent policy di- In many cases, monitoring systems are designed alogue among external partners and with gov- to meet donor data requirements--for exam- ernment. Given the PRSPs' broad scope and lack ple, through the progress reports, even when of prioritization, alignment would not neces- these exceed the country's needs or capacity. sarily entail changes in donors' programs. Nei- Since most low-income countries will probably ther donors nor the Bank have defined not be able to meet most of the Millennium De- specifically whether or how they should change velopment Goals (MDGs), PRSPs must be more the content of their programs to reflect PRSPs. effective in enabling countries to select and mon- Changes in donors' processes are more readily itor relevant indicators that can support do- evident. In the most positive cases, the PRS mestic decisionmaking on a sustained basis. process bolstered an already robust government role in aid coordination and increased partici- Weak analytical base for consideration of tradeoffs pation in, and greater coordination of, budget and linkages. It remains poorly understood how support instruments and attempts to stream- development policies and programs can best line performance monitoring. But overall, there x v 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 : A N I N D E P E N D E N T E VA L U AT I O N is still little evidence that donors have coordi- Credit (PRSC), to support PRSPs. PRSCs' longer nated and selected the majority of their pro- time frame and programmatic flexibility have the grams in response to PRSPs. Where the potential to match with medium-term PRSP government-donor dialogue was previously weak programs more effectively than does conven- or donors continued to drive the agenda, PRSPs tional adjustment lending. It is too early to tell by themselves have not noticeably promoted whether these design improvements will trans- donor coordination or increased government late into better outcomes. PRSCs in the case management of external assistance. study countries have helped to promote own- ership and to improve donor coordination. Effectiveness of World Bank Support Budget support is viewed as a vote of confi- As a proponent of the Initiative, the World Bank dence in the government's program and has accepted the responsibility to support the ap- processes, and the World Bank is seen as "join- plication of the PRSP's underlying principles and ing the team" of donors providing budget sup- to provide its assistance in support of govern- port. Bank task managers report that preparing ments' poverty reduction strategies. Thus, the and implementing PRSCs has led to more cross- PRSP entails a new way of doing business in low- sectoral interaction by staff, but some sector staff income countries for the World Bank. express worry that channeling more lending through the PRSC instrument could lower the Staff support during PRSP formulation. World Bank quality of technical inputs. staff have supported national stakeholders preparing PRSP documents by providing com- Economic and sector work. The World Bank's eco- ments at various junctures, while generally not nomic and sector work has added value to coun- imposing a particular World Bank viewpoint. try planning. Poverty Assessments and Public The Bank actively promoted participatory ac- Expenditure Reviews have been found most use- tivities by funding advisers and facilitating ful during PRSP formulation. The World Bank has arrangements for dialogue. Clients observed conducted fewer Poverty and Social Impact As- that decentralization of staff and authority to sessments than envisioned, even though it em- country offices has enhanced the quality and phasizes the importance of analysis of the usefulness of the Bank's support. impacts of policy reform on the well-being of the poor. In PRSP countries, the Bank has conducted Alignment of World Bank support. The content of more economic and sector work than it did be- Bank assistance strategies formulated subse- fore the Initiative. This additional work largely quent to PRSPs overlaps with the content of takes the form of fiduciary core diagnostics. Crit- PRSPs. But since PRSP programs are broad and ical gaps remain in linking policies and programs not well prioritized, this overlap has not entailed to poverty impact. major changes in Bank programs. The Bank should assist countries in improving the priori- Partnerships. The World Bank has promoted the tization of their PRSPs, and subsequently demon- PRSP as a coordinating framework for donor ac- strate how the content of its programs has tivities. The Bank has improved its own coordi- changed in light of these improved PRSPs. nation and relations with other external partners, Lending. Bank lending has increased for especially through its field-based staff. This im- countries with PRSPs, although these are also provement is not necessarily attributable to the the countries with higher-quality policies and PRS Initiative alone, because the Bank has been institutions, and they would ordinarily receive emphasizing decentralization and aid coordina- more resources under the International De- tion for several years. The Bank could still do velopment Association (IDA) policy of per- more to apply the partnership principle by en- formance-based allocations. The World Bank couraging government-led aid management and has introduced a new programmatic lending selectivity and integration of its own assistance instrument, the Poverty Reduction Support with the efforts of other partners. x v i S U M M A R Y O F F I N D I N G S A N D R E C O M M E N D AT I O N S Bank-Fund collaboration has improved, with tutions. The PRS process has not yet been in- more frequent discussions between staff of the two troduced in the most difficult country environ- institutions at the country level. While working ments, including those where ownership may be smoothly together in PRSP formulation and review most lacking. To increase the likelihood of achiev- in the case study countries, the Bank and the Fund ing gains in these more challenging countries-- had policy differences on the scope and sequenc- and to deepen the PRS process in existing PRSP ing of structural reforms in some countries that countries--the Initiative requires reorientation the PRSP framework alone could not overcome. along the lines noted below. In addition, the World Bank should deploy its economic and sec- Board review of the PRSP and the role of the Joint Staff tor work, financial support, and leadership Assessment (JSA). The Bank management's among donors more effectively to enhance the process for presenting a PRSP to the Board un- implementation of the Initiative and to keep the dermines country ownership. Stakeholders per- focus on process and results. ceive this practice as "Washington signing off" on a supposedly country-owned strategy. Executive PRS Process directors already have several mechanisms for au- thorizing Bank support to a country--notably · The World Bank's country activities under the their endorsement of Country Assistance Strate- Initiative should help foster greater emphasis gies (CASs) and their approval of individual lend- on improving country processes for planning, ing operations. In this context, their review of the implementing, and monitoring public actions PRSP appears redundant, as well as attenuating geared toward poverty reduction, and less on ownership in the eyes of most stakeholders. The completion of documents. Tailoring of the Ini- Review recommends that management develop tiative to country conditions should be actively a review procedure that is more transparently promoted, including establishment of goals supportive of ownership and more effectively and benchmarks that are appropriate to coun- linked to decisions about the Bank's program. try circumstances and that respond to do- The JSA was designed to provide the Bank and mestic priorities. Fund Boards with an assessment of the sound- · The World Bank should provide transparent and ness of the PRSP as a basis for support. It was also effective feedback to countries on their PRS designed to provide feedback on how a country's processes and develop a review procedure that PRSP could be improved over time. The JSAs is more supportive of ownership and more ef- have shortcomings that undermine achievement fectively linked to decisions about the Bank's pro- of these goals: mixed analytical quality and com- gram. It should strengthen the operational link prehensiveness, inadequate focus on the qual- between PRSPs and CASs and demonstrate more ity of process orientation, and limited awareness clearly how each CAS is anchored in the PRSP. of their findings and recommendations among stakeholders. Instead of being a comprehensive PRSP Content and Analytical review that underpins a discussion of selectivity Underpinnings and comparative advantage, the JSA is influenced by anticipated lending activities. For these rea- · The World Bank, in concert with other partners, sons, the quality of feedback on PRSPs should be should help countries build the capacity to more candid, transparent, analytically rigorous, address key analytical gaps about the poverty and comprehensive, and it should strengthen impact of policies and programs. Analysis at the partnership through greater involvement of country level should address areas such as other stakeholders. sources of growth, the quality of sector strate- gies, and the integration of the macro frame- Going Forward work with structural and social reforms. The low-income countries that have completed · Monitoring inputs, outputs, and outcomes of PRSPs are those with better policies and insti- ongoing activities is critical to a better under- x v i i 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 : A N I N D E P E N D E N T E VA L U AT I O N standing of what works and what does not. the use of the PRSP as a partnership frame- The Bank should assist in defining indicators work. Countries should be encouraged to that are specific to country priorities, realistic, make clear their expectations about how ex- and within current country capacity to moni- ternal partners--including the World Bank-- tor and use. The scope of monitoring and eval- should support their PRSPs. The Bank should uation can grow with capacity over time. lead by example in using the PRSP as the foundation for its assistance, taking into ac- Partnership and Alignment count other partners' activities and the Bank's comparative advantage, and in promoting · The World Bank, together with other donors, country leadership in managing external as- should assist PRSP countries in strengthening sistance. x v i i i 1 The PRSP Initiative: Objectives and Relevance T he Poverty Reduction Strategy Papers (PRSP) Initiative, introduced in 1999, has been under implementation for more than four years. To pro- vide the basis for assessing implementation progress over this period, this chapter first reviews the objectives and key design features of the Initia- tive and presents the implications of the Initiative for the Bank's work in low- income countries. The scope of the review and evaluative material are outlined, and the chapter concludes with an assessment of the Initiative's relevance and design, highlighting several inherent implementation challenges. What Is the PRSP Initiative? (MDGs) were proposed as time-bound and The PRSP Initiative--launched by the World outcome-based targets for poverty reduction. Bank and the International Monetary Fund, But how could better poverty outcomes be together known as the Bretton Woods Institu- achieved? The international community reached tions, or the BWIs--is intended to improve a consensus during the 1990s on a new develop- development outcomes. Stakeholder groups ment paradigm for enhanced aid effectiveness were disappointed by the poverty reduction and better development results. The key ingredi- impact of past development assistance, partic- ents for growth and poverty reduction were ularly the poor results in low-income countries. drawn from several sources, including examples In Sub-Saharan Africa, the $1 per day poverty seen as promising country experiences (for headcount index rose from 47 to 49 percent example, Bolivia, Mozambique, and Uganda). In from 1990 to 2000, as shown in figure 1.1. these cases, local ownership, improved donor Experience suggests that rapid and sustained coordination through country leadership, and growth is required for poverty reduction.1 The empowerment of the In Sub-Saharan Africa, the international community recognized the need poor were central. to reduce poverty more effectively in low- Messages that reinforced $1 per day poverty income countries, amid calls for greater these themes came from headcount index rose accountability from external partners. The work by the World Bank from 47 to 49 percent International Development Goals2 and the (for example, Assessing subsequent Millennium Development Goals Aid, Aid & Reform in from 1990 to 2000. 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 : A N I N D E P E N D E N T E VA L U AT I O N P o v e r t y R e d u c t i o n a n d G r o w t h b e l o w F i g u r e 1 . 1 E x p e c t a t i o n s i n t h e 1 9 9 0 s Poverty (percentage of population) Growth (percentage per annum) 60 7 50 6 5 40 4 30 3 20 2 10 1 0 0 ­1 ­10 ­2 ­20 East Asia South Asia Latin Middle East Sub- Europe and ­3 and Pacific America and Saharan Central and the North Africa Africa Asia Caribbean Population living on less than $1 a day (1990) Population living on less than $1 a day (2000) Real per capita GDP growth (1990­2000) Source: World Bank and IMF 2004; World Bank 2003a. Africa), international nongovernmental organi- to complete PRSPs to receive continued conces- zations (NGOs) (such as Oxfam 1995), academ- sional lending from the World Bank and IMF. ics (Sen 1999, for example), and the external These decisions made the PRSP a centerpiece review of the IMF's Enhanced Structural Adjust- for BWI relations with low-income countries. ment Facility (Botchwey and others 1998). These ingredients were packaged in the form of the Objectives. To achieve the overarching goal of Comprehensive Development Framework poverty reduction, the key objective of the (CDF), introduced by the World Bank.3 But while Initiative enunciated by the BWIs is to assist the CDF encapsulated "a way of doing develop- low-income countries in developing and ment," it lacked an operational link to decision- implementing more effective strategies to fight making in the World Bank and IMF. poverty (that is, strengthening the poverty The catalyst for this operational link came in impact of public actions, including policies and 1999. Following a successful campaign by a spending) through supporting and sustaining a network of NGOs and their supporters to country-driven Poverty Reduction Strategy strengthen the poverty impact of debt relief, the (PRS) process in low-income countries.4 BWIs modified the Heavily Indebted Poor In addition to this key objective, the policy Countries (HIPC) Initiative to require the 38 documents detail several subsidiary objectives qualifying HIPC countries to prepare and (box 1.1) that add to country-specific actions and implement PRSPs introduce goals for the Bank and other develop- The key objective is to assist as a condition for ment partners related to the CDF, the HIPC low-income countries in debt relief. They Initiative, aid effectiveness, and the effectiveness developing and implementing also required a of World Bank assistance. PRSPs are meant to larger group of operationalize the CDF, emphasizing the CDF's more effective strategies to more than 80 low- underlying principles in the various phases of the fight poverty. income countries PRS process.5 PRSPs should serve as the 2 T H E P R S P I N I T I AT I V E : O B J E C T I V E S A N D R E L E VA N C E framework for all external assistance geared reflected in an Annual Preparation of PRSPs and toward poverty reduction, and thus enhance Progress Report (APR). their endorsement by the public use of aid in general, and the use of debt Recognizing the consid- Boards of the BWIs are relief and World Bank assistance in particular. erable time required to needed for countries to The Bank and the Fund are to increase their formulate a full PRSP and coordination around the PRSP framework--for the need for HIPCs to gain access to resources. example, through joint missions and the drafting receive debt relief in the of Joint Staff Assessments. short term, debt relief is permitted to begin when a country completes an Interim PRSP.6 To obtain Key Features. There are two key features of the concessional resources from the BWIs, low- PRSP Initiative: five underlying principles and income countries are required to complete several process requirements that link PRSPs to PRSPs, or signal progress toward a full PRSP concessional BWI lending and HIPC debt relief. through an Interim PRSP or a PRSP Preparation The five underlying principles were drawn Status Report. The Boards of the BWIs must from the CDF and are set out in general terms endorse each of the required documents-- for countries to follow. They indicate the PRS Interim PRSPs, PRSPs, Annual Progress Reports, process should be: and PRSP Preparation Status Reports--on the basis of a Joint Staff Assessment (JSA) prepared by · Country-driven, involving broad-based par- Bank and Fund staff. ticipation · Comprehensive in recognizing the multi-di- Phases of the PRS Process. BWI policy papers, mensional nature of poverty and later guidance provided by the JSA · Results-oriented and focused on outcomes Guidelines, focus on how to prepare the PRSP that benefit the poor document, couched in a dynamic process of · Partnership-oriented, involving coordinated planning, implementation, monitoring, feed- participation of development partners back, and strategy adjustment (the PRS · Based on a long-term perspective for poverty process). The Guidelines explain that planning reduction. requires a comprehensive understanding of poverty and its determinants, choosing poverty The process requirements call for prepara- reduction objectives/indicators/targets, and tion of PRSPs and their endorsement by the formulating programs and policies for growth Boards of the BWIs for countries to gain access to and poverty reduction. This phase is expected resources. To obtain irrevocable HIPC debt relief, to culminate in completion of a PRSP countries are required to produce an initial PRSP document. Progress toward intermediate and and implement it successfully for a year, as longer-term outcomes is to be monitored B o x 1 . 1 S t a t e d O b j e c t i v e s o f t h e P R S I n i t i a t i v e Over-arching goal: Sustainable poverty reduction. Subsidiary objectives: · Operationalize the CDF. Key objective: Assist low-income countries in developing · Strengthen the link between debt relief and poverty reduction. and implementing more effective strategies to fight poverty. · Make the support of all development partners more effective. · EnhancethepovertyimpactoftheBank'sworkinIDAcountries. · Strengthen Bank/Fund collaboration. Source: World Bank and IMF policy papers (see Annex A). 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 : A N I N D E P E N D E N T E VA L U AT I O N during implementation and fed back to inform ties and coordination with other external partners adjustments, sustaining the long-term process. (discussed in Chapter 4). Participatory activities involving government, civil society, and external partners are What Is This Review Evaluating? envisioned throughout the PRS process. Scope. The objective of this review is to assess Implications for the Bank's Work. For the Bank, the the progress of the PRS process toward launch of the PRSP Initiative implied a new way of meeting the challenge of poverty reduction doing business in low-income countries. As one of and to assess the World Bank's role in support the proponents of the Initiative, the Bank of the process, with a view to informing and For the Bank, the launch accepted the responsibil- making recommendations to strengthen the ity to support the applica- implementation of the Initiative and to increase of the PRSP Initiative tion of the PRSP's the effectiveness of World Bank support. implied a new way of underlyingprinciplesand The review pursued three main evaluative doing business in low- to align its assistance with dimensions, as laid out in the Approach Paper:9 governments' poverty income countries. reduction strategies. · RELEVANCE OF THE OVERALL APPROACH: Are the Internal guidance outlined the significant changes design features of the PRSP framework aligned that would be required in the Bank's work to with the key impediments countries face in achieve the objectives of the Initiative. Policy 7 developing and implementing effective poverty Framework Papers (PFPs), which had previously reduction strategies? been the basis for BWI support, were replaced · APPLICATION OF THE UNDERLYING PRINCIPLES with country-owned PRSPs.8 The Bank's AND PRELIMINARY RESULTS: Have the underlying economic and sector work, financial assistance, principles been applied to the planning and assistance strategies, and interactions with other implementation processes of the PRSP? Have stakeholders were expected to help countries the process requirements met their objectives? strengthen PRS processes (discussed in Chapter What are the preliminary results of the process? 2) and to support implementation of the PRSP · EFFECTIVENESS OF WORLD BANK SUPPORT AND through alignment of International Development ALIGNMENT: How effectively has the World Bank Association (IDA) assistance with country priori- promoted the PRSP principles, and to what B o x 1 . 2 K e y E v a l u a t i v e M a t e r i a l f o r T h i s R e v i e w · Entry workshop in Washington, D.C., to receive input on · Four focus group sessions of a cross-sectoral mix of Bank scope and methodology of evaluation. task managers working in PRSP countries. · Ten country case studies, providing detailed evidence on the · In-country stakeholder surveys conducted in the 10 case application of PRSP processes and on the effectiveness of study countries. Bank assistance. Exit workshops were conducted at the · Existing evaluative material from both inside and outside the conclusion of each field visit to discuss preliminary findings. BWIs. · Two thematic studies on capacity enhancement and public · Cross-country workshop of national stakeholders in Addis expenditure management. Ababa, Ethiopia, to compare findings across the case studies. · Cross-country analyses, including reviews of PRSPs, JSAs, APRs, CASs, public expenditures, aid flows, and Bank lend- ing and non-lending activities. 4 T H E P R S P I N I T I AT I V E : O B J E C T I V E S A N D R E L E VA N C E extent is its assistance aligned to country pri- World Bank, discussions with staff, and analysis orities? of data from inside and outside the institution. Some common criticisms of the Initiative from OED constructed a country-level results outside the Bank are summarized in box 1.3. chain, shown in figure 1.2. This review assesses the inputs and outputs of the PRS process with An Assessment of Relevance respect to both the progress made and the The majority of stakeholders view the Initiative as effectiveness of the World Bank's support. In relevant, as evidenced by the national stakeholder addition, the evaluation assesses the interme- survey.12 Roughly 60 percent of stakeholders diate outcomes of the process, to the extent surveyed noted that they agree or completely possible. Since the process has been under way agree that the PRSP approach is a good model for for less than a full PRSP cycle, it is too early to addressing poverty reduction and that it improves evaluate the extent of its achievement of the on past modalities for donor assistance.13 The case desired impact--poverty reduction. Even after studies reveal that the underlying principles were a full PRSP cycle, it will be difficult in many seen as relevant in all 10 countries because they countries to evaluate the degree to which either addressed country-specific areas of poverty reduction has been achieved. weakness or reinforced areas of strength. The "fit" This review used a variety of evaluative between the PRSP approach and country circum- techniques as the basis for its findings and stances was most frequently noted with regard to recommendations. A list of key evaluative integrating existing strategies, emphasizing a material is provided in box 1.2. The 10 country sorely needed results focus, and providing an case studies provided evidence of how the organizing framework to guide (and manage) process has unfolded.10 The cross-country relations with external partners. In practice, analyses drew from a broader sample of PRSP- however, design of the Initiative presents several related documents and data.11 The review also tensions between application of the principles considered evaluation results from outside the and the process requirements. F i g u r e 1 . 2 A C o u n t r y - L e v e l R e s u l t s C h a i n Inputs Outputs Outcomes Outcomes Intermediate Longer term Country-driven · Country -owned long- · Policy adjustment · More effective resource preparation of PRSP, term, comprehensive, · Public spending adjusted use including: and monitorable to priorities · Higher aid quality · Broad-based strategy · Country-led alignment · Achievement of longer- participation · M & E arrangements of external support term PRSP targets Impact · Diagnosis of · Strategy integrated · Achievement of inter- · Evaluation and strategy Poverty poverty with budget processes mediate PRSP targets adjustment Reduction · Long-term · Framework for · Implementation · Sustainable growth planning alignment of external monitoring and greater · Consideration of support accountability tradeoffs · Realistic costing · Monitorable indicators · external support Evaluation scope 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 : A N I N D E P E N D E N T E VA L U AT I O N S o m e C o m m o n E x t e r n a l C r i t i c i s m s o f t h e B o x 1 . 3 P R S P I n i t i a t i v e Members of the development community, including bilateral · The current approach complements but does not replace donors, NGOs, and researchers, have reviewed country expe- past structural adjustment policies with short-term anti- riences with the PRS process. There is consensus that PRSPs poverty actions and longer-term distributional considera- represent an important innovation in development aid. Many have tions. observed that the process has the potential to transform poli- · Capacity constraints and limited government interest in en- cymaking and partner dialogue in a positive way, and that new gaging civil society have hampered the quality of participa- spaces for domestic policy dialogue have been created. tory processes, which, for the most part, have not been These reviews have also highlighted the following criti- institutionalized. cisms of the Initiative: · Macroeconomic and growth policies have tended to be neg- lected in PRSPs in favor of attention to the social sectors. · Political dimensions of ownership and sustaining public · Linkages between existing planning instruments of partner and parliamentary understanding are as critical to success governments and donors need to be strengthened. as technical commitment to PRS processes. · International support, through increased aid and greater · Civil society organizations have not been adequately in- market access, is as important as a greater domestic policy volved in consultations on the macroeconomic or structural effort. policies in PRSPs. Source: ActionAid (2004); CIDSE/Caritas Internationalis (2003); Danish Ministry of Foreign Affairs (2003); Maxwell (2003); ODI (2003); Sanchez and Cash (2003); SDC (2003); UNDP Evaluation Office (2003a); UNCTAD (2002); Wilks and Lefrançois (2002); World Vision (2002); Oxfam (2001, 2004); McGee and Norton (2000); Whaites (2000). Issues with the design of the Initiative. There is inherent vides inadequate guidance for adapting its tensioninaBWI-driveninitiativeinvolvingcondition- requirements to country conditions. All ality that is simultaneously meant to foster a country- countries, even those with very weak public driven PRS process. The BWIs expect each country's sector capacity, must submit documents accept- authorities to establish benchmarks and targets for able to the BWI staff and Boards to qualify for results of the PRSP Initiative, thereby reinforcing the HIPC debt relief and IDA/PRGF resources. principle of ownership. At the same time, the BWIs Because countries have a strong incentive to expect countries--if they want access to conces- quickly obtain these resources, many have relied sional resources--to complete PRSPs and their on external consultants during PRSP formula- related progress reports according to BWI-set tion--a reliance that in some cases inhibited guidelines and timetables. This expectation is capacity building. The ambitious timetable inconsistent with ownership. It is no surprise that initially set out by the World Bank and the Fund this tension has led to implementation problems as for completion of first-round PRSPs in HIPC and countries have tried to adapt the key features of the PRGF countries discouraged experimentation in Initiative (the underlying principles and process adapting the PRS process to country circum- There is inherent tension requirements) to their stances. In addition, the Initiative's underlying specific initial conditions principles have been applied somewhat in a BWI-driven initiative and manage the domestic uniformly to widely varying situations. Countries involving conditionality and external incentives have different needs and priorities in applying that is simultaneously they face. principles such as participation and partnership, Although the Initia- depending on their starting points. Yet the Initia- meant to foster a country- tive was designed to be tive does not adequately guide authorities to driven PRS process. country-owned, it pro- define the results and milestones they hope to 6 T H E P R S P I N I T I AT I V E : O B J E C T I V E S A N D R E L E VA N C E achieve from implementing these principles in The rest of the Authorities in some their countries. report is structured as countries have Tensions in design have also caused stakehold- follows. Chapter 2 asks incorporated features in ers to view the Initiative's purpose differently, as how the design features the strategy document summarized in box 1.4. In most case studies, outlined in this chapter, country authorities see the Initiative predomi- specifically the underly- that they thought the BWIs nantly as a condition for continued BWI support. ing principles and would prefer. Authorities in some countries have incorporated process requirements features in the strategy document that they summarized above, have been implemented thought the BWIs would prefer, especially an and assesses the Bank's support of these emphasis on expenditures in the social sectors.14 activities. Chapter 3 assesses the preliminary Limited dissemination of information about PRS results of the process. Chapter 4 reviews how architecture to the broad spectrum of interested well the Bank's assistance has been aligned to stakeholders has contributed to the gap in the PRSP. And Chapter 5 summarizes the main expectations. In the case studies, few beyond the findings of the evaluation and offers core group of government representatives that recommendations detailing how the Bank can worked closely with the World Bank were familiar improve its support and strengthen implemen- with these policy documents. tation of the Initiative. T h e P R S P I n i t i a t i v e M e a n s D i f f e r e n t B o x 1 . 4 T h i n g s t o D i f f e r e n t P e o p l e In most case study countries: · For many donors, the Initiative is an opportunity to enhance ac- countability and transparency in the use of external assistance. · Governments see the PRSP as added conditionality to give them access to resources through HIPC debt relief, or to These differences in expectations can interfere with col- continued concessional assistance from the BWIs. laboration across stakeholder groups toward the PRSP Initia- · NGOs in many countries see the PRSP as a means for gain- tive's goal of poverty reduction. Greater clarity and transparency ing greater political participation and voice. in the Initiative's focus and objectives are needed at the coun- try level to minimize this challenge. 7 2 Implementation of Design Features T his chapter assesses how the two key features of the PRSP Initiative-- its underlying principles and process requirements--have been im- plemented. The first section addresses the extent to which the five underlying principles of the PRSP approach have been applied, especially in PRSP formulation. The second section addresses the value added of three process requirements: Board endorsement of PRSPs, the Joint Staff As- sessments (JSAs), and the Annual Progress Reports (APRs). The analysis is based primarily on the in- stakeholders in each of the 10 countries, depth assessment of progress on the ground in targeted at individuals familiar with the PRS the 10 OED/IEO case study countries--a process in government, civil society, interna- representative sample covering roughly half tional NGOs, and other external partners; (2) a the PRSPs that had been discussed by the Board content review of PRSPs; (3) reviews conducted at the time of the evaluation's design (table separately by OED and IEO of JSAs of PRSPs; 2.1).1 The case study countries were chosen to and (4) a review of APRs. cover a variety of country situations and stages of PRSP implementation.2 To assess changes in Application of the Underlying Principles the process of national stakeholder interaction in working toward a country's development Country Driven goals, interviews were conducted with govern- Country ownership is a key tenet of the PRS ment officials, civil society, international NGOs, process because effective development requires and other external partners in the case study policies and priorities to originate from, and be countries. The interviews were supplemented driven by, national stakeholders. Three criteria by in-country stakeholder workshops. Finally, are used to assess the extent to which the PRS OED and IEO held a multi-country workshop process has been country-driven: 4 to discuss findings across all the case studies.3 Evidence for this chapter is also drawn from · Locus of the initiative: Country ownership is a other sources, which are mentioned explicitly What provided the im- key tenet of the PRS when used as a basis for findings in the text. petus for countries to These include: (1) a survey of national adopt a PRS process? process. 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 : A N I N D E P E N D E N T E VA L U AT I O N T a b l e 2 . 1 C a s e S t u d y C o u n t r i e s First Second Annual Annual Years of Progress Progress Region Country PRSP Date Implementation Reporta Reporta HIPC OED/IEO AFR Ethiopia 17-Sep-2002 1.7 17-Feb-2004 Yes OED Guinea 25-Jul-2002 1.9 Yes IEO Mauritania 6-Feb-2001 3.3 18-Jun-2002 18-Jul-2003 Yes OED Mozambique 25-Sep-2001 2.7 25-Jul-2003 Yes OED/IEO Tanzania 30-Nov-2000 3.3 27-Nov-2001 27-May-2003 Yes OED/IEO EAP Cambodia 20-Feb-2003 1.3 No OED Vietnam 2-Jul-2002 1.9 19-Feb-2004 Nob IEO ECA Albania 20-Jun-2002 1.9 10-Jul-2003 No OED/IEO Tajikistan 5-Dec-2002 1.5 No OED/IEO LCR Nicaragua 25-Sep-2001 2.7 11-Dec-2002 22-Jan-2004 Yes OED/IEO Note: Dates represent dates of discussion at Bank Board. a. As of May 2004. b. While eligible for debt relief under the Enhanced HIPC initiative, Vietnam is potentially sustainable without HIPC assistance. · Country commitment: How committed are HIPC Initiative, they soon embraced the stakeholders to the process, and to what extent approach, linking it to existing poverty has the PRSP been mainstreamed? reduction policies and ongoing analytical work · Participation: Has there been broad, sus- on poverty.5 In other cases, PRSP-related activi- tained, and purposive participation in the ties were cut back once BWI requirements for process? HIPC or a Fund Poverty Reduction and Growth Facility (PRGF) were met. In Guinea, the Locus of the initiative. One of the objectives of the government did not convene working groups PRSP Initiative was to serve as a framework for to continue relevant discussions after the PRSP access to concessional resources and debt was completed. In Tajikistan, consultations relief under HIPC. Borrowers in case study with civil society declined dramatically after countries undertook PRSPs largely based on government approval of the PRSP. these external conditionalities. The BWI origins The BWIs initially laid out an aggressive of the Initiative present an inherent tension in timetable for PRSP production, focused on fostering a country-driven process. Positive HIPC and PRGF countries. This pressure to examples in the case studies, however, show complete a PRSP under short deadlines in that this tension can be managed. This has order to access resources contributed to an been the case in Mozambique and Tanzania, excessive government focus on preparation of where the governments effectively broadened a document, rather than on a long-term their focus from document preparation and process to enhance a country's ability to plan, endorsement by the BWIs as a condition for implement, and monitor progress toward long- HIPC to implementation of a PRS process. term goals.6 Stakeholders in 4 of 10 countries Although the Mozambican authorities initially reported that they experienced pressure to viewed the introduction of the PRSP as complete PRSPs in time to meet BWI deadlines, additional conditionality being imposed on the including the Fund's PRGF cycle (e.g., country by the IMF and World Bank under the Cambodia, Mozambique, Nicaragua, Tanzania). 