82108 A Multi-Partner Evaluation of the Comprehensive Development Framework Implementation of Country-led Partnership in the context of the Comprehensive Development Framework Torgny Holmgren Expert Group on Development Issues, Ministry for Foreign Affairs, Sweden Charles Soludo African Institute for Applied Economics, Enugu; and University of Nigeria, Nsukka, Nigeria The findings, interpretations, and conclusions expressed here are those of the author(s) and do not necessarily reflect the views of the Board of Executive Directors of the World Bank or the governments they represent. The World Bank cannot guarantee the accuracy of the data included in this work. The boundaries, colors, denominations, and other information shown on any map in this work do not imply on the part of the World Bank any judgment of the legal status of any territory or the endorsement or acceptance of such boundaries. Contact: Operations Evaluation Department Partnerships & Knowledge Programs (OEDPK) e-mail: eline@worldbank.org Telephone: 202-458-4497 Facsimile: 202-522-3125 http:/www.worldbank.org/oed iii Table of Contents Abbreviations and Acronyms ......................................................................................... iv Executive Summary .......................................................................................................... v 1. Introduction ................................................................................................................. 1 2. Partnership: Theory and Expectations Under the CDF.......................................... 2 Partnerships in General................................................................................................ 2 Collective Action Problems........................................................................................... 3 Operational Definition of Development Partnership.................................................... 5 3. Development Partnerships ......................................................................................... 7 From “Donorship” to Partnership ............................................................................... 7 Different kinds of Development Partnerships ............................................................... 7 Among Stakeholders in Developing Countries ............................................................. 7 Among Actors in Donor Countries................................................................................ 8 Among Donors .............................................................................................................. 8 Between Donors and Partner Countries ....................................................................... 9 Regional Partnerships................................................................................................... 9 Multi-Country Partnerships .......................................................................................... 9 Two Key Factors for Sustainable Partnerships .......................................................... 10 Political Willingness and Commitment to Partnership............................................... 10 Country Ownership and Leadership ........................................................................... 11 4. Issues in the Implementation of Country-led Partnership.................................... 12 4.1 Conceptual Framework ................................................................................... 12 Incentive System ........................................................................................ 12 Asymmetrical Power Relations ................................................................. 13 Institutional Capacity................................................................................ 14 Legal-institutional Contexts in which Actors Operate.............................. 14 Asymmetric information or lack of proper flow of information................ 15 4.2 Issues in the Implementation........................................................................... 15 5. Survey Results ........................................................................................................... 19 5.1 Donors’ Views and Behavior .......................................................................... 20 5.2 Recipients’ Views and Behavior..................................................................... 27 6. Towards the Development of Partnership Performance Evaluation ........................... 33 7. Conclusions ................................................................................................................ 38 References ........................................................................................................................ 43 Terms of Reference ......................................................................................................... 46 Box 1: Evaluation Questions: Country-Led Partnership………………………….....59 iv Abbreviations and Acronyms ACP African, Carribean, Pacific countries AfDB African Development Bank AsDB Asian Development Bank CDF Comprehensive Development Framework CGIAR Consultative Group for International Agriculture Research CG Consultative Group CIDA Canadian International Development Agency CPIA Country Policy and Institutional Assessment (Index) DAC Development Assistance Committee of OECD DfID Department for International Development, UK EBRD European Bank for Reconstruction and Development EC European Commission EU European Union GDP Gross Domestic Product HIPC Heavily Indebted Poor Countries IDA International Development Agency IDB Inter-American Development Bank IDGs International Development Goals IMF International Monetary Fund NEPAD New Partnership for Africa’s Development MDGs Millennium Development Goals NGO Nongovernmental Organization ODA Official Development Assistance OED Operations Evaluation Department, World Bank OECD Organization for Economic Co-operation and Development PIU Project Implementation Unit PRSC Poverty Reduction Support Credit PRSP Poverty Reduction Strategy Paper SWAps Sector-Wide Approaches TA Technical Assistance UK United Kingdom UNDP United Nations Development Programme USAID United States Agency for International Development WB World Bank v Executive Summary 1. The study starts off with a review of the theoretical underpinnings of the partnership concept. We discuss different kinds of development partnerships and find that an effective country-led partnership is based on trust, attitude and contractual relationship. The World Bank Partnership Group (1998a:5) provides a useful operational definition of partnership: “Partnership is a collaborative relationship between entities to work toward shared objectives through a mutually agreed division of labor.” 2. We hypothesize that the actual implementation of effective partnership that is “led” by the recipients would largely depend on the incentive system faced by the actors to behave in a cooperative way, the power relations among them, the capacity of the anointed leader to provide leadership, the information set available to the partners, and the legal/institutional contexts in which the partners have to operate. We elaborate on each of these points, and show how they impact the implementation process. 3. Furthermore, we propose four hypotheses regarding the impact of donor behavior on the capacity of aid-receiving countries in implementing the country-led partnership principle. A survey was issued to selected donors and recipient countries on a number of key issues linked to the hypotheses. Based on the survey results and a review of existing literature we draw the following conclusions: 4. Our key finding is that there have been improvements in aid delivery process towards better coordination, country ownership and partnerships, but that these improvements have not been systematic nor can they be unambiguously attributed to the introduction of the CDF. Hypothesis 1. Asymmetrical power relations 5. “The higher the share of donor resources in the recipient country, the greater the donor power and hence the more difficult it is to truly have country-led partnership.” 6. Overall, the surveys confirm this hypothesis. However, development towards a more balanced relationship can be noticed in some aid dependent countries such as Uganda, with, for instance, local governments increasingly taking the lead in aid coordination. CDF has probably contributed to such adjustments in the CDF pilot countries, even though it is still problematic to fully deduce the role CDF has played compared to other factors. A more fundamental change in the relationship will depend mainly on transformed donor behavior. 7. Most donors and recipient countries believe that aid dependency and indebtedness have a negative influence on the quality of partnership. Aid dependency can, however, under certain circumstances also contribute to more constructive engagements between a developing country government and the donors. Several countries have established in– country aid coordination mechanisms. Further use of program and budget support will most likely result in improved coordination. Donors and recipient countries alike as the main obstacle to government-led coordination identified lack of local institutional capacity. Neither donors nor recipients see country-led partnership as contradictory to genuine country ownership, even if it occasionally can be unbalanced. vi Hypothesis 2. Donor legislation, guidelines and enforcement 8. “Partnership on the ground is hampered by disparate donor country and institutional legislation, guidelines and procurement rules guiding aid delivery.” 9. The hypothesis can be confirmed by the different approaches donors are taking. All donors in the survey claim that they have accepted the country-led partnership principle, but only two donors have changed their legal framework governing aid relationships. Most donors have introduced changes in their administrative rules and procedures governing development assistance and restructured their organizations in order to respond to the new aid relationship. Almost no donor has, however, established any explicit exit strategy guiding its withdrawal from a partnership. Current changes in donor behavior seem less grounded on legislation than on changes in working methods and habits. The only tangible evidence of changed behavior based on legal grounds is the untying of aid to the Least Developed Countries that was approved by OECD DAC in May 2001. The question thus remains whether legislative changes are needed in order for country-led partnership to take full effect in changes in donor behavior. Hypothesis 3. Donor policies and procedures 10. “The principle of country-led partnership is not an enforceable principle but rather a “moral issue” in which only converts apply it, and also determine the extent to which they do so. Without any possible sanction against non-compliance it is difficult to enforce it.” 11. The survey results support the hypothesis. Enforcement hinges almost exclusively on donor attitudes. Recipients have very few, if any, instrument to apply sanctions against non-complying donors. One possibility to create or strengthen that ability would be to introduce peer reviews where donors and recipients jointly review a specific donor. 12. Donors are increasingly moving away from projects to program support, sector- wide approaches and pooled funding mechanisms. Still, the large majority of donor-funded activities are project-related. There is no tendency to use program aid more frequently in CDF countries compared to non-CDF countries. The determining factor is the existence of a poverty reduction strategy or an economic reform program rather than the CDF. 13. Technical Assistance (TA), according to the recipients, continues to be supply- driven, hampering true ownership and the use of local capacities. Donors claim that they are changing the direction of their TA by abandoning freestanding TA projects, and by integrating TA into projects and programs. Donors continue to use Project Implementation Units, to a large extent, for carrying out programs and projects. 14. A number of initiatives have been taken recently among bilateral and multilateral donors to harmonize aid delivery procedures. So far the main emphasis of the work has been to find common international standards and principles at the institutional agency level, rather than to adapt donor behavior to the procedures given in individual developing countries. Furthermore, recipients find that donor headquarters' policies hinder the development of country specific harmonization. 15. Donors have only in a limited number of cases replaced own reporting requirements with regular government reports. Most evaluations undertaken are still vii agency related, assessing the partial impact of a particular donor’s efforts. Joint monitoring and evaluation of development programs rarely takes place. 16. Policy coherence issues are increasingly being discussed among the donor countries. Nevertheless, almost no country has implemented any changes in their overall policy framework to take development concerns into account when forming policies in other areas than development assistance. Hypothesis 4. Developing country policies and institutional capacity 17. “Improved partnerships are generally associated with better policy design and implementation. Causality could be either way, i.e. improved partnerships are easier in countries with better governance, policies and institutional structure. On the other hand, institutionalizing partnerships can strengthen institutional and governance structures, which impact on policy choice and implementation.” 18. The analysis supports the hypothesis. Local government leadership in management and coordination of aid resources has been strengthened in a number of developing countries over the last three years. The survey indicates that the cooperation among different local actors in developing countries have increased as a result of the introduction of the CDF principles. Almost all donors have, according to the responses, aligned their country assistance programs with the CDF countries’ development strategies (PRSP). Most donors find, contrary to the opinions of the recipients, that the partnership principle is viewed by stakeholders in the recipient country as imposed by donors. 19. Recipients and most donors believe that the implementation of country-led partnership principle will lower the transaction costs of administering aid programs. The expenses may, however, shift from the government towards the donors. Obstacles to Country-Led Partnerships 20. The factors that hamper country-led partnerships can be summarized to include: Among recipients • Insufficient or uncertain political commitment to poverty reduction/sustainable development (lacking ownership). • Limited implementation capacity. • Low standards of governance, including lack of transparency, weak monitoring and accountability mechanisms. • Aid dependence. Among donors • Lack of political will and legal-institutional framework to accept country-led strategies. • Preference to maintain traditional, and bilateral, patterns of interaction. • Distrust of government accountability mechanisms. • Differing donor agendas and desire to "plant the flag." • Frequent policy changes. • Lack of aid coordination. • Lack of harmonization of donor practices. viii • Provision of supply-driven Technical Assistance and continued use of Project Implementation Units. • Tying of aid. • Limited field presence. Broad observations 21. The study suggests the following broad observations regarding donor behavior: • General policy declarations about country-led partnership play a limited role unless they are followed up by new legislation or guidelines, supported and enforced by agency leadership. One good example is the OECD-DAC Recommendation on untying of aid. • Practical application of common aid instruments (SWAps, budget support) has had a great impact on donor behavior. Changes in performance have been more linked to the implementation of such aid instruments than to general policy declarations about the benefits of country-led partnership. The real test on donor commitment to country-led partnership takes place at the country level when aid strategies are transformed into practical use. • The wish by recipient countries to receive aid in the form of program and budget support will help change donor behavior. Donors would, by necessity, change their conduct if, and when, they redirect their funding from individual projects to common pools, and let go of agency specific solutions. • Increased use of program and budget support will promote country-led partnership through: - improved aid coordination - harmonization of donor procedures - joint monitoring and evaluation - restructured Technical Assistance • Peer pressure among donors works in changing policy directions. The efforts by OECD DAC and the Utstein group1, among others, to promote country-led partnership have considerably influenced donor policies. Continued work to bridge the gap between policies and practical implementation is still needed . • Recipient countries need to be closer involved in the development of donor policies. Too much policymaking is made in closed donor circles. Effective partnerships require that donors and recipients together form the future aid relationship. Recipients should, for instance, be involved in OECD DAC Peer Reviews of individual donors. Recommendations to Donors to Promote Country-Led Partnerships 22. Donors can best promote country-led partnership by: • Responding to local government leadership • Aligning donor assistance with government strategy • Accepting less high-profile programs and projects 1 Informal donor group consisting of Germany, the Netherlands, Norway and United Kingdom. ix • Supporting capacity building • Being transparent and providing timely information • Increasing the proportion of program and budget support (pooled funding) • Providing demand-driven Technical Assistance • Ending the use of Project Implementation Units • Improving aid coordination • Conducting joint analytical work (with local governments and among donors) • Harmonizing donor procedures and practices • Untying aid • Improving policy coherence • Increasing field presence and delegating more authority to country offices Partnership Performance Indicators 23. The CDF approach should be for the long-term. Ultimately we would like to observe the impact of country-led partnership on the Millennium Development Goals. We identify a number of cross-country indicators regarding the functionality and the development impact of the partnership principle. The indicators proposed could be further elaborated upon to form a partnership index to be used to measure intermediate or direct outcomes on the use of the partnership principle. We provide a skeletal listing of some issues/questions that might be part of the indicators, according to the following groupings: - Institutionalization of partnership - Recipient country policy and institutional stance - Donor country aid policies and procedures 24. The road to effective country-led partnership might be long, but the journey is doable. It requires not only political commitment to reforming the existing institutions and entrenched behaviors, but more so, continuing technical work. In other words, the agenda for country-led partnership is largely a work-in-progress. 1 1. Introduction 1.1 There is ongoing debate about the effectiveness of aid in addressing poverty and underdevelopment. Part of the reasons for the ineffectiveness is the defective delivery mechanism driven by disparate donor preferences, donor assistance strategies poorly aligned with country development strategies, wasteful and duplicative competition among externally-funded programs, and divergent and burdensome donor procedures and requirements that unnecessarily increase the transaction costs of delivering aid. In response to this ineffective mechanism, the CDF principles emphasize a new aid relationship based on country-led partnership, ownership, participation, donor coordination and harmonization. The implementation of these principles are expected to lead to greater efficiency in donor assistance, raise institutional quality in the recipient countries, and increase the impact on the attainment of the Millennium Development Goals (MDGs). 