A multi-partner evaluation of the Comprehensive Development Framework (CDF) : Ghana case study

82109 A Multi-Partner Evaluation of the Comprehensive Development Framework Evaluation of the Comprehensive Development Framework (CDF) Ghana Case Study 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 Acknowledgments ............................................................................................................ iv Abbreviations ................................................................................................................... vi Executive Summary ....................................................................................................... viii 1. Introduction..................................................................................................................1 2. The CDF Principles in Ghana - the Context and the History......................................3 Poverty in Ghana...................................................................................................3 Political Setting.....................................................................................................4 The Economic Reforms ........................................................................................6 Visions and Policy Frameworks ...........................................................................8 Social Sector Reforms.........................................................................................10 Agriculture ..........................................................................................................12 Public Sector Capacity and Public Sector Reforms............................................12 Aid Dependency..................................................................................................17 Introduction of the CDF in Ghana ......................................................................19 Summary.............................................................................................................21 3. The CDF and Ghana's Poverty Reduction Strategy...............................................23 The Evolution of the GPRS ................................................................................23 Long-term Holistic Development Framework....................................................26 Country Ownership.............................................................................................28 Partnership and Donor Behavior.........................................................................31 Results Orientation..............................................................................................33 4. Long-term Holistic Development Framework ........................................................37 5. Country Ownership ...................................................................................................41 Health..................................................................................................................42 6. Country-led partnership ...........................................................................................47 Country-led Strategic Coordination and Alignment of Donor Assistance .........47 Coordination of Concrete Assistance, Multi-Donor Approaches .......................49 Health..................................................................................................................49 Implementation Modalities, Project Implementation Units, Incentives .............53  iv Transaction Costs................................................................................................55 Decentralization of Donor Agencies...................................................................55 7. Results Orientation ....................................................................................................59 Health..................................................................................................................59 Economic Governance ........................................................................................60 Agriculture ..........................................................................................................61 8. Conclusions, Key Issues and Challenges..................................................................63 Annex 1: List of People Met............................................................................................67 Tables: Table 1: Incidence of Poverty by Region and Location in the 1990s..................................3 Table 2: Child Nutrition Status ............................................................................................4 Table 3: Macroeconomic Indicators 1983-2000 - Annual Averages...................................7 Table 4: Social Sector Indicators .......................................................................................11 Table 5: Comparative Aid Dependency Ratios (2000)......................................................17 Table 6: Distribution/Purpose of Aid Commitment to Ghana ...........................................19 Boxes: Box 1: Milestones in Ghana's Political Economy................................................................5 Box 2: Visions and Poverty Reduction Strategies in Ghana................................................9 Box 3: The Nettle of Public Sector Pay Reform in Ghana ................................................15 Box 4: Factors Contributing to the Failed Attempts of Public Sector Management Reforms..................................................................................................................16 Box 5: The "Work Out" .....................................................................................................21 Box 6: The Medium-Term Priorities and the GPRS Compared ........................................25 Box 7: The GPRS and the Millennium Development Goals .............................................35 Box 8: The Sector Wide Approach (SWAp) in Health - an Overview of Achievements .37 Box 9: Why the SWAp worked…A Donor's View ...........................................................49 Box 10: Insider's View of the Health SWAp.....................................................................50 Box 11: The Agriculture Sector Wide Approach Attempt ................................................51 Box 12: Delivery Mechanisms to Rural Communities ......................................................53 Box 13: The Challenge of Change.....................................................................................54 Charts: Chart 1: Total Program and Project Aid Disbursements ...................................................18 Chart 2: Program Aid.........................................................................................................19 Chart 3: Project Aid ...........................................................................................................19  v Acknowledgments The case study team would like to express its gratitude to the individuals whom they interacted with in Ghana, for their time and insights, and for the hospitality extended during the team’s District visits. Mr. Philip Owusu and Dr. Joe Amoako-Tuffour, of the Institute for Economic Affairs, provided very useful support in identifying key informants, and Ms. Genevieve Eba- Polley, Center for Policy Analysis, and Ms. Paula M. Lamptey, the World Bank Office in Accra, were invaluable in organizing meetings for the team, collecting documents, and arranging the final workshop. The workshop held at the end of the study generated very useful feedback from a wide array of stakeholders. The team appreciates especially the frank discussion with the Minister of Planning and Regional Development, Dr. P. Kwesi Ndoum. A short report from the workshop is included as an annex to this report. The study team is responsible for the report, including any errors or omissions. The team members worked individually on the preparation of the different chapters of the report, which was afterwards compiled and adjusted by the team leader. The team was composed of Nils Boesen, consultant (team leader); Anthony Killick, consultant (Overseas Development Institute, London); Laura Kullenberg, World Bank, Operations Evaluation Department (OED); Mirafe Marcos, OED; Abena D. Oduro, consultant (Center for Policy Analysis, Accra); and Julia Ooro, OEDCM, provided administrative support.  vi Abbreviations AAGDS Accelerated Agricultural Growth and Development AfDB African Development Bank AGSSIP Agricultural Services Sector Investment Program BPEMS Budget and Public Expenditure Monitoring System CAGD Controller and Accountant General’s Department CAS Country Assistance Strategy CDF Comprehensive Development Framework CEPA Center for Policy Analysis CG Consultative Group CIDA Canadian International Development Agency CSO Civil Society Organization CSPIP Civil Service Performance Improvement Program DACF District Assemblies Common Fund DFID Department for International Development ERP Economic Recovery Program FAO Food and Agricultural Organization GDP Gross Domestic Product GoG Government of Ghana GPRS Ghanaian Poverty Reduction Strategy GSS Ghana Statistical Service HIPC Heavily Indebted Poor Countries IFI International Financial Institution IMCRP Inter-Ministerial Committee on Poverty Reduction IMF International Monetary Fund MDA Ministries, Departments, and Agencies MDG Millennium Development Goals MLGRD Ministry of Local Government and Regional Development MoF Ministry of Finance MoH Ministry of Health MTEF Medium Term Expenditure Framework NDC National Democratic Congress NDPC National Development Planning Commission NGO Non-Governmental Organization NIRP National Institutional Renewal Program NPP New Patriotic Party OED Operations Evaluation Department PAMSCAD Program of Actions to Mitigate the Social Costs of Adjustment PER Public Expenditure Review PIU Project Implementation Unit PNDC Provisional Nation Defense Council PRSP Poverty Reduction Strategic Paper PUFMARP Public Financial Management Reform Program SAPRI Structural Adjustment Participatory Review Initiative SWAp ` Sector Wide Approach UK United Kingdom  vii UNDP United Nations Development Program UNICEF United Nations Children Education Fund US United States VAT Value Added Tax VIP Village Investment Project WB World Bank  viii Executive Summary 1. This report presents Ghana’s experience in implementing with the Comprehensive Development Framework (CDF) principles, based on an evaluation conducted by an international team in April-May 2002. The CDF concept draws together a set of principles in one package – long-term holistic development framework; country ownership; country-led partnership; and results orientation – principles that are distilled from development experience over the last five decades. The evaluation attempts to identify factors that favor or inhibit the way in which the principles are implemented as well as to extract lessons learned about their validity and applicability. 2. Ghana was selected as a case study for several reasons, including its status as a “CDF pilot country” (as agreed by Ghanaian and World Bank leadership in 1999) and the substantial aid flows to Ghana. The country study team sought the views of government officials at central and local levels, parliamentarians, civil society and private sector stakeholders, and local donor representatives. In addition to general “CDF-issues” about the relations between national and international actors, the team chose to focus on several specific sectors or themes: the Ghanaian Poverty Reduction Strategy (GPRS), health, agriculture, decentralization and public financial management. The study team focused on two main questions (1) has the designation of Ghana as a CDF pilot country caused improvements as the CDF principles have been implemented? and (2) have contemporary developments in Ghana been consistent with the CDF philosophy and what it seeks to achieve? Context 3. Ghana has on several accounts done well in the last decade. Most significant, the country adopted a parliamentary democratic system in 1992 and has held three multi- party elections, the last (in December 2000) resulting in the peaceful and orderly transition of power from one party (National Democratic Congress, NDC) to another (New Patriotic Party, NPP). After a long, stagnant period in the 1970s and the early 1980s, the economy resumed growth with average annual growth levels around 4 -5% in the last nearly 20 years. Poverty levels have decreased (from 52% in 1991/92 to 40% in 1998/99). Though average per capita income (US$390) remains lower than at independence, most of Ghana’s social indicators compare favorably with the averages for Sub-Saharan Africa. 