1 0 I M P L E M E N TAT I O N O F D E S I G N F E AT U R E S In Tanzania, the formulation of the PRSP took emphasize social sector During the preparation place in an extremely compressed timetable, expenditures in order to processes, Bank staff were under pressure to reach HIPC decision point, receive BWI endorse- supportive and keenly and the first Progress Report was rushed to ment. In Mozambique reach HIPC completion point.7 The PRSP was interested, but largely and Nicaragua, civil completed and submitted to the Cabinet just 6 society informants gave room for country months after the I-PRSP (Interim PRSP), expressed the view that ownership. compared with a current average time between government already I-PRSP and PRSP of 20 months.8 In Cambodia, knew what the BWIs the pressure to hastily produce an I-PRSP wanted--and knew they had to pass the BWI originated from the decision to convert the review process, including the JSA. The BWIs did Fund's Enhanced Structural Adjustment Facility not need to impose their agenda--holding the (ESAF) arrangement into a PRGF arrangement string to the HIPC purse was sufficient. Stakehold- which, under IMF operational guidelines, ers see these tensions with ownership as required links with a PRS process.9 This initial embodied in the submission of the PRSP pressure was in tension with the ongoing document to the Boards of the Bank and Fund, a preparation of the country's five-year socio- process perceived as "Washington signing off." economic development plan. Even though the BWIs influenced the timing of Country Commitment. Government commitment the process, national stakeholders in case study and country commitment are not the same. countries have confirmed that during the prepara- Country commitment for the PRSP goes beyond tion processes, interactions with Bank staff the decision of government to prepare a strategy generally did not compromise ownership. Bank to include the commitment of a wider range of staff were supportive and keenly interested, but national stakeholders. All case study govern- largely gave room for country ownership and did ments declared political commitment to the not impose a Bank viewpoint. This observation PRSP, and in many cases commitment comes was most prevalent in countries where strong from the highest levels of government, as in country-driven processes already existed, such as Tajikistan, where the PRSP was introduced by Tanzania, Ethiopia, and Mozambique, where the Presidential Decree. Such high-level "PRSP Bank was perceived to have changed its business Champions" play an important role in initiating model and culture to foster ownership. Clients and motivating the process. But interviews observed that decentralization of Bank staff and across a broad range of stakeholders in the case authority to country offices was especially helpful studies suggest that country commitment to the because staff in the country office are more PRS process varies considerably. Within govern- effective than sector missions from Washington at ment, commitment to the strategy is strongest considering alternative viewpoints and ceding among those closest to its preparation (typically control to the ministry. For example, the designa- the finance or planning ministries) and those tion of a PRSP point person in the Albania office who perceive the greatest benefit from the enhanced the quality of the Bank's role by provid- process--for example, through increased ing a real time and continuous interface through- resource allocations. Commitment fades in out the process. But in several case study sector ministries, especially ministries with low countries, the content of PRSP strategies was PRSP priority, and regional/local administrations. influenced by the BWI origins of the Initiative. The Civil society commitment depends on the extent Bank was closely involved both in the preparation to which the PRSP accommodates the particular of previous strategies and in analytical work such position of an organization. The involvement of as Poverty Assessments, Country Economic parliaments has been a particularly weak aspect Memorandums, and Public Expenditure Reviews of the process in the case study countries (see that fed into PRSPs. Government officials reported box 2.1). This dispersion in stakeholder commit- that they had considered that the PRSP needed to ment was echoed in survey results on percep- 1 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 : A N I N D E P E N D E N T E VA L U AT I O N tions of how country-driven the process was (see process, a key forum for policy debate, has figure 2.1). supported the PRSP analytically and has The integration of PRSPs into domestic embraced its principles, to the extent that it is processes such as elections, the budget, and now known as the participatory PER/PRS prior strategies is another marker of country process. Key policy decisions continue to be commitment. Although in a few countries the made outside the PRSP, notably in Ethiopia, PRSP feeds into existing processes, in most case where the government took a series of study countries it is yet to be mainstreamed important decisions in resettlement, into national policymaking. In Tanzania, the education, and technology immediately after participatory Public Expenditure Review (PER) the endorsement of the PRSP, outside of the B o x 2 . 1 P a r l i a m e n t s a n d t h e P R S P Lack of involvement of elected officials in participatory activi- This lack of engagement was often attributed to the domi- ties during PRSP formulation is well documented in the litera- nance of the executive branch in many PRSP countries and the ture (for example, Oxfam 2003). This was validated in the case low capacity of parliaments to engage in detailed policy dis- studies. In Albania, parliament was out of session during PRSP cussions. The Bank and others have recently undertaken activ- consultations, and Mozambique's National Assembly did not ities to address this issue, but the effects have not yet been felt discuss the PRSP. In Tanzania, stakeholders described parlia- in the case study countries. mentary approval of the PRSP as a rubber stamp. F i g u r e 2 . 1 W a s t h e P R S P r o c e s s C o u n t r y - D r i v e n ? Government Donor Civil Society Int'l NGO 1 2 3 4 5 Completely Disagree Completely Agree Note: The thick lines indicate mean score in the survey (from 1: completely disagree to 5: completely agree) on a composite of four questions on whether the PRSP process was country- driven. The box around the bands shows a 95 percent confidence interval. The mean score for the government respondents was significantly higher, at a 95 percent confidence level, than that of donors, and the means for both these groups were significantly higher than that for international NGOs and civil society. The difference in scores was not significant between the latter two groups. 1 2 I M P L E M E N TAT I O N O F D E S I G N F E AT U R E S PRSP framework. Where it has been outputs of consultations. Although in a few mainstreamed, the PRSP has proved to be more Participatory activities countries the PRSP feeds durable under political change. In Albania, the dropped off considerably into existing processes, in PRSP provided a useful fixed point for the civil after formulation of the service during a number of changes in govern- PRSP, as illustrated in box most case study countries ment. In Nicaragua, however, the PRSP was not 2.2. The Bank actively it is yet to be mainstreamed because the government promoted participatory mainstreamed into responsible for its formulation was voted out of activities, funding partici- power shortly after its completion. The new pation advisers in national policymaking. government proposed its own strategy outside countries with limited the PRSP framework to avoid jeopardizing experience (for instance, Cambodia and progress toward HIPC completion point. Tajikistan), facilitating arrangements for dialogue Subsequently, the government has worked (for example, Albania and Vietnam), or putting toward finalizing its strategy as an update to the civil society contacts in its local offices (for original PRSP. instance, Mozambique). Because of these efforts, the Bank was seen as opening the door, or Participation. The BWIs, as well as other donors, increasing the voice, for NGOs in these case study actively promoted broad-based participation as a countries. component of the Initiative's design. In all 10 case The impact of participation is difficult to studies, governments undertook extensive assess because of the lack of clear and shared consultations during formulation of PRSPs, objectives among stakeholders. including in several countries with little tradition of direct government-to-civil-society dialogue (for · For governments, country-driven objectives example, Albania, Cambodia, Tajikistan). A broad for participation were not well defined, and par- range of stakeholders was involved, with the ticipatory activities for composition of participants relatively consistent the PRSP were weakly The Bank was seen as across countries. NGOs and donors figured linked to established opening the door, or prominently in most processes, while organiza- domestic participatory tions located outside the capital, the private processes, where they increasing the voice, for sector, and political parties have been less existed. With govern- NGOs. involved. Feedback from private sector groups ments mostly report- suggests two explanations for their general lack ing on the size and coverage of consultations of engagement. First, they often have well- and not on the tangible outputs of the process, established channels of communication with it appears that consultations were, at least ini- government, which they prefer to use over formal tially, conducted mainly to fulfil a requirement participation in the PRSP. And second, they often for the PRS process. do not see the PRSP as being very relevant to their · Civil society expected participation to in- concerns about investment and growth. Consul- crease the transparency of decisionmaking, tations were generally based on draft PRSPs or and allow them to meaningfully contribute to sector strategies, and the agenda of broad discus- the formulation and implementation of the sions often focused more on poverty diagnostics strategy. By including nongovernmental stake- relative to macroeconomic policy and structural holders, the PRS process has opened up the di- reforms.10 Government officials and civil society alogue space, thereby meeting the first participants at the Addis Ababa workshop noted expectation to some degree in almost all the that the limited capacity of civil society to engage case study countries.11 But civil society has in detailed policy analysis was a barrier to greater been disappointed with regard to their sec- participation (for example, Mozambique, ond expectation. They perceive that they have Tanzania). Multi-stakeholder "sector working had little or no influence on the strategies,12 groups" were commonly used to organize formal with exceptions of bringing governance issues 1 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 : A N I N D E P E N D E N T E VA L U AT I O N P a r t i c i p a t i o n H a s L a r g e l y N o t B e e n B o x 2 . 2 S u s t a i n e d Consultations with civil society in PRSP formulation have been representatives queried the OED/IEO evaluation team for infor- sustained during implementation in a minority of the case study mation on post-PRSP progress. The mechanism for civil society countries. Positive cases include the creation of a participatory input during PRSP formulation (the NGO Counterpart Consortium) institutional framework for monitoring (the Poverty Observa- had faded in the absence of sustained donor funding. In Vietnam, tory in Mozambique and the Poverty Monitoring System in Tan- the government began to play a leadership role in the Bank-man- zania), and internalization of the independent facilitator's role aged participatory process during formulation, but overall the during formulation (the Carter Center in Albania). More often, di- government appears to have assigned low priority to the process alogue between the authorities and nongovernmental national since then. In Ethiopia, civil society's engagement in the PRS stakeholders dropped off after PRSP completion. During in- process fell off considerably following their close involvement country discussions in Tajikistan, NGOs and other civil society in the consultations for the PRSP. more prominently on the table (for example, reduction. The following section assesses Guinea and Mozambique) and canceling user whether PRSP strategies and processes have fees for primary schools (for example, Tanza- included two elements: nia).13 They are concerned about the decrease in engagement in some countries, attributing · Poverty Diagnosis: a multi-dimensional poverty the decline to a lack of genuine commitment profile with a sound analysis of the determi- to broad-based participation and low capacity nants of poverty on the part of governments. · Comprehensive Strategies that ensure consis- · The BWIs' objectives from the participatory tency between a country's macroeconomic, process, like those of governments, are not structural, and social policies and poverty re- well-defined. They are not noted in the JSA duction.15 guidelines, which stipulate that Bank and Fund staff should describe, but not assess, participa- Poverty Diagnosis. All the case study country tion (World Bank and IMF 2000b). The World PRSPs discuss the need to address poverty Bank's PRS Sourcebook (World Bank 2002a), comprehensively and include multidimen- though not technically Bank policy, ascribes the sional income and non-income poverty indica- following desired outcomes to participation: tors.16 Poverty data systems are at nascent enhancing accountability, transparency, and ef- stages of development in most countries, so ficiency of policymaking; increasing equity in poverty profiles were based on the best data policies, goals, and outcomes; and a shared available, which in some cases were long-term vision among all stakeholders (World incomplete or outdated. In many countries, Bank 2002a, p. 239). poverty analyses have drawn from Poverty Assessments conducted or supported by the In summary, views differ significantly across World Bank, and initiated prior to the launch stakeholder groups on the intent of participa- of the PRS process. In Cambodia, by contrast, tory activities. These polarized views are a large body of uncoordinated poverty illustrated in the survey results, shown in figure analyses led to problems in comparability and 2.2, in which the least agreement across uncertainty about the poverty baseline. In all stakeholder groups was seen for questions cases, the Initiative's focus on defining dealing with participation.14 poverty and its determinants has usefully highlighted data gaps and shortfalls in poverty Comprehensive analyses, and in Ethiopia and Tanzania has The PRSP aspires to provide a comprehensive resulted in efforts to improve the quality of view of poverty and a coherent strategy for its quantitative data. 1 4 I M P L E M E N TAT I O N O F D E S I G N F E AT U R E S F i g u r e 2 . 2 O p i n i o n s D i v e r g e o n P a r t i c i p a t i o n Your stakeholders were Final document was modified to Formulation of the adequately consulted during accommodate some of your macroeconomic framework was formulation viewpoints participatory 31% 33% 35% 38% 45% 44% 22% 25% 27% Agree Disagree Neutral Comprehensive Strategies. PRSPs in the case study and the regular BWI PRSPs in the case study countries focus largely on leveraging public tracking of expendi- countries focus largely on expenditures to reduce poverty and have not tures on health and uniformly delineated non-expenditure-related education in HIPC leveraging public policies or actions for enhancing growth or countries (OED 2003b); expenditures to reduce explored their impact on poverty reduction. second, the strong poverty. Consideration of Examples include macro-related areas such as donor preferences for tax/revenue policies and exchange rate manage- these sectors, also obstacles to growth is ment, plus the role of the private sector, trade, manifested in the weak. price policies, and privatization. Consideration relative dominance of of broader obstacles to growth is weak. Within social sector targets in the MDGs; third, the the domain of public expenditures, allocations to more likely "quick wins" in these sectors expanding service delivery in the social sectors (education, health, and social protection) T o p 1 0 P r i o r i t y dominate over investments in economic or T a b l e 2 . 2 S e c t o r s i n P R S P s productive sectors such as infrastructure, b y E x p e n d i t u r e agriculture, and rural development, as depicted in table 2.2. While full analysis of all areas is an 1 Social sectors unrealistic expectation for the early stages of the 2 Infrastructure Initiative, the quality of initial PRSPs highlights 3 Rural development the need for more comprehensive work to serve 4 Agriculture as the basis for poverty reduction strategies, 5 Public sector including in the areas of macroeconomic and 6 Private sector development growth policies and the non-social sectors. 7 Urban development The bias toward a social sector expenditure 8 Environment model of poverty reduction was created by a 9 Financial sector combination of influences: first, the emphasis in the enhanced HIPC Initiative on allocating 10 Economic management Source: Sectoral allocations of costed PRSP programs. "savings" from debt relief to the social sectors 1 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 : A N I N D E P E N D E N T E VA L U AT I O N because of more mature poverty-linked sector unrealistic, given the initial conditions of the strategies; and, fourth, the paucity of analytical country and recent track records, and, in some work in defining broad-based or pro-poor cases, long-term targets are more ambitious growth strategies, specific interventions to than the MDGs (OED 2003c). PRSPs can be used accelerate growth with an understanding of more effectively in enabling countries to select their poverty impact. and monitor relevant country-specific indicators that will provide a continued basis for informed Results Oriented decisionmaking. To what extent have countries adopted a results orientation in preparing and implementing a Prioritization and Costing. Many PRSPs have not PRSP? Desired features of a results orientation been reliably costed and fail to provide strategic are: prioritization, which is symptomatic of the weaknesses in public expenditure management · Setting Goals: with time-bound and moni- systems in these countries. A review of 25 torable indicators of progress completed PRSPs found 8 with serious deficien- · Prioritizing/Costing: public sector resources cies in the costing of programs--some had not and weighing tradeoffs to achieve the goals been costed at all (Albania, Kyrgyz Republic, · Monitoring Arrangements: providing a sys- Tajikistan, Uganda), and others had insufficient tem to monitor implementation and make nec- detail, inconsistencies, or a lack of clarity in essary adjustments. their costing (Benin, Guinea, The Gambia).17 In almost all PRSPs, the assumptions underlying Poor progress in these areas has contributed the costing are not discussed, nor is any attempt to results-orientation being the weakest of the made to relate the cost of activities to success in principles assessed in the case study countries. achieving desired outcomes. A number of Similar messages were echoed in OED's governments in case study countries described workshop in Addis Ababa. For example, the the costing exercise as a strain on capacity, and Guinea civil society representative stated that noted that they would have appreciated Bank "the PRSP is too theoretical, not linked to the support. The hard choices in prioritizing budget, and falls short of being an operational actions over the short to medium term have not road-map." been made in most PRSPs. In The Gambia and Kyrgyz Republic, priorities cover almost all areas Setting Goals. The PRSP has played an important of economic activity. Knowledge gaps about the role in encouraging countries to define indica- impacts of various public actions have impeded tors and quantitative and qualitative targets for tradeoffs with the result that the quality of poverty reduction. In Albania, for instance, the prioritization and costing depends on the second iteration of the PRSP includes assess- quality of underlying sector strategies. The ments by line ministries of their progress toward health and education strategies were the self-defined goals. Efforts to set targets have clearest in priorities and targets, while the been hampered by the lack of adequate data in weakest sector in almost all PRSPs was private many of the case study countries. A review of the sector development. In Ethiopia, strong health case study PRSPs, as well as broader reviews in and education strategies allowed the govern- The PRSP has encouraged the Bank (see Swanson ment to present well-prioritized and -costed and others 2002), shows PRSP programs in those sectors. countries to define a significant overlap indicators and between PRSP indicators Monitoring Arrangements. Progress in establishing quantitative and and the MDGs, although effective monitoring systems in the case study the quantitative targets countries has been slow, and has been qualitative targets for are often different. PRSP hamstrung by a lack of capacity in monitoring poverty reduction. targets are also often and evaluation (discussed in Chapter 3). A 1 6 I M P L E M E N TAT I O N O F D E S I G N F E AT U R E S number of PRSPs do not define clear strategies · Coordinated policy dialogue among donors for improving monitoring (such as Mauritania and with government and Guinea). Nicaragua has proposed a highly · Coordinated and country-led alignment of ambitious centralized tracking system that aims donor assistance to the PRSP to expand the state's presence in all 16 depart- · Reduced transactions costs for the government. ments or provinces. Mozambique and Tanzania are the only case study countries where Partnership orientation was the most mechanisms involving civil society have been positively assessed of the underlying principles established--the Poverty Observatory in in the survey.18 It received uniformly high Mozambique, and a Poverty Monitoring Commit- marks across stakeholder groups, including tee in Tanzania. Besides capacity issues, govern- international NGOs, for whom this principle ments in most countries are monitoring results was the only one that merited a score above as a requirement, and results are not being used neutral. Support for formulation, improve- to adjust strategies or to enhance accountability ments in donor coordination, and alignment of for performance. Albania is an exception, where assistance were rated highly, as shown in figure the measurement of indicators--for example, in 2.3. However, respondents felt that the quality the education sector--has enhanced the defini- of donor coordination was still poor. tion of programs, including the need to improve the quality of tuition and the maintenance and Donor Assistance during Formulation. During upgrading of infrastructure. formulation, donors have provided well- coordinated assistance to governments. In Partnership Oriented Ethiopia, the PRSP approach pushed donors to An increased partnership orientation would agree on a common substantial input in the involve: form of an Issues Note presented during the federal consultations (see box 2.3). In Tanzania, · Adequate and coordinated donor assistance donors were consulted at regular intervals during PRSP formulation during formulation so as not to interfere with P R S P P r o c e s s S c o r e s H i g h o n P a r t n e r s h i p F i g u r e 2 . 3 A s p e c t s , B u t M u c h L o w e r o n C u r r e n t Q u a l - i t y o f D o n o r C o o r d i n a t i o n Q25 56 24 20 Agree Q26 56 27 17 Neutral Q27 52 26 22 Disagree Q28 34 39 27 0% 20% 40% 60% 80% 100% Q25 - Donors supported the formulation of the PRSP without eroding country ownership Q26 - Donors aligned assistance to the PRSP priorities Q27 - Donor coordination improved since the launch of the PRSP process Q28 - Quality of donor coordination at present Note: On question 28, respondents ranked the quality of donor coordination at present from 1 (poor) to 5 (excellent). 1 7 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 : A N I N D E P E N D E N T E VA L U AT I O N the indigenous development of the PRSP. In the PRSP, the PRSP provided a common strategy Nicaragua, the United Nations Development around which to anchor their work. Donor Program (UNDP), the U.K. Department for coordination in Tajikistan is still at a nascent International Development (DFID), and the stage--the first in-country Consultative Group Inter-American Development Bank (IDB) meeting was held only in May 2003--but the provided funding for the consultative process PRSP has provided a useful starting framework. and the World Bank supported the technical drafting unit and contributed several analytical Alignment of Donor Assistance. Given the PRSPs' inputs during formulation. However, donor broad scope and lack of prioritization, inputs in Nicaragua were perceived as alignment would not necessarily entail overstepping the line between support and changes in donors' programs. Neither donors imposition, pushing the government toward a nor the Bank have defined specifically whether social sector emphasis in the PRSP. or how they should change the content of their programs to reflect PRSPs. Changes in donors' Policy Dialogue. The PRSP has provided an arena processes are more readily evident.21 In for transparent policy dialogue among external Mozambique and Tanzania, where govern- partners and with the government. Since the ment-donor coordinating mechanisms were completion of the PRSP in Ethiopia, govern- already strong, the PRSP has increased donor ment and donors have formed a joint task force participation in, and coordination of, budget on harmonization and have agreed on a support instruments and the streamlining of dialogue architecture to facilitate dialogue on their performance monitoring, as outlined in high-level policy issues. In Vietnam, the PRSP box 2.4. Initiative has provided focus to a group of "like- In countries where government leadership The PRSP has provided an minded" donors19 who in aid coordination around the PRSP has been meet regularly to weak, donors have not been able to coordinate arena for transparent exchange information around the PRSP. In Guinea, donors' percep- policy dialogue among and coordinate activi- tions are that the PRSP is not broadly owned. external partners and ties. Part of what defines Donors have been characterized as elusive, with the government. the group is the willing- taking a wait-and-see attitude, reflecting their ness and interest of its skepticism about the PRSP and the govern- members to align (to varying degrees) their ment's ability to implement it. In Tajikistan, activities with the PRSP.20 Working groups set four separate agencies, each with dozens of up around the PER process meet biweekly in sub-units, are responsible for donor coordina- Tanzania and are the main avenue for multi- tion and monitoring donor funding. In Maurita- stakeholder dialogue. Although these predate nia, external partners were initially enthusiastic about coordinating around the PRSP, but this C o o r d i n a t e d diminished as donors came to view the PRSP as D o n o r S u p p o r t f o r being increasingly controlled by the BWIs, B o x 2 . 3 P R S P F o r m u l a t i o n rather than by the government. i n E t h i o p i a In countries with a disharmonious donor context, the PRS process has been unable to The Development Assistance Group (DAG) has over 40 members, cov- enhance capacity for donor coordination. The ering the vast majority of donor agencies working in Ethiopia. In Sep- Albania government found donor coordination tember 2001 the DAG Core Group was established, with 17 members, the most challenging part of PRSP implementa- to support the PRSP in a coordinated manner through a pooled donor tion, because the fragmented and large donor fund. The Core Group also prepared a joint Issues Note as an input to community perceived the PRSP as a Bank- federal consultations and organized the preparation of a Joint Partner driven process and were unwilling to align their Review to complement the JSA. own approaches with it. UNDP pursued its own MDG-related activities in parallel with the PRSP, 1 8 I M P L E M E N TAT I O N O F D E S I G N F E AT U R E S P R S P H e l p s D o n o r s C o o r d i n a t e B u d g e t S u p - B o x 2 . 4 p o r t a n d S t r e a m l i n e P e r f o r m a n c e M o n i t o r i n g In Tanzania, the Poverty Reduction Budget Support Group coor- ditionality. In Mozambique, the G-11 donors that provide budget dinates members' conditionality using a single monitoring in- support are also developing a PAF, which is intended to better align strument (the Performance Assessment Framework, or PAF). The with the PRSP and harmonize donor conditions (including align- PAF uses Tanzania's PRSP as its guiding framework, and moni- ment with the PRGF and a Bank PRSC in the pipeline). In Vietnam, toring is synchronized with PRSP progress reports. With its first a number of donors are delivering a significant share of their aid PRSC, the Bank joined the group and now uses the PAF for its con- budgets through the World Bank Poverty Reduction Support Credit. which they viewed as a potential input into the There has been an Neither donors nor the achievement of the MDGs. A similar experience increase in the prevalence Bank have defined is noted in the Cambodia case study. After of budget support as an specifically whether or identifying monitoring indicators in their PRSP, aid delivery mechanism,25 the authorities were asked to engage in a and stakeholders perceive how they should change separate exercise to develop localized MDG that donors have aligned the content of their indicators. 22 Nicaragua represents a case where assistance to PRSP priori- programs to reflect PRSPs. a disharmonious donor context is exacerbated ties (figure 2.3). However, by the division of responsibilities for donor evidence of more coordi- Changes in donors' funding and public finances among four nated selectivity of donor processes are more agencies.23 Activities to coordinate donors are portfolios in support of readily evident. under way--for example, through the PRSP priorities is lacking. introduction of a programmatic approach in This is partially the result public sector reform--although progress has of the inertia from ongoing commitments and the been slow. time needed for donors to adjust strategies and Donor feedback from the case studies programs. Another challenge for gathering confirmed that the Bank has improved its evidence on alignment is the breadth of PRSP coordination and relations with other external priorities and programs that offer donors the partners. This is attributable not only to the latitude to formulate programs without country- PRSP, but also to decentralization and the driven parameters or constraints. In Tajikistan, the Bank's move into budget support groups, in the PRSP has so far failed to reconcile a substantial gap context of PRSC preparation. Yet the process between the government's preference for high has not been without problems between the spending on infrastructure investment and the Bank and other external partners. Only a third donors' preference to concentrate their finance on of donors agree that "the Bank is actively social sectors and safety nets. promoting and supporting coordination of donor assistance," compared with half of the Transaction Costs. In country case studies, the total survey respondents.24 In Cambodia, there rhetoric of donor alignment around the PRSP has was open conflict between the ADB, the UNDP, yet to reduce transaction costs for the govern- and the Bank during PRSP planning and ment. Governments reported that multiple donor formulation. In Nicaragua, there were relatively initiatives, including the PRSP, have actually poor relations between the IDB and the Bank. increased transaction costs in the short term. While many bilaterals appreciate and expect the These transaction costs range from formal report- Bank to play a leading role--with the PRSP as ing requirements and indicator tracking to the focal point--it is not so clear that the UNDP accommodating donor missions and attending and regional banks see things the same way. meetings. The PRSP has not yet led to alignment Donor alignment to the PRSP through changes of donor actions and government budget cycles, in lending/grant portfolios is difficult to measure. nor have donors harmonized their reporting 1 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 : A N I N D E P E N D E N T E VA L U AT I O N requirements with the APRs (except for timing of nate in Albania, with the Bank and Fund budget support instruments in Ethiopia and working together to build incentives for private Tanzania, and varied progress in joint perform- sector development into proposals for tax ance monitoring arrangements in Ethiopia, reform. In Ethiopia, Bank staff pointed to Mozambique, and Tanzania). Donor practices such increased cooperation in financial sector as reliance on project implementation units development. In three other countries, underminethepotentialforgovernment-leddonor however, there were policy differences collaboration. Representatives of African countries between the Bank and the Fund that the PRSP at the case study framework alone was unable to overcome Governments reported workshop in Addis Ababa (Mozambique, Tajikistan, and Vietnam). In all that multiple donor also noted that there are three cases, there were policy differences initiatives, including the multiple donor-driven between the BWIs on the scope and sequenc- initiatives that govern- ing of structural reforms (in financial sector PRSP, have actually ments must address development in Mozambique, the energy increased transaction alongside the PRSP Initia- sector in Tajikistan, and reform of state-owned costs in the short term. tive, such as the New enterprises in Vietnam). The existence of a Partnership for Africa's PRSP did not imply convergence of views, even Development (NEPAD) and the Strategic Partner- in cases where the process was considered ship with Africa (SPA), which in the aggregate highly country-driven, with clear indications of result in large transactions costs for African the government's views on reforms (Vietnam). countries. Bank-Fund Collaboration. Collaboration between Long term the Bank and Fund was expected to improve The long-term perspective for poverty under the PRSP Initiative, amid broader efforts reduction entails taking a strategic approach to to encourage systematic coordination between short-term programs with a view to meeting the two institutions (see World Bank and IMF long-term goals. This section asks whether 2001a, 2002a; World Bank 2002b). As part of the poverty reduction strategies have been defined PRS process, Bank and Fund staff have carried in the context of long-term goals. out joint missions--including the conduct of All case study countries had long-term JSAs--in connection with the PRSP in all 10 planning documents and processes in place case study countries. In addition, the two prior to the PRSP. The most common were the institutions have worked together to prepare country Vision 2020 initiatives, some of which progress reports on the implementation of the had become a central part of government PRSP Initiative that were presented jointly to planning (such as Tanzania Vision 2025; see the Development Committee.26 This joint work United Republic of Tanzania 1998). The PRSP has created opportunities for more frequent has been adopted alongside other long-term discussions among staff of the two institutions. planning processes, such as country Vision Bank staff noted greater interaction with the exercises, country MDG plans, and regional or Fund, including work in human development. global commitments such as NEPAD. But Working relations between the two institutions PRSPs, which are, on average, three-year were smooth in the case study countries: 52 programs for poverty reduction, do not analyze percent of respondents ranked the quality of how they will achieve longer-term goals, nor do Bank/Fund collaboration as good or excellent.27 they reconcile their own long-term targets with In two of the case study countries (Albania those of other long-term plans. In and Ethiopia), stakeholders viewed Bank-Fund Mozambique, a process has been started to collaboration as improving since the PRS define a Vision 2025, and it is unclear how this process was introduced. The PRSP provided a process will complement the PRS process. The framework for the two institutions to coordi- PRSP itself defines an overall poverty reduction 2 0 I M P L E M E N TAT I O N O F D E S I G N F E AT U R E S target for 2010, without specifying how the individual lending operations. In this context, country will reach this goal. In Tajikistan, the the Board review and endorsement of PRSPs PRSP defines targets for 2006 and 2015, but appear redundant, as well as attenuating does not provide any clear rationale for the ownership in the eyes of many national targets or how they can be met. stakeholders. The inconsistency of medium-term targets with long-term goals implies that PRSP strate- Joint Staff Assessments gies are in danger of not delivering countries' The JSA was designed to play two principal long-term visions. For instance, Tanzania's roles. According to the JSA Guidelines, it Vision 2025 aims to "create a critical mass of a should provide the BWI Boards with an assess- well-qualified human resource to face the ment of the soundness of the PRSP as a basis challenges of globalization." But PRSP targets for concessional assistance, and thus in education focus exclusively on primary represents an instrument of due diligence, education, which has also seen the highest although with differing roles (box 2.5). It budget increases in recent years. The formula- should also provide tion of a secondary education strategy has only "constructive feedback Many national recently been given attention in the 2004 Third to the country" on how stakeholders, especially Annual Progress Report. Albania provides a the PRSP strategy can be within civil society, good counterexample, however, because improved over time.28 consider that the Board European integration is the country's primary long-term objective and PRSP goals have been Analytical Quality. OED review process made consistent with requirements for the assessed all 28 JSAs undermines ownership stabilization and association process. associated with full and limits the scope for a PRSPs that had been The Value Added of Process endorsed by the Boards country-driven and Requirements of the Bank and Fund by country-defined strategy the end of June 2003.29 process. Board Consideration of PRSPs OED reviewed each After a country completes a PRSP, Bank and JSA's treatment of 11 Fund management present each PRSP to the issues, as listed in figure 2.4, to complement BWI Boards. Though BWI policy papers charac- the IEO's parallel review of how the same terize this process as endorsement of national group of JSAs handled 5 additional issues, as strategies, many national stakeholders, described in box 2.6.30 Among the 11 issues especially within civil society, see this process that OED rated, the JSAs in the aggregate as BWI approval that undermines ownership treated 5 at a satisfactory level, as shown in and limits the scope for a country-driven and figure 2.4 (where 3 out of 4 represents satisfac- country-defined strategy process (see, for tory). Assessment of endogenous and example, ActionAid USA/ActionAid Uganda exogenous risks and structural and sectoral 2004). They cite this practice as demonstrating policies were the best-handled topics, with 20 that the Boards are the key audience for the of the 28 JSAs rated satisfactory or better in PRSP and that Washington approval is the goal both areas. The JSAs were particularly weak in of the exercise. In practice, Board presentation their treatment of private sector participation of PRSPs also has little operational value, given and partnership issues. The quality of JSAs that the Board has endorsed all PRSPs thus far improved only slightly over the time period as a basis for concessional financing and that studied. The 13 earlier JSAs (reviewed by these discussions are generally separate from executive directors through the end of FY02) established mechanisms for authorizing World were rated 2.6 overall, and the 15 later JSAs Bank support to a country--notably the (reviewed during FY03) were rated 2.8. The endorsement of CASs and the approval of JSAs markedly improved their treatment of two 2 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 : A N I N D E P E N D E N T E VA L U AT I O N J S A s P l a y the participatory process acknowledged in the D i f f e r e n t R o l e s PRSP, and in Tanzania, the JSA did not address B o x 2 . 