1.2 “Partnership” in aid relationship is not a new principle. The principle was hardly practiced in reality. However, since 1999 the concept has been given added impetus within the context of the CDF.2 Three years have since passed. Has the principle been implemented, or changed the underlying aid relationships and effectiveness? This paper focuses on one of the principles—country-led partnership. More broadly, we focus on the relevance, efficacy and efficiency of the implementation of the country-led partnership principle both on the ground and at the policy level among donor and partner countries. Specifically, the study ascertains the extent to which the country-led partnership principle has been practiced, evaluates the factors that have facilitated or hindered its implementation; analyzes whether its implementation has been oriented to pursuing the overarching MDGs; and identifies a set of hypotheses and cross-country indicators regarding the ‘functionality’ and the development impact of the partnership principle. 1.3 The study was carried out in two phases. The first phase focused on donors and development agencies and assessed the extent to which donors’ behavior has changed to enable the implementation of the partnership principle. We evaluated the issues involved in developing an effective partnership and formulated some hypotheses regarding the implementation of the partnership. A survey on these questions was issued to donors and recipient countries. This phase also discussed the conceptual and methodological germane issues to the development of effective partnership performance evaluation. We also developed some cross-country indicators to regularly evaluate partnership performance. The second phase concentrated on country-specific aspects of the CDF, drawing on the country experience. This helped refine the preliminary findings from the 2 Within the context of the CDF, the characteristics of country-led partnership are: 1) Government leadership in the management and coordination of aid resources, including: analytical and diagnostic work; aligning external support on the basis of the country’s development strategy and development agency comparative advantage; harmonization of development agency procedures and practices, e.g. budget cycles, procurement, reporting, evaluation and safeguards. 2) Relations among government, development agencies, other stakeholders, marked by: mutual trust, consultation, and transparency; assumption of accountability for sound financial management and performance; effective, demand-led support for strengthening government management and coordination capacity and not undermining it. 2 first phase. The focus was primarily on the original 12 CDF pilot countries, and especially the six countries covered by the research evaluation of the CDF.3 1.4 The study draws on a review of current literature on partnerships among academia, international development organizations and bilateral development agencies. It is also based on the survey issued, as well as findings from the case studies in this CDF Research Evaluation and similar studies on implementation of the partnership principle. 1.5 The rest of the report is organized as follows: Section 2 sets the stage by clarifying some theoretical basis for partnership and outlines the expected characteristics of the country-led partnership under the CDF. Section 3 looks more specifically into Development Partnerships. In Section 4, we discuss the conceptual framework for implementation of country-led partnership. Furthermore we propose some hypothesis regarding the impact of donor behavior on the capacity of aid-receiving countries in implementing the country-led partnership principle, and briefly highlight some of the issues that promote or hinder effective partnership at the country level. In Section 5, we report on the survey results and general findings from the literature regarding donor and recipient country behavior. Section 6 evaluates the conceptual and methodological issues in the construction and application of indicators of partnership performance. A number of indicators are also developed. Section 7, finally, provides our main conclusions. 1.6 In the text, “recipient countries” are also referred to as partner countries or developing countries while “donors” comprise high-income countries, their specific development agencies, the international financial institutions and the international development organizations. “Partners” include all parties (countries, agencies, organizations) taking part in a partnership. “Stakeholders” refer to parties having an interest in obtaining results in a specific field, even if not being partners of formal partnerships. 2. Partnership: Theory and Expectations Under the CDF Partnerships In General 2.1 Partnership can generally be seen as a means to a common objective. "It is a collaborative relationship towards mutually agreed objectives involving shared responsibility for outcomes, distinct accountabilities and reciprocal obligations" (Picciotto 1998, p.1). Further, "..a partnership is justified only if the benefits exceed the costs and the net benefits are equitably assigned to the parties" (Picciotto 1998, p.4). To be effective, partnerships need to be based upon a common vision among the partners of what the partnership is about. There needs to be a common basis of shared values, and the 3 - The six countries included in the CDF Research Evaluation are: Bolivia, Ghana, Romania, Uganda, Vietnam, and one non-CDF pilot country, Burkina Faso. - The original 12 CDF pilot countries are: Bolivia, Cote d’Ivoire, Dominican Republic, Eritrea, Ethiopia, Ghana, Kyrgyz Republic, Morocco, Romania, Uganda, West Bank and Gaza and Vietnam. 3 objectives of the partnership should be equally established and owned by the participants. Second, there should be mutual benefits for the partners in the outcome. Third, there is also a need of clarity in task allocation and a balance in influence and fairness in allocation of costs and benefits. The design and operation of the partnership should be appropriate and the partnership must be managed efficiently. Furthermore, effective partnerships require efficient specialization, competence, coordination and motivation. (Piccolo 1998, pp.1, 4) 2.2 The partnership concept is heavily used in the commercial sector. Business relationships are normally based on contractual partnerships where performance criteria are explicit and measurable. Moreover, reporting requirements are specified, time frames and the limits on independent action are laid out and required and prescribed actions are delineated. (Maxwell, S. and Conway, T. 2000) 2.3 Successful partnerships are based on a number of prerequisites. First, there should be focused and shared objectives among the partners involved. Second, concrete outcomes and actions to aim for are needed. Third, full ownership of the partnership and its objectives are required at senior levels. Fourth, leadership from one partner or a group of agents is crucial to get the partnership going. Fifth, the partners must engage in broad- based support of the partnership goals. Sixth, capacity building is needed in order to secure the full participation by the weaker partners in a partnership. (World Bank, Partnerships Group 1998a; Picciotto1998) 2.4 Although the partnership approach to a certain extent is dominated by its process orientation, partnerships should in the end be judged by their results. Due to the difficulty in quantifying all monetary and non-monetary factors involved, similar partnerships might be best assessed by benchmarking while counterfactuals must be used when dealing with a unique partnership. (Picciotto 1998, p. 4) Collective Action Problems 2.5 Foreign assistance can be regarded as a collective action, where two or more agents coordinate their effort to accomplish an outcome. Development of a specific country can be regarded as such an objective where development assistance is an international public good, that in turn requires cooperation, coherence and coordination. (Picciotto 1998, p. 6). The following discussion on collective action is based on a paper by Sandler (2000), which analyses a number of issues that the implementation of the CDF brings about. 2.6 One overriding concern is free-riding. One mean to address this is to tie “agent- specific private benefits to the public benefits so that contributors have private motives to go beyond purely public goods benefits conferred to everyone for the support of the public good” (Sandler 2000, pp.4-5). Recognizing different donors’ support with name tags or flags is one way of handling this problem. A second escape is to organize the co- operation in large groups with smaller subgroups. Free-riders are easier to track and individual action is more recognizable when group size is smaller. (Sandler 2000, p. 5) 4 2.7 Provision of foreign assistance can also give “rise to joint products, where a single activity yields more than one outcome” (Sandler 2000, p.8). By providing foreign assistance, donors can get both individual benefits and “purely public benefits by improving the welfare and stability of developing countries” (Sandler 2000, p.8). “For the provision of a pure public good, a Prisoner’s Dilemma often applies where the dominant strategy is… to not contribute to the collective action" (Sandler 2000, p.8). Emphasizing contributor–specific benefits, can circumvent such free-riding. If one donor “derives a larger share of private benefits than others from providing a public activity with joint products there may be sufficient private inducements for it to contribute despite the actions of others” (Sandler 2000, p.6). The donor–specific benefits can consist of trade opportunities, political, security and immigration considerations, as well as prevention of the spread of diseases. 2.8 The provision of joint products may also be influenced by institutional arrangements. Having their efforts noticed can increase donors' benefits. The design of the Consultative Group meetings and the replenishment rounds of the concessional facilities of the international financial institutions, with donors openly stating their pledges, contributes to such kind of recognition. Utilization of the CDF matrix and the institutional design of the CDF structure in individual countries should probably have the same effect, since donors will become more aware of who is doing what and to what extent. This will improve chances to circumvent the free-riding problem associated with larger groups. 2.9 The total provision of a public good depends upon the efforts of the individual contributors. Decisive for this is the “technology of public good aggregation” or “the manner in which individual contributions are aggregated” (Sandler 2000, pp. 10-11). Some efforts could be characterized as “weakest-link technology, where the smallest effort level determines the effectiveness of everyone’s efforts” (Sandler 2000, p.11). One example would be hindering epidemics. “Another technology requires that a threshold level be obtained before any benefits are realized” (Sandler 2000, p.11). Potential contributors may view their contributions as the critical one putting all combined efforts over the threshold. A third technology has the largest effort to “determine the overall level of the public good, with smaller contributions adding nothing to the total quantity” but often “result in one single agent doing everything” (Sandler, 2000, p 11). The provision of public goods can, accordingly, be promoted by institutional arrangements that bring out the difference in the aggregation technologies. (Sandler 2000, p.12). 2.10 The leadership role varies with the contribution aggregator (Arce 2001, p.133). It is in some cases impossible to aggregate preferences of individuals without a leader whose preferences become the group's preferences. There is often a need of special incentives to cooperate, if one individual member of a large group should take the lead. The design of partnership is therefore critical (Picciotto 1998, p.4). In the weakest link case, “leadership should take on a co-coordinating function” (Arce 2001, p.133). Leadership can also provide stability in an unstable system with potential free riders when a political entrepreneur leads by example. Such a leader would unilaterally provide an intermediate amount of the public good, and would be willing to match other entities’ contributions by raising her own efforts. Leading by example “is able to stabilize the 5 system and support efficient outcomes for almost every contribution aggregator” (Arce 2001, p.133). Accordingly, there is a need to “understand the underlying contribution aggregator of a international public good prior to any prescription for collective action” (Arce 2001, p.133). 2.11 Asymmetric information may also cause collective action problems among the different parties receiving aid. Often the central government acts as an agent for the local governments and negotiates on their behalf. This may cause uncertainties for the local entities since they are not always aware of how much, or if, the central government is promoting their specific goals. (Sandler 2000, p.15). 2.12 To conclude, “the design of the CDF can have profound impact on the incentives of donors to support foreign assistance.” “[…] the CDF approach can assist the donors to coordinate their contributions, thus limiting the free riding and under-contributions. The CDF matrix can foster joint products, which in turn, can motivate giving by donors owing to donor-specific gains.” (Sandler 2000, p.17) The CDF matrix and approach will help in this by observing the lower sub-levels in the federal structure, where individual donor support is easier to single out and monitor. (Sandler 2000, p.17). Operational Definition of Development Partnership 2.13 There exist a number of definitions of partnership emanating from different disciplines. Depending on where to put the emphasis, some claim that a true partnership must build upon an equal relationship between the partners. However, the partnerships embedded in the aid relationship can never reach that level due to the inherent inequality, with one party possessing resources (financial, human, etc.) while the other party always is in an inferior position striving to get the resources at the best terms. On the other hand, others claim that a partnership very well can comprise a principal-agent relationship with one party superior in terms of access to resources. 2.14 There are definitely some basic differences between partnerships in the development field and in commercial relations. Typically, in a development venture, values and objectives differ sharply among participants and there is asymmetry of resources and capacities among them. This is why public consultations and participation are so crucial in the design and implementation of development projects and programs. Where one or more of the participants do not enjoy the organization and skills available to other participants, capacity building is a prerequisite of true partnership. 2.15 The partnership concept has been used in the field of development cooperation since the late 1960s. Lester Pearson suggested in 1969 that "partnerships should be an understanding between donors and recipients expressing reciprocal rights and obligations, directed to a clear objective, or finite enterprise, beneficial to both" (World Bank Partnerships Group 1998a).4 4 From L.B. Pearson (chairman) "Partners in Development, Report of the Commission on International Development," 1969. 6 2.16 During the 1990s many bilateral donors and international institutions came to use the concept more frequently and provide their own definitions and understanding of the concept. OECD DAC put the ideas of partnership at the center in its strategy document, Shaping the 21st Century, where responsibilities of the developing countries and external partners as well as joint responsibilities are identified (OECD DAC 1996). Several bilateral donors also came to define different criteria for partnership in the 1990s, with an emphasis on transparency of values and the idea of contractuality supported by a code of conduct. (Maxwell S., and Conway, T. 2000). 2.17 According to an international workshop on "Making Partnership Work on the Ground" in 1999, a broad definition of partnership should cover the following three dimensions: (i) general development frameworks: policies, principles and priorities for a country’s development, (ii) the relationship proper: the structure of the interaction between the parties, the division of labor and roles; and (iii) aid coordination. (Ministry for Foreign Affairs, Sweden 1999). 2.18 We conclude that a genuine partnership between two or more partners should build upon: • Trust • Attitude • Contractual relationship 2.19 In order to achieve sound productive partnerships there needs to be a basic trust among the partners involved. That requires an open attitude towards the relationship proper and the other partners’ views, assets and means they bring into the partnership. Changes in prevailing attitudes are needed in many cases. Shared commitment to the goals, the objectives and the results to be achieved is required as well as mutual accountability. Partnerships work probably best when they are based on reciprocal relationships with clear understandings about the roles and responsibilities of the different partners and when there is open, inclusive dialogue among the actors (OECD DAC 2001). For a partnership to work, the partners concerned need to possess a minimum of knowledge, resources and means to implement their part of the partnership. Capacity building will be needed to empower the weaker partners. There is therefore a need for a long-term perspective and for an incremental, sustainable approach to partnership (Maxwell, S. and Conway, T. 2000). In addition, to strengthen trust and commitment, partners need to regularly assess each other’s performance in meeting agreed responsibilities and obligations. 2.20 We find the following operational definition of partnership concept by the World Bank Partnerships Group very useful (1998a, p.5): “Partnership is a collaborative relationship between entities to work toward shared objectives through a mutually agreed division of labor.” 2.21 The definition is neutral in the sense that it does not dictate what policy objectives the partnership should aim at. The objectives should be country or sector specific, drawing on the local circumstances and conditions in each individual case. The main idea 7 here is not to focus on development outcomes but on the process to realize them, which in turn depends on the set of objectives that the individual actors have established for the development and thereby for the partnership. 3. Development Partnerships From “Donorship” to Partnership 3.1 Development cooperation needs to build upon a trustful relationship between the partners involved to be effective. The unequal relationship that have characterized aid relationships over the years, with one partner possessing not only the resources (economic, human, etc.) but also the power to direct the use of the resources, has to be changed if a trustful and sustainable relationship should occur. The power to formulate and direct the policies to govern the use of the resources should rest with the partner country. There is an apparent need for donors to let go of their overwhelming power in the aid relationship, the so-called “donorship” that has marked the official aid relationship since its beginning some 40-50 years ago. The aim is to increase aid effectiveness and to ultimately achieve poverty reduction, by establishing broader and more equal relationships between donors and recipients at all levels. This in turn requires that the developing country or domestic actor concerned assumes the responsibility for the policies and that the cooperation builds upon the priorities of the developing country/actor. The policies should not be imposed upon the recipient by some external partner. Accordingly, it takes ownership to create genuine partnership. Solid ownership is a necessary condition for the establishment of a true partnership. A true partnership, as just described, will at the same time strengthen the ownership. The two principles, ownership and partnership are, thus, mutually reinforcing. Different kinds of Development Partnerships 3.2 There exist a plentiful number of different kinds of partnerships in the field of development cooperation. The most common partnership is the one formed between a donor government/agency and a developing country government. The broadening of the development agenda as well as the poverty concept means that more actors are getting involved and new kinds of partnerships are emerging. A distinction can be made between partnerships within countries, between developing countries and their donors and among donors. 