26545 COUNTRY ASSISTANCE STRATEGIES: RETROSPECTIVE AND IMPLICATIONS OPERATIONS POLICY AND STRATEGY MAY 1, 2000 ACRONYMS ADB Asian Development Bank AfDB African Development Bank APL Adaptable Program Loan ARPP Annual Report on Portfolio Performance CAE Country Assistance Evaluation CAN Country Assistance Note CAR Country Assistance Report CAS Country Assistance Strategy CASE Country Assistance Strategies and the Environment Program CDF Comprehensive Development Framework CFAA Country Financial Accountability Assessment CG Consultative Group CODE Committee on Development Effectiveness CPIA Country Policy and Institutional Assessment CPAR Country Procurement Assessment Report CPPR Country Portfolio Performance Review DAC Development Assistance Committee DEC Development Economics Department EBRD European Bank for Reconstruction and Development ED Executive Director ESAF Extended Structural Adjustment Facility ESW Economic and Sector Work EU European Union FSD Financial Services Department GDP Gross Domestic Product HD Human Development HNP Health, Nutrition, and Population HIPC Heavily Indebted Poor Countries IBRD International Bank for Reconstruction and Development IDA International Development Association IDB Inter-American Development Bank IFC International Finance Corporation IMF International Monetary Fund LIL Learning and Innovation Loan MDB Multilateral Development Bank MIGA Multilateral Investment Guarantee Agency NEAP National Environmental Action Plan NGO Nongovernmental Organization OECD Organization for Economic Cooperation and Development OED Operations Evaluation Department OPS Operations Policy and Strategy PA Poverty Assessment PER Public Expenditure Review PR Progress Report (CAS) PREM Poverty Reduction and Economic Management Network PRSP Poverty Reduction Strategy Paper PSD Private Sector Development PSM Public Sector Management PSS Private Sector Strategy Annex QAG Quality Assurance Group SSP Sector Strategy Paper SSR Social and Structural Review TAL Technical Assistance Loan UNDAF United Nations Development Assistance Framework UNDP United Nations Development Programme WAEMU West African Economic and Monetary Union WBG World Bank Group WDR World Development Report COUNTRY ASSISTANCE STRATEGIES: RETROSPECTIVE AND IMPLICATIONS CONTENTS Page Executive Summary ...............................................................................................................iii I. Introduction ................................................................................................................. 1 II. Country Issues: Candid Depiction ............................................................................ 3 A. Country Focus ..................................................................................................... 3 1. Government Priorities ..................................................................................... 3 2. Country Development Context ....................................................................... 5 B. Participation ......................................................................................................... 6 C. CAS Disclosure .................................................................................................. 10 III. Network Issues: Comprehensive—but Prioritized—Diagnosis ........................... 11 A. Poverty Reduction .............................................................................................. 11 B. Traditional Areas................................................................................................ 12 C. Cross-cutting Issues............................................................................................ 14 1. Assessment .................................................................................................... 16 2. Summing Up ................................................................................................. 20 IV. Business Issues: Selective Program ........................................................................ 22 A. Selecting Strategically........................................................................................ 22 1. Magnitude of Potential Impact on the Ground.............................................. 22 2. Likelihood of Successful Government Action .............................................. 24 3. Bank Group Comparative Advantage ........................................................... 25 B. Calibrating Program Size ................................................................................... 32 1. Levels of Bank Support................................................................................. 33 2. Lending Programs ......................................................................................... 34 3. Triggers ......................................................................................................... 34 C. Managing Risks.................................................................................................. 36 D. Keeping Score .................................................................................................... 38 E. Checking Progress.............................................................................................. 41 V. Conclusions and Follow-up ...................................................................................... 44 Annexes Annex A. CAS List.................................................................................................... 51 Annex B. Near-CASs ................................................................................................ 52 Annex C. CAS Costs ................................................................................................. 54 Annex D. Cross-cutting Issues................................................................................... 55 Annex E. Sectoral Selectivity.................................................................................... 71 Boxes, Figures, and Tables Box 1. CASs over Time ......................................................................................................... 2 Box 2. External Environment ................................................................................................ 6 Box 3. Regional Issues........................................................................................................... 7 Box 4. Participation in the CAS............................................................................................. 8 Box 5. Poverty Reduction in the CAS ................................................................................. 13 Box 6. Human Development in the CAS............................................................................. 15 Box 7. CASs and the Environment ...................................................................................... 19 Box 8. Sector Strategy Papers and the CAS ........................................................................ 21 Box 9. Selectivity in the CAS .............................................................................................. 23 Box 10. Partnerships .............................................................................................................. 26 Box 11. Multilateral Development Institutions: Country Strategies ..................................... 28 Box 12. UNDAF Pilot Countries―Mali and Vietnam.......................................................... 29 Box 13. Track Record and Portfolio Management in the CAS ............................................. 29 Box 14. Integrating Findings from OED Evaluations into the CAS...................................... 30 Box 15. IDA Allocations and Country Performance ............................................................. 33 Box 16. Self-evaluation in the CAS....................................................................................... 40 Box 17. Evaluation of CAS Progress Reports ....................................................................... 42 Box D1. Gender in the CAS................................................................................................... 58 Box D2. Governance in the CAS ........................................................................................... 60 Box D3. Environment in the CAS.......................................................................................... 63 Box D4. Financial Sector in the CAS .................................................................................... 68 Box D5. Private Sector in the CAS ........................................................................................ 70 Figure 1. Cross-cutting Issues: Unsatisfactory CAS Treatment ............................................ 16 Figure 2. Historical and Projected GDP Per Capita Growth.................................................. 24 Figure 3. Content of Triggers by Type................................................................................... 35 Figure 4. Treatment of Risk in CASs..................................................................................... 36 Table E1. Sectoral Focus and Program Composition in FY98-99 CASs .............................. 75 Table E2. Sectoral Selectivity in CASs by Country Groups ................................................. 76 Country Assistance Strategies: Retrospective and Implications EXECUTIVE SUMMARY I. Introduction 1. This report takes stock of where we are in the CAS process, as a basis for learning lessons and for setting the agenda for future CAS evolution. It was discussed by the Executive Board Committee on Development Effectiveness on January 5, 2000, and has been revised to take into account the comments received at the meeting and subsequently. It examines the 37 CASs and 16 CAS Progress Reports that were discussed by the Board during the second half of FY98 and all of FY99. It builds on the first CAS Retrospective (CAS Retro I), which looked at the CASs discussed by the Board in FY97 and the first half of FY98.1 That document zeroed in on the inherent tension in the CAS between the need for comprehensiveness, on the one hand, and selectivity, on the other. It concluded that CASs needed to be diagnostically comprehensive, but programmatically and strategically selective. This document looks at how well we have done in managing this tension, and proposes strategic directions for future CASs. 2. Continuing Improvement. Continuing the trend identified in the last retrospective, CAS quality improved along several dimensions during the review period. CASs improved most in their treatment of participation and governance and of the social and political economy underpinnings of reform. Treatment of poverty reduction and human development also improved, especially in CASs for IDA-eligible countries. The treatment of gender, the environment, and the financial and private sectors has shown less progress. There were significant improvements in the building blocks of CAS program design and business strategy— portfolio management, lessons learned, triggers, and partnership. But CAS risk analysis has not improved. There is considerable scope for further improvements in individual CASs and for consolidation of the improvements across CASs; and there is a need for some sector boards to do better at getting their messages out to country teams. Efforts in these directions will need to continue. 3. Changing Context. However, the main conclusion of this review is that the CAS improvement agenda must begin to focus elsewhere. The challenge will be to “unpack” the CAS, which has become overloaded, and to refocus it sharply on the Bank Group program. The Comprehensive Development Framework (CDF)2 and the Poverty Reduction Strategy Paper (PRSP), though still in their very early days, are already changing the CAS’s operating environment. In countries where the CDF or a PRSP (or another country document) provides the country context, the CAS can become more of a Bank Group business strategy. This vision was spelled out in the CDF Internal Guidance Note3 earlier in 1999, and makes sense given both the 1 See Country Assistance Strategies: Retrospective and Outlook (SecM98-242), March 30, 1998. 2 The CDF is described in James D. Wolfensohn, A Proposal for a Comprehensive Development Framework, World Bank, January 21, 1999; and Joseph Stiglitz, Towards A New Paradigm For Development: Strategies, Policies and Processes, 1998 Prebisch Lecture, UNCTAD, Geneva, October 19, 1998. 3 See Comprehensive Development Framework: Internal Guidance Note (SecM99-244), April 12, 1999. iv increasing centrality of country focus and ownership in the development paradigm and the proliferation of CAS-like donor documents that is straining scarce government capacity. 4. Emerging Challenges. Such a shift will not happen overnight. A gradual transition period is envisioned during which learning by doing will take place, as countries prepare their PRSPs and/or CDF pilot programs, and CASs focus more on the Bank’s business strategy.4 The transition will require considerable work—including efforts to ensure the right balance among the PRSP/CDF, diagnostic economic and sector work (ESW), and the CAS itself. A stocktaking on ESW, now under preparation, will examine the Bank’s various ESW products and their uses, and will indicate the steps Management is taking to improve their quality, management, and strategic impact. It will provide an opportunity to consider which diagnostic products are needed to underpin the CAS and Bank lending programs, and in what timeframe. Care also will be needed to meet the IDA12 CAS commitments for IDA-eligible borrowers. But the strategic direction is clear: CAS documents should focus more on the problem areas in the diagnosis, on the selectivity whys and wherefores of the Bank Group’s program, and on the benchmarks for judging progress and Bank performance. 5. Organization of this Report. The organization of the report closely tracks CAS Retro I, with main sections on country focus, cross-cutting diagnosis, and program selectivity. This is in keeping with the advice provided by Executive Directors’ advisers and assistants at a technical briefing on CAS Retro II in April 1999. As in CAS Retro I, the focus is on CAS quality at entry, not on implementation. Broadly speaking, CASs were evaluated using the same criteria as in 1998; but in several areasincluding poverty, governance, and environmentmore demanding criteria were used this time, reflecting increasing expectations in these areas. The Networks and other central units contributed to the retrospective through desk reviews of CAS documents. II. Country Issues: Candid Depiction 6. The success of the CAS process in recent years owes much to the Bank’s renewed country focus under the Strategic Compact, and to the central role played by country directors in managing country relationships. This has led to a clearer representation in CASs of country concerns and aspirations and a better sense of how the Bank can help. 7. Country Focus. Most CASs now do a good job of presenting the government’s vision. Virtually all include a detailed discussion of the progress on reform and the evolution of government priorities over time. In general, there is a good fit between the Bank’s priorities and the government’s. Speaking on behalf of the governments of the countries they represent, most Executive Directors fully supported the CAS programs when they were presented to the Board. But there also are differences between the Bank and clients, which CASs increasingly acknowledge with candor—one-fourth of the FY99 CASs do so, up from 10 percent of FY97 CASs and 20 percent of FY98 CASs. The most often cited differences between the Bank and the government concern the pace and scope of reforms. In two cases where the CAS highlighted major differences between the government and the Bank, the respective Executive Directors registered concerns about the CAS and the CAS process. 4 For an example of this approach, see Republic of Ghana Development Strategy for Poverty Reduction (Report No. 20186-GH), March 8, 2000, and Ghana Country Assistance Strategy (Report No. 30296-GH), March 8, 2000. v 8. Participation. Participation has largely been mainstreamed in the CAS process. Three quarters of the CASs reviewed involved a significant level of participation, compared to about half in CAS Retro I. The quality of the participatory process also is improving, with country teams experimenting with consultation techniques: they are moving out of capital cities and into rural areas, and they are forming stronger ties with local groups and NGOs. But we need to better focus our consultations, moving beyond multiple open-ended consultations and becoming more issues-oriented, drawing on upstream stakeholder analysis. We also need to integrate participation into more of our activities. Participation is not a one-off CAS preparation activity; rather, it needs also to underpin the preparation and implementation of projects and ESW. Of course, with decentralization to country offices, this kind of activity becomes easier and is taking place to an increasing extent. But in pursuing this agenda, staff need to keep in mind Bank policywhich is to encourage but not mandate participation. Clearly, participatory approaches must be pursued with sensitivity and with the prior agreement of the government; better practice is for the government to lead the participatory process. 9. Disclosure. Starting in 1996, as CASs began to be prepared in a participatory way, pressures for disclosure arose; indeed, it was difficult to ask civil society groups for their inputs into a document they would never see. Reflecting these concerns, a number of governments wanted to share the CAS with civil society. Less than halfway through the period reviewed in CAS Retro II, the Board approved the CAS disclosure policy, which took effect on August 1, 1998,5 permitting the Bank to disclose CAS documents at the request of the government. After the new policy took effect, the Board discussed 28 CASs and Progress Reports in the CAS Retro II cohort; 75 percent of them were disclosed.6 III. Network Issues: Comprehensive—but Prioritized—Diagnosis 10. As a starting point for selectivity, the CAS diagnosis must be comprehensive—but even it must be prioritized. With the goal of poverty reduction as the key driver, the CAS diagnosis needs to reflect a broad view of country conditions, as a basis for assessing where the need for remedial action is greatest. In principle, all Network areas should be covered in the comprehensive diagnosis. In practice, they need not be treated equally. More attention must be given to the areas in which country performance most constrains the pursuit and achievement of sustained poverty reduction. If a topic is not a major development issue in a particular country, the CAS should say so, stating the reasons. 5 See Enhancing Board Discussion of Country Assistance Strategies (R98-199), July 17, 1998. That note built on the May 12, 1998, Board seminar on Country Assistance Strategies: Retrospective and Outlook (SecM98- 242), March 30, 1998. It also reflected comments made at the Board Steering Committee meeting on June 1, 1998, and comments received directly from Executive Directors individually and from Executive Directors’ staffs at a technical briefing on July 1, 1998. For IDA CASs, a new policy was approved in the context of Additions to IDA Resources: Twelfth Replenishment—A Partnership for Poverty Reduction. This policy, which took effect July 1, 1999, requires that IDA CASs be made public (with sensitive information appropriately handled in consultation with the government), unless justified by exceptional circumstances as agreed in each case by the Executive Directors. 6 The Bank’s disclosure policy is discussed in World Bank Policy on Information Disclosure: Issues Paper, (CODE2000-26), April 18, 2000. vi A. Poverty Reduction 11. The poverty focus of CASs has continued to improve. CAS Retro I found that CAS treatment of poverty reduction had improved over time. By the standards used in rating CASs in the last retrospective, the share of CASs judged to represent good or better practice has risen. However, even with these standards there are still one or two IBRD CASs each year that receive unsatisfactory marks for their treatment of poverty reduction. Meanwhile, with the Bank’s increasing emphasis on poverty outcomes, more stringent criteria for evaluating the poverty focus of the CAS have been adopted. Whereas in 1998, the CAS was rated on its diagnosis of poverty and its poverty-focused strategy, the new criteria adopted by the Poverty Sector Board assess whether poverty reduction is at the center of the policy dialogue, whether the CAS assesses the impact of the Bank-supported program on poverty, and whether the CAS includes poverty-focused benchmarks and monitoring indicators. Not surprisingly, with the higher bar, fewer CASs pass: whereas with the old criteria two thirds of FY99 CASs are rated good or better, with the new criteria that share is rated only satisfactory or better. Renewed efforts to improve Bank performance in this critical area are under way, in the context of the PRSP process.7 B. Traditional Areas 12. Coverage of the “traditional” mainstays of Bank programs—macroeconomic analysis and growth, infrastructure and energy, rural development, and human development—remains strong in this cohort of CASs. Virtually all present a thorough and focused diagnosis of country economic issues and report on recent economic events; most identify macroeconomic issues as a priority; and all contain lending and/or nonlending activities in this area. Over one half of CASs also identify infrastructure as a priority, and almost all have infrastructure lending activities. One third also identify energy as a priority, and 60 percent have at least one lending or ESW activity related to energy in the work program. The majority of CASs address rural issues, but the quality of coverage is variable. It is best where rural development is linked with poverty reduction, and where the country has a rural development strategy in place. In the area of human development, a central pillar of the poverty reduction effort, CAS treatment has been improving. Between FY98 and FY99, the share of CASs with satisfactory or better treatment increased from 80 to 96 percent—and IDA CASs were 100 percent satisfactory. This trend in part reflects the very active engagement of the HD Sector Board in the CAS process, building on its sector strategy papers (SSPs) for health, nutrition, and population. C. Cross-cutting Issues 13. In addition to these traditional areas of involvement, CAS diagnosis has increasingly focused on a set of cross-cutting areas that are essential for poverty reduction—gender, governance, environment, the financial sector, and private sector development. How have CASs progressed in these areas? In governance there has been significant and continuing progress, with more recent CASs doing an increasingly better job in addressing governance and corruption issues. But on gender and environment, CAS diagnostic treatment is broadly unchanged from 7 See Poverty Reduction and the World Bank: Progress in Fiscal 1999 (R2000-41), March 30, 2000; and Poverty Reduction and the World Bank: Progress in Fiscal 1999 Implementation Note (R2000-41/1, April 13, 2000. vii the CAS Retro I cohort’s, although many CAS matrices show more activities in these areas than their treatment in the text would suggest; nevertheless, better CAS coverage is clearly needed. CAS treatment of the financial sector has improved, but still varies widely; this remains an area in need of strengthening. The previous retrospective did not analyze treatment of private sector issues, and therefore we lack a firm benchmark. That said, it is quite clear that joint CASs address private sector development—and financial sector issues—much better than other CASs do. Going forward, early involvement of IFC and MIGA in the CAS process, and continuing efforts to seek deeper integration of country strategies and work programs in selected countries, will be essential. 14. Next Steps. The CAS reform agenda must include improving the diagnostic coverage of the cross-cutting issues that are the country’s key development priorities or constraints. The upcoming SSPs for gender, financial sector, and public sector management will be important vehicles for the Networks to identify those countries for which CAS treatment of these areas should be expected, and to clarify the kinds of activities that make the most sense for the Bank Group to support. For countries where performance on the cross-cutting issues is satisfactory, the CAS can give them less comprehensive treatment, explaining why. For countries with a CDF or PRSP, the comprehensive diagnosis should be included therein or in companion ESW, with the results briefly summarized in the CAS business strategy. IV. Business Issues: Selective Program 15. A comprehensive diagnosis of country problems and priorities is but the start of the CAS process, whose strategic thrust is really meant to answer the questions, What should the Bank Group do? Why? and How should Bank Group performance be measured? To this end—in moving from diagnosis of the problem to design of a Bank-supported solutionthe CAS must take into account the Bank Group’s comparative advantage, as reflected in the role of other partners, the Bank Group’s track record, and the availability of instruments. It must determine the size of the lending and nonlending programs and their breakdown into individual activities. It must set the benchmarks by which the Bank performance will be judged. A. Strategic Selectivity 16. Strategic selectivity is essential if the Bank Group is to maximize the development impact of its activities in supporting country actions and programs. CAS Retro I laid out a three- part framework to help country teams exercise selectivity and to help the Board and Management assess it. The framework emphasizes (a) the potential magnitude of the impact if the action is taken, (b) the likelihood of successful country action, and (c) the likely additionality of the Bank Group’s contribution. Although most CASs discuss one or more of these dimensions, many of them do not fully distill their implications for the design of the Bank Group program. Going forward, this is clearly an area for greater attention. viii 1. Impact on the Ground 17. In assessing the likely impact of government action, the bottom-line measure must be the implications of action for poverty reduction. Historically, CASs have generally not discussed the implications of the medium-term macroeconomic prospects for poverty reduction, or the sensitivity of these prospects to different policy and institutional reforms. This is changing. CAS poverty treatment has improved in some areas—indicator formulation and information strategy development—between FY98 and FY99. But we must do better, especially on the link between the pattern of growth and poverty reduction, and the connection between the Bank- and donor-supported program and poverty reduction. 2. Likelihood of Successful In-country Action 18. In assessing the likelihood of successful government action, the CAS should identify the key economic, political economy, and social and institutional constraints that will affect the government’s ability to take and sustain action in a particular area. Most CASs do a good job in identifying the economic constraints, but are weaker in discussing the social and political- economy underpinnings and institutional constraints. Yet these considerations are often the key determinants of the government’s room for maneuver and the likelihood of action in key areas. They also are important for determining the kind of intervention needed to support government action. 3. Bank Group Comparative Advantage 19. A proposed program or project may be a big-ticket item with a high probability of government action, but does the Bank have the comparative advantage to address it? How good is our track record in this area? What are our partners doing? How well do our CASs address these questions? While CASs increasingly use the vocabulary of strategic selectivity mentioning comparative advantage, track record, and so onthere is less evidence that CASs actually use selectivity itself as an analytic tool for focusing on certain sectors or themes and exiting from others. Furthermore, very few spell out why particular lending or nonlending instruments were selected. This, then, will be a major challenge going forward: to use the strategic selectivity framework (implementability in country, big potential impact, and Bank comparative advantage) to decide which activities the Bank should undertake, using which instruments, and which activities it should not undertake. 20. Track Record. CASs now do a much better job of reflecting lessons learned. In all cases where the Bank has recently been active in lending, the CAS assesses the Bank’s portfolio and portfolio management, and the quality and scope of these assessments have improved markedly since CAS Retro I. The FY99 ARPP9 highlights the improved performance of CASs in analyzing portfolio management issues, noting that portfolio management has now been mainstreamed in the CAS. There also is much more use of the findings of country assistance evaluations prepared by the Operations Evaluation Department (OED). However, there is still room for improvement. CASs can do a better job of linking policy, institutional, and governance issues to the diagnosis of portfolio performance; report more consistently and clearly on the use 9 See Annual Review of Portfolio Performance (ARPP), FY99 (CODE 99-80), October 13, 1999. ix of tools such as country financial management and country procurement assessments; and include specific, monitorable actions, with benchmarks, for addressing portfolio issues. 21. Partnerships and Comparative Advantage. CASs also have shown a marked improvement in their coverage of partnerships, which are now discussed in all CASs, up from 50 percent in CAS Retro I. Almost all include a discussion of partnership with the IMF and with regional development banks. Nearly 60 percent cover partnerships with nongovernmental organizations, and three quarters discuss donor coordination. However, only 38 percent indicate partners’ involvement in a sector or activity as the reason for the Bank to stand back. Closing the circle on partnerships, comparative advantage, and selectivity is clearly a priority for CASs going forward. 22. Instrument Choice. The choice of instruments is another dimension of selectivity. To date, few CASs have explicitly addressed the questions of the preferred combination of lending and nonlending activities, the types of lending instruments, and the sequencing of proposed instruments. They are particularly reticent on the ESW program. But as the CAS becomes more of a business strategy—and with the growing interest in instruments and the appropriate preconditions for their use—these issues will take on increased importance. Future CASs need to discuss the rationale for the mix of activities and the effects that changes in Bank scenarios would have on nonlending services. B. Calibrating Program Size 23. In preparing the CAS program, country teams need to take into account two different envelopes: one for the lending amount and the other for the administrative budget. For IDA countries, the lending program needs to be consistent with the IDA allocation, while for IBRD countries the lending envelope must reflect creditworthiness considerations. In terms of the CAS program, country teams need to ensure that the cost of undertaking the full set of lending and nonlending activities falls within the country budgetary allocation, in line with the recently adopted Management rule that the CAS base case must be fully funded. 