5 f o r t h e B a n k a n d the implications of a short consultation t h e F u n d process. Only for Mozambique did the JSA describe the limited role that parliament and The Bank and the Fund evaluate the government's poverty reduction civil society had played in that process. strategy in the JSA as a basis for concessional assistance. For the Bank, JSAs pay inadequate attention to existing the CAS is meant to build on the JSA to selectively design its assis- domestic processes that are relevant for a tance. The IMF does not have a clear equivalent to the Bank's CAS; the sustainable PRS process. For example, most JSA itself is intended to provide the operational link to the PRGF by "con- JSAs discuss the quality and realism of indica- tribut[ing] to the [determination of the amounts of assistance and de- tors and targets in the PRSP, but treatment of tailed design of the programs] through its assessments of the strengths the role and usefulness of the information for and weaknesses of the strategy." line ministries and other decisionmakers receives much less attention. Similarly, specific sector policies are assessed, but the policymak- challenging issues: balanced treatment of social ing process generally is not. This suggests the and growth sectors and treatment of alterna- scope of JSAs could be altered or amended to tive growth strategies. Box 2.7 illustrates how leverage this instrument to better support the two JSAs did a good job assessing--one PRS process. favorably and the other critically--the PRSPs' The JSA represents an assessment by Bank alternative growth strategies. staff about programs and strategies that the Stakeholders considered that the JSAs failed Bank has often contributed to developing to describe key weaknesses in the PRSP partici- through its support and comments during the patory process. Most JSAs provide only factual formulation not only of the PRSP, but also of its descriptions of participatory processes in building blocks, such as Public Expenditure conformity with the guidelines, which rule out Reviews (PERs) and Poverty Assessments (PAs). evaluative or normative statements on partici- This element of self-assessment in the JSAs may pation. But in 9 of the 10 case study countries, account for some of the analytical gaps and this factual treatment was overly optimistic or limited candor noted above. As a proponent of incomplete regarding the value and contribu- the PRSP Initiative and supporter of countries' tion of the participatory process. In Guinea, for PRSP formulation activities, the World Bank has example, the JSA did not mention limitations in an incentive to provide favorable assessments of the PRSP documents. All JSAs to date have concluded that the relevant PRSP provides an K e y F i n d i n g s o f adequate basis for concessional assistance B o x 2 . 6 I E O R e v i e w o f (although "adequate" is not defined). J S A s Usefulness of the JSA Feedback. All JSAs of case The IEO assessed the consistency of JSAs with five key elements of study PRSPs included advice and recommenda- the JSA Guidelines not covered by the OED review. The IEO review con- tions to country authorities and other stakehold- cludes that the JSA has met its objectives to a limited degree. In par- ers, which governments in Albania, Cambodia, ticular, the JSAs fail to address capacity constraints, are read by few and Tanzania, among others, found to be useful. decisionmakers, and reach formulaic conclusions. These character- But in general, only key officials in central istics limit their usefulness. OED concurs with these judgments. Both ministries are aware of the JSA content, which reviews note that there has been an incentive for BWI staff to reach has limited its usefulness to the broader positive findings on the adequacy of the PRSP in order to validate the audience of stakeholders. Governments in most strategy as a sound basis for World Bank and Fund concessional countries did not circulate the JSA to civil society, assistance. which is expected to play an active role in PRSP implementation. In the majority of cases, JSAs 2 2 I M P L E M E N TAT I O N O F D E S I G N F E AT U R E S F i g u r e 2 . 4 J S A s A r e W e a k i n S o m e A r e a s Treatment of 11 Selected Issues in 28 JSAs Highly satisfactory 4 3 Average 2.7 Rating 2 Highly 1 unsatisfactory and MTEF social policies issues in diagnosis of sectors short- strategies exongenous process managementof objectives constraints participation and sectoral Poverty strategy growth of budget treatment partnership and growthsector Capacity sector and and Integration long-term Donor public Private Balanced Alternative Structural Integration and Risks--endogenous Governance Issue G o o d P r a c t i c e T r e a t m e n t o f A l t e r n a t i v e B o x 2 . 7 G r o w t h S t r a t e g i e s The Senegal JSA was rated highly for its assessment of alter- "The PRSP lacks alternative frameworks or contingency native growth strategies in the PRSP. It described the three policies. Notably, the PRSP does not include an assess- macroeconomic scenarios and assessed their realism with re- ment of how the upcoming elimination of textile quotas spect to anticipated level of external financing, domestic re- will affect Sri Lanka's economic growth or the poor. For source mobilization for priority investments, and implementation this reason, staffs encourage the Government to develop capacity. It also recommended a specific case scenario and ad- an alternative, lower growth scenario and design con- vised authorities to provide an in-depth analysis of scenario tingency plans that fully reflect the envisaged risks." feasibility in the Progress Report. The Sri Lanka JSA makes a cogent argument for developing Source: Senegal JSA (December 2002) and Sri Lanka JSA (April alternative growth strategies: 2003). did not include the views of other stakeholders. an alternative to predominantly government and Some donors have called for a more inclusive BWI interactions. A more transparent and open assessment mechanism, similar to the joint process with other partners would be more donor reviews of performance assessment consistent with the partnership principle and frameworks associated with coordinated budget would enhance discussions of coordinated support found in Mozambique and Tanzania, as support, selectivity, and comparative advantage. 2 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 : A N I N D E P E N D E N T E VA L U AT I O N Annual Progress Reports monitoring and implementation systems. All 12 HIPC countries normally have to produce Annual countries reported their achievements in one or Progress Reports (APRs) that demonstrate one more priority sectors using quantitative indica- year of successful PRSP tors; 9 reported against either annual or medium- Progress reporting has implementation to reach term targets.34 Burkina Faso, for example, created strengthened the PRS the HIPC completion 31,870 water supply points in 2001 against a target process by increasing the point. Other countries of 31,000. Nine of the 12 APRs also reported on the have to produce APRs to country's progress in fulfilling qualitative policy relevance of the strategy retain access to conces- measures, but not always against a timeline. For and by providing a sional resources from instance, the 2003 Finance Law in Niger fulfills a yardstick for the IMF through PRGF 2002 PRSP target to adopt government accounting programs.31 APRs have and budget nomenclature in accordance with accountability. two objectives: first, to WAEMU guidelines. assess progress toward PRSP goals and intended policy/program reforms and, second, to allow Adjusting Strategies. The case studies reveal a governments to make adjustments in their number of examples of how progress reporting strategies in response to changing circum- has strengthened the PRS process by increas- stances.32 Twelve countries have completed at ing the relevance of the strategy and by provid- least one APR on the implementation of their ing a yardstick for accountability (see box 2.8). PRSP.33 This section reviews the extent to which In Ethiopia, production of an APR increased the APRs have served these two objectives in the case government's focus on monitoring and evalua- study countries and outlines several constraints tion (M&E) and resulted in a comprehensive that derive from the role of APRs as a BWI report- plan for monitoring. Donors interested in ing requirement. increasing budget support are also considering using the APR as the basis for a shared The constructive role Assessing Progress. Coun- conditionality framework and joint donor tries are using the APRs to performance assessments. In Mauritania and played by Annual report on progress in Tanzania, the APRs used detailed and updated Progress Reports is implementingtheirpover- household survey data that had not been compromised when they ty reduction strategies. completed in time for the PRSP. To be relevant Ten of the 12 APRs re- operational tools, APRs must be capable of are not connected to viewed describe the reorienting the PRSP strategy based on a sound domestic processes. status of poverty, develop- analysis of indicators. In Mozambique, the APR ments in the macroeco- provides a wealth of indicators, but it does not nomic framework, and the implementation of provide any guidance on reorienting the priority sector policies and performance, and the strategy to increase effectiveness. In Maurita- majority reported on their progress in developing nia, the first APR projected a dramatic increase N e w G o v e r n m e n t i n A l b a n i a U s e s B o x 2 . 8 A P R t o A d j u s t t h e P R S P Albania's first APR used newly released data from monitoring units and recent poverty reports to revise and up- date the PRSP's objectives. The APR was also used to improve the strategic alignment between the PRSP and the Millennium Development Goal and European Union accession processes. The APR included self-assess- ments from all line ministries. 2 4 I M P L E M E N TAT I O N O F D E S I G N F E AT U R E S in the costs of the PRSP program, but it did not processes. This lack of alignment implies explain the rationale for it. This was remedied duplication of effort and decreases the in the second APR, which presented more relevance of the APR as an operational tool. The realistic expenditure figures. costs of reporting requirements are high for The constructive role played by APRs is PRSP countries; many need to report to donors compromised when they are not connected to while continuing institutionalized or legally domestic processes such as budget or govern- mandated reporting requirements in-country. ment reporting, as described in box 2.9. In As currently constituted, the APRs are an almost all case studies, the timetables for APR additional strain on limited country capacity.36 production reflected BWI requirements,35 Stakeholders at the Addis Ababa workshop on rather than domestic monitoring needs. case study findings noted that APRs were not Stakeholders emphasized that the APR require- candid on progress in-country, another signal ment does not help to validate or strengthen that the APR is an external requirement rather existing institutionalized domestic monitoring than an operational document. A P R R e q u i r e m e n t s U n c o n n e c t e d t o B o x 2 . 9 D o m e s t i c P r o c e s s e s In Tajikistan, the government is obligated by law to prepare a national report for Parliament on the implemen- tation of the PRSP, with its own specifications, but these differ from those suggested in the BWI guidelines for APRs. In Albania the APR was completed in three months and did not include extensive consultations. In Mozambique, the need to fulfill PRGF requirements meant that the country's first APR submitted to the BWI Boards had only partial-year coverage and was done in parallel to regular reporting. This has since been corrected dur- ing preparation of the second APR, which is the same as the government's annual document. 2 5 3 Preliminary Results C hapter 2 assessed the application of the underlying PRSP principles and the value added of the APRs and JSAs. In the context of the results chain depicted in figure 1.2, this represents a review of the inputs and some of the outputs in the 10 case study countries. This chapter further assesses outputs and the intermediate and longer-term outcomes of the process. As noted in Chapter 1, the BWIs have left it Value Added in Planning and Links to to the country authorities to define the specific Implementation results and milestones expected from applica- tion of the underlying principles (for example, Planning what is expected from participatory activities). An increased focus on An increased focus on While the Bank and Fund review country poverty reduction and poverty reduction has experience in implementing the Initiative the need to base PRSPs annually through PRSP--Progress in on a sound poverty resulted in marked Implementation reports,1 they do not assess diagnosis has highlighted improvements in quality the contributions of the Initiative against the poor quality of data of poverty data. previously identified benchmarks. This chapter on poverty in many of assesses the implementation and preliminary these countries. In some cases, this has resulted in results of the process, with a focus on three marked improvements in the quality of those data. areas of progress that derive from the results For instance, the Tanzania PRSP was based on chain and early policy papers. First, are PRSPs dated sources--the most recent being a 1998 effective tools for planning and to what degree survey3--although subsequent adjustments to are they being implemented? Second, has the the strategy were able to use the more recently PRS process enhanced countries' capacities for completed Household Budget Survey of 2000/01. implementation, particularly for monitoring In Mozambique, the PRSP provided the first and evaluation? And third, what outcomes have detailed assessment of quantitative aspects of been achieved?2 poverty and revealed large regional disparities in 2 7 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 : A N I N D E P E N D E N T E VA L U AT I O N poverty incidence. The Mauritanian survey in most case studies the individual sector provided greater insight into how the poor strategies were not changed substantially in the perceive their participation in public or formulation stage. In Cambodia, Ethiopia, and community activities and the functioning and Tanzania, strategies in the sectors with the quality of public services. In addition, the process strongest donor presence--most commonly of placing sectoral programs under one strategy education and health--have tended to be the umbrella has made the link between sectoral most well-developed, but specific growth strategies and poverty reduction more tangible. strategies, especially for agricultural growth, Central ministries have begun to ask line have been neglected. In Tanzania, for instance, ministries to rank programs based on their likely where poverty is largely rural, development of poverty impact. In Albania and Nicaragua, the the agricultural sector strategy was not well- PRSP was helpful in spurring movement toward linked with the PRS process, and was not defining outcomes in poverty-related areas. In finalized until after the PRSP. The correspon- Guinea and Tanzania, the ding program was not ready at the time of the The process of placing PRS process comple- first APR. The process of aggregating individual sectoral programs under mented ongoing efforts sector strategies under one umbrella has thus to focus resources on highlighted shortcomings across sectors in the one strategy umbrella has priority sectors. Most case study countries, and in some cases a made the link between stakeholders agreed that process is in place to address these sectoral strategies and the PRS process was weaknesses. The collection of sector strategies focused on outcomes has also encouraged the central ministries of poverty reduction more that benefit the poor (see finance and planning to attempt coordination tangible. figure 3.1), and ranked of key programs across line ministries as a first that feature among the step toward cross-sectoral planning (see box five most positive in the survey, with a mean score 3.1). Cross-sectoral linkages begin from a weak of 3.6. starting point, however, and the poverty The PRS process has brought together reduction strategies do not give much consid- sector strategies in one policy document, but eration to tradeoffs among sector priorities. The poverty focus of the PRSP is focused more on expenditure allocations than on I s t h e P R S P defining medium- or long-term comprehensive P r o c e s s F o c u s e d F i g u r e 3 . 1 or individual sector strategies or policy o n O u t c o m e s T h a t measures. Sources of revenue and GDP growth B e n e f i t t h e P o o r ? have received uneven attention across the case study PRSPs. In Mauritania, for instance, the PRSP did not consider the options for using the 19% country's potential hydrocarbon revenues, although there are still some uncertainties about the true potential of the country's 55% revenues. The critical issue of trade policies 26% and their impact on growth have been mentioned in only 3 of the 10 case study PRSPs. Cambodia's PRSP is an exception: it integrated the trade policy dimensions that were Agree emerging around the country's accession to Neutral the World Trade Organization (WTO). Where Disagree details of revenue projections or resource Source: Stakeholder Survey, question 16, Annex B. environments are included, they lack realism. The case of Cambodia exemplifies this 2 8 P R E L I M I N A R Y R E S U LT S T h e P R S P H a s R a i s e d A w a r e n e s s a b o u t B o x 3 . 1 E x p l o r i n g S y n e r g i e s a c r o s s S e c t o r s In Albania, the APR played a useful role in providing links be- preparation process, guided by an inter-ministerial steering tween the line ministries and the core ministries responsible for committee and technical committee, enhanced inter-sectoral the PRSP, although this process was highly dependent on local focus and made ministries more mindful of key linkages be- consultants. Cambodia's PRSP process resulted in the Planning tween sectors. In Mozambique, the PRSP process brought greater Ministry and, subsequently, the Finance Ministry making a coherence to the government's poverty reduction policies, if valiant attempt to prioritize across sectors. This has resulted in not a major reshuffling of priorities. In Tajikistan, the organiza- line ministries starting to think beyond their individual sectors tion of PRSP consultations around nine working groups under in strategy development, which manifested itself in a cross- the direction of an overall steering committee helped to promote sectoral coverage of trade in the PRSP. In Ethiopia too, the PRS a comprehensive view that had not existed before. problem--total PRSP program costs amounted budget alignment with Sources of revenue and to $5 billion, against anticipated funding of only the PRSP. GDP growth have received $1.5 billion, half of which was already commit- The sample countries uneven attention. ted to ongoing projects. The Ethiopian PRSP differ little from other does provide an estimate of the rate of GDP low-income countries, growth that would have to be sustained to or indeed developing countries more generally, achieve the MDG for poverty reduction.4 None in their weakness of public expenditure of the other case study PRSPs goes this far. But management and lack of adequate integration even in Ethiopia, the full cost of the programs between budgets and medium-term instru- necessary to achieve PRSP targets exceeds the ments. But the PRSP has improved the resources envisaged in the macroeconomic prospects for development of public expendi- and fiscal framework. Stakeholders in Tanzania ture management (PEM) and for the linkages attributed the heavy focus on expenditures to between the PRSP and the budget. The focus on continuing gaps in the knowledge of the the actions necessary for strengthening the poverty impacts of macro and structural tracking of poverty expenditures--a result of policies. the HIPC Initiative--is responsible for some of the progress, such as the extensive PEM reform Implementation process in place in The priorities for expenditure in the PRSPs Mozambique. But non- The three-way linkage of have not been translated into priorities in the HIPC countries have the PRSP, the budget, and budget, because the three-way linkage of the also seen improvements the Medium-Term PRSP, the budget, and the Medium-Term in this area. For Expenditure Framework (MTEF) is typically example, Albania has Expenditure Framework weak or absent. In Cambodia and Guinea, made a start in linking is typically weak or spending plans for line ministries are being the PRSP with the absent. formulated with little regard to the PRSP. In budget and the MTEF, Tajikistan, the link between public spending with a shared institu- and policy objectives is very limited, and tional structure for the three processes. Indeed, constrained to a few sectors. Exceptions are the MTEF was set up explicitly to underpin the Tanzania, where the PRSP has served to identify PRSP. There has also been substantial reform in policies and priority programs that have PEM, albeit from a very poor initial condition. formed the basis for prioritizing the allocation The alignment of these processes is still in the of public expenditures, and Ethiopia, where early stages, but the structure is in place for its the government has taken steps to improve ultimate achievement. Public Expenditure 2 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 : A N I N D E P E N D E N T E VA L U AT I O N The PRSP has improved Reviews, which are increased by about 2.5 percentage points of the prospects for being done more GDP between 1999 and 2002, but these trends development of public regularly, are useful cannot be considered robust in light of defini- inputs into aligning tional concerns with the data aggregation. expenditure management PRSPs with budgets, OED analyzed sectoral trends in expenditures and for the linkages particularly if they are in 14 PRSP countries with sectoral expenditure between the PRSP and the carried out on a rolling data.7 In the four-year period prior to the basis (see box 3.2). completion of the PRSP, expenditures on budget. Have actual expendi- education and health had already been increas- tures changed post- ing as a percentage of GDP, while the agriculture PRSP? The concept of poverty-reducing and transport sector shares stayed constant. expenditures (PRE) has emerged from the This is consistent with OED's finding that HIPC HIPC expenditure-tracking exercises to resources have been targeted toward increased demonstrate greater budget alignment with expenditure in the social sectors (OED 2003a). the goal of poverty reduction.5 These expendi- Spending in the first year of PRSP implementa- tures are defined differently across countries,6 tion (designated as "T+1" in figure 3.2), and within countries their definition may vary available for five countries, continued patterns across time as countries adjust their priorities, of previous trends.8 Annual expenditure as a which makes time trend analysis very difficult. percentage of GDP was higher post-PRSP in the Two analyses report a significant enhancement education, health, and transport sectors of PRE since the launch of the PRSP Initiative. compared with the mean annual expenditure for The 2003 PRSP Progress in Implementation the pre-PRSP period. The largest absolute Report finds that poverty-reducing outlays increase is in education--from 3.0 percent to 3.8 increased by an average of 1.4 percentage percent of GDP--but these differences are not points of GDP and 3.9 percentage points of statistically significant for any of the four total government spending between 1999 and sectors.9 So far, there is no evidence that the 2001 in the 14 countries with available outturn PRSP has affected the pre-existing trend toward data. IEO, looking at data for 19 PRSP increasing education and health expenditures-- countries through 2002, also found that PRE a trend that emerged from the HIPC Initiative. G o o d P r a c t i c e : R o l l i n g P E R s A n a l y z e O p - B o x 3 . 2 t i o n s a n d P r o b l e m s i n P E M a n d D e t a i l e d B l u e p r i n t s f o r C h a n g e All the case study countries except Mauritania have had at which permits manageable subsets of issues to be tackled se- least one recent PER. In two cases, Ethiopia and Tanzania, PERs quentially. Although resource-intensive, this type of cumula- have been annual events. Increasingly, the diagnosis and rec- tive process seems appropriate in cases where a lengthy and ommendations of PERs have graduated from reiteration of best sustained effort at reform and capacity building is required. practice to a more nuanced view of "good enough" practice and PERs can also provide a convenient mechanism for coordi- feasible rates of change in the country-specific context. nation of donor activities, and of donor activities with govern- Recent PERs have also paid increased attention to institu- ment policy (whether or not that is included in a PRSP). Where tional issues. They have played a useful role in underpinning the the government has a well-developed MTEF, the PER can be PRSP when the country's medium-term apparatus is absent or aligned with that, or the PER can be an input into the improve- incomplete. ment of the MTEF. Finally, experience in Tanzania has led to the The experiences of Ethiopia and Tanzania suggest that these PER itself becoming participatory, and consequently identified benefits are enhanced when the PER is set up on a rolling basis, with the PRSP. 3 0 P R E L I M I N A R Y R E S U LT S P o s t - P R S P E x p e n d i t u r e s C o n t i n u e F i g u r e 3 . 2 P r e - P R S P T r e n d Expenditures (percentage of GDP) 4 3.5 3 2.5 2 1.5 1 0.5 0 T-3 T-2 T-1 T T+1 Education Health Note: T+1 represents the first year of PRSP implementation. Capacity Enhancement and for different reasons in each of the For the PRS to be sustainable, the process itself countries. must promote individual and institutional There are severe capacity problems in capacity development in PRSP countries. Has Cambodia, where a shortage of skills and the PRS process been successful in this task? experience at the staff level is compounded by While the need for expanded capacity is not weak and fragmented national institutions and limited to governments--the PRS process is a pervasive system of patronage that inhibits likely to be successful if domestic NGOs and staff performance. Albania also has a long way the private sector can play an effective role to go in establishing the through participation and monitoring--the capacity to support the Weak public sector focus is clearly on whether governments can PRSP, despite improve- capacity is a key design, implement, and monitor poverty ments in structures and impediment to the reduction programs. procedures within the In many of the case study countries, weak government. A rapid attainment of successful public sector capacity is the key impediment turnover in government PRSP outcomes. to the attainment of successful PRSP officials as a result of outcomes. Capacity constraints are widely low salaries and the acknowledged as posing obstacles to high number of political appointees has shaped implementation of the PRSP. The World Bank a civil service with poor skills and little motiva- PRS Sourcebook states that "evolution of a PRS tion to perform well. The lack of skills is even in a country will depend on the degree of more severe at the local level. Mozambique institutional and technical capacity to design faces immense challenges in translating plans and implement sectoral programs and policies into budgets, making budgeted funds available, to tackle poverty" (World Bank 2002a). Weak ensuring they are spent appropriately, and capacity has manifested itself in different ways enhancing monitoring and reporting. In all 3 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 : A N I N D E P E N D E N T E VA L U AT I O N case study countries, the modest capacity of of WBI activities is beyond the scope of this civil society to contribute to policy analysis is a evaluation, but evidence from the case studies significant constraint to widening the participa- suggests that both the activities and their value tory processes of the PRSP. The challenges of added varied considerably. The greatest building capacity at the sub-national level have awareness and appreciation for these services been noted in a number of cases (box 3.3). was in Tajikistan, a focus country for WBI. The PRS process has brought attention to Several stakeholders in government and civil capacity constraints that inhibit growth and society reported a value added from listening poverty reduction. It thus has the potential to to the experiences of other countries at help the authorities better prioritize capacity regional events held by WBI in Moscow and building activities and better clarify whether the Budapest. support being provided by the donor Regarding participatory activities, all case community is appropriate. The PRSP does not study governments undertook substantial appear to have been used in this way in most efforts to conduct consultations during case study countries, mostly because of a lack of formulation of their PRSPs. Through carrying comprehensive planning for capacity building. out these activities, capacity was gained, The clearest example of linking the planning of especially in countries with limited prior a poverty reduction experience. However, the tapering off of The PRS process has the strategy with the consultations after completion of the PRSP potential to help the planning of capacity document in many case study countries authorities better enhancement priorities suggests that those gains may not be sustained. was seen in Ethiopia, prioritize capacity where the PRSP Monitoring and Evaluation building activities, but provided an opportu- Most of the case study countries have limited does not appear to have nity for the government monitoring and evaluation capacity at the to articulate its capacity national and local levels, and a large share of been used in this way. building program as a the existing capacity tends to be associated crucial pillar of its with project implementation units (PIUs) for overarching policy framework. donor-supported projects. This capacity has The World Bank Institute (WBI) has had limited value, because PIU staff were not provided many activities geared toward responsible for developing indicators for the enhancing capacity in countries preparing projects, and the specialized data collection for PRSPs. These activities mainly center around projects is generally not connected to national WBI's Attacking Poverty program, which offers or sectoral databases. From this weak starting courses and seminars, both in Washington, point, the PRSP has spurred sustained interest D.C., and abroad. A comprehensive treatment in enhancing institutional capacity at the W e a k S u b - N a t i o n a l C a p a c i t y M a k e s B o x 3 . 3 I m p l e m e n t i n g P R S P s M o r e D i f f i c u l t In Ethiopia, Mozambique, and Tanzania, capacity weakness at sub- ing the capacity of the current system. In Albania, the government national levels of government was identified as a major challenge. has a decentralization program, but it is being implemented at a In all three cases, a wide range of ambitious reforms dealing with different pace and in a different way by each ministry, and there decentralization of planning, monitoring, and implementation is no clear roadmap to help put in place the local capacity needed. systems is being carried out. While there are high expectations The adequate sequencing of reforms, particularly in monitoring of the positive impact of these reforms on the implementation of and evaluation, will have a significant impact on the integration the PRSP, there is also great risk involved, as the reforms are strain- of PRSP M&E systems into national processes. 3 2 P R E L I M I N A R Y R E S U LT S national level in some of the countries studied. of Tanzania, which has developed a compre- In Albania, Tanzania, and Vietnam, the PRSP hensive national poverty-monitoring system, generated substantial movement. In Albania, problems are being encountered in data the PRSP crystallized the government's finding interested users. decision to move ahead with monitoring activi- Effective dissemination that brings ties in all ministries (with the personal interest understanding of the PRS process to the public of the prime minister). In Tanzania, a broadly at large stimulates demand for monitoring and participatory mechanism for monitoring has evaluation, but this effort has been quite weak been introduced, which has substantively thus far. In most case study countries, govern- enhanced national processes for poverty ments have not mounted effective public monitoring. relations campaigns. Press coverage of the Other countries have initiated efforts to PRSP has generally been limited to the actual improve capacity in monitoring, but progress launch of the document with stories for a day has been slow. A common stumbling block has or two and almost no been the decision to track a large number of follow-up. The potential Effective dissemination of indicators in spite of significant data challenges role of the media in (Guinea, Mozambique, and Tajikistan). PRSP monitoring PRSP the PRS process to the specifications of monitoring indicators exceed implementation has not public at large stimulates country capacity in the majority of cases. IDA been fostered in most demand for monitoring performance monitoring may introduce an of the case study and evaluation. additional series of indicators that countries countries. In general, have to track. Early difficulties in collecting and governments have not reporting data suggest that use of this informa- effectively promulgated the PRSP to the public. tion for decisionmaking is a much longer-term One exception is Tanzania, where a clear, lay- prospect. In Nicaragua, a centralized tracking person's version of the PRSP was published and system has been established and some indica- later translated into Swahili. It is notable that tors chosen, but the system is incomplete. The the government outsourced this work to an first APR in Ethiopia lays out a comprehensive NGO. plan for an M&E system, but progress in developing key indicators and clarifying how Preliminary Outcomes they will be monitored and how the M&E The short history of the PRS process limits system will work amid ongoing decentraliza- measurement of outcomes and attribution of tion reforms is still needed. A working group successes and failures. It is important to keep has been set up in Guinea to produce a list of in mind that this program was launched less monitoring indicators, but the group has not than five years ago. Nonetheless, OED yet put together a complete proposal. In attempted to measure preliminary outcomes in Tajikistan, a PRSP implementation unit has three areas: (1) achievement of quantitative been established, and a start has been made in PRSP targets; (2) improvement in policy and defining indicators. institutional frameworks; and (3) aggregate Efforts to build capacity have been highly changes in aid flows. supply-driven. A challenge to building capacity for monitoring and evaluation is ensuring that Achievement of PRSP targets. Nine of the 12 there is demand for data. Centralized, supply- countries with APRs report on quantitative driven monitoring systems may produce data targets for poverty reduction.10 Although each that does not feed back into policy or perform- country's progress is unique, common trends ance evaluation. For example, in Cambodia the emerge. Inputs and outputs have increased interest of policymakers is difficult to gauge, across sectors, as evidenced by sharp increases and the emphasis appears to have been on data in gross primary school enrollment rates, school gathering alone. Even in the more mature case construction, and the number of teachers; 3 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 : A N I N D E P E N D E N T E VA L U AT I O N improvement in the supply of drugs to health in a few countries, for a few years, or for a single centers and the number of vaccinations; and the indicator. expansion of water facilities. Outcomes are These limited data show modest improve- slower to respond. Maternal and infant mortality ments in literacy rates (including those of rates and prevalence of malaria and HIV/AIDs women), the control of TB, ozone depletion, have been stagnant in most countries with and the growth of new technologies. At the available data. The poor quality of service same time, infant mortality rates appear to have delivery in both education and health has been stagnated, a finding echoed in case studies and flagged. Information on poverty trends is in the recent World Bank and IMF Global inconclusive because of the limited availability Monitoring Report 2004.11 There are clearly of data. Where data are available, poverty rates problems with attributing progress over the have remained sluggish--in part, because of period to the introduction of the PRSP Initia- limited progress in tive because of the lack of a counterfactual; in Information on poverty agricultural growth and addition, the PRSP was introduced at different trends is inconclusive private sector develop- times in each of these countries. because of the limited ment. To supplement the Improvement in policy and institutional frameworks. availability of data. data from APRs on To assess whether the process of formulating achievement of PRSP and implementing PRSPs is associated with targets, OED analyzed progress in MDG indica- improvements in the policy and institutional tors in the same set of 12 PRSP countries that frameworks of low-income countries, changes have issued APRs. As table 3.1 demonstrates, in the World Bank's Country Policy and Institu- relevant data are available for fewer than a third tional Assessment (CPIA) ratings were consid- of the 49 specified U.N. indicators. In critical ered. The average CPIA rating for low-income areas such as extreme poverty, maternal health, countries improved from 1999 to 2003.12 This and combating HIV, very little can be said on finding accords with indicators of various progress to date, because data are only available dimensions of policy and institutional perform- P r o g r e s s i n M D G s f o r 1 2 P R S P C o u n t r i e s T a b l e 3 . 