OED IDA REVIEW REVIEW OF AID COORDINATION IN AN ERA OF POVERTY REDUCTION STRATEGIES IDA10-12 May 1, 2001 Operations Evaluation Department Document of the World Bank Director-General, Operations Evaluation : Mr. Robert Picciotto Director, Operations Evaluation Department : Mr. Gregory Ingram Operations Adviser, OEDDR : Mr. Nils Fostvedt Task Manager, IDA Review : Ms. Catherine Gwin REVIEW OF AID COORDINATION IN AN ERA OF POVERTY REDUCTION STRATEGIES IDA10-12 May 1, 2001 This paper is based on work by John Eriksson. Helpful inputs and comments were received from Mario Aguilar, Sadiq Ahmed, Ivar Andersen, Anupam Basu, Pamela Bigart, Shanta Devarajan, James Funna, Alan Gelb, Catherine Gwin, Engibert Gudmunsson, Gregory Hancock, Nagy Hanna, Nancy Happe, Keith Jay, Daniel Morrow, Sanjivi Rajasingham, Alison Rosenberg, Colin Scott, Gregory Toulmin, Hasan Tuluy and John Underwood. Patricia Laverly, Pierre-Joseph Kingbo, and Patrick Grasso prepared a thorough analysis of OED Country Assistance Evaluations, described in Annex 4. Valuable administrative assistance was provided by Julia Ooro. This is one of the series of background papers prepared for the OED IDA Review. In the IDA12 Replenishment Report, IDA Deputies requested OED to undertake an independent review of the IDA program during the IDA10­11 period and an interim review of IDA12. The Review concentrates on IDA's development contribution in six thematic development priorities: (i) poverty reduction; (ii) social development; (iii) private sector development; (iv) governance; (v) environmentally sustainable development; and (vi) gender. It also addresses four priority process reform objectives: (i) performance based allocations; (ii) enhanced CAS design and implementation; (iii) improved aid coordination; and (iv) participation. The findings, interpretations, and conclusions expressed in this paper are entirely those of the author. They do not necessarily represent the views of the Operations Evaluation Department or any other unit of the World Bank, its Executive Directors, the IDA Deputies or the countries they represent. i Contents Contents ................................................................................................................................i Abbreviations and Acronyms ..............................................................................................iii Executive Summary.............................................................................................................iv Recommendations..............................................................................................................vi 1. The Evolution of IDA's Aid Coordination Commitments ...............................................1 2. The Performance of IDA and Its Partners......................................................................3 Strengthen Coordination with Other Development Assistance Agencies.................................3 Increase Programmatic Selectivity.......................................................................................9 Move to Country-Led Partnerships....................................................................................10 Special Issues...................................................................................................................16 3. Lessons .........................................................................................................................19 Capacity and Commitment................................................................................................19 Mutual Responsibility and Distinct Accountabilities............................................................20 Aid Coordination Mechanisms and Processes.....................................................................23 The Dual Implications of Sectorwide Approaches...............................................................24 Implications of Country-Led Partnership for Conditionality and Monitoring........................25 4. Recommendations ........................................................................................................27 Implementation Challenges for IDA...................................................................................27 Challenges for IDA and Its Development Partners..............................................................27 Annexes Annex 1 Implementation Matrix for IDA10­12 Undertakings Related to Aid Coordination ...........................................................................................................................29 Annex 2 References ...........................................................................................................34 Annex 3 Persons Interviewed.............................................................................................38 Annex 4 Methodological Notes...........................................................................................40 ii Boxes and Figures Box 1.1 Key Passages on Aid Coordination in IDA Replenishment Reports.............................2 Box 2.1 IDA's Coordination in the Field ...............................................................................5 Box 2.2 Challenges for Programmatic Selectivity at the Country Level................................. 10 Box 2.3 Partner Views on Sectorwide Approaches...............................................................12 Box 2.4 Aid Coordination and Civil Society ........................................................................ 14 Box 2.5 Sector Programs and the Strategic Partnership with Africa...................................... 15 Box 2.6 A New Approach to IDA Allocations to Post-Conflict Borrowers............................. 18 Box 3.1 Donor Harmonization Required at the Country Level and the Corporate Level.......... 22 Box 3.2 High Transaction Costs Expected to Decline ........................................................... 23 Figure 3.1 Country Characteristics and Aid Coordination Arrangements.................................. 19 Figure 3.2 Development Assistance Characteristics and the Aid Coordination Environment...... 20 Figure 3.3 Relative Benefits and Costs of Aid Coordination Efforts......................................... 24 iii Abbreviations and Acronyms CAE Country Assistance Evaluation CAS Country Assistance Strategy CDF Comprehensive Development Framework CG Consultative Group CGAP Consultative Group for Assistance to the Poorest CODE Committee on Development Effectiveness CPIA Country Policy and Institutional Assessment DAC Development Assistance Committee (OECD) FRM Resource Mobilization Department HIPC Highly indebted poor countries IBRD International Bank for Reconstruction and Development IDA International Development Association IMF International Monetary Fund MDB Multilateral development bank NGO Nongovernmental organization OED Operations Evaluation Department OPCS Operations Policy and Country Services OECD Organisation for Economic Co-operation and Development PER Public Expenditure Review PIU Project implementation unit PRSC Poverty reduction support credit PRSP Poverty Reduction Strategy Paper SPA Strategic Partnership with Africa (formerly Special Program of Assistance for Africa) UNDP United Nations Development Programme UNHCR United Nations High Commissioner for Refugees UNICEF United Nations International Children's Education Fund iv Executive Summary i. Aid coordination is one of ten themes that the Operations Evaluation Department (OED) has assessed as part of its review of the International Development Association (IDA). The review covers IDA's implementation of recommendations made in the agreements that capped the tenth, eleventh, and twelfth replenishments of the association. Those agreements cover the periods of 1994­96, 1997­99, and 2000­02. As requested in the IDA12 Replenishment Report, the review examines both the extent of IDA's compliance with specific replenishment directives and the effectiveness of the actions taken.1 ii. Improved aid coordination has long been promoted for its potential contribution to the effectiveness of externally funded development programs and projects. A parallel tenet is that primary responsibility for aid coordination should lie with recipient countries, with donors directing their support to sound country-designed policies and programs. However, significant progress in advancing these complementary processes has remained elusive. iii. The relevant recommendations in the three recent replenishment agreements show a marked evolution. Whereas the IDA10 Replenishment Report urged greater coordination among development assistance agencies, IDA11 and IDA12 were more concerned with persuading IDA to follow the principle of comparative advantage in achieving programmatic selectivity. IDA12 emphasized the desirability of country-led partnerships that combine the objectives of country ownership and donor coordination. iv. Overall, this review finds that IDA has made substantial progress toward the commitments expressed in the replenishment agreements, particularly in the last few years. Progress has come through a sequence of Bank initiatives, including the "Partnership Initiative" of 1998 (World Bank 1998b) and the subsequent Comprehensive Development Framework (CDF) and Poverty Reduction Strategy Paper (PRSP) initiatives, followed by efforts to align operational policies and procedures with those initiatives. This is not to say that there is no room for improvement on IDA's part. This review identifies areas where additional actions are needed (see paragraphs xi- xiii below). Nonetheless, the major initiatives of the last few years signify a change in IDA's approach to aid coordination that is consistent with the thrust of replenishment recommendations to: (1) strengthen coordination with other development assistance agencies; (2) increase programmatic selectivity; (3) move to country-led partnerships; and (4) address special issues relating to coordination on public expenditure analyses and assistance for postconflict countries. v. Thanks largely to the CDF and PRSP initiatives, the principles of country ownership and partnership are now informing IDA's strategies and priorities on numerous fronts. The implications of both principles for IDA's role in aid coordination have not, however, been adequately spelled out. At the corporate level, a new Operational Memorandum on aid coordination and partnership is in the works but has yet to be issued. Meanwhile, IDA field performance varies considerably from country to country. In assessing compliance and effectiveness on the ground it is clear that it takes some multiple of "two to tango" to reform the long-standing practices of all development partners. To date, the move to genuine country-led partnerships that effectively combine ownership and partnership is being made in only a few IDA countries, typically in one or two sectors. Yet, the implementation of the principles of the CDF 1. The review draws heavily on a recent Operations Evaluation Department (OED) study, Review of Aid Coordination and the Role of the World Bank (World Bank 1999a). v and the success of the PRSPs will depend in large part on the extent to which country-led partnerships become a reality. vi. The experience to date in preparing full PRSPs has reinforced the significance of the capacity of the recipient country as a prerequisite for country-led partnership. Country commitment to sound development priorities and policies within a strategic framework oriented toward poverty reduction is also crucial. Strong country commitment and high institutional capacity tend to be associated with country-driven arrangements, which should be more effective in promoting greater development effectiveness than arrangements driven predominately by donors.2 vii. The characteristics of donors and international agencies also have a bearing on the relative ease or difficulty of moving to effective country-led aid coordination. Where donors and agencies are relatively few but give high priority to a development rationale for aid, the conditions for achieving country-led aid coordination are favorable. As the number of donors and projects increases, however, the environment for aid coordination becomes more challenging. Moreover, where nondevelopment motives for providing aid--such as the commercial interests of donors-- are dominant, effective country-led coordination is more problematic. viii. The foregoing considerations underscore the importance of obtaining acceptance by the parties of clear mutual responsibility and distinct accountabilities. Both the country and development agencies must accept their mutual responsibility for achieving development outcomes, while each partner agrees to be held accountable in certain areas. Development agencies must understand and accept that they are accountable for quality aid delivery--for coherence with country priorities, selectivity along lines of comparative advantage, efficient delivery of assistance, and support for country-led efforts to build capacity. Recipient countries, for their part, are accountable for developing sound policies and effective institutions that are underpinned by transparent and accountable systems of financial resource management. ix. Although IDA's coordination efforts have resulted in increased attention to coherence and selectivity at the country and corporate levels, implementation challenges at the country level and gaps at the corporate level remain. The support of IDA and other agencies for building recipients' capacity to coordinate aid has been weak. Development agencies have an obligation to help strengthen their recipients' capacity for transparent and accountable financial management in the interest of effective aid coordination. x. IDA-led Consultative Groups (CGs) and expanded local coordination forums have moved in the direction of country-led partnership, but challenges and tensions remain, including high front-end transaction costs. Sectorwide approaches continue to be seen as a way to combine country ownership and stronger donor coordination, as called for in an IDA12 commitment. By shifting the focus of accountability from the relationship between development agencies and their project implementation units (PIUs) to a sectoral relationship involving all external partners and the recipient government, sectorwide approaches can set the stage for partners' acceptance of mutual responsibility and distinct accountabilities. To the extent that development agencies work directly with a sectoral ministry rather than through individual project units, sectorwide approaches are more likely to contribute to country capacity than weaken it. However, experience also shows that several prerequisites must be present if sectorwide approaches are to realize those objectives. 2. For elaboration of the points in paragraphs vi­ix, see discussion in paragraphs 53-66 of the main text and in World Bank 1999a: 6, 23­24. vi Recommendations xi. The following recommendations can be implemented for the most part by IDA within its existing policy framework. Others that require IDA to act in concert with its partner agencies are highlighted below as special challenges. Several of these recommendations could be strengthened by being included in the IDA13 Replenishment Report. Implementation Challenges for IDA xii. The following recommendations represent changes that IDA can make within its existing policy framework. 1) Promote country-led aid coordination mechanisms and processes. In promoting country-led initiatives, IDA needs to be sensitive to the transaction costs that such initiatives can impose on the country and on other development agencies. Overlapping groups should be avoided and, where appropriate, a lead donor identified for each local group. Experience shows that adequately staffed and empowered country offices enhance IDA's role in supporting country- led aid coordination. Meeting this challenge requires adequate budgetary support (see recommendation xiii-3 below). Other development agencies that would assume a lead role in providing coordination and support must also be prepared to commit sufficient professional and budgetary resources to perform the role effectively. 2) Continue to strengthen and refine sectorwide approaches by incorporating the lessons of experience. Sectorwide approaches have the potential to strengthen country-led partnership and donor coordination, and at the same time to strengthen country capacity and shift the aid paradigm to one of mutual responsibility and distinct accountabilities. As IDA moves to more comprehensive programmatic support under the system of PRSPs, sectorwide approaches will likely continue to be an essential building block. 3) Develop and implement methods for monitoring and evaluating country-led partnerships and IDA's performance under the PRSP. Monitoring is essential, both to report on progress and to identify constraints to achieving it. IDA, working with its partners, should develop collaborative mechanisms and key indicators that can be used to monitor progress toward country-led partnership in the management and coordination of aid resources. Challenges for IDA and Its Development Partners xiii. The IDA Deputies have a dual role that is particularly important with respect to these recommendations, which have implications for the development agencies with which IDA works. The Deputies are in a unique position to convey the recommendations to donor agencies in their own countries as well as to IDA. The Deputies should: 1) Pursue programmatic selectivity at the country level and push the envelope at the corporate level. The challenge here is to ensure that selectivity proposed at the country level--in Country Assistance Strategies (CASs)--is consistent with what is set out at the corporate level in sector strategy papers and other documents. By the same token, the sector strategy paper process needs to take better account of resource and country considerations. Beyond the sectors, IDA needs to define more sharply its comparative advantage as an institution and carry through on the implications for its programs at the sector and country levels. vii 2) Ask IDA to work toward aid relationships in which all partners accept mutual responsibility for development outcomes, agree to be held accountable for their activities, and review the performance of each partner. A promising approach is being pursued by the Tanzania CG, which has established an independent review panel to assess donor performance and report its findings to the CG. A challenge for IDA in implementing this recommendation will be that of exercising leadership in concert with its partners without dominating the process or being perceived as doing so. This modus operandiis important even in those cases where strong donor-led coordination is required because country capacity and commitment are lacking. An effective partnership also requires that weaker partners be strengthened, so as to level the playing field. IDA should therefore formulate a joint strategy with other key partners to strengthen country capacity to manage and coordinate aid resources. The experience of the Partnership for Capacity Building in Africa (PACT) and other relevant initiatives should be drawn upon in formulating the strategy. 