24. Scenarios and Triggers. The CAS determines the overall level of Bank assistance under alternative scenarios, and triggers for moving between them. CASs do a reasonable job of linking the lending program to the identified development priorities. All CASs that contain more than one lending scenario have triggers for moving across scenarios; although the quality of the triggers has improved since CAS Retro I, they still need greater specificity. About two thirds of the CASs do a satisfactory or better job of linking the triggers with key issues and Bank lending. C. Managing Risks 25. Good CASs discuss the financial, creditworthiness, and reputational risks to the Bank. Slightly over one third of the CASs analyze the impact of key risks satisfactorily, and fewer than one third classify risks according to their perceived importance. Future CASs need to provide a more candid analysis of risks and establish a close link between risks and the discussion of medium-term scenarios. x D. Keeping Score 26. As the major instrument guiding the Bank’s strategy and program, the CAS should contain a monitoring and self-evaluation framework. The self-evaluation framework in most CASs consists of a large set of benchmarks and monitoring indicators that impose heavy information requirements and do not always distinguish between country and Bank performance. A smaller and more thoughtful set of core country and Bank performance benchmarks would help focus management attention on progress in achieving key objectives. The six self- evaluation pilot CASs completed during the review period contain useful preliminary lessons for improving CAS self-evaluation. They also provide a basis for moving ahead—albeit cautiously and on an experimental basis—to begin including in the Corporate Scorecard ratings of CAS strategic quality, with the CAS itself containing explicit benchmarks for self-assessment of Bank performance during implementation. E. Checking Progress 27. CAS Progress Reports are useful for keeping the Board abreast of country development issues and progress in implementing country strategies. Originally intended as very short documents for use in alternate years for large IBRD and IDA borrowers, they now are used more frequently and flexibly. Subject to a smaller set of requirements than the CAS, Progress Reports can be useful when country conditions are not propitious for the preparation of a full CAS, or as a bridge until a full CAS can be prepared. They also are useful products for reinforcing the message that progress needs to be judged against pre-agreed benchmarks. Not surprisingly, Progress Reports share the strengths of CASs, as well as their weaknesses―among which are the lack of recent information on the incidence of poverty and weak analysis of risks. Sharpening triggers and self-evaluation frameworks, as discussed above, should facilitate the subsequent assessment of the progress made toward reaching the CAS objectives when it comes time to prepare the CAS Progress Report. V. Conclusions and Follow-up 28. Clearly, improvement in the quality and coverage of CASs has continued, consolidating the improvements recorded in CAS Retro I. But with the changes under way in the Bank’s operating environment, it is now time to begin “unpacking” the CAS, allowing companion and supporting documents to play a larger role, and sharpening the CAS’s focus on the specific poverty reduction goals and the choice of Bank instruments for helping to achieve them. On this view, as progress on these other processes and documents unfolds, the CAS would increasingly become the Bank Group business strategy. To this end, the focus of the CAS agenda going forward should be on the following: • Increasingly relying on vehicles such as the CDF and PRSP to set out the country’s priorities and objectives for poverty reduction and other development goals, with the CAS focusing on the Bank Group’s business strategy in the context of this broader country agenda. Close coordination will be needed between the preparation of the CAS and a companion CDF/PRSP or other country-led document, and the consultation processes that underpin them. xi • Using cross-country indicators—such as the sectoral indicators the Networks are developing in the context of their SSPs—to help prioritize the cross-cutting challenges that the country faces. In the areas that most constrain development in a country, the CAS should include a full diagnosis; in areas where performance is better, the CAS can give a much more cursory treatment, explaining the reasons. • Increasing the reliance on ESW—public expenditure reviews, poverty assessments, social and structural policy reviews, and so on—for the CAS diagnosis and the country-led CDF/PRSP process. The forthcoming ESW Retrospective will discuss the emerging parameters for this work. • Ensuring that IFC and MIGA are involved early on in the CAS process, thus strengthening the coordination of Bank Group activities in supporting private sector development. • Using the strategic selectivity framework to assess which activities the Bank will undertake and which it will not, explicitly looking at the choice of instruments—both lending and nonlending—to be used for these activities. • Increasing the realism of CAS forecasts and risk analysis—both for country (economic growth and poverty reduction) and Bank business parameters. • Strengthening CAS monitoring and evaluation of Bank support for poverty reduction and other objectives as a basis for assessing Bank performance and enhancing accountability—and, over time, for including CAS ratings in the Corporate Scorecard. xii COUNTRY ASSISTANCE STRATEGIES: RETROSPECTIVE AND IMPLICATIONS I. INTRODUCTION 1. In 1998 the Bank conducted the first in-depth review of Country Assistance Strategies (CASs). That report (CAS Retro I) looked at all CASs discussed by the Board in the previous 18 months, assessing their coverage, content, and quality against Bank guidelines and best practice.1 This report (CAS Retro II) follows up on that work, looking at the 37 CASs and 16 CAS Progress Reports discussed by the Board in the second half of FY98 and in all of FY99—that is, between January 1998 and June 1999.2 Its objectives include the following: • assessing the progress made in achieving the goals laid out in the first retrospective; • identifying examples of best practices that could guide task teams in preparing future CASs; and • outlining an agenda for continuing improvement in CASs, as a basis for a discussion with Executive Directors on the way forward. 2. CAS Evolution. The role of and audience for the CAS have been evolving since it was first introduced in 1990. (Box 1 highlights some of the areas in which change has been most marked.) In the beginning, the CAS was exclusively a document for Board discussion of IDA country programs. The country authorities saw the CAS for the first time when their Executive Director shared it with them; they were not consulted during preparation. Over time this changed—both after the CAS was mainstreamed to cover all borrowers in 1992 and as the Bank’s operating environment changed. The first Operational Manual statement on preparing the CAS—OD 2.11, issued in November 1992—made no reference to outside consultation. Little over one year later, the revised statement (BP 2.11, issued in January 1994) stated that the CAS should be prepared in consultation with the government. By 1997 the value of “staff consultations with civil society, including NGOs, on the CAS” had been recognized, although it was also recognized that such consultations needed to “be pursued with sensitivity and with the prior general agreement of the government.”3 And since August 1998, it has been possible to disclose CASs to the public. Going forward, other changes are likely, especially as the Comprehensive Development Framework (CDF)4 and Poverty Reduction Strategy Papers (PRSPs) take root. The strategic content and the process of developing CASs have proven to be sufficiently valuable that other documents are taking the same approach (see Annex B). 1 See Country Assistance Strategies: Retrospective and Outlook (SecM98-242), March 30, 1998. 2 The CASs in this cohort are listed in Annex A. 3 See Enhancing the Board’s Discussion of Country Assistance Strategies (R97-73), April 18, 1997. 4 James D. Wolfensohn, A Proposal for a Comprehensive Development Framework, World Bank, January 21, 1999. 2 Box 1. CASs over Time The CAS instrument has been in use since 1990, long enough for several CASs to have been prepared for some countries. We examined a total of 28 CASs prepared for seven countries—Bangladesh, Brazil, India, Kenya, Mali, Mexico, and the Philippines―to see how the coverage of CASs has evolved over time. These CASs are divided into three groups: 11 from the early 1990s (1990-93), 10 from the mid-1990s (1994-96), and 7 from the late 1990s (1997-99). The coverage of several important issues has improved significantly since the early 1990s (see charts below). This assessment did not attempt to evaluate the evolving quality of coverage; however, the quality of the discussion in early CASs was weak by today’s standards. Country Diagnosis Government Priorities Implementation Capacity 100 100 100 Economic Government 80 80 priorities 80 Percent Percent 60 Percent 60 60 40 Social 40 40 Bank-government 20 Political economy 20 differences 20 0 0 0 Early 1990s Mid 1990s Late 1990s Early 1990s Mid 1990s Late 1990s Early 1990s Mid 1990s Late 1990s Comprehensive country diagnosis is a To ensure ownership, the Bank’s The borrower’s implementation prerequisite for meaningful program program should be consistent with capacity is a major factor in whether design. In the early 1990s, few of government priorities; but CASs also the Bank’s program will achieve these CASs discussed the country’s need to discuss candidly any impact. In this group of CASs, less political economy and social context; differences between the Bank’s and the than half of those from the early 1990s half or more did so in the mid-1990s; government’s viewpoints. In the early discussed the borrowing country’s and by the late 1990s, all of the CASs 1990s, three quarters of the CASs implementation capacity. This cover the social situation and most discussed government priorities; but all compares with 70 percent and 86 cover political issues. of them do so in the mid- and late percent, respectively, of those prepared 1990s. In the early 1990s, almost none in the mid- and late 1990s. of the CASs discussed differences; in the late 1990s, about 40 percent of them do. Choice of Bank Instruments Partnership Lessons Learned 100 100 100 80 80 80 Percent Percent Percent 60 60 60 40 40 40 20 20 20 0 0 0 Early 1990s Mid 1990s Late 1990s Early 1990s Mid 1990s Late 1990s Early 1990s Mid 1990s Late 1990s The Bank’s program should use a The Bank’s comparative advantage Lessons learned from past Bank combination of lending and nonlending vis-à-vis other donors should serve as a interventions need to be taken into activities to achieve best results. In the principle of selectivity in program consideration in program design. Few early 1990s, only 36 percent of the design. In this group of CASs, 64 of the CASs in this group that were CASs had separate discussions on percent of those prepared in the early prepared in the early and mid-1990s lending and ESW services; this 1990s mentioned the role of other discussed this area; today, all CASs in compares with 70 percent of the CASs donors, as compared with 90 percent of this group do so. of the mid-1990s and 100 percent of those of the mid-1990s and 100 percent those of the late 1990s. of those of the late 1990s. 3 3. Organization of the Report. This report follows closely the structure of CAS Retro I. Chapter II examines country issues related to CASs, Chapter III discusses CASs’ diagnoses of development challenges, and Chapter IV considers the design of the country program set out in CASs. Chapter V presents conclusions and recommendations stemming from the analysis. II. COUNTRY ISSUES: CANDID DEPICTION 4. This section discusses the progress made in grounding CAS priorities in the country context and in the government’s development agenda. It looks at how firmly the CAS is grounded in the country-specific context; how strong country ownership and (given the prior agreement of the government) broad stakeholder consultation are; how candidly the CAS acknowledges the areas where the government and the Bank “agree to disagree,” and how candid the discussion of the country’s situation, policy and institutional framework, and development needs is. A. Country Focus 1. Government Priorities 5. Good CASs identify the government’s development objectives, describe the major elements of the government’s strategy, and provide an even-handed assessment of reform progress. Almost all CASs have a separate section discussing the government’s agenda and priorities, and virtually all include a detailed discussion of the progress of reform and refer to the evolution of priorities over time. The discussion of government priorities is substantial in about 80 percent of the CASs (versus 60 percent in CAS Retro I). Some CASs (Kenya and South Africa) offer good examples of a discussion of the evolution of government priorities. About 40 percent of the CASs explicitly identify areas where reforms have stalled; of these, only about two thirds provide a satisfactory explanation for the lack of progress (the Senegal and Kenya CASs show good practice). In the future, CASs should discuss not only areas of good progress but also those in which progress has been lagging, so as to present a more balanced picture of reform progress and the challenges of implementation. 6. Government Support for CAS. All CASs are now prepared in consultation with government, and in most cases the process involves a number of agencies at different levels. But of course this approach does not guarantee accurate portrayal of government views. One way to gauge government agreement with the CAS is through the Board statement of the Bank’s Executive Director (ED) representing the country. For the CASs reviewed for this retrospective, with few exceptions, the EDs fully supported CAS programs as presented at the Board. In most cases, EDs remarked on the close consultations between Bank staff and government authorities during the CAS process, and welcomed the more participatory approach to the development of CASs. EDs frequently reported that their authorities welcomed joint IBRD/IFC CASs and others that benefited from the integration of IFC and MIGA in the strategy. In many cases, EDs stated that CASs were well aligned with the country’s own national poverty reduction strategy, and in some cases they noted that the CAS process had been instrumental in helping the government formulate its own national development strategy. 4 7. Consistency of Government and Bank Views. In general, CASs report a good fit between the Bank’s objectives and the government’s priorities. More than 70 percent of CASs link the Bank’s goals reasonably well to the government’s, and 15 percent―for example, the Senegal and Hungary CASs―do an excellent job, indicating the extent of Bank’s support and identifying differences between the government’s and the Bank’s views. The Panama CAS is an interesting case: as a result of a highly participatory poverty assessment, the government has officially adopted the analysis and recommendations of the assessment as its own priorities; the result has been a close alignment between the government’s objectives and the Bank’s. Relative to the CASs reviewed in CAS Retro I, there has been no significant change in the frequency and depth of the discussion of government-Bank differences. Differences between the government and the Bank are explicitly acknowledged in one fourth of the CASs, compared to 10 percent of the FY97 CASs, and 20 percent of FY98 CASs. The most often cited differences are on the speed (20 percent) and the scope (16 percent) of reforms. Differences in basic development priorities are reported in very few CASs. 8. Agreeing to Disagree. The explicit acknowledgment of differences in the CAS opens the way for a discussion of how the Bank is adapting its approach to make allowance for these differences. In the Kyrgyz and Vietnam CASs, for example, the Bank and government are in close agreement on most policies but differ on the pace, scope, and sequencing of a few further reforms. The Kyrgyz CAS mentions differences with respect to the reform of the pension system, the introduction of cost recovery for public services, and the government’s role in the industrial sector; to address them, the Bank proposes to intensify the policy dialogue, prepare sector analytical work, and provide technical assistance. In the CAS for Vietnam—where the Bank is urging an accelerated program of financial sector and enterprise reform and the Vietnamese authorities are expressing concern about the social costs of dramatic reform—the Bank proposes to launch a major program of advisory and technical assistance services to deepen support for the needed actions. In the Senegal CAS, where the issue is the pace of civil service reform, the Bank will allow more time for ownership to emerge and will support the process without trying to set the pace of reform. In the Gambia and Moldova CASs, which describe disagreements on the pace of reform, the Bank will use sector work to help build consensus. 9. Strong Disagreement. Two CASs highlight particularly strong differences between the government and the Bank on development priorities. In the case of the Uzbekistan CAS, these differences center on domestic and foreign trade liberalization. The Bank viewed a wide range of government controls as creating rent-seeking opportunities, with negative implications for the overall environment for private sector development, public sector accountability, and governance. The government stressed its commitment to reform, but cautioned that the speed and sequencing of these reforms would be a matter of careful and deliberate choice. The program that emerged from the CAS dialogue focuses on supporting efforts to create the incentives for deeper structural reforms through carefully selected investment operations in parallel with continuing policy dialogue. The Belarus CAS outlines the differences between the Bank and the government on the diagnosis of the economic situation and the reform strategy. The CAS includes a government memorandum with the authorities’ comments on specific aspects of the assistance strategy. 5 2. Country Development Context 10. With very few exceptions, CASs present a thorough and focused diagnosis of the country’s economic issues and a detailed account of recent economic developments. Some CASs also provide historical perspective that helps give an understanding of the root causes of the country’s development challenges (Guatemala, Lesotho, and Yemen). Other CASs show key indicators that help position the country’s present stage of development relative to its neighbors or other countries facing similar conditions (South Africa and Gabon). The Albania, Bolivia, Latvia, and South Africa CASs offer good examples with regard to country development context. 11. Social and Political Economy Context. To help explain the evolution of country policy and institutional outcomes, CASs increasingly describe the social and political-economy context, analyzing the strength of governmental, social, and economic groups and their preferences and the interactions among them. In contrast to the CASs of the early 1990s, over 95 percent of CASs today include some discussion of the social and political-economy context, with almost one third providing good examples (Latvia, Guatemala, Albania, South Africa, Gabon, and Lithuania). The Mali CAS provides a simple but effective method for addressing the position and strength of groups―identifying short-term “winners and losers” of specific policy initiatives and suggesting their likely response to policy changes. It must be noted, however, that in many CASs the social and political economy discussion is relatively weak. Given borrower sensitivities, it may not be possible to present in all CASs a candid assessment of the social and political context, and thus the document may only partially reflect the range of analytical insights available to the country team. 12. Government Commitment. The social and political economy analysis can help to assess government’s commitment to the proposed agenda—a condition that is increasingly recognized as necessary for the success of policy reforms but that is inherently difficult to gauge. More than half of the CASs attempt to assess commitment to current CAS goals by examining the government’s track record in implementing previous Bank-funded initiatives. This approach has obvious shortcomings—it attempts to assess the likelihood of future actions on the basis of a track record achieved under different political and social realities. Only 40 percent of CASs mention country commitment in connection with the proposed CAS program. As good practice examples, the India CAS concentrates support in states and programs for which there is a strong commitment to sustainable reform, and the Kenya CAS links broader participation to commitment throughout program implementation. 13. External Environment. External factors affect country performance in many dimensions, mainly through their economic impact on the balance of payments. Almost all CASs discuss external factors and associated risks. The discussion tends to focus rightly on specific factors that have the most country impact rather than on the external environment in general. The most commonly mentioned external factors in the FY98II-FY99 CASs are changes in the prices of oil or other commodities, volatility of foreign direct investment or portfolio flows, and contagion effects from crisis in neighboring countries or at the regional level (see Box 2). Several CASs discuss issues that are also relevant to other countries in the same region, such as the impact of a regional crisis or accession to the European Union (EU) (see Box 3). Trade issues clearly need to receive greater attention in future CASs. Looking forward, CASs should highlight relevant country issues brought about by ongoing global trade liberalization 6 efforts. The costs and difficulties faced by specific countries in implementing the Uruguay Round agreements should be discussed.5 Similarly, the impact of agricultural trade and services liberalization and issues of market access deserve more attention in the CAS. Box 2. External Environment Almost all the CASs mention several external constraints or risks that could impede or threaten the country’s development effort. About 80 percent have a separate section on external environment. The most frequently identified external constraints or risks are changes in the prices of key export or import commodities, movements in foreign capital flows and exchange rates, and economic developments in neighboring countries. Other factors mentioned include growth in OECD countries, regional stability, and trade barriers. The graph below summarizes these findings. External Environment in CASs (in percent) 100 97 External constraints/risks mentioned 78 80 60 43 41 40 30 19 20 16 14 14 0 Separate section Discuss external Oil/commodity FDI/foreign Crisis contagion Trade relations Growth in Regional stability Trade barriers on external factors price capital inflow with neighbors OECD countries environment Many CASs, however, mainly discuss economic issues—such as export/import performance or financing needs of the country—that could be affected by the external environment, rather than the external environment itself and its impact. The Thailand CAS provides a best practice example: it not only provides a clear and comprehensive picture of the external environment the country faces, but also discusses the impact of the external constraints and risks identified. B. Participation 14. Participation in the CAS process improves the Bank’s understanding of country circumstances and development priorities. It facilitates the diagnosis of social conditions and the identification of governance or institutional bottlenecks to implementation and service delivery. Moreover, participation can improve stakeholders’ understanding of the Bank’s goals, procedures, and expertise, and enhance country ownership of development programs. 15. Extent of Participation. Stakeholder participation in the CAS process has increased markedly over the past few years. Recent CASs have continued the trend toward more intensive civil society and NGO participation. Three-fourths of the CASs in the sample involved a significant level of such participation,6 compared to 52 percent in the last retrospective. In about 5 See The Uruguay Round and the Developing Economies (World Bank Discussion Paper 307), October 1995, and World Bank Support for Developing Countries on International Trade Issues (Sec M99-603), August 25, 1999. 6 This discussion of participation in the CAS process is based on a desk review of the 37 FY98II-99 CASs and an earlier in-depth review of 13 CASs, carried out by the Participation Thematic Group with the cooperation of the ESSD NGO Unit. 7 Box 3. Regional Issues Countries in the same region may face similar development issues and benefit from sharing experiences. In the CASs reviewed, three groups of countries provide examples: • For Lithuania, Latvia, and Hungary, accession to the European Union (EU) is a common driving objective. For Hungary, with a strong macroeconomic position and stable democracy in place, the focus of the government’s priorities is on institutional, social, and environmental issues that will prepare the country for future entry into the EU. For Lithuania, the CAS indicates clearly that strong social consensus for EU membership helps to make any protracted departure from the reforms highly unlikely. In the case of Latvia, in order to meet the EU accession requirement, the development agenda focuses on economic management, private sector development, reshaping the state’s role, the development of subnational government capacity, and service delivery in the social sectors. The Bank’s assistance strategies for these countries have been designed to support implementing the agenda and reforms and investments needed for EU accession. • For Indonesia, Malaysia, and Thailand—all countries that have suffered from the recent financial crisis—the strategies bear striking similarities. All three CASs focus on restoring macroeconomic stability, mitigating the social impact of economic crisis, and reviving growth. The Malaysia CAS indicates that the short-term challenges are to overcome the crisis and to maintain the country’s human capital, and the medium-term challenge is to rekindle growth. For Thailand, the first priority is to consolidate the still-fragile recovery and to revive growth. For Indonesia, the Bank’s strategy focuses on reinforcing social safety nets, stabilizing the economy, and strengthening institutions to support sustainable growth in the future. • A common theme for the Mali and Senegal CASs is regional integration. Both are members of the West African Economic and Monetary Union (WAEMU), which proposes mechanisms to introduce a common external tariff and a framework for investment incentives, and to harmonize economic policies among member countries. The Mali CAS highlights the need to reduce input and factor costs in order to compete favorably within the WAEMU. In the case of Senegal, WAEMU’s initiatives serve as an important incentive for Senegal to streamline its incentives further and improve its business climate. Both CASs also report on the reform programs adopted in the wake of the currency devaluation in 1994. Mali’s economy benefited significantly as a result of increased competitiveness in cotton, rice, gold, and livestock. To accompany the exchange rate adjustment, Senegal carried out a comprehensive medium-term adjustment and reform program aimed at creating the conditions for strong and sustainable growth and poverty alleviation. These reforms have led to a marked recovery of economic activity and a significant improvement in Senegal’s financial position; but the social cost was high. 40 percent of the CASs, preparation included a high degree of stakeholder participation and analysis by a representative mix of stakeholders; involved civil society in planning the consultations; made an effort to reach the poor in urban and rural areas; and involved high- quality follow-up of stakeholders’ original inputs. Among the earlier group of CASs, CAS Retro I found that only 20 percent of CASs met the standard for stakeholder participation, and only 10 percent had done so in FY96. The proportion of CASs with a low level of participation, or without significant stakeholder consultation, has dropped from 65 percent in FY96 to 25 percent in FY98-99. (Box 4 describes the criteria for evaluating CAS participation, summarizes the findings, and presents good practice examples.) Existing data do not allow for a systematic breakdown of the costs of participation in CASs. Available information suggests, however, that the additional budgetary costs of participation in CASs with significant or extensive civil society consultation were roughly $60,000, about 40 percent of the average cost of preparing a full CAS. (Annex C discusses CAS costs.) 16. Enhancing the Quality of Participation. Participation in the CAS process has, to a large extent, been mainstreamed. CAS Retro I asked the question, Are consultative methods being used in formulating Bank assistance strategies? Given the increasing use of participatory methods, the more appropriate question now is, What is the quality of the consultations being 8 Box 4. Participation in the CAS Criteria. The retrospective evaluated the extent of participation in CASs by ranking them in three groups: • CASs with broad and extensive civil society consultation, which involved participation by a wide array of stakeholders, involvement of civil society in the planning of consultations, major outreach into rural/poor areas, high-quality follow-up, etc. • CASs with significant civil society consultation, Civil Society Participation in CASs, FY96-99 which involved substantive participation of several (in percent) 100 groups of civil society stakeholders, but little or no FY96 FY97-98 I FY98 II/FY99 systematic civil society involvement in planning or 80 65 follow-up, no major outreach into rural/poor areas, 60 48 etc. 40 40 32 35 • CASs without significant civil society consultation, or 25 25 20 20 for which information was not available. 10 0 The results of this evaluation are shown in the graph. CASs with no significant CASs with significant civil CASs with extensive civil consultation society consultation society consultation Good Practices in CAS Treatment of Participation • Draw on participatory projects and existing consultative processes for the CAS. In Guatemala, the Bank has been using participatory methods since 1997: it has participated in Consultative Group meetings, consultations with Mayan elders, and government-sponsored open hearings. It has conducted annual Country Strategy Implementation Reviews and has used extensive consultations in the preparation and design of projects. This effort has helped to create an enabling environment for the CAS, reduce the expense of a participatory CAS, and consolidate the learning from several aspects of the country’s assistance program. • Institutionalize the participatory process in a country. In Bangladesh, NGO and civil society participation in development projects is standard practice. In preparing the CAS, the team used a participatory approach for each of the sector strategies, rather than a one-time CAS participatory event. To enhance the follow-up, sector leaders will remain in contact with various stakeholders as they carry out their work. • In accordance with the CDF, promote cooperative relations among development partners beyond complementary lending. Development programs are most effective when all the partners work together on a continuing basis. In Ghana, the Consultative Group created a “Mini-CG” of all the bilateral and multilateral agencies, which meets quarterly to facilitate coordination and oversees the CDF. Through the Mini-CG, the Government of Ghana is taking the lead in donor coordination. conducted in client countries? The depth of the participatory process depends on the level of development of civil society; thus, participation requires a country-specific approach. Good participation means a well-managed process with a broad spectrum of groups represented, not just the involvement of large numbers of people. The quality of the process also depends on the consultation methods used, client ownership of the consultative process, the extent of the disclosure of information, and the consultative mechanisms used to follow up with participants after the CAS has been drafted. 17. Analysis of Excluded Groups. A number of CASs build in a framework for including poor and excluded groups in regular consultation on proposed policy initiatives. The CASs for Vietnam, Bangladesh, Bolivia, Guatemala, Nepal, and the Philippines provide detailed information about the frequency and inclusiveness of the consultation process, and the fora in which consultations are held. These CASs are distinguished by the range of stakeholders consulted—villagers, scholars, business owners, peasants, union members, and so on—and by the linkage of the CAS program to the priorities and demands articulated in the consultation process. 9 18. Range of Methods. The range of consultation methods used in the CAS process has widened. Most consultations use survey methods and workshops held in the capital city with representatives of government, civil society, the private sector, other donors, and NGOs. In addition, CAS teams are beginning to include the rural population and vulnerable groups through focus groups, in-depth interviews, and village immersion programs. In Guatemala, besides holding workshops in the capital, the Bank team went to remote rural areas to hold focus groups with villagers and Mayan elders. These efforts led to an important sense of inclusion in a fragile post-conflict situation. In Mongolia, the CAS team commissioned a detailed study of stakeholder groups and their views before initiating the CAS consultations. Both the CAS team and the government used the study to focus workshop discussions. 19. Increasing Ownership. Drawing on the experience gained from participation in economic and sector work (ESW) and project implementation, CAS teams have widened the knowledge base for the CAS, lowered the cost of consultation, and controlled the expectations raised by high-profile consultations. In Bangladesh, for example, rather than conducting a high- profile consultation, the country team opted for more frequent, smaller-scale, and less expensive consultations at the sector level to provide a wide range of inputs to the CAS. To ensure country ownership of the CAS, the Bank has made wider use of local experts who can provide greater insights into stakeholder needs and greater access to rural and vulnerable groups. Consultations for the Philippines CAS, for example, were facilitated by a locally well known NGO with extensive local ties to groups that had never had direct contact with Bank staff. Participation and ownership can be increased by disclosure and feedback to CAS participants at different stages of the CAS process. 