1 w i t h A n n u a l P r o g r e s s R e p o r t s Number of Data available for...... indicators for Number Average Average Improvement Progress on Millenium monitoring of no. no. Between Development Indicators progress indicators countries years 1999 & 2003? Goal 1: Eradicate extreme hunger and poverty 6 1 3 2 Yes Goal 2: Achieve universal primary education 4 1 12 5 Yes Goal 3: Promote gender equality and empower women 4 3 8 4 Yes (2 of 3) Goal 4: Reduce child mortality 3 3 12 3 No (3 of 3) Goal 5: Improve maternal health 2 0 0 0 Unknown Goal 6: Combat HIV/AIDS, malaria and other diseases 10 1 8 3 Yes Goal 7: Ensure environmental sustainability 8 1 12 3 Yes Goal 8: Develop a global partnership for development 17 4 8 4 Yes Source: U.N. Statistics Division, http://millenniumindicators.un.org/unsd/mi/mi.asp and World Bank World Development Indicators. Data represent country averages with an equalling of the number of countries represented. Only indicators with data for more than one year on more than one country were included. 3 4 P R E L I M I N A R Y R E S U LT S ance compiled outside the World Bank (OED the differences between the performance of 2004). Improvements were achieved in each of 23 very early and early PRSP countries with the four areas that are assessed in the CPIA: over a year of implementation experience and economic management, structural policies, the 12 late PRSP countries, with PRSPs policies for social inclusion/equity, and public completed only in the last year. The PRSP sector management and institutions.13 countries with the longest implementation PRSPs have been completed in low-income time demonstrated a slight worsening in the countries with better policies and institutions quality of structural policies and economic (figure 3.3). This is demonstrated by a higher management since 1999, although the differ- 1999 CPIA average for the 35 countries that have ences are not statistically significant. Their completed a PRSP (3.36) than for the 31 non- overall CPIA average, however, has shown an PRSP countries (2.78). Both PRSP and non-PRSP improvement. For more recent PRSP countries improved between 1999 and 2003. countries, the data show improvements in The latter group lessened the performance gap each of the four CPIA dimensions between in structural policies, and PRSP countries 1999 and 2003. exhibited relative gains in public sector manage- ment and institutions (figure 3.3). The small Aid Flows to PRSP Countries. National stakeholders improvement in structural policies in PRSP in PRSP countries expect that formulating PRSPs countries is consistent with the Chapter 2 will attract more development aid. This expecta- finding that the benefits from initial PRSPs were tion is consistent with recent signals from the more from integrating existing sectoral strate- international community that achieving PRSP gies in a comprehensive strategy than from targets, and more broadly the MDGs, will require large advances in the quality of individual sector a substantial increase in external resources, plans. In 2003, average CPIA ratings were 3.47 along with continued policy and institutional for PRSP and 2.91 for non-PRSP countries. reform (World Bank 2003b). Official develop- Improvements in the quality of policies ment assistance (ODA) to low-income countries and institutions require time. OED looked at has been increasing from a 12-year low in 1997.14 Q u a l i t y o f P o l i c i e s a n d I n s t i t u t i o n s i n F i g u r e 3 . 3 P R S P a n d N o n - P R S P C o u n t r i e s Absolute Levels of CPIA - 1999 and 2003 Percentage Change in CPIA - 1999 and 2003 4.0 7% 3.5 6% 3.0 5% 2.5 4% 2.0 3% 1.5 1.0 2% 0.5 1% 0.0 0% Overall Rating Public Sector Policies for Economic Overall Structural Management Social Management Rating Policies 1999 2003 and Institutions Inclusion Equity 35 PRSP 31 Non-PRSP 3 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 : A N I N D E P E N D E N T E VA L U AT I O N Given that donors' aid disbursements require This is consistent with relatively higher policy some time to adjust, it is most appropriate to performance in the eight very early PRSP look for PRSP-related changes to aid flows in the countries. In terms of increases in ODA following countries that have had the longest amount of the PRSP, the limited information shows that There is no evidence at PRSP implementation ODA has increased for these very early PRSPs time. Figure 3.4 shows countries at roughly the same rate as for other this time that aid to PRSP that the ODA average for IDA countries. Thus there is no evidence at this countries has been each of the eight very time that aid to PRSP countries has been increas- increasing more than aid early countries, with ing more than aid to other IDA countries. More to other IDA countries. PRSPs completed by donors are granting aid to these eight countries 2001, is much higher in the form of budget support, however--up than for later PRSP and non-PRSP countries.15 from four donors in 1999 to 10 in 2002. F i g u r e 3 . 4 A v e r a g e N e t O D A F l o w s p e r C o u n t r y Millions of US$ 900 800 700 600 500 400 300 200 100 0 2000 2001 2002 8 Very Early 35 PRSP 31 Non-PRSP 3 6 4 Alignment of the Bank's Assistance T he PRS process is intended to be a framework for country-led man- agement of external assistance, including that of the Bank. External part- ners are expected to recognize country priorities and align assistance to support PRSP priorities and the PRS process more generally. This chapter examines the extent of change in the Bank's assistance strategies and finan- cial and non-financial assistance, both recent and planned, since the introduction of the PRSP Initiative. An internal guidance note from January 2000 PRSP framework. A new development serves to guide staff regarding the link between assistance instrument, the Poverty Reduction the PRSP framework and Bank operational Support Credit (PRSC), would support a programs (World Bank 2000a). The Bank's country's policy and institutional reform Country Assistance Strategies (CASs), program to help implement its PRSP. The which set out the strategy for alignment of IDA Bank's analytical work program was assistance, would take as their "vision" the expected to change significantly to underpin country's own description of its development the formulation of poverty reduction strategies goals and its strategy for achieving them, as set and to include a more systematic use of out in the PRSP. Building on the JSA and other diagnostic tools such as poverty assessments, analytical work, the CAS would detail the public expenditure reviews, and fiduciary and professional assessment of Bank staff of the safeguard reports. The Bank would train government's program and the Bank Group's country teams in key PRSP concepts and business strategy for supporting the PRSP. The relationships, offering in-depth follow-up sectoral composition of the CAS would be training for the development of sectoral selectively designed to support and comple- agendas related to the PRSP. ment country and partner efforts to reduce This chapter first assesses whether the poverty, taking into account the Bank's Bank's CASs are formulated with consideration comparative advantage. of PRS processes. Alignment of the Bank's The Bank's financial assistance was financial and non-financial assistance is then expected to align to the PRSP. Individual IDA discussed. Evidence is drawn from the case credits would indicate the links to the CAS and studies,1 analysis of World Bank lending and 3 7 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 : A N I N D E P E N D E N T E VA L U AT I O N non-lending activities for low-income PRSP four-year periods before and after the Initiative countries,2 and a review of 25 post-PRSP CASs. was launched (figure 4.1).4 IDA lending has decreased for the group of 31 non-PRSP Alignment of CAS Formulation countries that have not yet produced PRSPs. Through 2003, the Bank prepared CASs for 25 Annual lending volumes have been less volatile countries following PRSP completion.3 All CASs to the PRSP countries than to the non-PRSP describe the stakeholder consultations that countries over the FY96­CY03 period, suggest- contributed to the formulation of the CAS. ing that the PRSP provides a basis for more Sixteen CASs do not describe how the consul- predictable flows. tations for the PRSP affected the preparation of As shown in figure 4.1, the increase in total the CAS. In a few cases, extensive CAS consul- Bank lending to PRSP countries is the result of tations are described without mention of those an increase in investment lending. Adjustment done for the PRSP. Of the nine that do describe lending volumes over the two time periods how they built on preceding PRSP consulta- have remained flat in PRSP countries; the share tions and consensus building, CAS consulta- of adjustment lending out of total lending to the tions were designed to dovetail with the PRSP 35 PRSP countries fell from 30 percent to 24 consultations in several countries. In Benin, percent between the two time periods. Poverty Ethiopia, and Niger, the Bank proactively Reduction Support Credits (PRSCs), created connected CAS consultations directly with the specifically to support policy and institutional nationwide consultations around PRSP finaliza- reform in PRSP countries,5 account for an tion. In Guyana, the CAS explicitly draws on the increasing share of adjustment lending and content of PRSP consultations, and the comprise 35 percent of adjustment lending in planning for the PRSC was carried out during FY00­CY03. This programmatic lending instru- PRSP planning. The timing of the CAS is an ment potentially provides a better match with important factor in these cases, with a tighter medium-term PRSP programs because of its time frame allowing consultations for the PRSP longer intended term and more flexible to directly feed into the modality of financial assistance. The 25 available Annual lending volumes Bank's strategy. post-PRSP CASs indicate that adjustment have been less volatile to In a few cases, post- lending will rise to an average of 31 percent of PRSP CASs include country lending programs in PRSC countries, the PRSP countries than lending programs one with PRSCs accounting for 90 percent of total to the non-PRSP year longer than the adjustment lending programmed. The invest- countries, suggesting that three-year framework of ment lending record shows that there has pre-PRSP CASs. This has been a sharp increase in PRSP countries in the the PRSP provides a basis allowed a better match use of Adaptable Program Loans (APLs), which for more predictable with the two-to-five- have driven a 31 percent increase in total invest- flows. year horizon of PRSPs. ment lending between time periods (FY96­99; In measuring results, FY00­CY03). APLs accounted for nearly a the Bank has piloted results-based CASs (for quarter of total investment lending in example, the CAS for Mozambique) that FY00­CY03. enhance outcome measurement in alignment with the PRSP. PRSCs in the case study countries6 have helped to promote ownership. Budget support is seen Alignment of the Bank's Financial as a vote of confidence in the government's Assistance program, as outlined in the PRSP, because Lending has increased for PRSP countries, relative to external resources are given directly to the non-PRSP countries. Aggregate IDA lending had recipients to support their own programs. increased for the group of 35 countries with PRSCs have also improved donor coordina- PRSPs as of end-2003, when compared with the tion--the Bank has been seen as "joining the 3 8 A L I G N M E N T O F T H E B A N K ' S A S S I S TA N C E W o r l d B a n k L e n d i n g H a s S h i f t e d t o P R S P F i g u r e 4 . 1 C o u n t r i e s IDA Commitments to PRSP and Non-PRSP Countries $ billions 18 16 14 12 10 8 6 4 2 0 PRSP Non-PRSP PRSP Non-PRSP FY96­FY99 FY00­CY03 Investment Adjustment team" of donors providing budget support.7 In planning policies and Both the Bank and PRSP Tanzania and Mozambique, donors welcomed budgets. Since both the countries have much to the Bank's incorporation of PRSC conditionality Bank and PRSP countries learn about the poverty in the joint donor monitoring frameworks for have much to learn delivery of budget support; this increased about the poverty and and social impact of transparency and coordination, and decreased social impact of policy policy reform options. the government's transactions costs by reform options, the reducing the numerous bilateral discussions of Bank needs to maintain a balance between policy reforms. As shown in Ethiopia, however, financial management concerns and the need implementing greater donor coordination in for assessments of policy and institutional policy conditionality will require more work by constraints to poverty reduction and growth. the Bank, as there were limitations, given the sheer size of the Bank PRSC team. PRSCs as multisector policy lending. PRSCs are meant to encompass the entirety of the Bank's support Focus on financial management. Since budget for policy reform. The PRSC has replaced Sector support requires adequate fiduciary safeguards, Adjustment Credits and is less sectorally concen- a key challenge in many countries is balancing trated than previous adjustment lending in the 35 the move to budget support with the realities of PRSP countries (figure 4.2). Four-fifths of the weak PEM systems. Fiduciary weaknesses are commitments in the 14 PRSCs approved through prevalent in PRSP countries, 8 underscoring the CY03 cover public sector governance, financial and importance of the Bank's fiduciary assessments private sector development, human development, as an adequate basis for programmatic lending. and social protection, with the latter two areas The planned increase in PRSCs will require the showing large increases compared with their share Bank to keep doing considerable fiduciary in earlier adjustment lending. Thus the introduc- economic and sector work (ESW) in PRSP tion of the PRSC has been associated with a shift countries. But because fiduciary ESW is mainly toward more social sector emphasis within adjust- concerned with the controls and procedures for ment lending. expenditures, it does not directly support PRSP Bank task managers report that preparing and formulation by filling the analytical gap in implementing PRSCs has led to more cross- 3 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 : A N I N D E P E N D E N T E VA L U AT I O N P R S C s H a v e L e s s S e c t o r a l C o n c e n t r a t i o n F i g u r e 4 . 2 T h a n P r e v i o u s A d j u s t m e n t L e n d i n g t o P R S P C o u n t r i e s A. Thematic Focus of Pre-PRSP Adjustment Lending Trade and integration 5% Social protection Urban development 1% Social development/ and risk 5% gender Economic management 1% 17% Rural development 1% Environment and Rule of law natural resources 2% 3% Public sector governance 23% Finance and private sector Human development development 40% 2% B. Thematic Focus of PRSCs Economic management 0% Urban development Environment and 4% Trade and integration natural resources 4% 6% Finance and Social protection private sector development and risk 19% 21% Rural development Human development 7% 13% Public sector governance 26% sectoral interaction among Bank staff compared previous adjustment lending. But some sector with their involvement with Policy Framework staff underscored the risk that the PRSC could Papers and previous eventually water down sectoral inputs by the Bank task managers national strategies. Bank. Bank task managers also noted that existing report that preparing and Macroeconomists inter- institutional constraints to working cross- implementing PRSCs has act more with sector sectorally are manifest in interactions around led to more cross-sectoral staff, and sector staff PRSCs. They emphasized that the PRSP Initiative report more participa- has significantly increased the role of the Poverty interaction among Bank tion in PRSC design and Reduction and Economic Management (PREM) staff. implementation than in Network, leading to tensions with other sectors. 4 0 A L I G N M E N T O F T H E B A N K ' S A S S I S TA N C E Within current budget processes in the Bank, this overlaps in these areas. Comparison of aggregate has increased competition in task management No CAS program pro- Bank lending programs and control over resources. Many focus group poses assistance to a participants noted that internal incentives for sector that falls outside from CASs before PRSPs budget processes may play an undue role in those discussed in the with those after shows influencing decisions on Bank lending around PRSP. Comparing sectoral that there have been no PRSCs. expenditure shares, while major shifts in sectoral PRSPs allocate 43 percent Lending alignment in post-PRSP Country Assistance of total costed programs allocations. Strategies. The 25 post-PRSP CASs provide a to the social sectors, post- basis for assessing how the content of Bank PRSP CASs earmark only 25 percent of IDA commit- assistance strategies is consistent with that of ments for these sectors.11 The post-PRSP CAS figure PRSPs. In each country, the pair of CAS represents a slight decline from the 27 percent documents before and after the PRSP, as well as average for pre-PRSP CASs. Public sector reform, in the PRSP document itself, have been reviewed contrast, figures much more prominently for the to identify changes in Bank CAS programs and Bank than for countries' PRSP programs. While in the relationship to priorities established in shares are not expected to match, the differences the PRSP.9 are infrequently discussed in CASs. The content of the Bank's CASs formulated subsequent to PRSPs Changes in Bank lending in relation to PRSPs. overlaps with the content of PRSPs. But since PRSP Comparison of aggregate Bank lending programs programs are broad and not well prioritized, this from CASs before PRSPs with those after shows that overlap has not entailed major changes in Bank there have been no major shifts in sectoral alloca- programs. tions. Planned lending to public sector manage- These comparisons capture only some ment has increased somewhat, while lending to aspects of content alignment of the Bank's productive sectors, and to a lesser extent social lending because they do not account for sectors, has declined (figure 4.3).10 All CASs ongoing Bank programs. The differences in describe the broad pillars of the PRSP strategy, and priorities can also be explained by the Bank mostemphasizethattheBank'sproposedprogram exercising selectivity in its support, relative to S e c t o r a l A l l o c a t i o n i n B a n k L e n d i n g F i g u r e 4 . 3 a n d P R S P s CASs before and after PRSPs CAS compared with PRSP Percentage of Commitments Percentage of Commitments/Program 50 50 40 40 30 30 20 20 10 10 0 0 Social Growth Public Urban Rural Other Social Growth Public Urban Rural Other Sector Sector Pre-PRSP CASs PRSPs Post-PRSP CASs 4 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 : A N I N D E P E N D E N T E VA L U AT I O N Most CASs do not other donors' programs. work. This type of ESW includes five main explicitly ground Bank Some CASs do explain products, and completion of all five is required programs within the the Bank's absence from at least every five years.12 Country Financial certain sectors by citing Accountability Assessments (CFAAs) were the larger context of other the presence of other most commonly done products, as shown in external assistance. donors, but most do not figure 4.4. The intent of the focus on core explicitly ground Bank diagnostic ESW is to support PRSPs by covering programs within the larger context of other the main economic developments and policy external assistance in particular sectors. agenda designed to achieve sustainable growth and poverty reduction. Alignment of the Bank's Analytical Work The Bank's analytical work has added value to The Bank increased economic and sector work country planning. The Bank's ESW has been (ESW) in countries preparing PRSPs. In the 35 useful to the countries preparing PRSPs, countries that completed PRSPs between July especially in its support of their poverty 1999 and end-2003, the Bank did about one- diagnosis. Poverty Assessments (PAs) were a third more ESW than it had in the preceding major basis of the poverty profile in most case four years, roughly three more ESW reports per study countries. A noteworthy exception is country. In other IDA-eligible countries, the Tanzania, where the last PA was conducted in Bank did 11 percent less ESW--during the 1997 and authorities noted that the PRSP was same time period, about one less report per significantly handicapped in the absence of country. Since PRSP countries are, in the more recent data. Country authorities also aggregate, better performers on CPIA, both the underscored that Bank-supported PERs were Bank's ESW and its lending have shifted to particularly helpful in planning individual line countries with stronger policy frameworks. ministry programs in the PRSP and advancing Most of the increase in ESW in PRSP links to the budget. In general, they found countries took the form of core diagnostic CFAAs and Country Procurement Assessment C o r e E S W I n c r e a s e d i n P R S P C o u n t r i e s , F i g u r e 4 . 4 a n d T h e y R e c e i v e M o r e C o r e D i a g n o s t i c E S W T h a n N o n - P R S P C o u n t r i e s Shift in Types of ESW Prepared in PRSP Percentage of Countries with Core ESW Countries Conducted in Past Five Years No. of ESW Tasks Percentage of Countries 140 100 120 90 80 100 70 80 60 50 60 40 40 30 20 20 10 0 0 CFAA CPAR PER PA CEM/DPR 1996­99 2000­04 Core ESW Types Core ESW Other Diagnostic PRSP (35) Non-PRSP (31) Note: CFAA= Country Financial Accountability Assessment; CPAR= Country Procurement Assessment Review; PER= Public Expenditure Review; PA= Poverty Assessment; CEM/DPR= Coun- try Economic Memorandum/Development Policy Review. 4 2 A L I G N M E N T O F T H E B A N K ' S A S S I S TA N C E Reviews (CPARs) less directly useful, perhaps poverty reduction The Bank's ESW has been because of their lack of familiarity with these strategy. A noteworthy useful to the countries products, given their later introduction by the exception to this preparing PRSPs, Bank, and their stronger focus on implementa- practice can be seen in especially in its support of tion dimensions of the strategy. The World Nicaragua, where the Bank Quality Assurance Group's reviews Bank undertook an their poverty diagnosis. corroborate these favorable impressions of unprogrammed study of ESW, with 90 percent of recent core diagnostic growth in the agriculture sector in response to ESW rated satisfactory. government priorities. The Bank has conducted fewer Poverty and But critical gaps remain in linking policies and Social Impact Assessments (PSIAs) than programs to poverty impact. The Bank's reduction envisioned, even though the Bank emphasizes in sector-specific analytical work to accommo- the importance of analysis of the impacts of date core diagnostics has left key analytical gaps. policy reform on the well-being of the poor. Only Survey respondents ranked the timeliness and 18 PSIAs had been completed in PRSP countries relevance of analytical and advisory inputs by the end of 2003, roughly one for every two lowest among World Bank performance issues PRSPs completed. The limited evidence of distri- (see figure 4.5).13 In Mozambique, for example, butional analysis to The Bank has conducted adding a CFAA and CPAR to the analytical work inform PRSP design was of the Bank was relevant to the governance noted in the March 2002 fewer Poverty and Social objectives of the PRSP, but the absence of the Joint Review and reiter- Impact Assessments than activities dropped and delayed to accommodate ated in the most recent planned. these projects left gaps in the Bank's support for Progress in Implement- building on important pieces of the PRSP. In ing the PRSP report (World Bank and IMF). Of Cambodia, the Bank engaged in assisting the the 25 post-PRSP CASs available at the end of development of the country's agricultural and 2003, only three included a formal PSIA. The rural development strategies, though its inputs Bank increased its budget to support PSIA work were not completed in time for inclusion in from July 2002 to June 2003, however, and these critical components of the country's completed a PSIA Toolkit in April 2003. There has P R S P A l i g n m e n t o f t h e B a n k ' s A n a l y t i c a l W o r k W a s N o t R a t e d a s H i g h l y a s F i g u r e 4 . 5 A l i g n m e n t o f F i n a n c i a l A s s i s t a n c e o r S t r a t e g y Alignment of country assistance Financial assistance supports PRSP Analytical and advisory activities strategy priorities provide relevant and timely inputs to PRSP process 29% 30% 37% 48% 61% 58% 10% 12% 15% Agree Disagree Neutral 4 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 : A N I N D E P E N D E N T E VA L U AT I O N The Bank needs to also been increased formal ESW pieces programmed in each one. uptake from the Poverty No CAS contains ESW that lies outside of broad intensify its work on Reduction Strategy Trust PRSP priorities, but a large amount of non- policy and institutional Fund for PSIA activi- lending resources are devoted to core fiduciary constraints to poverty ties.14 ESW, which are required to be updated every reduction and growth. Looking at the non- five years. They are the dominant focus of many lending programs in CAS non-lending programs, including those of post-PRSP CASs, the alignment with PRSPs is small countries with limited ESW budgets. In difficult to assess. Country non-lending light of critical gaps in linking policies and programs are either too limited to cover many programs to poverty impact, and the current areas, or too diverse and large to determine levels of resources devoted to analytical work, strategic focus. The Nepal, Sri Lanka, and the Bank needs to intensify its work on policy Tajikistan CASs are examples of highly diversi- and institutional constraints to poverty fied non-lending programs, with as many as 20 reduction and growth. 4 4 5 Findings and Recommendations T he findings of this Review reflect progress in the work of the PRS Ini- tiative. In some cases, the findings validate staff understandings--for example, as reflected in joint Bank/Fund Progress In Implementation Reports. They also provide benchmarks and suggest gaps to be filled to assess the Initiative's progress in the future. The recommendations are geared to- ward enhancing the effectiveness of the Bank's support of the PRS Initiative, either directly, through actions by the Bank, or through other actions that will improve the PRS process and thus enhance the Bank's work through its sup- port of the PRSP. The findings and recommendations are presented in three areas of particular relevance to the Bank's work in low-income countries: the PRS process, PRSP content and analytical underpinnings, and partnership and alignment. The Initiative's contribution so far has varied countries. Most PRS strategies deal primarily with widely across countries, depending mainly on the composition of public expenditure, while the initial conditions in each country's public treating growth and other poverty-reducing sector planning and monitoring capacity, govern- actions lightly, or not at all. External partners ment-partner relationships, and relations among have supported the PRSP process during donors. The PRSP has added the most value in formulation, but have not yet systematically countries where government leadership and aid adapted their assistance programs to support management processes were already strong. It country priorities. has had less effect in countries with weak public The PRS Initiative is an improvement over sector capacity or with donor-dominated aid the Policy Framework Papers of the 1990s, but relationships. Because the Initiative includes a remains a work in progress and has not yet uniform requirement--completion of an accept- fulfilled its full potential to enhance poverty able document--it has not been sufficiently reduction efforts in low-income countries. The tailored to match the full range of capabilities World Bank should continue to support it, and public policy priorities found in low-income while making changes in three main directions: 4 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 : A N I N D E P E N D E N T E VA L U AT I O N · Reduce or eliminate uniform requirements partnership, and a long-term perspective. and foster better customization. Countries typically report on steps they have · Encourage PRSPs to explore a wider range of taken to put these principles into practice. But policy options, including those aimed at en- what results should a country expect from their hancing growth. successful implementation? There are no · Help define clearer partnership frameworks intermediate indicators for what the principles around PRSPs, with accountabilities for both are expected to achieve, which makes it countries and partners. difficult to assess progress toward the Initia- tive's objectives. This lack of clarity has led to PRS Process different and sometimes incompatible expecta- The Initiative demands countries make tions among stakeholders. The experience with ambitious and complex changes that will take broad-based participation illustrates the time to deliver poverty reduction results. Costs problem. Stakeholders do agree that the have also been significant, and the intended extensive consultations conducted in most long-term nature of the PRS process is not yet PRSP countries during strategy formulation clearly established with all national stakehold- have brought new actors into the development ers. Risks to sustainability are thus significant. dialogue and increased transparency. On this Success depends on countries seeing criterion, participation has been a success. But demonstrated benefits in their own processes, civil society organizations consider that this as well as from higher-quality external openness has had only limited impact on the assistance. design of PRSPs or on domestic policy, with the exception of greater attention to governance More tailored implementation and clarity about concerns in some countries. Moreover, partici- expected country-specific process achievements. patory activities have sometimes waned once There is an inherent tension in designing a the PRSP was formulated, suggesting that BWI-driven initiative involving conditionality governments were concerned more about that is simultaneously meant to foster a fulfilling donor requirements than about country-driven process. The policy papers were achieving systemic change. Along these circumspect in important areas to promote dimensions, participation has not achieved all ownership. This has led to two key problems its aims. during implementation. First, there is no mechanism or guidance to adapt the Initiative's Recommendation 1: The Bank should help foster processes and requirements to differing better customization of the Initiative to country country conditions, especially to the weak circumstances and more focus on improving long- public sector capacity found in many low- term processes. income countries. In practice, countries have understandably focused on completing · Greater emphasis should be placed on im- documents that give them access to resources. proving country processes for planning, im- This attention to requirements has often been plementing, and monitoring public actions at the expense of adaptation of the PRS process geared toward poverty reduction, and less on to unique country circumstances. The BWIs' completion of documents. Toward this goal, the initially ambitious timetable for completion of Bank should help countries identify what di- first-round PRSPs in HIPC and PRGF countries mensions of domestic processes present major reinforced the tendency to generate standard- constraints to poverty reduction and establish ized strategies in early PRSPs. milestones for specific improvements in these Second, the PRS process is expected to areas, including capacity building. Clarification follow five principles--country ownership, of these milestones will help to equilibrate results orientation, comprehensiveness, stakeholder expectations and promote a longer- 4 6 F I N D I N G S A N D R E C O M M E N D AT I O N S term view and results orientation. Country au- strengthen partnership through greater involve- thorities should establish indicators and re- ment of other stakeholders. port them in PRSPs, APRs, and second-round PRSPs. Recommendation 2: The Bank should provide transpar- · The Bank should not simply permit--but ent and effective feedback to countries on their PRS should actively promote--tailoring of domes- processes and develop a review procedure that is tic processes to country conditions. It should more supportive of ownership and more effectively ensure that the timing of progress reporting de- linked to decisions about the Bank's program. rives from local processes and that their con- tent links to domestic decisionmaking. The · Bank management should develop a proce- Bank (and IMF) should demonstrate flexibility dure for Board review of the PRSP that is more in the Initiative's handling of initial conditions, transparently supportive of ownership and including cases where constraints suggest build- more effectively linked to decisions about the ing a minimum capacity in key areas before Bank's program. This would both reduce the completing a PRSP. perception of "Washington signing off" and promote ownership. The assessment of Bank Board review of the PRSP, and role of the JSA. Bank and Fund staff of the soundness of the PRSP management's process for presenting a PRSP could be directly incorporated in the CAS and to the Board undermines ownership. discussed directly in the context of the pro- Stakeholders perceive this practice as posed program. This could strengthen the op- "Washington signing-off " on a supposedly erational link between the assessment and the country-owned strategy. Executive directors CAS and could allow the CAS to demonstrate already have several mechanisms for authoriz- more transparently how the Bank's planned as- ing Bank support to a country--notably their sistance is derived from the PRSP. endorsement of CASs and their approval of · The Bank should provide feedback to the coun- individual lending operations. In this context, try on its PRSP in a form that is candid, trans- their review of the PRSP appears redundant, parent, analytically rigorous, and comprehensive. while it also attenuates ownership in the eyes The Bank should also provide feedback in a of most stakeholders. manner that strengthens partnership by in- The JSA was designed to provide the Bank and volving other stakeholders. Fund Boards with an assessment of the · These objectives should be met either through soundness of the PRSP as a basis for support. It a major redesign of the JSA instrument and was also designed to provide feedback on how a process or through discontinuation of the JSA country's PRSP could be improved over time. The and reliance on more interactive means, such JSAs have shortcomings that undermine achieve- as stakeholder workshops, management notes, ment of these goals: mixed analytical quality and or other such devices. comprehensiveness, inadequate focus on the quality of process orientation, and limited PRSP Content and Analytical awareness of their findings and recommenda- Underpinnings tions among stakeholders. JSAs involve a degree PRSPs, in comparison with PFPs, have stronger of self-assessment, because the Bank has usually country ownership, improved poverty focus, assisted with its formulation. Rather than being a and give more attention to outcomes. But their comprehensive review that underpins a discus- analysis of the links between policies/programs sion of selectivity and comparative advantage, the and poverty reduction has gaps, and their JSA is influenced by anticipated lending activities. scope is narrow. For these reasons, the quality of feedback on PRSPs should be more candid, transparent, Improved poverty focus. The Initiative has analytically rigorous, and comprehensive and improved the poverty focus of development 4 7 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 : A N I N D E P E N D E N T E VA L U AT I O N strategies. The importance of basing PRSPs on these long-standing analytical gaps. But the PRS sound poverty diagnoses has highlighted the process could be used more effectively to poor data on poverty in many countries and, in identify the key analytical gaps and to develop some cases, has resulted in marked improve- a research agenda to address them. ments in data quality. The PRS process has played an important role in emphasizing multidimen- Inadequate attention to growth policies. PRSPs to sional income and non-income poverty indica- date have not considered the full range of tors. The process of gathering sector programs policy actions required for growth and poverty under one strategic umbrella has helped policy- reduction. They focus largely on public makers appreciate the links between sector expenditures and have not fully explored or strategies and poverty reduction, and sector specified non-expenditure-related policies or expenditure plans are being challenged to meet actions for enhancing growth and poverty poverty reduction criteria. reduction. Among public expenditures, social sector spending has received more attention in Added attention to monitoring, but results focus relation to the poverty reduction potential of remains weak. The PRS process has drawn more spending in other areas. attention to a results orientation. In practice, however, the development of country-specific Recommendation 3: The Bank, in concert with other goals and indicators has a long way to go. partners, should assist countries to strengthen Monitoring systems are just emerging and are analysis of the poverty impact of policies and often not institutionalized, with requirements programs, and give more attention to growth. that are frequently donor-driven and exceed country capacity. Many PRSP targets are unreal- · The World Bank, in concert with other partners, istic, given the initial conditions of the country should help countries build the capacity to and its recent track record; some are even address key analytical gaps about the poverty more ambitious than the global MDGs. PRSPs impact of policies and programs. Analysis must be more effective in enabling countries to should address areas such as sources of growth, select and monitor relevant country-specific the quality of non­social sector strategies, and indicators that will provide a continued basis the integration of the macro framework and for informed decisionmaking. structural and social reforms. · Monitoring inputs, outputs, and outcomes of Weak analytical base for consideration of tradeoffs ongoing activities is critical to a better under- and linkages. The PRS process has drawn standing of what works and what does not. attention to the lack of knowledge of the The Bank should assist in defining indicators linkages between policies/programs and that are specific to country priorities, realistic, poverty-related outcomes, and it underscores and within current country capacity to moni- the need for more analytical work on these tor and use. The monitoring and evaluation issues. This knowledge gap has not been scope can grow with capacity over time. reduced greatly during the formulation process--there is a general lack of evidence of Partnership and Alignment robust consideration of policy options and tradeoffs. Linkages of individual sector strate- External partners need to support and align with the gies to poverty reduction are uneven, and most PRS process. Most partners have accepted that PRSPs do not adequately integrate the macro the PRSP has the potential to be an overarching framework and the sector strategies through a framework for aid management, and many have rigorous consideration of the sources of provided well-coordinated assistance to govern- growth, the social impact of macro policies, or ments for their formulation. PRSPs generally various macro-micro linkages. It would be provided a constructive framework for transpar- unreasonable to expect first-round PRSPs to fill ent policy dialogue among external partners and 4 8 F I N D I N G S A N D R E C O M M E N D AT I O N S with government. Given the PRSPs' broad scope willingness to adjust to country-specific timeta- and lack of prioritization, alignment would not bles and processes. For example, the CAS necessarily entail changes in donors' programs. should be completed after the PRSP and adopt Neither donors nor the Bank have defined its time frame. specifically whether or how they should change · The Bank should also define which process the content of their programs to reflect PRSPs. areas it will assist the country in addressing and Changes in donors' processes are more readily the nature of its assistance in a transparent evident. In the most positive cases, the PRS manner, including coordination with other process bolstered an already robust government stakeholders. role in aid coordination and increased participa- · The Bank should help countries improve the tion in, and greater coordination of, budget prioritization and costing of PRSP programs support instruments and attempts to streamline and demonstrate the alignment of its assis- performance monitoring. But overall, there still tance with these improved PRSPs, taking into is little evidence that donors have coordinated account other partners' activities and the and selected the majority of their programs in Bank's comparative advantage. support of PRSP priorities. Where the govern- ment-donor dialogue was previously weak or Looking Forward donors continued to drive the agenda, PRSPs by The above findings and recommendations of themselves have not noticeably promoted this Review highlight the need for the next donor coordination or increased government round of PRSPs to assess improvements in management of external assistance. processes, in prioritization, in demonstrating The Bank has been largely responsible for ownership, and in defining a clear partnership leveraging the PRSP as a coordinating framework framework with accountabilities for both for external assistance. The Bank has improved its countries and partners. This is a critical stage in coordination and relations with other external the Initiative: some countries with mature partners, especially through its local representa- PRSPs have reached the end of their first PRSP tion. But this improvement is not necessarily attrib- cycle and are evaluating their experience and utable to the PRSP Initiative alone, given the Bank's adjusting their strategies (as in Burkina Faso earlier and continued attention to decentralization and Tanzania). This is also important for and aid coordination. The Bank could still do more embedding the Initiative in low-income to apply the partnership principle by encouraging countries that have yet to complete an initial government-led aid management and selectivity PRSP. For the Bank, continued focus on evalua- andintegrationofitsownassistancewiththeefforts tion and review is needed to enhance support of other partners. for the Initiative. The piloted instruments of CAS completion reports and results-based Recommendation 4: The Bank, in concert with other CASs promise to be useful tools that will help donors, should assist the country in defining a to assess the Bank's alignment and support for partnership framework under the PRSP, with more country-level PRS processes. Independent explicit reference to prioritization and the expected evaluation also has a role to play going forward, role of external partners. including assessment of the Bank's develop- ment effectiveness at the country level in · The Bank should be a leader in supporting OED's Country Assistance Evaluations and CAS country-led aid management by demonstrating completion report reviews. 