3.3 This study will primarily focus on the basic partnership between governments (donor and partner country governments). In order for development to take place, partnerships among actors and stakeholders in the two country categories are required and therefore also reflected upon. The same goes for partnerships among donors, and between donors and different stakeholders in the partner countries. Among Stakeholders in Developing Countries 3.4 Partnerships between governments and stakeholders in developing countries aiming at common development objectives are becoming more common in line with 8 increased emphasis on participatory approaches. A more active participation by different actors strengthens the national ownership of development policies. Ongoing democratization and a stronger role played by national parliaments do also contribute to this development. The experience from Uganda illustrates this, where a key factor in the country’s development was a national consultation conference held in 1989 as a device to establish and strengthen ownership of the reform program. By opening up for other voices to be heard it helped open up the development process to other parties. (Devarajan et al. 2001) This is especially critical when going from “first-generation reforms” (mainly of macro economic nature) to “second-generation reforms”, like pubic sector reform, where participation by a much broader spectrum of actors and individuals is needed in order to achieve long lasting results. Among Actors in Donor Countries 3.5 Aid has in many countries been provided by civil society organizations, like missionary or emergency relief organizations, long before any public and tax-based official aid emerged. Since the late 1950s and early 1960s, at the dawn of the independence among many developing countries, the public aid system has grown in both financial terms and in number of actors. More recently, the private sector has come to play an increasingly critical role in providing finance and knowledge to developing countries. Since the early 1990s the private flows exceed the public flows. The relation of private to public flows has changed from 1:1 ten years ago to 3:1 today. New kinds of partnerships are emerging between governments/development agencies and non- governmental organizations and the private sector in the OECD countries. An increased interest in the private sector to take part in development has initiated a number of public- private partnerships. In addition, political parties and national parliaments in high-income countries have shown a greater interest in development matters since the end of the Cold War. 3.6 The number of actors (agencies, ministries, NGOs, etc.) involved in development matters and in providing aid has increased a lot over the last decade. In forming a coherent development assistance policy it is of great importance for the individual donor country to draw on the combined knowledge and resources of all its domestic actors. This is especially true when the relationship with a particular developing country is being designed. Policy coherence requires a participatory process. Among Donors 3.7 A number of partnerships exist among donors, especially government development agencies. The Development Assistance Committee (DAC) of the OECD might be seen as a partnership among bilateral donors. The replenishment rounds for the international development banks' concessional facilities also form a kind of partnership among the participating donors. The multilateral institutions cooperate in partnerships based on int.al. Memoranda of Understanding between the World Bank and the regional development banks (AfDB, AsDB, IDB) and the World Bank, EBRD and the Commission of the European Communities. Task Forces and Working Groups are also 9 formed for specific topics among the donors. Common for most arrangements is that the donors/institutions set their policies without the inclusion of developing countries. The plentiful networks among donor country NGOs also fall into this category as does policy- making groups such as Eurodad, Jubilee 2000 etc. Between Donors and Partner Countries 3.8 The most common partnership is the one formed between the governments of a donor country and of a developing country. The relationships often comprise formal bilateral state-to-state agreements but can also include other international or non- governmental entities. A myriad of relationships exists between the governments and nongovernmental organizations on both sides. Donor country NGOs can cooperate with public organizations in the developing country, and donor country governments might wish to cooperate with nongovernmental actors in the developing country concerned. Extensive direct contacts between public donor agencies and grassroots entities in developing countries can be complex and need to be structured in accordance with the overall policy framework governing the relationship between the two states. Regional Partnerships 3.9 The regional dimension of development is a good cause for forming regional partnerships. For certain issues, regional consensus through regional or sub-regional bodies may be more advantageous in forming ownership than pure domestic concerns (World Bank, Partnerships Group 1998b, Annex V). By the same token regional partnerships might be more beneficial than global alliances of bilateral partnerships with donors. 3.10 The New Partnership for Africa's Development (NEPAD) is an excellent example of a regional partnership. Based on African commitment and ownership, and developed by African leaders, NEPAD is a holistic, comprehensive integrated strategic framework for the socioeconomic development of Africa. It is also a framework for a new partnership with the rest of the world. NEPAD was highlighted as an example of a promising regional partnership and basis for a partnership between a continent and her donors at the Monterrey Conference on Financing for Development in March 2002. 3.11 The European Union and a number of states in the African, Caribbean and Pacific (ACP) area have formed a special partnership, based on an integrated approach focusing on economic and trade cooperation. The most recent agreement from 2000 aims at establishing new trading arrangements, through economic partnership agreements, where regional economic cooperation will play a critical role for the integration of developing countries into the world economy and for poverty reduction. (Cotonou agreement 2000). Multi-Country Partnerships 3.12 Many cross-country and global issues require partnerships at the global level. The emerging development agenda has multi-country and global issues. There are a number 10 of cross-border issues - like financial flows, trade and labor standards, rainforests and climate change - that require joint action across countries and by widely divergent set of actors. In addition there are a range of development concerns for which attention and resources need to be mobilized internationally. Some of the most successful partnerships in development are to be found around well-focused multi-country or global concerns. One example is the Consultative Group for International Agricultural Research (CGIAR), which brings together 53 public and private sector members, which support a network of 16 research centers. 3.13 Where and how a particular global public good should be produced will vary according to the nature of the specific good. For the time being there are about 20 global issues that cannot be tackled outside a framework of a global collective action. When it comes to resolving these issues, the current global governance set-up has not always been up to the task. Treaties and conventions are generally too slow for global issues requiring urgent response. UN conferences are too short on follow-up mechanisms. The multilateral institutions are often not able to handle global issues alone and G7/8 type mechanisms also suffer from limitations. New ideas being proposed on alternative global governance approaches are non-permanent intergovernmental groups, Global Issues Networks (permanent groups including public and private sector and civil society) and Global Public Goods Task Forces. (Ministry for Foreign Affairs, Sweden 2001, pp. 39- 45; Rischard 2002, pp. 61-64, 171-180). Two Key Factors for Sustainable Partnerships 3.14 The focus of the study is how country-led partnership can contribute to development and more effective development assistance. We found in Section 2 that trust and changes in attitudes as well as practices among both donors and recipient partners are key in building strong partnerships. Two key requirements to achieve those changes are the commitment and political willingness among both donors and partner countries to form country-led partnerships as well as ownership and leadership by the developing country partners. Political Willingness and Commitment to Partnership 3.15 A key feature of the country-led partnership principle is the willingness of the parties involved to enter into the relationship of the kind that is envisaged. It means adapting to the counterpart’s priorities and leaving behind an attitude to only focus on one’s own interests. The partner countries should "own" their development agenda. "Donors have to make the provision of aid–including the identification of priorities, negotiation of aid programs, delivery of aid resources, and evaluation of aid efforts – more supportive of the country's ownership." (World Bank, Partnerships Group 1998a, p.10). 3.16 Donors have to show a willingness to let go of their direction of the country's development program and projects. They need to adapt to new circumstances where the partner country's development plan should be decisive for the aid relationship. Leaving 11 behind eagerness to disburse aid and take over specific projects in order to let them be as effectively implemented as possible will require huge changes in attitude and practices for many donors. A true partnership might lead to slower disbursements and, in the short term at least, less effective implementation. 3.17 All partners in a partnership need to be transparent and open about their objectives with the partnership and demands on the co-partner. The nature of those commitments should be made explicit in the development of new partnership arrangements. Equally important is to establish under what circumstances the partnership might end. From a donor's perspective this means putting on paper the exit strategy for its aid involvement in an individual country be it due to certain time-limits being reached or due to other economic or political reasons (country reaches certain income level, obstruction of democratic and human rights principles, etc.). It is essential that both parties are aware of these conditions before entering into a formal partnership. 3.18 To be successful, partnerships need to be monitored and evaluated, according to some agreed benchmarks. It is in this context, also of great essence, which party decides whether a partnership is on track. The acid test of the strength of a partnership could be a situation where a recipient actor has based its policies on a process of national dialogue about its development strategy, but the donor disagree with the policies proposed. 3.19 The formation of a partnership itself requires some kind of a partnership in its making. Imposed partnerships will not be long-lived. What can developing countries expect in return for their commitment to a partnership bargain, if proposed of an external partner? There is an apparent risk that partnership under such circumstances might be viewed upon as a reward for compliance with the donor-initiated strategies. Country Ownership and Leadership 3.20 The developing country government must show political will to assume ownership of its development agenda. The government should lead the dialogue on how to formulate and implement its national development agenda. Donors, on their part, should step back from any leadership position and accept the government’s leadership by adjusting their own assistance strategies and activities to the government’s strategy. (Ministry for Foreign Affairs, Sweden 1999). 3.21 Lack of government ownership is manifested by the fact that it is often the donors and not the government who identifies the donor-supported projects and programs. Replacing the conventional project approach with program and budget support, based on the partner country's priorities, will strengthen government ownership. Reducing or eliminating the level of tied aid will also contribute to improve the country’s ownership and leadership. 12 4. Issues in the Implementation of Country-led Partnership 4.1 Conceptual Framework 4.1 As evident from Section 2, partnership is a principle with several ramifications, and one whose operational requirements can differ depending on the context. In the specific context of the CDF, it is required that the partnership be “country-led” in the sense of the recipient countries being on the “driver’s seat” with regards to both their development programming and coordination of aid relationships. This is the expectation of what “ought to” or “should” happen under the CDF principles. 4.2 But the reality can be dramatically different. We hypothesize that the actual implementation of effective partnership that is “led” by the recipients would largely depend on the incentive system faced by the actors to behave in a cooperative way, the power relations among them, the capacity of the anointed leader to provide leadership, the information set available to the partners, and the legal/institutional contexts in which the partners have to operate. We elaborate on each of these points, and show how they impact the implementation process. Incentive System 4.3 What is the incentive for cooperative behavior? Without appropriate incentives- rewards system or threats of sanctions - to behave in a cooperative way, it is difficult to enforce partnership. Both donors and recipients must not only share a common objective, but also have incentives to pursue them in a given way, under the leadership of the “country” for the country-led partnership principle to work well. But unlike in the case of corporate partnership where the partners might be driven by a common objective of profit motive, the development partnership is not always driven by the aim to “develop.” According to Easterly (2001: 115-6) “Lenders face incentives that cause them to give loans even when the conditions of the loans are not met. Recipients face incentives that cause them not to make reforms even when they get conditional loans. Many kinds of incentives cause these problems.” 4.4 CDF principles are predicated on the assumption that aid is primarily and almost exclusively directed to “development goals.” But there are other (sometimes dominant) reasons why donors give “aid.” Some aid are for strategic national interests of donors; sometimes new loans are given “to enable the old loans to be paid back” (Easterly 2001, p.117); sometimes the staff of the aid agencies simply give aid because their power, prestige and size of budget depend on how much disbursements they make, and in some cases “loan pushing” occurs because the donor staff is positively rewarded depending on the size of the loans he/she makes. Foreign aid creates and sustains a huge consultancy industry (jobs) in the OECD countries—amounting in the case of aid to Africa to about US$4 billion per annum. Of course, some aid is surely for “development” and directed to cure poverty. The recipients know this - that aid will come regardless of its behavior - at least one of the aforementioned factors would push lenders to give out aid anyway. Thus, in the context of disparate objectives of aid and diverse incentives faced by the actors, it 13 can sometimes be difficult to get the actors to behave in a cooperative way, especially if doing so would threaten their dominant objectives. It is therefore of little surprise that far little progress has been made on the partnership principle since it featured in the development literature in 1969. Even under the PRSP processes that are supposed to be driven by the recipients, donors usually finance such expensive undertakings, while in some cases also providing technical assistance. If donors are to submit their programming to the outcomes, the likely scenario is that they would either try to influence the “internal processes” that produce the “agenda” or back out altogether. Asymmetrical Power Relations 4.5 The nature of partnership depends on the power relations among the partners. Power in this instance is a function of the financial resources available to the partners. The asymmetrical power relations in most partnerships (donors and recipients; husband and wife; corporations, etc.) largely determine whether there are “senior” or “junior” partners. As the saying goes, he who pays the piper dictates the tune. Under the partnership principle in the CDF, we hypothesize that the implementation of the principle would largely depend on the relative shares of the donors in the country’s budget or as share of its GDP. In very low-income countries, with very high aid dependency and low capacity, it is indeed difficult for the countries to “lead”—especially where the donor actors are not mandated to cooperate in any special ways by their laws or international conventions. Aid is based on altruism, and if donors have complete control over the size and timing of the disbursement of aid, then a highly aid-dependent country cannot be but in a beggarly position vis-à-vis the donors. Effective country-led partnership in this circumstance can be somewhat difficult. Partnership must at the same time be distinguished from coordination. Coordination and cooperation especially by donors (i.e., among relatively equal partners)—to avoid waste and duplication—is certainly different from partnership in the sense of letting the recipient drive the process. 4.6 Another form of asymmetrical power relationships pertains to the recipient governments on the one hand and the domestic NGOs, and stakeholders on the other. Governments can consult these organizations but partnership is a tall order. Most evaluations of the CDF principle so far bemoan the fact that “open and wide consultation process for establishing the national development agenda” have often ended up as one-off events. This is not surprising. First, there is no threat of sanctions from the local stakeholders if the government fails to partner with them. Second, while the recommendation for “consultation” is reasonable, the “partnership” aspect is a bit controversial. In a democracy where the government has been given a mandate by the majority of the population, supposedly based upon its proposed programs and policies, it becomes tricky to require its “partnership” with NGOs and “stakeholders” in the design and implementation of its programs. Also, it is not clear where the NGOs derive their legitimacy, or the basis for selecting those that should have seats on the table.5 5 ”Consultation” should be the right word, and this should be in a context where the government is not bound to take the views of the NGOs or the “stakeholders.” Once the government has the mandate of its population through popular suffrage, it will be risky to require it to be held captive by special interest groups—through mechanisms for “partnership.” Furthermore, there is a proliferation of NGOs in most 14 Furthermore, the “stakeholders” might be special interest groups that may have lost their case in a national election. It is difficult to make a convincing case for another “wide consultation process” where there is a parliament—representatives of the people—and where the parliament approves government programs and the budget. Negotiations or bargaining with other groups that undermine the authority of the parliament derails the development of democratic institutions whereby government is held accountable to the people through their elected representatives. What should be required is that the development agenda is debated in the parliament, and the NGOs and other interest groups would be invited to make presentations in open hearing sessions. Institutional Capacity 4.7 Leadership is about capacity to provide vision and mobilize resources, including development assistance to accomplish the vision. It is also about managing complex relationships, such as the donors. Here, there could be a two-way causal link between partnership and aid relationships. Improved partnerships are easier in countries with better governance and institutional structure and hence better policy choices and implementation. On the other hand, institutionalizing partnerships can orchestrate and strengthen civil society and governmental institutional and governance structures, which impact on policy choice and implementation. Evidence shows that donors are more easily persuaded about the merit of partnership where there is a clear national development strategy. But developing such clear national strategy requires capacity. Evidence also suggests a high correlation between the level of development of a country and its institutional capacity. Thus, more effective partnerships have better prospects in more advanced economies than in the least developed countries. Legal-institutional Contexts in which Actors Operate 4.8 The legal and administrative rules guiding aid relationships are determined by individual aid agency. Their respective country laws, administrative guidelines and budgetary appropriations guide bilateral donors. Actors in the field adhere to such rules and laws. Thus, without changes in the legal and procedural rules guiding aid relationships, it is difficult to provide the “sanctions” against non-compliance or "rewards" for compliance. According to the World Bank CDF Secretariat (2001a: 4) “Progress in aligning the country assistance strategies of external partners to the national strategy is still low. Lack of clear signals from headquarters is partly responsible for this along with commercial interests and insistence on donor visibility.” Such a result is not surprising, since in the end, the actors cannot flout their country procurement procedures and rules. Here, partnership can often be confused with token cooperation, based essentially on the personal chemistry and whims and caprices of the donor staff. developing countries since the late 1980s as individuals resorted to forming NGOs as “survival” mechanisms devised to receive aid money. This phenomenon has given rise to what many pejoratively refer to as “NGIs”—that is, nongovernmental individuals. In that case, what will be the criteria to select the NGOs or NGIs to sit on the table to “partner” with the governments. 