20. Improving Relationships. The CAS process enables the Bank to form a wide variety of lasting relationships with civil society groups, including NGOs, local networks, and individual organizations in remote areas. In India and Bangladesh, the CAS team was able to reinforce Bank ties with various local groups. In Vietnam, the CAS team concentrated on developing initial relations with civil society groups and also formed a strong working relationship with the private sector. The CDF and PRSP, with their focus on enhancing country ownership, will continue to strengthen this process: CASs conducted in CDF countries (for example, Bolivia, Ghana, and Dominican Republic) run parallel to and draw upon the CDF, and the PRSPs provide the country context for CASs and Bank lending levels. 21. Managing Participation. Successful consultation requires time, resources, physical presence, knowledge, and expertise. Stakeholder participation is most productive and cost- effective when it is part of strong field-based relations with stakeholders and when the task team, country office staff, and headquarters country team staff collaborate closely in planning and coordinating participation. Involving other donors in the process can help to coordinate development efforts and can lead, in some cases, to cost sharing. Skillful management can help to prevent unrealistic expectations or loss of focus: clearly defining the goals, modalities, and expected outputs of participation, and using them consistently to structure and steer consultations, can help to manage expectations. Widely divergent views can often be handled without losing focus or ignoring dissent, by explicitly recognizing important challenges to policy positions taken in the CAS, while clearly explaining why the CAS takes these positions. Finally, feedback is a crucial element to a successful CAS process. It is natural that participants who have contributed to the process will want to see the CAS document and engage in further discussion. 10 22. Next Steps. To improve stakeholder involvement in the CAS, it is important to move away from multiple open-ended consultations toward more focused participation based on upstream stakeholder analysis; adapt participatory methods to the context in which they are used; and develop a continuous process of participation as part of all Bank activities. Participation should be an iterative process with cumulative results. Individual consultations can contribute to a central body of knowledge and can be used by successive country teams in formulating CASs and specific projects. C. CAS Disclosure 23. At the time of CAS Retro I, CASs were not made available to the public. On July 30, 1998, the Executive Directors approved the CAS disclosure proposals made in that retrospective,7 and the new disclosure policy for CASs took effect on August 1, 1998. The new policy has two parts: • First, at the government’s request, the Bank makes the CAS (or CAS Progress Report) available to the public. • Second, with the government’s consent, the Bank issues a CAS Public Information Notice (CPIN), which summarizes key elements of the country’s economic and social situation, its development agenda, and the Bank strategy.8 24. Experience with the New Policy in FY99. In FY99, the Board discussed 28 CASs after the policy took effect on August 1, 1998. Of these, 19 (68 percent) have been made public.9 Nine CASs were not made public.10 Seven FY99 CASs have been translated into the national language;11 in another five countries with disclosed CASs, English is the official language. Of the CASs discussed by the Board in FY99 after August 1, 15 were for IDA-eligible borrowers— both IDA-only and blend borrowers—and 13 were for IBRD borrowers. Eleven (73 percent) of the CASs for IDA-eligible borrowers have been, disclosed; eight (62 percent) of the CASs for IBRD borrowers have been disclosed. Eighteen CPINs were published.12 7 See Enhancing Board Discussion of Country Assistance Strategies (R98-199), July 17, 1998. This note built on the May 12, 1998, Board seminar on Country Assistance Strategies: Retrospective and Outlook (SecM98-242), March 30, 1998. It also reflected comments made at the Board Steering Committee meeting on June 1, 1998, and comments received from Executive Directors and Executive Directors’ staffs at a technical briefing on July 1, 1998. 8 On January 12, 1999, the IDA Executive Directors approved the Report on IDA12, which recommended that “[IDA] CASs should be made public, starting in July 1999 (with sensitive information appropriately handled in consultation with the Government), unless justified by exceptional circumstances as agreed in each case by the Executive Directors.” This policy was approved by the IDA Governors on April 9, 1999, and took effect on July 1, 1999. 9 Those for Croatia, Gambia, Indonesia, Lao PDR, Lithuania, FYR Macedonia, Malawi, Moldova, Nepal, Panama, the Philippines, Rwanda, Russia, South Africa, Sri Lanka, Trinidad and Tobago, Ukraine, Vietnam, and Yemen are available on the Web and in the InfoShop. 10 Those for Argentina, Belarus, Bosnia, Gabon, India, Kenya, Malaysia, Mexico, and Pakistan. 11 Those for Gabon, Nepal, Macedonia, Panama, Ukraine, Vietnam, Yemen, and Lao PDR. 12 Those for Croatia, Gabon, Gambia, Indonesia, Lao PDR, Lithuania, FYR Macedonia, Moldova, Nepal, Pakistan, Panama, the Philippines, Rwanda, South Africa, Sri Lanka, Trinidad and Tobago, Ukraine, and Vietnam. 11 III. NETWORK ISSUES: COMPREHENSIVE—BUT PRIORITIZED—DIAGNOSIS 25. As a starting point for selectivity, the CAS diagnosis must be comprehensive—but prioritized. With the goal of poverty reduction as the key driver, the CAS diagnosis needs to present a broad view of country conditions, as a basis for assessing where the need for remedial action is greatest. In principle, all Network areas should be covered in the comprehensive diagnosis. In practice, they should not be treated equally. More attention must be given to the areas where country performance is the most constraining to the pursuit and achievement of sustained poverty reduction. If the topic is not a major development issue in a particular country, the CAS should say so, stating the reasons. A. Poverty Reduction 26. As the Bank’s overarching objective, poverty reduction must be at the core of the CAS. Following up on the 1990 WDR and a policy,13 the Bank moved during the 1990s to develop knowledge about poverty and its causes as a basis for shaping CASs. As the quality of poverty data and analysis has improved, this approach―basing strategy development on a diagnosis of poverty―has been largely mainstreamed in the CAS. The introduction of Poverty Reduction Strategy Papers (PRSPs) will accelerate this trend.14 The next steps are to strengthen that connection and begin to establish a framework for evaluating progress toward poverty outcome goals. 27. Progress Since CAS Retro I. There has been continuing improvement in the poverty focus of CASs. Using the CAS Retro I criteria, the percentage of CASs judged to represent good or better practice increased between FY96 and FY99 from 20 percent to 67 percent, and the percentage of CASs judged to be unsatisfactory declined from 28 percent to 8 percent. CAS Retro I found the share of CASs with a satisfactory or good (not just acceptable) treatment of poverty had continually increased, to more than half the CASs in the first half of FY98. For the FY98II-99 CASs, that share now exceeds 60 percent. CAS Retro I found that poverty reduction was treated inadequately in very few CASs. This was also true in this cohort. No IDA CASs were found to be unsatisfactory in their treatment of poverty. 28. Raising the Bar. Meanwhile, the Poverty Sector Board has raised the standards used to evaluate the poverty focus of CASs. The new criteria extend and deepen the previous criteria: under the old criteria CASs were judged by the quality of their poverty-focused diagnosis and strategy, but the new criteria also look at the centrality of poverty reduction to the CAS and the policy dialogue, how well the CAS spells out the country’s own poverty reduction strategy, whether the CAS includes country and Bank performance goals and benchmarks, and whether the CAS discusses the poverty reduction impact of the Bank-supported program. Using these more stringent criteria, 61 percent of the FY98-99 CASs (64 percent of IDA CASs, 56 percent of 13 See Assistance Strategies to Reduce Poverty, World Bank Policy Paper (Washington, D.C.: World Bank, 1991). 14 PRSPs will summarize the outcomes of a country-led, results-oriented, and participatory strategy development process and will provide the context for debt reduction and concessional assistance programs for low-income countries. The general characteristics of PRSPs, and the principles guiding PRSP design and implementation, are set out in Poverty Reduction Strategy Paper―Operational Issues, Joint IMF/World Bank paper (R99-24), December 10, 1999. 12 IBRD CASs) are judged to contain a satisfactory or better poverty focus. (Box 5 describes the new criteria and shows the results of the evaluations.) 29. FY98 and FY99 Compared. The Poverty Sector Board also compared the CAS ratings under the new criteria for FY98 and FY99, to see whether there has been any progress on the poverty focus of the CAS between these two years. This comparison suggests that the poverty focus has continued to improve, with the share of CASs considered to be satisfactory or better increasing from 57 to 66 percent. (The improvement is stronger for IBRD CASs, with the share of satisfactory or better CASs increasing from 40 to 67 percent. For IDA CASs the satisfactory or better share remained at about 65 percent both years, but no IDA CAS was considered unsatisfactory in FY99.) Most of the progress has been in the area of indicator formulation and information strategy planning. In some areas, however, there was little improvement between FY98 and FY99: the discussion of the linkages between growth and poverty reduction continues to be weak, as does the assessment of the impact of past Bank programs on poverty reduction. These are clearly areas that deserve further attention going forward. 30. Next Steps. For the future, CASs will need to report on the main characteristics of poverty and its determinants; trends over time; links between growth—and the patterns of growth—and poverty reduction; and the implications of macroeconomic policies, government programs, and the Bank program for poverty reduction, including benchmarks for assessing future performance. For low-income countries in the process of preparing PRSPs, CASs will need to focus on helping the country―together with other development partners―build substantial in-country capacity to analyze poverty, put together a monitoring system, design policies and programs, and establish an ongoing consultative and participatory process to support PRSP design and implementation. And, once a satisfactory PRSP is in place, strategic selectivity in the CAS will be aided by the poverty reduction priorities arrived at through the PRSP process, and results will be measured according to indicators consistent with the poverty reduction targets agreed in the PRSP.15 B. Traditional Areas 31. In the effort to address poverty reduction, CASs have traditionally focused the Bank’s activities in a few areas—typically, macroeconomic and fiscal sustainability, infrastructure, energy, rural development, and health and education. This section briefly reviews CAS coverage of these areas; the next section will turn to the CASs’ treatment of less traditional cross-cutting topics. 32. Macroeconomy, Infrastructure, and Energy. The macroeconomy, infrastructure, and energy have long been mainstays of CAS coverage. The economic context has been a strong 15 For Bank actions to enhance poverty focus, see Poverty Reduction and the World Bank: Progress in Fiscal 1999 Implementation Note (R2000-41/1), April 13, 2000; Poverty Reduction and the World Bank: Progress Report in Fiscal 1999, (R2000-41), March 30, 2000; Poverty Reduction Strategy Papers: Internal Guidance Note Revised, (R99-239/1), January 19, 2000; and Poverty Reduction Strategy Paper―Operational Issues, Joint IMF/World Bank paper (R99-24), December 10, 1999. See also Heavily Indebted Poor Countries (HIPC) Initiative: Strengthening the Link Between Debt Relief and Poverty Reduction (DC/9924), September 17, 1999; and Building Poverty Reduction Strategies in Developing Countries (DC/99-29), September 22, 1999; Poverty Reduction Strategy Papers (PRSPs)—Progress Report (SecM2000-168), March 29, 2000. 13 Box 5. Poverty Reduction in the CAS Criteria. The following criteria were used in CAS Retro I to evaluate the poverty content of CASs: • Is the diagnosis of poverty adequate and does it incorporate the findings of the poverty assessment or other relevant analysis, if available? • Given the country’s poverty diagnosis, is the CAS appropriate? Under that particular country’s circumstances, have the key elements of the Bank’s poverty reduction strategy―labor-intensive growth, human capital development, and safety nets―been addressed? The CAS Retro I evaluation followed a relative standard: given what is known about poverty in a particular country, does the CAS seem to be doing the “right” thing? A CAS covered those elements that were appropriate given country circumstances and taking into account partnerships and the Bank’s comparative advantage. The results follow: Figure A. CAS Poverty Treatment over Time: Figure B. Poverty Treatment over Time, Old Criteria IDA and IBRD CASs: Old Criteria (in percent) (in percent) 100 100 FY96 FY97 FY98 FY99 CAS Retro FY97-98I CAS Retro FY98II-99 80 71 80 67 62 60 52 56 60 52 52 56 44 37 38 38 40 40 31 28 25 25 20 20 7 8 20 10 13 4 4 0 0 0 Unsatisfactory Acceptable/minimally Best practice/good Unsatis. Accept/min Best Unsatis. Accept/min Best acceptable accept prac/good accept prac/good IDA IBRD The criteria now used by the Poverty Board to evaluate the poverty focus of CASs cover the following: • Does the CAS explicitly state that poverty reduction is its overall goal? • Does it contain monitorable indicators and performance goals for both country and Bank and quantifiable and time-bound benchmarks of progress that reflect the multidimensional nature of poverty? • How good is the poverty diagnosis? Does it provide statistical and qualitative information on the poor; their occupational, educational, and gender characteristics; their geographic location; and changes in poverty over time? Does it refer to the findings of a participatory poverty assessment? • How good is the poverty reduction strategy? Does the CAS discuss clearly how the strategy will help achieve the poverty reduction objective and how the Bank’s strategy fits into the country’s strategy? Does the CAS discuss how different patterns of growth and of Bank assistance will affect the poor? • How good is the CAS’s analysis of the effect of past Bank activities on poverty reduction? The results of the evaluation under the new criteria follow: Figure C. CAS Poverty Treatment over Time: Figure D. Poverty Treatment in FY98-99 New Criteria IDA and IBRD CASs: New Criteria (in percent) (in percent) 100 100 FY98-99 FY98 FY99 All countries IDA IBRD 80 80 58 60 50 53 60 50 47 39 40 33 40 29 28 28 26 21 17 17 20 11 10 13 11 10 13 20 11 11 6 6 0 0 Unsatisfactory Marginally satisfactory Satisfactory Highly satisfactory Unsatisfactory Marginally satisfactory Satisfactory Highly satisfactory Bottom Line. A necessarily rough comparison between the assessments under both sets of criteria shows the extent to which the “bar has been raised.” Under both sets of criteria, progress in the CAS poverty treatment is evident. Under the previous criteria 92 percent of FY99 CASs were rated acceptable or better, while under the new criteria 66 percent are rated satisfactory or better. 14 building block of the CAS since the early 1990s; as discussed in Chapter II, virtually all CASs present a thorough and focused diagnosis of country economic issues and a detailed account of recent economic events. Over 90 percent of CASs identify macroeconomic issues as a priority area, and all CASs contain either a lending or a nonlending activity in this area. Macroeconomic issues account for one fourth of the ESW activities in CAS programs. Infrastructure and energy are also well established as areas of focus in Bank CASs. Of the FY98II-99 CASs, 57 percent identify infrastructure as a priority, and almost all have activities in this area; infrastructure accounts for one fourth of the number of lending activities in CASs. Energy is identified as a priority in one third of the CASs, and about 60 percent contain at least one lending or nonlending activity in this area. 33. Rural Development. The majority of CASs address rural issues. A review by the Rural Sector Board of the CAS treatment of the key rural development topics—poverty reduction, widely shared growth, food security, and sustainable natural resource management—rated three fourths of the CASs as fair to satisfactory. It also found wide variability in the treatment of the individual themes, with the quality of the poverty reduction discussion the highest among the four topics. The quality of the treatment of rural themes is better in CASs for countries that have rural development strategies than for those without such strategies. This suggests that in countries where rural development is a priority, there are benefits to preparing a rural development strategy as a prelude to the CAS. 34. Human Development. Of the traditional areas of CAS focus, human development (HD) is most central to the Bank’s poverty reduction efforts. A well-articulated Bank country program to support progress in education, health, and nutrition should be a key component of a poverty- focused CAS. To evaluate progress in HD over the last two full years, this review considered all FY98 (not just FY98 second half) and FY99 CASs presented to the Board. The review found that the treatment of HD issues in the FY99 CASs improved relative to FY98. About 52 percent of the CASs completed in FY99 contain a good analysis of HD issues, up from 43 percent in FY98. About 44 percent are satisfactory (37 percent in FY98), and only one CAS (4 percent of the total) has poor coverage of the HD sectors. These results show a substantial improvement over FY98, when 20 percent of the CASs were rated poor in their treatment of HD issues. (Box 6 sets out the criteria by which the CASs were evaluated and describes the HD features of good practice CASs.) C. Cross-cutting Issues 35. In addition to these traditional areas of involvement, CAS diagnosis has increasingly focused on a set of cross-cutting areas that are essential to poverty reduction: gender, governance and corruption, environment, financial sector, and the private sector. In assessing whether CASs provide a satisfactory treatment of these areas, there are two approaches that give very different pictures. (See Figure 1.) • The first measures all CASs against an absolute best-practice standard and rates their treatment as more or less satisfactory by that absolute standard. This approach, which serves to highlight excellence in the treatment of these areas, also results in a higher number of “unsatisfactory” ratings. Network contributors to the 15 Box 6. Human Development in the CAS Criteria. The retrospective used three criteria to evaluate the focus of each CAS on human development (HD) issues―that is, issues of education, social protection, and health, nutrition, and population: • Do qualitative and quantitative country indicators show any clear deficits in HD requiring special attention? Does the CAS link recent economic trends with social expenditures and changes in the HD sector? • Does the CAS discuss the priority given to the HD sectors? Is Human Development Focus this discussion balanced and integrated with the strategy to (in percent) achieve macroeconomic growth? 80 FY98 FY99 • How good is the CAS’s coverage of HD issues? Is the coverage 60 52 43 43 consistent with the priority given to HD? Does the CAS discuss 40 37 how specific instruments are used to target specific problems 20 and population groups? 20 4 0 The results of this evaluation are shown in the graph. Poor Fair Good Good Practices in CAS Treatment of Human Development • Use of country indicators to highlight deficits in HD. The Panama CAS highlights the fact that, despite the stable macroeconomic situation, the country’s high income inequality has led to a concentration of health resources in high-income urban areas and in curative care. The Malawi CAS identifies the urgent need to improve delivery of social services. The Gabon CAS presents evidence that the HD indicators (except for safe urban drinking water) are comparable to those in the rest of Sub-Saharan Africa. The Macedonia CAS highlights deficiencies in health and in education, including skills related to the labor market. • Priority given to HD sectors. The Malawi CAS focuses on education and health. The Gabon CAS makes the improvement of social welfare a main objective. The Macedonia CAS focuses about 70 percent of FY01 lending on the social sectors, building on the first generation programs of the previous CAS. • Quality of HD coverage. The Panama CAS proposes interventions targeting indigenous communities, unemployed women, pregnant and lactating women, and youth at risk from violence. The Malawi CAS targets youth and people at risk of contracting AIDS, recommending improvements in the quality of primary education, promotion of secondary and vocational education, and family planning and information campaigns for the prevention of AIDS. The Gabon CAS concentrates on sector strategies, minimum lending projects, and coordination of the work of other donors working in human development. One objective of the Macedonia CAS is to assist in the development of human capital; it supports the government in developing a comprehensive long-term education strategy and improving the quality of educators, particularly in high- poverty rural areas. The CAS also supports health reforms and measures to improve labor market flexibility. Improving CAS treatment of HD. The diagnosis should indicate the priority of various HD issues, and the country’s overall ranking in terms of HD. The CAS should contain information on total HD and social expenditures. Indicators should be used consistently: each intervention should have specific targets and expected outcomes, and benchmarks should be, to the extent possible, quantitative and outcome-based. The HD Network plans to prepare a standard set of indicators to help task teams improve CAS treatment of HD issues. retrospective used this approach. (Annex D presents further details on the criteria used, the findings, and the best practices.) • The second approach builds on the Networks’ assessments but applies the principle of selectivity to interpret the Network scores. This approach is based on the premise that the CAS diagnosis need not—and cannot—cover all areas in the same degree of detail: comprehensive diagnosis should not necessarily mean equally comprehensive coverage. It thus zeros in on the question: What share of CASs cover a given area in an unsatisfactory way in countries for which that area is a major development? By 16 weeding out the CASs for countries where performance on the particular issue is satisfactory, the measure focuses attention on the CASs where there is both a country performance issue and an unsatisfactory treatment of that issue. It thus provides a more realistic view of the percentage of CASs whose treatment of a specific area would need to be improved.16 Figure 1. Cross-cutting Issues: Unsatisfactory CAS Treatment (in percent) 100 Absolute standard assessment (unsatisfactory CASs) 80 Priority-based assessment (unsatisfactory CASs in countries with a CPIA of 3 or less) 60 57 53 48 43 43 36 38 40 24 25 25 25 22 19 20 16 10 33 19 19 20 19 22 16 13 6 10 0 5 5 6 5 0 All IDA IBRD All IDA IBRD All IDA IBRD All IDA IBRD All IDA IBRD Gender Governance Environment Financial sector Private sector From these analyses, three key areas emerge in which future efforts should be concentrated: gender, governance, and the financial sector. 1. Assessment a. Gender 36. Using the absolute standard, the PREM Network rated more than half of all CASs as satisfactory or better in their coverage of gender issues—about the same as CAS Retro I; it rated 36 percent of the CASs as unsatisfactory and 11 percent as marginally satisfactory. Using the CPIA for gender as a proxy for the extent to which gender issues are an important development 16 As a rough indicator of priority development issues in the CAS countries, the retrospective used the ratings from the Bank’s annual Country Policy and Institutional Assessment (CPIA). The CPIA evaluates each borrower country’s policy and institutional environment in 20 dimensions covering economic management, structural policies, policies for social inclusion and equity, and public sector management and institutions. Thus it can help identify areas of strength and weakness at the country level upon which to base an assistance strategy. The CPIA is used in this report as a means to assess whether those policy issues that pose major development constraints are being adequately addressed in the CAS. For each item the CPIA gives a score from 1 (low) to 6 (high). The CAS Retro II evaluation considered any area to be a development priority for a country that received a 1999 CPIA rating of 3 (moderately unsatisfactory) or below. It assumed that lighter CAS treatment of such areas in other countries (those with a CPIA score higher than 3 in that area) was appropriate, and therefore by definition not “unsatisfactory.” Thus it focuses on the narrower set of CASs— those that should have covered an area adequately and did not. This treatment can be seen as the first step toward sharpening the analysis of selectivity. The CPIA is discussed in Country Performance and IDA Allocations (IDA/SecM98-502), September 16, 1998, and IDA Country Performance Rating Process—Annual Report 1999 (IDA/SecM2000-58), February 10, 2000. The CPIA methodology and coverage have been periodically refined, drawing on experience and reflecting changes in the development paradigm. The CPIA methodology still provides an imperfect picture of priorities across countries, and further work is planned to improve it. 17 constraint, the priority-based assessment found 19 percent of FY98II-99 CASs unsatisfactory.17 While the absolute standard suggests far worse performance in IBRD CASs, once the severity of the issue is taken into account via the priority-based assessment, IDA and IBRD CASs perform about equally. 37. Next Steps. By any measure, there is clearly need for improvement in the treatment of gender in CASs, and the Network analysis has indicated a number of ways it intends to address this area. Going forward, CASs need to explicitly address whether gender is a country issue or not, and strengthen the links between analysis and follow-up activities. In countries where gender disparities are a priority development issue, the CAS should support activities in this area, subject to the Bank’s track record and partners’ programs. If the Bank’s assistance does not include specific activities in this area, the CAS should explain why. Also, in many cases, the gender agenda needs to be broadened from interventions relating to women’s health and education to a wider range of sectors that address women’s multiple capacities in a society— such as agriculture, water and sanitation, transport, and finance. A recent review of the treatment of gender issues in poverty assessments completed for IDA countries during the FY97-99 period found that quality improved during this three-year period. This improvement in the analytical base can be expected to translate into better treatment of gender in CASs. The Gender Sector Board is preparing a strategy paper that will address these issues and that should help country teams to better address gender in CASs. b. Governance and Corruption 38. Consideration of governance and corruption in CASs is a relatively new area for the Bank, but the results of the Network evaluation suggest that these issues are being mainstreamed in Bank work. Most CASs discuss governance and corruption issues, although by an absolute standard 43 percent are rated “unsatisfactory.” A priority-based assessment shows that in fact CASs are taking into account the importance of the problem for the country—CASs for countries where governance and corruption issues present significant risks tend to cover these issues more, and only 16 percent provide unsatisfactory treatment for such countries. Furthermore, the quality of treatment improved even over the period reviewed for this retrospective: CASs prepared in the second half of the period treat governance and corruption issues better than their predecessors. 39. Next Steps. The guidelines in the Bank’s Anticorruption Action Plan indicate several areas for improved CAS treatment of governance and corruption.18 The CAS should contain a good diagnosis of a country’s governance and corruption issues. Bank ESW, particularly the results from public expenditure reviews and country procurement and financial system assessments, may help to strengthen the CAS’s diagnosis of corruption. For countries where poor governance and corruption pose significant developmental risks—as reflected, for example, in the CPIA indicators—the CAS should outline the country’s agenda for improving governance and combating corruption and should include a monitorable program to assess progress. The choice and design of instruments should reflect a good understanding of existing institutions and the best way to reform them. For these countries, the CAS should discuss the conditions under 17 CPIA question no. 11 asks the extent to which laws, policies, or institutions limit equality of economic opportunity for women as well as other economic groups. 18 Anticorruption Action Plan for FY99 (SecM98-900), November 9, 1998. See also Helping Countries Combat Corruption, the Role of the World Bank (Washington, D.C.: World Bank, September 1997). 18 which particular instruments should be used. For example, the Bank may consider providing general (untied) external financing only if the country has a credible anticorruption strategy and its financial management and public sector accountability systems meet minimum standards, so the Bank can reasonably expect that its resources will likely be used for development purposes. The content and structure of the lending program should depend on progress in these areas, which should be measured by specific benchmarks. c. Environment 40. Using the absolute-standard assessment, the Network found that most CASs contain a brief assessment of the country’s environmental problems and a description of Bank instruments used to address them, but that the coverage of environment is unsatisfactory in about 16 percent of CASs. The priority-based assessment found that only 5 percent of CASs contain an unsatisfactory treatment of the environment in countries where, on the basis of the CPIA, environment is viewed as a priority.19 41. Next Steps. The Network analysis has indicated a number of ways the CAS treatment of environmental issues can be improved in countries where environment is a priority. Environmental concerns should be mainstreamed by linking them to economic and sector issues. The link between poverty and the environment can be strengthened by focusing on policies that reduce inequality and environmental degradation; supporting interventions that improve environmental conditions affecting the health of poor people; and ensuring that natural resource management contributes to the livelihood of the rural poor. Environmental indicators and analyses of environmental trends need to be better incorporated in these CASs, by drawing, for example, on the environmental indicators collected for the World Development Indicators or on environment-related ESW.20 The Network is working with Regional management and country teams to support these improvements. In addition, a pilot program was launched in October 1998 to test approaches and tools on how to integrate environment in CASs. As the program is completed, its lessons will provide valuable insights on the best way to proceed. (For a summary of the findings so far, see Box 7.) d. Financial Sector 42. CAS coverage of the financial sector has improved, but still varies widely. Some CASs contain very comprehensive and insightful discussions of the sector; but the absolute-standard assessment found that 43 percent of CASs contain an unsatisfactory treatment. The priority- based assessment shows that only 22 percent of CASs give unsatisfactory coverage of the financial sector in countries where the area is a priority. There is evidence of selectivity in coverage: CASs for countries that have suffered financial crisis, and more recent CASs in general, cover the financial sector better. In addition, joint Bank/IFC CASs, on average, cover the financial sector better than non-joint CASs. In addition, joint Bank/IFC CASs, on average, cover the financial sector better than non-joint CASs. 19 The CPIA on the environment assesses the extent to which the country’s economic and environmental policies foster the protection and sustainable use of resources and the control of pollution. 