4 9 ANNEXES ANNEX A: SUMMARY OF THE BWI POLICY PAPERS ON THE PRS INITIATIVE Objectives Intermediate outcomes Process / outputs 1) Strengthening the Link Between Debt Relief and Poverty Reduction, August 1999 (World Bank and IMF 1999a) Enhance the link between HIPC and Outcome indicators An outcome-based strategy that is mon- poverty reduction 1) Provide useful insights into the relative ef- itored frequently and entails 3 steps: 1) Establish, during the short time-hori- ficiency of social spending 1) Selecting outcome indicators (the selection zon of the Initiative, with the participation 2) Provide a coherent framework for the de- and monitoring of outcome and indicators of civil society, a durable process that sign of social spending plans over the medium should be done in a broad-based partici- would yield sustainable growth and term (including resource allocation decisions) patory process) poverty reduction over the longer term (as 3) Offer an additional advantage in the mon- 2) Identifying determinants of poverty debt relief is irrevocable at completion itoring of the implementation of the program 3) Establishing priorities for public action point) 4) Enhance transparency and accountability and related external assistance 2) Ensure that HIPC debt relief is an inte- Link to HIPC Contents gral part of broader efforts to implement 5) An impetus to prioritize the allocation of 1) Consistency between a country's macro- outcome-oriented poverty reduction strate- all available resources in line with targets re- economic, structural, and social policies and gies lated to poverty reduction the goals of poverty reduction and social de- 3) Demonstrate HIPC is geared to poverty 6) Improve efficiency of poverty-reducing velopment reduction expenditures 2) Should serve as the basis for designing 4) Demonstrate higher aid flows are jus- 7)Target social spending to the poor Bank and Fund lending operations, and as a tified by the effective use of all assis- Broad-based participatory process framework with which all ESAF and Bank- tance 8) Create a sense of ownership (lessons supported programs should be consistent from Higher-Impact Adjustment Lending) 3) Must be produced in a way that includes 9) Potentially increase information flow avail- transparency and broad-based participation able to design and implement a PRS (not via in the choice of goals, the formulation of poli- a poverty fund--need to ensure "poverty cies, and the monitoring of implementation -- fund" is fully integrated into wider budget) with ultimate ownership by the government 2) Building Poverty Reduction Strategies in Developing Countries, September 1999 (World Bank and IMF 1999b) Strengthen the impact of public action 1) A comprehensive understanding of poverty Elements of Framework on poverty and its determinants at the country level 1) A comprehensive understanding of poverty Additional Objectives 2) Prioritizing public actions with the high- and its determinants 1) Promote progress toward the Interna- est poverty impact: 2) Choosing public actions that have the high- tional Development Goals on a country basis · Sustained growth and access to economic est poverty impact 2) Operationalize CDF in a way that would opportunities for the poor 3) Outcome indicators that are set and mon- systematically link diagnosis and public ac- · Improving delivery of essential services itored using participatory processes tions to poverty outcomes · Increasing empowerment and participation 3) Provide the basis for ensuring that HIPC · Reducing vulnerability and insecurity debt relief is an integral part of poverty re- 3) Participatory setting and monitoring of duction efforts outcome indicators 4) As part of a broader effort, enhance the impact of the Bank's work on poverty reduc- tiontopromotesustainablegrowthandtoen- sure that its benefits are reaching the poor 5 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 : A N I N D E P E N D E N T E VA L U AT I O N Objectives Intermediate outcomes Process / outputs 3) PRSPs ­ Operational Issues, December 1999 (World Bank and IMF 1999c) To assist countries in developing and 1) Help to encourage and track country Five Principles: implementing more effective strate- progress on four fronts: 1) Country-driven gies to fight poverty · Development of appropriate diagnostics 2) Results-oriented · Deepening of a shared vision across civil 3) Comprehensive To make: society 4) Partnerships 1) Actions at the country level and · Setting of priorities and design of public 5) Long-term perspective 2) The support of development partners actions to achieve desired poverty reduc- Questions to Consider When Designing more effective in bringing sustainable tion outcomes a PRSP: poverty reduction · Development of participatory processes 1) What are the key dimensions of poverty? used for the setting of poverty reduction 2) What are the obstacles to poverty reduc- goals and monitoring of implementation tion? 2) Enhance collaboration between Bank and 3) What are the objectives and targets? Fund in order to help country authorities pro- 4) What are the priority public actions, and duce a poverty reduction strategy in which how can public spending be made more ef- policy actions to raise growth and reduce ficient? poverty are integrated into a coherent macro 5) What is the framework for monitoring structural and social framework progress? 6) How to ensure the transparency of public expenditure? 7) How does external assistance and the ex- ternal environment support or affect the country's efforts? 8) How did the participatory process affect formulation and what did it consist of? 4) PRSPs ­ Internal Guidance Note, January 2000 (World Bank 2000a) 1) Help countries to fill gaps in poverty re- Gaps in Poverty Reduction Strategies: Five Principles: duction strategies and to improve their 1) Do not always focus on monitoring re- 1) Country-driven design and implementation (Sharpen na- sults 2) Results-oriented tional poverty reduction strategies and 2) Do not always make clear the causal links 3) Comprehensive make them more effective) between public action and poverty reduction 4) Partnerships 2) Enhance Bank/Fund collaboration (by 3) Do not always reflect a participatory ap- 5) Long-term perspective helping the country authorities to produce proach a poverty reduction strategy that appro- Analytical Basis for Implementation: priately balances financial/macroeconomic 1) Empowering the poor through participa- and structural/social considerations) tion and public institutions that are ac- countable to the poor 2) Providing for security through reducing the impact of shocks 3) Expanding opportunities for the poor through the creation of physical and human assets and a sustainable growth path 5 4 ANNEX B: GUIDELINES FOR COUNTRY CASE STUDIES The Operations Evaluation Department (OED) source of evaluative material. The methodology of the World Bank is undertaking an evaluation will combine two types of case studies identi- of the Poverty Reduction Strategy Paper (PRSP) fied in the evaluation literature, "program im- process. The objective of the review is to assess plementation" (primarily explanatory) and progress of the PRS process toward meeting the "illustrative" (mainly descriptive) (see Datta challenge of poverty reduction and to assess the 1990; Morra and Friedlander 1999). The case World Bank's role in support of the process, studies will be "program implementation" case with a view to informing and, where necessary, studies as they intend to investigate the im- making recommendations to strengthen the im- plementation of the PRS process in each coun- plementation of the Initiative and to increase the try relative to its underpinnings. They will also effectiveness of the World Bank's support. The be illustrative in nature as they will add in- evaluation will be conducted in collaboration depth examples to other sources of evaluative with the IMF's Independent Evaluation Office material, including external evaluations, the- (IEO), which is carrying out a parallel evaluation matic studies, cross-country analysis, and sur- of the Poverty Reduction and Growth Facility vey work. (PRGF) process. Together these evaluations will To capture the diversity of country experi- provide an independent assessment of BWI ef- ences with the PRSP Initiative, the OED review fectiveness in support of the PRS process to the has sought to conduct a high number of case Boards of the two institutions. The issues that the studies. OED plans to conduct eight case stud- evaluations will cover are discussed in detail in ies for the evaluation, four of which will be con- the Issues Paper from IEO (http://www.imf.org/ ducted jointly with IEO. IEO is planning to external/np/ieo/2002/prsp/index.htm) and the conduct another two case studies independ- Approach Paper from OED (http://www.world ently. In total, OED and IEO will perform an in- bank.org/oed/prsp). depth investigation of 10 of the 23 countries This note proposes guidelines for conducting with completed PRSPs through December 2002. country case studies answering the evaluative ques- A range of criteria were considered in se- tions laid out in the OED Approach Paper. It is based lecting the PRSP countries for case studies: heavily on related work between OED and IEO as geographic balance; HIPC and non-HIPC; range primary input for both evaluations. While the indi- of country conditions and PRSP experience; vidualevaluationswilleachfocusoninstitutionalper- and both mature and recent programs. Coun- formance, the OED and IEO teams will conduct try selection also took into account the coun- jointdeskandfieldworkinoverlappingcountrycase try coverage of previous or ongoing work.1 studies.Itisexpectedthatcollaborationbetweenthe The following countries were selected based units will result in a single country case study report on the criteria set out above: Albania, Cam- for each of the joint case studies. bodia, Ethiopia, and Mauritania (OED) and Mozambique, Nicaragua, Tajikistan, and Tan- Approach zania (joint OED-IEO). IEO will be conducting Given the country-specific nature of the PRS an additional two case studies in Guinea and process, case study analysis will be used as a key Vietnam. 5 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 : A N I N D E P E N D E N T E VA L U AT I O N Details on Methodology on the emerging findings of the case study from The following are guidelines for conducting the a representative group of stakeholders, including individual case studies by OED, including those stakeholders involved in the interviewing. conducted jointly with IEO. They define the A stakeholder survey jointly designed by OED basic parameters for the case studies, and can be and IEO is being conducted in each country to pro- further tailored to reflect country characteristics. vide input for both the OED and IEO studies. The The guidelines have been informed by experi- survey will supplement information gained from ence from lead missions in Albania (OED) and key informant interviews and will target each cate- Tanzania (joint OED-IEO). goryofstakeholderinvolvedinthePRSprocess.Cov- A local consultant will be sought to assist erage will include the central government ministries in each of the case studies. The local consultant (Planning and Economy/Finance) and the Central will be expected to contribute to the team in line Bank, line ministries, donors, NGOs, and repre- with their level of expertise and knowledge. This sentatives of the private sector and other civil soci- is expected to yield a range of local consultant ety groups (for example, trade unions, labor unions, assistance across the case studies, from assisting academia, media, and faith organizations). The goal in scheduling meetings and providing local of the survey is to obtain institutional responses knowledge, to preparing background papers (not individual views) on the PRSP and on Bank and and briefs. Local consultants will also be used to IMF performance in supporting the process. As a re- coordinate implementation of the country stake- sult, the criteria for respondents includes good holder survey (see below). knowledge of the PRS process and familiarity with Key informant interviews will be held both their respective institution's positions on specific is- at Headquarters and in the field. Interviews will sues. A core set of stakeholders has been identified be conducted with relevant World Bank and IMF that each country survey must target, including key staff who have been associated with the country ministries involved in the PRS process, but outside team and/or the PRS process. This will be fol- of these parameters, the specific coverage and num- lowed by in-country consultations to be con- ber of responses is likely to vary from country to ducted for a period of at least two weeks. These country. Local consultants will be used to adminis- consultations are the primary input for the field- ter surveys in each case country, further enhancing work and will be targeted at a broad range of the independence of the exercise. stakeholders, including key ministries (central Case studies will use multiple methods for and line), key government units involved in the data collection, including reviews of relevant PRS process, local governments, bilateral and documents and databases, in-depth interviews of multilateral donors, local and international NGOs, key informants both in Washington, D.C., and in private sector representatives, media, and, wher- the country, and an in-country workshop. Rele- ever possible, parliamentarians. Interview modal- vant documents may include official govern- ities will vary and may include group ment data and reports, Bank and Fund analytical consultations. Finally, an exit workshop (see and advisory work, related policy documents, below) will be held that will target stakeholders staff reports to the Board, internal documenta- consulted during the mission. tion, prior OED and IEO evaluations, external In-country field visits outside of capital will evaluations, and materials from research insti- be conducted in each case country, unless not tutes and civil society. warranted, or not possible given country con- An extensive analysis of the data will be carried ditions. Field visits were conducted in both Al- out using the technique of triangulation, in which bania and Tanzania and provided considerable the reliability of findings will be developed through value added. The modalities of field visits will vary multiple data sources, and findings will be validated from country to country. throughagreementacrossmultipletypesofdata.An An informal exit workshop will be held in example would be assessing process changes in each case country at the conclusion of the mission. public expenditure management in a country. The The purpose of the workshop is to seek feedback country's Public Expenditure Review may yield an 5 6 A N N E X B : G U I D E L I N E S F O R C O U N T R Y C A S E S T U D I E S initial finding that could then be verified with mul- ings to key internal stakeholders at a World Bank tiple data sources, such as interviews with Bank or IMF brown bag lunch, after which the case staff in the Regional Public Expenditure thematic study report will be finalized. group, officials in the Ministry of Finance, and civil society. Economic data on budget execution could Final Report be cross-checked against budget laws and qualita- In joint case study countries, OED and IEO will tive data on disbursements and procurement; that coordinate to produce a single report as a key is, data would be validated across multiple types of input to both evaluations. Some sections of the data. Likewise, to assess the realism of macroeco- report will be under the exclusive responsibility nomic projections made by the IMF, one may assess of either IEO or OED, while others will convey theextentofdeviationsfromrecenttrends,compare a joint assessment. The suggested outline for projections with outcomes, and cross-check that case study reports is as follows: against views expressed at the time by the authori- ties, review departments, and the World Bank. I) Country Background: · Poverty, political context, economic policy Phases and Key Deliverables and performance, national strategies and their formulation Phase 1 ­ Desk Review: A desk review will be con- · The PRSP Process: ducted in advance of field work in each of the ­ Relevance and value added of the PRSP ap- case countries. The desk review will analyze and proach compared with past modalities for summarize main findings from available research strategy formulation and donor assistance materials and collate the necessary data. A desk ­ Application of underlying principles and brief will be prepared, following a standard out- assessment of what has changed line to provide an initial basis of information for ­ Preliminary evidence on process results, the country study. including alignment of budget to the PRSP II) World Bank Effectiveness Phase 2 ­ Washington-Based Interviews and Mission · Bank support during PRSP formulation Planning: Phase 2 will build on the desk review · Bank support and alignment during im- through interviews with Bank and Fund staff. plementation Key persons for interviewing in the field will be III) The PRGF and IMF Support identified, evaluative questions will be tailored IV) Bank-Fund Cooperation and the JSA around the emerging evaluative hypotheses, and · Role of the JSA arrangements for the workshop and field inter- · Bank-Fund cooperation views will be finalized. V) Main Conclusions Phase 3 ­ Country Visit: The field work will investi- Staffing of Joint Case Study Teams gate the key evaluative questions in-country and For joint case studies, teams will consist of at least seek to address open issues. Structured interviews one core evaluation team member from each with key informants and an in-country workshop evaluation unit, along with at least one senior in- will form the basis for consultations. As noted country national consultant. Either OED or IEO above, an in-country stakeholder survey will be will be assigned overall responsibility for coor- applied after field work has been completed. dinating the case study. The team will also be sup- ported by junior consultants, to be assigned by Phase 4 ­ Final Washington-Based Analysis: In Phase each unit as deemed necessary. The OED and 4, the case study team will agree on the main find- IEO teams will nominate a case study coordina- ings and supportive evidence, submit them to a tor. His/her responsibilities include overseeing new round of triangulation with HQ-based staff coordinating interviews, arranging the in-coun- from the Fund and the Bank, and draft a report. try workshop, overseeing the production of in- The case study teams may present their find- puts, and production of the final report. 5 7 ANNEX C: PRSP STATUS OF 81 IDA-ELIGIBLE COUNTRIES AS OF MAY 2004 Enhanced Date of Presentation to the Board of Post-PRSP HIPC Executive Directors of the World Bank CAS at end- Country Eligibility I-PRSP PRSP APR1 APR2 APR3 2003 A) Countries with a PRSP at End-2003 (35) Very early (PRSP completed in 2000 or 2001) Uganda E 2-May-00 31-May-01 23-Jul-02 9-Sep-03 X Burkina Faso E 30-Jun-00 6-Dec-01 5-Nov-02 25-Mar-04 Tanzania* E 4-Apr-00 30-Nov-00 27-Nov-01 27-May-03 Mauritania* E 6-Feb-01 18-Jun-02 18-Jul-03 X Bolivia E 27-Jan-00 5-Jun-01 Nicaragua* E 21-Dec-00 25-Sep-01 11-Dec-02 22-Jan-04 X Mozambique* E 6-Apr-00 25-Sep-01 25-Jul-03 X Honduras E 6-Jul-00 11-Oct-01 26-Feb-04 X Early (PRSP completed in 2002) Niger E 20-Dec-00 7-Feb-02 19-Nov-03 X Zambia E 4-Aug-00 22-May-02 Albania* NE 8-Jun-00 20-Jun-02 10-Jul-03 X Vietnam* PS 12-Apr-01 2-Jul-02 19-Feb-04 X The Gambia E 14-Dec-00 16-Jul-02 X Guinea* E 22-Dec-00 25-Jul-02 X Yemen, Rep. PS 27-Feb-00 1-Aug-02 X Rwanda E 21-Dec-00 6-Aug-02 X Malawi E 21-Dec-00 29-Aug-02 23-Oct-03 X Ethiopia* E 20-Mar-01 17-Sep-02 17-Feb-04 X Guyana E 14-Nov-00 17-Sep-02 X Tajikistan* NE 31-Oct-00 5-Dec-02 X Senegal E 20-Jun-00 23-Dec-02 X Cambodia* NE 18-Jan-01 20-Feb-03 Kyrgyz Republic NE 4-Dec-01 27-Feb-03 X Late (PRSP completed in 2003) Mali E 7-Sep-00 6-Mar-03 X Benin E 13-Jul-00 20-Mar-03 X Sri Lanka NE 1-Apr-03 X Azerbaijan NE 5-Jul-01 7-May-03 Ghana E 24-Aug-00 8-May-03 Cameroon E 10-Oct-00 31-Jul-03 X Mongolia NE 27-Sep-01 9-Sep=03 5 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 : A N I N D E P E N D E N T E VA L U AT I O N Enhanced Date of Presentation to the Board of Post-PRSP HIPC Executive Directors of the World Bank CAS at end- Country Eligibility I-PRSP PRSP APR1 APR2 APR3 2003 Georgia NE 11-Jan-01 6-Nov-03 Chad E 25-Jul-00 13-Nov-03 X Nepal NE 18-Nov-03 X Madagascar E 19-Dec-00 18-Nov-03 X Armenia NE 22-May-01 20-Nov-03 B) Other IDA-eligible countries (31) Kenya PS 1-Aug-00 6-May-04 Serbia & Montenegro NE 8-Aug-02 16-Mar-04 Pakistan NE 4-Dec-01 11-Mar-04 Congo, Dem. Rep. E 11-Jun-02 Guinea Bissau E 14-Dec-00 São Tomé & Principé E 27-Apr-00 Sierra Leone E 25-Sep-01 Burundi E 22-Jan-04 Central Afr. Rep. E 18-Jan-01 Comoros E Congo, Rep. of E Côte d'Ivoire E 28-Mar-02 Laos E 24-Apr-01 Togo E Angola PS Afghanistan NE Bangladesh NE 17-Jun-03 Bhutan NE Bosnia- Herzegovina NE 00-Oct-02 Djibouti NE 27-Nov-01 Eritrea NE Haiti NE Indonesia NE Lesotho NE 6-Mar-01 Moldova NE 19-Dec-00 Nigeria NE Papua New Guinea NE Solomon Islands NE Timor-Leste NE Uzbekistan NE Zimbabwe NE 6 0 A N N E X C : P R S P S TAT U S O F 8 1 I D A - E L I G I B L E C O U N T R I E S A S O F M AY 2 0 0 4 Enhanced Date of Presentation to the Board of Post-PRSP HIPC Executive Directors of the World Bank CAS at end- Country Eligibility I-PRSP PRSP APR1 APR2 APR3 2003 C) Small Island Economy Exceptions, Inactive Countries and India (15) Liberia E Myanmar E Somalia E Sudan E Cape Verde NE 8-Apr-02 Dominica NE 19-Dec-03 Grenada NE India NE Kiribati NE Maldives NE Samoa NE St. Lucia NE St. Vincent NE Tonga NE Vanuatu NE *OED/IEO case study country. Enhanced HIPC Eligibility: E ­ Countries eligible for E-HIPC; PS ­ Countries potentially sustainable under traditional debt relief mechanisms; NE ­ Countries not eligible for E-HIPC. Date of Presentation to the Board of Executive Directors of the World Bank: This date may differ from the date of completion of the PRSP in-country. OED used the latter date for dividing countries into Very Early, Early, and Late grouops. 6 1 ANNEX D: SURVEY OF NATIONAL STAKEHOLDERS IN THE 10 CASE STUDY COUNTRIES As part of the OED and IEO evaluations of the asked to define the nature of their involvement PRSP process and the PRGF, a survey of PRSP in the PRSP process and their level of familiarity stakeholders was administered in each of the with the PRSP document, the Bank, and the IMF. 10 countries where a case study was undertaken. Given the targeted nature of the survey, re- The objective of the survey was to obtain per- spondents who were "not aware" of the PRSP ceptions of the PRSP process and the role of process were excluded from the results. The the World Bank and IMF in supporting the Ini- specific samples were selected using three main tiative. inputs: information gained through the country A standard survey of 39 questions was ad- case study mission; participants listed in the ministered in each country. The survey consists PRSP document; and input from the local con- of four main components: information on re- sultant. In some cases, samples were circulated spondents; the PRSP process (covering owner- to obtain broader input on their composition. ship, results orientation, comprehensiveness, The study teams also identified a set of highly rel- partnership-orientation, and long-term per- evant respondents in each country for whom a spective); World Bank performance; and the survey response was required. These included role of the IMF. In most cases, respondents were core ministries and agencies, key PRSP-related asked to indicate the extent of their agreement ministries, and major donors. Survey question- with statements on a five-point scale.1 The sur- naires were tracked in order to ensure that re- vey was translated into local languages, where sponses were obtained from key groups, but necessary, and pre-tested. A local consultant individual respondents could choose to remain with survey experience was engaged in each anonymous. country to assist with administration of the sur- The following section presents aggregate find- vey. Survey results were coded by the local con- ings from the survey from all 10 countries. Sec- sultant and sent back to Washington. An outside tion A provides an overview of the survey contractor, Fusion Analytics, was hired to analyze respondents, including the nature of involve- the data. ment and familiarity with the process. Section B Roughly 1,000 questionnaires were sent out provides an aggregated snapshot of stakeholder to stakeholders in the case study countries. perceptions of the PRSP process across each of There were a total of 779 responses, spread five main sub-categories. Section C provides the evenly across stakeholder categories. The survey mean results for all questions concerning the role was targeted at key groups within the three main and effectiveness of Bank and Fund support. categories of PRSP stakeholders: government, Section D presents results for questions with civil society, and international partners.2 Within the most positive and negative responses and each group, the survey sought to obtain an in- questions where there was the greatest con- stitutional view and was targeted at the most sensus or disagreement on issues. The survey knowledgeable individuals. Respondents were questionnaire is attached in Section E. 6 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 : A N I N D E P E N D E N T E VA L U AT I O N A. Respondent Information 1. Composition of Respondents (n = 779). Government 35% Civil society 42% Int'l NGO 10% Donor 13% 2. Nature of Involvement (%) Involved in both Strategy and 20% Implementation/Monitoring Involved in Implementation/ 16% Monitoring only Direct Contribution to Strategy 14% Consulted during Strategy Only 14% Not Involved, but Aware 32% Not Aware 4% 3. Level of Familiarity PRSP document 56% Bank 51% IMF 33% 6 4 A N N E X D : S U R V E Y O F N AT I O N A L S TA K E H O L D E R S I N T H E 1 0 C A S E S T U D Y C O U N T R I E S B. The PRSP Process Relevance 3.48 Partnership-Oriented 3.33 Comprehensive and 3.20 Long-Term Country-Driven 3.13 Results-Oriented 3.12 C. Bank and IMF Q34: World Bank assistance supports 3.69 PRSP priorities Q32: World Bank involvement has 3.65 been very helpful Q33: World Bank strategy is aligned 3.62 with PRSP Q36: World Bank promoting coordination of donor assistance 3.53 Q35: World Bank activities provide relevant inputs 3.45 World Bank Q37: IMF involvement has 3.35 been very helpful Q39: Design of PRGF program 3.27 indicates more flexibility Q38: Gov't-linked budgets with PRGF is more pro-poor and growth 3.23 than before IMF Note: The five-point scales used in most questions--from 1 (completely disagree) to 5 (completely agree). 6 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 : A N I N D E P E N D E N T E VA L U AT I O N D. Composite Results Table Agree or Disagree or Don't completely Neutral completely know or agree (%) (%) disagree (%) Standard unsure (%) Question (4 or 5) (3) (1 or 2) Mean deviation (0) Most positive Q8 - Relevance: PRSP is a good model 59 30 11 3.69 1.06 7 Q9 - Relevance: PRSP adds value 58 28 14 3.64 1.08 9 Q10 - Relevance: PRSP improves on past modalities 57 30 13 3.61 1.05 11 Q25 - Partnership-oriented: Donors supported formulation 56 24 20 3.57 1.15 16 Q16 - Results-oriented: Outcomes benefit poor 55 26 19 3.56 1.15 8 Most negative Q18 - Results-oriented: Structure to monitor results 28 31 41 2.84 1.17 21 Q19 - Results-oriented: Results feed back 31 32 37 2.96 1.16 24 Q22 - Comprehensive: Macroeconomic framework participatory 38 27 35 3.03 1.25 17 Q17 - Results-oriented: Realistic targets and plans 38 28 34 3.14 1.16 9 Q21 - Comprehensive: Alternatives fully explored 35 32 33 3.01 1.12 18 Most polarized Q15 - Country-driven: Gov't continues to engage stakeholders 48 19 33 3.23 1.42 13 Q13 - Country-driven: Your stakeholders were consulted 45 22 33 3.20 1.33 11 Q14 - Country-driven: Final document was modified to accommodate viewpoints 44 25 31 3.16 1.29 19 Q22 - Comprehensive: Macroeconomic framework participatory 38 27 35 3.03 1.25 17 Q12 - Country-driven: PRSP driven by national stakeholders 42 28 30 3.24 1.25 8 Greatest consensus Q30 - Partnership-oriented: Quality of Bank/Fund collaboration 52 32 16 3.44 1.04 46 Q28 - Partnership-oriented: Current donor coordination 34 39 27 3.03 1.02 18 Q29 - Partnership-oriented: Coordination b/t World Bank and IMF improved 46 35 19 3.36 1.04 48 Q10 - Relevance: PRSP improves on past modalities 57 30 13 3.61 1.05 11 Q8 - Relevance: PRSP is a good model 59 30 11 3.69 1.06 7 Most unfamiliar Q29 - Partnership-oriented: Coordination b/t World Bank and IMF improved 46 35 19 3.36 1.04 48 Q30 - Partnership-oriented: Quality of Bank/Fund collaboration 52 32 16 3.44 1.01 46 Q19 - Results-oriented: Results feed back 31 32 37 2.96 1.16 24 Q27 - Partnership-oriented: Donor coordination improved 52 26 22 3.43 1.11 22 Q18 - Results-oriented: Structure to monitor results 28 31 41 2.84 1.17 21 Note: % Agree, % Neutral, and % Disagree are calculated as a percentage of those who responded to each question. % Don't Know or Unsure is calculated as a percentage of total re- spondents to the survey (779). Most Positive: were chosen based on highest percentage who agreed or completely agreed and top give means. Where there were differences among the top five that resulted, the question with the larger number of respondents was chosen. Likewise, Most Negative were selected based on highest percentages who disagreed or com- pletely disagreed and the five lowest means. The Most Polarized questions and those with Greatest Consensus were those questions with the five highest and lowest standard devia- tions respectively. Most Unfamiliar were questions that had the highest percentage of total respondents to the survey answering that they Didn't Know or were Unsure (0). 6 6 A N N E X D : S U R V E Y O F N AT I O N A L S TA K E H O L D E R S I N T H E 1 0 C A S E S T U D Y C O U N T R I E S E. Survey Questionnaire OED Review of the PRSP Process / IEO Evaluation of PRSPs and the PRGF Country Stakeholder Survey The Operations Evaluation Department (OED) of the World Bank and the Independent Evaluation Office (IEO) of the International Monetary Fund (IMF) are independent units reporting to their respective institutions' Executive Boards. Both units are collaborat- ing in some aspects of their respective evaluations of the Poverty Reduction Strategy Paper (PRSP) Initiative. The objective of these evaluations is to assess the Initiative's relevance, design, and implementation, as well as the prospects for achieving the imme- diate objectives and development goals. The evaluations will also seek to identify possible needs for improvement. An area of par- ticular focus will be the role and effectiveness of the World Bank and the IMF (including through the Poverty Reduction and Growth Facility ­ PRGF) in support of the PRSP process. The following survey is designed to provide input into a series of full country case studies being conducted for the evaluations. The survey will take between 10 and 15 minutes to complete, consisting of 39 multiple choice questions grouped into three cate- gories: I) Respondent information II) The PRSP process A) Relevance B) Application of the underlying principles ­ Country-driven ­ Results oriented ­ Comprehensive and long-term ­ Partnership-oriented ­ Intermediate effects and outcomes III) World Bank and IMF role and performance. The evaluation teams welcome additional comments beyond the scope of the survey questions. These can be sent by mail, elec- tronic mail, or through the evaluation Web sites. Operations Evaluation Department Independent Evaluation Office PRSP Review Core Team PRSP Evaluation 600 ­ 19th Street N.W. International Monetary Fund Washington, D.C. 20433 700, 19th street NW http://www.worldbank.org/oed/prsp Washington, D.C. 20431 ieo@imf.org 6 7 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 : A N I N D E P E N D E N T E VA L U AT I O N Instructions Please mark the appropriate response for each question. Individual responses to the survey will not be shared outside of the evaluation teams and results will be presented only in aggregate form. Follow-up questions, noted in italics, are optional and may be skipped in case time pressures so require. I) RESPONDENT INFORMATION Country: _____________________ Date: _____________________ Name (optional): _____________________ Organization (optional): _____________________ E-mail Address (optional): _____________________ 1. Please indicate from the choices below the category you belong to: Government i. Central Government : Economy, Finance, Planning, Central Bank............ . . . . . . . . . . . . . . . . . . . . . 3 ii. Central Government: other (sector ministry, agencies, etc.). . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3 iii. Local Government . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3 iv. Parliament . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3 Civil Society v. Local nongovernmental organization . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3 vi. Business sector . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3 vii. Labor union . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3 viii. Academia. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3 ix. Media . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3 x. Religious organization . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3 xi. Political party. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3 xii. Other . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3 International Partner xiii. Donor . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3 xiv. International NGO . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3 2. What has been the nature of your involvement in the PRSP process? Mark all that apply. Made direct Aware of Consulted contribution Participate Participate Unaware of the process, during strategy during strategy during during the PRSP process but not involved formulation formulation implementation monitoring 3 3 3 3 3 3 3. How knowledgeable are you of the PRSP document? Unfamiliar Very knowledgeable 1 2 3 4 5 4. How familiar are you with the work of the World Bank in the country? Unfamiliar Very knowledgeable 1 2 3 4 5 6 8 A N N E X D : S U R V E Y O F N AT I O N A L S TA K E H O L D E R S I N T H E 1 0 C A S E S T U D Y C O U N T R I E S 5. How familiar are you with the work of the IMF in the country? Unfamiliar Very knowledgeable 1 2 3 4 5 6. Have you had interactions with the World Bank or been involved with Bank-supported projects/programs or analytical/advi- sory work? Yes 3 No 3 7. Have you had interactions with the IMF? Yes 3 No 3 II) THE PRSP PROCESS A) Relevance of the PRSP process 8. The concept of a PRSP is a good model for addressing poverty reduction. Don't know Completely Completely or Unsure Disagree Agree 0 1 2 3 4 5 9. The PRSP process has added value to the country's response to the challenges of poverty reduction, compared to previous strategies/policy frameworks. Don't know Completely Completely or Unsure Disagree Agree 0 1 2 3 4 5 10. The PRSP framework improves on past modalities for donor assistance. Don't know Completely Completely or Unsure Disagree Agree 0 1 2 3 4 5 11. The benefits and strengths of the PRSP process have so far outweighed costs and weaknesses. Don't know Completely Completely or Unsure Disagree Agree 0 1 2 3 4 5 B) Application of the underlying principles Country-driven 12. The PRSP process is driven by and managed by national stakeholders. Don't know Completely Completely or Unsure Disagree Agree 0 1 2 3 4 5 13. The group of stakeholders you belong to (as defined in the first section) was adequately consulted in formulating the PRSP strategy. Don't know Completely Completely or Unsure Disagree Agree 0 1 2 3 4 5 6 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 : A N I N D E P E N D E N T E VA L U AT I O N 14. The final PRSP document was relevantly modified to accommodate some of the viewpoints of your stakeholder group. Don't know Completely Completely or Unsure Disagree Agree 0 1 2 3 4 5 15. Government has continued to engage your stakeholder group in the implementation and monitoring of the PRSP. Don't know Completely Completely or Unsure Disagree Agree 0 1 2 3 4 5 Results-oriented 16. The PRSP process is focused on outcomes that benefit the poor. Don't know Completely Completely or Unsure Disagree Agree 0 1 2 3 4 5 17. The strategy includes realistic targets and action plans. Don't know Completely Completely or Unsure Disagree Agree 0 1 2 3 4 5 If you answered 1 or 2, please indicate cause (multiple choices allowed): Lack of prioritization. . . . . . . . . . . . . . . . . . . . . . 3 Financing constraints . . . . . . . . . . . . . . . . . . . . . 3 Implementation capacity constraints. . . . . . . . . 3 Inadequate diagnostic . . . . . . . . . . . . . . . . . . . . 3 18. An effective structure to monitor and evaluate results has been established. Don't know Completely Completely or Unsure Disagree Agree 0 1 2 3 4 5 19. Results from monitoring and implementation feed back into the policy formulation process. Don't know Completely Completely or Unsure Disagree Agree 0 1 2 3 4 5 Comprehensive and long-term 20. The PRSP document provides an adequate diagnosis of the causes of poverty. Don't know Completely Completely or Unsure Disagree Agree 0 1 2 3 4 5 21. During strategy formulation policy alternatives were/are fully explored. Don't know Completely Completely or Unsure Disagree Agree 0 1 2 3 4 5 7 0 A N N E X D : S U R V E Y O F N AT I O N A L S TA K E H O L D E R S I N T H E 1 0 C A S E S T U D Y C O U N T R I E S 22. The formulation of the macroeconomic framework of the PRSP has been sufficiently participatory. Don't know Completely Completely or Unsure Disagree Agree 0 1 2 3 4 5 23. The PRSP document strikes an appropriate balance between the promotion of growth and poverty alleviation (e.g., through social expenditures and/or redistributive measures). If you answered 1 or 2, it is mainly because: Don't know Completely Completely or Unsure Disagree Agree 0 1 2 3 4 5 The PRSP puts too much emphasis on growth . . . . . . . . . . . . . . . 3 The PRSP puts too much emphasis on poverty alleviation . . . . . . 3 24. The PRSP document provides an adequate road-map for achieving long-term goals and targets. Don't know Completely Completely or Unsure Disagree Agree 0 1 2 3 4 5 Partnership-oriented 25. Donors supported the formulation of the PRSP in a manner that did not erode country ownership. Don't know Completely Completely or Unsure Disagree Agree 0 1 2 3 4 5 26. Donors are aligning their assistance to the PRSP priorities. Don't know Completely Completely or Unsure Disagree Agree 0 1 2 3 4 5 27. Donor coordination has improved since the launch of the PRSP process. Don't know Completely Completely or Unsure Disagree Agree 0 1 2 3 4 5 28. Please rate the quality of donor coordination at present. Don't know Completely Completely or Unsure Disagree Agree 0 1 2 3 4 5 29. Collaboration and coordination between the IMF and the World Bank have improved since the start of the PRSP process. Don't know Completely Completely or Unsure Disagree Agree 0 1 2 3 4 5 30. Please rate the quality of Bank/Fund collaboration at present. Don't know Completely Completely or Unsure Disagree Agree 0 1 2 3 4 5 7 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 : A N I N D E P E N D E N T E VA L U AT I O N Intermediate effects and outcomes 31. The PRSP process has brought about significant changes for the better in the following areas (mark all that apply). Ownership of economic policies . . . . . . . . . . . . . . . . . . . . . . . . . .3 Clarity and openness of socio/economic policy debate . . . . . . . .3 Transparency and accountability . . . . . . . . . . . . . . . . . . . . . . . . . .3 Business environment . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .3 Public services delivery . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .3 Decisionmaking processes . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .3 Budget processes and allocations . . . . . . . . . . . . . . . . . . . . . . . .3 III) WORLD BANK AND IMF ROLE AND PERFORMANCE WORLD BANK ROLE 32. World Bank staff involvement in the PRSP process has been very helpful. Don't know Completely Completely or Unsure Disagree Agree 0 1 2 3 4 5 If you answered 1 or 2, please indicate the main reason(s) why (mark all that apply): Not enough involvement............................ . . . . . . . . . . .3 Too much involvement .............................. . . . . . . . . . . .3 Involvement wrongly focused in substance...... . . . . . . . . . . . . .3 Involvement wrongly targeted...................... . . . . . . . . . . .3 Other (please specify) _______________________________ 33. The World Bank's strategy for country assistance is aligned with the PRSP. Don't know Completely Completely or Unsure Disagree Agree 0 1 2 3 4 5 34. The World Bank's financial assistance supports PRSP priorities. Don't know Completely Completely or Unsure Disagree Agree 0 1 2 3 4 5 35. World Bank analytical and advisory activities provide timely and relevant inputs to the PRSP process. Don't know Completely Completely or Unsure Disagree Agree 0 1 2 3 4 5 36. The World Bank is actively promoting and supporting government-led coordination of donor assistance. Don't know Completely Completely or Unsure Disagree Agree 0 1 2 3 4 5 7 2 A N N E X D : S U R V E Y O F N AT I O N A L S TA K E H O L D E R S I N T H E 1 0 C A S E S T U D Y C O U N T R I E S IMF ROLE 37. IMF staff involvement in the PRSP process has been very helpful. Don't know Completely Completely or Unsure Disagree Agree 0 1 2 3 4 5 If you answered 1 or 2, please indicate the main reason(s) why (mark all that apply): Not enough involvement............................ . . . . . . . . . . .3 Too much involvement .............................. . . . . . . . . . . .3 Involvement wrongly focused in substance...... . . . . . . . . . . . . .3 Involvement wrongly targeted...................... . . . . . . . . . . .3 Other (please specify) _______________________________ 38. The government budget linked with the PRGF-supported program is more pro-poor and pro-growth than before. Don't know Completely Completely or Unsure Disagree Agree 0 1 2 3 4 5 39. The design of the PRGF-supported program indicates that the Fund is showing more flexibility in considering country-driven policy alternatives. Don't know Completely Completely or Unsure Disagree Agree 0 1 2 3 4 5 7 3 ANNEX E: REVIEW OF JOINT STAFF ASSESSMENTS OF PRSPS Purpose ology, and findings of a parallel content review Reviews of Joint Staff Assessments (JSAs) are conducted by OED. The last section outlines the being carried out as part of the evaluation of issues and rating criteria used for the review. the Poverty Reduction Strategy process. Ac- cording to the JSA Guidelines, the JSA is to be Scope jointly prepared by the staffs of the Fund and the JSAs for 28 countries were selected. These rep- Bank to provide an assessment for their re- resent all JSAs and associated PRSPs reviewed by spective Boards as to whether or not "the strat- the Boards of the Bank and the Fund between egy presented in a PRSP constitutes a sound the first PRSP reviewed by the Bank's Board in basis for concessional assistance from the Fund May 2000 (Uganda) and the last PRSP reviewed and the Bank" (World Bank 2000a). Thus, the JSA in FY03 (Azerbaijan). Only JSAs for full PRSPs should provide a window on an important aspect were reviewed. of the process: staff views of the PRSP, its Eleven issues were identified for assessment. strengths and its weaknesses. In addition, ac- They include three "key questions" delineated by cording to the guidelines, the JSA is to "provide the JSA Guidelines. These issues were not cov- constructive feedback to the country about how ered in the IEO review: it might improve its strategy over time." Two parallel JSA reviews have been under- 1) Governance and public sector management taken for the PRSP Process Evaluation. One re- 2) Poverty diagnosis view has been conducted by the Independent 3) Structural and sectoral policies. Evaluation Office (IEO) of the International Mon- etary Fund (IMF). The IEO review covers the In addition, eight additional issues were in- following "key questions" from the JSA Guide- cluded: lines: (1) country ownership through participa- tory processes; (2) analysis of risks to 4) Alternative growth strategies implementation; (3) assessments of targets, in- 5) Balanced treatment of social and growth sec- dicators and monitoring arrangements;.(4) as- tors sessments of priority action plans (including 6) Capacity constraints macroeconomic framework, fiscal choices, and 7) Donor partnership issues financing plan); and (5) consideration of coun- 8) Integration of short- and long-term objec- tries' starting points. In addition, the IEO re- tives view examines the following JSA process issues: 9) Integration of strategy into MTEF and budget (1) time lag in drafting; (2) collaboration be- process tween the IMF and the World Bank; (3) interac- 10) Private sector participation tions with local stakeholders in preparation of the 11) Risks ­ endogenous and exogenous.1 JSA; and (4) value added of JSAs in terms of use- fulness to the IMF Executive Board, to authori- These issues emerged from a review of a ties, and to local stakeholders. pilot sample of JSAs and associated Bank Board This annex summarizes the scope, method- comments, as well as from the 10 country case 7 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 : A N I N D E P E N D E N T E VA L U AT I O N studies conducted as part of the PRSP Evalua- were more than marginally above satisfactory tion.2 (3.0). For only six issues did more than half of the JSAs receive a satisfactory or better rating and Methodology on only one issue (risks) was the satisfactory Each JSA was assessed against each issue, using proportion above 75 percent. This less than im- a framework of criteria and scale factors. The cri- pressive performance suggests considerable teria indicate the best practice elements for each room for improvement. issue. The scale factors indicate which criteria Guidelines issues in satisfactory range. The need to be satisfied to meet one of the four rat- three issues covered by the JSA Guidelines-- ing points: highly unsatisfactory, unsatisfactory, structural and sectoral policies, poverty diag- satisfactory, and highly satisfactory. nosis, and governance and public sector The criteria and scale factors for the first three management--are rated in the lower satisfac- issues mentioned above draw heavily on the tory range, between 2.9 and 3.3. The majority of "Key Questions" posed for these issues by the JSA other issues get lower ratings. Guidelines (section 6). Criteria and factors for the Poor "learning." The 28 JSAs were divided other eight issues were developed from a re- into two categories: (1) those reviewed by the view of the pilot sample of JSAs and Board doc- Bank's Board in FY01­02 (13 JSAs); and (2) those uments and country case studies.3 In addition to reviewed in FY03 (15 JSAs). The purpose of this the numerical ratings, narrative comments were division was to see what change in ratings may prepared for each issue and each JSA, and are have occurred between these two periods. The available in the longer background paper. third and fourth columns of the table indicate that Methodological Limitations. The main limi- there has been marginal improvement--from an tation of the results presented below in table E.1 overall rating of 2.6 in FY01­02 to 2.8 in FY03. is that they do not cover five "key questions" de- Clear improvement has taken place between scribed in the JSA Guidelines and covered in the two periods for 6 of the 11 issues. But the the IEO JSA review. Another potential limitation degree of change has been modest, with two ex- to drawing inferences from averages would be ceptions: "balanced treatment of sectors" and "al- high variation across countries. But, as shown in ternative growth strategies," where improvement table E.1, country ratings for any given issue has been marked, but not enough to lift the lat- vary by less than one standard deviation among ter issue to a satisfactory, or 3.0, level. countries. Ratings for two of the three lowest-rated is- sues--"capacity constraints" and "private sec- Findings tor participation"--remained about the same or Marginally satisfactory quality. As the second increased modestly. "Donor partnership," al- column of table E.1 shows, JSAs for 28 coun- ready at an unsatisfactory level, experienced a sig- tries manifest an overall average rating between nificant decline. The number of JSAs per issue "unsatisfactory" and "satisfactory" on the 11--2.7 receiving a satisfactory rating for each of these on a 4-point scale. Only five issues had ratings three issues varied between only one and three in the satisfactory range, and, of these, only two JSAs in each period. 7 6 A N N E X E : R E V I E W O F J O I N T S TA F F A S S E S S M E N T S O F P R S P S Table E.1. Ratings of Joint Staff Assessments by 11 Selected Issues (figures in parentheses = number of JSAs rated 3.0 & above) 5/00 thru 6/03 Standard 5/00 thru 6/02 7/02 thru 6/03 Issue (ranked by overall rating, 2000­03) (28 JSAs) deviation (13 JSAs) (15 JSAs) 1. Risks--endogenous and exogenous 3.4.(26) +/­ 0.5 3.5.(13) 3.4.(13) 2. Structural and sectoral policies 3.3.(20) +/­ 0.7 3.1.(8) 3.4.(12) 3. Poverty diagnosis 3.1.(19) +/­ 0.7 3.2.(9) 3.1.(10) 4. Integration of strategy into MTEF and budget process 3.1.(19) +/­ 0.8 3.1.(9) 3.1.(10) 5. Balanced treatment of social and growth sectors 3.1.(19) +/­ 0.8 2.7.(6) 3.4.(13) 6. Governance and public sector management 2.9.(17) +/­ 0.8 2.8.(7) 3.0.(10) 7. Integration of short- and long-term objectives 2.6.(12) +/­ 0.5 2.6.(5) 2.7.(7) 8. Alternative growth strategies 2.4.(10) +/­ 0.8 2.0.(2) 2.7.(8) 9. Capacity constraints 2.2.(6) +/­ 0.6 2.1.(3) 2.2.(3) 10. Private sector participation 2.1.(5) +/­ 0.7 1.9.(2) 2.3.(3) 11. Donor partnership issues 1.8.(3) +/­ 0.7 2.1.(2) 1.5.(1) Overall average 2.7.(14) +/­ 0.7 2.6.(6) 2.8.(8) Note: Ratings are on a 4-point scale: 4=highly satisfactory, 3=satisfactory, 2=unsatisfactory, 1=highly unsatisfactory. See Annex B for more explanation. Issues, Rating Criteria, and Scale Factors 2. Poverty Diagnosis Rating scales: Criteria: (first three criteria are synthesized (1) = Highly unsatisfactory from section B of JSA Guidelines): JSA provides (2) = Unsatisfactory (a) an assessment of quality and comprehen- (3) = Satisfactory siveness of poverty data presented by the PRSP; (4) = Highly satisfactory (b) qualitative data, including data directly from the poor (as in Participatory Poverty Assessments); 1. Private Sector Participation in PRSP (c) analysis of nature and determinants of poverty; Preparation and Implementation (d) an assessment of growth and distributional im- Criteria: JSA provides (a) a description of pacts of past policies and programs; and (e) link- participation in PRSP preparation; (b) a de- ages to proposed strategy clearly indicated. scription of participation in PRSP implementa- tion; (c) feedback from the private sector about Scale Factors: its participation; (d) proposed measures to ad- (1) JSA provides only a cursory summary of dress private sector concerns; and (e) a plan for poverty data presented by the PRSP, with future involvement of private sector. little or no assessment of its quality. (2) Poverty data described and at least one of the Scale Factors above criteria met. (1) JSA does not mention whether the PRSP (3) Two of above criteria met, including candid refers to private sector participation in PRSP assessment of data quality. preparation. (4) At least three of above criteria met, includ- (2) Cursory reference to private sector partici- ing realistic recommendations. pation in PRSP preparation. (3) Analysis of at least 2 of the 5 above criteria. 3. Alternative Growth Strategies (4) Analysis of at least 3 of the 5 above criteria, Criteria: JSA provides (a) a description of at including realistic recommendation(s). least one alternative growth strategy and (b) an 7 7 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 : A N I N D E P E N D E N T E VA L U AT I O N assessment of implications and realism of alter- (more than 90 percent of space given to native strategies. one of the two clusters and less than three paragraphs for all sectors). Scale Factors (2) Coverage is somewhat skewed (more than (1) JSA does not refer to alternative growth two-thirds of space given to one of the two clus- strategies, whether or not they are men- ters) and skimpy (less than six paragraphs). tioned in the PRSP. (3) Coverage is fairly balanced (about the same (2) Cryptic reference to alternative strategies-- coverage given to each cluster) and com- possibly to a "contingency plan" or "macro- prehensive (six-to-eight paragraphs). economic scenario"--as presented in the (4) At least two of the above criteria are met, in- PRSP or as lacking in the PRSP. cluding realistic recommendations. (3) Clear description of at least one alternative growth strategy. 6. Structural and Sectoral Policies, Including (4) Assessment of one or more alternative Policies for Social Inclusion and Equity growth strategies, including realistic rec- Criteria: (drawn from sections D.5 and D.6 ommendations. of the JSA Guidelines) JSA assesses PRSP analy- sis of (a) impact of sectoral policies and pro- 4. PRSP Integration into MTEF and Budget grams on poverty; (b) prioritization and Processes sequencing, cross-sectoral issues, and measures Criteria: JSA provides an assessment of (a) to mitigate negative impacts; and (c) special is- prioritization and costing in the PRSP; (b) the sues, including, but not necessarily limited to, MTEF and its use in the budget process or, if no gender, social protection and labor, HIV/AIDS, MTEF, a cogent argument supporting an MTEF; and environmental sustainability. and (c) other related aspects of the budget process, such as expenditure controls. Scale Factors (1) JSA reveals relatively cryptic descriptions of Scale Factors policies, but no assessment of the quality of (1) Cursory coverage of only costing or priori- PRSP analysis, including impact on the poor. tization issues. (2) Comprehensive coverage, but weak assess- (2) Both costing and prioritization issues cov- ment of PRSP analysis. ered, but only in a cursory manner. (3) Full assessment of PRSP analysis, including (3) Clear, comprehensive assessment of cost- social inclusion issues, such as those listed ing and prioritization issues. above in criterion 6-(c). (4) Comprehensive assessment of the budget (4) Clear and candid assessment of PRSP analy- process, including costing and prioritiza- sis, including cross-sectoral issues, prioriti- tion, the existing and/or potential role of an zation and sequencing, and/or measures to MTEF, and other spending issues, such as ex- mitigate negative impact, including recom- penditure controls. mendations. 5. Balance Between (i) Social Sectors and (ii) 7. Governance and Public Sector Management Economic Growth. Criteria: (drawn in part from section D.7 of Criteria: JSA provides (a) balanced cover- the JSA Guidelines): JSA candidly assesses PRSP age of the two clusters (about 50 percent each); analysis of (a) measures to address financial (b) comprehensive coverage of the two clusters management and procurement problems; (b) (more than 8 paragraphs); and (c) prioritization plans to improve service delivery, including by within and among clusters. local governments; (c) steps to improve trans- parency and accountability vis-à-vis needs of the Scale Factors poor; (d) measures to improve efficiency and fair- (1) JSA gives very skewed and skimpy coverage ness of the legal and judicial system; and (e) ef- 7 8 A N N E X E : R E V I E W O F J O I N T S TA F F A S S E S S M E N T S O F P R S P S forts to improve civil service performance and (2) Mentions but does not analyze short-term address corruption. and long-term targets. (3) Analyzes relation between longer-term ob- Scale Factors jectives, including MDGs, and PRSP short- (1) JSA provides very cryptic description of PRSP term targets. coverage of one of the above-mentioned (4) Assesses PRSP analysis and makes clear and sub-issues. realistic recommendations. (2) Two issues covered, but little or no assess- ment of PRSP quality in these areas. 10. Capacity Constraints (3) Candid assessment of two of the above issues. Criteria: The JSA assesses the PRSP with re- (4) Systematic and candid assessment of at least spect to (a) analysis of capacity constraints; (b) three of the above sub-issues, including rec- analysis of policy and institutional as well as ommendations. human capital formation dimensions; (c) meas- ures that need to be taken that incorporate the 8. Donor Partnership Issues lessons of past experience; and (d) appropriate Criteria: JSA assesses PRSP analysis of (a) donor support. role of other donors in PRSP preparation process; (b) current and expected donor strategic align- Scale Factors ment with the PRSP; (c) degree of harmonization (1) JSA makes no reference to capacity issues. of other donor policies and procedures; and (d) (2) Only fragmentary and cursory reference to current and expected extent of country leader- capacity constraints. ship of these partnership dimensions. (3) Systematic assessment of PRSP analysis of ca- pacity constraints. Scale Factors (4) Systematic assessment, which also recom- (1) JSA provides only a cryptic reference, if any, mends measures that need to be taken to other donor participation in PRSP con- and/or role for donor support. sultation or preparation process. (2) Provides assessment of donor strategic align- 11. Endogenous and Exogenous Risks ment with PRSP. Criteria: JSA (a) describes both exogenous (3) Assessment of other aspects of harmoniza- and endogenous risks; (b) distinguishes between tion and/or country leadership of partnership. JSA and PRSP-identified risks; (c) indicates likely (4) Meets all criteria, including realistic recom- impact if risk is realized; (d) suggests mitigation mendations for next steps. strategies. 9. Integration Between Short-term PRSP and Scale Factors Long-term Development Strategies (1) No mention of such risks. Criteria: JSA assesses how PRSP (a) analyzes (2) Cursory and/or scattered reference to such linkages between ST, MT, and LT objectives; (b) risks. assesses need for longer-term strategies; (c) rec- (3) Clear and logical description of risks, in- ommends measures to be taken. cluding distinction between JSA and PRSP- identified risks. Scale Factors (4) Clear and candid description, as well as in- (1) JSA does not address this issue. dication of impact and mitigation strategies. 7 9 ANNEX F: MANAGEMENT RESPONSE Although some countries achieved modest re- for engagement with low-income countries by ductions in poverty in the 1980s and 1990s, over- the Bretton Woods Institutions (BWIs), other all progress in reducing poverty was less than had multilateral institutions, and most bilateral been hoped for, especially in low-income coun- donors. By end-June 2004, 42 countries had tries. This disappointment led to a critical ex- completed full PRSPs, another 14 had launched amination of what policies best promote the process with Interim PRSPs (I-PRSPs), and economic growth and reduce poverty in low- 23 countries had completed at least one An- income countries, and a realization that the de- nual Progress Report (APR).1 As the PRSP ap- livery of external support required significant proach gains momentum, and more recent improvement. There was also explicit recognition PRSPs build on the experience of earlier ones, that conditionality without ownership was inef- there is evidence that both participatory fective, and a desire to link provision of debt re- processes and content are improving. But chal- lief under the Enhanced Heavily Indebted Poor lenges remain. The Initiative has benefited from Countries (HIPC) Initiative to comprehensive a wealth of critical commentary by donors, civil country-owned strategies to reduce poverty. As society organizations, and academics.2 The an- a consequence, in December 1999, the Boards nual PRSP implementation progress reports of the World Bank and the IMF approved a new (prepared jointly by the Bank and the Fund) approach to the challenge of reducing poverty have also been candid in describing the chal- in low-income countries based on poverty re- lenges countries face in developing and imple- duction strategies (PRSs) produced by the coun- menting PRSPs. For example, the most recent tries themselves. annual report highlighted that the multiple ob- jectives of the PRS Initiative result in inherent The PRS Initiative. This approach involved the de- tensions that manifest themselves along sev- velopment of country-driven, results-oriented, eral dimensions, including the need for (a) comprehensive poverty reduction strategies that greater cohesion between PRSPs and other plan- are long-term in perspective and foster domestic ning documents; (b) a balance between ambi- and external partnerships in line with the princi- tion and realism in setting PRSP targets; (c) ples that underpin the Comprehensive Develop- greater attention to strengthening weak public ment Framework (CDF). These strategies were to expenditure management systems and effec- be embodied in Poverty Reduction Strategy Papers tive links to the budget in light of limited coun- (PRSPs), which were expected to serve as the try capacity; (d) better prioritization and focus; framework for development assistance beyond the and (e) improved donor alignment and har- operations of the Fund and Bank. monization around national strategies.3 Progress in Implementation. The PRS Initiative in- OED Review. Management welcomes the review troduced an explicit poverty focus, and em- by the Operations Evaluation Department (OED) phasized the need to link diagnostics, policies, of the Poverty Reduction Strategy (PRS) process. budgets, and results. As the Initiative enters its It raises issues similar to those the Bank has fifth year, the PRS has become the central model highlighted in its progress reporting. 8 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 : A N I N D E P E N D E N T E VA L U AT I O N Context. It is useful to start by placing the review that lack of clarity with regard to country-specific in the context of the evolving PRS process. First, process achievements creates tensions, and it rec- the PRS Initiative seeks to fundamentally change ommends establishing milestones for what own- the dynamics of external assistance to develop- ership, partnership, and results focus are ing countries and must be seen explicitly as a expected to achieve. It finds that countries have long-term undertaking. The process is still young. not used PRSPs effectively to prioritize their ca- The 40 countries with full PRSPs have had an av- pacity-building needs. It also suggests that Board erage implementation period of just less than two review of the PRSP and Joint Staff Assessment years. Because the PRS process has built-in mon- (JSA) undermines ownership because it creates itoring mechanisms, it evolves over time; but a perception of "Washington signing off." the OED review is mainly a snapshot of a single point in time. In addition, the review is based pri- Flexibility of the PRS Process. Management believes marily on 10 country case studies. The richness that the current architecture provides countries of the findings would likely have benefited from with the flexibility to tailor the PRSP process complementary thematic reviews based on all and content to individual circumstances. For ex- PRSP countries. ample, the guidelines for the JSA note that "the specific content of PRSPs will vary widely among Partnership Orientation. The Bank works closely countries" and that "staff should take into con- with the IMF and other partners in supporting the sideration the country's starting point." Indeed, PRS Initiative. Concurrently with the OED review, Uganda developed the first full PRSP by building the IMF's Internal Evaluation Office (IEO) also on its Poverty Eradication Action Plan (PEAP), completed an assessment to which IMF manage- which itself had incorporated most of the CDF ment is responding. While the two institutions principles. The OED review finds that the 10 work together, the comments in this paper reflect countries it reviewed in its case studies did not solely the views of World Bank Management; they use the flexibility inherent in the PRS process, but do not purport to represent the IMF's position, it does not examine either why this was the case or the views of the Bank's development partners in these countries or whether this was true also in other agencies or client countries. in other countries that had prepared PRSPs. Management believes that at least two factors OED Findings and Management could be at play in the case study countries. Comments First, in these countries as in some others, there The OED review concludes that the PRS Initia- may have been an initial rush (because of the link tive has the potential to enhance poverty re- to the HIPC Initiative) that set certain early prac- duction in low-income countries and should tice examples that were then emulated else- continue to be supported by the Bank. It also where as good practice. Second, the ambiguity finds that country experiences vary widely, and in the instrument may have led both country of- that countries face significant challenges in ef- ficials and Bank staff to favor a technique that fectively engaging in a PRS process. The review drew on early experience as approaches that recommends enhancements in some areas, par- worked. With the passage of time, however, ticularly in the areas of process, content, and PRSPs and APRs have better reflected country donor alignment. This section examines OED's specificities. It is also important to note that ini- main findings and provides Management's com- tial conditions and country-specific priorities ments on the analysis. are likely to have played a part in the relatively long time that many countries took to complete Process Issues a full PRSP. Of the 36 countries that prepared I- The OED review finds that the mechanisms for PRSPs before finalizing a full PRSP, the average adapting PRSPs to different country conditions time between the I-PRSP and the full PRSP was are weak, and that linkages to domestic planning just over 22 months, ranging from under 7 and budget processes are inadequate. It notes months to more than 44 months. 