15 4.9 The principle of country-led partnership, like other CDF principles, is being preached as a “moral issue” in which only converts apply it, and also determine the extent to which they do so. It is not an enforceable principle. Since aid is based on altruism, the recipients cannot report any donor that fails to abide by the principle. Without any possible sanction against non-compliance, or international convention rules guiding aid delivery, or changes in the donor rules, guidelines and laws, the dynamics of donor- recipient partnership in each country depends on the quality of governance and institutional capacity in each country, and the personal (chemistry) relationships between the donor actors and government officials. Related to the legal and institutional framework, is the concept of a “common pool” approach to donor fund management— whereby an international aid convention requires that all donor funds be pooled together, and administered from a common pot. This would be one effective mechanism of orchestrating true partnership as all donors would be forced to cooperate, coordinate and partner with the recipient countries. Asymmetric information or lack of proper flow of information 4.10 A critical issue in the operationalization of partnership is information flows. In partnerships, each participant is entitled to every information regarding the enterprise. Relating this to the donor-recipient partnerships, both parties should have complete information regarding the objectives/expected outcomes and the financial flows. Under the current “partnership” arrangement, donors devise and conduct their own performance evaluation based on their own values and indicators. These are not necessarily the performance/monitoring indicators the recipients would have liked to see. On the other hand, there seems to be asymmetric information in the sense that a lot of the information set, which the recipients would have liked to have regarding donor behavior, is not available. For partnership to work, recipients would like to know how many financial resources to expect from donors over a reasonably long time and the timing of the disbursements. They would like the donors to set their programs and priorities for their particular countries on the basis of the recipients’ own agenda or a mutually agreed agenda. On the contrary, the recipients hardly know what to expect next or how the donors allocate their resources. 4.2 Issues in the Implementation 4.11 The main adaptations to country-led partnership seem to rest with the donors, since they have been in the driver’s seat for so long in the aid relationship. Our discussion primarily focuses on how donors and development agencies are practicing the partnership principle in different dimensions after the launching of the CDF. Based on the discussion about the conceptual framework above we propose four hypotheses regarding the impact of donor behavior on the capacity of aid receiving countries in implementing the country- led partnership principle. Under each hypothesis we discuss a number of key issues to implement development partnerships. The issues are drawn from earlier experience, findings in current literature, work undertaken by OECD DAC in formulating partnership 16 principles6 as well as the survey conducted in this study. The issues will help us assess what changes, if any, have taken place in donor approaches over the last few years. In Section 5 we will then report upon donors’ and partner countries’ views on the different issues according to the survey. Hypothesis 1. Asymmetrical Power Relations 4.12 “The higher the share of donor resources in the recipient country, the greater the donor power and hence the more difficult it is to truly have country-led partnership.” Aid Dependency 4.13 Aid dependency can be measured as the share of aid in the country budget, aid/ GDP or as aid per capita. Country size matters since aid dependency can be expected to be larger in smaller economies. Aid Coordination 4.14 Aid coordination is closely linked to country-led partnership. Responsibility for the coordination should reside with the partner country government since the purpose of the coordination should be “to integrate external assistance with the development priorities of the country concerned” (World Bank OED 2001b, p. xv). Donors should stimulate and help strengthen such partner-led aid coordination. This may be improved by donor’s own delegation of decisionmaking authority to country offices. 4.15 The country commitment and institutional capacity to manage and coordinate aid are strongly related with aid coordination patterns. Strong country commitment and high institutional capacity tend to be associated with country driven arrangements. On the other hand, the weaker the country commitment and the institutional capacity, the more likely donors will be to insist on donor-driven aid coordination in the interest of development effectiveness. The characteristics of the donors and international agencies also influence the relative ease or difficulty of moving to effective country-led aid coordination. The conditions for achieving country-led aid coordination are more favorable when donors have a high development orientation in their assistance programs and are relatively few. As the number of donors and projects increases the environment becomes more challenging. Country-led aid coordination becomes problematic when non-development motives (commercial and security concerns, etc.) are the main reason for donors to provide aid. (World Bank OED 2001b, pp.6-7). 4.16 The aid relationship has not by itself encouraged coordination. Donors have competed for projects in their desire to disburse aid. Recipient countries have occasionally gained in terms of funding arrangements by not encouraging coordination. Besides, line ministries have had no incentives to coordinate with other ministries in their search for funds. Coordination efforts can also displace necessary elements of efficient competition among donors. Lack of competition has often implied that any potential gain from division of labor and specialization among donors has been undermined. Every 6 A working checklist to guide efforts toward improved partnerships was formulated by OECD DAC in 1998, and “Guidelines for Poverty Reduction” issued in 2001 articulate priorities for bilateral agencies in working with partners to reduce poverty (OECD-DAC, 1998 and OECD-DAC, 2001). 17 single donor has been doing everything in each country where it is active. (Easterly W., 2002, pp. 16-21) Coordination must, accordingly, take its origin in a demand by the recipient in a way that fosters competition and division of labor rather than duplication and monopolistic behavior among the donors. Country Ownership 4.17 Country ownership was found in Section 3 to be a key factor for sustainable partnerships. The developing country should formulate the priorities and the plans that they seek the assistance for from international donors. Donors and especially the international financial institutions have on their side frequently conditioned their support (lending) on different measures being undertaken by the recipients partners. The use of conditionality has been criticized since the 1990s for having both imposed measures on the recipients and for not having lead to intentions aimed at, or results looked for. Hypothesis 2. Donor Legislation, Guidelines and Enforcement 4.18 “Partnership on the ground is hampered by disparate donor country and institutional legislation, guidelines and procurement rules guiding aid delivery.” Donor Legislation and Working Methods 4.19 Donor relations with sovereign nations are directed by a number of international and domestic instruments including legislation, government guidelines, budget appropriations etc. In order to fully apply a country-led partnership approach changes might be needed in all of those instruments. Some changes, like new fiduciary standards, might require legislative changes while others like strengthened field presence and delegation of authority to local offices can be decided at agency level. It is of interest to learn if any such have taken place among donors in order to encompass the notion on country-led partnership, or to use common donor-language, is the country-led partnership principle “owned “ by the donors concerned? Transparency and Exit Strategies 4.20 Transparency of donor and recipient partner interests is needed to build successful partnerships. The donors need to be open about under what circumstances they are ending their assistance (exit strategy), be it planned or due to breach of the contractual relationship. Tied Aid 4.21 Practices involving tied aid are prominently identified among procedures that can impair local ownership and capacity building, with substantial economic costs. Aid can become up to 25 percent more cost effective if it is completely untied, according to World Bank estimates. Untying of aid will in most cases require changes in legislation or guidelines. Furthermore, a demand-based aid system built upon developing country responsibility for results requires that aid is untied. Hypothesis 3. Donor Policies and Procedures 4.22 “The principle of country-led partnership is not an enforceable principle but rather a “moral issue” in which only converts apply it, and also determine the extent to which 18 they do so. Without any possible sanction against non-compliance it is difficult to enforce it.” Modify Aid Delivery from Project to Program and Budget Support 4.23 Evidence suggests that in order to improve development impact, donors should end the proliferation of projects and provide their aid increasingly in form of program aid and budget support. The assistance should be integrated into the recipient country’s budget. Donors should adopt, to the greatest extent possible, a multi-year time frame for programming and funding as a complement to multi-year partner government fiscal planning and budgeting. Technical Cooperation 4.24 The existing modalities for technical cooperation often appear expensive end excessive, hampering true ownership including the use and development of local capacities. Evaluations often show a low success rate for free standing technical assistance (TA). Technical assistance works when it responds to genuine demand and is well managed, but a lot of TA has been supply-driven and has not responded to the partners' requirements. In order to implement donor-supported projects, donors have also in many cases established parallel administrative mechanisms, like Project Implementation Units (PIUs). Policy Coherence 4.25 OECD countries need to ensure that decisions by their governments in all policy areas affecting developing countries should be coherent with the Millennium Development Goals. Policy decisions in areas such as trade, agriculture and migration have a huge impact on developing countries and are in many cases more important than aid in reducing poverty. Estimates by the OECD and the World Bank indicate that OECD tariffs and subsidies for agriculture and manufacturing may cause annual losses to the developing countries in the same order of magnitude as their total ODA receipts. Improving the coherence among donors’ development cooperation policies and their other policies affecting recipient partners is therefore increasingly important to help the developing countries move toward reduced dependence on aid and increased market access. Harmonization and Adjustment to Local Procedures 4.26 Partner country governments often single out harmonization of donor policies and procedures as the first thing donors should do to improve aid coordination. Donors should reduce the burden that development assistance creates for local partners by combining efforts (joint missions, common assessments, etc.) and ease administrative requirements (harmonizing procedures and paperwork). External partners should adjust more to local procedures. In some cases there is a need for donors to harmonize their procedures according to common international standards, like procurement procedures. Host governments, in their turn, need to improve national systems to reach international standards and best practices. Donors ought to align not only their procedures but also their assessments and policies. Poverty assessments should for instance be done in tandem between donors. Joint assessments will probably become even more important in step with increased use of program support. 19 Monitoring and Evaluation 4.27 Evaluation is essential for assessing impact and progress as well as learning from experience in any partnership relation. In order to assess partnership performance the different partners’ efforts should best be monitored and evaluated jointly by donors and developing country partners alike. To strengthen trust and commitment, the partners need to assess each other’s performance in meeting agreed responsibilities and obligations. (OECD-DAC 2001a, pp. 56-57). Hypothesis 4. Developing Country Policies and Institutional Capacity 4.28 “Improved partnerships are generally associated with better policy design and implementation. Causality could be either way, i.e. improved partnerships are easier in countries with better governance, policies and institutional structure. On the other hand, institutionalizing partnerships can strengthen institutional and governance structures, which impact on policy choice and implementation.” Country-Led Development Strategies 4.29 The country-led partnership approach requires the existence of satisfactory national development and poverty reduction strategies. Ultimately the annual budget ought to play the role of policy document upon which donors should base their support. Capacity Building 4.30 Appropriate domestic institutional capacity is necessary to run a homegrown development program. Effective partnerships require that the capacities of both governments and civil society organizations are enhanced and their structures strengthened. Institutional capacity is needed to identify, design and implement the development agenda, but also to monitor its implementation, to coordinate and evaluate individual aid supported activities (World Bank, Partnerships Group 1998a, p.10). One objective could be to increase “the capacity of [recipient] countries to have public expenditure management systems that meet the levels of standards with which external partners can work” (World Bank CDF 2001b, p.36). 4.31 Donors often overstretch and undermine the capacity of recipient governments by applying different rules and procedures regarding procurement, disbursement, reporting, accounting and auditing, and by the use of Project Implementation Units. Capacity building could improve by a closer focus on outputs and outcomes of capacity building activities rather than on inputs and budgets. More and better use of local capacity would also be a step in that direction. This is especially true for analytical work. 5. Survey Results 5.1 Have donors and partner countries’ behavior – policies, procedures, and practices – changed to enable implementation of the country-led partnership principle since the CDF was launched? The discussion below is based on questionnaires issued to donor and developing countries alike, and on general findings about recent changes in the behavior of the donor community at large. 20 5.1 Donors’ Views and Behavior 5.2 The donors were asked to focus their answers on the cooperation with countries applying the CDF principles. Responses were received from six bilateral donors (DfID- UK, CIDA-Canada, Danida-Denmark, Netherlands, Norway and USAID-USA) and one multilateral institution (European Commission). Most bilateral donors provided answers from headquarters as well as country offices. In total, 13 replies were received. This is a selective sample that naturally does not represent the views of the full donor community. However, we believe that the survey gives a good indication of what donors are undertaking in terms of their implementation of the partnership approach and as such will cast valuable light on the main issues being addressed in this paper. The discussion follows the structure of the four hypotheses and the corresponding issues put forward in Section 4.2. Hypothesis 1. Asymmetrical Power Relations Aid Dependency 5.3 Most donors in the survey believe that aid dependency and indebtedness have a negative influence on the quality of partnership. It makes it harder for government to say “No” to donor ideas, and diminish donor confidence in national ownership. Partnerships in unequal relationships are difficult and the recipient countries tend to agree with donors on all issues to avoid conflicts. Some respondents find that the PRSP is seen by many recipient country actors as merely HIPC conditionality. Aid Coordination 5.4 There are signs of improved aid coordination in many countries since the late 1990s. Developing country governments are increasingly playing a more active role and the local donor coordination has been intensified. Consultative Group meetings, which used to be held in western capitals, are now more and more being arranged in the capitals of the partner countries. The new format strengthens the local government’s role and involves a broader spectrum of government officials, elected representatives, private sector associations and civil society organizations. 5.5 Several countries have now established in-country coordination mechanisms. According to the CDF Secretariat at the World Bank there are adequate in-country coordination mechanisms in more than half of the 46 countries that were either part of the original CDF group or have prepared some kind of PRSP. (World Bank CDF 2001b, pp. 12-13). 5.6 All but one donor in the survey believed that aid coordination has improved in the CDF countries after the introduction of the CDF/CDF-like principles. USAID found that there was either good coordination predating the CDF, or that improved coordination was more due to donors' involvement with the PRSP process. Aid coordination was for instance already very active and intense in Bolivia before the CDF, based on the OECD/DAC’s “Shaping the 21st Century” principles. The introduction of the CDF has enabled the Bretton Woods institutions to participate more fully in the coordination process. However, in recent years, both donors and the government have found growing 21 coordination demands extremely onerous and have advocated that the number of existing joint working groups be reduced. In addition, coordination has almost come to a standstill in 2002 because of the upcoming elections. 5.7 Donors are divided on the issue whether the division of labor has increased among donors in the CDF countries. Some think that this is the case while others think that it is not noticeably as a result of CDF implementation but more likely due to improved coordination before the CDF was introduced or due to the PRSP process. 