4. Ghana has since the early 1980s implemented a number of far-reaching but gradualist reforms. They have included exchange rate and trade reforms, dismantling of extensive price and distribution controls, and divestiture in telecommunications and banking. 5. There are, however, a number of considerable challenges ahead. Though some level of macroeconomic stability was achieved in the previous decades, economic growth has occurred more because of public than private investment, the former largely financed through unsustainable domestic borrowing levels and generous donor assistance,  ix crowding out private borrowing and pressing interest rates upwards. Further, economic growth and poverty reduction have been very unevenly distributed, concentrating around Accra while other regions, particularly those in the north, have experienced an increase in poverty levels. 6. The previous government (NDC) did not complete the comprehensive reform agenda. The privatization program has, in ten years, only covered 25% of public enterprises (by value). Public sector employment remains very high, and civil service reform, including re-establishment of proper incentives for performance, has not advanced. Comprehensive and costly initiatives to enhance public sector financial management have not produced the desired results. Decentralization was put on the agenda in 1988 with the creation of District Assemblies, but the policy framework for decentralization is not entirely consistent, and progress has been limited. 7. The new government (NPP), which took office in January 2001, was elected on a liberal platform promoting a free enterprise economy as the guarantee for growth and prosperity. In addition to the unfinished reform agenda, the government also inherited a serious macroeconomic imbalance, fostered by excessive election-year spending, and sharply deteriorating terms of trade in 1999, when cocoa and gold prices dropped and oil prices increased. The new government has succeeded in reversing inflationary tendencies through prudent fiscal management and a tight monetary policy, thereby reducing inflation from a peak 42% when it took office to 21% by the end of 2001, and with a likely year-end scenario of 10% for 2002. 8. Some steps have been taken to revitalize the reform agenda in key areas like divestiture, utility pricing, public financial management, and civil service reform. However, it is too early to see concrete results of these initiatives, and the strength of the government’s commitment to reform in these highly sensitive areas is still untested. 9. Ghana continues to be highly dependent on foreign assistance, which finances roughly 90% of public investment expenditures. Even a strong political commitment to civil service and public financial management reform will not develop the capacity of the public sector in the short run. Similarly, at a more fundamental level, the previously dominant patronage elements of the political system, though apparently weakened, will not disappear rapidly, and will continue to seek expressions in favoritism, ghost-workers in the public sector, and other forms of corruption. 10. The application of the CDF principles takes place in this context. In any context, their full application will often imply in-depth changes and restructuring within both donor and government agencies—changes that require political support, capacity, and continuous nurturing. In the context of aid dependency, weak public sector capacity, and an overloaded reform agenda, it is therefore especially important not to apply unreasonably demanding standards to any initial period. Introduction of CDF in Ghana 11. Ghana was designated a CDF pilot country in May 1999 by the World Bank. The immediate effects in Ghana were: i) the formation of 14 government-led thematic “CDF-  x groups” for coordination and dialogue between the government and donors, ii) the institutionalization of quarterly mini-consultative group meetings (mini-CGs), iii) hosting of regular CG meetings in Ghana, and iv) monthly working lunches between donor representatives. The CDF was strongly promoted by the World Bank office in Accra and accepted by the previous government as a means to strengthen collaboration across all the development partners. However, available evidence indicates that the CDF was first understood as a (matrix-type) framework with corresponding general co- ordination mechanisms, rather than as wider principles for behavior of and relations between national and international stakeholders. 