8 2 A N N E X F : M A N A G E M E N T R E S P O N S E Alignment of the PRS Process with Domestic Planning evant issue, therefore, in terms of changes in the and Budgeting. Management agrees that continued architecture of the PRS process, including pro- efforts are needed to further align the PRS cedures for Board review of PRSPs and JSAs, is process with domestic planning and budgeting how they help mitigate this tension. Manage- processes. Management notes that good practice ment believes that the key to enhancing coun- examples are emerging. As noted above, Uganda try ownership is to for PRSP countries to further built its PRSP on its PEAP, and it uses its existing integrate the PRS into domestic planning and Poverty Status Report as the APR. Yemen intends budget processes, as well as to institutionalize to synchronize its next PRSP with the country's participation--including by better integrating Five Year Plan. Mongolia and Ethiopia have spe- Parliaments and other domestic stakeholder cific plans for aligning the PRS to domestic groups into the process. To support this, Man- processes. Mozambique uses its existing annual agement agrees that it would also be beneficial budget review as its APR. Management notes to improve the JSA process so as to reduce the that continuing to strengthen the link to do- perception of "Washington signing off" on coun- mestic processes may be a key element in rein- try-owned documents. However, Management forcing the PRS as a country-driven process. does not agree with OED's suggestion that dis- Management will encourage PRSP countries to continuing Board presentation of JSAs would move further in this direction and, along with the be more supportive of ownership and facilitate IMF, will monitor progress in this regard. more candid and transparent feedback to coun- tries on PRSPs. Neither this suggestion nor the Degree of Clarity in the Process. To facilitate tailoring proposal that the operational link be made solely the PRS Initiative to country circumstances and pri- through the Bank's Country Assistance Strategy orities, the PRSP architecture was not specific (CAS) will help resolve the tension with country about expected process achievements. Further- ownership. Moreover, in Management's view, more, the PRSP is an instrument charged with this recommendation could undermine the rel- multiple objectives, many of which imply ten- evance of the PRS process sions--for example, the range of proposals emerg- ing from the participatory process versus Improvements to the JSA Process. Management is prioritization. PRSPs need to reflect choices and working with the IMF to introduce changes to the compromises across all these objectives. Man- JSA process that would support a more country- agement believes that this will inevitably result in driven process while reducing the perception of a less than ideal level of performance on some as- the Bank and the Fund driving the process and pects, and that countries' abilities to manage the contributing to the PRS approach being a more tensions inherent in the process will depend cru- effective framework for partnership at the coun- cially on their technical and administrative capac- try level. These modifications also seek to en- ity and on their political institutions. Strengthening hance the value of the JSA instrument in such capacity will take time and will require the co- providing candid and prioritized feedback to ordinated and sustained support of donors. countries as well as the Board. Specific changes, which will be proposed in the forthcoming an- Board Review of the PRSP and the JSA. The review nual PRSP implementation progress report (pre- concludes that the Bank should continue sup- pared jointly with the IMF), will seek to achieve porting the PRS Initiative. In management's view, these aims by clarifying the objectives and au- continued Bank support implies the need for dience of the JSA, emphasizing the need for maintaining an operational link between the greater selectivity, and allowing for more nu- PRSP and Bank assistance. Maintaining such a link ance and greater candor in the staffs' views. leads to an inherent tension with country own- ership of the PRS process, because it requires that Annual Progress Reports (APRs). Although APRs are Bank staff provide feedback on the strengths not specifically discussed in the OED findings, the and weaknesses of the country's strategy. The rel- analysis in the review seems to suggest that APRs 8 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 : A N I N D E P E N D E N T E VA L U AT I O N have played a constructive role. In Management's Beyond the Social Sectors and Expenditure Policies. view, these contributions could be enhanced by PRSPs and APRs are showing increasingly broad more closely linking APRs with the budget, re- sectoral coverage, not just in the discussion, but ducing the strain they impose on country ca- also in the targets and indicators. Of the 32 full pacity, and ensuring that their timing is driven by PRSPs that were completed by July 2003, all in- the country, rather than by the BWIs, including cluded a target or indicator related to economic by considering elimination of the requirement growth; 94 percent included one on macroeco- for Board presentation of APRs and associated nomic stability; 94 percent one on water; 81 per- JSAs other than in exceptional cases. In this way, cent one on roads; 78 percent one on APRs could also help in reducing the residual ten- governance/accountability; and 72 percent one sion between country ownership and the link be- on sanitation. While all had indicators on ma- tween the PRS process and Bank assistance. To ternal health and school enrollment and atten- this end, Management will present specific dance, coverage of other key social sector changes to the APR and associated JSA process indicators was more mixed. PRSPs have univer- in the forthcoming PRSP implementation sally addressed social sector issues, while also in- progress report. These changes will be aimed at creasingly including more balanced attention to better supporting country efforts to integrate growth and other sectors. the APR into domestic processes, reducing the reporting burden on countries, and adjusting Attention to Outcomes. Management agrees that the JSA process to address a similar set of con- the PRS process has helped to promote a results cerns as those identified for the full PRSP. orientation, but that countries still need help strengthening their country-specific goals and PRSP Content indicators and their monitoring systems. PRSPs The OED review finds that PRSPs' poverty focus often have long lists of indicators, with actual in- and attention to outcomes have improved, but dicators sometimes not well defined and baseline that more work is needed. It also suggests that data not always available. JSAs and associated PRSPs have focused excessively on social sector sector dialogue provide feedback on the realism expenditures. It notes the need for better ana- of targets as well as weaknesses in the proposed lytics to underpin PRSPs, particularly in the areas indicators or the institutional framework for col- of growth, linkages between policies and poverty lecting and using data. Countries have tended to outcomes, and trade-offs. It also argues that the refine their indicator lists in their APRs, suggest- PRSP process could have been used more ef- ing that as implementation proceeds, countries fectively to identify key analytic gaps and de- become more selective in emphasizing indicators velop a research agenda to address them. that are more directly linked to policy actions in the PRSP. The Bank and its donor partners are as- Poverty and Social Sector Focus. Management sisting countries in developing monitoring and agrees with OED's findings that the PRS process evaluation systems through a variety of instru- has (a) helped increase the focus on poverty; (b) ments (for example, through the Trust Fund for heightened awareness of the need for improv- Statistical Capacity Building and the STATCAP ing monitoring systems; and (c) pulled sectoral program), as well as with financial and technical programs together under one strategic frame- support for sectoral and household surveys and work, making the link between sectoral strate- strengthening of monitoring and evaluation sys- gies and poverty reduction more tangible. tems. The Bank is working to deepen cross-coun- Management also acknowledges that some early try learning by preparing a set of country case PRSPs focused primarily on the social sectors-- studies designed to outline the strengths and probably for many reasons, ranging from the weaknesses of different institutional frameworks link to the HIPC Initiative to the fact that relatively for monitoring and evaluation systems. These stronger diagnostics were readily available in case studies will be used to prepare good prac- the social sectors. tice and options notes for developing countries 8 4 A N N E X F : M A N A G E M E N T R E S P O N S E to consider as they strengthen their poverty mon- strengthen country ownership of these goals and itoring systems. to continue to enhance the PRSP as an operational and country-led framework for achieving them. Analytic Base. Management also agrees that there Furthermore, donors will also need to make is a need to continually upgrade the analytic demonstrable progress in providing more and base on which PRSPs are developed and imple- better aid to support countries with sound PRSPs. mented. The annual joint Bank-Fund PRSP im- plementation progress reports have identified a Donor Alignment. Management shares OED's view number of areas in which PRSPs' analytic un- that donor programs need to be better aligned derpinnings have been weak: for example, bet- with PRSP priorities, especially in countries where ter understanding the determinants of growth the government-donor dialogue was weak before and micro-macro linkages and trade-offs. The the PRS process. Management also agrees that OED review identified the same issues. While re- better prioritization of PRSPs would help ensure cent PRSPs have shown much greater apprecia- effective alignment of the content of donor pro- tion of the centrality of growth to poverty grams. Management is nonetheless encouraged reduction, the analytical basis to understand the that the observed increase in coordinated donor policy levers for accelerating growth and for budget support--in Tanzania and Uganda in par- making it more pro-poor have been more elusive. ticular--constitutes evidence of improved donor Strengthening analysis of these issues will re- alignment in these countries that has been fa- quire long-term capacity building in countries, cilitated by the PRS process. Such coordinated development of more quantitative tools, and budget support is enhancing selectivity by fo- sustained support from development partners. cusing assistance on key expenditure, sectoral, On growth, with several partners the Bank is and institutional policy reforms, increasingly using 14 country case studies and cross-country moving away from the retail physical invest- empirical analysis to better understand how ments of the past and facilitating changes in country conditions and growth strategies affect both the content of donor programs and in the distributional impact of growth on the poor. donor processes. Management also notes that The insights will be used to strengthen country- there is emerging evidence of reduced transac- specific analytic support. The Bank is also re- tion costs of donor assistance in these coun- viewing growth experiences of the 1990s to tries, as noted in other donors' evaluations of the better understand the growth process and pro- PRS process. In Cambodia, Tanzania, and Uganda, vide more effective country-level advice. On the Bank's work with several donors on joint trade issues, the Bank and other donors are sup- assistance strategies should help to reduce trans- porting diagnostic work through the Integrated action costs for government officials. The ex- Framework for Trade and Development. The panding use of SWAps in some PRSP countries Bank is also working to help countries main- also holds promise for reducing transaction stream Poverty and Social Impact Assessments costs. Management acknowledges, however, that (PSIAs) through a combination of staff training very substantial progress in donor harmonization programs, knowledge sharing, and financial sup- and alignment has so far been limited to a hand- port through trust funds and dedicated budget ful of PRSP countries, and that significantly more allocations. As a result, some 65 Bank-supported effort is required to make it more of a reality else- PSIAs are being carried out in around 48 IDA-el- where. Management agrees that PRSPs by them- igible countries. selves will not solve the donor harmonization problem, but they provide a useful and increas- Partnership and Donor Alignment ingly used tool around which to build greater Management agrees that continued efforts are donor alignment and harmonization. Efforts are needed to further reinforce the results focus of already under way in some countries to assist in the PRS process. With a global compact to achieve defining partnership frameworks under PRSPs, the MDGs by 2015, a key challenge is to with support from the Bank and other donors. 8 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 : A N I N D E P E N D E N T E VA L U AT I O N Management will seek to expand these efforts to ment finds a reasonably robust link between a others. In parallel, Management has a major pro- country's PRSP and the Bank's lending program. gram at the corporate level to promote global It is true, as the OED review points out, that harmonization efforts to facilitate implementa- the contents of many Bank programs have not tion of the partnership frameworks. With other changed radically in response to PRSPs. The ex- donors, the Bank is monitoring progress at the panding use of PRSCs in a number of countries country level. does, however, reflect efforts to leverage the Bank's assistance for PRSPs by using an instru- Bank Alignment with PRSPs ment better suited to clients' multisectoral Management would like to raise an issue with re- poverty reduction challenges. As noted above, gard to the alignment of Bank assistance pro- Management agrees nonetheless that PRSPs need grams with PRSP priorities. Management agrees to be better prioritized and remains committed with the review's finding that the Bank has been to supporting this outcome. responsible for leveraging the PRSP as an agreed coordinating framework for donor activities. At OED Recommendations the same time, the review's recommendations Detailed responses to OED's specific recom- imply that the Bank is not yet aligning CAS timeta- mendations are attached in the Management bles and timeframes to PRSPs; that CASs do not Action Record. yet take into account other partners' activities; and that the Bank has not yet indicated how the Conclusions content of CAS programs should be adapted to Bank Management welcomes the OED review of PRSPs. In this connection Management notes the PRS process and the opportunity to discuss that since July 2002 the Bank has required that the process as it moves forward. Monitoring and CASs for IDA-eligible countries be prepared on evaluation are built into the process, and the the basis of PRSPs, and that they discuss how the OED review and companion IEO review con- content of the proposed program supports the tribute to our ability to continue enhancing the PRSP. Of the 28 countries that have completed PRS Initiative. Management agrees with the broad PRSPs/I-PRSPs since then, 27 of them had CASs conclusion that the PRSP holds the potential to approved after the PRSP completion date. Like- enhance poverty reduction in low-income coun- wise, CASs increasingly contain an explicit dis- tries and should continue to be supported by the cussion of selectivity in the Bank's assistance, Bank. Bank Management is fully committed to considering other donors' programs and Bank continuing to support the PRS process, en- comparative advantage. CASs demonstrate couraging countries to tailor the process and alignment by explaining how specific aspects of content to country conditions, supporting the in- the assistance programs support PRSP pillars; stitutionalization of participation and the deep- and, as the OED review notes, the content of ening of the analytic underpinnings, notably post-PRSP CASs overlaps with the content of with regard to the growth agenda, and working PRSPs. On the basis of its own review of post- to support more robust mechanisms for donor PRSP CASs for IDA-eligible countries, Manage- alignment. 8 6 A N N E X F : M A N A G E M E N T R E S P O N S E Management Action Record OED Recommendation Management Response 1: The Bank should help foster better customiza- Management agrees that the PRS Initiative should be tailored to country cir- tion of the Initiative to country circumstances and cumstances, and the PRS architecture already provides for this. Manage- more focus on improving long-term processes. ment notes that, through the JSA and other mechanisms, the Bank already · Greater emphasis should be placed on im- provides feedback to countries on where PRSPs are weak and suggests pri- proving country processes for planning, im- orities for the coming year. However, Management believes that it would plementing, and monitoring public actions be against the spirit of the PRS process to introduce further process condi- geared toward poverty reduction, and less on tionality. In the interest of helping countries customize the PRS process, the completion of documents. Toward this goal, the Bank will continue to assist countries that want to improve their planning, Bank should help countries identify what di- implementation, and evaluation processes. If they wish, countries could es- mensions of the PRS process present major tablish milestones for specific improvements on process constraints. Man- constraints to poverty reduction and establish agement agrees that initial conditions are important, but, because the PRS milestones for specific improvements in these framework is a process, it is feasible for countries to prepare PRSs in the areas, including capacity building. Clarification absence of full diagnostics and to use the PRS process to identify and, over of these milestones will help to equilibrate time, fill gaps. Management believes that setting a "minimum capacity" stan- stakeholder expectations and promote a longer- dard before allowing countries to complete PRSPs would undermine coun- term view and results orientation. Country au- try ownership. Management will, together with the IMF, take up issues related thorities should establish indicators and report to improvements in the JSA and APR processes in the forthcoming PRS im- them in PRSPs, APRs, and second-round PRSPs. plementation progress report (also see next point). The modifications to the · The Bank should not simply permit--but APR will include: supporting countries in linking it more explicitly to domestic should actively promote--tailoring of the PRS decision-making processes; considering elimination of the requirement that process to country conditions. It should ensure APRs and associated JSAs be presented to the Boards other than in ex- that the timing of progress reporting derives ceptional cases; and reducing the associated reporting requirements by fo- from local processes and that their content cusing it on improving underlying processes. Once the agreed improvements links to domestic decision-making. The Bank are introduced, Management will consider this action as complete. (and IMF) should demonstrate flexibility in the Initiative's handling of initial conditions, including cases where constraints suggest building a minimum capacity in key areas be- fore completing a PRSP. 2: The Bank should provide transparent and ef- Management agrees that the Bank should provide transparent and effective fective feedback to countries on their PRS feedback to countries on their PRS process and believes that it is doing so processes and strengthen the operational link be- through a variety of written and interactive instruments, including JSAs, CASs, tween PRSPs and assistance. analytic work, and project and policy dialogue. Many of these activities in- · Bank management should develop a procedure volve other stakeholders. Of course, like all processes, this one could be fur- for Board review of the PRSP that is more ther strengthened, and the Bank will look for opportunities to do so. transparently supportive of ownership and Management also believes that there is a strong operational link between more effectively linked to decisions about the PRSPs and Bank assistance programs in countries with PRSs, which is rein- Bank's program. This would both reduce the forced by the existing review process that examines the links between the perception of "Washington signing off" and CAS and the country's program. Again, this process can be strengthened; this promote ownership. The assessment of Bank is one of the aims of the results-based CAS, which is likely to be mainstreamed, and Fund staff of the soundness of the PRSP following the review upcoming in the first half of FY05. Management ac- could be directly incorporated in the CAS and knowledges an inherent tension between country ownership and the PRSP's 8 7 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 : A N I N D E P E N D E N T E VA L U AT I O N discussed directly in the context of the pro- operational link to BWI concessional assistance. However, Management be- posed program. This could strengthen the op- lieves that shifting Bank assessment of the PRSP to the CAS will not resolve erational link between the assessment and the this tension. Management does not believe that discontinuation of the JSA CAS and could allow the CAS to demonstrate would promote ownership and is concerned that this could undermine the more transparently how the Bank's planned as- relevance of the PRS process. Management believes that Board presenta- sistance is derived from the PRSP. tion of PRSPs and JSAs is a valuable step in the PRS process and should be · The Bank should provide feedback to the coun- continued. As noted above, Management does agree that there is scope to try on its PRSP in a form that is candid, trans- consider improvements to the JSA and APR processes. On the JSA instru- parent, analytically rigorous, and ment, in particular, the adjustments aim to: clarify its objectives and audi- comprehensive. The Bank should also provide ence; enhance the candor and nuance of the feedback it provides to countries feedback in a manner that strengthens part- and the Board; and enhance its contribution to the use of the PRS process nership, by involving other stakeholders. as a country-led partnership framework. Management will consider that · These objectives should be met either through this action is complete when the agreed changes are introduced. a major redesign of the JSA instrument and process or through discontinuation of the JSA and reliance on more interactive means such as stakeholder workshops, management notes, or other such devices. 3: The Bank, in concert with other partners, should Management agrees with the need to work in concert with other partners assist countries to strengthen analysis of the in strengthening analysis on the poverty impact of policies and programs poverty impact of policies and programs, and give and to give attention to growth in its assistance to PRS countries; and the more attention to growth. Bank is doing all this. The Bank works closely with other partners to help · The World Bank, in concert with other partners, countries strengthen the analytical underpinnings of the PRSP. The Bank is should help countries build the capacity to working to help PRS countries mainstream Poverty and Social Impact As- address key analytical gaps about the poverty sessments through a combination of staff training programs, knowledge shar- impact of policies and programs. Analysis ing with countries, and financial support--either through country budgets should address areas such as sources of or from trust funds. On growth, the Bank is carrying out a work program jointly growth, the quality of non-social sector strate- with several partners that will use 14 country-case studies and cross coun- gies, and the integration of the macro frame- try empirical analysis to better understand growth strategies and their ef- work and structural and social reforms. fect on distributional impacts. The Bank is reviewing growth experiences · Monitoring inputs, outputs, and outcomes of of the 1990s to draw lessons that will help it better understand the process on-going activities is critical to a better un- of growth and provide more effective country-level advice. On trade is- derstanding of what works and what does sues, the Bank and other partners support diagnostic work through the In- not. The Bank should assist in defining indi- tegrated Framework for Trade and Development. Assistance to countries to cators that are specific to country priorities, improve the results aspects of PRSPs continues to be a priority, with sup- realistic, and within current country capacity port provided through several key instruments. These include the Trust Fund to monitor and use. The monitoring and eval- for Statistical Capacity Building and STATCAP programs, as well as finan- uation scope can grow with capacity over time cial and technical support for sectoral and household surveys and for strengthened monitoring and evaluation systems. Results-based CASs help to focus country and Bank attention on country M&E systems, because of the need to monitor joint commitments. One of the elements that Man- agement looks for in a results-based CAS is a discussion of the country's results framework and how the Bank and other donors are supporting the country in this regard. It is likely that results-based CASs will be mainstreamed after the upcoming review of experience to date. JSAs and associated sec- 8 8 A N N E X F : M A N A G E M E N T R E S P O N S E tor dialogue already provide feedback on the realism of targets as well as on the quality of the proposed indicators and the institutional framework for collecting and using data. Management will consider this action as complete when the cross-country empirical analysis on growth is com- pleted, when all active IDA-eligible countries have completed a full set of core diagnostic ESW, expected by early FY06, and when results-based CASs are in place in at least 75 percent of active IDA-eligible borrowers (as- suming that results-based CASs are mainstreamed.). 4: The Bank, in concert with other donors, should Management agrees. This process already under way in several countries assist the country in defining a partnership frame- will be facilitated by the major ongoing worldwide efforts at donor align- work under the PRSP, with more explicit reference ment and harmonization. Management agrees, nonetheless, that significant to prioritization and the expected role of external challenges remain in making greater harmonization and alignment a real- partners. ity in most PRSP countries. The March 2005 Second High-Level Forum on · The Bank should be a leader in supporting Forum will consider time-bound targets for achieving key stages of align- country-led aid management by demonstrat- ment and harmonization. The Bank seeks to support country-led aid man- ing willingness to adjust to country-specific agement, in part by requiring that CASs follow PRSPs and, where PRSCs are timetables and processes. For example, the part of the country support program, increasingly aligning PRSCs to the coun- CAS should be completed after the PRSP and try budget cycles. As noted above, results-based CASs--which help to adopt its timeframe. strengthen selectivity, alignment, and coordination with other donors--are · The Bank should also define which process likely to be mainstreamed after the upcoming review. The Bank already pro- areas it will assist the country in addressing vides support to help countries improve the prioritization and costing of PRSP and the nature of its assistance in a trans- programs, notably through its country-level work with clients on develop- parent manner, including coordination with ing sectoral program budgets and medium-term expenditure frameworks. other stakeholders. Management notes that in addition to donor technical support, an individ- · The Bank should help countries improve the ual country's progress on prioritization requires a strong domestic consen- prioritization and costing of PRSP programs, sus around difficult choices, which is not easily influenced by external and demonstrate the alignment of its assis- actors such as the Bank. Management remains committed to aligning Bank tance with these improved PRSPs, taking into assistance to PRSPs, and will continue to demonstrate such alignment, in account other partners' activities and the particular through the CAS's articulation of how the content of the Bank pro- Bank's comparative advantage. gram supports PRSP priorities. Management will consider this action com- plete when results-based CASs based on PRSPs are in place in at least 75 percent of active IDA-eligible countries (assuming that results-based CASs are mainstreamed). 8 9 ANNEX G: CHAIRMAN'S SUMMARY: COMMITTEE ON DEVELOPMENT EFFECTIVENESS (CODE) On July 19, 2004, the Committee on Develop- the PRS Initiative as Washington-driven rather ment Effectiveness (CODE) met to discuss the than country-driven, particularly since the PRSP OED Review of the Poverty Reduction Strategy document is discussed by Executive Directors; (PRS) Process and Draft Management Response (iii) quality and process shortcoming of Joint to the OED Review of the Poverty Reduction Staff Assessments (JSAs) prevent this instrument Strategy (PRS) Process. The OED evaluation was from adequately fulfilling its important role of conducted in parallel with an evaluation by the providing feedback to the country and to the IMF's Independent Evaluation Office (IEO) en- Board; and (iv) donor alignment with PRSPs has titled Report on the Evaluation of Poverty Re- so far entailed few observable changes in the se- duction Strategy Papers (PRSPs) and the Poverty lectivity of donors' programs given that most Reduction and Growth Facility. The IEO re- PRSPs have a broad scope and limited prioriti- port, as well as Summaries of Ten Country Case zation. OED recommended improvements to Studies Undertaken as Part of the IEO Evalua- the Initiative to strengthen country ownership, tion of the PRSP/PRGF and the OED Review of foster flexible and customized implementation, Poverty Reduction Strategy (PRS) Process pre- improve the content of national strategies, and pared jointly by OED and IEO, were circulated enhance the partnership framework. as background information. Management Response. Management in general OED Evaluation Findings. The evaluation, built on agreed with OED's findings, which confirm those OED's earlier evaluation of the Highly Indebted of the PRSP implementation progress reports Poor Countries and of the CDF, found that the prepared by Bank and Fund staff for the Devel- PRS Initiative has made significant contributions opment and IMF Committees since 2000. Man- in many low-income countries, including on: agement highlighted that the PRS process is improving the poverty focus of national strate- relatively new, it is an evolving process, and the gies, integrating sector programs in an overall average length of implementation is less than strategy, and increasing attention to monitor- two years in countries that have completed PRSPs. ing. Notwithstanding the important progress There has also been continued progress in sev- achieved, the evaluation also found that the Ini- eral countries since the completion of OED re- tiative's potential has not been reached, and im- view. Management agreed that work is needed to provements are needed to foster sustainable strengthen the PRS process but differed on some processes at the country level to have a long-run of OED's recommendations on how to do so. impact. OED emphasized four specific areas for Specifically, Management believed that: (i) the PRS improvement: (i) the requirements and prac- architecture already allows flexibility and the pos- tices that the Bank has attached to the Initiative sibility of tailoring to country circumstances and have focused on the PRSP documents as the what is needed is more support for countries; (ii) main product, rather than on improving the there are several options for strengthening the planning, implementation, and evaluation process of discussion of the PRSP and JSA at the processes used by countries in their effort to re- Board, but discontinuing those discussions would duce poverty; (ii) many national stakeholders see be problematic, particularly in terms of main- 9 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 : A N I N D E P E N D E N T E VA L U AT I O N taining the operational relevance of the PRS ap- monization and coordination around the PRSPs, proach; and (iii) the PRS has quickly become taking into account OED findings and recom- central to the engagement of the Bank and other mendations related to the broad scope, and lim- external partners with client countries and is in- ited prioritization and costing of PRSP programs creasingly reflected in CASs as well as the support and their impact on donor alignment. programs of other donors. The Committee specifically requested Man- agement to improve the JSA process while look- Conclusions and next steps. Committee members ing for ways to strengthen country ownership, welcomed the OED review, which they found to and to present some suggestions in consultation be a frank and constructive assessment with with Fund staff in the annual PRSP Progress Re- some clear messages to be addressed by Man- port to be prepared for the Development and agement, donors and the Board. Some mem- IMF Committees for the 2004 Annual Meetings. bers commented that the OED review could That report will again emphasize the centrality have put more emphasis on the positive contri- of growth for poverty reduction in low-income butions of the PRS process given the existing cir- countries. Management also noted that the PRSP cumstances, although this does not mean that Progress Report in 2005 will include Bank and Management should be complacent. Overall, Fund staff assessments as well as the views of members found the Draft Management Response other stakeholders on a wider range of issues of to be balanced but felt it could have been more PRS content and process. forthcoming in addressing some of the issues Among the specific issues raised during the raised. Much of the discussion focused on per- Committee meeting were: ceptions of dilution of country ownership, the importance of growth to poverty reduction, part- Flexibility in Implementing the PRS Initiative. Several nership and donor alignment, and on ways to speakers supported OED's recommendation for improve the process and quality of the JSA. efforts by the Bank to promote customization and Management noted the recommendations flexibility to reflect country circumstances. Speak- from CODE members. Management will revise ers acknowledged the need to increase the em- its response to address more clearly several spe- phasis of the Initiative on public actions toward cific points identified at the meeting. The Com- reducing poverty rather than to focus on com- mittee proposed that the Management response pletion of documents and procedures. There clarify the following main issues: (i) how to en- was a broad sentiment in favor of reducing the hance country ownership, keeping in mind the process time, watching the costs involved, sim- limited support for eliminating the Board dis- plifying the PRS process, and linking it to the cussion of PRSPs. The present process of the country's longer-term development strategy and staff presenting the PRSP on behalf of the coun- the MDGs. One speaker indicated that the Bank try and then the JSA enhances the perception of should also encourage the MICs to address the "Washington signing" on a country-owned PRSP; PRS process. Some speakers noted that achieve- (ii) how to improve the JSA and staff's feedback ment of desired changes in the PRS process to countries and to the Board; (iii) how to make would require significant changes in the Bank's the PRS process more results-focused and assist internal culture, and the way it works with part- countries in understanding the poverty impact ners. They highlighted the importance of im- of policies, including policies to enhance growth; proving the PRS process in an ongoing dialogue (iv) how to apply the process more flexibly rather both within countries and between the Bank and than in a standardized way with uniform re- countries. Management responded that the Bank quirements; and (v) what are the other benefits is undergoing an important process of change of of the PRS process, which were mentioned by attitudes and internal organizational culture. participants but not adequately reflected in the OED report. Consideration should be given to PRS country ownership and JSA discussion. While the issue of donor agencies' support for har- many speakers agreed that the Bank (and Fund) 9 2 A N N E X G : C H A I R M A N ' S S U M M A R Y: C O M M I T T E E O N D E V E L O P M E N T E F F E C T I V E N E S S ( C O D E ) process for presenting a PRSP to the Board on be- better aligned with PRSP priorities and provide half of the country and also presenting the JSA may better support to, among other things, the do- undermine country ownership in the eyes of na- mestic planning and budgeting processes. They tional governments, other speakers stressed the commented on the need to strengthen the link- importance of Board discussion of PRSPs. Overall, age between the PRSP and the annual budgeting speakers expressed their chairs' openness to mod- process and medium-term expenditure frame- ify the current Board procedure to address this work. Many speakers also encouraged better issue to better reflect the perception of country alignment of donors' programs and a proactive ownership. Speakers made a range of suggestions role for the Bank in harmonization. A member to improve the process, including for the Execu- stressed the importance of the expected reduc- tive Director representing the country or the Gov- tion of transaction costs, and questioned whether ernment--via video conference--to present the these costs have been lowered. A few members PRS, and focusing the Board discussion on the JSA. observed that the OED review and management There was little support for discontinuing Board response seem to imply that alignment of donor discussion of JSAs, though a few questioned the assistance around the PRS means participation extent to which JSA is adding value to the process. in budget support. They pointed out that donor One speaker agreed with OED on assessing the alignment is possible by means of budget sup- soundness of PRSP within the CAS. Other speak- port or project aid, and accordingly suggested ers concurred with Management that it is impor- that the wording be reexamined. Management tant to elaborate on the feedback on PRSPs, in appreciated OED's comment on the need to im- particular to HIPC countries. One speaker stressed prove content alignment of the Bank's own pro- that the JSA should address alignment with the gram and to improve the partnership framework, country processes instead of focusing on or align- and indicated that they will explore this matter ing with the Bank-Fund processes. Management further. welcomed OED's message on the need, over time, for greater ownership, and OED and members' Participation. Several speakers highlighted the comments on improving the JSA process. importance of enhancing participation, while recognizing the diversity of countries' political PRSP Content and Analytical Underpinnings. Several structures, and the variety of approaches to con- members agreed with the OED finding that sev- sultation including the involvement of parlia- eral PRSPs have focused mainly on public ex- ments. At the same time, one member noted penditure and spending on social programs rather that the consultation process risked weakening than on broader policy actions for growth and key institutions of civil society. Another speaker poverty reduction. A few members asked the echoed OED's concerns regarding the lack of Bank to address the critical gaps that remain in clarity of the objectives of participatory activities, linking policies and programs to poverty impact. as understood by governments, civil society, and Several speakers suggested that the Bank focus on the Bretton Woods Institutions. While partici- capacity building, including analytical capacity; pation is beneficial and critical to the PRS process, strengthening the results focus; and promoting OED outlines concerns related to the sustain- private sector development, infrastructure, rural ability of participation, and the need to define the development, job creation, and revenue policy re- concept more transparently in order to inform form. Management noted that PRSPs, notably expectations across stakeholder groups. more recent ones, as opposed to I-PRSPs, have had more focus on policies for growth and that the Bank-Fund Collaboration. Several speakers stressed Bank will continue to support the necessary an- the importance of coordination among the Bret- alytical work at the country level. ton Woods Institutions to reduce the perception of "Washington signing" on a supposedly country- Partnership and Alignment. Members agreed with owned strategy. A few speakers commented on the OED findings that donor programs need to be need to synchronize the approaches of the Bank 9 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 : A N I N D E P E N D E N T E VA L U AT I O N and the Fund, improving the linkages between the suggested that the institutions should reduce the macroeconomic framework and the PRSP, and in- burden of conditionality on their counterparts in tegrating the short-term macroeconomic frame- the governments. Another member felt the issue work with long-term planning. Others cautioned of Bank-Fund collaboration was not well addressed that collaboration should not lead to increased em- in the OED review and needed to be more explicit phasis on short-term stabilization strategies at the on what each institution should do individually and expense of development dimensions. A speaker what they should do jointly. Chander Mohan Vasudev, Chairman 9 4 ENDNOTES Summary of Recommendations and Throughout this report, Poverty Reduction Strategy Findings Paper (PRSP) refers to the strategy document itself, 1. The OED evaluation was conducted in collabo- while Poverty Reduction Strategy (PRS) process refers ration with the Independent Evaluation Office (IEO) to the overall process of planning, implementing, and of the IMF, which is producing a parallel evaluation fo- monitoring/evaluating/adjusting the strategy. cused on the Fund's support to the Initiative. The key 5. The PRSP documentation explained that it would evaluative material includes 10 country case studies be part of a broader effort to "enhance the impact of (of which 4 were conducted jointly with IEO), a na- the Bank's work on poverty" by "promot[ing] sus- tional stakeholder survey conducted in the case study tainable growth and ensur[ing] that its benefits are countries, background papers, cross-country analyses, reaching the poor" by "practical operationalization" of and focus group sessions of Bank task managers, as the CDF (World Bank and IMF 1999b). See CDF Sec- well as existing evaluative material from within and out- retariat (2003) for a discussion of the distinctions be- side the World Bank. tween the CDF and the PRSP, highlighting the latter's links to resource conditionality and characteristic as an Chapter 1 "action plan" (rather than "a way of doing business"). 1. During the 1990s, significant poverty reduction 6. Interim PRSPs are intended to outline the gov- on a regional basis was achieved only in South and East ernment's commitment to, and strategy for, poverty Asia, which were also the only regions where real per reduction and to record commitments to a timeline and capita growth exceeded 3 percent. a consultative process leading to a full PRSP, which will 2. The International Development Goals first ap- have a more thorough discussion of country's progress peared in the OECD DAC report, Shaping the 21st Cen- toward poverty reduction (World Bank 2000a). tury: The Contribution of Development Cooperation 7. Implications of the Initiative for the Bank's work (OECD DAC 1996). These goals were updated in the in low-income countries were outlined in two policy Millennium Declaration in 2000, signed by 189 coun- papers: World Bank 2000; World Bank 2000c. tries (United Nations 2000). 