20 See 2000 World Development Indicators, World Bank, March 2000. 19 Box 7. CASs and the Environment Pilot Program. The CAS, as a strategic Bank document, represents a key point of intervention for integrating environmental concerns into Bank operations. The objective of the Country Assistance Strategies and the Environment (CASE) program, which started in October 1998, is to test tools and approaches to integrate environment in CASs, in the belief that mainstreaming environmental issues will lead to improved development outcomes. The program has two components. The first is to test approaches to “greening” the CAS through a series of case studies in which CASE team members participate as full members of the CAS team. The second is to prepare a small retrospective study of the linkage between environmental management and development outcomes in selected countries. In seeking to improve the treatment of environment in the CAS, the goal is clearly not to “green” each and every CAS. However, where there are significant environmental issues in a country, the CAS should treat them in a substantive manner or explain why it does not. Case Studies. To date the CASE program has produced four case studies (Dominican Republic, Zambia, Azerbaijan, and Tunisia) and two retrospective studies (Thailand and Botswana). Two more case studies (Pakistan and Argentina) are slated for completion in 2000. A final report documenting lessons learned and best practice will be prepared at that point. Early Findings. Two preliminary findings have emerged from the case study work. First, the establishment of a clear link among environment, resource issues, and economic outcomes facilitates the integration of environment in the CAS program. Second, a structured approach to presenting the issues facilitates communication among the CAS team members. The key element is a “CAS analytical matrix,” which arrays information about country-specific environmental sectors and issues against a summary of issues and an identification of general driving forces. The matrix is also a useful diagnostic tool to delineate the links among environmental issues, macroeconomic, sectoral, and project-level actions in the CAS, and policy and institutional issues. The use of environmental indicators with regional and income-level comparators has proved to be an effective tool to raise the profile of environmental issues with both country teams and government officials. 43. Next Steps. The analysis suggests a number of ways to improve CAS coverage of the financial sector in countries where the area is a priority. Such CASs should provide a diagnosis of the financial sector and its role in supporting private sector development, and should lay out the progress made since the last CAS. They should identify major financial sector issues, including an analysis of the system’s stability and efficiency, and the ease of access to credit by different parts of the economy. The diagnosis should be based on a financial sector assessment or financial sector chapters of ESW, and should provide references. In terms of the sector strategy, these CASs should describe the policy dialogue with the government and set out how the issues identified will be addressed and by whom. When the financial sector is not a priority, the CAS should explain why. The CAS should clearly describe the division of labor within the Bank Group among IBRD, IFC, and MIGA. e. Private Sector 44. Because the private sector can be involved in many—or all—segments of a country’s economy, it is a very difficult subject to discuss within the limits of a CAS. In general, CASs should provide a brief profile of the sector and focus on constraints. According to the absolute— standard analysis, about 22 percent of the CASs provide unsatisfactory coverage of the private sector. Here again, however, selectivity is apparently being applied: the priority-based assessment indicates that only 5 percent of CASs contain an unsatisfactory treatment of the private sector in countries where the sector is a priority. 45. Joint CASs. The joint CAS program was introduced in FY97 to enable the Bank Group to strengthen coordination and cooperation between the Bank and IFC/MIGA. Sixteen of the 37 20 FY98II-99 assistance strategies are joint CASs, of which eight include a Private Sector Strategy annex (PSS). Joint CASs, and particularly those with PSSs, stand out as providing a much more substantive coverage of the private sector and a joint discussion of the Bank and IFC programs. Most PSSs describe progress made, identify priority issues, and discuss the Bank Group’s role. Most PSSs also identify the financial sector, infrastructure, and legal and regulatory frameworks as major bottlenecks for private development. However, one fourth of the PSSs fail to provide a Bank Group strategy with respect to private sector development. The joint CAS program has increased the degree of consultation between Bank and IFC staff during the preparation of private sector development strategies and has brought about a closer relationship between Bank country teams and their IFC counterparts in designing the strategy. 46. Next Steps. The evaluation suggests a number of ways to improve CAS treatment of the private sector in countries where it is a priority. Such CASs should provide a clear profile of the private sector—its depth, structure, and major constraints—and should describe the Bank Group’s strategy to address these constraints. A key area for improvement is to fully coordinate the Bank program and IFC/MIGA private sector activities. The Bank Group’s private sector development strategy aims to create the synergies necessary to ensure that the mix of products and services offered to member countries meets their needs most efficiently.21 It proposes moving beyond agreement between the World Bank and the IFC on priorities in the CASs, to a joint agreement on work programs that reflect these priorities—and, in selected cases, joint review of individual operations. A PSS will likely be useful for a country’s first joint CAS, or when a private sector development (PSD) agenda is particularly important. In particular, because it is difficult to adequately treat the range of PSD issues within the confines of a short CAS document that covers many sectors and issues, the preparation of PSSs should remain a useful focus for Bank-IFC dialogue. The joint CAS process has worked best when IFC has been engaged at a very early stage in analyzing growth and private sector development issues, delineating the most important elements of the strategy, and consulting with government and other stakeholders. MIGA’s earlier engagement in the CAS process will further enhance the synergies of the Bank Group program. 2. Summing up 47. Going forward, it will be important for CASs to ensure that diagnostic coverage of the cross-cutting issues is commensurate with their importance to development in the country. CASs will need to give complete coverage to those issues in countries where they are problems, and the Networks will need to work closely with Regional management and country teams to support this improved coverage. For future CASs, sector strategy papers to be developed by the Networks will help alert the country teams for countries where any of these areas are known to be major development constraints. The upcoming gender strategy paper, financial sector strategy paper, private sector development strategy, and other SSPs will offer important opportunities for identifying those countries for which CAS treatment of these areas should be expected, and clarifying the kinds of activities that make the most sense for the Bank Group to support. (See Box 8 on SSPs and CASs for a summary of recent experience.) For countries where any of these issues is a moderate or low priority, future CASs can give it less comprehensive treatment and explain why. 21 World Bank Group Private Sector Development Strategy (R99-175, IFC/R99-154), September 10, 1999. 21 Box 8. Sector Strategy Papers and the CAS Purpose. SSPs are Network products that review the Bank’s operational experience, identify sectoral priorities for the Bank at the global level, and propose an implementation plan for pursuing these priorities. They have the potential to be important building blocks for CASs. They define sectoral indicators that can serve as a basis for assessing country progress and performance, and for identifying priority country programs for Bank involvement. They can lay out other business implications and progress indicators for the Bank, and facilitate greater selectivity at the corporate and country levels. They can also help to assess Bank performance in implementing its CASs. Since Rural Development—the SSP prototype—was completed in 1997, Networks have prepared SSPs on health, nutrition, and population (the first full-fledged SSP), education, the environment and energy, and urban and local government.a Several others are in the pipeline, including SSPs on public sector institutions and governance, the financial sector, social protection, energy, and information infrastructure. SSP Value to CAS. SSPs are increasingly identifying critical countries and programs, prioritizing Network involvement in CAS preparation. In the case of environment and energy, the SSP discussed ways in which the Network Family will raise the awareness of country directors and task team leaders in countries that are important to the environment/energy agenda, and then become involved in their CAS work. The urban and local government strategy provides a list—drawn from urban action plans prepared by each Region—of the countries that are potential candidates for special focus. The choices reflect not only the importance of the country to the urban and local government agenda but also the state of the dialogue within the context of country strategy. Network Inputs to CAS. Some Networks are also realizing that there is still a significant distance to go before they can exert a sustained and strategic influence over the sectoral agenda at the country level. Some Network Anchors are doing more to engage operational staff Bankwide in the preparation of SSPs, as well as reaching beyond the SSP process to provide strategic inputs upstream in the CAS process. For instance, in FY98, the HD Network encouraged each Region to select two countries identified in the HNP SSP as high priorities, and to develop best practice CASs with input from the HD anchor. In FY99, following a review of the experience in FY98, the Network decided to selectively strengthen its attention to CASs even further by developing HD indicator sheets for countries to help draw attention to areas that require special attention. Over the last few years, the PREM Network has developed a system of CAS review in which the Regional coordinators within the Anchor keep track of the CAS pipeline, and then concentrate their efforts before the concept review stage. More recently, ESSD has focused on 10-12 priority CASs, with additional services provided for nonpriority CASs on demand. In the meantime, some Regional staff have indicated that for SSPs to be more useful as a CAS building block, they must provide value added in the form of good practice sectoral policies and practices. Regional staff would also like SSPs to focus much more candidly on implementation issues such as policy, institutional, and capacity constraints in countries, as well as staffing and other business constraints within the Bank. ___________________________________ a Rural Development: From Vision to Action (R97-0025), February 18, 1997; Sector Strategy Paper: Health, Nutrition and Population (R97-0168), July 22, 1997; Sector Strategy Paper: Education (R99-68), April 28, 1999; and Sector Strategy Paper: A Strategic View of Urban and Local Government Issues: Implications for the Bank (R99-191), October 8, 1999. 22 IV. BUSINESS ISSUES: SELECTIVE PROGRAM 48. Building on the diagnosis of the country’s development challenges, the CAS sets out a selective Bank Group program of lending and nonlending activities. This chapter looks at the treatment of program design and implementation in CASs. It covers two closely interrelated areas: (a) the choice of Bank activities and the analysis presented to explain the choice, and (b) the level of assistance and how it responds to country performance. It also covers how the Bank’s program set out in the CAS is assessed through self-evaluation and through CAS Progress Reports. A. Selecting Strategically 49. Strategic selectivity involves systematically examining trade-offs among possible Bank group activities, assessing the relative impact of the activities, and prioritizing among them while taking resource constraints into account. CAS Retro I outlined a framework to help country teams exercise selectivity when they choose potential interventions. It covers three dimensions: • The potential magnitude of impact. How important is the activity? If the activity is fully implemented, what will be its impact on overall country performance and sustainable poverty reduction? • The likelihood of successful country action. What is the likelihood that the activity will be successfully completed and its full impact realized, taking into account country ownership, capacity, and commitment to act? • Additionality of Bank contribution. What is the Bank’s specific role and contribution in this activity or sector, taking account of its track record and comparative advantage vis-à-vis other partners? And—though CAS Retro I did not include this aspect— what are the appropriate instruments? 50. Assessment. Most CASs cover one or more dimensions of the framework―impact, track record, and comparative advantage. But often they do not fully distill the implications for the composition of the assistance program. In general, about one half of the CASs do a satisfactory job in conceptualizing their program of activities in the key areas of Bank intervention by addressing one or more of the components of the selectivity framework and by outlining a program of activities in these areas that is consistent with the degree of priority nominally given to them. (Box 9 describes some good examples of selectivity.) In terms of number of stated priorities, sectoral coverage of CASs, and concentration of activities per area, the FY98II-99 CASs are similar to those analyzed in CAS Retro I (see Annex E). 1. Magnitude of Potential Impact on the Ground 51. In assessing the likely impact of Bank-supported government action, the bottom-line measure must be the implications of that action for poverty reduction. It is well established that the level of growth is a key driver of poverty reduction. But the pattern of growth also matters. 23 Box 9. Selectivity in the CAS Criteria. The framework for selectivity involves three questions: • Potential magnitude of impact. How important is the activity, and what will its impact be on overall country performance and sustainable poverty reduction? • Likelihood of successful country action. What is the likelihood that the activity or sector program will be successfully completed and its full impact realized, taking into account country capacity and commitment to act? • Additionality of Bank contribution. What is the Bank’s specific role and contribution in this activity or sector, taking into account its track record and comparative advantage? Selectivity in CASs (in percent) Scores. As the graph shows, many CASs discuss one of the 100 dimensions of selectivity―the decision not to be involved in certain 80 areas. 60 45 43 38 40 20 Good Practices in CAS Treatment of Selectivity 0 Indicates areas Gives reasons Donor • Big potential impact. The Kenya CAS identifies economic where Bank involvement as governance as the most important bottleneck for Kenya’s will not be the reason for development. As a result, it shifts proposed nonlending activities involved no Bank activity from analytical and prescriptive work to activities that facilitate broad-based efforts to tackle economic governance. • Implementability in country. The Moldova CAS does not provide financing for infrastructure because it recognizes implementation difficulties. The CASs for Bulgaria and Lithuania do not propose lending activities in certain areas because of lack of government commitment or interest. Selectivity criteria developed by country teams. Several country teams have developed their own criteria for selectivity in their CASs, some of which are similar to the three criteria listed above. In general, the use of such criteria reflects a country team’s understanding of selectivity taking into account country circumstances. Such criteria broaden and enrich our understanding about different aspects of selectivity. • Bangladesh: “IDA will exercise selectivity by choosing to back activities that have strong client commitment and that will achieve key institutional capacity, policy reform, and service delivery goals. Selectivity in this context also has a time dimension, requiring perseverance once a specific institutional development activity has been selected.” • Yemen: “Yemen’s needs are so great and so diverse that almost every sector looks like a deserving candidate for Bank assistance. To have the greatest impact, however, the strategy should selectively address sectors and issues which are the most binding constraints on development: (a) where the Bank has the capacity to make improvements; (b) where the Government places priority and has control over outcomes; and (c) which other donors, private investors, and the Government are not able to handle without Bank involvement.” • Mexico: “Three main selectivity criteria: The first is marginal impact on poverty reduction...avoiding the adverse macroeconomic shocks that would likely hurt the poor—first and most severely. The second is the client’s expressed needs. WBG support will be sensitive and proactive toward the authorities’ sweeping reforms and pressing needs, such as the decentralization process or environmental degradation, and will be careful not to give the government grounds for disappointment in the face of unachievable requests. The final criterion is comparative advantage. The WBG is equipped to undertake support in a wide array of areas, yet needs to focus where its capabilities can produce the largest return for Mexico.” • Thailand: “The Bank will be active in areas where it has a clear comparative advantage; proposed lending and non-lending activities must fully coincide with Thailand’s primary development objectives; the Bank’s proposed assistance must have a champion at a high level of Government, and a participatory approach must ensure sustained broad-based support at the working level and within civil society; the Bank’s assistance will be restricted to areas where the Government on its own, or with other partners, cannot more productively or efficiently achieve the same objective.” 24 Good policies, stable macroeconomic environments, open trade regimes, protected property rights, and efficient bureaucracies that can deliver essential public services—all are important for growth. The CAS must establish the linkages between these and the poverty reduction strategy. 52. Medium-term Scenarios. Historically, CASs have linked macroeconomic policies and growth scenarios to their impact on poverty in a linear way, with higher growth identified with more poverty reduction. They have not typically looked at the implications of the pattern of growth for poverty reduction. Nor have they treated the sensitivity of poverty outcomes to different policy and institutional reforms. At the same time, CAS scenarios tend to be over- optimistic. In the CASs reviewed, 30 percent of countries are projected to grow at 3 to 4 percent per year, whereas historically only 4 percent of the countries have grown at this rate. More significantly for risk mitigation, only 5 percent of the CAS countries are projected to grow at 1 percent or less, whereas historically more than 55 percent of countries have grown at this rate (see Figure 2). Thus, CASs may not adequately grapple with the consequences of likely slow growth. They must do better, especially on the links between pattern of growth and poverty reduction and between the Bank- and donor-supported program and poverty reduction. Figure 2. Historical and Projected GDP Per Capita Growth (in percent) 40 1965-97 actual Long-term projections 30 20 10 0 Under 0.0 0.00 -1.00 1.01 - 2.00 2.01 - 3.00 3.01 - 4.00 Above 4.00 GDP per capita growth rates (%) 2. Likelihood of Successful Government Action 53. In assessing the likelihood of successful government action, the CAS should identify the key economic, political economy, social, and institutional constraints that will affect the government’s ability to take and sustain action in a particular area. Most CASs do a reasonable job in identifying the economic constraints. But they are weaker in their coverage of the social and political economy underpinnings and institutional constraints, which are often the key determinants of the government’s room for maneuver and the likelihood of action and also are important for determining the kind of support the Bank should provide. 54. Institutional and Organizational Analysis. Development outcomes depend on the strength of the country’s institutions. Virtually all CASs assign priority to strengthening institutions, but there is considerable overlap and conflation of organizations (capacity) and institutions (rules of the game, allocation arrangements). Most CASs clearly articulate the objectives of increasing government capacity (e.g., to reduce corruption and increase 25 transparency, efficiency, and targeting), but few provide country-specific details about institutional constraints that could serve as the basis for a focused strategy. The Latvia CAS, among many that make institutional reform central, stands out because it provides country- specific information about institutional constraints. It shows that the rules governing budget allocation and borrowing provide incentives that erode the capacity of crucial central ministries over time. This approach enables the Bank program to focus on Latvia’s particular institutional environment, rather than on generic capacity-building measures. 3. Bank Group Comparative Advantage 55. A proposed program or project may be a big-ticket item for poverty reduction with a high probability of government action, but still the Bank needs to be sure it has the comparative advantage to address it. What are development partners doing in the area? How good is the Bank’s track record in this country and sector? While many CASs use the vocabulary of strategic selectivity—comparative advantage, track record, appropriate instruments, and so on— they are only beginning to use selectivity itself as an analytic tool for focusing on certain sectors or themes and exiting from others. Furthermore, very few spell out why particular lending or nonlending instruments were selected. Future CASs will need to do a much better job, using the strategic selectivity framework to decide which activities the Bank should undertake, using which instruments. a. Partnerships 56. Partnerships, Coordination, and Comparative Advantage. As the Bank’s key country strategy instrument, the CAS should contain a discussion of the roles of various donors, the coordination among the donors, and the effects of the Bank’s program.22 The review of the FY98II-99 CASs found significant improvement in the discussion of partnerships and donor coordination. (Box 10 shows the results of the evaluation of CASs and cites good practice examples.) Coordination across sectors with other partners is discussed in three quarters of the CASs, and coordination within sectors (often health and education) in 60 percent. Almost all CASs discuss the Bank’s comparative advantage vis-à-vis other partners, but fewer than 40 percent discuss it in connection with program design for key areas. All the CASs that analyze the Bank’s comparative advantage mention the role of other MDBs and bilateral donors. Further improvement is needed, particularly in CASs for IDA countries, where severe donor coordination problems persist.23 All CASs should include a table showing the sectors in which the Bank and other donors are active. In addition, future CASs should give priority to strengthening the discussion of the Bank’s comparative advantage with respect to other partners, including the private sector. 22 See Partnership for Development: Proposed Actions for the World Bank (SecM98-421), May 21, 1998, and Partnership for Development: From Vision to Action (SecM986/789), September 23, 1998. 23 A recent OED review compared the discussion of aid coordination in the last two CASs for a set of 15 countries. The earlier CASs were prepared in FY94-96 and the later set in FY97-98. The results suggest that CASs are increasingly linking development effectiveness to aid coordination, and that they are more clearly delineating the Bank’s comparative advantage and selectivity with respect to other donors. However, the CAS discussions of country capacity for aid coordination, and of the donor efforts to build capacity and harmonize procedures, were found to be weak. See Review of Aid Coordination and the Role of the World Bank (Report No.19840), Operations Evaluation Department, October 28, 1999. 26 57. Key Partnerships—IMF. Virtually all the CASs (94 percent) include a substantive discussion of partnership between the Bank and the IMF. The relationship is particularly close in countries implementing substantial reform programs and in transition economies. The Mongolia CAS, for example, describes the extensive and frequent consultations under the ESAF program: IDA works closely with the IMF in supporting Mongolia’s stabilization and structural adjustment efforts, public expenditure management, and tax reform, through parallel missions, collaborative technical assistance, and updating the Policy Framework Paper (PFP). The Bangladesh CAS discusses Bank and IMF collaboration on assisting the government with integrating macroeconomic and structural reforms. In about 60 percent of CASs, satisfactory macroeconomic performance under an IMF-supported government program is a trigger for different lending scenarios. For countries preparing PRSPs, Bank-Fund collaboration will be even closer. Box 10. Partnerships Criteria. The retrospective examined the extent to which CASs discussed the roles of various donors, the coordination among donors, and the effects of the Bank’s program. The results of this evaluation are shown in the graph. CAS Treatment of Donor Coordination (in percent) 100 FY97-FY98I FY98II-FY99 80 60 40 20 0 Donor section Thorough donor Donor table Bank and IMF Intrasector Cross-sector Bank comp. Bank and NGO Bank comp. Flows affect section collaboration donor donor advantage by collaboration advantage by Bank selectivity coordination coordination sector instrument Quality of treatment Main issues discussed in CAS Good Practices in the CAS Treatment of Partnerships • Coverage of partnerships. The Senegal, Kyrgyz Republic, Macedonia, Philippines, and Lithuania CASs provide detailed information on partners’ activities in the matrix. • Coordination and comparative advantage. The Bolivia CAS describes how the Bank has coordinated the efforts of the government and donor community to use the same set of indicators to monitor progress in achieving the country’s development objectives. The Lithuania and Hungary CASs are especially noteworthy in their treatment of donor coordination in countries preparing for EU accession. The CASs for Mongolia, Thailand, the Philippines, and Macedonia analyze previous impact and client survey results. The Hungary and Macedonia CASs discuss the Bank’s comparative advantage and use it to define the Bank’s assistance program. The Latvia CAS, to leverage limited Bank funds, proposes using the Bank’s comparative advantage to develop sector frameworks for donor funding. 27 58. Regional Development Banks. Most CASs (85 percent) also discuss the Bank’s relationship with regional development banks, with many CASs reporting specific actions to improve coordination―for example, more frequent high-level meetings, joint portfolio reviews, and joint missions. (Box 11 summarizes the MDB work on coordinating and harmonizing their country strategies.) The Trinidad and Tobago CAS provides detailed information on IDB activities in the country―including IDB’s lending amount, which is larger than the Bank’s. It indicates that, in moving forward, the Bank will step up coordination with the IDB in order to increase the pace of implementation of structural reforms in crucial sectors. The Lao PDR CAS points out that the ADB currently provides more resources to Lao PDR than IDA does. The CAS discusses IDA’s collaboration with the ADB at the project/sector level in the areas of education, health, rural development, forestry, and transport; it also mentions intentions to collaborate with ADB in the financial sector. 59. Other Partnerships. Almost 60 percent of CASs discuss NGO activities (as compared with 30 percent in CAS Retro I) and highlight direct NGO involvement in the execution of Bank projects, particularly in the health, education, and environment sectors. The Bangladesh CAS, for example, discusses NGO or stakeholder involvement in almost all IDA-assisted projects; and the Vietnam CAS discusses the participation of an NGO in poverty monitoring, analysis, and targeting. Recent CASs also focus on partnerships with the private sector and civil society in subproject development and implementation. The Moldova and Bulgaria CASs highlight partnerships between the Bank and private foundations and institutes, and the Philippines CAS describes the Bank’s partnerships with the chamber of commerce. Other CASs report on the Bank’s collaboration with the EU, bilateral agencies, and UN agencies (see Box 12 on UNDAF). In almost all CAS matrices, the involvement of bilateral donors in each area is clearly specified. b. Track Record and Portfolio Management 60. A candid discussion of the Bank’s track record in the client country, particularly since the last CAS, is essential. This discussion should draw on several sources, including self-evaluation, client surveys, and OED evaluations, and should highlight what worked well and what worked less well. A frank evaluation of the progress toward achieving the goals set in the previous CAS also contributes to effective monitoring and accountability management. 61. Coverage of Track Record and Lessons Learned. All the CASs reviewed for this report include a discussion of the Bank’s track record. All discuss portfolio performance, a key ingredient of track record, but only 78 percent apply relevant lessons from the implementation of the previous CAS. (Box 13 shows the results of this evaluation and cites some good practice CASs.) The FY98II-99 CASs continue the trend of strengthening strategy design by discussing implementation lessons from the previous CAS and by taking them into account in the formulation of future strategy. However, task teams do not seem to have made full use of all available Bank sources and seem to underuse, in particular, the body of work from OED and the Quality Assurance Group (QAG). 