3) Endorse a global role for IDA in donor harmonization. The forces working against donor harmonization are strong. The World Bank, as a preeminent multilateral development institution, is in a unique position to exercise global leadership in working with other development assistance agencies toward meaningful harmonization. The harmonization agenda goes beyond procedural issues--important as these are--to critical issues of policy and strategy such as selectivity and performance-based allocation. These challenges need to be pursued at both the country and corporate levels, where the agencies of the United Nations need to be included as well. This is not just an issue for IDA, but one that needs the support of all development agencies, taking into account the policies, procedures, and capacities of client countries. IDA should convene a conference of the leaders of multilateral and bilateral development assistance organizations to seek agreement on the major harmonization issues that impede country-led partnership and agree on a program of action for harmonization.3 4) Provide adequate resources to promote country-led partnership. Assuming the leadership roles described above will take dedicated allocations of IDA budget and staff--in resident missions as well as at headquarters. It also will require a realignment of staff and management incentive structures, to recognize and reward adequate attention to the needs of partnership. Without such efforts, the opportunity provided by the CDF and PRSPs to put into practice the lessons of 50 years of development cooperation is likely to be lost. 3. Such a conference would acknowledge and take into account ongoing harmonization efforts, including the meetings of the presidents of the multilateral development banks and related harmonization forums and working groups, the Informal Group of Multilateral Secretariats convened by the chair of the Development Assistance Committee of the Organisation for Economic Co-operation and Development, and the Development Assistance Committee's Task Force on Donor Practices. 1 1 The Evolution of IDA's Aid Coordination Commitments 1. An effective development partnership requires the partners to agree on the objectives of aid and on their respective responsibilities and areas of accountability in realizing those objectives. It also requires them to support capacity development, an essential element in ensuring that weaker members are able to participate fully in development partnerships (Picciotto 1998). 2. Ideally, recipient countries assume primary responsibility for aid coordination, while donors recognize and support sound country-designed policies and programs. But, progress in advancing these complementary processes has remained elusive. Given stagnating flows of officialdevelopment assistance, disappointing development performance, and increased demands for better accountability and results, the international development community has over the last five years renewed and intensified its calls for strengthened aid coordination. That concern is reflected in the recommendations that followed the last three replenishments of the International Development Association (IDA10, IDA11, and IDA12). 3. Shortly after the IDA12 agreement was issued, the President of the World Bank proposed the Comprehensive Development Framework (CDF), which incorporates country ownership and partnership as two of four central organizing principles for development effectiveness (Wolfensohn 1999).4 Pilot programs were launched in eight IDA countries.5 A year later, the CDF principles were being put into practice in a much larger number of countries as the Bank and the International Monetary Fund (IMF) joined to assist IDA countries in the preparation of Poverty Reduction Strategy Papers (PRSPs).6 As will be elucidated in subsequent chapters, both initiatives have major implications for IDA's role in aid coordination. 4. Aid coordination is one of ten themes being examined in the Operations Evaluation Department's (OED) review of IDA. This paper draws heavily on a recent OED study, "Review of Aid Coordination and the Role of the World Bank" (World Bank 1999a, identified hereafter as the 1999 OED study),7 as well as on recent reports on the CDF and the PRSP process (World Bank 2000e, 2000m, and 2000n). The analysis has been supplemented by a review of related documents and selected interviews to update the 1999 OED study and cover certain IDA replenishment issues not treated in depth by that evaluation. It also draws on the findings of eight country consultations undertaken for the IDA review.8 4. The other two principles are a long-term, holistic approach and a focus on results. 5. Pilot CDF programs are underway in eight IDA countries: Bolivia, Côte d'Ivoire, Eritrea, Ethiopia, Ghana, Kyrgyz Republic, Uganda, and Vietnam. Non-IDA pilots include Dominican Republic, Morocco, Romania, and West Bank and Gaza. Several nonpilot IDA countries are also consciously implementing the CDF, including Bhutan, Tanzania, and Zambia (World Bank 2000m: 53­54). 6. The relationship between the CDF and PRSPs is explained in a "Joint Bank-Fund Note on CDF-PRSP Linkages" issued in April 2000 (World Bank 2000n: 22). 7. The reference above is to a version of the study discussed by the Committee on Development Effectiveness (CODE) of the Bank's executive directors in November 1999. The main focus of the study is IDA countries with Bank- convened consultative groups. The same study, under the title "The Drive to Partnership: Aid Coordination and the World Bank" (World Bank 1999b), was discussed at a Development Partnership Forum jointly sponsored by the Development Assistance Committee, the United Nations Development Programme, and the World Bank. The forum was held in Paris, December 7­8, 1999. 8. Bangladesh, Bolivia, Cambodia, Ghana, India, Kyrgyz Republic, Mozambique, and Vietnam. 2 5. Eighteen commitments made in the IDA10­12 replenishment agreements explicitly address some aspect of aid coordination (annex 1). Most fall into one of three clusters that refer to important dimensions of aid coordination. · Strengthening coordination with other development assistance agencies. · Increasing programmatic selectivity. · Moving to country-led partnerships. Related special issues include improving coordination of work on public expenditure analyses and cooperating on a comprehensive assistance approach for postconflict countries. 6. Key passages from the respective IDA replenishment reports show an evolution in perspective from increasing coordination among development assistance agencies in IDA10 to achieving greater programmatic selectivity, reflecting IDA's comparative advantage, and, in IDA12, to country-led partnerships that combine the objectives of country ownership and donor coordination. IDA12 also contains specific commitments intended to strengthen coordination between IDA and the IMF, and between IDA and regional multilateral development banks (box 1.1).9 Box 1.1 Key Passages on Aid Coordination in the IDA10­12 Replenishment Reports IDA10 (March 1993) The [IDA] Deputies attach particular importance to the role that IDA plays in coordinating the efforts of donors in specific regions, countries, and sectors. These efforts have helped to improve the overall effectiveness of aid and to develop a common view among donors of priority development needs. Given the constant real level of official development assistance, there is an even greater premium on the efficient use of such resources and the Deputies look to IDA's leadership in this regard. (IDA 1993: paragraph 48) IDA11 (April 1996) In planning the assistance strategy, IDA takes into account the activities and plans of nongovernmental organizations, bilateral donors, United Nations agencies, and regional development banks, consulting with the main donors in each case, with a view to maximizing the complementarity and effectiveness of donor assistance overall. (IDA 1996: paragraph 9) IDA12 (December 1998) Partnership with borrowing countries entails strengthening capacity for coordination based on their own development strategies and a readiness to adjust to local circumstances, to build capacity and to systematically involve civil society. Forging strong partnerships with other donors and stakeholders in borrowing countries is critical, given that IDA has limited resources and that it is only one source of aid to low-income countries (IDA 1998: paragraph 56). Source: International Development Association 1993, 1996, 1998. 7. The next chapter assesses IDA's compliance with the commitments described above. The final chapters draw lessons and make recommendations based on the findings of that assessment. 9. Other specific aid coordination commitments in IDA10 refer to IDA's relationships with bilateral agencies and with UN agencies (IDA 1993: 19). 3 2 The Performance of IDA and Its Partners 8. Overall, the International Development Association (IDA) has made substantial progress toward the commitments expressed in the agreements that capped its tenth, eleventh, and twelfth replenishments (IDA10, IDA11, IDA12) (IDA 1993, 1996, and 1998). Progress has come through a sequence of Bank initiatives, including the Partnership Initiative of 1998 (World Bank 1998b) and the subsequent Comprehensive Development Framework (CDF) and Poverty Reduction Strategy Paper (PRSP) initiatives, followed by efforts to align operational policies and procedures with those initiatives. This is not to say that there is no room for improvement on IDA's part. This review identifies areas in which IDA can take additional actions. Nonetheless, the major initiatives of the last few years signify a change in IDA's approach to aid coordination that is consistent with the thrust of replenishment recommendations related to aid coordination. Moreover, it must be recognized that IDA can go only so far on its own. In assessing compliance and effectiveness on the ground it is clear that it takes some multiple of "two to tango" to reform long-standing practices. 9. The following sections are organized by the broad commitment clusters delineated in the previous section. The discussion of IDA's undertakings draws in part on a matrix presented in annex 1. The assessment of compliance and effectiveness on the ground draws from a range of sources, including the published literature, surveys, interviews, workshops, and field visits analyzed in the 1999 OED study of aid coordination. As mentioned in section 1, additional sources specific to the IDA review include reports of visits to the eight IDA countries selected for consultations and a review of a substantial number of recent documents and interviews with selected Bank staff (annexes 2 and 3). Annex 4 describes the methodologies employed for three analyses undertaken expressly for this paper: (1) an assessment of the treatment of aid coordination in IDA Country Assistance Strategies (CASs); (2) a review of findings related to aid coordination in recent OED Country Assistance Evaluations (CAEs) of IDA programs; and (3) a review of aid coordination in recent IDA Public Expenditure Reviews (PERs). Strengthe n Coordination with Other Development Assistance Agencies 10. The seven IDA commitments in this cluster are concerned with IDA's relationships with the following categories of development assistance agencies: bilateral donors (at the country level), the United Nations (UN) system, the International Monetary Fund (IMF), and regional multilateral development banks. IDA10 commitments were concerned with the first two categories; IDA12 commitments dealt with IDA's coordination with the IMF and regional development banks. Strengthen country level coordination with bilateral donors 11. Although IDA management did not issue any guidance to staff with respect to this IDA10 commitment (or any other specific IDA10 commitment), it did include a special section on aid coordination activities--including local aid coordination--in its implementation report on the tenth replenishment (World Bank 1997: 25­28). The subsequent Partnership Initiative of 1998, followed by the CDF and the advent of the PRSPs, heightened the importance given to country- level coordination. Findings from a wide range of sources confirm that IDA's field managers have in recent years taken steps to improve coordination at the local level, such as convening or participating in periodic meetings with localdonor and government representatives.10 Interest has 10. Sources include reports from country consultations conducted for this review; an analysis of 23 recent OED Country Assistance Evaluations for IDA countries, described in annex 4; the OED aid coordination study (World Bank 1999a); the March 2000 mid-term progress report on the CDF (World Bank 2000d: ii, 8­10); and the August 2000 report on country experience (World Bank 2000m: 16­17). 4 been strong in holding Consultative Group (CG) meetings in IDA countries. Although the question of in-country meeting venues has pros and cons, experience suggests that in-country locations facilitate country ownership. This issue is revisited in the section of this paper that deals with IDA12 commitments on partnership and aid coordination.11 12. Placement of IDA country directors in the field, with accompanying delegations of authority, has reinforced these developments and is generally applauded by the broad spectrum of IDA's development partners. However, as illustrated in box 2.1, complaints persist about insufficient delegation of authority to the field and inadequate consultations with other donors at the country level, especially on the part of missions from headquarters, and about IDA's shortcomings in sharing and seeking information and in providing feedback in planning processes, such as for the CAS (World Bank 1999a: 21).12 11. See paragraphs 39-40 below and the accompanying footnote. 12. This finding is somewhat at variance with the observation in the above-mentioned IDA10 commitment that while "visiting missions made efforts to contact other donors, more could be done by IDA's Resident Representatives to ensure that donors were kept informed between missions" (IDA 1993: 19). The 1999 OED study reports a common criticism from governments, and especially from local donor representatives, of arroganceand insensitivity by visiting Bank missions (World Bank 1999a: 21). Reports from the country consultations conducted for this review contain similar criticisms. 5 Box 2.1 IDA's Coordination in the Field Representatives of local development agencies and governments differ in their views on the effectiveness of IDA's aid coordination efforts. · "The split between headquarters and [the field office] is creating problems.... IDA is [limited] by a critical lack of field staff with decisionmaking authority and by poor communication between headquarters and the field." · "Focus groups ... note a positive change in the Bank's attitude from a secretive to a more open behavior. However, donors are the most critical.... Sector specialists in particular complain about not having been invited to participate in a meaningful way." · "IDA missions consulted donors regularly to keep them informed of developments." · "Several donors argued that IDA was sometimes too much of a protagonist and maintained that they would be more ready to collaborate if the government, rather than IDA, did the leading." · "Some donors feel threatened by IDA's dominant position. HQ staff arrive in too much of a hurry and are insufficiently sensitive to local conditions. They need to be given more time, be more patient, and recognize that reform cannot be rushed." · "In a number of areas the Bank has yielded the lead donor role to other institutions willing and able to fulfill that role." · "IDA is criticized for not consulting with its partners, but rather informing them after the fact of its decisions." · "Donors agreed that the local consultative group has institutionalized a mechanism for sharing information among the donor agencies, but underscored that they are far from effective donor cooperation." Sources: Reports from country consultations conducted for OED IDA review; OED Country Assistance Evaluations for IDA countries (annex 4); World Bank 1999a; World Bank 2000d; World Bank 2000m. Make more effective use of the capabilities of the United Nations system 13. This IDA10 commitment refers specifically to better use by IDA of "the skills and experience available in the UN system in support of its country programs, especially in human resource and institutional development" (IDA 1993: 19).13 Consultation and coordination between IDA and UN agencies at leadership and headquarters levels have reportedly increased since the beginning of the IDA10 period in 1994. The IDA10 implementation report refers to cooperation with UN agencies having "intensified" (World Bank 1997: 25), and Bank staff posted in New York point to closer coordination with the UN system having occurred during the IDA11 period. After a hiatus of a decade, the Bank recently renewed its resident presence in Geneva. The "Partnership Initiative" and the Comprehensive Development Framework, as well as strong interest by top leadership of IDA, UN agencies, and the UN secretary-general, have all contributed to closer coordination. Another contribution has come from the substantial increase in attention given by IDA over the last three years to postconflict assistance and the increased coordination with UN agencies that this has necessitated, especially with the office of the United Nations High Commissioner for Refugees.14 13. The commitment also recommends cooperation with the UN Commission on Sustainable Development, which was particularly extensive during the IDA10 and IDA11 periods (Martin-Brown interview 2000). 