5.8 All bilateral donors find that there are effective government-led mechanisms for aid coordination in the CDF countries. EC believes that this is the case only when a PRSP is in place. Among examples given, the Consultative Group mechanism in Bolivia has grown from an annual affair to a yearlong activity of coordination and discussion. In Vietnam the local PRSP has the potential to promote government led aid coordination, as donors are increasingly accepting it as a framework for their own development assistance. 5.9 All countries and organizations have taken active steps over the last two years to actively support partner countries in strengthening their capacity to lead aid coordination. Many donors support key ministries (Ministries of Finance, Planning and Investment) in strengthening their general functions and to help them take a lead role in the production and implementation of the PRSP. One important aspect of donor support is peer pressure on other donors to let the government take the lead. 5.10 Donors indicated the following reasons for lacking government-led aid coordination. The reasons are ranked 1-5, where 5 gives the highest weight. The number of replies given under each rank is shown in the brackets. Constraints Rank Lack of local institutional capacity 1 [0 ] 2 [1 ] 3 [5 ] 4 [2 ] 5 [5 ] Government unwillingness to lead 1 [5 ] 2 [3 ] 3 [1 ] 4 [2 ] 5 [1 ] Constraints on donors by their respective l 1 [1 ] 2 [2 ] 3 [5 ] 4 [5 ] 5 [0 ] and procurement procedures Donor preferences differ 1 [0 ] 2 [3 ] 3 [7 ] 4 [2 ] 5 [1 ] Poor mechanism for donor coordination in 1 [3 ] 2 [1 ] 3 [2 ] 4 [0 ] 5 [0 ] country 5.11 Most donors as the key constraint indicated lack of local institutional capacity. Other constraints mentioned in addition to the listing above were donor and government fatigue with mounting aid coordination demands and reluctance of some donors to let go of some of their own organizational imperatives. Much of the donor hesitation to coordination may be caused by political reasons, wishing to retain visibility or freedom to act, rather than by legal or formal procedural grounds. 5.12 Two donor country offices find that closer alignment of donor and governmental priorities could come at the expense of civil society-government relations in recipient 22 countries. Others mean that this should not occur if donors are careful to insist on government accountability to domestic constituencies. Country Ownership 5.13 Most respondents do not find that country-led partnership development is contradictory to genuine ownership. It may however be unbalanced, especially if the country is a "poor performer." Donors need also be self-restraining enough to suppress their own preferences where necessary to ensure genuine ownership. Experience tells that when the government in place has shown leadership and has had strong ideas about what is best for the country, they have succeeded in building strong and genuine partnerships. 5.14 Some donors have increased their cooperation with different nongovernmental actors (civil society, private sector) in the recipient countries since the introduction of the CDF principles. But the majority of donors claim that the CDF itself has not been the main reason for closer cooperation with other actors. 5.15 The use of conditionality has been criticized for having been ineffective and for having imposed measures on governments that they do not want. Donors, especially the International Financial Institutions (IFIs) have also overburdened their support with conditions. To be effective, conditionality must focus on measures that the government believes in. It should also be used in the most appropriate phase of reform undertakings, when it can strengthen a government commitment to reform (Devarajan et al. 2001, pp 34-35). The IMF and the World Bank, the external partners making most use of conditionality, have stated they intend to draw down the number of conditions in each loan. The use of post-conditionality and single-tranche loans has become more frequent. Several bilateral donors have simultaneously introduced "governance " conditionality by demanding that certain democracy and human rights principles and standards be applied in the recipient countries. This has occasionally led to conflicting opinions with the IFIs, which have continued to provide funding to countries where the bilaterals have ceased to provide aid due unacceptable democracy and human rights records. In other cases bilateral donors have supported a government with satisfactory political records but weak financial policies, where the IFIs have withdrawn their support. Hypothesis 2. Donor legislation, guidelines and enforcement Donor Legislation and Working Methods 5.16 All donors responded that they had accepted the country-led partnership principle. USAID has supported the principle of partnership prior to the CDF. The partnership concept applied by the U.S. goes beyond host country governments and includes civil society, private sector, and other stakeholders. The CDF principles have also been important for EU but not as important as the PRSP process. 5.17 Only two donors (Denmark and Norway) claimed that they had changed the legal framework governing aid relationships to better deal with country-led partnerships. Denmark adopted a new strategy for its development policy in 2000, “Partnership 2000.” Laws are not governing aid relationships in Norway, but changed “rules” have been proposed by the government and approved by the parliament. The Netherlands replied 23 that there has been no need to change the legal basis for its development cooperation to comply with the CDF-principles, since they were already applied. 5.18 Most donors have introduced changes in their administrative rules and procedures governing development assistance in order to implement country-led partnership. EU has introduced a new process for Country Strategy Papers, anchored in nationally owned strategies for poverty reduction. CIDA has a number of decentralized funds that are managed from the field and in close partnership with the domestic governments. In addition CIDA is one of the few donors to have a domestic director for a Cooperation Office (in Bolivia). 5.19 All donors have restructured their organizations to better respond to the “new” aid relationships. All countries have or foresee an increased delegation of authority, and will decentralize more staff to in-country offices in developing countries. It has also become more common to hire local professionals. The World Bank, which used to be very concentrated to its headquarters, now has many of its country directors located in the partner countries. USAID has traditionally followed a highly decentralized operation, with in-country mission directors holding substantial authority and responsibility for in- country implementation. EU’s major restructuring of its aid administration, launched in 2000–2001, also contains increased delegation to the field. The Netherlands decentralized its bilateral development cooperation assistance to the local embassies already in 1996. A change in competence of personnel can be expected in the Netherlands, from project officers to policy discussion managers. Only macro-support is still decided centrally but that will probably be decentralized in the next few years. CIDA's Policy Branch has doubled in size between 2000-2002. Norway has strengthened the macroeconomic competence in its aid agency, NORAD. DfID has increased its delegated authority over the last two years, and further decentralization has taken place from regional to country offices. Transparency and Exit Strategies 5.20 There are very few OEDC countries, if any at all, which have explicit criteria established for when they are going to withdraw from a partnership/development co- operation relationship. The Cotonou agreement between the EU and the ACP states is an exception since it includes mechanisms for handling situations under which the purpose of the agreement risks being breached. Such mechanisms could be consultation procedures and different measures regarding human rights, democratic principles and the rule of law. (Cotonou Agreement 2000). Tied Aid 5.21 After more than 25 years of continuous discussions and three years of negotiations the bilateral donors in the OECD DAC finally came up with a Recommendation in May 2001 to untie aid (in most sectors) to the Least Developed Countries by the year 2002. Almost all agencies and donors have had to change their legislation (equivalent) concerning tied aid since the Recommendation was approved. The European Commission applies partial untying. All forms of aid covered by the Cotonou agreement with the ACP countries are untied for procurement among the EU members and the participating countries. Similar applications are now being discussed to 24 be introduced for other regions that receive aid from the Commission (Asia, Latin America, the Mediterranean and Central and Eastern Europe). Among the bilateral donors the UK has untied all its development assistance from April 2001 and a Swedish Parliamentary Commission has recently proposed to the government to untie all its aid. Hypothesis 3. Donor policies and procedures Modify Aid Delivery from Project to Program and Budget Support 5.22 Donors are increasingly moving away from individual projects to program support, sector-wide approaches and pooled funding mechanisms. To some extent this is due to the fact that many bilateral donors have increased the number of recipient countries and broadened the amount of supported activities. It has in some cases become necessary, due to administrative reasons, to pool resources together to be able to deliver, i.e. disburse funds. New administrative arrangements have also been introduced, like silent partnerships where one donor channels its funds via others. 5.23 The switch to program support can mainly be noted in developing countries that have come furthest in the process of introducing country-led policy frameworks, such as poverty reduction strategies. The more committed a government is to such plans, and the more developed they are, the better the prospects for donors to pool their resources together. There is no general tendency among donors to use program aid more frequently in CDF countries compared to non-CDF countries. The determining factor is the existence of a poverty reduction strategy or an economic reform programme rather than the CDF. 5.24 USAID has traditionally used budget support but experience with weak government institutions in some countries, and general statutory restrictions on some funding accounts, have limited the use of this mode. Denmark does not give budget support but Danida is moving from project support to sector support. Still, the large majority of donor-funded activities are project related. Program lending carries some problems for donors, like measuring results and linking own efforts to the outcome of the activity. Some donors are for that reason not taking part in direct budget support. Program and budget support also requires more sophisticated analyses and information systems for measuring results. Technical Cooperation 5.25 Most Technical Assistance (TA) continues to be supply-driven but donors are gradually changing the direction of their technical assistance by abandoning separate TA projects, and by integrating TA into projects and programs. Local capacities are being used more frequently, especially as the support moves toward general budget support. For the most part donors have, according to the survey, ended or are trying to end their use of Project Implementation Units in the conduct of aid programs. This is increasingly being done when there is a shift towards sector and budget support. However when the local capacity for the implementation of programs is lacking some donors still use implementation units. 25 Policy Coherence 5.26 The donor countries are increasingly addressing policy coherence issues. OECD DAC has produced an illustrative checklist on policy coherence for poverty reduction (OECD DAC 2001a, pp. 103-107). The need for coherence is enshrined in the treaty of the European Communities. United Kingdom, Norway and Sweden have addressed policy coherence in recent government and parliament papers. The United States is more and more using inter-agency working groups to set international policies. The US administration is encouraging greater USAID interaction with other agencies in addressing policies affecting low-income countries. In the Netherlands the Ministry of Foreign Affairs is in charge of all aspects of relations with developing countries. A specific coherence unit will be set up within the ministry in 2002. Switzerland has issued specific guidelines, and Germany examines all new legislation for its coherence with development policy (OECD DAC 2001a, pp. 93-94). However, almost no country responding to the questionnaire has implemented changes in their overall policy framework to take development concerns into account when forming policies in other areas than development assistance. Harmonization and Adjustment to Local Procedures 5.27 OECD DAC established in late 2000 a Task Force on Donor Practices with the objective to strengthen partner countries' ownership through identifying and documenting donor practices, which could reduce the burden of the capacities of partner countries to manage aid, and lower the transaction costs involved. The Task Force is specifically examining financial management and accountability; reporting and monitoring; and the pre-implementation phase. A group of developing countries is engaged to provide advice on various aspects of the harmonization work (OECD DAC 2001b). 5.28 The Development Committee of the World Bank and IMF requested in April 2001 that the development partners work together to develop an overall framework including a time-bound action plan for further progress in harmonizing operational policies and procedures. Three priority areas were singled out: financial management, procurement and environmental assessment (Development Committee 2001a, p. 2). A number of Multilateral Development Bank (MDB) Working Groups, established as a result of a 1996 MDB Task Force recommendation, are now looking into harmonization more explicitly in the three areas mentioned above and in evaluation cooperation. (Development Committee 2001b, pp. 2-3). Parallel to these efforts by the IFIs, the United Nations (UNDG Management Group) is working on harmonizing and simplifying operational policies and procedures among UN funds and Programs (Development Committee 2002, p.8). 5.29 The main emphasis of the work of the different working parties has been to find common international standards and principles at the institutional agency level, rather than to adopt donor behavior to the procedures given in individual developing countries. At the country level there are so far only few donors that are pursuing creative approaches (Development Committee 2001b, p.2). In the end, progress in individual countries will prove if the efforts are effective and successful. One interesting case could be Vietnam, where the donors are planning to assist the government to develop a harmonization program (Development Committee 2002, p.11). 26 5.30 Harmonization is likely to increase around local governments' standards/rules as donors move towards general budget support. The PRSP process is facilitating this. Respondents to the questionnaire find no difference in harmonizing procedures with CDF countries compared to non-CDF countries. Four of the seven donors find that headquarters’ policy hinders the development of country specific harmonization. Harmonization varies by country, but headquarters’ policy sets limits for flexibility. General policies aimed at increasing field presence and the promotion of more collaborative, knowledge-based and programmatic work will help harmonization efforts. Monitoring and Evaluation 5.31 A slight majority of the donors in the survey have replaced some of their own reporting requirements with regular government reports in any CDF country over the last two years. The European Commission aims to make normal annual PRSP reviews the core documents for their annual country program review, avoiding any demand for additional material from governments. Reporting requirements in Norway already follow regular government reporting practices to a large extent. CIDA has adopted the PRSP indicators jointly endorsed by the donor community in Bolivia and accepts regular progress reports by other donors and the government on these indicators. The Netherlands has in one case agreed to use the yearly report to the local Parliament as the report to donors. 5.32 Few initiatives have been taken to coordinate evaluation work among donors and recipient countries. Most evaluations undertaken are still agency related, assessing the partial impact of a particular donor’s efforts. Little attention has been given to capacity building for undertaking joint analytical work. All donors in the survey undertake some joint monitoring and evaluation (with recipient governments) of development programs since the CDF was launched, especially of budget support. USAID mentions that their policy requires a results orientation approach—why it is standard USAID practice to collaborate closely with partner countries in developing scopes of work for evaluations, reviewing findings, and determining actions to address issues. Hypothesis 4. Developing country policies and institutional capacity Country-Led Development Strategies 5.33 Almost all donors have, according to the survey, aligned their country assistance programs with the CDF countries’ development strategies (PRSP). Donors also support the countries in developing the PRSPs. Many countries and institutions, like Norway, UK, the European Commission and the World Bank, base their agreements/memoranda of understanding with individual partner countries on the country’s PRSP. The UN has established the Development Assistance Framework as an instrument to support national development strategies. However, a lot of work remains for donors to align their assistance strategies to the developing countries strategies. Only a third of the 46 CDF/PRSP-countries mentioned earlier are experiencing the benefits of greater donor selectivity. Donor alignment has probably made most progress at the sectoral level where sector working groups under government leadership are under way in a number of countries (World Bank CDF 2001b, pp. 13, 30-31). 27 5.34 No donor in the survey consults directly with the national Parliament in the recipient country when preparing his or her country assistance strategy. Consultations are held with the government. Some donors find they should not consult with the national parliaments but rather urge the local governments to do so. 5.35 Most donors find that the partnership principle is viewed by stakeholders in the recipient country as imposed by donors. Some stakeholders are inveterately suspicious of government and donors alike, others welcome the imposition, and yet others see it as a natural development of previous rhetoric. In Uganda, however, the PRSP (Poverty Eradication Action Plan) is very much seen as Uganda principles by domestic actors. Capacity Building 5.36 All donors, except Denmark and USA, believe that implementation of the Country-led Partnership principle will result in lower transaction costs of aid programs. According to the USAID the transaction costs will not likely be lowered due to weak local institutional capacity in most partner countries. In the long run the costs may decrease but in the short run with a mix of new and old approaches running side by side, it may mean that transaction costs actually increase, both for governments and donors. The costs for the donors would certainly increase given increased dialogue with government, civil society and other donors as well as country specific harmonization efforts. Success will also depend on the effectiveness of donor coordination. 