12. Though useful, neither the CDF-groups nor the quarterly mini-CGs (which pre- dates the CDF) can claim to have had broad or significant impact on the behavior and procedures of government and donors on the ground. Other sector working groups, pre- dating the CDF and linked to specific programs, have proven more effective than the CDF-groups, some of which never became effective. The much-cited sector wide approach (SWAp) in health pre-dated CDF (but fed into development of CDF thinking). The profile of the CDF has receded with the launch of the PRSP/GPRS, however CDF thinking clearly influenced the global design of PRSPs, whose processes and activities were intended to embody the four CDF principles. In Ghana there is a low awareness of the CDF concept outside donor and some government circles, and to avoid any identification between CDF and the previous government, the World Bank office has, since the change of government, downplayed the use of the term. 13. The most significant institutional effect of launching the CDF has been to change the orientation and image of the Accra World Bank office, which has very actively promoted the principles and sought to work in a more collaborative manner both with government and with other donors. The Bank has, for example, redefined its Country Assistance Strategy as the Bank’s “business plan to support government projects and programs” in an attempt to move away from the notion of World Bank projects and programs. The Bank has further decentralized authority to the country office. The Country Director shifted most responsibility for task management to national staff, and reported significant improvements in the quality of the Bank’s portfolio as a result. The Ghana Poverty Reduction Strategy 14. The Ghana Poverty Reduction Strategy (GPRS) has become the centerpiece process and product, which today embodies the CDF principles at the macro-level. It has attracted strong attention from all donors, and the draft GPRS was the key document in the most recent CG meeting in Accra. 15. Ghana’s GPRS was preceded by other attempts at comprehensive planning, including “Ghana – Vision 2020, the First Step” (1995), the “Accelerated Poverty Reduction Strategy” presented to the 10th CG meeting in 1999, and an interim PRSP presented to the Bretton Woods institutions in July 2000. In the same month, the process of preparing the broader, more consultative GPRS was launched. Meanwhile, as elections approached, the envisaged preparation of a medium-term plan building on “Vision 2020” never materialized and the GPRS process came to a temporary halt as a new government took possession.  xi 16. The new NPP government entered with a long-term political vision expressed in the party manifesto “Agenda for Positive Change.” Notably, the manifesto makes no reference at all to Vision 2020, or to the GPRS process. But President Kufuor’s new administration was quick to say it would continue with the GPRS process, but wished to put its own stamp on it. Accordingly, the timetable was extended into 2002. As first priority, the new government took measures to re-establish macroeconomic stability. After only three months, it reversed previous policy by deciding to apply for HIPC debt relief—a development that increased the already considerable financial stakes attached to the GPRS. 17. Preparation of the GPRS had started in July 2000 and continued in a special Task Force within the National Planning Commission. Consultants (mainly Ghanaian) were hired to synthesize the inputs and produce a full GPRS for Cabinet discussion in late 2001. The document would then be submitted to the CG meeting in April 2002, by which time costings for the various investments in the GPRS were to have been finalized. 18. The drafting process was given very few resources and had initially limited contact with other significant policy processes in government. The Secretariat drafted a GRPS, which is very comprehensive, combines economic growth and social equity concerns, and includes a large part of the reform agenda that has been on the table for a decade. The GPRS demonstrates the need for wide-ranging change and multiple initiatives on many fronts, if satisfactory poverty reduction is to be achieved. The preliminary costing of the GPRS indicates a total cost of US$8.3 billion, roughly five times the estimated available US$1.6 billion over 3 years. The GPRS process thus generated an un-prioritized list of investments costing far in excess of realistic resource possibilities. 19. The GPRS is undoubtedly locally owned in the sense of being authored by Ghanaians. The consultative processes included civil society organizations, regional workshops and presentations to political parties and parliament. The majority of important donors, including the World Bank, tried to encourage national ownership of the GPRS by maintaining a clear hands-off position during the preparation period (though a few donors reportedly sought to ensure that their particular programs and projects were included). 20. The draft GPRS suggests that the consultative process may have constrained the taking of hard choices (e.g. there was little discussion of trade-offs between market-led growth and poverty reduction; or the need to address interpersonal inequalities). The GPRS can show that it responded to a wide range of comments offered but some policy issues (such as water privatization) were not part of the discussion. 21. Only in early 2002, when key donors made it abundantly clear to the government that the GPRS would be the exclusive reference document on which donors would base their pledges, did the GPRS move into the mainstream of government policy formation. Prioritization of the broad GPRS was essential and the Cabinet decided to use this scaling-down process to assert their own medium-term priorities, which proved to owe far more to the NPP’s election manifesto than it did to the GPRS. The result was a program, led by large investments in power plants and highways, that seeks to achieve an  xii accelerated growth led by a revived private sector but which is at odds with stated donor- preferences and international donor consensus about the key ingredients of “pro-poor growth.” 22. Considerable efforts have been made to link the GPRS process to the medium term expenditure framework (MTEF) budget process. In broad terms, reallocations in the 2002 budget were consistent with an increased focus on poverty reduction priorities, although the real implication of these reallocations (from a contingency fund) is unclear. The correlation between the envisaged medium-term priority expenditures of the GPRS, the “ordinary” budget expenditures, and the use of the HIPC-initiative funds is not yet clear. The preparation of the 2003 budget and three-year annual forecast will be a critical test of both the poverty reduction priorities of the government and the relevance of the broader GPRS document. 