14. A 1998 memorandum of understanding between the World Bank and the UNHCR adds to those signed over the years with UN agencies such as UNDP and FAO (World Bank 1998h). 6 14. However, coordination experience in the field has been mixed. Given the central development role of the United Nations Development Programme (UNDP) among UN agencies, especially with respect to human resource and institutional development, its relationship with IDA is important. In an effort to clarify their respective aid coordination roles, the two institutions concluded three agreements between 1986 and 1996. Although the 1999 OED study found examples of positive collaboration, it also recorded continued reports from several local donor representatives of counterproductive tensions in the field (World Bank 1999a: 19­20). As a response to the above-mentioned IDA12 commitment calling for Board review of proposals to strengthen partnership, IDA submitted a paper in November 1999 that cites several initiatives and good practice examples regarding World Bank­UNDP coordination at the headquarters and country levels (World Bank 1999c: 5­6). 15 The CDF has stimulated more intensive contacts between the Bank and the UN system through the "UN focal points network" and exchanges with other UN bodies (World Bank 2000d: 12, and 2000m: 4). Additional impetus to improved coordination has come from the PRSP process. At the same time, the nature of the relationship between the Bank's two key initiatives and the UN's Development Assistance Framework could benefit from further clarification. Both the Bank and the Fund have recently entered into a formal agreement to cooperate with the UNDP in assisting countries in the preparation of PRSPs, and similar agreements with the International Labour Organization and UNICEF are under discussion (World Bank 2000n: 30). Strengthen coordination with the International Monetary Fund 16. The IDA12 commitment to improve coordination with the IMF noted the "widespread use of joint World Bank­IMF missions" and recommended that such missions be increased (IDA 1998: 21). Guidelines for collaboration between the two institutions, including division-of-labor issues, have been promulgated from time to time for three decades. Historically, experience with joint missions and Bank-Fund coordination has varied considerably from one region to another, but the mandate for joint support for the PRSP process has given the coordination objective new prominence and impetus.15 The leadership of both institutions has promoted closer coordination and cooperation in implementing that initiative and the related initiative for highly indebted poor countries (HIPC). The PRSP process has been accompanied by a series of joint policy statements and guidelines, as well as an expansion in joint committees, working groups, and joint or closely coordinated missions, all of which adds up to a high degree of compliance by IDA over a relatively short period of time. 17. By the beginning of 2000, two sets of PRSP guidelines called for joint missions between the Bank and the Fund (World Bank 1999e: 14­15 and 2000a: 4).16 The latest progress report describes close Bank-Fund collaboration in all aspects of work on PRSPs, including the preparation of joint staff assessments that accompany papers for discussion by the respective 15. Given the close relationship between the objectives of the two institutions, it is not surprising that the history of joint guidelines for collaboration goes back to 1966. General guidelines have been issued at occasional intervals by the President of the Bank and the Managing Director of the Fund (Development Committee 1998). Those were supplemented by guidance on collaboration on public expenditure issues in 1995 and on strengthening financial sectors in 1997 and 1998 (for references, see Development Committee 1998: 2). Interviews with IDA staff indicate that joint Bank-Fund missions have been more common and routine, and coordination more harmonious, in Central America than in Africa. 16. The Bank's Internal Guidance Note on PRSPs recognizes that not all missions need be joint, stressing that policy dialogue and all documents are to be shared and that close collaboration is required in the preparation of the "Joint Staff Assessment" (World Bank 2000a: 4). The first PRSP progress report, issued in April 2000, cites nine joint PRSP­ related missions to the Latin America and Caribbean Region and ten joint missions to the Africa Region (World Bank 2000e: 8). 7 boards (World Bank 2000n: 15). A Bank-Fund Joint Implementation Committee, established in May 2000, meets biweekly to monitor progress, coordinate reports and briefings, and resolve policy and process issues.17 18. Perhaps inevitably, tensions have emerged as the two institutions have sought to coordinate their work on PRSPs. The tensions have been ascribed to differences in "organizational culture," such as a reported ability of the IMF to implement decisions more expeditiously than the Bank.18 Notwithstanding the guidelines and such mechanisms as the Joint Implementation Committee, coordination at the operational level has reportedly been of mixed quality. The current progress report cites a recent review of early experience prepared by the Department for International Development in the United Kingdom that "offers mixed views from field staff on coordination between the Bank and Fund and other donors ... basically related to a failure to change behaviors (`business as usual')" (World Bank 2000n: 28). Management at the Bank and Fund recognizes the need for a change in organizational culture and is expanding and revising internal staff training efforts and exercises accordingly.19 19. In sum, it would appear that the PRSP process is bringing about an unprecedented degree of coordination between IDA and the IMF. That collaboration will ultimately stand or fall on the extent to which it results in consistency on the policy issues of primary concern to the two institutions--macroeconomic in the case of the Fund and structural and sectoral in the case of the Bank--and on how the two institutions' views should be incorporated into a coherent poverty reduction strategy.20 Strengthen coordination with regional development banks 20. Management took steps that anticipated or accompanied several IDA12 commitments regarding coordination with regional development banks. These and subsequent actions by IDA indicate a high degree of compliance with the commitments. 21. In response to the 1996 report of the Task Force on Multilateral Development Banks of the Development Committee of the Boards of Governors of the IMF and the World Bank (Development Committee 1996), the banks have sought greater collaboration and coordination on a broad range of fronts. Seven working groups have been established at the semi-annual meetings of bank heads. Of the seven, the two mentioned in the IDA12 Replenishment Report, procurement and evaluation, are the oldest and most active and have the most to show for their work.21 After more than two years of intense effort, the procurement working group has produced 17. The Joint Implementation Committee has set up three task groups dealing with (1) governance/public expenditure issues; (2) Bank-Fund roles regarding country issues and policy, including approaches to conditionality; and (3) options for accelerating decisions pertaining to highly indebted poor countries. The committee takes the lead in designing initiatives for capacity building, training, and measures to "enhance transparency and information dissemination" (World Bank 2000n: 15). 18. The Fund's ability to implement decisions more quickly than the Bank was mentioned in interviews with both Bank and Fund staff. 19. "Clinics" covering the PRSP process have been held for Bank and Fund staffs, and internal training is reportedly stressing "the core principle of country ownership" so as to "dispel perceptions of dominance of the process, or `business as usual'" (World Bank 2000n: 24). 20. Respective areas of "traditional expertise, mandate, and responsibility" for the Bank and the Fund are elaborated in some detail in Poverty Reduction Strategy Papers: Operational Issues, jointly prepared by the Bank and the Fund in December 1999 (World Bank 1999e: 13­14). Some areas--including policy environments for private sector growth, trade liberalization, and financial sector development--are identified as needing to be "shared between the two staffs." 21. The other working groups are financial sector fundamentals and reform; governance, corruption and capacity building; environmental assessments; private infrastructure; and poverty issues (World Bank 2000c: 3). 8 standard bidding documents for procurement of goods, consultant services, and civil works. Documents setting forth harmonized procurement policy and guidelines for combating corruption in procurement are in preparation. The evaluation cooperation group has agreed on standard evaluation criteria, and members are working to harmonize their respective guidelines with the standard criteria. The group is also working on statements of good practice with respect to evaluation procedures for private sector and public sector operations, as well as for the evaluation of institutional development impact. The recently formed group on poverty issues (with special reference to the PRSP process) produced a paper that was well received at the Bank-Fund annual meetings in September 2000. A working group on harmonized project processing is under discussion. This new group would seek agreement on synchronized missions and reporting procedures, among other things. 22. The importance attached to harmonization efforts by key elements in IDA's authorizing environment is demonstrated by two requests for progress reports that emerged from the September 2000 annual meetings: one from the G-7 finance ministers, and the other from the Development Committee. Moreover, efforts to harmonize the work of the regional development banks are to be closely coordinated with bilateral harmonization efforts, which have been given new impetus as a result of an October 2000 decision by the Development Assistance Committee of the Organisation for Economic Co-operation and Development (OECD) to establish a two-year bilateral task force on the harmonization of donor practices.22 23. An IDA12 commitment called for the World Bank and the African Development Bank to clarify respective institutional roles in order to "ensure complementary approaches and effective use of assistance resources" (IDA 1998: 21). A closely related IDA12 commitment states the need to "coordinate concessional assistance flows to Africa from multilateral development banks to ensure optimal development impact" (IDA 1998: 15). Consultations beginning in 1997 between the two institutions led to the commissioning of a team of independent experts in mid- 1998 that produced a study for a strategic partnership between the two institutions (Team of Independent Experts 1999). Drawing from this study, the management of the two banks initiated a consultative process that resulted in an aide-mémoire and a subsequent memorandum of understanding signed by both institutions in March 2000 (African Development Bank­World Bank 2000). Those developments are consistent with an IDA12 commitment calling for the two banks to clarify respective institutional roles in order to "ensure complementary approaches and effective use of assistance resources" (IDA 1998: 21). 24. The memorandum of understanding identifies areas for cooperation between the two institutions at the country, sector, and thematic levels and with regard to capacity building and institutional strengthening. Although the memorandum did not interpret the division of labor as clearly as suggested by the experts, it does provide a basis for broad-ranging cooperation. An implementation monitoring matrix, with agreed actions and target dates, has been jointly prepared. Monitoring and evaluation mechanisms will be essential for assessing the effectiveness of the memorandum of understanding. 25. To enhance implementation of the understanding between the World Bank and the African Development Bank, plans are underway to introduce a partnerships training module for the staffs 22. The first meeting of this task force was held at the OECD on January 25­26, 2001, and was followed by a World Bank­convened forum for multilateral financial institutions, which included the Chair of the bilateral DAC Task Force, on February 1, 2001. The forum focused on the harmonization of operational policies, strategies, and practices. An ongoing "Informal Group of Multilateral Secretariats" convened by the chair of OECD's Development Assistance Committee has focused on monitoring "key changes in agency procedures and behavior needed to implement country- led partnerships" (OECD 2000). Also see World Bank 2000p. 9 of the two institutions. A possible second stage would broaden participation to government and nongovernment stakeholders from the region. 26. Paralleling the cooperation between the Bank and the African Development Bank, senior management of the Asian Development Bank and the relevant regions of the World Bank have recently signed a joint letter calling for enhanced strategic collaboration (Asian Development Bank­World Bank 2000). On February 9, 2001 the MDB Presidents issued a joint statement that committed MDBs to closer collaboration to reduce poverty based on the CDF principles. Increase Programmatic Selectivity 27. Three commitments--one in IDA11 and two in IDA12--convey the Deputies' growing concern that IDA and other donors should achieve greater programmatic selectivity. Management has clearly conveyed to staff the need to address selectivity at the country level and provided guidance through such vehicles as directives on CASs and sector strategy papers (notably in the social sectors). Staff response, as reflected in CAS documents, has been mixed. Retrospective analyses issued in 1998 and 1999 by the Bank's Operations Policy and Strategy Vice Presidency (now Operations Policy and Country Services) found that although most CASs discussed donor coordination and comparative advantage in broad terms, fewer than 40 percent discussed selectivity in connection with program design for key areas. The 1999 retrospective notes that "further improvement is needed, particularly in CASs for IDA countries, where severe donor coordination problems persist" (World Bank 1999d: 24; also 1998a). Reviews conducted for the 1999 OED study and for the current review confirm that recent CASs have taken increased note of the activities of other donors (World Bank 1999a: 52­59), but only a few go beyond a description of what IDA and other donors are supporting to make the link between IDA's comparative advantage and its strategy and program, including the prioritization of IDA activities "across and within sectors and by instrument" as proposed by one IDA12 commitment.23 28. The Bank's recent report on country experiences under the CDF indicates increasing awareness of the selectivity issue in the eight pilot IDA countries but also notes that progress in achieving greater selectivity on the ground has been limited. The report comments that "individual countries' concerns about visibility, and about their freedom to frame their development cooperation policies in the context of their overall foreign and trade policies, often preempt better partnership" (World Bank 2000m: 18). Difficult technical and organizational implications of pursuing comparative advantage for individual partners, including the Bank, are also noted. The report calls for innovative approaches and provides the example of a recent discussion between the government of Tanzania and donors, in which it was suggested that an "independent group ... be established to judge donor performance" (World Bank 2000m: 19). Following a recommendation by the June 2000 consultative group meeting for Tanzania, an independent panel along these lines is now being planned. 29. The gap between top management's emphasis on the need for greater programmatic selectivity and the staff's performance on the ground results in part from the resistance of other donor partners, implying that IDA's efforts alone are not sufficient; IDA also needs to "comply through others" to ensure that a country receives adequate support among and within sectors. As suggested in box 2.2, poor selectivity can result in gaps as well as overlaps in donor assistance. 23. Thirty-eight CAS documents issued for IDA countries in the last eight years were reviewed for this paper. The methodology employed is explained in annex 4. 10 Box 2.2 Challenges for Programmatic Selectivity at the Country Level The 1998 CAS for Albania provides a straightforward indication of what IDA will and will not do: "Our assistance to Albania will continue to complement those of other donors. We will not be engaged in new lending to the transport sector, where the EU, the EBRD and Italy are already involved. We will also not be engaged in new lending to the power sector, where EBRD, Japan, Switzerland and Austria are already financing the rehabilitation of power generation. We will not be involved in lending to the telecommunications sector, where EBRD has a lead role. In the financial sector we are supporting the restructuring and privatization of the state banks, while the EU, the IMF, Italy and the U.S. are providing the institutional and infrastructural support. We will complement other donors' efforts in public administration and judicial reforms." (World Bank 1998c: 19) However, the 1998 OED Country Assistance Evaluation for Albania concludes that some priority sectors have been relatively underfunded by donor assistance, including public administration reform, with 4 percent of total assistance, and infrastructure, with 11 percent, even though this sector accounts for 45 percent of IDA assistance. (World Bank 1998f: 44) The 1998 country assistance review for Mozambique reported that donors believed they could provide necessary financial resources for the Health Sector Recovery Program, but saw IDA's intellectual and policy leadership as important. IDA is said to have shown substantial intellectual leadership in the energy sector but seemed timid in implementation, even though few other donors were present (Landau 1998). One of the country consultations undertaken for this review revealed that the concentration of donor assistance in a few sectors was increasing the burden on the corresponding government units, which faced more projects with different procedural requirements. At the same time, some sectors, such as agriculture, were being neglected. One IDA Country Assistance Evaluation reports that "the presence of virtually all donors in all sectors has placed an overwhelming burden on the Government and on aid coordination" (See paragraph 7, Annex 4). 