8. PFPs were a tripartite agreement between the 3. For a full discussion of the origins of the CDF, government, Bank, and Fund, summarizing a country's see the recent OED-led evaluation, Toward Country- medium-term economic framework and covering main Led Development: A Multi-Partner Evaluation of the reform areas. The process of writing PFPs was widely rec- CDF (CDF Secretariat 2003). ognized as largely Washington-driven, with initial drafts 4. The PRSP Initiative was introduced in BWI pol- locally written in a minority of cases, but even then, they icy documents around the time of the 1999 Fall Meet- were often redone by IMF and World Bank staff. ings of the World Bank and IMF (World Bank and IMF 9. The Approach Paper for this review was en- 1999a,b). Further development of its objectives and dorsed by executive directors in February 2003 and is design features, as well as guidance for the support available at the OED external Web site. of the Bank and Fund, continued in several BWI-pro- (http://www.worldbank.org/oed/prsp/) (OED 2003a). duced documents in late 1999 and early 2000 (World 10. The OED evaluation was conducted in collabo- Bank and IMF 1999c, World Bank 2000a). Annex A pro- ration with the Independent Evaluation Office (IEO) of vides a detailed review of each paper's contribution the IMF, which is producing a parallel evaluation focused to the cumulative architecture of the PRSP Initiative. on the Fund's support to the Initiative (see the forth- 9 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 : A N I N D E P E N D E N T E VA L U AT I O N coming Report on the Evaluation of Poverty Reduction partnership, from which it was not statistically differ- Strategy Papers (PRSPs) and the Poverty Reduction ent. See Annex D. These results reinforce similar find- and Growth Facility (PRGF) (IEO forthcoming)). The ings from Toward Coountry-led Development: A case study methodology, developed jointly by OED Multi-Partner Evaluation of the CDF (CDF Secre- and the IEO, is presented in Annex B. Joint case stud- tariat 2003). ies were done for Mozambique, Nicaragua, Tajikistan, 14. This was also noted prominently in the case and Tanzania. These assess both World Bank and IMF studies of Toward Coountry-led Development: A support. OED conducted the case studies for Albania, Multi-Partner Evaluation of the CDF (CDF 2003). Cambodia, Ethiopia, and Mauritania, and IEO con- ducted the case studies for Guinea and Vietnam. Chapter 2 11. From the 81 IDA-eligible countries, OED es- 1. A volume of summaries of the 10 country case tablished a sample of 66 for analyzing data across studies is available at Individual case studies are available upon countries, and India. Blend countries were included. request. For purposes of comparing countries with completed 2. Selection criteria were maturity of the PRSP PRSPs with other low-income countries, these 66 coun- process, geographic balance, coverage of non-HIPC tries were divided into two groups: 35 countries that countries, and country initial conditions. Coverage pro- had completed a PRSP by the end of 2003 (referred to vided by case studies in recent or ongoing OED work in the report as "PRSP countries"), and 31 that had not was also a factor in the selection. A Multi-Partner ("Non-PRSP countries"). For some analyses, the PRSP Evaluation of the CDF (CDF Secretariat 2003) covered countries were subdivided into "Very Early," "Early," and the PRSP process in Bolivia, Burkino Faso, Ghana, "Late," depending on the date they completed their Uganda, and Vietnam. Debt Relief for the Poorest ­ An PRSPs. Annex C provides details of the PRSP status and OED Review of the HIPC Initiative (OED 2003b) con- OED's categorization of all 81 IDA-eligible countries. ducted case studies in Cameroon, Guyana, Malawi, 12. As mentioned in box 1.2, a survey of national Uganda, and Zambia. stakeholders was conducted in each of the 10 case 3. A representative from civil society and one from study countries. Roughly 1,000 questionnaires were the government were invited from each of the 10 sent out to stakeholders in the 10 case study countries. countries to attend the January 2004 multi-country The sample was targeted at individuals familiar with workshop in Addis Ababa. the PRSP process in government, civil society, inter- 4. The criteria for ownership draw on Borrower national NGOs, and external partner groups, and was Ownership of Adjustment Programs and the Political derived from information from the country case study Economy of Reform (Johnson and Wasty 1993). mission, participants listed in the PRSP document, 5. This was a consistent message that the evalua- and input from local consultants. There were 779 re- tion team heard from senior government officials in sponses, spread evenly across stakeholder categories. Mozambique, as well as from IMF and World Bank staff. Questionnaires were tracked at the stakeholder-group Respondents to the survey in Mozambique agreed level to ensure a response from key groups, but in- that the "PRSP is a good model" and that it "improves dividuals could choose to remain anonymous. A stan- on past modalities," but respondents were more neu- dard questionnaire was sent to each participant, and tral on aspects of participation--most notably the ad- the survey was self-administered. Local survey con- equacy of broad consultations with stakeholders. sultants and firms were used to collect and code the 6. These findings corroborate evidence from the data. Survey analysis was conducted in Washington by CDF evaluation (CDF Secretariat 2003) that the re- Fusion Analytics, in consultation with OED and IEO. source conditionality tied to PRSPs was a major mo- The survey, a description of the methodology, and an tivation in their preparation, causing rushed overview of the results can be found in Annex D. formulation and subsequent shortcomings in quality 13. Respondents assessed relevance of the PRSP ap- (for example, in participatory processes, analytical proach more highly than any other dimension. Its work, and overall institutionalization of the process). mean score of 3.5 was significantly higher than the This was observed through both HIPC conditionality composite scores for the five principles, excluding and IDA and PRGF linkages (CDF Secretariat 2003). 9 6 E N D N O T E S 7. At decision point, a country qualifies for debt re- (for example) the financing requirements for educa- lief and receives interim relief under the HIPC Initia- tion tripling during the process. For Mozambique, re- tive. At completion point, the country receives source requirements had (to a certain degree) been irrevocable debt relief. introduced under the Action Plan for the Reduction 8. As of September 2003 (World Bank and IMF of Absolute Poverty (PARPA). This was facilitated by the 2003, p. 2). ongoing Medium-Term Budget Framework, which 9. The formal policy position linking PRSPs to helped cost sector strategies. PRGF requests and reviews is set out in Poverty Re- 18. The mean score for partnership-oriented (3.33) duction and Growth Facility--Operational Issues was significantly higher, at a 95 percent level, than (IMF 1999a) and was endorsed by the Board in the con- those of the other principles. cluding remarks from that discussion (IMF 1999b). 19. The group is currently composed of donor 10. See IEO (2004) for a full discussion of the agency representatives from Canada, Denmark, Fin- shortcomings of participation in the macroeconomic land, Japan, Netherlands, Norway, Sweden, Switzer- framework. land, and the United Kingdom, with Germany and 11. Albania, Ethiopia, Mozambique, Nicaragua, Australia as regular attendees. The "like-minded" Tajikistan, Tanzania, and Vietnam. donors group has its origins in the "Utstein" group- 12. Civil society representatives in the case study ing of Norway, Netherlands, Germany, and the United countries commented that the BWIs tended to equate Kingdom. Representatives of these countries working consultation with participation. In a submission to in Vietnam began to meet informally in 2000 in an ef- the OED evaluation based on a survey of civil society fort to leverage "shared objectives centered on poverty views in 10 countries, CIDSE/CI, an alliance of Catholic reduction, improved aid effectiveness, and a collab- development organizations in Europe and North orative style of working." The group became a natu- America, reported that while there has been a slight ral forum for discussing PRSP issues. change in attitude and practice of Bank staff toward 20. Members reserve the right to depart from the participatory activities, the World Bank approaches par- PRSP framework in response to political and/or human ticipation as information sharing, or at best as con- rights concerns. sultation (CIDSE/CI 2003). 21. An independent evaluation of Swiss engage- 13. The decision to eliminate user fees in Tanzania ment in the PRSP process found that in some countries, was influenced both by NGOs, who had been lobby- PRSPs are influencing program planning. For instance, ing intensely for this during formulation of the PRSP, in Mozambique, processes have changed (the PRSP is and Bank staff. an indicator in the monitoring of the Swiss country pro- 14. See Annex D (Section D: Composite Results gram). In Vietnam, both process (using pooled finance Table). and working under the umbrella of a government min- 15. These elements are derived from BWI policy istry) and content (choice of geographic areas and sec- papers. tors for support) have changed. But in some cases, 16. Definitions of poverty include both income alignment of content is "coincidental" because of the and living standard measures such as the incidence, broad scope of the PRSP, as in Tajikistan, where the PRSP depth, and severity of income poverty; access to and SDC have both independently prioritized gover- health, education, and potable water; the quality of nance and health. The evaluation recommends that SDC housing; and poverty by demographic, marginalized country offices should be explicit about how program groups, and regions. content and the approach they are taking are appro- 17. This is based on the review of 25 PRSPs for coun- priate in the context of the PRSP (SDC 2003). tries in which the Bank subsequently completed a 22. A recent assessment of MDG Country Reports Country Assistance Strategy. In six of the case studies, by the UNDP's Evaluation Office reaches consistent costing and prioritization were either not done or findings for a broader group of countries. Their report deemed incomplete or very poor. Even in the strongest concludes, "There is need for convergence and cases, achievements are very preliminary. In Maurita- stronger links between the monitoring and reporting nia, the PRSP was very overoptimistic in what could processes of MDGs, Poverty Reduction Strategy Papers be achieved or financed, and also very volatile, with (PRSPs), and other comprehensive national develop- 9 7 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 : A N I N D E P E N D E N T E VA L U AT I O N ment frameworks and reporting instrument." (UNDP ism of the PRSP's analysis, and recommend future Evaluation Office 2003b). work. 23. The Ministry of Foreign Relations coordinates 30. The IEO review of JSAs covered an additional IDB and other donor funding, the Ministry of Finances five key questions raised in the JSA Guidelines, as supervises the use of some recurrent expenditures and well as JSA process issues mainly related to the IMF. is in charge of relations with the World Bank, the Na- 31. The use of APRs as a condition for IDA resources tional System of Public Investment handles invest- is less systematic. In one case study country, the APR ment expenditures and a considerable share of was used as a requirement for PRSC tranche release. recurrent spending, and the Central Bank registers for- 32. These two objectives are taken from the PRS eign currency transactions. Sourcebook (World Bank 2002a). 24. This question received the lowest mean score 33. As of February 2004, 18 APRs had been pro- from donors and was the only question in which their duced covering 12 countries (Albania, Burkina Faso, response was lower than that of international NGOs Ethiopia, Honduras, Malawi, Mauritania, Mozambique, and civil society. Nicaragua, Niger, Tanzania, Uganda, and Vietnam). 25. See the data on aid flows presented in the pre- Burkina Faso, Mauritania, Nicaragua, Tanzania, and liminary outcomes section of Chapter 3. Within the Uganda have multiple APRs. Three of the 10 case case studies countries, increased budget support was study countries have not completed an APR, though noted in Ethiopia and Mozambique. In Tanzania, the the case study for each includes discussion of an on- amount of Bank budget support lending has increased; going APR process. See Annex C for more details on as a share of total lending, however, it has declined. completion of APRs. Complete data on aid flows in Tanzania were not avail- 34. Targets do not cover all priority sectors: Tanzania able. has indicators in income, human development, and 26. The Bank and the Fund prepared PRSPs-- survival; Uganda in two of four pillars, income and qual- Progress in Implementation reports in April 2000, ity of life; Niger in education and health; Burkina Faso September 2000, April 2001, September 2001, Sep- in education, health, and water; and Mozambique in tember 2002, and September 2003 and conducted a education, health, roads, and water. joint review of the PRSP approach in April 2002. 35. Four countries prepared APRs in order to reach 27. The survey question on the quality of Bank- completion point under HIPC, and in Tanzania the sec- Fund collaboration ranked seventh-highest among ond APR was a required prior action for the PRSC. the 30 questions in the survey on the PRSP process, 36. For countries such as Cambodia and Tajikistan, both in terms of its mean score and the percentage the formulation of the PRSP consumed a large portion answering in the top two rankings (4 or 5). of available capacity in the planning and finance min- 28. The assessment of these two objectives in this istries and government at large, and the APR presents section is based on two in-depth content reviews of a similar challenge. The consultative processes re- JSAs conducted by OED and IEO, interviews with quired for the APR are an additional challenge for all Bank staff, evidence in the country case studies, and stakeholders. The level of consultations in APRs was feedback from PRSP stakeholders. weak in Albania, Mozambique, and Tanzania due in part 29. Annex E contains a discussion of the scope, to time constraints and/or the high costs of staging methodology, and findings of the JSA review. The large consultations. aim was to assess, using a four-point scale, how well each JSA identified the strengths and weaknesses of Chapter 3 its PRSP and signaled specific directions for im- 1. Joint Bank-Fund progress reports were presented provement. The assessment covered each JSA's treat- to the BWI Boards semi-annually from April 2000 to Sep- ment of 11 issues, as listed in figure 2.4. The issues tember 2002. Reports are now made on an annual were selected based on their importance in tran- basis, with the most recent in September 2003. scripts of Board discussions of a sample of JSAs and 2. This chapter draws from two background papers, in the course of the case studies. To obtain a high "Public Expenditure Management Linkages to the score, the JSA had to describe how the PRSP handled PRSP Process" (Bevan 2004) and "Capacity Enhance- the issue, assess the comprehensiveness and real- ment and the PRSP Process" (Kavalsky 2004)--and is 9 8 E N D N O T E S supplemented by cross-country reviews of the com- from just under 3 percent pre-PRSP to 4 percent of position of public expenditure in the 23 countries GDP post-PRSP. with more than a year's experience in implementation 10. See Annex C for the 12 countries that have is- of a PRS, a review of monitoring and evaluation in 12 sued APRs. countries that have completed Annual Progress Re- 11. The Global Monitoring Report 2004 notes that ports on PRSP implementation, and evidence from the prospects for achieving the MDGs are gravest in 10 OED/IEO country case studies. health. On current trends, the goals of reducing child 3. The others were the 1991/92 Household Budget and maternal mortality will not be attained in most Re- Survey, the 1993 Human Resource Development Sur- gions, and only a small proportion (15­20 percent) of vey, and the 1995 Participatory Poverty Assessment. countries appears to be on track. The risks of failure 4. One scenario assumes future constancy of the re- to halt the spread of HIV/AIDs are also daunting cent growth elasticity of poverty; another assumes that (World Bank and IMF 2004). the distributional properties of growth can be improved. 12. Analysis in this chapter uses the PRSP and non- 5. Data on actual "poverty reducing expenditures" PRSP country samples defined in Annex C. CPIA rat- are available from 1999 to 2002 from the World Bank ings are available only for 28 out of the 31 IDA and IMF HIPC units. The labeling of an expenditure countries without a PRSP. Afghanistan and Timor-Leste category as "poverty reducing" originated as a means did not have ratings for either 1999 or 2003, and Ser- to track HIPC debt relief. Initially figures on "social bia and Montenegro had ratings for 2003 only. spending" were reported in the annual World Bank and 13. The CPIA includes 20 equally weighted dimensions IMF HIPC ­ Status of Implementation reports for in the overall rating of policy encompassing economic countries that have reached decision point under the management (comprising economic policies), struc- HIPC Initiative (see, for example, World Bank and tural policies (broadly covering trade, financial sector, and IMF 2001b, Appendix tables 6 and 7). With the intro- regulatory policies that determine the enabling climate duction of the PRSP, staff report on "poverty-reducing for the private sector, as well as policies for environmental expenditures" (see, for example, World Bank and IMF sustainability), policies for social inclusion/equity (cov- 2003b, Appendix tables 4 and 5). ering policies for human resource development, gender, 6. Some countries include only primary education social protection, and equity of resource use), and pub- and basic health, while others include agriculture, lic sector management and institutions (comprising rural roads, and water and sanitation, and one coun- public financial management, quality of public admin- try incorporates infrastructure, governance, agricul- istration, control of corruption, and rules-based gover- ture, and macroeconomic management. nance) (World Bank and IMF 2004). 7. OED researched data on the 23 countries that 14. Data on aggregate aid flows are from the OECD- had completed a PRSP by end-2003, and thus had an DAC, which provides data on the volume, origin, and opportunity to implement their first post-PRSP budget. types of aid and other resource flows to over 180 re- Expenditure data were most widely available for the cipients. It includes both bilateral and multilateral ODA. education, health, agriculture, and transport sectors The database is available online at http://www.oecd.org/ in 14 countries of the total set of 23. dataoecd/50/17/5037721.htm. Analysis was conducted 8. To compare pre- and post-PRSP years, we on data available from the Web site in January 2004, cov- adopted a timeline in which observations were ag- ering the period through 2002. gregated in years ranging from T-6 to T+3, where 15. See Annex C for a list of the eight very early T+1 represents the first year of PRSP implementation countries. in each country. Expenditure data covering PRSP im- plementation were available only for Ethiopia, Guinea, Chapter 4 Mozambique, Tanzania, and Uganda, and consistently 1. The effectiveness of Bank support to the PRSP available only for four years before and one year after process was addressed in the eight case studies con- implementation. ducted either solely by OED or jointly with the IEO 9. Using all available data points from T-6 to T+3 (see table 2.1). The case studies of Guinea and Viet- for the same set of five countries, the mean annual ex- nam were conducted solely by IEO and did not address penditure on education showed a greater increase, Bank effectiveness. 9 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 : A N I N D E P E N D E N T E VA L U AT I O N 2. Analysis on Bank lending and non-lending in this program for the PRSP. In 8 of the 25 countries, the cost- chapter uses the PRSP and non-PRSP country samples ing in the PRSP was too vague or contained inconsis- defined in Annex C. tencies that prevented comparison of sectoral 3. See Annex C for a list of the countries. alignment, leaving a sample of 17 PRSP countries that 4. Total IDA lending to the 35 countries with PRSPs have all three sets of data available. by end-2003 rose from $13.1 billion in FY96­99 to 10. Of the 17 countries in the sample, 15 had multi- $15.9 billion in FY00­CY03. By comparison, total IDA sector adjustment credits in their pre- or post-PRSP lending to 31 other active IDA-eligible countries fell base case programs. Commitments for these credits from $13.1 billion in FY96­99 to $11.0 billion in were divided equally among the sectors covered. FY00­CY03. PRSPs were prepared at different times 11. Social sectors are defined as education, health, during the four years after the launch of the Initiative, social protection, and direct/targeted poverty inter- but for cross-country comparison purposes, OED ag- ventions. Growth includes infrastructure, PSD, agri- gregated the post-PRSP four-year period. culture, economic management, and financial. 5. In May 2001, the Bank created PRSCs as a new 12. There are four types of ESW: (1) Core Diagnostic programmatic lending instrument designed to support ESW: Country Economic Memorandums and Devel- IDA countries' medium-term policy and institutional opment Policy Reviews, Poverty Assessments, Public reform programs, as reflected in PRSPs, and to com- Expenditure Reviews, and, since FY01, Country Fi- plement the Fund's PRGF. PRSC programs include a nancial Accountability Assessments and Country Pro- series of adjustment operations that deliver budget curement Assessment Reviews; (2) other Diagnostic support based on prior policy actions and require ESW that addresses sector-specific or thematic issues: core diagnostic economic and sector work, including Institutional and Governance Reviews, Social Protec- an assessment of the country's development policies tion, Health, and Education Sector Reviews, and the and its public financial accountability arrangements. like; (3) Country Advisory Reports, which consist of The guidelines mention that the share of IDA lending more specialized products relevant to specific clients through the PRSC instrument may be as high as half, and specific topics: Commodities Studies, Debt and depending on country circumstances. Creditworthiness Studies, and so forth; and (4) other 6. There were approved PRSCs in 3 of the 10 case Non-Formal or Non-Assigned ESW: Policy Notes, work- study countries either before or during the bulk of the shops, and conferences. country-specific work (Albania, Tanzania, and Viet- 13. This question received a mean score of 3.45, but nam). Dialogue on preparation or appraisal of PRSCs this was not significantly lower than scores on other was ongoing during case work in an additional three Bank questions. countries (Ethiopia, Mozambique, and Nicaragua). 14. According to a World Bank presentation at a 7. There is limited information on co-financing in workshop for bilateral donors in October 2003, PSIAs PRSC program documents. In Sri Lanka, the Asian are ongoing in 38 PRSP countries. Development Bank is expected to contribute US$400 million during 2003­06. In Vietnam, four donors (Den- Annex B mark, the Netherlands, Sweden, and the United King- 1. For example, the extensive country studies by dom) contributed $49 million for the first PRSC, and the Multi-Donor Evaluation of the CDF covering coun- the first three contributed $33.7 million for the sec- tries with completed PRSPs (Uganda, Bolivia, Viet- ond PRSC. The second PRSC for Burkina Faso was co- nam, and Burkina Faso) contain a significant amount financed by Belgium specifically for the of information on the PRSP process in these countries. implementation of the Health Development Plan within the overall framework of the PRSP. Annex D 8. As documented in the joint Bank/IMF bench- 1. The five-point scales used in most questions of- marking of PEM systems in HIPC countries, Actions fered a range from 1: Completely Disagree to 5: Com- to Strengthen the Tracking of Poverty-Reducing Pub- pletely Agree. Respondents could also mark 0 for lic Spending in HIPCs (World Bank and IMF 2002b). Don't Know or Unsure. 9. The CAS base case scenario by sector is consid- 2. Fourteen stakeholder groups were identified: ered for Bank lending and the medium-term sectoral Government - central government, line ministries and 1 0 0 E N D N O T E S sector agencies, local government, Parliament ­ Civil African country (Mauritania, February 2001; Cameroon, Society ­ local NGOs, business sector, labor unions, ac- July 2003), a recent Asian JSA (Nepal, November 2003), ademia, media, religious organization, political party, and a Latin American JSA (Honduras, September other ­ International Partner ­ donor, international 2001). NGO. Results at the stakeholder group level will be pre- 3. The framework for the rating system is similar sented in the aggregate analysis across all countries. to the one employed in the IEO review. Annex E Annex F 1. While not one of the "key questions" of the JSA 1. These numbers refer to PRSPs, I-PRSPs, and Guidelines, JSA's are asked by the Guidelines to assess APRs completed by countries and transmitted to the risks to successful implementation of PRSPs (para- Bank and IMF. Not all have been discussed by the graph 4). Both the OED and IEO reviews assessed JSA Boards. treatment of risks, but the OED review used somewhat 2. See, for example, Oxfam 2004; Booth 2003. The different criteria, including how JSAs have assessed en- Bank's PRSP Web page currectly provides links to dogenous risks (see Annex B for more details). over 50 such reviews. 2. The sample of JSAs and Board comments was 3. Updates can be found at http://poverty.world generated by selecting an early and recent JSA for an bank.org/index.php?view=sub&id=3789. 1 0 1 REFERENCES ActionAid USA/ActionAid Uganda. 2004. Re- Reduction Strategy Papers (PRSPs) and thinking Participation--Questions for Civil Poverty Reduction Growth Facility (PRGF). Society about the Limits of Participation in Washington, D.C.: IMF. PRSPs. Discussion Paper. Washington, D.C. IMF (International Monetary Fund). 1999a. *Bevan, David. 2004. "Public Expenditure Man- Poverty Reduction and Growth Facility-- agement Linkages to the PRSP Process." Back- Operational Issues. IMF Staff Paper SM/99/293. ground Paper to OED Review of the PRS Washington, D.C. Process. OED, Washington, D.C. ------. 1999b. Concluding Remarks by the Booth, David. 2003. "Are PRSPs Making a Differ- Chairman of the IMF's Executive Board, ence? The African Experience." Development Poverty Reduction Strategy Papers ­ Opera- Policy Review 21. tional Issues and Poverty Reduction and Botchwey, Kwesi, Paul Collier, Jan Willem Gun- Growth Facility ­ Operational Issues. ning, and Koichi Hamada. 1998. External BUFF/99/154. Washington, D.C. Evaluation of the ESAF: Report by a Group of Johnson, John, and Sulaiman Wasty. 1993. Bor- Independent Experts. Washington, D.C.: IMF. rower Ownership of Adjustment Programs and CDF Secretariat. 2003. Toward Country-led De- the Political Economy of Reform. World Bank velopment: A Multi-Partner Evaluation of Discussion Paper No. 199. Washington, D.C. the Comprehensive Development Framework. *Kavalsky, Basil. 2004. "Capacity Enhancement Washington, D.C: World Bank. and the PRSP Process." Background Paper to CIDSE/CI (Coopération Internationale pour le OED Review of the PRS Process. OED, Wash- Développement et la Solidarité/Caritas Inter- ington, D.C. nationalis). 2003. "Submission to IEO/OED Maxwell, Simon. 2003. "Heaven or Hubris: Re- Evaluation of Poverty Reduction Strategy Pa- flections on the New Poverty Agenda." De- pers." Compiled and edited by Trócaire, velopment Policy Review 21(1). Maynooth, Ireland. McGee, Rosemary, and Andy Norton. 2000. Par- Danish Ministry of Foreign Affairs. 2003. "Re- ticipation in Poverty Reduction Strategies: view of Nordic Monitoring of the World Bank A Synthesis of Experience with Participatory and the IMF Support to the PRSP Process." Approaches to Policy Design, Implementa- Washington, D.C. Photocopied. tion and Monitoring. Institute of Develop- Datta, Lois-Ellin. 1990. Case Study Evaluations. ment Studies Working Paper 109. Surrey, U.K. U.S. General Accounting Office, GAO/PEMD- Morra, Linda, and Amy C. Friedlander. 1999. 91-10-19. Washington, D.C.: U.S. Government "Case Study Evaluations." OED Working Paper Printing Office. No. 2. Washington, D.C. IEO (Independent Evaluation Office, IMF). Forth- ODI (Overseas Development Institute). 2003. coming. Report on the Evaluation of Poverty "Are PRSPs Making a Difference? The African *Available through the OED Help Desk, by e-mail< eline@worldbank.org> or through the Web site 1 0 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 : A N I N D E P E N D E N T E VA L U AT I O N Experience." Development Policy Review 21 Development Economics, Development Data (2). Group. Washington, D.C.: World Bank. OECD DAC (Organisation for Economic Co- SDC (Swiss Agency for Development and Co- operation and Development, Development operation). 2003. Independent Evaluation Assistance Committee). 1996. Shaping the of SDC's Bilateral Engagement in the Poverty 21st Century: The Contribution of Develop- Strategy Paper (PRSP) Process. Evaluation ment Co-operation. Report adopted at the 2003/1, Evaluation and Controlling Division. Thirty-Fourth High Level Meeting of DAC. Berne: Gerster. Paris. United Nations. 2000. United Nations Millen- OED (Operations Evaluation Department, the nium Declaration. Adopted by the General World Bank). 2004. 2003 Annual Review of De- Assembly at the 8th Plenary Meeting in New velopment Effectiveness ­ The Effectiveness of York. http://www.un.org/millennium/declara Bank Support for Policy Reform. Robert J. tion/ares552e.htm. Anderson, task manager. OED Study Series. UNCTAD (United Nations Conference on Trade Washington, D.C.:World Bank. and Development). 2002. Economic Devel- ------. 2003a. "Approach Paper--OED Review of opment in Africa. From Adjustment to the Poverty Reduction Strategy Paper (PRSP) Poverty Reduction: What is New? New York Process." Available at http://www.world and Geneva: United Nations. bank.org/oed/prsp/#approach_paper. UNDP (United Nations Development Program) ------. 2003b. Debt Relief for The Poorest--An Evaluation Office. 2003a. Evaluation of OED Review of the HIPC Initiative. Madhur UNDP's Role in the PRSP Process. Volumes I Gautam, task manager. OED Study Series. and II. New York. Washington, D.C: World Bank. ------. 2003b. Millennium Development Goals ------. 2003c. 2002 Annual Review of Devel- Reports: An Assessment. New York. opment Effectiveness--Achieving Develop- United Republic of Tanzania. 1998. The Tanza- ment Outcomes: The Millennium Challenge. nia Development Vision 2025. Dar es Salaam. Soniya Carvalho, task manager. OED Study Se- Whaites, Alan. 2000. PRSPs: Good News for the ries. Washington, D.C.: World Bank. Poor? Social Conditionality, Participation Oxfam. 2004. From `Donorship' to Ownership? and Poverty Reduction. Milton Keynes, U.K.: Moving Towards PRSP Round Two. Oxfam World Vision International. Briefing Paper 51. Washington, D.C.: Oxfam In- Wilks, Alex, and Fabien Lefrançois. 2002. Blind- ternational. ing with Science or Encouraging Debate? ------. 2001. Making PRSPs Work: The Role of How World Bank Analysis Determines PRSP Poverty Assessments. Washington, D.C.: Oxfam Policies. London: Bretton Woods Project and International. World Vision International. ------. 1995. The Oxfam Poverty Report. Kevin World Bank. 2003a. World Development Indi- Watkins, ed. Oxford: Oxfam U.K. cators. Washington, D.C. Sanchez, Diana, and Katherine Cash. 2003. Re- ------. 2003b. Supporting Sound Policies with ducing Poverty or Repeating Mistakes? A Civil Adequate and Appropriate Financing. Wash- Society Critique of Poverty Reduction Strat- ington, D.C. egy Papers. Church of Sweden Aid, Diakonia, ------. 2002a. A Sourcebook for Poverty Re- Save the Children Sweden, and the Swedish duction Strategies--Core Techniques and Jubilee Network. Brussels: EURODAD. Cross-Cutting Issues. Volume 1. Jeni Klugman, Sen, Amartya. 1999. Development as Freedom. ed. Washington, D.C. New York, N.Y.: Alfred A. Knopf. ------. 2002b. "Operationalizing Bank-Fund Col- Swanson, Eric Valdeman, and others. 2002. Tar- laboration in Country Programs and Condi- gets and Indicators for MDGs and PRSPs: tionality--Staff Guidance Note." In What Countries Have Chosen to Monitor. Strengthening IMF­World Bank Collabora- 1 0 4 R E F E R E N C E S tion on Country Programs and Condition- ------. 2002a. Strengthening IMF­World Bank ality--Progress Report. Washington, D.C. Collaboration on Country Programs and ------. 2001. Assessing Aid : What Works, What Conditionality--Progress Report. Washing- Doesn't, and Why. Washington, D.C. ton, D.C. ------. 2000a. "Poverty Reduction Strategy Papers: ------. 2002b. Actions to Strengthen the Track- Internal Guidance Note." Washington, D.C. ing of Poverty-Reducing Public Spending in ------. 2000b. "Guidelines of Joint Staff Assess- HIPCs. Washington, D.C. ment of a Poverty Reduction Strategy Paper." ------. 2002c. "Senegal Joint Staff Assessment." Available at http://poverty.worldbank.org/files/ Washington, D.C. jsa%20guidelines.pdf. Washington, D.C. ------. 2001a. Strengthening IMF­World Bank ------. 2000c. Supporting Country Develop- Collaboration on Country Programs and ment in Low- and Middle-Income Countries. Conditionality. Washington, D.C. Washington, D.C. ------. 2001b. Heavily Indebted Poor Coun- ------. 1998. Aid and Reform in Africa: Lessons tries (HIPC) Initiative--Status of Implemen- from Ten Case Studies. New York: Oxford Uni- tation. Washington, D.C. versity Press for the World Bank. ------. 1999a. Strengthening the Link Between World Bank and IMF. 2004. Global Monitoring Debt Relief and Poverty Reduction. Wash- Report 2004. Washington, D.C. ington, D.C. ------. 2003a. Poverty Reduction Strategy Pa- ------. 1999b. Building Poverty Reduction pers: Detailed Analysis of Progress in Imple- Strategies in Developing Countries. Wash- mentation. Washington, D.C. ington, D.C. ------. 2003b. Heavily Indebted Poor Coun- ------. 1999c. Poverty Reduction Strategy Pa- tries (HIPC) Initiative--Status of Implemen- pers--Operational Issues. Washington, D.C. tation. Washington, D.C. World Vision. 2002. Masters of Their Own De- ------. 2003c, 2002c, 2001c,d, 2000a,b. PRSPs-- velopment? PRSPs and the Prospects for the Progress in Implementation. Washington, D.C. Poor. Alan Whaites, ed. Milton Keynes, U.K.: ------. 2003d. "Sri Lanka Joint Staff Assessment." World Vision International. Washington, D.C. 1 0 5 OED PUBLICATIONS Study Series 2003 Annual Review of Development Effectiveness: The Effectiveness of Bank Support for Policy Reform Agricultural Extension: The Kenya Experience Assisting Russia's Transition: An Unprecedented Challenge Bangladesh: Progress Through Partnership Bridging Troubled Waters: Assessing the World Bank Water Resources Strategy The CGIAR: An Independent Meta-Evaluation of the Consultative Group on International Agricultural Research Debt Relief for the Poorest: An OED Review of the HIPC Initiative Developing Towns and Cities: Lessons from Brazil and the Philippines The Drive to Partnership: Aid Coordination and the World Bank Financial Sector Reform: A Review of World Bank Assistance Financing the Global Benefits of Forests: The Bank's GEF Portfolio and the 1991 Forest Strategy and Its Implementation Fiscal Management in Adjustment Lending IDA's Partnership for Poverty Reduction Improving the Lives of the Poor Through Investment in Cities India: The Dairy Revolution Information Infrastructure: The World Bank Group's Experience Investing in Health: Development Effectiveness in the Health, Nutrition, and Population Sector Jordan: Supporting Stable Development in a Challenging Region Lesotho: Development in a Challenging Environment Mainstreaming Gender in World Bank Lending: An Update The Next Ascent: An Evaluation of the Aga Khan Rural Support Program, Pakistan Nongovernmental Organizations in World Bank­Supported Projects: A Review Poland Country Assistance Review: Partnership in a Transition Economy Poverty Reduction in the 1990s: An Evaluation of Strategy and Performance Power for Development: A Review of the World Bank Group's Experience with Private Participation in the Electricity Sector Promoting Environmental Sustainability in Development Reforming Agriculture: The World Bank Goes to Market Sharing Knowledge: Innovations and Remaining Challenges Social Funds: Assessing Effectiveness Uganda: Policy, Participation, People The World Bank's Experience with Post-Conflict Reconstruction The World Bank's Forest Strategy: Striking the Right Balance Zambia Country Assistance Review: Turning an Economy Around Evaluation Country Case Series Bosnia and Herzegovina: Post-Conflict Reconstruction Brazil: Forests in the Balance: Challenges of Conservation with Development Cameroon: Forest Sector Development in a Difficult Political Economy China: From Afforestation to Poverty Alleviation and Natural Forest Management Costa Rica: Forest Strategy and the Evolution of Land Use El Salvador: Post-Conflict Reconstruction India: Alleviating Poverty through Forest Development Indonesia: The Challenges of World Bank Involvement in Forests Uganda: Post-Conflict Reconstruction Proceedings Global Public Policies and Programs: Implications for Financing and Evaluation Lessons of Fiscal Adjustment Lesson from Urban Transport Evaluating the Gender Impact of World Bank Assistance Evaluation and Development: The Institutional Dimension (Transaction Publishers) Evaluation and Poverty Reduction Monitoring & Evaluation Capacity Development in Africa Public Sector Performance--The Critical Role of Evaluation Multilingual Editions Allègement de la dette pour les plus pauvres : Examen OED de l'initiative PPTE Appréciation de l'efficacité du développement : L'évaluation à la Banque mondiale et à la Société financière internationale Determinar la eficacia de las actividades de desarrollo : La evaluación en el Banco Mundial y la Corporación Financiera Internacional Côte d'Ivoire : Revue de l'aide de la Banque mondiale au pays Filipinas: Crisis y oportunidades Reconstruir a Economia de Moçambique : http://www.worldbank.org/oed TMxHSKIMBy359723zv":;:^:(:# THE WORLD BANK ISBN 0-8213-5972-X