23. Given the limited reform capacity in Ghana, there is also a need to extend time horizons beyond the three-year GPRS and to sequence and prioritize reforms carefully in a longer time perspective. The forthcoming, constitutionally required medium-term plan is an opportunity to provide this as it spans a 3-5 year horizon, and the intention to develop a new long-term vision statement is to be welcomed in this perspective. Ghana’s previous experiences with visions and plans should, on the other hand, lead to rather modest expectations about the effectiveness of longer term holistic planning at macro- level. 24. The 1999 strategy, the interim PRSP of 2000, and the GPRS of 2002 were authored by the government, and in that respect, are expressions of country ownership. However, the incentives to prepare these strategies were linked to Ghana’s aid dependency and to donor demands, particularly those of the Bretton Woods institutions. Both the PRSP-I and the GPRS are requirements for getting access to financing mechanisms (PRGF and HIPC), and the 1999 strategy was largely prepared for the benefit of the World Bank Country Assistance Strategy preparation process. 25. The GPRS preparation process also illustrates the elusiveness of the ownership concept. On one hand, the superimposition of the new government’s medium-term priorities was an assertion of ownership on the part of the president and his Cabinet, and resulted in a strategy that was further away from the wishes of some major donors. On the other hand, the new priorities were asserted in an entirely non-consultative manner and in a way that much reduced the value of earlier consultations. Whereas the original text of the GPRS had a reasonable claim to be fairly widely owned by organized stakeholders within the country, it had weaker commitment from the government. Conversely, while the new priorities are strongly owned by the top political leadership they have fewer claims to broader ownership across society. 26. The donors have participated in various consultative meetings leading to the GPRS, though, as noted above, they have largely left the preparation of the GPRS to national actors. Generally, donors welcome the GPRS draft, but questions have been raised on the poverty orientation of the medium-term priorities. The impact on the quantity and composition of aid is still unclear. The broader GPRS version will on one  xiii hand allow the inclusion of virtually any donor-funded activity, while the medium-term priorities may meet considerable difficulties in attracting funding. 27. The GPRS includes specific poverty reduction targets for the coming three years. However, the available capacity to implement the plan and deliver results is a key constraint, also in relation to the medium-term priorities. The GPRS itself mentions that the strategy will most likely fail without civil service reform—a reform that will need a decade to become effective. 28. The GPRS has clearly been an additional policy process driven in most respects by its link to external funding opportunities and timetables. Nor has the GPRS been well aligned with other democratic and constitutionally demanded processes. The government has been fairly consistent in sticking to its growth-based platform, which it believes to be the best way to poverty-reduction. This raises questions of the balance between and legitimacy of electoral processes and consultative/participatory processes respectively and how far broad consensus can and should be expected in the two processes. 29. Considering the history and context of the GPRS, the study team finds that neither donors nor national actors should focus exclusively on the GPRS or make unreasonable demands for its inclusiveness and resilience, nor build up special purpose monitoring schemes narrowly focusing on the GPRS. By doing so, they risk overlooking other important policy processes and institutional capacity development processes, and the GPRS needs to be seen as part of a wider scheme. Long-term holistic development framework 30. While the GPRS and other macro-level strategies intend to provide an overall development framework, efforts have also been made to establish holistic frameworks— or comprehensive policies—in various sectors and cross-cutting areas, including agriculture, health, decentralization and economic governance (these are the sectors and themes reviewed by the study team). Especially in health, the successive five-year sector strategies have served as framework for both government and (most) donor actions. In agriculture, the policies have been fairly stable, but actions have not been strongly aligned with the policy framework. In education, there is reportedly no comprehensive policy framework. 31. Efforts to enhance the comprehensiveness and coherence of reform in key areas (economic governance, civil service reform, decentralization, and divestiture) are urgently needed but yet to emerge. In recognition of this, government and aid partners have recently taken initial steps to intensify analytical work and dialogue in several areas (agriculture, economic governance, and decentralization). 32. In summary, and including the GPRS, there is progress toward a long-term holistic development framework in Ghana, but it is still limited. This is true whether a framework is understood as a policy/strategy framework or as an operational matrix for concrete actions. This limited progress invites caution about how to interpret and operationalize the holistic framework-principle in a situation where both national policy formulation and implementation capacity is limited.  xiv Country Ownership 33. It was popular pressure that led to the reintroduction of multiparty democracy in Ghana in 1992. Some rather timid attempts in participatory policymaking were made during the previous government, but the present government is undoubtedly marked by greater openness and tolerance, and greater commitment to consultation. Civil society and the private sector are more fully engaged in consultations at various levels, also outside the scope of the GPRS processes. 