30. Strategic selectivity at the corporate level has received little attention. A recent exception is a paper prepared by the Operational Policy and Country Services Vice Presidency, which recognizes that: Underpinning these country menus is the Bank's overall corporate menu. At a time of scarce corporate resources, when the development paradigm emphasizes cooperation among development partners for greater development impact, it behooves the Bank to apply the principle of strategic selectivity to the design of its corporate menus. Should the Bank offer everything--all sectors and all products? Or should its offerings be guided by a division of labor with other [international financial institutions] and partners? (World Bank 2000j: 11) 31. The paper suggests that the Bank is about to tackle the issue of strategic selectivity at the corporate level through sector strategy papers, which are expected to "increasingly set out the menu of lending and nonlending products to be offered by the Bank, along with a consideration of how the Bank's menus relate to those of other [international financial institutions]." The task of bringing about consistency in what is said about selectivity in CASs and what is said in a sector strategy paper has reportedly begun in the health, nutrition, and population sector, but there is a long way to go before all major sectors are addressed. The issue of intersectoral selectivity-- or IDA's overall comparative advantage at the corporate level--has yet to be delineated. Move to Country-Led Partnerships 32. The four IDA12 commitments dealing with the related issues of ownership and partnership were negotiated in 1998, in parallel with the launching of the Bank's Partnership Initiative. The 11 CDF and PRSP initiative have considerably expanded the boundaries of the 1998 initiative, both conceptually and operationally. Brief discussions of each of the four commitments follow. Aim through capacity building to ensure that countries can assume the central role in aid coordination 33. Capacity has emerged as a major constraint to country-led preparation of full PRSPs by IDA countries.24 The 1999 OED study finds that the practices of IDA and other donors (practices such as independent project implementation units) can undermine country capacity, and that efforts to support capacity building have tended to be supply-driven and ineffective (World Bank 1999a: 16­17, 22­23). The 1999 review recommended that IDA's approach to capacity development be demand-driven and not undermine existing capacity. Management has recently prepared a draft Operational Memorandum containing language that is consistent with the OED recommendation, although not as specific.25 Explore practical mechanisms to combine country ownership with improved donor coordination 34. IDA is grappling with the challenge entailed by this commitment. A paper prepared by the Operations Policy and Country Services Vice Presidency, "Supporting Country Development: World Bank Role and Instruments in Low- and Middle-Income Countries," concedes that assuming full responsibility for aid coordination "continues to pose problems, particularly for small states and low-income countries with limited capacity" (World Bank 2000j: 6). 35. The burdens imposed by disparate donor-supported activities and donors' divergent policies, procedures, and practices constitute a major obstacle to country leadership where capacity is weak. Over the last decade, IDA has participated in sectorwide approaches in several African countries and, to a more limited extent, in Asia. Although certain prerequisites must be met before sectorwide approaches can be effective (box 2.3), such approaches continue to be seen as a practical way of simultaneously increasing country ownership and improving donor coordination.26 New sectorwide initiatives being pursued by IDA in Cambodia and Vietnam are explicitly justified by expectations that they will simultaneously result in enhanced country capacity and ownership, more coherent donor support to country sector strategies, and greatly reduced aid delivery transaction costs (World Bank 2000b and 2000f). 24. The current progress report indicates that earlier schedules for production of full strategy papers are likely to be significantly delayed owing to capacity constraints (World Bank 2000n: 5­6, 21). The Bank's recent report on country experience under the CDF also indicates that in the eight IDA CDF pilot countries, "lack of government capacity is continually cited as a major constraint" (World Bank 2000m: 13). 25. This draft statement, which responds to a recommendation of the 1999 OED study calling for new operational policy, is intended to replace the existing OP 14.30 on "Aid Coordination Groups." This is in accordance with the Management Response of February 2001 to the OED Study. 26. See the 1999 OED study for a summary of pertinent literature on sectorwide approaches (World Bank 1999a: 17­ 19). Stakeholders interviewed by visiting IDA review teams viewed such approaches in health (in Bangladesh) and in education (in Mozambique) as having led to increased country leadership. One feature of effective sectorwide approaches that emerges from these visits as well as from experiences examined in the 1999 OED study is the need to proceed at a pace that permits the country to learn from experience and not be pushed by the implementation and disbursement deadlines of donors, including IDA. 12 Box 2.3 Partner Views on Sectorwide Approaches Rationale. Government officials from two IDA countries--Ghana and Mozambique--saw the initial rationale for sectorwide approaches as being to develop coherent sector strategies and subsequently to streamline donor assistance projects. They held that government ownership of several sectorwide approaches was "complete," and donors broadly concurred with their assessment. An improvement on the traditional approach. Donor representatives affirmed that sectorwide approaches were a significant improvement over the traditional approach of separate, uncoordinated projects, although they were harder to design. They noted that sectorwide health project in Bangladesh took five years to develop and that groundwork for Ghana's health program began in the late 1980s. IDA's role. IDA's role is generally considered to be supportive and catalytic, although its performance during implementation drew considerable criticism due to perceptions of unwillingness to modify "protracted and centralized" procedures. A country-owned "code of conduct". The Mozambique Ministry of Health has established a code of conduct for donors participating in the sectorwide health program. The Ministry signed first and was able to convince all other donors to sign. The Ministry stresses that its health program is a Mozambique-owned initiative. Source: Country consultations conducted for the OED IDA review. 36. IDA has not encouraged or promoted sectorwide approaches at the corporate level. The current PRSP progress report indicates "broad support" from donors for making the PRSP the focal point of donor assistance programs and for improved coordination in their implementation. When it comes to country proposals to donors to provide assistance in the form of "broad-based sectoral and budgetary support," reactions are more qualified, with some donors indicating an intention to "explore modalities for supporting PRSPs through more conventional vehicles" (World Bank 2000n: 14). 37. "Supporting Country Development" proposes a new lending instrument--a poverty reduction support credit--to support PRSP programs (World Bank 2000j: 13).27 The new credit would parallel the Fund's Poverty Reduction and Growth Facility and would thus enable better synchronization of the two institutions' conditions for the respective instruments. The poverty reduction support credits could have sectoral components but would also address key cross- sectoral constraints, such as the need for an adequate medium-term expenditure framework. However, sound sector strategies will need to undergird any move toward broad programmatic support. 38. Approaches and instruments that will facilitate harmonization at the country level should continue to be pursued. But progress on harmonization will depend as much, if not more, on actions taken by development agencies at headquarters levels and working through mechanisms such as the previously described harmonization working groups and task forces established by OECD's Development Assistance Committee and the multilateral development banks. Take a leadership role in aid coordination and strengthen CGs as a coordinating mechanism 39. This commitment also calls for increased involvement of civil society and nongovernmental organizations in the consultation process and for holding consultative group meetings in the recipient country (IDA 1998: 21). Although there has not been a major shift of meetings to the recipient country, movement in this direction is now discernible, with IDA 27. The poverty reduction support credit is described as a programmatic structural adjustment credit that focuses on poverty reduction. The purpose of programmatic structural adjustment credits is to provide medium- to long-term support for a country's policy reform and capacity building program, focusing on outcomes and results. 13 countries such as Bolivia, Ghana, Tanzania, and Uganda now (or soon to be) the regular venue for their consultative group meetings, and several other IDA countries rotating between the country and Paris, Tokyo, or Washington.28 Choice of location for "Development Forums," as apex-level aid coordination meetings are called in the South Asia Region of the Bank, are client- driven. The 2001 Pakistan meeting was held in Islamabad, for example, as was the 1998 meeting. 40. A possible interpretation of the Deputies' injunction to strengthen CGs might be that IDA itself should take a more dominant role in consultation processes. But this would run counter to the central principle of country ownership under the CDF and PRSP processes. The 1999 OED study contends that progress toward country ownership would be accelerated if each Bank country team that supports a CG were to work with the government and other development partners to formulate a strategy explicitly intended to help the country move toward aid coordination leadership. This need not imply the absence of a mechanism resembling the CG, but it does connote a process organized, convened, and chaired by the recipient country. The current draft policy statement on development partnership would commit the Bank to support country-led coordination processes and mechanisms but stops short of calling for the formulation of country- by-country strategies to help implement this objective. 41. The 1999 OED study reported that the "Bank has given substantially more attention to cooperation with civil society--particularly nongovernmental organizations (NGOs)--and the private sector in recent years, but is still in a tentative, experimental mode when it comes to actively seeking their increased involvement in aid coordination processes" (World Bank 1999a: 11). Participation by a broad range of stakeholders in preparing, implementing, and monitoring is set forth as a major requirement of the PRSP process; it is also a challenge for the transition from interim to full PRSPs (World Bank 2000n: 9, 12­13, 17-18). Recipient government representatives tend to have distinctly mixed views about development agencies inviting civil society groups to participate in aid coordination processes, especially if this is seen to be circumventing government (box 2.4).29 28. Of 209 CG meetings held and chaired by the Bank between 1992 and 1999, only 16, or about 8 percent, were held in the borrowing country. The 16 in-country meetings all have occurred since 1995, with the annual number reaching a peak of 7 in 1998 but dropping to 3 in 1999 (and increasing to just 4 or 5 in 2000). The September 2000 CG meeting for Bolivia, held in Paris, decided to hold subsequent meetings in Bolivia after donors gave assurances that they would continue to send senior representatives. The pros and cons of holding in-country CG meetings are discussed in the 1999 OED study (World Bank 1999a: 12­14). 29. For evidence, see the 1999 OED study (World Bank 1999a: 10­13). A similar picture is reported for the eight IDA pilot countries (World Bank 2000m: 16­17) and from the country consultations conducted for this review. 14 Box 2.4 Aid Coordination and Civil Society After considerable debate, participants --most of whom were IDA country officials --in a February 1999 workshop on aid coordination came to the following conclusions about involving civil society. · In spite of differences in the relative size, composition, and traditions of civil society in different countries, governments should be encouraged to experiment with different means of involving civil society in priority setting and aid coordination. · Involving civil society in the process involves both risks and opportunities. Risks include the potential for international manipulation and the effect of prolonged consultations on the timeliness of decisionmaking and implementation. Opportunities include consensus building that creates societal support for needed reforms, strengthened commitments to implementation, and increased transparency. · Donor engagement with local civil society should occur with the concurrence, presence, and participation of government, and not independent of it. · Participants also recommended that politicians as well as government officials be involved in the dialogue. Subsequent forums, including the December 2000 plenary meeting of the Strategic Partnership with Africa, have noted that coordination with parliaments has been weak (Strategic Partnership for Africa 2000a: 6). (For further discussion of this point, see the background paper on participation prepared for the current review.) Source: "Aid Coordination in an Era of Partnerships," unpublished proceedings of a workshop for aid coordination officials, February 10­11, 1999, pp. 10­11. The workshop was sponsored by OED in cooperation with the Swiss Agency for Development and Cooperation. 42. With leadership and support from IDA, the Strategic Partnership with Africa (SPA) has provided an important informal forum for coordination among bilateral and multilateral donors at the regional level. A comprehensive, independent evaluation of the partnership, commissioned by donors as part of a mid-term review in 1998, found that it was "unique in its focus on enhancing support for adjustment efforts and fostering the spread of improved donor practices" (World Bank 1998i: 107). The evaluation found that the partnership had made good progress in untying and streamlining procurement practices for import-support aid programs. It also highlighted the evolution of sector investment programs as "an interesting example of how the SPA mechanism works to develop a consensus and convert it into operational modalities" (World Bank 1998i: 10- 11). Among the evaluation's more critical findings were: (a) "little evidence that SPA conclusions directly informed [consultative group] or [roundtable] processes or local aid groups operating in recipient countries" and (b) "relatively marginal" African participation within the SPA process (World Bank 1998i: 107 and 115). These findings guided subsequent efforts to adapt the modus operandi of the partnership. Some additional background on SPA, including its role in tracking sector programs and monitoring the institutionalization of the PRSP, is highlighted in box 2.5.30 30. Box 2.5 is drawn from information provided by the secretariat of the Strategic Partnership with Africa, includinga presentation to the partnership's December 2000 plenary meeting, and from the evaluation cited in the main text (World Bank 1998i). The last observation on the PRSP country studies is drawn from the chairman's summary of the December 2000 plenary report (Strategic Partnership with Africa 2000b: 6). The partnership appears to use the term "sector program" in the same way that "sectorwide approach" is used in this paper. According to the plenary presentation, a sector program can encompass a continuum from project assistance to program support (Strategic Partnership with Africa 2000a: 24). 