5.2 Recipients’ Views and Behavior 5.37 In seeking the views of recipient countries we draw on the findings of the case studies and replies to a questionnaire from country case study teams in the CDF Research Evaluation (Bolivia, Burkina Faso, Ghana, Romania, Uganda and Vietnam). A reply to the questionnaire was also received from the local World Bank team in Mozambique, a non-CDF country. Even though the seven countries together represent a small sample, we believe that the results give some interesting indications on the implementation of the country-led partnership approach from the recipient partners' perspective. Hypothesis 1. Asymmetrical power relations Aid Dependency 5.38 The quality of the partnership can be negatively influenced by the extent of indebtedness and aid dependency, but also the other way round. The high quality of partnership in Uganda has led to constructive engagements between government and donors on the issues of indebtedness and aid dependency. One important outcome has been Uganda’s active role in the design of the HIPC debt relief initiative. Ghana's decision to formulate a Poverty Reduction Strategy was partly influenced by the need to secure HIPC debt relief and continuing access to concessionary credits from IMF and the World Bank. While many Bolivians feel the weight of HIPC and lingering complex aid conditionalities, none of Vietnam's major reforms have been influenced by conditionality. In Mozambique, on the other hand, aid dependency has inevitably put the government in an uncomfortable position. To access funds it is often led to undertake actions, which it 28 would rather not do, even though the government probably is better than most sub- Saharan Africa aid-dependent countries in saying “no.” Aid Coordination 5.39 It is hard to tell that aid coordination, generally, has improved after the CDF was introduced. In some cases it has—like in Bolivia, Vietnam and Uganda—while in other countries, such as Romania, it is not noteworthy. CDF has probably helped improve more coordination efforts that were already underway than it has triggered new coordination initiatives. 5.40 The division of labor among donors has increased in a number of countries since the CDF was launched. In Ghana this is mainly due to better information flows about each donor's activities and strengths. In Uganda, there is greater selectivity in sectors where SWAps are in full effect. Progress can also be noticed in Mozambique, even though slower. The multilateral institutions are cooperating much more closely and the division of labor has become clearer. In Bolivia, coordination has made it possible to bypass obstacles created by some donors' rigid procedures to some extent. There is, however, a tendency among donors to crowd into attractive sectors. In Romania it is hard to discern any movement, one way or the other. Parallel coordination processes are being used in Burkina Faso, in the transformation from the EU-led Conditionality Test program to the World Bank-led PRSP process. One process gathering the European donors in providing budget support is separate from the process adopted by the IMF and the World Bank. Twenty partnership groups formed in Vietnam, so far mainly chaired by donors, have contributed towards sector strategies. Donors also feel a stronger obligation to consult with each other before taking new initiatives. The groups can also contribute to greater selectivity among donors since all donors do not have the capacity to effectively take part in all groups. The partnership approach therefore implies that individual donors, particularly the smaller ones, need to actively build expertise and be more selective in how they engage. 5.41 Lack of local institutional capacity and differing donor preferences are specified as the main reasons for lacking government-led aid coordination. Most countries can demonstrate some effective government-led mechanisms for aid coordination. This is the case in Uganda with the Ministry of Finance at the institutional level and SWAps at the sectoral level. The Bolivian government leads 17 subgroups, established to work on the local PRSP. Despite the many coordination mechanisms, aid is highly dispersed and fragmented with more than 850 donor-funded projects and programs. The multiple forums for dialogue also often steer clear of current controversial issues. This is also the case in Ghana, where locally held CGs and mini-CGs do not result in in-depth discussions of serious development issues. The opposite experience can be observed in Vietnam where significant improvements can be noticed in the way the government and the donor community interact in an open and frank way on matters of national and sector policies. CG meetings, being held in Vietnam, include representatives from the private sector and NGOs. The capacity for aid coordination is still weak, especially at the provincial level. New coordination initiatives are now being taken in Ghana since the coordination mechanisms created following the launch of the CDF failed. While donors have increasingly been coordinating among themselves, the level of coordination with the 29 government remains weak. The capacity of the Ghanaian government to lead donor coordination is also limited and some ministers are reluctant to stronger coordination, for a fear of donor "ganging-up". Different sector programs and the Conditionality Test framework in Burkina Faso have created a precedent for more effective aid coordination. Country-led partnership is also developing positively at the local level where provincial Development Committees are effective vehicles for coordinating various development programs. In Romania donors are not very enthused about government-led coordination, and there is no existing apex mechanism, such as a Consultative Group. 5.42 Recipients often find that closer alignment of donor and governmental priorities come at the expense of government-civil society relations in the recipient countries. This is conceivable in Romania where private sector leaders feel they are ignored because senior donor officials head right for the government and ignore other actors on their short visits to the country. In the case of Uganda, on the other hand, government priorities have taken into account civil society’s voices and concerns. Country Ownership 5.43 Country-led partnership is commonly not seen as contradictory to genuine country ownership. Partnerships can only be effective if the partners respect one another and recognize respective and mutual achievements. Trust means a lot for the shape of the partnership, exemplified in Bolivia where the trust between the donors and the government declined a lot in 1998-2001 because of power shifts within the government and increased corruption. The case study of Burkina Faso indicates that donors need to recognize the country's achievements and respect its talent and potential rather than remarking on its inability to manage its affairs. Hypothesis 2. Donor Legislation, Guidelines and Enforcement Donor Legislation and Working Methods 5.44 It appears that donors, by gradual changes in their working methods, have accepted the CDF principle of country-led partnership in several countries (Bolivia, Mozambique, Vietnam and Uganda), while that it is not fully the case in other countries, like Ghana and Romania. Donors have increased their budget support in Uganda and are moving to ensure alignment with the Poverty Eradication Action Plan when providing project support. The main planning processes in Mozambique (PRSP, SWAps) have been government-led and are well accepted by donors. Donors have demonstrated their commitment to country-led partnership in Bolivia by support to the HIPC and to the preparation of the PRSP. In Romania, however, donors are ambivalent about country-led partnership. They have concerns about the capacity and style of the most logical ministry to lead partnership and tend to prefer bimodal relationships with pertinent line ministries and to coordinate informally with other donors as the need arises. Donors have in "principle" accepted the partnership approach, but there is a widely held belief (among donors and government officials) that the government has not assumed the necessary leadership needed for effective donor coordination. 5.45 Several donors have strengthened their presence in recipient countries, for example in Ghana and Vietnam. In the case of Vietnam donors that have assigned greater 30 autonomy to their field offices have got a clear advantage in engaging in partnership initiatives. Still most donors continue to draw significantly on headquarter-based staff. Hypothesis 3. Donor Policies and Procedures Modify Aid Delivery from Project to Program and Budget Support 5.46 Donors are trying to align their assistance programs with the recipient country’s development strategy in most countries where a PRSP exists (Burkina Faso, Bolivia, Mozambique and Uganda). However, in Ghana it is difficult to affirm if donors' assistance has been aligned to priorities of the government. In Romania, lip service is given to alignment with the Governing Action Program but this is not done in a rigorous way. 5.47 It emerges that recipient countries prefer donors to provide program and budget support rather than project aid. Donors are responding by gradually shifting from project to program and budget support. Increased use of SWAps in several countries indicates an enhanced acknowledgment of country ownership and recipient-driven partnership. The government of Vietnam has, however, mixed views about SWAPs. Some ministry officials fear that SWAPs could increase their burden if a large number of donors simply crowd into a sector without at the same time harmonizing or simplifying their procedures. In Uganda the use of SWAps has led to a relatively high degree of budget support. In the education and health sectors, 50% of aid is in the form of budget support. Eight bilateral donors are channeling budget support through the joint donor program and conducting joint annual reviews in Mozambique. Substantial changes have been seen in Bolivia over the last decade, with a shift towards sectoral approaches and gradual alignment of ODA with national priorities. Several NGOs are also giving up traditional "island" projects and are instead advocating influencing public policies. Donors are providing more budget support in Burkina Faso as well, but there are different opinions in the donor group on the provision of budget and sector support. Some donors have, moreover, decreased their field presence and are channeling more funds via international or national NGOs. This has created problems when those programs do not fit with the local PRSP in Burkina Faso and also require complementary government resources for their implementation. There seems to be a trend towards higher proportion of budget support also in Ghana. However, weak public finance management and implementation capacity are constraining further use of such support. Technical Cooperation 5.48 Technical assistance (TA) still appears to be supply-driven, hampering true ownership and the use of local capacities. Recipient country governments would like to get more control over TA, including taking part in selection and reviewing costs. Donors engage a lot in micromanagement of tasks that could be carried out in the normal course of public administration. There is, however, like in Ghana, an increasing shift towards the use of national experts instead of expatriates. Changes in government and donor behavior can often be linked to the introduction of SWAps and the shift towards more budget support. In Uganda, for instance, some success can be registered in the government's emphasis on attracting demand-driven technical assistance where SWAps are in full effect. 31 5.49 Donors continue to use Project Implementation Units (PIUs) to a large extent for carrying out programs and projects. Their persistence is of particular concern for the government of Uganda, which has stated a policy to phase them out in the near future. In Romania a lot of aid is provided this way. Six out of 22 World Bank funded projects are delivered through PIUs. There is also a wide range of PIUs being used in Bolivia, Ghana and Mozambique. Donors keep on to give incentives for "their" projects and programs in Bolivia, establishing PIUs and bypassing national procedures in order to achieve short- term efficiency. Two thirds of the donors use PIUs in Ghana. In addition, a "parallel civil service" is emerging in several areas, funded by donors and taking over strategic functions of the government. Harmonization and Adjustment to Local Procedures 5.50 Some progress can be seen in a few countries, regarding donors' adaptation to local country-led rules and procedures. This is more so in sectors where full SWAps or similar joint programs are in effect. Efforts to harmonize aid-giving procedures seem mainly to be a donor-driven process even if Uganda (among the countries appearing in our sample) has been consulted about DAC guidelines on harmonization of donor procedures and practices. Donor headquarters (HQ) policy are in general believed to hinder the development of country specific harmonization. In Uganda some donors have lagged behind others in aligning their rules and procedures with government and other donors because of HQ regulations that prohibit such moves. Initial steps have been taken to harmonize procedures in Vietnam. The government has promulgated a new law on procurement and the multilateral banks have harmonized their procurement procedures. Changes in procedures would eventually require radical decisions at donor and IFI headquarters level, and even changes in national legislation. "Cooperation islands" are common in Bolivia, with programs run by donor headquarters staff, at times disregarding national government policies. One major difficulty in harmonizing procedures in Bolivia has been the immense differences between the World Bank and the IDB. The slow progress in harmonization efforts in Ghana is illustrated by the fact that only three donors have adopted procurement rules of other donors. Monitoring and Evaluation 5.51 There are very few cases where donors have replaced own reporting requirements with regular government reports and where joint monitoring and evaluation of development programs are taking place. Joint sector reviews have been established in Uganda in the education and health sectors. Donors have accepted the report of the reviews as fulfilling their own requirements. In Mozambique and Vietnam there are some examples of joint analytical work. While donors in Ghana increasingly appear more willing to come together in the early stages of project and program formulation, the desire to maintain a bilateral role is evident in the few number of joint monitoring and evaluation missions donors undertake. While a few donors report an increasing trend in joint monitoring and evaluation missions, the majority feels that the trend has remained flat over the last five years. Bilateral donors and IFIs associate themselves with the local PRSP in Burkina Faso but they do not take advantage of coordinating their follow-up work. Bilateral donors evaluated, for instance, the PRSP in the summer of 2002 while the Bretton Woods institutions' staff assessment of the same PRSP will take place a few months later, in the fall. 32 Hypothesis 4. Developing Country Policies and Institutional Capacity Country-Led Development Strategies 5.52 Local government leadership in management and coordination of aid resources has been strengthened in a number of countries over the last three years. In some cases this can be attributed to the application of the CDF or CDF-like principles. There is a strong tradition of government leadership in Vietnam – the government determines the content, sequence and pace of the development program and takes the lead in aid coordination. A special decree was issued in May 2001 outlining the basic legal framework for ODA management. CG meetings have emerged as benchmarking events guided by the CDF principles. Various processes have also been introduced in Bolivia, such as a national dialogue, to institutionalize government leadership and consultations. Five of the six countries in the CDF Evaluation can today provide a satisfactory country strategy under government leadership (Bolivia, Burkina Faso, Ghana, Uganda and Vietnam). The evolution of the PRSP is a step in the direction of establishing country owned policies, but many governments find it hard to link their national strategic thinking with the PRSP process. In addition, there are only a handful of countries where the medium-term strategies are closely linked to a consistent budgetary framework (World Bank CDF 2001b, p. 16). 5.53 There is some evidence that the cooperation among different actors - government, civil society organizations, private sector – in individual countries has increased as a result of the introduction of the CDF principles. Donors and the government in Vietnam report significant improvements in the substance and process of cooperation. Extensive consultations on the 10-year development plan are now, for the first time, held outside the party structure. Uganda has been deliberately applying CDF-like principles for more than a decade, and this has resulted in strong partnerships among different actors. The government has actively engaged donors and Civil Society Organizations in formulation, implementation, and continuous revision of its national and sector strategies. A limited impact can be noticed in Romania, where the “CDF Consultations” in 1999 led to consultative approaches to formulation of EU accession strategy and to a ministry (Development and Prognosis) in the new government committed to CDF principles. In Mozambique, not being a CDF pilot country, the PRSP, "Vision 2025" and joint donor program exercises have played similar roles. 5.54 National parliaments are increasingly playing a more active role in preparation of national strategies and budgets. The role of the parliament has been evolving in Uganda, but the significance of parliament’s involvement in the budget process is viewed/appreciated differently among various actors within government and parliament. A consultative process in Romania in 2000 that led to approval of the Medium-Term Economic Strategy submitted to the EU involved intensive discussions with parliament and their ultimate approval. The current “Governing Action Plan” was endorsed by parliament. 5.55 The implementation of the country-led partnership principle is not, according to the replies received, viewed by stakeholders in the individual countries as imposed by donors. The principle is "home grown" in Bolivia and many activities in the area were not 33 only launched in Bolivia before the CDF concept was put forward, it also probably influenced the World Bank to adopt the CDF approach. In Vietnam the CDF is seen as a compact between the government and donors on a new way of doing business. Capacity Building 5.56 Recipients find that the implementation of country-led partnership principle will result in lower transaction costs of administering aid programs. It may, however, shift expenses away from the government and towards the donors. In Uganda, the transaction costs of administering aid programs have been lowered in sectors where full SWAps are in effect. The transaction costs of delivering and administering aid in Ghana remains fairly high, but few positive and tangible shifts can be observed in donors’ behavior that, if maintained, can reduce procedural and administrative burdens for both government and donors. The Bolivian experience shows that increased coordination also risk raising transaction costs. There are clear trends toward "donor crowding" around attractive issues for international cooperation, which could raise the costs in that particular sector; making coordination more efficient can lower the costs. There is, finally, a widespread distrust among donors in Burkina Faso on the capacity of Burkina government to track current expenditures. Some donors have therefore introduced their own systems of follow-up on particular sector investments. 6. Towards the Development of Partnership Performance Evaluation 6.1 The continuing imbalance in aid relationships is manifest in many ways. An important and previously neglected dimension of the problem is the imbalance in performance monitoring as between donors and recipients. Whereas the behavior and performance of low-income developing countries is measured and assessed in ever- increasing detail within the international community, the behavior and performance of their donor “partners” receives only cursory attention, except at an aggregate level which is of little operational usefulness to individual recipients. When it comes to performance monitoring, as in so many other spheres, the powerful (the donors and the international financial institutions) still call all the shots. Genuine partnership in development requires the monitoring, by independent assessors, of individual and collective donor performance at the level of individual aid-recipient developing countries. Do donors live up to their rhetoric and their promises? In what measurable ways? It is not difficult to devise measures of donor performance at the recipient country level; and some have been suggested above. Instituting systems of donor performance monitoring at the recipient country level can assist in improving understanding of aid effectiveness; promote the new forms of partnership of which there is so much talk; and, most important, assist policymakers in low-income countries in their difficult task of promoting poverty reduction and development. It is long overdue. It is time for it to be done (Helleiner 2000, pp.19-20). 