34. There is strong Ministry of Health ownership in Health SWAp and the draft SWAp II espouses more of a domestic partnership approach. The strength of Ghanaian leadership has been crucial. Leadership plus careful preparation (as well as like-minded aid partners who were willing to take risks) have been the key elements of the success of Health SWAp. In agriculture, the overall policy framework seems solidly anchored in the Ministry of Agriculture, while ownership in the area of institutional reform and economic governance reform is more doubtful. 35. Donors have generally been supportive of a shift toward country ownership at policy level (GPRS, the Health SWAp, support for CSOs & dialogue processes). Notably, donors have sought to build capacity in think-tank type NGOs to stimulate national policy dialogue. 36. However, by general agreement, the quality of the domestic dialogue leaves room for improvement. Documentation is inaccessible or arrives late (leaving insufficient time for consideration), or is not available in languages or concepts that can be easily understood by the general populace. There is a tendency toward formalism—to read ‘consultation’ as dissemination and validation of preset government intentions. And CSOs themselves admit they often lack the capacity and resources to engage seriously on policy issues that require expertise and research (the country’s think tanks could be tapped more for this purpose, as well as to augment capacities in government and parliament). 37. The machinery of consultation/dialogue still tends to be ad hoc, and is not institutionally embedded. Parliamentary Committees and District Assemblies are not systematically included in the processes, thereby fostering a potentially dangerous divide between consultative processes and representative democracy. 38. The government is itself sometimes too passive vis-à-vis the donors. For government to take effective charge it needs to have a clear, convincing program around which coordination can occur. On the other hand, donors can also be passive, hiding behind arguments of national ownership to avoid insisting that attention be given to critical but difficult and politically sensitive areas such as civil service reform. Options for gradual, joint construction of country ownership—which has proven successful in other areas—may thereby be lost. 39. While the realities of high aid dependence (with 40% of the budget coming from ODA) and limited local capacities inevitably condition what it is possible to achieve, there are possibilities of enhancing national ownership through the further deepening of  xv democratic values, institutions, and accountability within the political culture. One way forward would be to open up policy processes further, with a rapid transition from ‘validation’ to ‘consultation’ to greater ‘participation.’ Such openings should be carefully integrated into the democratic system, to avoid that participation become an ad hoc event for those already privileged by a strong, often donor-funded voice. Country-Led Partnership 40. The performance of the donors is mixed. There have been positive moves in desirable CDF-like directions; more open CGs, mini-CGs, improved information flows, greater acceptance of group processes leading to peer pressures for change and selectivity, increases in the share of jointly funded projects and programs, shifts toward sectoral and non-sectoral budget support, increases in joint donor preparation/appraisal missions. But, outside the health sector, there is limited evidence of a strong general trend toward joint funding schemes, harmonization of reporting and procurement rules, lower transactions costs and ownership-transfer. There are also questions about whether donor attempts to maintain an arms-length approach to encourage local ownership will be sustainable. 41. GoG capacity to lead donor coordination is severely limited both in central ministries, some sector MDAs and at district level. Weak systems, ad hoc approaches and disincentives to share basic information weaken transparency. Stalling of key reforms, such as the national procurement reform, is putting in doubt government’s commitment to push through fundamental reforms in public financial management that ultimately impacts donor behavior. 42. Coordination does not extend much beyond Accra. Many government ministers are ambiguous about whether they want stronger coordination, as they fear this might lead to donors ‘ganging-up’ and involving themselves more deeply in policy or program management issues. 43. The Health SWAp has many virtues: good local ownership, reduced transaction costs, improved coordination, better monitoring, and strengthened local capacities. But it is far from a pure case. The impact of a triple-layer donor structure on efficiency/transactions costs is unclear. Also, despite much attention to expenditure and impact monitoring, the effects of the SWAp on health indicators are still inconclusive. 44. In other sectors—agriculture, economic governance, and decentralization— practical coordination on the ground (of approaches, planning, interlinkages, timing and reporting) is not efficient, and transactions costs for government remain unacceptably high. 45. Donors persist with “bad practices” that are well known to have adverse affects on the public sector—such as circumventing the government budget, providing supply- driven assistance directly to districts without coordinating through the center, poaching the most talented government staff, topping-up salaries of preferred personnel, etc. In several areas a donor-funded ‘parallel civil service’ has been created, essentially taking over the functions of government in the area. The study team heard of few successes of  xvi building sustainable capacity in the public sector, while accounts of donor-funded activities undermining capacity were frequent. 