15 Box 2.5 Sector Programs and the Strategic Partnership with Africa The Strategic Partnership with Africa (SPA), originally known as the Special Program of Assistance for Africa, was launched in 1987 as a donor-coordinated response to the debt and development cris is in Africa. Chaired by the World Bank, it is an informal forum for donors and multilateral institutions involved with Africa. A primary focus of the partnership has been the mobilization of quick-disbursing resources in support of economic reforms, but it has also been a forum for sharing knowledge and experience to enhance the effectiveness and efficiency of overall resource use, including aid. In recent years the partnership has emphasized "sector programs." In 1999 SPA's members agreed to establish a mechanism for tracking sector programs; by mid 2000 information had been collected on 30 operations in 6 sectors involving 17 countries and 13 donors. A component of the tracking involves a quality assessment of aspects considered by stakeholders to be strong and weak. Strong aspects include design, sector priorities and objectives, ownership (stakeholder involvement), and dialogue quality. Weak or mixed aspects include budget management (especially at the decentralized level), implementation capacity, and monitoring and evaluation of results. SPA members see the agenda and trends emerging from the tracking as enhancing the partnership's purpose and goals. In May 2000, SPA and the Netherlands sponsored a workshop on donor harmonization and adjustment of financial management and control procedures in sector programs. The members of the partnership recognize the need to improve linkages with the PRSP process. They also realize that sector programs need to fit into and support PRSPs, which cannot substitute for sector programs (Strategic Partnership with Africa 2000a: 8). SPA recently launched country studies on institutionalization of the PRSP process. Preliminary results include findings of broad acceptance and government commitment, with weak institutional capacity--including for data collection and use--as a constraining factor. Source: World Bank 1998i, Strategic Partnership with Africa 2000a and 2000b. In consultation with UNDP, prepare for board review proposals for strengthening partnership 43. In November 1999, IDA staff prepared "Comprehensive Development Framework (CDF): Strengthening Partnership" as a response to this commitment (World Bank 1999c). IDA has sought to keep UNDP informed about the Partnership Initiative, the CDF, and the PRSP process, and to seek its views. The commitment also called for such proposals to take into account the then ongoing OED study on aid coordination and to "address the need for consultative meetings focused on government policy priorities, led as much as possible by the host government" (IDA 1998: 20). Although the totality of efforts has been impressive, IDA has not yet issued a revised aid coordination policy or a strategy in support of country-led mechanisms and processes. The preparation of full PRSPs is to be a participatory process involving development assistance agencies and civil society, with the recipient government in the lead. This should lead to meaningful country-led aid coordination, but the extent to which the PRSP process will realize this objective remains to be seen. 44. The four IDA12 commitments grouped under the "country-led partnership" heading suggest a potential tension between the IDA Deputies' desire to move to country-led and country- owned aid coordination processes, on the one hand, and a perceived need for IDA management to take a leadership role in aid coordination, on the other. A related tension is framed in the following terms by OED's 1999 Annual Review of Development Effectiveness: "How should donor demands for financial accountability be reconciled with client-led partnership, particularly when countries lack capacity?" (Hanna and others 1999: 20­21) 45. The challenge, as noted in the review, is to successfully manage the tension. In the present case, the challenge is to delineate a leadership role for IDA that is consistent with facilitating movement toward country-led partnership. This challenge will be revisited in the concluding chapter. 16 Special Issues 46. Coordination of public expenditure analyses and of financial assistance to postconflict countries were the subjects of two IDA12 commitments. Those subjects raise special challenges for aid coordination. Increase coordination on public expenditure analyses 47. This IDA12 commitment explicitly mentions "donors, the IMF, and the regional development banks" (IDA 1998: 27). Over the years, IMF staff members have occasionally joined PER missions, but for the most part, the Fund draws on Bank-led reviews.31 A recent OED assessment of impact, based on a sample of 35 reviews prepared between fiscal 1994 and fiscal 1998, found that few reviews solicited the views of other stakeholders or involved significant interaction with other donors prior to or during the review exercise, nor were many completed PERs widely disseminated to external stakeholders (World Bank 1998e: 16). A review of PERs prepared for IDA countries over the last three years yielded similar results, with a couple of significant exceptions.32 The South Asia Region of the Bank reports good examples of donor participation in recent years (for example, in Nepal and Pakistan) and now requires that all new PERs are to involve both other donors and the government in their preparation. 48. IDA has supported the Strategic Partnership with Africa in preparing "Guidance for Strengthening Public Finance Management in Sub-Saharan Africa" (Strategic Partnership with Africa 1999). This recent publication emphasizes the importance of improved coordination with other donors in the preparation of PERs, as well as the involvement of recipient governments and stakeholders. Bank guidance and reports on PRSPs have emphasized the importance of public expenditure analyses as key diagnostic tools to help estimate the cost implications of alternative poverty reduction measures (World Bank 2000a: 8­9, and 2000n:8, 18). The logic of the CDF principles and the PRSP process imply a need for close coordination of work on public expenditure analyses. 31. The so-called 1989 Concordat between the Bank and the Fund delineates a still-prevailing basic division of labor that assigns primary responsibility to the Bank for "the composition and appropriateness of development programs and ... priorities" (Development Committee 1998: 2­3). That phrase has been interpreted to include public expenditure analyses. A "Guidance Note for Bank-Fund Collaboration on Public Expenditure Issues" promulgated in 1995 by the President of the Bank and the Managing Director of the Fund identified various ways in which staffs of the two institutions should maintain and strengthen their collaboration in this area (World Bank­International Monetary Fund 1995). A joint paper on operational issues in PRSPs produced in December 1999 indicates that Bank staff will "take the lead on advising on how to improve the effectiveness and poverty-orientation of public expenditure (through Public Expenditure Reviews and the like)." (World Bank 1999e: 14). 32. By far the most extensive example of donor participation was the November 1999 Public Expenditure Review for Ethiopia, which reported participation of five other donors on the review team, including the IMF, European Union, African Development Bank, the Netherlands, and the United Kingdom, as well as workshops held in Addis Ababa and in Brussels. This effort followed the 1997 Public Expenditure Review for Ethiopia, which included six other donor representatives on the team. See annex 4 for further discussion. The above-mentioned OED assessment cites Ethiopia, Peru, Russia, and Uganda as exceptions to the general lack of consultation and coordination with other partners in the preparation of Public Expenditure Reviews (World Bank 1998e: 10, 16). 17 Work with international financial institutions and donors to develop a comprehensive approach to guide financial assistance to postconflict countries 49. IDA's efforts to develop and implement a strategy of assistance to postconflict countries were underway when this IDA12 commitment was negotiated.33 A Board-approved "Framework Paper" led to the establishment in 1997 of a postconflict unit that has provided support to Bank operations in many postconflict countries (World Bank 1998j). An operational policy statement, "Development Cooperation and Conflict," was approved in November 2000 after extensive Board review (World Bank 2000g). The drafting process involved extensive consultations with bilateral and multilateral donors as well as with civil society and NGOs. Reactions by these stakeholders to the statement and the consultation process have reportedly been positive.34 The statement also incorporates key recommendations of the 1998 OED review, "The World Bank's Experience with Post-Conflict Reconstruction" (Kreimer and others 1998). 50. The new policy statement is one of several efforts of IDA and other partners to develop a comprehensive approach for financial assistance to postconflict countries. These efforts include: · Following the April 1998 memorandum of understanding, the President of the World Bank and the United Nations High Commissioner for Refugees in 1999 chaired two meetings of donors to discuss the gap between humanitarian relief assistance and development assistance to postconflict countries. While this so-called Brookings Process has not resulted in the creation of a new "global fund" for postconflict countries, it has raised the awareness of donors and development agencies of the poor alignment that tends to prevail between agency budget mechanisms and the needs of postconflict countries.35 · The Bank's Development Research Department is undertaking research on conflict risk indicators and their implications for programming and policy toward a country at risk of conflict. This work should have important implications for a coordinated approach to assistance to postconflict countries and countries vulnerable to conflict or a resumption of conflict. · A third element deals with the allocation of IDA resources to postconflict countries. The outline of a proposal being prepared as a contribution to the IDA13 negotiations is summarized in box 2.6. 33. See IDA 1998: 14. The commitment further recommends that "lending and non-lending should be provided selectively to post-conflict countries, commensurate with government progress in implementing recovery policies and donor willingness to provide concerted support." 34. A total of 13 consultation meetings were held on the draft OP/BP in the United States, Europe, Africa, and Asia, between April and June 2000. Reactions are reported in the minutes of the June 12 CODE meeting (World Bank 2000k: 10­13). The Bank's Post-Conflict Unit has undertaken a comparative analysis in which the draft Operational Policy/Bank Procedure compares favorably with the 1997 guidelines on conflict, peace, and development of OECD's Development Assistance Committee. 35. Kuroda interview 2000. 18 Box 2.6 A New Approach to IDA Allocations to Postconflict Borrowers In the IDA13 negotiations, the Deputies will discuss a proposal to better align IDA allocations with the exceptional financial needs of postconflict IDA countries, while at the same time maintaining performance. The proposal delineates four aspects of an allocation process for postconflict borrowers: · Eligibility criteria to select relevant countries for exceptional treatment. · Re-engagement criteria to determine the size of the initial allocation. · Postconflict progress indicators to measure performance in postconflict countries eligible for exceptional treatment. · Limits on the duration and volume of exceptional allocation. The postconflict progress indicators would consist of a consolidated version of the Country Performance and Institutional Assessment (CPIA) indicators, refined and regrouped to address the realities of countries affected by conflict, and a new cluster of public security, reconciliation, and demobilization and disarmament indicators. Source: Draft unpublished proposal prepared by the Resource Mobilization Department, World Bank, October 2000. 51. The IDA12 Replenishment Report stresses the financial aspects of postconflict reconstruction, including eligibility criteria and arrears clearance or rescheduling (IDA 1998: 14). For some postconflict countries, trust funds have been established to which donor countries have contributed with the express purpose of clearing IDA arrears. This mechanism has been critical to the resumption of Bank operations in Bosnia and Herzegovina, Cambodia, Haiti, and Rwanda. "Rights Accumulation" and "Fifth Dimension" programs are other possibilities.36 But these approaches are not adequate for countries with large accumulated arrears, which would entail an unacceptably long interval before lending could resume. An alternative noted in the IDA12 commitments is to provide limited IDA grant financing prior to clearance of the arrears. But this is seen only as a "last resort and as part of a concerted international assistance effort to assure positive flows to postconflict countries as soon as performance warrants" (IDA 1998: 14). This alternative has yet to be employed. 52. Arrears to IDA typically occur in countries that are also in arrears to the IMF. Any approach to resolving arrears problems thus requires close coordination. The IMF provides limited flexibility for arrears clearance in postconflict countries, but only where the country is committed to reform and other international financial institutions show comparable flexibility. This circumscribed approach is justified by the IMF in terms of protecting against moral hazard and undermining the IMF's preferred creditor status (see IMF 1999a: 65­67 and IMF 1999b). Following a commitment made by the Bank's President and the Fund's Managing Director at the September 2000 annual meetings, a joint Bank-Fund working group has been established to develop options to address arrears problems faced by highly indebted, poor, postconflict countries and to report to the Spring 2001 joint meetings.37 36. Rights accumulation occurs when a borrowing country's allocation increases on the basis of a review of a "shadow" adjustment program, but no lending takes place until the arrears are, in effect, offset by the "rights accumulation." The Fifth Dimension program involves the extension of an IDA credit, covered by Bank income, equivalent to the interest payment due on an IBRD loan extended to an IDA country, with the understanding that the country will then make the required interest payment. 37. If a country is judged to have started an armed conflict against its citizens or against another country, the only option now available to IDA is to refrain from making new loans. Including a provision in future agreements to suspend disbursements in such circumstances would give IDA a potentially significant additional option. Although this suggestion is not directly related to aid coordination, such an option would be most effectively exercised in close coordination with other donors. 19 3 Lessons 53. Thanks largely to the CDF and PRSP initiatives, the principles of country ownership and partnership are informing IDA's strategies and priorities on numerous fronts. However, the implications of both principles for IDA's role in aid coordination have not been adequately spelled out, and implementation by key partners on the ground is spotty, varying considerably from country-to-country. At the corporate level, a new Operational Memorandum on aid coordination and partnership is in the works but has yet to be issued (replacing OP14.30 "Aid Coordination Groups", 1989). On the ground, genuine country-led partnerships, effectively combining ownership and partnership, are being realized in only a few IDA countries, typically in one or two sectors. The success of the PRSP will depend in part on the extent to which country- led partnership is realized. 54. The experience described in the previous section leads to the following overarching conclusion: Effective country-led partnership requires capacity and commitment in the recipient country, and acceptance of mutual responsibility and distinct accountabilities. These two elements are discussed in turn. Capacity and Commitment 55. The experience to date in preparing full PRSPs has reinforced the significance of weak capacity of the recipient country as a constraint to country-led partnership. The recipient's commitment to sound development priorities and policies within a strategic poverty reduction framework is also crucial. The 1999 OED study suggests a pattern of relationships between country capacity and commitment, on the one hand, and aid coordination arrangements on the other. Strong country commitment and high institutional capacity tend to be associated with country-driven arrangements, which should promote greater development effectiveness than arrangements driven predominately by donors. Conversely, the weaker the recipient country's commitment and institutional capacity, the more likely donors will insist on donor-driven aid coordination in the interests of development effectiveness. These hypothesized relationships, generally validated by empirical experience, are illustrated in figure 3.1 below (see World Bank 1999a: 6, 23­24). Figure 3.1 Country Characteristics and Aid Coordination Arrangements Country commitment Institutional capacity High Low High Country-driven Joint sponsorship Low Country-driven Donor-driven (with institutional strengthening) 56. The core of the PRSP approach is country leadership. One would therefore expect to find a successful PRSP case in the upper-left, "country-driven" quadrant of the figure. But, as noted in the previous section (paragraph 33), formulating a successful, concerted strategy to support the strengthening of institutional capacity has proven elusive. The 1999 OED study concluded that Bank leadership was required at the global or corporate level in order to work jointly with other partners to develop such a strategy (World Bank 1999a: 31). An issue that emerges from the country consultations is whether IDA is prepared to accept the timeframe implications of capacity building through country-led planning and design, as opposed to imposing its own schedule (for 20 example, a PRSP production schedule), even while giving rhetorical emphasis to "ownership." The 1999 OED study recognized that building capacity for country-led aid coordination would likely take five to ten years in most IDA countries (World Bank 1999a: 39).38 57. How can a country move to the country-driven quadrant of figure 3.1? Depending on its current situation, the answer may be to strengthen country commitment, strengthen capacity, or do both. But the characteristics of donors and international agencies also have a bearing on the relative ease or difficulty of moving to effective country-led aid coordination. When donors and their agencies give high priority to a development rationale for aid and are relatively few in number, the conditions for achieving country-led aid coordination are favorable. As the number of donors and projects increases, however, the environment for aid coordination becomes more challenging. Moreover, as nondevelopment motives for providing aid--notably the commercial interests of donors--become dominant, effective country-led coordination becomes problematic. While the end of the Cold War has reduced the dominance of the ideological motive for providing aid, geopolitical considerations are still at work.39 Figure 3.2 illustrates the influence of development assistance characteristics (drawn from World Bank 1999a: 8). Figure 3.2 Development Assistance Characteristics and the Aid Coordination Environment Development orientation Number of donors High Low Low Favorable Relatively difficult High Challenging Very difficult Mutual Responsibility and Distinct Accountabilities 58. These considerations suggest that progress toward country-led partnership will require acceptance by the parties of mutual responsibility and distinct accountabilities. The recipient country and development agencies must accept mutual responsibility for achieving development outcomes, while each set of partners agrees to be held accountable for distinct tasks. · Development agencies must be accountable for delivering high-quality assistance that is consistent with country priorities, selective along lines of comparative advantage, efficient, and complementary to country-led efforts to build capacity. 