28 Box 11. Multilateral Development Institutions: Country Strategies Roundtable. To encourage dialogue among bilateral and multilateral development institutions, Operations Policy and Strategy (OPS) organized a Roundtable on Operational Policies and Strategies in June 1999. In the session on country strategies, representatives from each institution shared information about country strategy formulation and execution, and about issues their institution faces. Points of Comparison. The country strategies of the major multilateral development banks (MDBs)―AfDB, ADB, EBRD, IDB―take an approach to key areas that is similar to the Bank’s. Most use country strategies as a central business planning tool and a focal point of Board/Management relations and donor coordination. The AfDB, ADB, and World Bank increasingly use them as a key instrument for Management oversight of staff. All of the strategies cover roughly the same ground, focusing on country context, government priorities and strategy, previous program implementation, comparative advantage of the institution, and other donors’ programs. The strategies analyze risks and specify assistance triggers and lending scenarios. Most also spell out programs of support that include both lending and nonlending activities. Many of the institutions do not require that strategies be fully costed and budgeted for implementation, although the World Bank is moving in that direction. Most do include a mechanism for monitoring and evaluation, as well as independent evaluation along the lines of OED’s. Finally, only the ADB routinely discloses its country strategies to the public, although the EBRD, IDB, and World Bank are reviewing their disclosure policies. Coordination. Overall, as the MDBs increase their focus on donor coordination under a central framework such as the CDF, they are seeking to better coordinate their country strategies. Of particular interest is how internal similarities and differences in the management of policies, practices, and strategies affect their ability to collaborate in support of country-level and regional development efforts. To facilitate this discussion, participants in the MDB roundtables have formed a standing committee that will meet twice yearly to identify areas in which to promote collaboration and harmonization. The matrix below compares some of the features of MDB country strategies. Item AfDB ADB EBRD IDB UNDP World Bank Group Country Strategies: Formulation Name of document? Country Country Country Country Paper Common Country Strategy Operational Strategy (CS) (CP) Country Assistance Paper (CSP) Strategy Assessment Strategy (CAS) (COS); (CCA); UN Country Development Assistance Assistance Plan (CAP) Framework (UNDAF) Client government consultation Yes Yes Yes Yes Yes required? Client government approval required? Yes No No No No Civil society consultation required? Yes Yes No Encouraged Encouraged Donor partner consultation required? No No No Encouraged Encouraged Country Strategies: Content Recent country political and social Yes Yes Yes Yes Yes Yes developments reviewed? Previous assistance strategy reviewed Yes Yes Yes Yes Yes and performance evaluated? Portfolio performance reviewed? Yes Yes Yes N/A Yes Assistance triggers and scenarios Yes Yes Yes (non- N/A Yes specified? quantitative) Monitorable performance indicators Yes Yes (in COS) No No Yes Increasingly specified? Country Strategies: Disclosure Routine full public disclosure? Partial Yes No; under No Yes IDA yes; review IBRD optional Press release after Board discussion? No No No N/A Yes 29 Box 12. UNDAF Pilot Countries―Mali and Vietnam The UN Development Assistance Framework (UNDAF) and the Bank’s CAS fulfill similar functions. To harmonize work under these two instruments, the Bank and the UN have agreed to pilot their parallel preparation in a few countries. The FY98II-99 CAS cohort included two of the UNDAF pilot countries, Mali and Vietnam. Both CASs have shown positive experiences with the pilot. • The Mali CAS indicates that country office staff have actively contributed to the thematic groups on governance and the social and productive sectors as part of UNDAF preparation. In return, Bank staff will continue to keep the UN agencies abreast of the CAS process through regular updates at the monthly UN interagency coordination meetings. • The Vietnam CAS indicates that the coordinator of the UN system in Vietnam and heads of the various UN agencies have been heavily engaged in CAS discussions. Also, as a result of the joint discussions on the UNDAF, the Bank and UN agencies are exploring the possibility of collaboration in the preparation of two targeted poverty projects. (Strengthening Partnerships, a report requested by the IDA12 Deputies, also indicated that the Bank has played an active role in the development of the UNDAF document in Vietnam.) Box 13. Track Record and Portfolio Management in the CAS Criteria. The retrospective used the following criteria to evaluate the CASs’ treatment of the Bank’s track record and of portfolio management: • Does the CAS discuss portfolio performance? Does it examine the track record beyond portfolio performance? • Does the CAS apply relevant lessons from implementation of the previous CAS? Does it apply lessons from the track record beyond portfolio performance? • Does it use the insights from various sources to integrate portfolio management into the CAS? If it does, what is the depth of that integration? The results of this evaluation are shown in the graphs. Lessons Learned Portfolio Management―Integrating Lessons (in percent) (in percent) 100 100 Incorporates portfolio lessons 100 95 FY97-FY98I FY98II-FY99II 95 Analysis of other areas of track record 80 84 80 Link between other track record and portfolio management 75 78 Overall rating 65 73 60 60 57 19 51 62 40 40 45 43 16 35 20 20 21 22 14 14 0 30 14 14 16 0 0 Portfolio Record beyond Apply portfolio Apply lessons Use of OED Use of QAG Unsatisfactory Marginally Satisfactory Highly satisfactory performance portfolio lessons beyond work work satisfactory portfolio Good Practices in CAS Treatment of Portfolio Management • Overall. The CASs for Albania, Bulgaria, Lithuania, the Philippines, Vietnam, and Yemen stand out for their good treatment of portfolio management issues. • Lessons of experience. The CASs for India, Mexico, and the Philippines were rated particularly good in incorporating the lessons of experience. 30 62. Use of OED Findings. OED’s Country Assistance Reviews (CARs) and Country Assistance Notes (CANs) provide important lessons about the implementation of previous Bank country strategies.24 CAS Retro I found that about only one in five CASs explicitly discussed OED findings. For the FY 98II-99 CASs, that share increased to one in four. This improvement resulted from the increased number of OED reviews completed during this period, as well as from the efforts made to systematically schedule these reviews upstream of the CAS to provide the task teams with the benefit of OED’s findings. A closer coordination of schedules has helped to mainstream OED findings into the CAS and to link evaluation of the impact of Bank activities with design of the near-term strategy. However, there is a continuing problem, in that OED’s results tend to be delivered too late to provide much upstream guidance. (Box 14 describes the links between CASs and CARs/CANs.) Box 14. Integrating Findings from OED Evaluations into CASs OED Country Assistance Reviews (CARs) and Country Assistance Notes (CANs) provided an independent evaluation of the Bank’s assistance program in a country over the previous years. CANs were mini-CARs. Both drew lessons from past experience and were intended to serve as inputs to future Bank strategy. Nine of the ten FY98II-99 CASs, and four of the five PRs, for which an OED evaluation was available discuss the findings from the evaluation. The CASs that discuss the findings from the corresponding OED evaluations are those for Albania, Bangladesh, Bolivia, Kenya, Malawi, Nepal, Thailand, the Philippines, and Yemen; and the PRs are those for Indonesia, Sri Lanka, the Philippines, and Ukraine. Good Practices in Integrating Findings from OED into CASs • Assessment of Bank assistance. The Yemen CAS uses OED findings to discuss the evolution of Bank assistance since 1980, assess its impact, and draw lessons. The OED review found that the Bank could improve the impact of its program by (a) setting clear priorities and ensuring that its activities are consistent with those priorities; and (b) ensuring the sustainability of investment. Four areas of emphasis are suggested: removing constraints to growth, reducing the gender gap, strengthening water management, and improving governance. Several of OED’s suggestions on how to improve the sustainability of Bank interventions are echoed in the CAS, such as strengthening institutional capacity, phasing out autonomous project implementation, simplifying project design, and taking more time to prepare operations. • OED CARs help strengthen in-country strategies. Several of the strategic elements in the Bangladesh CAS were supported by the findings of OED’s CAR, such as the need to address institutional constraints and increase political commitment to institutional change. The CAR made several recommendations to improve the relevance and implementation of IDA programs that are closely aligned with those in the CAS. The Philippines CAS also reflects the results of the CAR, placing renewed emphasis on the financial sector, a dialogue on the judicial system, the importance of an integrated Bank Group strategy, a new rural strategy, and enhanced partnerships with donors and civil society. The lessons identified in the Albania CAR, such as the need to prioritize assistance efforts, most notably in the financial sector, also contributed to shaping the CAS. c. Instrument Choice 63. The choice of instruments is another dimension of selectivity: What is the preferred combination of lending and nonlending activities for a particular sector, given our knowledge of the area and the level of policy dialogue? What types of lending instruments should be used, given the country policy and institutional setting? 24 OED no longer differentiates between these two products; all such reviews are now referred to as Country Assistance Evaluations (CAEs). 31 64. Composition of the Program. To date, CASs have not explicitly addressed these issues. They are particularly silent on the program of nonlending activities. While they discuss changes in the level of lending in response to policy performance, they generally do not do the same for the ESW program. This does not mean, of course, that country teams did not take these issues into account during CAS preparation. Nevertheless, future CASs should explicitly address the question of the rationale for the mix of lending and nonlending activities and the effects that changes in Bank scenarios would have for nonlending services. They also should mention the relative cost of the Bank’s involvement in an activity or sector. This criterion is the one least employed by FY98II-99 CASs: few discuss the relative cost or trade-offs of Bank involvement. The Kenya CAS is an exception; and the Bulgaria, Hungary, and Latvia CASs discuss cost dimensions in the context of the design of their programs. 65. Choice of Lending Instrument. Few CASs discuss the choice of lending instruments.25 The Trinidad and Tobago CAS explicitly links lending instruments with particular objectives. Learning and innovation loans (LILs) will be used where the scope of reform effort is unclear and where the Bank can play a catalytic role in support of partnerships; technical assistance loans (TALs) will be used to strengthen institutional capacity in key sectors; small investment loans will support sectoral reforms; and, finally, partial risk guarantees will be used to foster private sector participation in the water sector. The Bolivia CAS indicates that although lending will consist mainly of project loans and technical assistance, the Bank will use sector investment loans to support a time-slice or subset of government investments in a sector and use adjustment loans to promote significant structural reforms. The CAS also discusses the use of LILs and adaptable program loans (APLs).26 LILs are proposed for use where alternatives need to be tested and the Bank can play a catalytic role in building development partnerships. In areas where sector knowledge is strong, long-term objectives are clear, and country ownership of the process established, APLs would be the preferred option. The Thailand CAS discusses the conditions under which a single-tranche operation would be preferred to a multitranche operation. It concludes that small single-tranche operations are more effective than large multitranche operations in the context of Thailand’s rapidly evolving situation, provided they are embedded in a coherent medium-term reform program. Several CASs use experience from Bank operations to recommend changes in lending instruments, such as LILs and APLS, as well as greater reliance on nonlending services. The Gambia, Philippines, and Yemen CASs are good examples in this area. 66. Implementation Capacity. An important element in the choice of instruments is compatibility with the country’s implementation capacity. About 90 percent of CASs discuss the country’s implementation capacity, but fewer than two-thirds link that discussion with program design. To address institutional capacity constraints, nearly one half of the CASs propose technical assistance programs, while about one fourth suggest less complex project designs. CASs may also propose adjusting the timing or sequencing of reforms, making more use of 25 See Programmatic and Emergency Adjustment Lending: World Bank Guidelines (R98-249), October 2, 1998, which provides references on the various types of lending, and discusses programmatic and special adjustment loans. Complementing this discussion, the paper, World Bank Policy-Based Guarantees (R99-53), April 2, 1999, provides an account of the evolution of the Bank’s guarantee instruments—including IBRD partial credit and partial risk guarantees and IDA partial risk guarantees—and provides the background for the recent approval, on a pilot basis, of IBRD policy-based guarantees. See also World Bank Lending Instrument: Resources for Development Impact, OPS, April 2000. 26 See Adaptable Lending—New Investment Instruments (R97-203), August 5, 1997. 32 NGOs, postponing lending, increasing capacity in the field, and using different instruments. The Bangladesh, Kenya, Moldova, and Trinidad and Tobago CASs offer good examples of discussions of capacity constraints. Given the Bank’s emphasis on decentralization, it is important to analyze local-level capacity to identify governance challenges and to develop a concrete set of measures to address them. Few CASs do this. One exception, the Nepal CAS, is highly focused on local capacity, largely reflecting the government’s emphasis on decentralization. It highlights the importance of adequate information for excluded groups, such as farmers, as well as for rural groups whose members are likely to be unaware of government initiatives for service delivery. 67. ESW Underpinnings. Strategy design benefits strongly from the existence of a body of analytical information that diagnoses the major economic issues and outlines policy options to address them. About 80 percent of CASs (86 percent of IDA CASs) had at least one piece of ESW available (a PER, poverty assessment, or country economic memorandum) to assist in CAS formulation. However, for only 50 percent of the CAS countries (62 percent of IDA CASs) was a PER prepared within the two years preceding the CAS; and for only 46 percent of the countries (57 percent of IDA countries) was a poverty assessment (PA) prepared in the previous two years. PERs and PAs are the Bank’s major analytical pieces in these two key areas. CASs use the results of available PAs, but could do better. More than 60 percent of the CASs link the proposed Bank program to the country’s poverty challenges in a satisfactory or better manner, but fewer CASs relate the reasons why a certain program should be preferred over another. As discussed above, a major deficiency of the CAS poverty focus is the lack of recent data. The absence of up-to-date information implies that our knowledge on the poverty situation in a specific country is outdated and incomplete, so that the poverty reduction strategy is likely to be weak. The same is true for public expenditure reform initiatives. (Of course, alternative sources of information can narrow potential information gaps; for example, IMF work and staff policy notes could fill some information requirements.) These results suggest that to strengthen the CAS, the Bank needs to carry out regular analytic work in core areas, and then use it in country dialogue and in strategy design. The upcoming ESW Retrospective will take up the issue of strengthening the country-based product portfolio, and the appropriate periodicity for diagnostic ESW. B. Calibrating Program Size 68. CAS programs need to take into account two types of budget envelopes―the amount of lending and the administrative budget. In IDA countries, the allocation of IDA resources is based on performance, but it also takes into account such other considerations as absorptive capacity (see Box 15). In IBRD countries, lending amounts take into account the Bank’s institutional risk-bearing capacity and the country’s creditworthiness. But the CAS program also needs to take into account administrative budget constraints. During CAS preparation, the full set of CAS activities is fully costed (lending, nonlending, and supervision activities) to ensure that the CAS program is consistent with budgetary allocations. Taking into account these two types of budget constraints, the CAS sets out a realistic program of activities. The following discussion covers program design and levels of support. It discusses how assistance scenarios should be linked to medium-term projections and should include specific triggers related to policy performance. 33 Box 15. IDA Allocations and Country Performance To promote aid effectiveness, IDA links allocations to performance, concentrating resources where they are likely to have the greatest impact and ensuring consistent treatment among IDA-eligible countries. The figure shows the correlation between country performance and actual lending to IDA-only countries under IDA11: better performing countries get more IDA resources per capita. Relationship between IDA’s country performance ratings are derived from two IDA Lending and Performance, FY97-99 assessments: 25 • The annual Country Policy and Institutional Assessment Annaual Average Per Capita IDA Lending (CPIA),a which assesses the quality of each borrower’s 20 policy and institutional performance in areas relevant to economic growth and poverty reduction. The CPIA 15 consists of 20 criteria covering economic management, (US$) structural policies, policies for social inclusion/equity, and public sector management and institutions. The 10 country is rated on a scale of 1 (low) to 6 (high). • A measure derived from the implementation performance 5 ratings for IDA projects. The resulting assessment is lowered if country governance 0 conditions are very weak—the “governance discount.” Poor Excellent Average Performance Rating Country IDA allocations derived from the CPIA anchor CAS lending programs, with actual lending programs calibrated to reflect special country conditions. For example, all things being equal, allocations would be reduced for countries with absorptive capacity constraints and increased for post-conflict countries. Within the intervening years of the 2- to 3-year CAS cycle, the IDA envelope is adjusted in line with country performance, per the CAS triggers. In exceptional cases, when changes in country conditions exceed what was envisaged in the CAS triggers, and the revised IDA allocations and lending scenarios would fall outside the CAS scenarios, a CAS Progress Report or a new CAS is warranted to inform the Board of the implications of the new developments. Lending levels in the CAS, including the outcome of the allocation process and the criteria and country-specific judgments on which these allocations are based, are discussed with the country authorities. However, the details of the process linking country performance and IDA allocations have generally been less systematically discussed. For the future it will be important to ensure that country authorities understand the methodology underlying the CPIA, the rating process, the criteria linking performance to IDA allocations, the rationale for their country ratings, and the specific links to their IDA allocations. This must be part of the Bank’s dialogue with the country authorities. _________________________________________ a See Country Performance and IDA Allocations (IDA/SecM98-502), September 16, 1998, and IDA Country Performance Rating Process—Annual Report 1999 (IDA/SECM2000-58), February 10, 2000. 1. Levels of Bank Support 69. From a business perspective, the CAS should determine the overall level of Bank support under alternative economic and policy scenarios―typically a base-case scenario plus a low-case scenario; sometimes a high-case scenario is also used. In most CASs the base case represents the starting position and the most likely scenario, and should be linked to the country’s medium-term prospects. CAS Retro I recommended that future CASs, especially those for emerging market economies, include a thorough analysis of macro-fiscal and financial risks, including debt- sustainability risks, if important. The CAS should establish a set of quantitative and qualitative triggers that specify in a transparent manner how the Bank’s assistance will vary in response to country policy performance. In addition, the CAS should clearly identify the risks associated 34 with the strategy and, to the extent possible, highlight the risks that are associated with events that are most likely to occur and that will have the largest country impact. At the country level, CAS lending levels, triggers, and risk assessments determine the lending envelope and set the key parameters for all Bank lending activities in the country. 2. Lending Programs 70. This review found that about nine out of ten CASs do a satisfactory or better job of linking the lending program to the identified priorities. Most IDA CASs explicitly address how the IDA allocation relates to country performance, including governance needs and capabilities. 71. Lending Scenarios. All CASs specify the lending program covering the CAS period for the scenario that is considered the most likely. The base case is generally considered most likely, but in a few CASs the starting point is the low case (Kenya and Belarus) or the high case (Bulgaria, Albania, and Macedonia). About 63 percent of the CASs include three lending scenarios, and 30 percent include two lending scenarios (usually a base and a low case). On average, low-case lending levels are 36 percent below base-case volumes, and high-case levels are 52 percent above base-case volumes. In some CASs no lending is contemplated in the low case, and thus the difference between the low and the base cases is even greater. The steepness of the performance-lending curve remains essentially the same as the one estimated in CAS Retro I, ranging from 41 percent below the base case for the low-case scenario, to 68 above the base case for the high-case scenario. Most CASs discuss how the composition of the lending program would change in response to changes in the level of Bank assistance. Although all CASs contain a detailed description of nonlending services, only a few make clear how the level and the composition of such services would change in parallel with changes in lending levels. In this area, little progress has been made since the previous retrospective. 3. Triggers 72. All CASs that have more than one lending scenario contain a set of triggers indicating how the Bank would move from one assistance scenario to another in response to country performance during CAS implementation. Good CASs should have tight links among the set of specified triggers, the key issues that are identified, and the major areas of Bank activity that are outlined in the CAS. Overall this review found that about two thirds of the CASs do a satisfactory or better job at linking these different areas of the assistance strategies. In IDA CASs these linkages are a little tighter―three fourths of the IDA CASs were found to be satisfactory or better in linking triggers, issues, and lending, in comparison with about one half of the IBRD CASs. Many CASs present the triggers in a separate table. To ensure consistency in the messages about country performance, the diagnostic tools used to identify priority issues (such as the CPIA) should inform the content of the triggers. 73. The most common triggers found in CASs are structural reform and macro-fiscal reform (83 percent and 81 percent of the CASs, respectively). Portfolio performance triggers are included in 71 percent of the CASs. The use of portfolio triggers is continuing to increase in parallel with the Bank’s growing emphasis on portfolio improvement. In general, IDA CASs place a stronger emphasis on structural and macro-fiscal reform and portfolio performance. About 90 percent of IDA CASs include macroeconomic triggers (67 percent for IBRD), 90 percent include structural reform triggers (75 percent for IBRD), and 80 percent include portfolio 35 performance triggers (60 percent for IBRD). These differences may reflect not only differences in country circumstances, but also the generally poorer condition of the portfolio in IDA countries. Not surprisingly, in view of the disclosure issues surrounding the CPIA, no CAS explicitly refers to the CPIA in the formulation of triggers. 74. Precision of Triggers. Clear and precise triggers are essential for monitoring country performance and guiding Bank decisions on lending levels during CAS implementation. In the CASs reviewed, 80 percent of the macroeconomic triggers are of a qualitative and general nature (e.g., satisfactory macroeconomic performance); 66 percent of the sectoral triggers are qualitative but specific (e.g., adoption of a particular policy measure or piece of legislation); and 94 percent of the portfolio triggers are quantitative (e.g., achieving a specific target under a portfolio implementation plan). This composition is broadly similar to that found in CAS Retro I. The degree of precision of the triggers has improved somewhat: structural reform triggers have become more qualitative-specific and portfolio triggers more quantitative. However, macroeconomic triggers continue to be general and qualitative. There are no significant differences between IDA and IBRD CASs in terms of the precision of their macroeconomic and structural triggers, although IDA CASs tend to a have a slightly higher share of qualitative-specific and quantitative triggers. Because quantitative and qualitative-specific triggers can usually be monitored more objectively than general quantitative triggers, future CASs should―to the extent possible―use these types of triggers. 75. Content of Triggers. As in CAS Retro I, this report found that among the broad spectrum of possible indicators of macro-fiscal, structural reform, and portfolio performance triggers, most CASs use a relatively narrow subgroup of triggers. The most commonly used macroeconomic triggers are satisfactory performance under an IMF-supported program and fiscal balance; for structural reform, privatization is the most frequently used trigger. (Figure 3 shows the results of the analysis of triggers.) With respect to the macroeconomic triggers, there is a sharp difference between IDA and IBRD countries, with 80 percent of IDA CASs containing performance under an IMF package as a trigger while only 30 percent of IBRD CASs do so. This difference reflects the fact that many more IDA programs are implemented in parallel with a stand-by or ESAF arrangement. In comparison to CAS Retro I, these CASs give less emphasis Figure 3. Content of Triggers by Type (in percent) 100 FY97-98I FY98II-99 80 60 40 20 0 Macro/fiscal Fiscal IMF Revenue Structural Financial Privatization Trade Educ/health Portfolio IP/DO Disbursement Effectiveness Project performance balance package target reform reforms reform expenditures problem lag delays specific projects Macro-fiscal management Structural reform Portfolio performance 36 to trade triggers. In addition, 39 percent of the FY98II-99 CASs contain institutional reform triggers (47 percent for IBRD CASs) and 14 percent contain governance triggers (20 percent for IDA CASs). In countries―and particularly IDA countries―where weak governance, including corruption, is a significant development constraint, CASs should contain governance triggers. C. Managing Risks 76. The CAS should discuss financial, creditworthiness, and reputational risks. It should differentiate between mainly country risks and mainly Bank risks. It should also distinguish between risks that are endogenous and can be attenuated by adequate government or Bank actions, and risks that are exogenous and not subject to risk mitigation. Equally, the CAS should acknowledge candidly the risks that remain after the mitigation options have been exhausted, and should explain how such residual risks may affect creditworthiness. Particularly for large borrowers, the CAS should discuss exposure risk―the management of exposure and the factors influencing country exposure and risks―and should discuss how changes in external environment could affect the country’s access to external financial markets and creditworthiness. 77. Identification of Constraints and Risks. Virtually all CASs list several external and domestic factors that could constrain or threaten the country’s development effort. The most frequently identified risks are political risks (political uncertainty), adverse external events, and backsliding in implementing structural reforms. The most frequently identified external factors are changes in the prices of key export or import commodities, movements in exchange rates and capital flows, and economic developments in neighboring countries. External risks are identified more prominently in recent CASs than in those reviewed for CAS Retro I; this trend largely reflects the turbulent events in the last few years in world financial markets, as well as softness in some key commodity prices. (Figure 4 shows the kinds of risks covered in CASs, and the findings on the quality of the treatment.) Weak implementation is identified as a risk in 47 percent of all CASs (57 percent of IDA and 31 percent of IBRD CASs), and a weak financial sector in about 30 percent of all CASs (47 percent of IBRD and 14 percent of IDA CASs). Corruption and governance are identified as risks in 21 percent of the CASs. The majority of CASs treat financial vulnerability inadequately (as was also true in CAS Retro I), although the Figure 4. Treatment of Risk in CASs (in percent) 100 FY97-98I FY98II-99 80 60 40 20 0 Includes risks Risk Satisfactory Best Political Non-macro Slippage in External Implementation Corruption/ Financial section classification risk practice risk environment reform macro-fiscal environment capacity governance sector as H/M/L analysis analysis backtracking policy development Quality of treatment Main risks identified in CAS 37 major middle-income borrowers—such as Mexico, Thailand, and Malaysia—do handle the issue satisfactorily. Although the identification of major risks is virtually universal (as it was in CAS Retro I), the assessment of the constraints and risks varies from comprehensive (in a few cases— Bolivia, Mexico, Thailand, and Vietnam), to cursory (in more than half the CASs). The discussion could be enriched by a discussion of how other partners or private observers assess the country’s outlook, but no CAS does this. 78. Quality of Risk Assessment. Essentially all FY98II-99 CASs contain a discussion of risks. But the quality of treatment is very uneven and in many cases it falls short of satisfactory. In general there seems to have been little improvement in this area since CAS Retro I. Less than one-third of CASs classify risks according to their perceived importance―high, medium, and low. Slightly more than one in every three CASs does a satisfactory job in analyzing the impact of key risks, about the same share that was found to do so in CAS Retro I. There are no significant differences between IDA and IBRD CASs with respect to the quality of their risk treatment. About 60 percent of CASs discuss mitigation measures (44 percent of the IBRD CASs and 71 percent of IDA CASs), but very few discuss residual risks. How risks affect creditworthiness remains poorly addressed in CASs: only half of the IBRD CASs contain a satisfactory analysis of how country creditworthiness is affected if events associated with the risks materialize. Although differences in methodologies and objectives make comparisons of risk assessment difficult, it is clear that this is an area that deserves further attention in the CAS. Future CASs should provide a more candid analysis of the risks―country risks as well as those risks associated with the implementation of the strategy―and should link them closely with the discussion of the medium-term economic scenarios. 79. Impact and Mitigation of Identified Risk Factors. In the majority of CASs, the risks are expressed in general terms, such as “global economic slowdown,” “export price fluctuations,” or “lack of progress on the reform program.” This level of generality makes it difficult to assess the likelihood that a problem will occur, the potential impact on the economy of such a problem, or the possibilities for mitigating these effects. In this area, there has been little improvement since CAS Retro I. Having identified key risks in the internal or external environment, CASs should also discuss the economic implications for the assistance strategy if the events associated with these risks materialize. A few CASs represent good practice on specific elements of risk. The Gabon CAS assesses how GDP growth would be affected by a specified drop in oil export prices; the Macedonia CAS proposes a contingency plan if more than a specified number of refugees enter from Kosovo; the Mali CAS assesses the potential impact of the central bank’s gold sales on gold prices and on its terms of trade; and the Bolivia CAS contains a comprehensive sensitivity analysis that projects growth on the basis of future domestic policies and external environments. 