6.2 Expectedly, the CDF approach should be for the long-term, and therefore it might be too early to expect concrete impacts on the ultimate objectives—the Millennium Development Goals (MDGs). But whether we can observe measurable long-term impacts depends on a clear definition of what to expect, and the tools for the evaluation. 34 6.3 “Measurement” of the ”degree of implementation of partnership,” and ultimately the impact of such “partnership” on the MDGs is a key aspect of any evaluation. So far, some ad-hoc measures exist which are used to monitor the implementation of the CDF principles. For example, the CDF Secretariat has recently constructed data on rankings of the extent of implementation of CDF principles based on surveys from 38 countries. On the partnership principle, the monitoring measures include evaluations pertaining to aid coordination, donor alignment with country strategy, and harmonization (World Bank CDF 2001b, Annex A). DAC has also compiled a list of key parameters for donors in building effective poverty reduction partnerships (OECD DAC 2001, p.59). Furthermore, there are suggestions that the World Bank’s Country Policy and Institutional Assessment (CPIA) index be used as a proxy for the quality of processes, and work has begun on better linking CPIA with the CDF and PRSP processes (World Bank CDF 2001b, p.37). These are very important positive steps, which need to be refined and improved upon. Currently, the CPIA is essentially World Bank internal and confidential ranking, based upon the staff members’ subjective views of policies and institutions in the countries. Much of the CDF surveys are based on a “YES” or “NO” kind of response which, for purposes of measuring progress, can sometimes be unhelpful. 6.4 There is therefore a need to begin to think in the direction of building a set of quantifiable indicators for the purpose of comparisons (over time and across countries) of the degrees of implementation of the partnership principle. Not only should such indicators be quantifiable, they should also be readily available to the general public as part of the process of transparency. 6.5 It must be pointed out at the outset that “quantification” in the assessment of qualitative behavior is a difficult task, involving sometimes arbitrary judgments regarding choice of ranking scale to be used; benchmarks; base period; assignment of weights to specific questions or modules of questions; and the aggregation of the weights to produce composite indicators (see Soludo, 1998). The construction of such indicators cannot be a one-off exercise, as it requires continuing refinements of the methodology. But a start is necessary. What is presented here is an attempt to draw attention to the kinds of issues that could guide such an effort. 6.6 The construction and use of the indicators to monitor progress would involve at least three sequential steps. First, it would be useful (despite all the problems) to construct a summary “index of development partnership” between the donors and recipient country governments—to gauge the extent of implementation of the principle. Of necessity, the construction of such an index will throw up the issues/factors that promote or hinder the implementation of the principle (as discussed in Section 4). Needless to say, such an index of partnership will also encompass the other CDF- principles of ownership, participation and coordination. A second step would be to examine the correlation between the index of partnership and summary indicators of the policy environment (however measured) including for example the CPIA index.7 This is on the hypothesis that improved partnership would lead to better policy design and 7 For the CPIA to become a credible measure, it has to be subjected to public scrutiny of its methodology, and the countries being assessed have to know the criteria for the assessment. 35 implementation, or that a better policy environment necessarily leads to better donor- recipient partnership. 6.7 Third (and the most tricky) is the inference as to the impact of partnership on the ultimate objectives—MDGs. This faces two clusters of problems. In the first instance, it needs to be established the extent to which the “improved” policy environment (such as the CPIA measure) is the result of improvements in donor-recipient partnership or through other “exogenous” improvements in the economic governance of the country. Second, and assuming we can unambiguously ascertain the direction of causality and size of effects between the “policy environment” index and the “partnership index”—and show for example, that improved partnership significantly causes improvements in the policy environment—then the next step is to ascertain the extent to which the policy environment impacts on the ultimate objectives—MDGs. In effect, partnership works on the ultimate objectives mostly through its impact on intermediate variables. Except of course where we can directly link development partnership to some of the Millennium Development Targets (which are intermediate outcomes), it may be difficult to make a direct pitch between partnership and the ultimate outcomes. A potential way out is to use the crude method of before-and-after approach whereby we evaluate the outcomes before and after the introduction of the CDF principles. But this would be a very long shot given the methodological problems of controlling for other variables, gauging the counterfactuals, etc. On the other hand, one could use the control-group approach of comparing the CDF countries with the non-CDF countries. Again, the methodological problems are messy, as we need to control for the initial conditions, etc. 6.8 A potentially rewarding approach is to construct the “partnership index” for individual countries, and then assess the relationship between changes in such an index (or set of indicators) and some measures of aid efficiency—cost savings resulting from partnership; increased resources for operations rather than administration; increased aid; reduction in the volatility of aid flows; increased targeting of the poor, etc. In other words, the overall impact of partnership on the ultimate objectives can be assessed in a rather indirect manner. 6.9 Furthermore, the construction of the index could have several modules, including a very important module that focuses on donor behavior. This, as Helleiner points out in the beginning quote, should be a very important contribution. So far, donors do the entire evaluation of both recipient country policies and donor countries' aid behavior themselves. Recipient countries neither sit at the DAC, nor are they involved in any evaluation of the donor behavior. OECD DAC evaluations are based on “peer reviews” that are not based on any standardized criteria. This contradicts basic elements of performance evaluation—whereby the institution rather than its clients evaluate it for performance. We suggest a change in approach whereby the evaluation of OECD DAC derives essentially from a survey of recipients. Among other things, such a sub-index of donor behavior not only provides donor agencies with a mirror to assess themselves, it also provides useful information to the recipients. Just like competitive ranking of the recipient countries based on the policy and institutional indices could put pressure on them to deepen reforms, the periodic publication and ranking of donors’ extent of implementation of the partnership principle could also provide an indirect pressure on 36 them to change their behavior. Such a donor ranking could be aggregated at the global level, but with survey results from individual recipient countries. This framework could be a very powerful instrument to push forward the implementation of the partnership principle. 6.10 We propose that the construction of the indicators should build upon the existing evaluations—to improve upon them and make them quantifiable. For example, a dramatic methodological change could be to structure the questions away from the “YES” or “NO” categories to rank-ordered responses as in the CPIA. Respondents could be asked to rank-order their responses on a scale of, say, 1 to 5 where 5 indicates “excellence” in partnership. That way, it would be easy to quantify the degrees of partnership, and also rank-order the recipients and donors. Some of the questions will no doubt require quantitative responses such as “how many formal partnership agreements have been entered into, extended or finalized?”. Coding such responses, as well as others would require a sense of benchmarking. Out of the total number of aid agencies operating in the particular country, should one hundred percent be ranked 5, or should 81-100 percent be ranked 5, while 61-80 be ranked 4, and in that order? These are some of the kinds of issues that require judgment calls in the construction of the indicators. 6.11 There are certainly several kinds of questions and issues that could become part of the indicators. In the foregoing section we reported on the results of the survey of donor and recipient country views and behavior with respect to a number of the issues germane to making some inference regarding the implementation of the partnership principle. The results were mostly qualitative in nature and designed to get some “rough” idea of the extent of the implementation. Overall, the results of the survey provided useful insights and also strengthened the case for improving on the kind of empirical evidence provided in this study to construct quantitative indicators. Below, we provide a skeletal listing of some issues/questions that might be part of the indicators, according to the following groupings: Institutionalization of Partnership • How many formal partnership agreements have been entered into, extended or finalized? • Existence of government department with responsibility to coordinate all aid relationships in the country. • Capacity and effectiveness of such government aid agency. • Extent of donor support in strengthening the capacity and effectiveness of the government aid agency. • Regularity of meetings of donors and government aid agency. • Standardization of donor reporting formats. • Regularity and comprehensiveness of donor reporting on their activities to the recipient country governments. • Movement from project aid to program aid. • Proportion of aid going through government budget.8 8 Helleiner (2000) observes that in Tanzania, where strenuous efforts have purportedly been made to transfer “ownership” of development programs from aid donors to the government, only 30 percent of 37 • Proportion of tied to total aid. • Number of joint reports and mission analysis among donors. • Number of “common funding pool” established. Recipient Country Policy and Institutional Stance • Existence and quality of poverty reduction strategy. • Alignment of development strategy with the budget or medium-term expenditure framework. • Institutional capacity for monitoring and evaluation. • Commitment and effort in implementation - extent to which policymakers meet intermediate targets within their control. Donor-Country Aid Policies and Procedures • Policy coherence among different policies affecting developing countries. • Alignment of assistance with recipient’s development strategy. • Donor priorities in country being set at headquarters or driven by country specific requirements. • Predictability of aid flows over a medium-term planning frame. • Alignment of assistance and disbursement to recipient countries' budgeting and expenditure cycles. • Degrees of shortfall of ODA from the commitments or promises. • Tying of procurement. • Extent of use of recipient country experts as opposed to donor country experts. • Extent of availability of contingent and compensatory financing against shocks. • Growth in donor country commitments/disbursements. • Extent to which donors meet the agreed financing gap in the recipient’s budget. • Extent of adjustment of donor administrative and financial accounting requirements to the recipient’s procedures. • Extent of “new” legislation, guidelines and rules requiring increased partnership and coordination. • Extent of internal reorganization of donor agency to respond to the new requirements of partnership. 6.12 As indicated earlier, the above listing of variables are indicative and expected to motivate ideas and debate towards the eventual construction of composite indicators of partnership. It would, for instance, require the development, pre-testing and administration of questionnaires to both donors and recipients. Such an exercise is beyond the scope of the present study. ODA was estimated to flow through the government budget in fiscal year 1999. 38 7. Conclusions 7.1 One of the key questions motivating this study is whether the introduction of the CDF in 1999 has changed the underlying aid relationships and implementation, especially in relation to country-led partnerships. Our key finding, both from the review of existing literature and the survey of selected donors and recipients, is that there have been improvements in aid delivery process towards better coordination, country ownership and partnerships, but that these improvements have not been systematic nor can they be unambiguously attributed to the introduction of the CDF. There is the question of attribution. To draw any robust conclusions about the impact of the CDF alone, a larger sample and a control group would have been needed. Prior to the introduction of CDF, and even after its introduction, there have been other initiatives to promote greater coordination and country ownership including for example, OECD-DAC "Shaping the 21st Century Principles and Guidelines," and the PRSP process. No doubt, the CDF and PRSP have led to more active participation by the Bretton Woods institutions in aid coordination. 7.2 The literature on partnerships suggests that effective partnership be based on trust, attitude and contractual relationship. However, in practice, the kind of partnership envisaged under the aid relationships - to be "country-led" in the sense of the recipient countries being on the drivers' seat - is beset with a number of challenges. We have framed some of the major issues in the form of hypotheses, and the inferences on them from the survey results have been insightful. 7.3 First, we argued that the asymmetrical power relationships between donors and recipients could impact on country-led partnerships. That is, the higher the share of donor resources in the recipient country, the greater the donor power and hence the more difficult it is to truly have country-led partnership. Overall, the surveys confirm this hypothesis. However, development towards a more balanced relationship can be noticed in some aid dependent countries such as Uganda, with, for instance, local governments increasingly taking the lead in aid coordination. CDF has probably contributed to such adjustments in the CDF pilot countries, even though it is still problematic to fully deduce the role CDF has played compared to other factors. A more fundamental change in the relationship will depend mainly on transformed donor behavior. 7.4 Most donors and recipient countries believe that aid dependency and indebtedness have a negative influence on the quality of partnership. Aid dependency can, however, under certain circumstances also contribute to more constructive engagements between a developing country government and donors. Several countries have established in– country aid coordination mechanisms. Further use of program and budget support will most likely result in improved coordination. Donors and recipient countries alike as the main obstacle to government-led coordination have identified lack of local institutional capacity. 7.5 The second hypothesis pertains to donor legislation, guidelines and enforcement. In essence, it is contended that partnership on the ground is hampered by disparate donor country and institutional legislation, guidelines and rules guiding aid delivery. All donors 39 in the survey claim that they have accepted the country-led partnership principle, but only two donors have changed their legal framework governing aid relationships. It appears that donors, by gradual changes in their working methods, are increasingly accepting the CDF principle of country-led partnership. Most donors have introduced changes in their administrative rules and procedures governing development assistance and restructured their organizations in order to respond to the new aid relationship. Almost no donor has, however, established any explicit exit strategy guiding its withdrawal from a partnership. Accordingly, current changes in donor behavior seem less grounded on legislation than on changes in working habits. The only tangible evidence of changed behavior based on legal grounds is the untying of aid to the Least Developed Countries by the year 2002 that was approved by OECD DAC in May 2001. The question thus remains whether legislative changes are needed in order for country-led partnership to take full effect in changes in donor behavior? 7.6 Third, the survey results support the hypothesis that the principle of country-led partnership is not an enforceable principle but rather a "moral issue" in which only converts apply it, and also determine the extent to which they do so. Without any possible sanction against non-compliance it is difficult to enforce it. Recipients have very few, if any, instrument to apply sanctions against non-complying donors. One possibility to create or strengthen that ability would be to introduce peer reviews where donors and recipients jointly review a specific donor. 7.7 Donors are increasingly moving away from projects towards program support, sector-wide approaches and pooled funding mechanisms. Still, the large majority of donor-funded activities are project-related. There is no tendency to use program aid more frequently in CDF countries compared to non-CDF countries. The determining factor is the existence of a poverty reduction strategy or an economic reform program rather than the CDF. The survey also indicates that recipient countries prefer donors to provide program and budget support instead of project support. 7.8 Technical assistance, according to the recipients, continues to be supply-driven, hampering true ownership and the use of local capacities. Donors claim that they are changing the direction of their Technical Assistance by abandoning freestanding TA projects, and by integrating TA into projects and programs. In countries where such adjustments are taking place, they can be linked to the introduction of SWAps and the shift towards more budget support. Donors still use Project Implementation Units, to a large extent, for carrying out programs and projects, even in countries with improved aid coordination like Uganda. 7.9 A number of initiatives have been taken among bilateral donors and multilateral institutions to harmonize aid delivery procedures. So far the main emphasis of the work has been to find common international standards and principles at the institutional agency level, rather than to adapt donor behavior to the procedures given in individual developing countries. Furthermore, recipients find that donor headquarters' policy hinder the development of country specific harmonization. In the end it is progress in individual countries that prove that the efforts are effective and successful. Some progress can be noticed in a few countries where full SWAps or similar joint programs are in effect. 40 7.10 Donors have only in a limited number of cases replaced own reporting requirements with regular government reports. Most evaluations undertaken are still agency related, assessing the partial impact of a particular donor’s efforts. Joint monitoring and evaluation of development programs rarely take place. Where they exist, they are linked to the implementation of SWAps or budget support. 7.11 Policy coherence issues are increasingly being discussed at the policy level among the donor countries. Nevertheless, almost no country has implemented any changes in their overall policy framework to take development concerns into account when forming policies in other areas than development assistance. 