38. The 1999 OED study also pointed out that institutional capacity needs to be "unbundled" or separated into three major components: organizational capability, internal and external rules, and incentive structures (organizational values and reward systems). Merely providing training and technical assistance, for example, will not be effective if the real problem is inadequate or perverse rules and incentive structures, as when promotion and perquisites are awarded solely on the basis of loyalty to supervisors, spending or cutting the budget, or other input -related criteria, as opposed to results-based criteria. Disincentives to more effective aid coordination range from issues of corruption and transparency to power dynamics and conflicts over priorities. They affect recipient countries and donors alike, complicating effective aid coordination (World Bank 1999a: 7, 34). 39. Another donor characteristic is diversity and complexity of procedural and accountability requirements, which can also impede greater country leadership. However, the greater the development orientation, the greater the willingness of donors to harmonize procedural requirements. The same comment applies to the internal incentive structures of donor agencies that mitigate against coordination. Several replies to a joint questionnaire sent by UNDP and the Bank to donors mentioned all these factors as potential barriers to effective aid coordination. 21 · Recipient countries must be accountable for developing sound policies and effective institutions supported by transparent and accountable systems of financial resource management. Quality aid delivery 59. While IDA's coordination efforts have resulted in increased attention to coherence and selectivity at the country and corporate levels, implementation challenges at the country level and gaps at the corporate level remain. 60. IDA has made significant progress over the last seven years. It has sought to increase the coherence and selectivity of its assistance relative to that of other development agencies. Its efforts have ranged from information sharing and consultation to strategic coordination, and to activities undertaken jointly with other partners. At the same time, IDA10­12 commitments related to aid coordination have moved progressively from information sharing and consultation to strategic coordination. 1) IDA's performance in information sharing and consultation has improved at both country and corporate levels. However, complaints of inadequate consultation at the country level persist. Decentralization of authority to the field has tended to improve the situation. Several relevant corporate level agreements with United Nations (UN) agencies have been concluded or are under way in connection with the PRSP initiative. 2) Country Assistance Strategies (CASs) and related documents provide evidence of increased strategic coordination at the country level, as evidenced by more attention to issues of programmatic selectivity and comparative advantage. However, evidence from the countries piloting the Comprehensive Development Framework (CDF) suggests that substantial progress toward selectivity on the ground will require strong headquarters support from all development agencies with significant portfolios in a country. 3) At the IDA corporate level, the emerging sector strategy papers in principle provide a framework for intrasectoralselectivity, but the task of bringing about consistency between what is said about selectivity in CASs and what is said about selectivity in the sector strategy papers is just beginning. The issue of intersectoral selectivity--or comparative advantage--at the institutional level has yet to be addressed. 61. The Poverty Reduction Strategy Paper (PRSP) initiative has led to an unprecedented intensity of coordination between IDA and the IMF involving consultation, strategic coordination, and joint work. The ultimate test of this coordination will be continued evidence of consistent views on what constitutes a coherent poverty reduction strategy. 62. The preparation of key analytical documents, such as Public Expenditure Reviews (PERs), with a few exceptions, has involved relatively limited consultation with, or participation by, other development agencies. The PRSP initiative may lead to increased coordination of these efforts. 63. IDA has also made impressive progress toward procedural harmonization with the other multilateral development banks. A more difficult challenge for IDA, given the divergent authorizing environments involved, is that of promoting greater procedural harmonization with other development agencies, including bilateral donors and UN agencies. 64. The 1999 OED study found that disparate and burdensome donor policies, procedures, and practices generated the most concern among recipient governments. The list ranges from 22 contradictory conditions to divergent policies on local cost financing, budget and procurement procedures, reporting and audit requirements, and a plethora of uncoordinated visiting missions.40 Harmonization was singled out as the first thing donors should do to improve aid coordination. At the same time, government and donor respondents to surveys were pessimistic about the progress and promise of existing aid coordination efforts (World Bank 1999a: 15). 65. Progress toward harmonization is being made by some donors in some sectors and in some countries, particularly through sectorwide approaches.41 Further progress could be made on a country-by-country basis by adopting the approach to accountability suggested above. However, as donors at the country level have noted (box 3.1), this is an issue that must also be tackled at the corporate level. The Bank has made impressive progress with its multilateral lending partners in harmonizing certain procedures. As part of its global public policy role, the Bank should support harmonization by bilateral donors through dialogue with them at headquarters levels and in such forums as the OECD Development Assistance Committee. A significant initiative by the Bank was to host a conference of representatives of multilateral agencies and the Chair of the OECD- DAC Bilateral Task Force on Donor Practices on harmonization issues on February 1, 2001. Box 3.1 Donor Harmonization Required at the Country Level and the Corporate Level Consultations in selected IDA countries yielded the following observations on the need for donor harmonization. · Government officials across countries lamented the high transaction costs of disparate donor reporting requirements and procedures; many donors agreed that the need for harmonization of procedures is acute, especially in the context of sectorwide approaches. · One observer in Bangladesh noted that "the system is trying to solve these issues on a case-by-case basis... [but] they can only be adequately addressed on the institutional level." · Respondents from the Kyrgyz Republic, Bolivia, and Ghana held that many of the constraints to better partnership, such as procedural requirements and accountability standards, cannot be addressed at the country level and argued that there was insufficient coordination on these issues at the institutional level among donor agencies. Source: Country consultations conducted for the OED IDA review. Country transparency and accountability 66. The support of IDA and other development agencies for building borrowers' capacity to coordinate aid has been weak. Establishing transparent and accountable financial management systems is the responsibility of the recipient country, but development agencies, including IDA, have an obligation to help strengthen the country's capacity for financial management in the interest of aid effectiveness and country-led coordination. There are few positive examples to draw from, but several negative lessons have emerged from IDA experience, such as avoiding supply-driven practices of technical assistance, and project implementation units (PIUs) which tend to undermine country capacity. 40. The harmonization "net" covers a wide range. In addition to the procedural requirements and practices mentioned above, divergent policies dealing with safeguards can also create an unnecessary burden for recipient countries and a challenge for aid coordination. Safeguard policies are discussed in detail in the environment background paper prepared for this review. 41. See the discussion of sectorwide approaches in Ethiopia and Ghana in the 1999 OED study (World Bank 1999a: 18­19). Also see the mixed picture of limited harmonization progress at the country level in the eight IDA countries in the CDF pilot as summarized in the recent report on country experience with CDF (World Bank 2000m: 20). 23 Aid Coordination Mechanisms and Processes 67. IDA-led Consultative Groups (CGs) and growing local coordination forums have moved in the direction of country-led partnership, but challenges and tensions remain, including high transaction costs. One result of improved aid coordination, with the country in the lead, should be a reduction in the transaction costs of aid delivery. However, more intensive aid coordination efforts themselves entail transaction costs. The participants in CDF pilots and sectorwide approaches clearly experience these costs, but they believe they will decline and that benefits will become more apparent (box 3.2). Box 3.2 High Transaction Costs Expected to Decline The country consultations undertaken for this review yielded the following observations on the subject of transaction costs. Donors and some government officials in Ghana, Bolivia, and Vietnam held that "the inputs to the process are heavy now because we are in the start-up phase" and that "despite the steep initial costs ... transaction costs will eventually decline and benefits will become more apparent." In the short run, some Bolivian officials maintain that the CDF has "slowed things down," making aspirations to establish ownership and partnership even more difficult, and that "the CDF has not been internalized fully by the government, IDA, or other donors." Donors in Vietnam agreed that if the CDF were to be successful, it could greatly improve the effectiveness of aid; if it were not well managed in the future, however, the CDF could result in excessive meetings of low productivity. Source: Country consultations conducted for the OED IDA review. 68. The 1999 OED study suggests a rough sequence of aid coordination efforts (figure 3.3) in which measures with relatively lower costs to donors but high benefits to countries would be tackled first. Those measures include alignment with sector strategies, common reporting procedures, and joint missions. Measures of great benefit to countries but tougher and more costly for donors (such as common procurement procedures) would be tackled subsequently (World Bank 1999a: 20­21). As local groups become more focused on coordinated efforts involving agencies committed to a development orientation, participants should start to perceive benefits that exceed the costs of coordination efforts. This is happening in some of the more successful sectorwide approaches, such as in health in Ghana. 24 Figure 3.3 Relative Benefits and Costs of Aid Coordination Efforts Benefits to recipients Costs to donors Low High High · Untying technical assistance · Untying projects · Untying import aid · Unearmarked contributions to common fund · Common procurement and disbursement procedures Low · Better information exchange · Increased selectivity (comparative advantage) · Joint missions · Earmarked common fund contributions · Common reporting procedures · Increased coherence with country priorities · Common budget cycles The Dual Implications of Sectorwide Approaches 69. Sectorwide approaches seem to have dropped out of the corporate lexicon of IDA. At the country level, however, IDA and other development agencies continue to view them as a way to combine country ownership with stronger donor coordination. Sectorwide approaches continue to be pursued in several Sub-Saharan African countries, as well as in Bangladesh and more recently in Cambodia and Vietnam. By moving away from the typical project-based accountability paradigm, sectorwide approaches also can promote capacity building and relationships of mutual responsibility and distinct accountabilities. To the extent that development agencies work directly with a sectoral ministry rather than through individual PIUs, capacity building is more likely to occur. In such circumstances, the accountability shifts from one between a PIU and a donor to one between development agencies and the recipient government, and between government and citizens.42 70. Experience shows that several conditions must be met if sectorwide approaches are to realize these objectives. In addition to overarching preconditions such as macro-level stability and a clear sector strategy, a minimum core of donors must be prepared to support the strategy and to harmonize their procedures sufficiently for the sectorwide approach to yield added value. Notwithstanding the available literature and documentation on sectorwide approaches, there is little rigorous evidence of their effectiveness in terms of development outcomes and impacts, as compared with traditional project approaches (Garner and others 2000). What difference have they made to development outcomes on the ground? Have sectoral patterns emerged?43 It is in IDA's interest and within its capacity to pursue these questions. 42. These implications are elaborated in a recent paper prepared for the Canadian International Development Agency by Mark Schacter, Sector Wide Approaches, Accountability and CIDA: Issues and Recommendations. (Schacter 2000: i­iii, 24). The author also develops the implications of sectorwide approaches for the accountability relationship between the donor agency and its own government and public. 43. The apparent lack of effectiveness of sectorwide approaches in agriculture has been ascribed to divergent views regardingappropriate public and private sector roles (Jones in World Bank 1999b: 32, note 20). 25 Implications of Country-Led Partnership for Conditionality and Monitoring 71. Making specified policy changes a condition for disbursement of structural adjustment credits is the main way in which IDA has sought policy reform in recipient countries. But policy changes and reforms brought about by conditionality will not necessarily be sustained.44 Such key CDF and PRSP principles as country ownership, participation, partnership, and a long-run orientation to results arguably require a new approach to conditionality. 72. A paper prepared as background for the OED 1999 Annual Review of Development Effectiveness foresaw that successful implementation of the CDF required a new approach to conditionality. The paper characterized the traditional approach as "a blueprint package of policy conditions attached to a tranche release of single loans" and often as an attempt to "force externally designed policy changes on unwilling governments" (Branson and Hanna 1999: 1). The authors discuss several variants of a new approach built around a concept of mutual commitment to a policy compact by the Bank, its partners, and the recipient country. Key features of conditionality built around mutual commitment would include: · Viewing conditionality as part of an evolving process rather than a "one-off event." · Building commitment for the broad directions but allowing flexibility and adaptation in implementation. · Moving toward progressively greater flexibility for a "credible performer," including government choice of sequence of reforms and tranche release. · Involving other stakeholders, so as to reduce the risks of coercion inherent in "secret negotiations" and minimize the risks of shirking or policy reversals.45 73. The PRSP for Burkina Faso reports on a relevant effort, known as the "Conditionality Reformulation Test Exercise," that was initiated in 1997 in conjunction with the Special Program of Assistance for Africa. That effort, coordinated by the European Community, has brought together government, donors, and lenders to agree on measurable performance indicators for key sectors. Once mutual agreement exists on sector strategies, targets, and indicators, the government will assume control for the means of implementation, and for the pace and sequence of reforms. The indicators are to be used as the basis for decisions regarding aid disbursement, "preferably in the form of budget support" (World Bank 2000h: 58). 74. Implicit in the approach to conditionality suggested above is the requirement that partners regularly monitor, discuss, and review performance, as well as agree on adjustments as needed (Branson and Hanna 1999: 6). The framework of mutual responsibility and distinct accountabilities implies that the agenda for such sessions should be reciprocal in nature. Time should be set aside for the participants to discuss donors' performance in the delivery of 44. Johnson and Wasty identify four factors affecting degree of "ownership" of structural adjustment program reforms: (1) the borrower as locus of initiative for program design and implementation; (2) level of intellectual conviction among key policymakers; (3) expression of political will by top leadership; and (4) efforts toward consensus-building among various constituencies (Johnson and Wasty 1993: 5). Also see Assessing Aid: What Works, What Doesn't, and Why (World Bank 1998g: 50­53). 45. Branson and Hanna 1999: 1­2, 7­9. The authors draw evidence from an empirical analysis of results under the Higher Impact Adjustment Lending initiative in Sub-Saharan African countries. They find that countries adopting the initiative tend to be associated with more positive policy outcomes in a number of areas than do other Sub-Saharan countries (Branson and Hanna 1999: annexes 1 and 2). 26 assistance, and for discussion of country performance.46 This also implies that the partners should agree on objectives and benchmarks for enhanced donor performance, in parallel with the approach for assessing country performance.Pursuant to a recommendation of the June 2000 Tanzania CG meeting, an independent panel is being established to assess donor performance in providing quality aid delivery to Tanzania. 46. The meetings convened to review progress under the Marshall Plan of assistance to Western Europe after World War II included an agenda item for discussion of donor performance. 