80. Linking Country Outlook and Bank Scenarios. In a good CAS, the scenarios of Bank support are clearly defined, triggers are specific and transparently presented, and the risks are accurately and candidly identified. The choice of triggers and the assessment of risks are focused on the key development challenges and the likelihood that Bank support will help to address them effectively. The base/low/high-case lending scenarios presented in virtually all CASs are intended to describe how the Bank’s strategy would respond to better or worse performance by the government in managing the economy. But government actions are not necessarily the only variable determining economic outcomes. The different scenarios for the country strategy thus capture only a narrow range of possible eventualities. All CASs should 38 identify the two or three major risk variables affecting the economic outlook, and should develop at least partial sensitivity estimates on key outcomes. The implications should be reflected in the strategy and proposed risk mitigation measures. Of the CASs reviewed, about one third describe low- and high-case economic (as distinct from lending) scenarios that would result from specific changes in the environment (Bolivia, Croatia, Hungary, Latvia, Lesotho, Moldova, and Yemen, among others). A few CASs—those of Gambia, Gabon, Senegal, and Uzbekistan—go farther and project how GDP growth and other economic outcomes would be affected by the alternative scenarios. However, only in some cases are the implications of these results integrated into the story line of the CAS. 81. Linking External Environment and Lending. The external environment exerts a major influence on Bank lending. Historically, changes in Bank lending volumes and mix have tracked changes in client countries’ external environments. IBRD adjustment lending tends to move countercyclically with portfolio flows to developing countries, while IDA flows tend to be more affected by commodity prices. IBRD lending, especially adjustment lending, has tended to increase as countries’ current account deficits have increased and portfolio flows have declined, and vice versa. Total IDA lending and IDA adjustment lending have moved countercyclically with real commodity prices. Thus changes in the external environment have repercussions at the country level as well as on the level and composition of lending Bankwide. With the integration of financial markets and the increase in contagion risks, it is important that CASs in general, and those for emerging markets in particular, include a discussion of the implications these risks hold for lending scenarios. Such a discussion would strengthen the link between risk analysis and the strategic approach. D. Keeping Score 82. As the major instrument guiding the Bank’s strategy and program, the CAS should contain a self-evaluation framework for assessing Bank and country performance. As CAS Retro I suggested, a key contribution of a forward-looking self-evaluation is to set benchmarks for assessing CAS implementation, with particular attention to Bank performance. A self- evaluation framework helps to clarify the objectives of the Bank strategy; over time it helps to guide adjustments in the Bank’s lending and nonlending activities. It also helps, during the preparation of a subsequent CAS, to evaluate how well those objectives were achieved and to draw lessons for the future. Such lessons could be complemented by inputs from other sources, such as client surveys and reviews by OED. 83. Features of a Good Self-evaluation Framework. To ensure client ownership of the evaluation process, country benchmarks should be determined in a participatory way. The Bank’s benchmarks (as distinct from country benchmarks) should focus on measuring what was achieved in the areas where the Bank has played an important role.27 They should be, to the extent possible, specific and quantitative, and associated with a set of readily available monitoring indicators that permit the tracking of progress. The CASs reviewed for this 27 Outputs, the direct product of an organization’s activities, are distinguished from outcomes, which are what is achieved in relation to objectives. A school building program, for example, may result in the construction of a certain number of schools in a given district (output) but may not necessarily lead to the desired increase in child literacy rates in that district (outcome). See OED’s paper 1998 Annual Report on Operations Evaluation (Report No. 18579), November 1998. 39 retrospective were evaluated for these features. (Box 16 describes the evaluation criteria and results and gives some examples of good practice.) 84. Areas for Improvement. The self-evaluation framework in most CASs consists of a large set of benchmarks and monitoring indicators that impose heavy information requirements. A smaller and more thoughtful set of core country and Bank performance benchmarks would reduce the information requirements and help focus attention on progress toward achieving key objectives. Only 22 percent of the CASs reviewed for this retrospective present a core set of benchmarks, most of them country benchmarks, as in the CASs for the Philippines, Yemen, Lao PDR, Malaysia, and Bolivia. Indicators in the remaining CASs lack prioritization and specificity and tend to be broad statements of objectives that are diffuse and not easily monitored. Only 35 percent of CASs reviewed explicitly discuss the information strategy to be used to assess implementation progress. Most CASs—such as those for Nicaragua, Thailand, Gabon, and Panama—leave that task to project supervision and CPPRs, supplemented in some cases by client feedback surveys (Lao PDR and Lithuania). The South Africa CAS, with its strong emphasis on analytical and advisory services, proposes that an external review be conducted every two years to follow up on earlier evaluations. 86. Client Surveys. The Strategic Compact called for more systematic surveying of key stakeholders to gauge the quality and effectiveness of the Bank’s products and services. In response, a client survey program was piloted in FY98 and mainstreamed in FY99. Two years of experience have shown that the client survey can be a valuable feedback and evaluation tool for country teams, assisting them in identifying strong and weak points in their work with clients, and in allocating effort accordingly. A total of 33 surveys, 12 in FY98 and 21 in FY99, covering all Regions, have been completed since the program began. Of these, 16 fed directly into preparation of CASs. Among the CASs that reported on client surveys were those for Azerbaijan, Croatia, Lao PDR, Lithuania, Macedonia, Mongolia, the Philippines, and Vietnam. The most frequently mentioned area of client interest is the degree of Bank focus on the country’s top development priorities. Generally, the CASs report that clients are satisfied with Bank performance in this area. Areas needing improvement are flexibility in adjusting to changing country circumstances (particularly the application of operational policies), Bank use of local expertise, and improving capacity at the local level. 87. Self-evaluation Pilots. Nine CASs were selected at the time of the last retrospective for a pilot self-evaluation program. Six of these self-evaluations were completed during the FY98II- 99 period—Bolivia, Lithuania, Mali, Macedonia, the Philippines, and Yemen. Although all CASs prepared during FY98II-99 include benchmarks and/or monitoring indicators, the self- evaluation pilots stand out for the richness and completeness of their self-evaluation frameworks. Four of the six pilots meet the self-evaluation criteria outlined in Box 16 in a satisfactory or better manner, compared to only 16 percent for the non-pilot CASs. All the pilot CASs differentiate between country and Bank benchmarks, and all satisfactorily evaluate the outcome dimension, compared to 29 and 55 percent of the non-pilot CASs, respectively. 40 Box 16. Self-evaluation in the CAS Criteria. The retrospective used the following criteria to assess CASs’ self-evaluation frameworks: • Does the CAS differentiate between country and Bank benchmarks? • Do the benchmarks attempt to measure development impact? • Does the CAS select core benchmarks on which to base the evaluation? • Were the benchmarks set in a participatory way, and with explicit reference to Bank analytical work, cross- country evidence, or other empirical evidence? • Does the CAS information strategy ensure that timely quantitative information will be available to guide CAS implementation? Scores. Three broad conclusions emerge from this review. First, although all CASs include benchmarks, the overall quality of the self-evaluation frameworks varies widely and in general is weak. This is to be expected, given the novelty of this dimension in the CAS and the difficulty of setting up quantitative, outcome-oriented benchmarks, determining their baseline values, and ensuring that monitoring indicators are available. These problems were a major reason for the introduction of the pilot self-evaluation program described in para 87. Second, as could be expected, the quality of the self-evaluation frameworks of the pilot CASs stands out. Third, there has been a slight improvement over time in the quality of self-evaluation frameworks overall, possibly as a result of the dissemination of best practices from the pilots. Self-evaluation by CAS Group Self-evaluation over Time (in percent) (in percent) 100 100 All CASs Pilot Non-pilots FY98 II FY99 I FY99 II 80 80 60 46 48 60 50 54 42 35 33 33 33 35 40 30 40 25 25 19 16 18 14 18 20 5 20 7 9 0 0 0 0 0 Unsatisfactory Marginally Satisfactory Highly Unsatisfactory Marginally Satisfactory Highly satisfactory satisfactory satisfactory satisfactory Good Practices in Self-evaluation Frameworks. The Bolivia and Philippines CASs distinguish between Bank and government benchmarks and propose a set of core benchmarks. In the Bolivia CAS, for example, the core benchmarks include the poverty headcount ratio, poverty gap ratio, prevalence of child malnutrition, infant mortality rate, and maternal mortality rate. Highlighting the participatory dimension, the CAS indicates that the development outcomes were suggested to the government by the donor community at a recent CG meeting, and that the specific quantitative targets were proposed by the Bank and will be discussed with government and donors in the near future. The Philippines CAS also includes a list of country performance indicators and targets, as well as a core set of benchmarks. For both countries, the CAS matrices show a wide set of quantitative and time-bound, outcome- oriented benchmarks. The Bangladesh CAS relates the country benchmarks to the OECD DAC goals for 2015. 88. Corporate Scorecard. The CAS self-evaluation pilots provide useful, albeit preliminary, lessons for mainstreaming CAS self-evaluation. However, full assessment of the pilot program will have to wait until the next CAS cycle, when the self-evaluations will compare benchmarks with actual outcomes. In the meantime, with the ultimate objective of including CAS ratings in the Corporate Scorecard, we intend to begin addressing the practical problems involved in refining and operationalizing the underlying methodology. For example, the fact that only a small number of CASs is completed each year will present problems of statistical significance―how to interpret changes in CAS ratings over time and across units. Over time, we plan to include three levels of CAS strategic quality rating: 41 • Design—rated by the quality of country focus, poverty focus, diagnostic quality, and strategic selectivity. • Implementation—judged by Bank performance against the benchmarks set out by the country team in the self-evaluation framework of the CAS, including the results of the client surveys. • Impact—assessed by OED, in the context of Country Assistance Evaluations. E. Checking Progress 89. Progress Reports (PRs) keep Management and the Board apprised of the implementation of country programs. This function is routinely carried out every other year for major IBRD and IDA borrowers.28 But PRs have also been used in situations where political events or elections necessitate adjustment of the CAS cycle. A PR is expected to document progress in implementing the strategy and confirm that the CAS objectives and strategy remain appropriate. It should report on the status of the Bank Group’s dialogue with country authorities and key Bank group partnerships, and should offer an update on economic and political developments, prospects, and risks. It should focus on key developments since the last CAS, following the general outline of a CAS but in a shorter format. Although a PR may update and fine-tune the lending program, it does not reassess the Bank’s overall lending and nonlending strategy in the country. 90. Background. PRs were not covered in CAS Retro I, and therefore there is no ready baseline for comparison. During FY98II-99, 16 PRs were presented to the Board, of which two were for Argentina (February 1998 and November 1998) and two for Ukraine (May 1998 and April 1999). About two-thirds of the PRs were prepared in the context of the normal CAS cycle. Other PRs were prepared in lieu of a new CAS, which was postponed because of a major election (Ukraine 1999 and the Philippines), or were prepared in the context of Board presentation of a major lending operation (Pakistan and Argentina). The average elapsed time between the CAS and a PR for a country was about 20 months—although the actual time varied between one year for most PRs (Argentina 1998, Bosnia and Herzegovina, India, and Mexico) and three years for Pakistan and Ukraine 1999. 91. Evaluation. Given the small number of PRs presented to the Board during the period under review, the conclusions of this review are tentative. (Box 17 describes the criteria used to evaluate the PRs and sets out some good practice examples.) In general, most PRs do a good job of discussing recent economic and political developments, progress in structural reforms, progress in implementing the Bank Group’s program, and portfolio issues. To some extent, the PRs mirror the CASs—that is, some of the areas identified here as needing improvement are the same areas highlighted for the CASs. Improvements are needed in the discussion of recent social and poverty reduction developments, the use of benchmarks and triggers to evaluate progress, the analysis of risks, and the integration of IFC and MIGA into the Bank Group strategy. 28 See Interim Guidelines for Preparing CASs and CAS Progress Reports, OPS, January 9, 1998 (revised April 9, 1998), and “CAS Guidelines for IDA12,” in Additions to IDA Resources: Twelfth Replenishment—A Partnership for Poverty Reduction (IDA/R98-195), December 23, 1998. 42 Box 17. Evaluation of CAS Progress Reports Criteria. The retrospective used the following criteria to evaluate the Progress Reports: • Are descriptions of relevant economic and noneconomic events since the CAS adequate? • How well does the PR discuss implementation of the country program? • How well does it show linkages with the CAS? Does it discuss the relevance of the CAS objectives, the lending program in the context of the lending scenarios and the triggers for moving between them, the reasons for any failure to meet the CAS objectives, and the use of benchmarks to evaluate performance? • How well does it discuss portfolio management and risk? The results of this evaluation are shown in the graphs. Assessment of Reform Progress Link with CASs (in percent) (in percent) 100 100 Relevance of CAS objectives Reform progress discussed 80 Position of lending scenarios w/CAS triggers 75 80 75 Reasons for falling short of objectives Reasons for lack of progress discussed Use of benchmarks to evaluate performance 57 58 60 56 60 51 40 40 29 29 25 19 17 19 17 19 21 20 20 14 6 6 8 0 0 0 0 0 0 0 Unsatisfactory Marginally Satisfactory Highly satisfactory Unsatisfactory Marginally Satisfactory Highly satisfactory satisfactory satisfactory Good Practice in Progress Reports • Recent developments. The Argentina 1998, Pakistan, and Sri Lanka PRs provide good discussions of recent economic and political issues. • Poverty reduction. The Indonesia PR uses survey data to analyze the social impact of the recent economic crisis. The data underline the complex ramifications of the crisis: urban areas were hit harder than rural areas, some regions fared better than others, and informal labor markets helped to absorb some effects of the crisis. The survey’s insights about the impact of the crisis on the poor and near-poor were used to strengthen the design of the Bank’s activities, particularly in the area of the social safety net. • Progress of reform. The Pakistan, Ukraine, and Sri Lanka PRs present a balanced view of sectors leading and lagging in reform, and do a good job of linking this discussion to the CAS. The Pakistan and Indonesia PRs contain good discussions of governance and corruption. • Use of OED and QAG findings. The Bosnia and Herzegovina PR incorporates the results of an OED assessment of Bank experience with post-conflict reconstruction. The Philippines, Sri Lanka, and Ukraine PRs include a discussion of several recommendations contained in OED’s evaluations for these countries. 92. Recent Developments. Most PRs contain a satisfactory or better discussion of recent events. The discussion of recent economic and political issues is satisfactory in about 80 percent of the PRs; the discussion of social issues is generally weaker. Overall, about two-thirds of the PRs contain a review of recent economic, political, and social events that provides the necessary background for discussion of the implementation of the country strategy. 93. Poverty Reduction. About 40 percent of the PRs contain an unsatisfactory discussion of progress made in poverty reduction. This problem is due largely to a lack of recent data. More than half of the PRs contain no data on the incidence of poverty, and 13 percent present data that are more than three years old. The problem of data obsolescence is clearly acknowledged in some cases (India), and plans put forward to correct it. However, most PRs do not lay out a plan to set up systems that periodically update poverty information. 43 94. Progress of Reform. Almost all PRs contain a satisfactory or better discussion of reform progress since the CAS, and include concrete examples. About 75 percent of the PRs identify areas where no progress has been made, and some (Ukraine and Sri Lanka) provide a substantive and clear discussion of the reasons for the lack of progress. Most PRs mention governance and corruption, but the discussion is often weak. Where governance and corruption are identified as issues, only one third of the reports contain a satisfactory discussion of their impact on Bank projects. Almost all PRs mention the financial sector, and most (70 percent) discuss it in a substantive manner. 95. Links with the CAS. To assess the linkages between the CAS and PR, this review looked at four key dimensions of the PR: (a) the extent of the discussion of the relevance of the CAS objectives; (b) the depth of discussion of the lending program in the context of the CAS scenarios, with particular emphasis on the use of the triggers; (c) the extent of the discussion, if appropriate, of the reasons why the CAS has fallen short of its objectives; and (d) the use of benchmarks to evaluate performance. About three-fourths of the PRs provide a satisfactory or better discussion of the first three areas. The discussion of benchmarks is generally weaker. Although most PRs attempt to use CAS benchmarks to evaluate progress, so many of the benchmarks are unprioritized and unspecific that meaningful evaluation is often not possible. In some instances, such as the Indonesia and Russia PRs, country circumstances have changed so radically that there is little point in trying to use the CAS program benchmarks to evaluate country performance. Lack of specificity also poses problems in the use of triggers to evaluate the positioning of the lending program in the context of the base, high, or low lending scenarios. Some PRs (25 percent) include a table or matrix to discuss the positioning of the Bank lending in the context of performance and triggers. 96. Portfolio Management. The treatment of portfolio issues in PRs is generally satisfactory, mirroring the results of the CAS evaluation. All the PRs but one include a discussion of the portfolio, and about 70 percent of these discussions are satisfactory or better. The Pakistan and Indonesia PRs are examples of good treatment of portfolio issues. Most PRs (80 percent) do a satisfactory or better job of applying past lessons to portfolio management issues. As with the CASs, however, lessons from other areas are not fully incorporated into portfolio management. All but one PR contain an action plan to address identified problems. More than half of the PRs that identify governance and corruption as a major issue discuss their impact on Bank projects, but not all PRs outline concrete steps to be taken. Like the CASs, many PRs do not fully make use of OED and QAG findings. 97. Bank’s Future Program. Although most PRs find that the CAS objectives remain valid, many (63 percent) recommend changes in lending scenarios. The justifications provided for such a change are satisfactory in 60 percent of the cases, but in some cases―mainly because the triggers are too general and qualitative―it is difficult to clearly discern the link between performance and triggers so as to support the change from one lending scenario to another. The most often used triggers relate to macroeconomic, structural reform, and portfolio issues. More specificity and selectivity in the design of triggers is needed to enable a more substantive and persuasive discussion of changes in lending scenarios. 98. Risks. All but one PR include a risk section. The risks most commonly identified are political risks, backtracking on structural reforms, slippage on the macro-fiscal front, and adverse external developments. However, they are not differentiated by probability of occurrence or 44 severity of impact. Similarly, most PRs for IBRD countries do not consider the impact on creditworthiness in the event that the risks materialize. Fewer than half of the PRs propose mitigating measures and discuss residual risks. The discussion of risks in PRs, like that in CASs, needs to be strengthened. 99. Next Steps. The Progress Report is a useful and flexible instrument to inform Management and the Board about country development issues and the progress made in implementing country strategies. It does not have to meet all the CAS requirements and can be prepared relatively quickly for use as part of the CAS cycle, when country conditions are not propitious for the preparation of a full CAS, or as a bridge until a full CAS can be prepared. Because a key objective of a PR is to assess CAS implementation, it is necessary that the parameters to be used in the evaluation be clearly laid out in the CAS. More selective benchmarks for assessing Bank as well as country performance, and explicit triggers, will facilitate the PR task of assessing progress in CAS implementation. Fuller discussion of recent developments on poverty and social development is warranted, notwithstanding the fact that such developments move slowly. V. CONCLUSIONS AND FOLLOW-UP 100. The last CAS retrospective outlined an ambitious reform agenda that included, among other things, further improvements in participation; the disclosure of the CAS; strengthening CAS diagnosis, particularly in cross-cutting areas; improving the poverty focus of the CAS; moving forward with strategic selectivity; and strengthening the discussion of scenarios, triggers, and risks. How far have we moved in implementing this agenda? 101. Country Issues. Most CASs now do a good job of presenting the government’s development agenda. Participation has largely been mainstreamed, and the quality of the participatory process also is improving; but we need to better focus our consultations, moving beyond multiple open-ended consultations and becoming more issues-oriented, drawing on upstream stakeholder analysis. The new disclosure policy for CASs has been well received. On the external environment, however—an important CAS section for borrowers—we still need to do a better job, especially in presenting trade concerns and regional issues. 102. Network Issues. The poverty focus of CASs has continued to improve—although the standards by which we judge CASs have changed even faster. There is room for improvement, especially on the link between the pattern of growth and poverty reduction, and the connection between the Bank- and donor-supported program and poverty reduction. The discussion of areas that are the mainstays of CAS coverage―macro-fiscal, infrastructure and energy, rural development, and human development―remains strong. The CAS diagnosis also focuses on a set of areas that are vital to the poverty reduction effort: gender, governance and corruption, environment, financial sector, and private sector. In these areas, there has been an overall, but very uneven, improvement. Looking at only the CASs for countries where issues in these areas are priorities, the number of “problem” CASs decreases, but does not disappear. Going forward, it will be important to ensure that even these lower percentages improve. CASs will need to give more coverage to the cross-cutting issues where country performance is weak, as indicated, for example, by CPIA scores or other indicators. The Networks also will need to do a better job in 45 highlighting the problem countries and in providing good ESW and lending products for solving problems in different country settings. 103. Strategic Selectivity. While many CASs use the vocabulary of strategic selectivitymentioning comparative advantage, track record, choice of instruments, and so onfew actually use selectivity itself as an analytic tool for focusing on certain sectors or themes and exiting from others. Furthermore, very few spell out why particular lending or nonlending instruments were selected. The dimension of portfolio management has now been mainstreamed in CASs, and CASs continue to strengthen strategy design by discussing implementation lessons from the previous CAS and from OED findings. Virtually all CASs in the cohort discuss the Bank’s track record and partners’ roles, but many do not follow up these considerations in the design of the program. Progress has been made in the area of lending scenarios and triggers, but not in risk analysis. The elements of the selectivity framework need to be fully considered in the formulation of the Bank’s program. A major challenge going forward will be to use the strategic selectivity framework to decide which activities the Bank should undertake using which instruments, and in which areas the Bank should not be involved. 104. Agenda Going Forward. Clearly, improvement in the quality and coverage of CASs has continued, consolidating the improvements recorded in CAS Retro I. But with the changes under way in the Bank’s operating environment, it is now time to begin “unpacking” the CAS, allowing companion and supporting documents to play a larger role, and sharpening the CAS’s focus on the specific poverty reduction goals and the choice of Bank instruments for helping to achieve them. On this view, as progress on these other processes and documents unfolds, the CAS would increasingly become the Bank Group business strategy. To this end, the focus of the CAS agenda going forward should be on the following: • Increasingly relying on vehicles such as the CDF and PRSP to set out the country’s priorities and objectives for poverty reduction and other development goals, with the CAS focusing on the Bank Group’s business strategy in the context of this broader country agenda. Close coordination will be needed between the preparation of the CAS and a companion CDF/PRSP or other country-led document, and the consultation processes that underpin them. • Using cross-country indicators—such as the sectoral indicators the Networks are developing in the context of their SSPs—to help prioritize the principal cross-cutting challenges that the country faces. In the areas that most constrain development in a country, the CAS should include a full diagnosis; in areas where performance is better, the CAS can give a much more cursory treatment, explaining the reasons. • Increasing the reliance on ESW—public expenditure reviews, poverty assessments, social and structural policy reviews, and so on—for the CAS diagnosis and the country-led CDF/PRSP process. The forthcoming ESW Retrospective will discuss the emerging parameters for this work. • Ensuring that IFC and MIGA are involved early on in the CAS process, thus strengthening the coordination of Bank Group activities in supporting private sector development. 46 • Using the strategic selectivity framework to assess which activities the Bank will undertake and which it will not, explicitly looking at the choice of instruments—both lending and nonlending—to be used for these activities. • Increasing the realism of CAS forecasts and risk analysis—both for country (economic growth and poverty reduction) and Bank business parameters. • Strengthening CAS monitoring and evaluation of Bank support for poverty reduction and other objectives as a basis for assessing Bank performance and enhancing accountability—and, over time, for including CAS ratings in the Corporate Scorecard. ANNEXES ANNEXES CONTENTS Page Annex A. CAS List ............................................................................................................ 51 Annex B. Near-CASs......................................................................................................... 52 Annex C. CAS Costs.......................................................................................................... 54 Annex D. Cross-cutting Issues.......................................................................................... 55 Annex E. Sectoral Selectivity ........................................................................................... 71 51 ANNEX A CAS LIST Country Report Report No. Board Date AFRICA Cameroon PR IDA/R98-14 03-13-98 Gabon CAS 18415 10-15-98 Gambia CAS 18361 09-10-98 Kenya CAS 18391 09-24-98 Lesotho CAS 17751 06-04-98 Malawi CAS 18349 08-27-98 Mali CAS 17775 05-26-98 Rwanda CAS 17478 04-07-98 Rwanda PR IDA/R99-135 06-29-99 Senegal CAS 17269 01-29-98 South Africa CAS 18995 03-23-99 EAST ASIA and PACIFIC China PR R98-107 05-28-98 Indonesia PR R99-18 03-09-99 Lao PDR CAS 19098 04-20-99 Malaysia CAS 18812 03-25-99 Mongolia CAS 17604 06-02-98 Philippines CAS 19137 05-04-99 Philippines PR R98-41 03-24-98 Thailand CAS 18002 07-09-98 Vietnam CAS 18375 09-22-98 EUROPE and CENTRAL ASIA Albania CAS 18161 07-30-98 Belarus CAS 18726 01-29-99 Bosnia PR IDA/R98-111 08-06-98 Bulgaria CAS 17655 04-28-98 Croatia CAS 19280 06-03-99 FYR Macedonia CAS 18162 08-04-98 Hungary CAS 17257 01-27-98 Kyrgyz Rep. CAS 17641 05-07-98 Latvia CAS 17706 05-19-98 Lithuania CAS 19135 05-11-99 Moldova CAS 18896 05-06-99 Russia PR R98-288 12-22-98 Tajikistan CAS 18075 07-30-98 Ukraine PR R98-39 [Rev.] 05-21-98 Ukraine PR R99-75 05-20-99 Uzbekistan CAS 17376 03-10-98 SOUTH ASIA Bangladesh CAS 17453 03-31-98 India CAS 17241 IN 01-15-98 India PR IDA/R99-10 02-18-99 Nepal CAS 18578 12-15-98 Pakistan PR R98-37 01-21-99 Sri Lanka PR IDA/R98-190 01-19-99 MIDDLE EAST and NORTH AFRICA Yemen CAS 19073 05-25-99 LATIN AMERICA and THE CARRIBEAN Argentina PR R98-36 03-17-98 Argentina PR IFC/R98-233 11-10-98 Bolivia CAS 17890 06-16-98 Brazil PR IFC/R98-87 06-02-98 Guatemala CAS 18036 07-14-98 Mexico CAS 19289 06-08-99 Mexico PR IFC/R98-30 03-26-98 Nicaragua CAS 17496 04-09-98 Panama CAS 18421 11-19-98 Trinidad and Tobago CAS 19052 04-15-99 52 ANNEX B NEAR-CASS 1. Other Board documents also contain CAS-like Bank strategies. They typically describe recent economic developments and place the Bank’s strategy in that context. • The Special Program of Support for Brazil1 paper was prepared in response to Brazil’s request for Bank financial support in light of the country’s reduced access to international capital markets following the Asian and Russian crises. The paper provides an overview of recent economic developments and the adverse effects of the crises, and describes the Government’s program. It proposes that the Bank’s support be delivered through special sectoral adjustment loans amounting to US$4.5 billion over three years. The paper also discusses the external financing framework for Brazil, and the benefits and risks of the program. • The 1999 interim strategy for Burundi was prepared as the country was emerging from an episode of violent conflict that erupted in 1996.2 Since 1996, the Bank has processed no new operations, and the interim strategy is designed to cover a period of 18 to 24 months. Its goals are to prevent worsening of poverty in the country and to stabilize the economy as the peace process evolves. Based on the Bank’s experience in dealing with conflict and post-conflict situations, the strategy addresses the country’s situation along three fronts: to promote governance and ownership, through participation in rehabilitation and reconciliation, to create productive employment and rural purchasing power, and to restore key imports to support private sector activity and restore minimum levels of essential social services. • The Bolivia Comprehensive Development Framework (CDF) paper3 presents the key implementation features of Bolivia’s pilot CDF. The CDF is a logical extension of the 1998 CAS in that it has allowed the Bolivian Government and the Bank to focus on key issues related to societal transformation and the institutional aspects of the development challenge. The paper discusses recent economic developments, progress in implementing the CAS, and the key features of implementing the CDF in Bolivia―what is already in place, and what more is required under the CDF, for the government as well as for the Bank. The pilot will test the “what” of the CDF―the holistic approach within a long-term strategy―and the “how” of the CDF―country ownership, and partnership in development. After 18 months of implementation, the success of the pilot will be measured by the extent to which specific goals have been reached, and through client and staff surveys. • The Transitional Support Strategy for Kosovo4 covers an 18-month period and includes a proposal to establish a Trust Fund for Kosovo. This trust fund will enable the Bank to respond to the UN’s call for a coordinated effort to support Kosovo’s post- conflict reconstruction and recovery. The transitional strategy for Kosovo includes a ________________________ 1 The Framework Paper: Special Program of Support for Brazil (SecM98-943), November 25, 1998. 