46. In some areas (e.g. capacity building in districts, investment delivery mechanisms to communities) there are several ‘enclave’ approaches leading to duplication of efforts, diversity in approaches, limited communication of best practices, and proliferation of workshops. Capacity building efforts focus primarily on training and workshops, rather than addressing structural constraints (incentives to performance). Some donors still provide assistance, which is largely supply-driven, earmarked or tied exclusively to goods and services purchased in their home countries. Technical assistance still comes mostly in the form of international consultants. And here again there is no clear improving trend. 47. The combination of donors’ continuing preference for quick-disbursing mechanisms and rapid visible results (separate projects, PCU/PIUs) and GoG reluctance to refuse offers is persistent obstacles to change. In many cases, the government also has incentives to seek unsustainable, but short-term efficient TA solutions instead of relying on civil servants. 48. Donor’s flexibility varies, including their willingness to delegate authority to local representatives and national staff. The World Bank, among others, has moved in this direction by moving the Country Director from Washington to Ghana and giving national staff more authority for task management. These kinds of shifts are much appreciated by Ghanaians. 49. The timeliness and usefulness of donor reporting to GoG is improving but is still seriously inadequate, partly because of donor inability (and/or unwillingness) to provide information, and partly because of a fragmented, unsystematic approach of the GoG when seeking information. Government utilization of the information provided is also in need of improvement. 50. The above shortcomings underline the need for the types of changes advocated by the CDF: more country ownership, improved country-led coordination, and reduced transactions costs. Translating these principles into everyday operations means moving from project to program aid, greater decentralization of donor agencies to enable flexible decisionmaking and increased dialogue with the government and the wider society. And such changes are premised on a government willing to articulate a clear program and to assert leadership, and on donors finding the program sufficiently poverty-oriented. Results Orientation 51. Despite efforts, there is not much progress here, and the monitoring situation is in general quite weak. But some efforts are being made. There are active plans to strengthen the monitoring capabilities of NDPC, especially in connection with GPRS, but present staffing and budget of NDPC remains quite inadequate to the task.  xvii 52. The Health SWAp arrangements are strong on the development of performance monitoring and the independent review of progress, but there remain large data problems and considerable inertia. Plans for monitoring SWAp II are currently being strengthened. 53. Attempts to build the MTEF into budgetary processes are also a positive factor, although effectiveness is constrained by MoF resistance, unpredictability of MDA resources and the plethora of other public expenditure reform initiatives. The basic machinery required to foster results-orientation is still quite weak. It is not clear how far these initiatives will improve accountability toward the public and the electorate, and to what degree they only focus on technical or administrative aspects of public financial management. Incentives fostering political accountability, which in the end are key to results-orientation, are still weak, and Ghana’s past political culture and history plus factors like low literacy, point to serious challenges for advancement in this area. Sequencing and Relations between the CDF Principles 54. The story of CDF in Ghana demonstrates the intimate relationship between the four principles: A long-term holistic development framework is of limited value if consensus among stakeholders—including donor partners—is not broad enough to make it fairly robust over time. And consensus and ownership at a “framework-level” is of little significance if operational behavior remains uncoordinated and if donors, seeking to produce visible results and to protect their particular investment, apply implementation practices that are destructive to sustainable capacity building. Finally, results orientation requires not only country monitoring capacity but, fundamentally, broad implementation and accountability capacity, including key areas linked to budgeting/planning, execution of budgets and financial and political accountability. Conclusions 55. The direct impact of launching the CDF-principles in Ghana was limited. There were some cases where CDF principles could be seen in action and showed real movement forward (such as the health SWAp, parts of the GPRS and CG processes). Many of these initiatives pre-dated the CDF, though they undoubtedly gained some momentum by the World Bank’s strong global advocacy of the CDF. And it is clear that the CDF pilot period saw positive changes in the image and practices of the World Bank in Ghana. 56. Looking more broadly at how far CDF principles have been adhered to generally in the country, there are mixed results. 57. Firstly, the Ghana experience demonstrates that capacity parameters—including leadership—define how CDF shall or can be applied. Though strong political leadership at the top certainly exists in Ghana, it is not linked to broader capacity in the public sector. Building this capacity will be a long-term effort, and it will demand that donors radically change their approaches to implementation. Looking back, the willingness to do so—and to push for the necessary civil service reform—has been quite low, and it is not easy to see imminent changes in the incentive structures for either government or donors that would put the capacity issue on the top of the agenda.  