27 4 Recommendations 75. The following recommendations can be implemented for the most part by IDA itself and within its existing policy framework. Several recommendations that would require IDA to act in concert with its development agency partners are highlighted as special challenges. Several of these recommendations would benefit from being incorporated into commitments under IDA13. Implementation Challenges for IDA 76. The following recommendations represent challenges that IDA can implement within its existing policy framework. 1) Promote country-led aid coordination mechanisms and processes. In promoting country-led initiatives, IDA needs to be sensitive to the transaction costs that such initiatives can impose on the country and on development agencies. Overlapping groups should be avoided and, as appropriate, a lead "support donor" identified for each local group. Experience shows that adequately staffed and empowered country offices enhance IDA's role in supporting country- led aid coordination. Meeting this challenge requires adequate budgetary support. (See recommendations in paragraph 77.) The same conclusion applies to other development agencies that would assume a lead role in coordinating support: They must be prepared to commit sufficient professionaland budgetary resources to perform the role effectively. 2) Continue to strengthen and refine sectorwide approaches, incorporating the lessons of experience. Sectorwide approaches continue to be initiated and supported in the field by IDA and other development agencies. They have the potential to strengthen both country-led partnership and donor coordination, and at the same time to reinforce country capacity and shift the aid paradigm to one of mutual responsibility and distinct accountabilities. As IDA moves to more comprehensive programmatic support under the Poverty Reduction Strategy Papers (PRSPs), sectorwide approaches will continue to be an essential building block. 3) Develop and implement methods for monitoring and evaluating progress in country-led partnerships, including IDA's performance under the PRSP. Monitoring is essential, both to report on progress and to identify constraints to achieving it. IDA, working with its partners, should develop collaborative mechanisms and key indicators that can be used to monitor progress toward country-led partnerships for the management and coordination of aid resources. Challenges for IDA and Its Development Partners 77. The IDA Deputies have a dual role that is particularly important with respect to these recommendations, which have implications for the development agencies with which IDA works. The Deputies are in a unique position to convey the recommendations to donor agencies in their own countries as well as to IDA. The Deputies should: 1) Pursue programmatic selectivity at the country level and push the envelope at the corporate level. This challenge includes ensuring that selectivity proposed at the country level in CASs is consistent with what is set out at the corporate level in sector strategy papers and other documents. By the same token, the sector strategy process itself needs to take better account of resource and country considerations. Beyond the sectoral dimension, IDA needs to define more sharply its comparative advantage as an institution and carry through on the implications for its programs at the sector and country levels. 28 2) Ask IDA to work toward aid relationships in which all partners accept mutual responsibility for development outcomes, assume their distinct accountabilities, and review the performance of each partner. A promising approach is being pursued by the Tanzania Consultative Group (CG), which has established an independent review panel to assess donor performance and report its findings. A challenge for IDA in implementing this recommendation will be exercising leadership in concert with its partners without dominating or being perceived as dominating the process. This modus operandiis important even in cases where low country capacity and commitment require donor-led coordination. An effective partnership also requires that weaker partners be strengthened, so as to level the playing field. IDA should therefore formulate a joint strategy with other key partners to strengthen country capacity to manage and coordinate aid resources. The experience of the Partnership for Capacity Building in Africa and other relevant initiatives should be drawn upon in formulating the strategy. 3) Endorse a global public policy role for IDA regarding donor harmonization. The forces working against donor harmonization are strong. The World Bank, as a preeminent multilateral development institution, is in a unique position to exercise global leadership in working with other development assistance agencies toward meaningful harmonization. The harmonization agenda goes beyond a wide range of policy and procedural issues--important as these are--to critical policy and strategic issues such as selectivity and performance-based allocation. These challenges need to be pursued at both the country and corporate levels, where the United Nations (UN) agencies need to be included as well. This is not just an issue for IDA, but one that needs the support of all development agencies, taking into account the policies, procedures, and capacities of client countries. IDA should convene a conference of the leaders of multilateral and bilateral development assistance organizations to seek agreement on the major harmonization issues that impede country-led partnership and agree on a program of action for harmonization.47 4) Provide adequate resources to promote country-led partnership. Assuming the leadership roles described above will take dedicated allocations of IDA budget and staff--in resident missions as well as at headquarters. It also will require a realignment of staff and management incentive structures, to recognize and reward adequate attention to the needs of partnership is recognized and rewarded. Without such efforts, the opportunity provided by the Comprehensive Development Framework and PRSPs to put into practice the lessons of 50 years of development cooperation is likely to be lost. 47. Such a conference would acknowledge and take into account ongoing harmonization efforts, including the meetings of the presidents of the multilateral development banks and related forums and working groups, the informal group of multilateral secretariats convened by the chair of OECD's Development Assistance Committee, and the Development Assistance Committee's Task Force on Donor Practices. 29 ANNEX 1: IMPLEMENTATION MATRIX FOR IDA10­12 UNDERTAKINGS RELATED TO AID COORDINATION Introduction The IDA12 Replenishment Report requested OED to undertake an independent review of the IDA program during the IDA10-11 period and an interim review of IDA12. The Review concentrates on IDA's development contribution in six thematic development priorities: 1) poverty reduction, 2) social sector development, 3) private sector development, 4) governance, 5) environmentally sustainable development, and 6) gender. It also addresses four priority process reform objectives: 1) performance based allocations, 2) enhanced CAS design and implementation, 3) aid coordination, and 4) participation. The following matrix lists the main undertakings of the IDA10,11 and 12 Replenishment Reports related to Aid Coordination and presents comments on the extent of IDA's implementation. The comments focus on actions taken by IDA during the respective replenishment period and, where appropriate, give a sense of subsequent or on-going actions. The matrix reports on the extent of compliance, not effectiveness in terms of outputs or outcomes from IDA's actions. The wide variation in the nature of the undertakings ­ranging from encouragement of broad redirections in areas of operations to specific calls for reports ­ created a challenge for arriving at aggregate ratings. The findings on IDA's degree of compliance as reported in the IDA Review's report and this background study made use of a system of ratings on individual undertakings, reviewed with management. This system also served as input into the substantive discussion of implementation that is summarized in the text in this Annex. 30 Undertakings Comments Strengthen Coordination with Development Assistance Agencies IDA has significantly strengthened its coordination with other agencies over the IDA10­12 period. With bilateral donors most progress has come in the field. In the case of United Nations (UN) agencies, other multilateral development banks, and the International Monetary Fund (IMF), by contrast, improvements have been achieved largely at the headquarters level, while experience in the field is mixed. IDA10: Report on any new IDA aid coordination activities in IDA10 A special section on aid coordination activities (including local aid coordination efforts) was included in implementation reviews. the IDA10 implementation report. IDA10: Strengthen country-level coordination with bilateral aid agencies.IDA's More frequent placement of IDA country directors in the field has reinforced in-country coordination, as resident representatives should do more to ensure that donors are kept has greater use of in-country Consultative Groups (CGs) and other local meetings involving donors and informed between missions. government. However, complaints persist about IDA's inadequate consultation with other donors at the country level, lack of transparency and timeliness in sharing information, and inadequate feedback. IDA10: Review coordination with the UN system. Explore greater use of UN Consultation and coordination between IDA and UN agencies at the corporate level has intensified skills and capacity, especially in human resource and institutional development. during IDA10­12. After initial tensions, the Comprehensive Development Framework (CDF) and Increase cooperation with the new UN Commission on Sustainable Poverty Reduction Strategy Papers (PRSP) initiatives are resulting in closer coordination in the field. Development. Coordination with the UN Commission on Sustainable Development was particularly extensive during IDA10 and IDA11. IDA10: Draw on the work of OECD committees in identifying priorities for aid With the Partnership Initiative, CDF, PRSP, and 1999 OED study, collaboration on aid coordination with coordination. the Development Assistance Committee of the Organisation for Economic Co-operation and Development (OECD) has intensified. IDA10: Explore supplementing existing regional approaches in the Caribbean Pilot regional CGs in Central America were dropped in the mid-1990s for lack of interest, but were and Central America. recently reintroduced by the Inter-American Development Bank. The Caribbean Group for Cooperation and Economic Development functions more as a regional policy dialogue group than as a means of strategic coordination. Strategic coordination tends to occur more at country and subregional levels. IDA12: Joint World Bank­International Monetary Fund (IMF) missions should The PRSP has been accompanied by a series of IDA -IMF joint policy statements and guidelines, as be increased and coordination strengthened. well as an expansion in joint committees, working groups, and joint or closely coordinated missions. However, reports are mixed regarding the quality of IDA -IMF field coordination on PRSPs. 31 Undertking Comments IDA12: (a) High priority should be given to collaboration with regional IDA has given high priority to increased harmonization of procurement, evaluation, and other development banks on harmonization of evaluation procedures and on procedures among multilateral development banks--and it is achieving solid results. IDA and the strengthening and harmonizing procurement rules and documents. African Development Bank concluded a memorandum of understanding in March 2000 that provided (b) Concessional assistance flows from multilateral development banks to some clarification of respective roles. A mechanism to monitor and assess implementation of the Africa should be coordinated to ensure optimal development impact. understanding is being established. The harmonization issue is now on the agenda of the Development Committee. (c) A memorandum of understanding should be prepared to clarify the roles of the World Bank and the African Development Bank. Increase Programmatic Selectivity IDA management has clearly conveyed to staff the need to addressstrategic selectivity at the country IDA11: In planning the Country Assistance Strategy (CAS), IDA should consult level via guidance in CASs and sector strategy papers. However, full on-the-ground compliance on with main donors in order to maximize complementarity and effectiveness of selectivity also requires compliance through other partners with the requisite finance, capacity, and donor assistance. commitment. IDA12: IDA should work closely with other donors in the social sectors, taking Only in the last two years have CASs begun to indicate clearly what IDA will and will not do based on its into account their assistance efforts and acknowledging their areas of emphasis comparative advantage. A recent report on CDF pilot country experiences notes that progress in and comparative advantage. achieving greater selectivity on the ground has been limited. IDA12: CASs should include a thorough discussion of IDA's coordination and Sector strategy papers are expected to explain how the Bank's lending and nonlending products relate collaboration with external partners and a table of current and planned donor to those of other international financial institutions. The task of achieving consistency between the aid that identifies those sectors and activities in which IDA will take the lead discussions of selectivity in CASs and sector strategy papers has begun in the health, nutrition, and and those in which other donors will take the lead. population sector, but all other major sectors remain to be covered. The issue of intersectoral selectivity --or IDA's comparative advantage at the corporate level--has yet to be addressed. Move to Country-Led Partnership Propelled by the CDF and PRSP initiatives, the principles of country ownership and partnership are now informing IDA's strategies and priorities on numerous fronts. A new Operational Memorandum on partnership and aid coordination is being prepared. The move to effective country-led partnerships requires the reform of long-standing practices and is being realized in only a few IDA countries, typically in one or two sectors. 32 Undertaking Comments IDA12: Through capacity building, ensure that countries can assume the The revised draft policy statement on partnership contains general language on capacity building. The central role in coordinating aid. IDA-supported Partnership for Capacity Building in Africa seeks to apply the lessons of past support for capacity building, which typically has been poor. IDA12: Continue to explore practical mechanisms that effectively combine IDA has participated in sectorwide approaches in several African countries and, to a lesser extent, in increased recipient country ownership with improved donor coordination. Asia. Sectorwide approaches are seen as a practical way of increasing country ownership and improving donor coordination. Experience shows that their effectiveness depends on the fulfillment of several prerequisites. IDA12: Management should take a leadership role in aid coordination, IDA-led CGs and expanded local coordination forums have moved in the direction of country-led exploring ways to strengthen the role of CGs and to increase coordination and partnerships. Although there has not been a major shift of consultative group meetings to recipient involvement of civil society and nongovernmental organizations (NGOs) in the countries, movement in that direction is discernible. Among the challenges to be overcome are the consultation process. One such method is to hold group meetings in the initially high transaction costs of expanded efforts and finding more effective ways of involving NGOs recipient country. and civil society. IDA has exercised major leadership in bringing about improved donor coordination at the regional level through the Strategic Partnership with Africa, but the role of recipient countries in the partnership needs to be expanded. IDA12: Management, in consultation with the United Nations Development The totality of efforts under the Partnership, CDF, and PRSP initiatives is impressive. However, IDA has Programme (UNDP), should prepare for Board review proposals for not yet issued a revised aid coordination policy or a strategy in support of country-led mechanisms and strengthening partnership, taking into account OED's ongoing evaluation of processes. The preparation of full PRSPs is designed to be a participatory process, involving CGs. Proposals should address the need for in-country policy dialogue and development assistance agencies and civil society, with the recipient government in the lead. The consultative meetings focused on government policy priorities, led as much as process should lead to meaningful country-led aid coordination, but the extent to which it will do so possible by the host government. remains to be seen. Address Special Issues IDA12: Coordination with donors, the IMF, and regional development banks Recent assessments of PERs indicate that in most cases coordination is not yet present. However, should be strengthened in the preparation of Public Expenditure Reviews IDA, acting as the secretariat for the Strategic Partnership with Africa, has issued guidance applicable (PERs) and other expenditure analyses. to the Africa Region, calling for improved coordination with other donors in the preparation of PERs. One Bank region (South Asia) has called for all reviews to involve other donors and governments. The logic of the PRSP process implies a need for close coordination of work on public expenditure analyses, tracking, and related country capacity building. 