2 An Interim Strategy for Burundi: 1999-2001 (Report No. 19592-BU), July 30, 1999. 3 Bolivia: Implementing the Comprehensive Development Framework (Report No. 19326-BO), May 21, 1999. 4 Transitional Support Strategy for Kosovo (SecR99-178), September 16, 1999. 53 ANNEX B detailed discussion of partnerships, programs of other donors, and donor coordination. The scope of the strategy is limited and takes into account the scarcity of special, nonregular financing resources required to support a nonmember. In light of the expected financing role of other donors such as the EU, the Bank will focus on using its comparative advantage in policy advice, institution building, and aid coordination to ensure the sustainability of the reconstruction and recovery effort. • The status report on the Trust Fund for Gaza and the West Bank5 discusses the implementation of the existing portfolio and includes a request for replenishment of the trust fund to finance projects in the pipeline for FY00 and FY01. In terms of nonfinancial activities, the document describes the Bank’s role in administering two major trust funds―the Holst Fund and the Technical Assistance Trust Fund―and its role in aid coordination and policy analysis. Looking forward, the document proposes that the Bank concentrate its efforts in areas where it has comparative advantage and can work in partnership with bilateral and multilateral donors, the private sector, and NGOs. It proposes the selective use of lending in sectors where it can help to build institutions for good governance, decrease poverty, improve access to safe water and a clean environment, develop human resources, and promote the private sector. Moreover, the Bank will continue to provide leadership in aid coordination activities and to provide policy advice and technical assistance in a range of complex areas. ________________________ 5 Trust Fund for Gaza and the West Bank: Status Report and Requirements for Additional Funds (SecM99-499), July 15, 1999. 54 ANNEX C CAS COSTS 1. Since CAS Retro I, the average CAS completion cost decreased slightly, from $132,000 in FY97 to $125,000 in FY98, but increased to $151,000 in FY99.1 CAS costs are still quite a small share of the country budget. As CAS Retro I indicated, average CAS costs represented 2.5 percent of the country budget in FY97-98I. In FY98II-99, the share declined to 2 percent. CAS Retro I reported that the cost of most individual FY97-98I CASs was less than 4 percent of the country budget; however, in the CAS Retro II period, the variation in cost is higher―30 percent of CASs cost more than 4 percent of the country budget. The average cost share for small countries―3.4 percent―is twice as high as that for large countries―1.7 percent. Average CAS Completion Cost, FY 92-99 175 150 125 $ thousands 100 75 50 25 0 FY92 FY93 FY94 FY95 FY96 FY97 FY98 FY99 2. Reasons for Cost Increase. Several factors may have contributed to the increase. First, participation: 75 percent of FY98II-99 CASs have involved a significant consultative process, compared with 52 percent in FY97-98I and 35 percent in FY96. An informal survey by the Bank’s NGO unit shows that the average cost of participation is $60,000, but the variation is high. A second factor contributing to the increased costs may be joint CASs: 16 FY98II-99 CASs are joint CASs, compared with 8 FY97-98I CASs. A third factor is that CAS coverage has become much more complex and the average quality of CASs has increased significantly (see Box 1 of the main report). 3. Data Caveats. As CAS Retro I indicated, pre-FY95 cost data on CASs are especially unreliable, and are probably low, so that the more accurate data for recent years exaggerate the size of the CAS cost increase. However, two factors may have caused an underestimate of the CAS cost increase. First, the data presented here are for the administrative budget only, and recent CASs with extensive participation may have used trust funds in addition to the administrative budget. Second, the country budget data for FY98II-99 also include non-country- specific work, which is not part of the actual budget available to the country directors. ________________________ 1 The source of cost data is Corporate Resources Management (CRM). Figures exclude costs incurred by units not currently integrated in CRM’s CAS cost recording system, such as the Legal Department, Finance Group/Resource Management, and IFC. All cost data refer to total direct completion costs of CASs; that is, they exclude loading factors for Bankwide overheads (leave, training, etc.) and include costs incurred in the year of Board presentation of the CASs as well as in previous years. Recorded total CAS costs include “product” costs (recorded against a task-specific budget code) and “process” costs (recorded with the activity identified as “CAS” for a particular country and year). 55 ANNEX D CROSS-CUTTING ISSUES 1. This annex presents details of the analysis of CAS coverage in five cross-cutting areas, measured using an absolute standard of best practice. Largely carried out by the Networks, the analysis is not tempered by consideration of the degree of priority that should be given to the area in a given country. An analysis taking priority into account was also carried out, and its results are discussed in Chapter III. For comparison, the figures on unsatisfactory CASs from the priority-based assessment are displayed in this annex with the figures from the absolute- standard analysis. A. Gender 2. There are many countries where the health of women and girls is substantially worse than that of men and boys, where girls have significantly less access to education than boys, where maternal mortality is high, and where women are denied equal economic opportunities. Discrimination by gender may be an important form of exclusion. In such settings, focusing on gender concerns can have high developmental impact in terms of poverty reduction and economic growth. An analysis of gender issues thus helps determine whether or not gender should be a CAS priority; serious disparities should be analyzed in the CAS. If gender is not a serious development constraint, the CAS should say so, briefly summarizing the evidence. 3. Coverage. The Network rated more than half of all CASs satisfactory or better in their coverage of gender issues—about the same as CAS Retro I. Gender is identified as a high priority in less than a quarter of all CASs—about the same as in the last retrospective―and as a moderate priority in more than half. Priority ratings, however, do not always imply action: while all CASs that rate gender as a high priority follow up with specific proposed actions, fewer than half of those that rate gender as a moderate priority do so. However, many CASs do address gender issues systematically: 46 percent propose some action to address gender inequalities. (Box D1 sets out the criteria for evaluating CAS coverage of gender and describes best practices.) 4. Prioritizing Gender in the CAS. There is a general overlap between countries where gender is rated a high priority in the CAS and countries where there are serious gender gaps. For example, Gambia, Nepal, Rwanda, and Yemen—all of which are among the ten countries with the highest maternal mortality or child mortality gap1 in the world—all accord high priority to gender issues in the CAS. Attention to gender varies across sectors. For CASs with gender activities spelled out in the action matrix, activities in the traditional health and education sectors account for 39 and 27 percent, respectively, of all gender activities. There is limited explicit reference to women’s potential contributions to development in other sectors—such as agriculture, infrastructure, and finance—and to the synergies between promoting gender equality and other development goals. Few CASs explain why gender action is, or is not, needed. 5. Next Steps. There is room for improvement in the treatment of gender in the CAS. Assistance strategies need to explicitly address whether gender is a country issue or not, and the link between analysis and follow-up need to be strengthened. In countries where gender is ________________________ 1 The child mortality gap measures the female disadvantage in mortality, ages 1-4; the biological norm shows a female advantage in this age group. 56 ANNEX D Box D1. Gender in the CAS Criteria. The retrospective used the following criteria to evaluate the gender focus of each CAS: • Are gender issues covered? • Does the CAS discuss why gender issues are relevant or not relevant to the CAS? • Are actions proposed, or if they are Gender Focus of CASs not, does the CAS discuss why no (in percent) actions are needed? Figure A. Absolute Standard • To what extent do the proposed 80 interventions cut across sectors? 60 The results of the overall evaluation are 53 shown in the graphs. Figure A shows 40 36 33 33 the assessment of CASs by an absolute 29 24 20 20 20 best-practice standard. Figure B is 20 11 14 7 derived by sorting the CASs rated unsatisfactory on the absolute standard 0 All IDA IBRD All IDA IBRD All IDA IBRD All IDA IBRD into two categories—those for countries with CPIA scores of 3 or below, and Unsatisfactory Marginally satisfactory Satisfactory Highly satisfactory others. For the second group, the assumption is that the issue is not a Figure B. Priority-based serious development constraint, and 80 therefore a lighter diagnostic treatment 60 in the CAS is not unsatisfactory; these 55 CASs have been moved to the 40 33 33 “marginally satisfactory” cate-gory in 28 29 Figure B. The “satisfactory” and 20 19 19 20 19 20 20 “highly satisfactory” categories remain 7 unchanged. With this priority-based 0 approach, the share of “unsatisfactory” All IDA IBRD All IDA IBRD All IDA IBRD All IDA IBRD CASs is 19 percent. Unsatisfactory Marginally satisfactory Satisfactory Highly satisfactory Good Practices in CAS Treatment of Gender. The following features of the Senegal CAS represent good practice. • Mainstreaming of gender into the poverty analysis. The CAS links gender issues, such as illiteracy among women, to economic progress and the sustainability of development operations. It includes an inventory of gender action in all Bank assistance, to help position the CAS recommendations in a meaningful context. • Inclusion of women in the consultative process. The CAS is based on extensive consultations with women’s NGOs and civil society groups to help identify gender objectives in a variety of areas, assist in prioritizing them, and build a local constituency for the action program. • Analysis of gender across sectors. The CAS focuses on cross-cutting gender gaps and constraints such as lack of access to land and credit, legal status, and female genital mutilation. It draws on national priorities articulated in the government’s National Action Plan for Women. • Recommendations for specific actions. The CAS anchors gender firmly in the action program and the action matrix. • Use of gender indicators to assess progress. The CAS includes specific indicators to assess gender progress during the CAS implementation period. These indicators cover (a) improved access of women to land, credit, and extension services; (b) improved female access to education; (c) strengthened health and family planning services; and (d) improved legal status for women. • Use of gender action as a trigger for lending. The CAS includes three gender triggers in the lending program: (a) increase in girls’ share of school enrollment; (b) increase in contraceptive prevalence; and (c) decline in maternal mortality. 57 ANNEX D identified as an issue but the Bank’s assistance does not include actions in this area, the CAS should explain why. Moreover, there is a need to broaden the agenda from limited interventions relating to women’s health education to a wider range of sectors that address women’s multiple capacities in a society such as agriculture, water and sanitation, transport and finance. A recent review of the treatment of gender issues in poverty assessments completed for IDA countries during the FY97-99 period found that quality improved during this three-year period. The improvement in the analytical base can be expected to be translated into better treatment of gender in CASs. The Gender Sector Board is currently preparing a strategy paper that will address some of these issues and that should help country teams to address gender in CASs. B. Governance and Corruption 6. The Bank views corruption as a symptom of underlying weaknesses in public sector institutions. In the last few years, the Bank has openly recognized the risks that corruption poses to its operations, and the importance of good governance to the success of operations. Since the Executive Directors approved the Board paper on anticorruption in September 1997,2 CASs have been expected to discuss the effects of corruption and its implications for Bank operations. Guidelines issued in the Anticorruption Action Plan3 in January 1999 formalized the role of CASs, making explicit that they should give attention to anticorruption issues in the design of the assistance strategy. 7. Treatment in CASs. The results of the CAS Retro II evaluation suggest significant progress in mainstreaming the issues of governance and corruption in Bank work. Of the CASs reviewed, 78 percent discuss governance and corruption issues, and about 65 percent make an effort to mainstream governance conditions and corruption risks into country assistance strategies, in keeping with the guidelines issued in January 1999 in the Bank’s Anticorruption Action Plan. This result reflects a range of efforts made since approval of the 1997 strategy report, including meetings with CAS task team leaders and communications with country directors upstream of the CAS process. Given that 70 percent of the FY98II-99 CASs were presented to the Board before January 1999, these efforts appear to have been largely successful. (Box D2 summarizes the findings on governance and describes some good practice examples.) Most of the CASs mention audits in particular sectors or projects, but few have performance benchmarks or triggers for governance conditions or corruption risks. In the Kenya CAS, which does include such triggers, one is the government’s performance on several economic governance reforms, including the strengthening of accountability institutions—with a minimum requirement the adoption of a comprehensive anticorruption strategy. The Azerbaijan CAS also includes a set of triggers related to progress in support of public sector anticorruption reforms. Several other CASs (Bangladesh, Moldova, Nepal, and Mexico) focus on progress in public sector reform and governance. ________________________ 2 See Helping Countries Combat Corruption—The Role of the World Bank (R97-201/1), September 4, 1997. 3 Anticorruption Action Plan for FY99 (SecM98-900), November 9, 1998. 58 ANNEX D Box D2. Governance in the CAS Criteria. The retrospective used the following criteria to evaluate each CAS’s treatment of governance: • Does the CAS discuss the issues of governance and corruption? Does it demonstrate an awareness of the Bank’s governance and corruption guidelines? Does it discuss accountability agencies? • How well does the CAS diagnose the causes of corruption? Does it analyze the institutional environment, assess the governance conditions that create opportunities for corruption, and identify problem areas? • How well does the CAS diagnosis reflect current Bank policy? • How well does the CAS assess the risks of corruption to Bank projects? Does it propose measures to protect the projects from corruption? Scores. Some of the results of this absolute-standard evaluation are shown in the graphs below. Treatment of Governance in CASs Governance Qualitative Criteria (in percent) (in percent) 100 100 78 Yes No Unsatisfactory Marginally satisfactory 80 80 65 Satisfactory Highly satisfactory 60 60 43 46 35 40 40 27 27 30 24 22 22 16 14 19 16 16 20 20 0 0 Mentions governance Reflects current Bank Diagnosis of causes of Diagnosis reflects Bank Assessment of risks for conditions and guidelines corruption guidelines Bank projects corruption risks The review also considered the treatment of governance issues in CASs for countries where these issues are a serious development problem, to see whether this treatment had improved between the first half and the second half of the review period. Some of the results of this evaluation are shown below. Governance Qualitative Criteria: Trends over Time (January 1998-June 1999) (in percent) 100 First half Second half First half Second half First half Second half 80 60 56 56 38 40 31 33 33 33 33 31 25 25 25 25 25 25 17 13 17 17 17 20 6 6 6 6 0 ly ly ly ry ry ry ry ry ry y y y y y y Unsatisfactory Marginally Satisfactory Highly Unsatisfactory Marginally Satisfactory Highly Unsatisfactory Marginally Satisfactory Highly Satisfactory satisfactory Satisfactory satisfactory Satisfactory satisfactory Diagnosis of causes of corruption Diagnosis reflects Bank guidelines Assessment of risks 59 ANNEX D Box D2. Governance in the CAS (continued) Figures A and B below compare the results of an absolute-standard evaluation and a priority-based evaluation. In the absence of an overall rating on governance, the ratings on the two diagnosis criteria were used as a proxy. A CAS was rated “unsatisfactory” or “marginally satisfactory” if it received such a rating under either criterion (Figure A). Figure B is derived by sorting the CASs rated “unsatisfactory” into two groups—those for countries with a CPIA score of 3 or below, and others. For the second group, the assumption is that governance is not as serious a development constraint, and therefore a lighter treatment in the CAS is not unsatisfactory; these CASs have been moved to the “marginally satisfactory” category in Figure B. Thus, under the priority-based approach, the share of “unsatisfactory” CASs is 16 percent. Figure A. Absolute Standard Figure B. Priority-based (in percent) (in percent) 100 100 80 80 62 57 60 51 60 43 37 40 40 25 24 24 25 16 19 20 13 20 0 0 All IDA IBRD All IDA IBRD All IDA IBRD All IDA IBRD Unsatisfactory Marginally satisfactory Unsatisfactory Marginally satisfactory Good Practices in CAS Treatment of Governance. In terms of general diagnosis, of a diagnosis that reflects Bank guidelines, and assessment of risks for Bank projects, good practice CASs include those for Bangladesh, Bulgaria, Latvia, Bolivia, Thailand, Albania, Kenya, and Mexico. 8. Qualitative Evaluation. Beyond evaluating whether a specific topic had been covered, the review also assessed the quality of the treatment of governance and corruption in the CASs (see Box D2). To establish best practices that can be used as benchmarks for future CASs, the review set high standards for CAS treatment of governance and corruption issues. Only 41 percent of the CASs can be considered satisfactory or better in their diagnosis of the causes of corruption, in the sense that they analyze the institutional environment in which corruption occurs and assess the governance conditions that create opportunities for corruption. Fewer than 40 percent of the CASs explicitly assess the risks of corruption to Bank projects and propose remedial measures, if pertinent. About one third of the CASs use information from Country Procurement Assessment Reports (CPARs), Country Financial Accountability Assessments (CFAAs), PERs, or ESW in their diagnosis of governance conditions and risks of corruption, and about the same number integrate anticorruption activities into technical assistance and lending programs. CASs that do a good job of diagnosis and integration include those for Bangladesh, Bulgaria, Latvia, Bolivia, Thailand, Malawi, Albania, Kenya, and Mexico. 9. Countries at Risk. For countries where governance and corruption issues present significant risks—on the basis of a set of external indicators, as well as Bank assessments—this review also looked at whether CASs treat those issues in greater depth than CASs for countries where those risks were seen to be less important. About three out of four CASs prepared during FY98II-99 fall in the risk category. CASs for those countries scored higher in terms of coverage in all four categories (discussion of risks, adherence to Bank guidelines, discussion of the role of auditing agencies, provision of benchmarks and triggers) than the average CAS. In qualitative 60 ANNEX D terms, however, there is little difference between the complete sample and the subset of CASs for countries where poor governance and corruption pose significant risks. 10. Progress over Time. Because CAS Retro I did not undertake a qualitative evaluation of governance in its cohort of CASs, there is no clear benchmark for assessing progress in this area. To evaluate progress over time, the FY98II-FY99 CASs for countries considered at risk were arranged chronologically by Board date, and divided into two groups that roughly represented the first three quarters and the last three quarters of the FY98II-FY99 period. The conclusion is that there has been improvement, particularly in nonqualitative/coverage criteria. The results are more mixed for the quality dimension. Two areas in which there has been noticeable improvement are the quality of the diagnosis, and the use of the Bank’s country corruption and governance assessments. This improvement may be due to investment in country knowledge, although the CASs probably contain only a small portion of accumulated country knowledge. 11. Next Steps. The guidelines for the Anticorruption Action Plan indicate several areas for improved CAS treatment of governance and corruption. The CAS should contain a good-quality diagnosis of a country’s governance and corruption issues. Bank ESW―particularly the results of PERs, CPARs, and CFAAs―may help to strengthen the CAS’s diagnosis of corruption. For countries where poor governance and corruption pose significant developmental risks— determined by a set of indicators collected by the Poverty Reduction and Economic Management Network—the CAS should outline the country’s agenda for improving governance and combating corruption, and include a monitorable program to assess progress. The choice and design of instruments should reflect a good understanding of existing institutions and the best way to reform them. For countries at risk, the CAS should discuss the conditions under which particular instruments should be used. For example, the Bank may consider providing general (untied) external financing only if the country has a credible anticorruption strategy and its financial management and public sector accountability systems meet minimum standards, so the Bank can reasonably expect that its resources will likely be used for development purposes. The content and conditionalities of the lending program should depend on progress in these areas, which should be measured by specific benchmarks. C. Environment 12. The growing body of evidence on the cost and long-term impact of environmental degradation demonstrates the need to address environmental concerns—industrial emissions, pollution of ground and surface water resources, untreated waste, soil erosion, deforestation, and similar issues—in the CAS, and to mainstream these concerns in Bank activities.4 This review assessed the treatment of environmental concerns in the FY98II-99 CASs. (Box D3 presents the criteria used in the evaluation and good practice examples.) 13. Problem Identification. Almost every CAS gives some consideration to the environment, but the quality of treatment varies widely. Most contain a brief assessment of the ________________________ 4 Information on the cost of environmental externalities can be found in, inter alia: Clear Water, Blue Skies— China’s Environment in the New Century, World Bank (1997); Rains Asia—An Assessment Model for Acid Deposition in Asia, World Bank (1997); Emerging Asia—Changes and Challenges, Asia Development Bank (1997); and UN Human Development Report, United Nations (1998). 61 ANNEX D country’s environmental problems and a description of Bank instruments used to address them, but only 55 percent identify and analyze the problems in a satisfactory or better manner. Of these, 72 percent propose specific lending, nonlending, and grant instruments as interventions. Environmental indicators are seldom used. National Environmental Action Plans (NEAPs) and environment-related ESW are the most common source of information for the CASs. In most cases, the diagnosis and strategy analysis are better in the CAS policy matrix than in the text. 14. Environmental Objectives. Most countries have some form of ongoing or proposed Bank support for environmental activities, but problem identification and diagnosis in the CAS does not always reflect the actual amount of environmental work taking place in the country. Some CASs include attention to the environment as a main objective. The Vietnam CAS, for example, has the objective of “deepening the quality and sustainability” of development, and incorporates environmental protection as part of its rural development strategy. 15. Mainstreaming Environmental Analysis. More than half the CASs do a satisfactory or better job of mainstreaming environmental concerns, usually by incorporating them into agricultural and rural sector activities, water sector strategies, or energy reforms. Mainstreaming of environmental concerns into macroeconomic and trade issues is rare. Several CASs mainstream a concern for the environment in innovative ways. The Mongolia CAS identifies and analyzes the link between GDP growth and natural resource depletion. Its treatment of the risks of ignoring resource depletion should be replicated in the CASs for other resource- dependent countries. The Gambia and Senegal CASs treat gender, capacity building, and environment as cross-cutting issues, and include a matrix to indicate how these issues should be addressed in all projects. Another dimension of mainstreaming environmental concerns is taking them into account in designing policies and incentive mechanisms. About 40 percent of the CASs do a satisfactory or better job of linking policy reforms and environmental change, while 35 percent do a satisfactory or better job of linking environment to the narrower policy Box D3. Environment in the CAS Criteria. The retrospective used seven criteria to evaluate the environmental aspects of each CAS: • Does the CAS identify key environmental problems? Does its description of the country’s overall development context include environmental as well as economic and political issues? • Does the CAS propose specific interventions for each environmental problem identified? • Does it mainstream environmental concerns into other activities? • Does it link the environment and poverty in the country? Does it analyze the feedback mechanisms among rural and urban poverty, environmental degradation, and economic slowdown? • Does it link specific policies and environmental change? Does it take into account the costs of resource depletion when formulating macroeconomic and sector policies? Environment in CASs (in percent) • Does it analyze how incentives affect natural resource management? Does it evaluate how various policies, 80 Identify issues Treat issues Mainstream issues including property rights and pricing issues, affect 60 49 natural resource management? 41 41 40 32 • Does it include indicators of progress? 27 24 22 19 16 20 11 14 Scores. Some of the results of the absolute-standard 5 evaluation are shown in the graph at right. 0 Unsatisfactory Marginally Satisfactory Highly satisfactory satisfactory 62 ANNEX D Box D3. Environment in the CAS (continued) Figures A and B below compare the results of an absolute-standard evaluation and a priority-based evaluation. In the absence of an overall rating on environment, the ratings on the identification and treatment criteria were used as a proxy. A CAS was rated “unsatisfactory” or “marginally satisfactory” if it received such a rating under either criterion (Figure A). Figure B is derived by sorting the CASs rated “unsatisfactory” into two groups—those for countries with a CPIA score of 3 or below on environment, and others. For the second group, the assumption is that the environment is not as serious a development constraint, and therefore a lighter treatment in the CAS is not unsatisfactory; these CASs have been moved to the “marginally satisfactory” category in Figure B. Thus, under the priority-based approach, the share of “unsatisfactory” CASs is 5 percent. Figure A. Absolute Standard Figure B. Priority-based (in percent) (in percent) 100 100 80 80 60 50 60 38 40 29 40 27 31 25 24 16 20 5 5 6 20 10 0 0 All IDA IBRD All IDA IBRD All IDA IBRD All IDA IBRD Unsatisfactory Marginally satisfactory Unsatisfactory Marginally satisfactory Good Practices in CAS Treatment of Environment • Identification of key environmental problems. The Rwanda CAS uses a text box in the introductory pages to note that (a) large-scale movements of population and livestock have increased deforestation of and encroachment on protected park lands; (b) population encroachment is threatening a fragile ecological zone; and (c) the massive destruction of livestock in some parts of the country during the genocide has starved the soil of nutrients previously provided by animal manure. • Proposal of specific interventions. The Uzbekistan CAS identifies the dessicating Aral Sea, and the rising salinization and desertification of river deltas in the densely populated Aral Basin, as striking symptoms of an unsustainable model for water and land management. It includes support for the multidonor-sponsored Aral Sea Basin Program, comprising 8 interlinked program elements and 18 projects. • Mainstreaming of environmental concerns into other activities. The Mongolia CAS notes that traditional national accounting methods do not adequately measure the cost of depletion of mineral resources and pastureland. It calculates the cost of that depletion as the value that extracted copper and gold would have in the ground, plus the environmental damage to forest and pastureland as a result of agriculture and livestock production. It argues that while the rapid use of natural resources may be reasonable in the short run to ease the transition, continuing to dissipate wealth in this manner will depress future incomes. The CAS mainstreams natural resource depletion into macroeconomic policies by establishing a technical assistance program on national accounts, and by identifying mechanisms to develop a long-term strategy for fiscal management of copper revenues. • Linkage between the environment and poverty. The Lesotho CAS notes that rural poverty is both a cause and a consequence of the country’s serious environmental problems. It links rural poverty to soil erosion, diminished soil fertility and crop yields, deforestation, and overgrazing of rangeland. It proposes support, along with other donors, for development of a comprehensive Agriculture Sector Investment Program to address these issues. The CAS also links urban environmental degradation to health problems, and proposes support for government programs in potable water, sewerage, and solid waste disposal to upgrade the basic infrastructure of the poor. • Linkage between specific policies and incentives and environmental change. The Panama CAS summarizes the effects of different instruments to bring about environmental change: (a) a project to increase rural incomes and stem environmental degradation; (b) a complementary GEF grant to conserve the Atlantic biological corridor; (c) a project to stem land degradation and support farmers; (d) technical assistance to combine environmental and privatization goals; (e) a road project to strengthen environmental impact assessment capacity; and (f) a project with TA components on institutional capacity building, sustainable forestry, and water management. 63 ANNEX D concerns of pricing and property rights. Policy issues in CASs generally reflect the concerns shared by countries in a given region. In some Latin America CASs, the major concern is to strengthen environmental policies and regulations; while in the CASs for Hungary, Lithuania, and Bulgaria, accession to EU is a major driver of environmental regulations. In South and East Asia CASs, particularly, there is an effort to link better environment to sector and macroeconomic strategies—water reforms in the case of Vietnam and India, budget reforms in the case of Mongolia, energy policies in the case of the Philippines. 