7.12 Finally, the analysis support the hypothesis that improved partnerships are generally associated with better developing country policy design and implementation. Causality could be either way, i.e. improved partnerships are easier in countries with better governance, policies and institutional structure. On the other hand, institutionalizing partnerships can strengthen institutional and governance structures, which impact on policy choice and implementation. 7.13 Local government leadership in management and coordination of aid resources has been strengthened in a number of developing countries over the last three years. The survey also indicates that the cooperation among different local actors in developing countries have increased as a result of the introduction of the CDF principles. Almost all donors have, according to the responses, aligned their country assistance programs with the CDF countries’ development strategies (PRSP). Most donors find, contrary to the opinions of the recipients, that the partnership principle is viewed by stakeholders in the recipient country as imposed by donors. 7.14 Recipients and most donors find that the implementation of country-led partnership principle will lower the transaction costs of administering aid programs. Weak local capacity in recipient countries may impede the potential cost savings. Expenses may also shift away from the governments towards the donors. 7.15 Thus, the factors that hamper country-led partnerships can be summarized to include: Among Recipients • Insufficient or uncertain political commitment to poverty reduction/sustainable development (lacking ownership). • Limited implementation capacity. • Low standards of governance, including lack of transparency, weak monitoring and accountability mechanisms. • Aid dependence. Among Donors • Lack of political will and legal-institutional framework to accept country-led strategies. 41 • Preference to maintain traditional, and bilateral, patterns of interaction. • Distrust of government accountability mechanisms. • Differing donor agendas and desire to "plant the flag." • Frequent policy changes. • Lack of aid coordination. • Lack of harmonization of donor practices. • Provision of supply-driven Technical Assistance and continued use of Project Implementation Units. • Tying of aid. • Limited field presence. 7.16 Furthermore, some other broad observations can be deduced from the analysis as follows: • General policy declarations about country-led partnership play a limited role unless they are followed up by new legislation or guidelines, supported and enforced by agency leadership. One good example is the OECD DAC Recommendation on untying of aid. • Practical application of common aid instruments (SWAps, budget support) has had a great impact on donor behavior. Changes in performance have been more linked to the implementation of such aid instruments than to general policy declarations about the benefits of country-led partnership. The real test on donor commitment to country led partnership takes place at the country level when aid strategies are transformed into practical use. • The wish by recipient countries to receive aid in the form of program and budget support will help change donor behavior. Donors would, by necessity, change their conduct if, and when, they redirect their funding from individual projects to common pools, and let go of agency specific solutions. • Increased use of program and budget support will promote country-led partnership through: - improved aid coordination - harmonization of donor procedures - joint monitoring and evaluation - restructured Technical Assistance • Peer pressure among donors works in changing policy directions. The efforts by OECD DAC and the Utstein group,9 among others, to promote country-led partnership have considerably influenced donor policies. Continued work to bridge the gap between policies and practical implementation is still needed. • Recipient countries need to be closely involved in the development of donor policies. Too much policymaking is made in closed donor circles. Effective partnerships require that donors and recipients together form the future aid relationship. Recipients should, for instance, be part of the OECD DAC peer reviews of individual donors. 9 Informal donor group consisting of Germany, the Netherlands, Norway and United Kingdom. 42 7.17 Consequently, the following main recommendations are made for donor actions to support country-led partnerships. Donors Can Best Promote Country-Led Partnership by: • Responding to local government leadership. • Aligning donor assistance with government strategy. • Accepting less high-profile programs and projects. • Supporting capacity building. • Being transparent and providing timely information. • Increasing the proportion of program and budget support (pooled funding). • Providing demand-driven Technical Assistance. • Ending the use of Project Implementation Units. • Improving aid coordination. • Conducting joint analytical work (with local governments and among donors). • Harmonizing donor procedures and practices. • Untying aid. • Improving policy coherence. • Increasing field presence and delegating more authority to country offices. 7.18 Finally, the study argues that partnership and indeed the CDF principles should aim to impact on the MDGs. It is not enough to preach the principles, progress towards greater effectiveness would require a mechanism for performance monitoring and evaluation. The stakeholders—donors and recipients—must have a way of monitoring the degree of implementation of the partnership, and also over time, the impacts of the principles on the ultimate objectives—the MDGs. In Section 6 of the paper, we proposed a framework for thinking about the development of effective performance evaluation indices. Such indices, consisting of about 30 different variables are grouped into three major headings: • Institutionalization of partnership. • Recipient country policy and institutional stance. • Donor country aid policies and procedures. 7.19 The road to effective country-led partnership might be long, but the journey is doable. It requires not only political commitment to reforming the existing institutions and entrenched behaviors, but more so, continuing technical work. In other words, the agenda for country-led partnership is largely a work-in-progress. 43 References - Arce, M., Daniel G. 2001. Leadership and the aggregation of international collective action, Oxford Economic Papers 53 (2001), 114-137. - Cotonou agreement. 2000. ACP-EU Partnership agreement signed in Cotonou on 23 June, 2000, Supplement to the Courier, Special Issue, September 2002. - Devarajan, S., Dollar D., and Holmgren T. 2001. Aid and Reform in Africa – Lessons from Ten Case Studies, World Bank, Washington, D.C. - Development Committee. 2002. Harmonization of Operational Policies, Procedures and Practices: Second Progress Report, Document SecM2002-0143, March 7, 2002, Washington, D.C. - ____________. 2001a. Communiqué, April 2001. - ____________. 2001b. Harmonization of Operational Policies, Procedures and Practices: Experience to Date, document DC 2001-0006, April 13, 2001, Washington, D.C. - Easterly, W. 2002. The Cartel of Good Intentions: Bureaucracy versus markets in foreign aid, Working Paper Number 4 (revised May 2002), Center for Global Development, Washington, D.C. - ____________. 2001. The Elusive Quest for Growth: Economists' Adventures and Misadventures in the Tropics, Cambridge, MA: MIT Press. - Foster, M. and Naschold, F. 2000. Expenditure Framework and Partnership, World Bank OED Working Paper Series No 9, World Bank, Washington, D.C. Summer 2000. - Hanna, N. and Agarwala, N. 2000. Toward a Comprehensive Development Strategy, World Bank OED Working Paper Series No 16, World Bank, Washington DC, Summer 2000. - Helleiner, G. 2000. “Towards Balance in Aid Relationships: Donor Performance Monitoring in Low-Income Developing Countries”. Draft. Paper prepared for Festschrift in honor of Lance Taylor. - Maxwell, S. and Conway, T. 2000. Perspectives on Partnership, World Bank OED Working Paper Series No 6, World Bank, Washington, D.C. Summer 2000. - Ministry for Foreign Affairs. 2001. Making Globalisation Work for the Poor- Conference Report, Kramfors, Sweden. 20-21 June 2001. 44 - ____________. 1999. Making Partnerships Work on the Ground, Workshop Report, Ulvsunda Castle. Stockholm, Sweden. 30-31 August 1999. - OECD DAC (Development Assistance Committee of the Organization for Economic Cooperation and Development). 2001a. DAC Guidelines on Poverty Reduction, OECD DAC, Paris. - ____________. 2001b. Task Force on Donor Practices, Progress Report, December 2001, www.oecd.org/dac/donorpractices. - ____________. 2000, Partnership for Poverty Reduction: From Commitment to Implementation, Statement by the DAC High Level Meeting, Paris, 11-12 May 2000, OECD DAC, Paris. - ____________. 1996, Shaping the 21st Century, The Contribution of Development Co-operation, OECD DAC, Paris, May 1996. - OED (Operations Evaluation Department). 2001a. 2000 Annual Review of Development Effectiveness, From Strategy to Results, World Bank, Washington, D.C. - _____________. 2001b. The Drive to Partnership, Aid Co-ordination and the World Bank, World Bank, Washington, D.C. - _____________. 1999a. 1999 Annual Review of Development Effectiveness, World Bank, Washington, D.C. - - Picciotto, R. 1998. The Logic of Partnership, A Development Perspective, World Bank, Operations Evaluation Department, Photocopy. - Rischard, J.F. 2002. High Noon, Twenty Global Problems, Twenty Years to Solve Them, Basic Books, New York. - Sandler, T. 2000. On Partnerships, CDF and Collective Action, World Bank, Operations Evaluation Department, Photocopy. - Soludo, C. C. 1998. “Monitoring Economic Performance in Africa: Towards the Construction of Performance and Policy Stance Indices”. A methodological Note for the construction of the UN-ECA Indices. Mimeo. UN-ECA, Addis Ababa. - Wolfensohn, James D. 1999. Proposal for a Comprehensive Development Framework, World Bank, Washington, D.C. January 21, 1999. - World Bank, CDF Secretariat. 2001a. Comprehensive Development Framework, Implementation Experience in Low- and Middle-Income Countries, Progress Report, World Bank, Washington, D.C. April 26, 2001. 45 - _____________. 2001b. Comprehensive Development Framework, Meeting the Promise? Early Experience and Emerging Issues, World Bank, Washington, D.C. August 27, 2001. - _____________. 2001c. Comprehensive Development Framework, Questions and Answers, March 2001. - _____________. 1999. Comprehensive Development Framework, Country Experience, March 1999 – July 2000, World Bank, Washington, D.C., September 2000. - World Bank, Partnerships Group. 1998a. Partnership for Development: Proposed Actions for the World Bank, Partnerships Group, Strategy and Resource Management, World Bank, Washington, D.C. May 1998. - World Bank, Partnerships Group. 1998b. Partnership for Development: From Vision to Action, Partnerships Group, Strategy and Resource Management, World Bank, Washington, D.C. September 1998. 46 Terms of Reference Multi-Stakeholder Evaluation of the Comprehensive Development Framework (CDF) Terms of Reference for Country-led Partnership Thematic Study November 2001 Background 1. The development assistance system has expanded far beyond the scope and structures envisioned by its architects at the mid-point of the last century. Yet, in spite of notable successes in some countries and regions, poverty has increased and concessional flows have shrunk, even as demands for official resources have multiplied. Public concern with the efficacy of aid has grown. It was in this context that the Comprehensive Development Framework (CDF) was articulated by the World Bank president in January 1999. 2. The CDF comprises a set of principles that a number of developing countries and development assistance agencies have been seeking to put into practice over the last two years. The principles on which the CDF is based are: long-term, holistic development framework; country ownership; country-led partnership; and results orientation. The evaluation is supposed to ascertain the extent to which CDF principles have been practiced and whether their implementation has been oriented to pursuing overarching development objectives, such as the Millennium Development Goals (MDGs). The evaluation will deal with the relevance, efficacy, and efficiency of the overall development assistance system in selected countries, both on the ground and at the policy level, including linkages to the MDGs. 3. The CDF principles are intended to help resolve persistent problems in the development assistance system, by: • Bringing about a better balance between macroeconomic and financial issues, on the one hand, and structural, social, and human dimensions, on the other hand; • Shifting from donor-driven development processes to full ownership by the country; • Eliminating ineffective coordination marked by aid poorly aligned with country strategies, wasteful competition among externally-funded programs, and burdensome donor procedures, and; • Using results information to improve the performance and accountability of aid programs. 4. The evaluation will be conducted over a 12-month period and will be completed by January 2003. It consists of the following main components: • Case studies of five CDF pilot countries (including Bolivia, Ghana, Romania, Uganda, and Vietnam) and one non-CDF pilot country (Burkina Faso); • Four thematic studies and a cross-cutting econometric study; and 47 • A synthesis report. 5. Country-specific and thematic studies are to be guided by the same conceptual model. This will enhance cross-fertilization and comparability, though there ought to be room for useful differentiation in terms of methodology of analysis, and tailor-made features. The thematic and country studies are being implemented in an iterative manner. Three country studies (Bolivia, Uganda, and Vietnam) are being implemented as soon as possible and their initial results will inform the thematic studies. Similarly, initial results of the thematic studies, which will be desk studies, will be shared with the country case study teams for potential in-depth exploration. The remaining three country studies (Burkina Faso, Ghana and Romania) will be conducted and completed by March 30, 2002. 6. The draft findings and conclusions of the synthesis report will be the subject of a major dissemination conference in November 2002 and will be presented to the Evaluation Steering Committee and the World Bank’s Executive Directors in January 2003. Objectives and Scope of Work 7. The objectives of the evaluation are to: identify the factors that have facilitated implementation of CDF principles and those that have hindered it; assess the extent to which CDF implementation has affected intermediate outcomes, and to the extent possible, longer-term development outcomes, such as the Millennium Development Goals (MDGs); promote learning and capacity development in countries where CDF principles are being implemented; and create an ongoing mechanism for periodic assessment of CDF processes and feedback to decisionmakers. 8. The five tables of “core evaluation questions” (attached) have been framed with these objectives in mind. The tables provide an overall conceptual framework for guiding the analysis, especially in terms of the formulation of evaluation questions and identification of, or need to construct, measurable performance indicators. 9. The thematic study will be done in two phases. The first phase of the analysis, which is mainly a desk study, should focus on donors and other development agencies, especially in terms of assessing the extent to which their behavior has changed sufficiently to enable the implementation of CDF principles to make an impact at the country level. In addition, this phase also includeS impact analysis in the context of the cross country experience using secondary data. The next interrelated phase for the thematic papers should be to expand the focus to examine issues related to country- specific aspects of the CDF, where the country case study experiences could be used for sharpening the application of the “logframe” and the “balance-sheet” frameworks in the formulation of hypotheses (also attached). 48 10. The study should therefore be based on the following broad terms of reference: First phase: 1) Undertake an analytical review of the newly emerging consensus (on the concepts of CDF) on development and the role of the Country-led Partnership in the new development agenda. 2) Expand and elaborate upon the “balance sheet” list (Box 1) of the potential positive and negative consequences of alternative ways of implementing the Country-led Partnership on the processes/intermediate outcomes as well as development goals. 3) Assess the quality (nature) of development aid and the extent of change in donors’ behavior, following the recent consensus on the CDF-like development strategies. 4) Formulate relevant hypotheses on the potential impact of the quality (nature) of development aid on the capacity of aid-receiving countries in implementing the Country-led Partnership. 5) If possible, build or identify cross-country indicators (from available data or sources) on: the extent of implementation of the Country-led Partnership, the quality (nature) of policies and institutions as a proxy for processes; and assess the appropriateness of these indicators for hypotheses testing. Undertake basic cross-country analysis utilizing the data, focusing on the Country-led Partnership.10 6) If possible formulate relevant hypotheses on the potential impact of CDF implementation principles (on processes and outcomes) and to the extent feasible, examine the possible trade-offs due to the Partner-led Partnership. Second phase: 1) Drawing from country experience refine cross-country indicators developed in the first phase as well as cross-country econometric analysis. 2) If possible, use the case of a particular country (for which broadly similar data may be available: e.g. from a PRSP or SPA-based survey) to assess the role of the Country-led Partnership in the development discourse of the chosen country on how they affect development goals. 10 Cross country is used liberally to mean comparative analysis with varying degrees of country coverage and methodologies. 49 Box 1: Evaluation Questions: Country-Led Partnership (Design Paper for a Multi-Partner Evaluation of the CDF, September 2001, Table 3.) What has been done to implement this principle and when, how, and why did it happen? 1. What changes have occurred in the alignment of external aid with country development strategy? When, how, and why? 2. What changes have occurred in country-external partner relations re: analytical and advisory activities, policy dialogue, and participatory approaches? When, how, and why? 3. What changes have occurred in selectivity in aid provision so as to avoid overlap and duplication? When, how, and why? To what extent has the comparative advantage of each partner been taken into account? 4. What changes have occurred regarding country leadership in the coordination of donor activities? When, how, and why? 5. What changes have occurred regarding harmonization of donor procedures and practices (e.g. reporting, budget cycles, procurement, disbursement, evaluation, safeguards)? When, how, and why? What are the obstacles? 6. What role has the country played in harmonization efforts? Local donors? What actions can only be taken by donor headquarters? 7. To what extent and how well have sector groups served as mechanisms to improve aid effectiveness (e.g. increase coherence, selectivity, harmonization)? What have been the intermediate results in terms of changes in behaviors and practices? 8. How has quality of partnership influenced, and been affected by, the extent of indebtedness and by aid dependency? 9. How effective are in-country forums for partner coordination in terms of formulating medium-term strategies and resource mobilization? How do partners perceive the benefits and costs? 10. To what extent have partners recognized their responsibilities and accountabilities through such mechanisms as joint reviews of their performance? 11. How has intensified partnership affected consultations by major donors with other partners? 12. How has intensified partnership between government and external partners affected relations among levels of government, civil society, and the private sector (e.g. associations, philanthropic organizations? 13. To what extent has aid strengthened or undermined country capacity to manage and coordinate aid? 14. What impact has implementation of the partnership principle had on aid delivery transaction costs?