33 Undertakings Comments IDA12: The World Bank and IMF, in cooperation with other major creditors, are IDA has taken several actions that represent a high degree of compliance with this commitment. Chief encouraged to continue development of a comprehensive approach to assisting among those actions are the following: post-conflict countries. Such an approach would include refinement of eligibility criteria and timely use of available financial instruments. · After extended internal and external reviews, a new operational policy statement, "Development Cooperation and Conflict," has been approved (Operational Policy/Bank Procedure 2.30). · A forthcoming paper, "Determining IDA Allocations and Monitoring Progress in Countries Affected by Conflict," will be shared with IDA Deputies and Bank Executive Directors early in 2001. Among other things, the paper proposes a mechanism to better align IDA allocations with the exceptional financing needs that often face postconflict countries, and eligibility criteria for obtaining such financing. · A joint Bank-Fund working group has been established to develop options to address arrears problems faced by poor, highly indebted, postconflict countries and to report those options to the joint Bank-Fund meetings in spring 2001. Source: For a more detailed analysis on which this annex is based, see Chapters 2 and 3 of the main text. 34 Annex 2 References African Development Bank and the World Bank. 2000. "Memorandum of Understanding." March 14. Abidjan and Washington, D.C. Asian Development Bank and the World Bank. 2000. "ADB-WB Cooperation: Framework and Structure." January 28. Manila and Washington, D.C. Branson, William, and Nagy Hanna. 1999. "Ownership and Conditionality." Unpublished background paper for the 1999 Annual Review of Development Effectiveness. Operations Evaluation Department, World Bank, Washington, D.C. Development Committee (of the Boards of Governors of the International Monetary Fund and World Bank). 1996. Serving a Changing World--Report of the Task Force on Multilateral Development Banks. Washington, D.C.: International Monetary Fund and World Bank. ______. 1998. "Report of the Managing Director and the President on Bank-Fund Collaboration." Development Committee document DC/98-21, September 29. Garner, Paul, Walter Flores, and Shenglan Tang. 2000. "Sector-wide Approaches in Developing Countries." British Journal of Medicine 321 (July 15). International Development Association. 1993. "Additions to IDA Resources: Tenth Replenishment." IDA document IDA/SecM93-80, March 1. World Bank, Washington, D.C. ______. 1996. "Additions to IDA Resources: Eleventh Replenishment." Internal report. World Bank, Washington, D.C. ______. 1998. "Additions to IDA Resources: Twelfth Replenishment--A Partnership for Poverty Reduction." Internal report dated December 23. World Bank, Washington, D.C. International Monetary Fund. 1999a. Annual Report 1999. Washington, D.C. ______. 1999b. "Fund Assistance to Post-Conflict Countries." Internal report dated March 19. Washington, D.C. Hanna, Nagy, Ramgopal Agarwala, William Battaile, Deepa Chakrapani, Ruchira Corcoran, and Federico Mini. 1999. 1999 Annual Review of Development Effectiveness. Operations Evaluation Department, World Bank, Washington, D.C. Johnson, John H., and Sulaiman Wasty. 1993. Borrower Ownership of Adjustment Programs and the Political Economy of Reform. World Bank Discussion Paper 199. World Bank, Washington, D.C. Kreimer, Alcira, John Eriksson, Robert Muscat, Margaret Arnold, and Colin Scott. 1998. "The World Bank's Experience with Post-Conflict Reconstruction." Operations Evaluation Department, World Bank, Washington, D.C. Landau, Luis. 1998. "Rebuilding the Mozambique Economy--Country Assistance Review." Operations Evaluation Department, World Bank, Washington, D.C. 35 Organisation for Economic Co-operation and Development. 1986. "Guiding Principles for Aid Co-ordination with Developing Countries." Internal report. OECD Development Assistance Committee, Paris. ______. 1996. Shaping the 21st Century: The Contribution of Development Co-Operation. Paris. ______. 2000. "Strategic Monitoring of Key Changes in Agency Procedures and Behaviour Needed to Implement Country-led Partnerships." Memorandum from chairman of the Informal Group of Multilateral Secretariats dated October 30. OECD Development Assistance Committee, Paris. Picciotto, Robert. 1998. "The Logic of Partnership: A Development Perspective." Internal report dated September 29. Operations Evaluation Department, World Bank, Washington, D.C. Schacter, Mark. 2000. "Sector Wide Approaches, Accountability and CIDA: Issues and Recommendations." Memorandum prepared for Policy Branch, Canadian International Development Agency, October. Institute on Governance, Ottawa. Strategic Partnership with Africa. 1999. "Guidance for Strengthening Public Sector Management in Sub-Saharan Africa." World Bank, Africa Region, Washington, D.C. ______. 2000a. "Chairman's Summary." Delivered at plenary meeting of Strategic Partnership with Africa, Paris, France, December 5­7. Africa Region, World Bank, Washington, D.C. ______. 2000b. "Tracking of Sector Programs in Africa 2000: Progress Report." Paper presented at plenary meeting of Strategic Partnership with Africa, Paris, France, December 5­7. Africa Region, World Bank, Washington, D.C. Team of Independent Experts. 1999. "A Study on Strategic Partnership Between the World Bank and the African Development Bank." Internal report, July. Washington, D.C. Wolfensohn, James D. 1999. "A Proposal for a Comprehensive Development Framework." Discussion draft, January 21. World Bank, Washington, D.C. World Bank. 1997. IDA in Action 1993­1996. The Pursuit of Sustained Poverty Reduction. Washington, D.C.: World Bank. ______. 1998a. "Country Assistance Strategies: Retrospective and Outlook." Internal report dated March 30. Operations Policy and Strategy Vice Presidency, World Bank, Washington, D.C. ______. 1998b. "Partnership for Development: Proposed Actions for the World Bank: A Discussion Paper." Board document SecM98-421, May 21. Partnerships Group, Strategy and Resource Management, World Bank, Washington, D.C. ______. 1998c. "Country Assistance Strategy for the Republic of Albania." Report 18161-ALB. July 8. World Bank, Washington, D.C. ______. 1998e. "The Impact of Public Expenditure Reviews: An Evaluation." Report 18573. November 13. Operations Evaluation Department, World Bank, Washington, D.C. ______. 1998f. "Country Assistance Evaluation--Albania." Report 18030. Operations Evaluation Department, World Bank, Washington, D.C. 36 ______. 1998g. Assessing Aid: What Works, What Doesn't, and Why. A World Bank Policy Research Report. New York: Oxford University Press. ______. 1998h. "A Framework for Cooperation between United Nations High Commissioner for Refugees and the World Bank." Letter dated April 17 and annex signed by James D. Wolfensohn and Sadako Ogata. Washington, D.C. ______. 1998i. "The Special Program of Assistance for Africa (SPA): An Independent Evaluation." Operations Evaluation Department, World Bank, Washington, D.C. ______. 1998j. Post-Conflict Reconstruction--The Role of the World Bank. Washington, D.C. ______. 1999a. "Review of Aid Coordination and the Role of the World Bank." Report 19840. Operations Evaluation Department, World Bank, Washington, D.C. ______. 1999b. "The Drive to Partnership: Aid Coordination and the Role of the World Bank." Operations Evaluation Department, World Bank, Washington, D.C. ______. 1999c. "Comprehensive Development Framework (CDF): Strengthening Partnership." IDA document IDA/SecM99-654, November 15. World Bank, Washington, D.C. ______. 1999d. "Country Assistance Strategies: Retrospective and Implications." Discussion draft dated December 6. Operations Policy and Strategy Vice Presidency, World Bank, Washington, D.C. ______. 1999e. "Poverty Reduction Strategy Papers: Operational Issues." Draft report prepared in conjunction with the International Monetary Fund, December 10. World Bank, Washington, D.C. ______. 2000a. "Poverty Reduction Strategy Papers. Internal Guidance Note." Internal report dated January 18. Operations Policy and Strategy Vice Presidency, World Bank, Washington, D.C. ______. 2000b. "Memorandum of the President of the IDA and the IFC to the Executive Directors on a Country Assistance Strategy of the World Bank Group for the Kingdom of Cambodia." Report 20077-KH, February 7. World Bank, Washington, D.C. ______. 2000c. "Report on Progress in MDB Collaboration." Report prepared for the Meeting of the Presidents of Multilateral Development Banks, Manila, February 25. World Bank, Washington, D.C. ______. 2000d. "Comprehensive Development Framework (CDF): Mid-Term Progress Report." Internal report from the CDF Secretariat dated March 30. World Bank, Washington, D.C. ______. 2000e. "Progress Report on Poverty Reduction Strategy Papers (PRSPs)." Internal report dated April 13 prepared in conjunction with the International Monetary Fund. World Bank, Washington, D.C. ______. 2000f. "Vietnam: Country Assistance Strategy Progress Report." Document number IDA/R2000-66, May 8. World Bank, Washington, D.C. ______. 2000g. "Development Cooperation and Conflict." Operational Policies and Bank Procedures 2.30. World Bank, Washington, D.C. 37 ______. 2000j. "Supporting Country Development: World Bank Role and Instruments in Low- and Middle-Income Countries." Discussion draft dated July 21. Operations Policy and Strategy Vice Presidency, World Bank, Washington, D.C. ______. 2000k. "Minutes of Meeting held on Monday, June 12, 2000." CODE document CODE/M2000-16, July 26. Committee on Development Effectiveness, World Bank, Washington, D.C. ______. 2000m. "Comprehensive Development Framework: Report on Country Experience. March 1999­July 2000." Board document SecM2000-474, August 10. CDF Secretariat, World Bank, Washington, D.C. ______. 2000p. "`Harmonization' of Policies and Procedures--A Status Report for Heads of MFI Meeting, September 24." Operations Policy and Strategy Vice Presidency, World Bank, Washington, D.C. World Bank and International Monetary Fund. 1995. "Bank-Fund Collaboration on Public Expenditure Work." Joint memorandum from Michel Camdessus and James D. Wolfensohn dated June 27. Washington, D.C. 38 Annex 3 Persons Interviewed James Adams, Country Director, AFC (video conference) Mario Aguilar, OCS Ivar Andersen, FRM Ian Bannon, LCC Anupam Basu, African Department, IMF Pamela Bigart, OCS Nat Colletta, Head, Post-Conflict Unit, SDV Louis Dicks-Mireaux, Development Issues Division, Policy Development and Review Department, IMF Clare Fleming, EXT, New York (telephone) James Funna, CDF Secretariat, OPS Engilbert Gudmunsson, FRM and CDF Secretariat, OPS Sanjeev Gupta, Expenditure Policy Division, Fiscal Affairs Department, IMF Pablo Guerrero, Head, CDF Secretariat, OPS Gregory Hancock, Partnership Group, AFR Nancy Happe, Chief, Development Issues Division, Policy Development and Review Department, IMF Keith Jay, FRM Orsalia Kalantzopoulos, Country Director, LCC Shigeo Katsu, ECAVP Basil Kavalsky, Country Director, ECA (telephone) Kasuhide Kuroda, SDVPC Michael Lav, FRM and OED Elizabeth Littlefield, CGAP Secretariat (telephone) Joan Martin-Brown, SDV (telephone) John Mitchell, CDF Secretariat, OPS Daniel Morrow, PRM Ohene Nyanin, Head, Partnership Group, AFR Jorma Paukku, FRM Alison Rosenberg, Partnership Group, AFR Allison Scott, African Department, IMF Colin Scott, SDVPC Gregory Toulmin, CDF Secretariat Hasan Tuluy, Country Director, AFC Harold Wackman, Country Director, AFC (telephone) Loreen Yap, retired (formerly SAC and FRM) Giuseppe Zampaglione, FRM 39 Legend: AFC, Africa Country Director Groups; AFR, Africa Regional Office; CDF, Comprehensive Development Framework; CGAP, Consultative Group for Assistance to the Poorest; ECAVP, Office of the Regional Vice President, Europe and Central Asia Region; EXT, External Affairs; FRM, Resource Mobilization Department; IMF, International Monetary Fund; LCC, Latin American and Caribbean Region Country Units; OCS, OperationalCore Services Network; PRM, Poverty Reduction and Economic Management Network; SAC, South Asia Country Units; SDV, Social Development Department; SDVPC, Postconflict Unit, Social Development Department. 40 Annex 4 Methodological Notes 1. This annex explains the methodologies and gives the results of reviews of coverage of aid coordination in three types of IDA documents: · Country Assistance Strategies (CASs); · Public Expenditure Reviews (PERs); and · Country Assistance Evaluations (CAEs) conducted by OED. 2. The reviews of CASs and PERs assess the quality of treatment of aid coordination in these documents. The review of CAEs brings together the findings of recent CAEs regarding key aspects of aid coordination and IDA's role. The findings of these reviews are referenced and drawn on in the main text. Quality of CAS Treatment of Aid Coordination 3. The treatment of aid coordination in selected CASs for IDA countries was assessed during two periods: one near the beginning of IDA10 (1992­93), and the other near the beginning of IDA12 (1998­99). On a scale of "0" to "3" (with "3" being highest), the average rating for the 15 countries was "1.1" in the earlier period and "2.4" in the later period. 4. The main criterion used to judge the treatment of aid coordination issues in the CAS was the extent to which the discussion went beyond a general description of what other donors were doing--and the assertion that IDA intended to coordinate with them--to a discussion of specific ways in which IDA sought complementarity with other donors. A distinguishing characteristic of CASs that rated a "3" was a forthright discussion of selectivity or comparative advantage and an indication of what IDA would not do because it was deemed to lack comparative advantage. There has clearly been an increase in quality from the initial period to the final period--from unsatisfactory to satisfactory or better. 5. A subsequent examination of CASs issued during fiscal 2000 (through July 2000), suggests some further improvement in the treatment of aid coordination issues, including programmatic selectivity. Eight full or complete CASs and eight partial CASs (progress reports, updates, addenda) were rated. The average for the full CASs was "2.6", while the average for the partial CASs was "1.9". Degree of Consultation and Coordination with Other Donors in Preparation of PERs 6. PERs for nine IDA countries were rated on a scale of "0" no consultation or participation of other donors to "3" (extensive participation). The average rating for the nine countries was 1.3. By far the most extensive example of donor participation was the November 1999 PER for Ethiopia, which reported participation of five other donors--including the International Monetary Fund, European Community, African Development Bank, the Netherlands, and the United Kingdom--and workshops in Addis Ababa and Brussels. That effort followed the 1997 Ethiopia review, which included six other donor representatives on the team. 41 Findings of CAEs on Aid Coordination 7. CAEs prepared by OED over the last four years for 23 IDA (and IDA-IBRD) countries were reviewed for their findings on five criteria of effective aid coordination and IDA's role. (a) Adequacy of consultation with other partners by IDA Good practice, or "highly satisfactory," would be characterized by a situation where partners view IDA staff in the resident mission and from headquarters as being consistently and fully consultative, both at the beginning of a process or mission and in providing feedback at or near completion. Consultation by IDA is also viewed as highly satisfactory in terms of style and timeliness (e.g. sufficient advance notice of a visit and sufficient time to respond to a request for comments). (b) IDA attention to selectivity and comparative advantage Good practice: IDA has included an analysis of its comparative advantages, leading to identification of sectors in which to concentrate its assistance and those in which other donors have comparative advantage. IDA will have also drawn out the implications of this analysis for the activities it supports, the instruments it uses, and the priority it gives to these interventions. Moreover, IDA demonstrates that it is actually implementing its portfolio in accordance with these priorities. (c) IDA effectiveness in promoting harmonization of procedures among donors Good practice: IDA has successfully worked with other donors to bring about harmonized procedures for budgeting, procurement, and reporting in key sectors, as well as highly coordinated or joint visiting missions for design, supervision and evaluation, thereby reducing transactions costs in the delivery of aid and the consequent burden on the country. (d) IDA's assistance to strengthen the capacity of the country to manage and coordinate aid resources Good practice: IDA has pursued a comprehensive strategy to strengthen country aid management and coordination capacity. This has included (1) the provision of effective, demand-led technical assistance for this purpose; (2) development of such capacity by drawing on country analytical capacity and encouraging the government to take the lead in preparing, convening, chairing, and following up on aid coordination meetings, including the CG and local coordination meetings; and avoiding, or correcting, practices, such as Project Implementation Units (PIUs), that tend to undermine country capacity. It has encouraged other donors to follow suit. (e) IDA has sought cost-effective approaches to aid coordination mechanisms Good practice: IDA has worked with the government and other partners to establish aid coordination mechanisms that participants perceive as being worth the direct and opportunity costs they entail, including "apex" level meetings, such as CGs, held in the country. IDA has also worked to structure local meetings to include both information sharing and reaching decisions on actions to which the participants attach significance and value. 8. Only a few examples reported by CAEs came close to "good practice," as defined above. The most frequent evidence of good practice was for the first criterion, "consultation." The least satisfactory performance was reported for the second and fourth criteria, "selectivity" and "strengthen capacity," respectively. The 2000 Tanzania CAE is a noteworthy example of a CAE that provided pertinent, but not all "good practice," findings for all the above criteria. A matrix 42 that presents the findings culled from each of the 23 CAEs can be obtained from the Corporate Evaluation and Methods Group of OED (OEDCM) on request.48 48The findings represent the views of OED evaluators and are not necessarily shared by the respective Bank Regions. IDA's performance against each of the above criteria is not assigned ratings in the matrix because the coverage of each criterion varied from one CAE to another, reflecting the fact that most CAEs undertaken during the period did not include such criteria in their designs. The CAE analysis was prepared by Patricia Laverly, Pierre-Joseph Kingbo, and Patrick Grasso of OED.