16. Poverty and the Environment. The Environment Sector Board finds that the link between poverty and environment is not adequately addressed in the FY98II-99 CASs. Less than half attempt to link poverty and environmental concerns, and many of these do not propose specific interventions—although there may have been some interventions that both helped the poor and stemmed environmental degradation that were not adequately linked in the CASs. The Nicaragua CAS, for example, clearly spells out the connection between poverty alleviation and the rational exploitation of natural resources. In fact, more CASs link poverty to natural resource degradation than to the health impacts of water and air pollution. Overall, environment-poverty links need to be better understood and better operationalized, and future CASs should offer concrete operational proposals. 17. Assessment Over Time. CAS Retro I did not directly review environmental content of CASs, but relied on the results of a 1997 Bankwide review,5 which used different criteria and ranking scales. Thus a direct comparison of the results of CAS Retros I and II is not possible.6 Nevertheless, the picture that emerges from comparing criteria that could be matched (environmental problems and opportunities assessment, poverty and economic policies) is that no significant change has occurred in the quality of treatment of the environment in the CAS. 18. Next Steps. As the discussion indicates, there is scope for learning from examples of best practice. Environmental concerns should be mainstreamed by linking them to economic and sector issues. This can be done by taking into account the possible environmental impact of key policy reforms; ensuring that countries dependent on natural resources for exchange earnings develop sustainability criteria to manage resource use; identifying prices, property rights, and other incentive mechanisms that influence environmental considerations; and using these mechanisms to correct environmental externalities and improve efficiency of resource allocation. The link between poverty and the environment needs to be strengthened by focusing on policies that reduce inequality and environmental degradation; by supporting interventions that improve environmental conditions affecting the health of the poor; and by ensuring that natural resource management contributes to the livelihood of the rural poor. Environmental indicators and analyses of environmental trends need to be better incorporated in the CAS. For example, the environmental indicators collected for the World Development Indicators could be included in the CAS annexes, and environment-related ESW should be integrated in the CAS. ________________________ 5 Ekbom, A., and J. Bojo, Mainstreaming Environment in Country Assistance Strategies, Discussion Paper No.1, Environment Group, Africa Region, World Bank (1997). This is a review of 34 CASs prepared during the FY 1996-98 period. 6 The Ekbom-Bojo (1997) review used a three-level scale for ranking CASs, while this review uses four levels. The rankings used by Ekbom and Bojo are 1=not discussed at all, 2=mentioned but not elaborated, and 3=well elaborated. 64 ANNEX D D. Financial Sector 19. The financial sector—whether it functions well or not—plays a critical role in economic development. When it functions well, it mobilizes scarce capital resources and intermediates them to their most productive uses. How well countries allocate their resources and manage risks is a key determinant of their growth performance and, consequently, of their success in poverty reduction. An ill-functioning financial sector causes resources to be wasted and can eventually throw countries into financial crises and recession, raising unemployment and setting back hard-won achievements in poverty reduction. The East Asia crisis offers a strong example of the forces unleashed by a breakdown of the financial sector. 20. Role of the Financial Sector. Given its important role, the financial sector has been recognized as a key priority of the Bank work in recent years7 and should be fully addressed in the CAS. The CAS should identify the key financial sector issues in the country, develop a strategy to address deficiencies, and prescribe the role of the Bank Group. Where the financial sector is not considered to be a high priority, the CAS should explicitly state this fact and explain why. When the financial sector is weak but the country and the Bank disagree on the strategy, scope, or speed of reforms, the CAS should outline how the Bank Group will try to advance the dialogue or influence the debate in the country. 21. Findings. Overall, CAS coverage of the financial sector has improved, but it varies widely. Some CASs contain very comprehensive and insightful discussions of the sector, but 16 percent of CASs still have no discussion of the sector. While the matrices of most CASs give a clear diagnosis and strategy, many CASs do not reflect these insights in the main text. CASs for countries that have suffered financial crisis―and more recent CASs in general―have a better coverage of the financial sector. For example, the 1998 Thailand CAS contains a good discussion of the crisis and its impact on the real sector and outlines the Bank program for the sector in the short and medium term, laying out the roles of the Bank, the IFC, and other partners. This is a marked improvement relative to the previous CAS, which contained little reference to the sector. 22. Coverage and Strategy. The majority of the CASs discuss major issues in the financial sector, but only 43 percent cover the progress made in the sector since the previous CAS. Few CASs base their discussions on a financial sector assessment or financial sector ESW. The length of the section on the financial sector varies from 13 paragraphs to none, with an average of 3 paragraphs. Only 57 percent of the CASs present a strategy for the financial sector, and only 40 percent integrate the strategy with issues in the enterprise/corporate/real sectors. Many of the CASs do not go beyond providing a list of the activities that the Bank Group will undertake. About 40 percent contain a joint strategy with IFC, and 76 percent mention the roles of other partners. ________________________ 7 Markers in this process include the issuance of World Development Report 1989: Financial Systems and Development, the establishment of the Financial Sector Development Department in 1992, the increasing resources devoted to the sector in the Strategic Compact and New Spending Authority, the establishment of the Special Financial Operations Unit in 1997, and the 1999 creation of the Financial Sector Operations Vice Presidency. 65 ANNEX D 23. Joint Bank/IFC CASs. Joint Bank/IFC CASs8 have on average better treatment of financial sector than non-joint CASs. The eight joint CASs that have Private Sector Strategy (PSS) annexes contain the most comprehensive discussions of the financial sector. The data show clearly that the improved CAS treatment of the financial sector can be attributed to the Bank/IFC collaboration.9 The CASs for Mali and the Philippines, in particular, stand out. Some non-joint CASs prepared for crisis countries—Malaysia and Thailand—also have very extensive treatment of the financial sector. All four of these CASs do a good job of laying out the issues, particularly their consequences for the real sector, and present the Bank’s program in the short and medium term. Moreover, these CASs discuss the roles of IBRD, IFC and MIGA, as well as for other donors. (Box D4 sets out the criteria for evaluating CASs and presents good practice examples of CAS treatment of the financial sector.) 24. Areas for Improvement. A major deficiency in the treatment of the financial sector in many CASs is the absence of a strategy to address financial sector issues. In many of those that do include a sector strategy, the link between the strategy and the real sector could be strengthened. It is expected that, as the number of joint CASs grows, the treatment of the sector will improve over time. In general, a satisfactory financial sector section should cover several areas. It should provide a diagnosis of the financial sector and its role in supporting private sector development, and it should lay out the progress made since the last CAS. It should identify major financial sector issues, including an analysis of the system’s stability and efficiency, and the ease of access to credit by different parts of the economy. The diagnosis should be based on a financial sector assessment or financial sector ESW, and should provide references. In terms of the sector strategy, the CAS should describe the policy dialogue with the government and set out how the issues identified will be addressed and by whom. When the financial sector it is not a priority, the CAS should explain why. The CAS should clearly describe the division of labor within the Bank Group among IBRD, IFC, and MIGA. E. Private Sector 25. Sustained economic growth is essential for poverty reduction, and the private sector is the engine of growth in the world economy. The Bank Group can offer a broad range of instruments through the Bank, IFC, and MIGA to promote private sector development. These instruments can play an important role in helping countries create an enabling environment for private sector development by supporting the establishment and growth of private enterprises, stimulating efficient domestic capital markets that provide investment financing and working capital, and catalyzing external capital flows. ________________________ 8 The joint Bank/IFC CAS pilot program is described below under “Private Sector.” 9 One hypothesis was that the joint CASs contain better treatment of the financial sector because they are prepared for emerging market economies, where private sector and financial issues are more prominent. However, the findings did not bear out this hypothesis: the overall coverage of the financial sector is only slightly better in IBRD CASs than in IDA CASs, and in one area—progress in the sector—the coverage of IDA CASs is better than that in IBRD CASs. In addition, IDA countries account for about half of joint CASs as well as half of total CASs reviewed. Thus it is clear that the “jointness” is the factor that accounts for the quality of the financial sector discussion in joint CASs. 66 ANNEX D Box D4. Financial Sector in the CAS Criteria. The criteria for evaluating CAS coverage of the financial sector look at two key aspects: • Coverage of the financial sector. Does the CAS discuss progress made in the sector? Does it analyze the issues remaining in the sector? Does it anchor the analysis in a financial sector assessment or financial sector ESW? • Coverage of the financial sector strategy. Does the CAS present a strategy for the sector? Is the strategy integrated with issues in the enterprise/corporate/real sector? Is there a joint Bank/IFC private sector strategy? Does the CAS define the roles of donors and the coordination among them? Scores. The results of the evaluation of CAS coverage are shown in the graph. Financial Sector in CASs (in percent) 100 100 PSS annex Joint CASs All CASs 100 100 94 86 88 88 88 81 75 75 76 80 69 57 56 60 43 41 41 40 20 0 Mentions progress made Identifies issues Presents a strategy Strategy integrated with A joint strategy with IFC Mentions roles of donors real sector Coverage of sector Coverage of strategy Figures A and B below compare the results of an absolute-standard evaluation and a priority-based evaluation. The absolute-standard rating (Figure A) was derived for each CAS on the basis of the criteria above. Those CASs rated unsatisfactory are divided into two groups―countries with a “weak” financial sector (CPIA score of 3 or below) and countries with a “strong” financial sector (CPIA score above 3). In Figure B, the CASs rated “unsatisfactory” under the absolute standard that are for countries with a low CPIA keep their “unsatisfactory” rating, while those for countries with a CPIA above 3 are moved to the “marginally satisfactory” category. With this priority-based approach, the share of “unsatisfactory” CASs is 22 percent. Figure A. Absolute Standard Figure B. Priority-based (in percent) (in percent) 100 100 80 70 80 60 60 48 43 48 38 38 33 34 40 40 27 22 19 20 20 6 0 0 All IDA IBRD All IDA IBRD All IDA IBRD All IDA IBRD Unsatisfactory Marginally satisfactory Unsatisfactory Marginally satisfactory Good Practices in CAS Treatment of the Financial Sector • Coverage of the sector. The CAS for Malaysia provides an overview of the sector, describes recent measures taken in financial restructuring, and outlines progress made so far. It also lays out key issues facing the sector. The Philippines PSS describes recent progress in commercial banking, nonbank financial intermediaries, equities market, and the market for securitized term debt. It also makes reference to ESW. • Coverage of the strategy. The Thailand CAS does a good job of describing the financial crisis and its consequences for the real sector, and notes the Bank’s program in the short and medium terms. The Malaysia CAS discusses the strategy for the financial sector and corporate sector integrally. Both CASs describe a joint strategy with IFC and mention the roles of different donors. 67 ANNEX D 26. Coverage. A good discussion of private sector issues in CASs provides a clear profile of the private sector—its depth, structure, and major constraints. Very few FY98II-99 CASs provide a profile of the sector; most focus on identifying constraints. The most frequently identified constraints are a weak financial sector, inadequate infrastructure, and lack of proper legal and regulatory frameworks. Property rights and reform of state-owned enterprises are priority issues for transition economies. Some CASs identify priority sectors for private sector development such as agriculture, construction, energy, mining, and tourism. Overall, the coverage of the private sector is weak. Only 60 percent of CASs have a separate section or separate discussions on private sector; the rest only discuss specific private sector issues such as privatization and competition. Only about a fourth of CASs contain a detailed discussion of private sector issues. (Box D5 sets out the criteria used in the evaluation and provides good practice examples.) 27. Bank and IFC Strategy. Besides presenting the profile and discussing constraints to the sector, the CAS should also describe the Bank Group’s strategy for addressing the constraints— clarifying the respective roles of the Bank and IFC and the degree of complementarity of their activities. Most CASs contain such discussion; but in more than half of the CASs the Bank’s and IFC’s assistance programs are discussed separately without providing a connection between the two. Only 46 percent of the CASs discuss IFC’s activities together with the Bank’s program at the sectoral or project level, and most do not go beyond a simple statement on the complementary role of the Bank and IFC or MIGA. 28. Joint CASs. The joint CAS program was introduced in FY97 to enable the Bank Group to strengthen the coordination and cooperation among the Bank and IFC/MIGA.10 Sixteen of the 37 FY98II-99 assistance strategies are joint CASs, of which eight include a Private Sector Strategy annex (PSS). Joint CASs, and particularly those with PSSs, stand out as providing a much more substantive coverage of the private sector and a joint discussion of the Bank and IFC programs.11,12 Most PSSs describe progress made, identify priority issues, and discuss the Bank Group’s role. Most PSSs also identify the financial sector, infrastructure, and legal and regulatory frameworks as major bottlenecks for private development. However, one fourth of the PSSs fail to provide a Bank Group strategy with respect to private sector development. ________________________ 10 Under the program, countries were classified into three groups according to the relative priority of private sector development for the country’s overall development goals and the scope for Bank/IFC collaboration. For the first group of countries, identified by the Bank and IFC as high strategic priorities (category A), a joint CAS with a private sector strategy (PSS) annex was prepared. For the second group, for which private sector development was identified as a priority and there was substantial scope for Bank/IFC collaboration (category B), the CAS was prepared jointly and the private sector strategy reflected in the CAS. For the last group, where private sector development was not a CAS priority or the scope for collaboration was modest (category C), IFC was less involved in drafting the CAS. The selection of CASs to be jointly prepared is reviewed annually by the Bank and IFC, and includes all Comprehensive Development Framework (CDF) countries where IFC has a significant role. 11 It could also be true that joint CASs have been prepared in countries where coverage and cooperation are already in place. In any case, there is a definite indication that joint CASs do better than non-joint CASs. 12 In its strategic focus paper, MIGA states its intention to become more involved in the CAS process. Initially MIGA will focus on a limited number of countries, but as it gains more resources it will increase the range of its involvement. MIGA will work with the rest of the Bank Group to ensure that its collaboration on projects allows for more effective use of its resources. See MIGA’s Guiding Principles: Strategic Focus Paper (MIGA/SecM99-7), April 5, 1999. 68 ANNEX D Box D5. Private Sector in the CAS Criteria. The retrospective evaluated three dimensions of CAS coverage of the private sector: CAS Coverage of the Private Sector • Does the CAS provide a clear description of the private (in percent) sector in the country—its depth, structure, and progress to All CASs Joint CASs PSSs 100 date? 88 80 75 75 • Does it identify major constraints to the development of 62 the sector? 60 46 40 • Does it explain the role of the Bank Group in addressing 40 the constraints? 20 0 Substantive coverage of the Integrated discussion of Bank Scores. Some of the results of the evaluation are shown in the sector and IFC activities graph at right. Figures A and B below compare the results of an absolute-standard evaluation and a priority-based evaluation. The absolute-standard rating (Figure A) was derived for each CAS on the basis of the criteria above. Those CASs rated unsatisfactory are divided into two groups – countries with a “weak” private sector (CPIA score of 3 or below) and countries with a “strong” private sector (CPIA score above 3). In Figure B the CASs rated “unsatisfactory” under the absolute standard that are for countries with a low CPIA keep their “unsatisfactory” rating, while those for countries with a CPIA above 3 are moved to the “marginally satisfactory” category. With this priority-based approach, the share of “unsatisfactory” CASs is 5 percent. Figure A. Absolute Standard Figure B. Priority-based (in percent) (in percent) 100 100 88 80 80 63 60 60 49 40 32 40 22 25 22 19 20 13 20 10 5 0 0 0 All IDA IBRD All IDA IBRD All IDA IBRD All IDA IBRD Unsatisfactory Marginally satisfactory Unsatisfactory Marginally satisfactory Good Practices in CAS Treatment of the Private Sector • Sector profile. The Mali Private Sector Strategy annex (PSS) gives a clear picture of the role of the private sector in the economy. It shows the percentage of GDP for which the private sector is responsible and how much employment the private sector provides. It also indicates in which areas the private sector mainly operates and in which geographic areas it is concentrated; it also discusses the growth rates of the private sector and some major development events in recent years. • Discussion of constraints and priorities. The PSSs for Bolivia, Mexico and Nicaragua provide comprehensive discussions of the major constraints and priority issues in the private sector: a weak financial sector, inadequate infrastructure, and lack of a proper legal and regulatory framework. The Trinidad and Tobago CAS also identifies labor market rigidities as a major constraint. • Joint strategy. The Macedonia CAS provides concrete discussions of how the Bank and IFC coordinate and collaborate at both the strategic and the project level. It also mentions collaboration with other donors. 69 ANNEX D 29. Joint CAS Assessment. The joint CAS program has increased the degree of consultation between Bank and IFC staff during the preparation of private sector development strategies. While the level of cooperation remains uneven, it is evolving toward a closer relationship between Bank country teams and their IFC counterparts in designing the strategy. Where IFC has been fully engaged in the CAS process, results have been more satisfactory—in terms of both a shared diagnosis and the creation of an agenda for action for private sector development. Joint CAS preparation has yet to have a full impact on country work programs, which have continued to be prepared separately. The Bolivia and Vietnam CASs are examples of best practice in joint preparation, stemming from close working relationships in the field between the Bank and IFC. Thus, the best way to improve the content of joint CASs is to improve the quality of Bank/IFC interaction. Joint CASs have been a useful vehicle for dialogue on IFC’s role and strategy in a country, beyond individual operations. They have been particularly valuable for countries nearing graduation from Bank assistance, and in post-crisis situations. The scale of the joint CAS program over the last 18 months has struck a reasonable balance between the benefits of closer collaboration and the costs of staff time to engage in CAS preparation. 30. Preparation of a Joint Strategy. A key area for improvement is to fully coordinate the Bank program and IFC/MIGA private sector activities. In FY98II-99, only a quarter of PSSs are reflected and cited in the main text of CASs. To accelerate the pace of private sector development, the Bank Group needs to be more effective at leveraging Bank and IFC programs. The CAS provides the basis for such deeper collaboration. The Bank Group’s private sector development strategy13 aims to create the synergies necessary to ensure that the mix of products and services offered to member countries meets their needs most efficiently • From strategy to work programs. The Bank Group’s private sector strategy proposes moving beyond agreement between the World Bank and the IFC on priorities in the CASs, to a joint agreement on work programs that reflect these priorities—and, in selected cases, joint review of individual operations. Initially, this deeper integration of country strategies and work programs should be attempted in a few countries. Such collaboration will help ensure that the Bank Group’s country work programs include adequate economic and sector work to diagnose private sector development issues, as well as the necessary advisory and lending support to help implement the proposed reforms. Joint CASs will also indicate how IFC’s country work programs will support the broader development goals. In some sectors (telecommunications, oil and gas, mining, advisory services, small and medium enterprises), combined departments will ensure collaboration in implementing CASs at the policy and transaction level; in other sectors (power, water, transport), strengthened Bank Group sector boards will carry out this role. Both approaches will strengthen the impact of joint CASs on Bank Group work programs in the area of PSD. • Format. The need to prepare separate Bank and IFC private sector strategies will continue to depend on whether separate strategies will improve the quality of the overall CAS. However, given the space pressures in the main CAS document, the continued use of a PSS annex will likely be useful for a country’s first joint CAS, or ________________________ 13 World Bank Group Private Sector Development Strategy (R99-175, IFC/R99-154), September 10, 1999. 70 ANNEX D when a PSD agenda is particularly important. In particular, because it is difficult to adequately treat the range of PSD issues within the confines of a short CAS document that covers many sectors and issues, the preparation of PSSs should remain a useful focus for Bank-IFC dialogue—and such strategies provide the Board with a fuller picture of IFC’s role. • Early engagement. The joint CAS process has worked best when IFC has been engaged at a very early stage in analyzing growth and private sector development issues, delineating the most important elements of the strategy, and consulting with government and other stakeholders. This highlights the need to bring IFC in at an early stage of CAS preparation. IFC is currently fully engaged in the pilot CDF process, which will help to strengthen this upstream collaboration. An earlier engagement from MIGA in the CAS process will further contribute to enhancing the synergies of the Bank Group program. 71 ANNEX E SECTORAL SELECTIVITY 1. CAS Retro I undertook a detailed analysis of sectoral selectivity that provided the baseline for a comparison over time. Whether or not CASs became more selective depends on how selectivity is defined. Like CAS Retro I, this report looks at three dimensions of selectivity: (a) sectoral focus―how many areas does an average CAS specify as a priority for Bank intervention? (b) sectoral coverage―do actual program choices match the stated priorities? and (c) sectoral concentration―what share of all lending and ESW activities is accounted for by the four areas with the most activities? 2. Sectoral Focus. In terms of stated priorities, there is a strong similarity across the FY98II-99 CASs. At least 70 percent of the CASs identify the same seven areas of priority for Bank involvement: rural development, macro-fiscal, poverty, public sector management, education, private sector development, and health. More than 90 percent of the CASs cite rural development and macro-fiscal management as a priority (see Table E1, Col. 2). The FY97-98I CASs identified the same seven areas as priorities, plus environment and infrastructure. 3. Sectoral Coverage. Although the number of priorities stated in the CASs has been reduced, has the sectoral coverage of Bank activities been adjusted accordingly? Do actual program choices in the FY98II-99 CASs match stated sectoral priorities? A comparison between the top seven stated priorities and the top seven activity-intensive areas shows some mismatches (see Table E1, Cols. 2 and 3). Infrastructure and environment, which are not among the top seven stated priorities in the CASs, are among the most activity-intensive areas. For example, while environment is given as a priority area in only 51 percent of the CASs, 78 percent of the CASs include a loan or an ESW task on environment. This finding is a change from FY97-98I CASs, in which actual program choices broadly matched stated priorities. In sum, the reduced number of priorities has not been matched by the actual program choices; FY98II-99 CASs continue to have relatively broad sectoral coverage, and, in this sense, limited selectivity. 4. Sectoral Concentration. In terms of sectoral concentration, four areas―macro-fiscal, infrastructure, rural development, and public sector management (PSM)―account, on average, for more than 50 percent of the lending and nonlending activities in the CAS. In an average CAS, the two areas with the largest share of ESW activities are macro-fiscal and PSM; the two areas with the largest share of lending are infrastructure and rural development. The three areas with the smallest average share of Bank activities per CAS are gender, social protection, and energy―essentially the same areas as in FY97-98I CASs. In terms of sectoral concentration, the CAS programs are rather selective. 5. Selectivity Results by Country Group. Neither the average number of priority areas identified in the FY98II-99 CASs nor the sectoral coverage of CAS programs (the number of areas with lending or ESW activities) changes significantly when CASs are grouped by country population size, per capita income, IBRD or IDA membership, or performance. However, on average, large country programs include close to three times as many lending operations, twice as many ESW activities, and in total twice as many Bank activities per covered area as small countries. CAS programs for well-performing countries contain about 20 percent more lending operations and 38 percent more ESW activities per covered area than CAS programs for poorly performing countries. In sum, then, CASs for large countries and, to a lesser extent, CAS 72 ANNEX E programs for well-performing countries seem to take advantage of intrasectoral economies of scale. At the same time, CAS programs are slightly more concentrated, and in this sense are more selective, for poorly performing countries than those for well-performing countries. These differences are sharper for ESW than for lending activities. The pattern of greater concentration of areas of activity in CAS programs for poorly performing countries, relative to programs for well-performing countries, is in line with expectations and in the right direction. (Table E2 illustrates these results.) Overall, these results are similar to those obtained in CAS Retro I. 6. Caveats. Activity counting is a rough instrument for analyzing selectivity. Several activities are necessarily multisectoral, and it is rather difficult to unbundle them across sectors. Poverty, for example, accounts for a small average share of lending and nonlending activities per CAS, but this does not mean that poverty is treated inadequately. Many nonlending services, such as public expenditure reviews, necessarily cover many areas that have a direct bearing on poverty; while poverty-related lending activities fall, in many cases, under such other categories as health or education. The mapping of activities into Network Families is not always clear-cut, and some activities can reasonably be classified in more than one Network Family. Thus, it cannot be inferred solely from the number of Bank activities in a sector whether the Bank plays a large or a small role in that sector, and much less what the likely impact of its program will be. 73 ANNEX E Table E1. Sectoral Focus and Program Composition in FY98-99 CASs (in percent) Share of Average share of activities Share of CASs with Areas of Bank CASs that in area per CAS involvement by Network identify Loans or Loans and Family area as a Loans only ESW only Loans or ESW in ESW in Loans ESW priority in area in area ESW area areas Sector areas Macro-fiscal 92 100 43 8 49 18 9 26 Infrastructure 57 92 41 3 49 14 23 6 Education 81 89 22 14 54 7 9 6 Rural 97 84 8 16 59 11 13 9 FSD 68 81 24 27 30 7 5 8 PSM 84 76 19 14 43 10 10 11 Health 70 65 19 32 14 6 8 5 PSD 76 65 19 32 14 5 4 5 Energy 35 59 19 22 19 4 4 3 Social protection 59 54 11 30 14 3 3 4 Cross-cutting areasa Environment 51 78 8 30 41 6 6 6 Povertyb 86 81 0 57 24 6 3 9 Gender 16 11 11 0 0 0 0 1 Social development 38 46 16 16 14 3 3 2 a Stated CAS objectives in these cross-cutting areas are met largely through lending or ESW tasks in other sector areas. Thus, the percent shares shown in the table for lending or ESW specifically in these areas capture only a small part of all Bank activities that help meet CAS objectives in these areas. b Many operations in the Program of Targeted Interventions broadly target poverty, so they are not included among the specifically poverty-oriented interventions captured in this table. 74 ANNEX E Table E2. Sectoral Selectivity in CASs by Country Groups Selectivity indicator All Countries Countries Countries Countries IBRD IDA countries with large with small with good with poor populationa populationa performancea performancea Number of stated priority areasb 7 7 6 6 7 6 7 in more than 70 percent of CASs Number of areasb covered in Bank program Areas with lending or ESW 10 12 10 10 10 10 10 Areas with lending 6 7 6 7 6 6 7 Areas with ESW 8 11 8 8 8 9 8 Number of Bank activities Lending and ESW activities 30 58 24 31 25 35 28 Lending activities 13 28 10 14 12 12 14 ESW activities 17 30 14 18 13 23 14 Average number of Bank activities per covered areab Lending and ESW activities per area 3.0 5.0 2.5 3.1 2.5 3.4 2.8 with lending or ESW Lending activities per area with 2.1 4.1 1.6 2.1 1.9 2.1 2.0 lending ESW activities per area with ESW 2.1 2.8 1.8 2.2 1.7 2.6 1.9 Four-sector concentration ratioc(percent) Lending and ESW 53 55 53 51 59 54 53 Lending 54 59 53 55 59 54 57 ESW 55 53 55 53 61 57 55 a Large countries are those with populations of at least 30 million, small countries those with populations of less than 30 million. Countries with aggregate 1997 country performance ratings of at least 3.0 are classified as good performers, those with ratings below 3.0 as poor performers. b Bank activities are grouped into fourteen Network Family areas: macro-fiscal management, poverty alleviation, gender, public sector management, health, education, social safety support, private sector development, banking and financial sector, energy, infrastructure, environment, social development (includes social assessments, civil society, participation, cultural heritage, post-conflict resolution), and rural development. c The four-sector concentration ratio is the total percentage of Bank activities (lending, ESW, or lending and ESW) accounted for by the four Network Family areas with the most activities.