xviii 58. The Health SWAp came about after both national and international stakeholders perceived a crisis in the health system and in previous donor support to the sector. Ghana as such is in no such crisis that could prompt major changes in the partnership relations– on the contrary, the country is now facing the risks and opportunities of becoming a donor favorite. Rather than promoting the application of the CDF principles, this may lead to increased competition and crowding among donors. It will be hard for the government to refuse the gentle offers, but it may be worthwhile to establish minimum standards for what kind of support (and strings attached) are acceptable. The government could also commit itself to a plan for decreasing overall levels of aid. 59. This analysis suggests that change toward flexible CDF-like behavior is likely to be only gradual, and not consistent. The most important contributing factor will be the further deepening of democratic values, processes and institutions relative to the forces of clientelism, including the growth of civil society, with its associated momentum in favor of greater accountability. For the donor community, the key challenge is to patiently align policy dialogue and aid negotiation methodologies, size of aid as well as implementation modalities to this complex process. 1. Introduction 1.1 This report presents the findings of an international team that has reviewed the experience of Ghana in implementing the principles of the “Comprehensive Development Framework” (CDF). Ghana is one of six country case studies1 being conducted as part of a global multi-stakeholder evaluation of implementation of the CDF. 1.2 The CDF concept, which was first articulated by the World Bank president in January 1999, comprises a set of principles that a number of developing countries and development assistance agencies have been seeking to put into practice to improve the effectiveness and impact of the global aid system. The principles on which the CDF is based—long-term, holistic development framework; country ownership; country-led partnership; and results orientation—are distilled from development experience over the last five decades. The multi-stakeholder evaluation aims to identify the factors that favor and inhibit their successful implementation in a number of countries, and to distill learning about the validity and limitations of their application. 1.3 Ghana was selected as a case study for several reasons, including its status as a “CDF pilot country” (as agreed by Ghanaian and World Bank leadership in 1999) and the substantial aid flows to Ghana. 1.4 The study evaluated the application of the four CDF principles separately and together. Preliminary hypotheses and research topics were defined on the basis of the Design Paper for the CDF Evaluation.2 Through interviews involving key stakeholders from the government, the political system, civil society, and international cooperation partners, and through a survey focusing on transaction costs administered to donors present in Ghana, facts and viewpoints were sought regarding: • the extent to which CDF principles are applied and the causes thereof; • the specific impact on development processes and on the quality of the work processes of international cooperation; • evidence of ultimate impact in terms of development targets and objectives, and on resource allocation and mobilization; and • emerging lessons learned in order to enhance the impact of poverty reduction efforts, including international cooperation efforts. 1.5 The study team worked in Ghana from April 2-9 and from May 6-24, 2002. During the April visit a number of issues and themes were selected for closer review based on advice from stakeholders. These included the Ghana Poverty Reduction Strategy (GPRS), health, agriculture, decentralization and economic governance. The broader issue of public sector capacity constraints and attempts to address these— 1. The other countries are Burkina Faso, Ghana, Romania, Uganda, and Vietnam. 2. World Bank, 2001: Design Paper for a Multi-Partner Evaluation of the Comprehensive Development Framework, CDF Secretariat 2001.  2 especially civil service reform—was added in May, as it was found to be of central importance for an assessment of the implementation of CDF-principles in Ghana. 1.6 The report is structured as follows: first, the Ghanaian historical and socio- economic context is summarized, focusing on factors relevant to the issues and themes chosen for the study. This is followed by a discussion of the inauguration of Ghana as a CDF pilot country and an assessment of the immediate outcomes. The GPRS and the processes leading to it were found to be so central for the evaluation that it merited a separate chapter. The remaining findings are structured around the four principles of the CDF, followed by a concluding chapter.  3 2. The CDF Principles in Ghana - the Context and the History 2.1 Ghana has on several counts done well in the last two decades. Most significant, the country adopted a parliamentary democratic system in 1992 and has held three multi- party elections, the last (in December 2000) resulting in the peaceful and orderly transition of power from one party (National Democratic Congress [NDC]) to another (New Patriotic Party [NPP]). After a period of economic stagnation during the 1970s and early 1980s, the economy resumed growth and has achieved average annual growth levels around 4 to 5 percent over the last decade-and-a-half. Poverty in Ghana 2.2 Poverty levels have decreased from 52% in 1991/92 to 40% in 1998/99. Though average per capita income (US$390) remains lower than at independence, most of Ghana’s social indicators compare favorably with the averages for Sub-Saharan Africa. But the favorable trend in the averages masks worrying patterns of poverty in Ghana. The first is that approximately two-thirds of the poor in Ghana are below the so-called “lower poverty line,” i.e. they are living in extreme poverty.3 The second is that of significant differences in the spatial distribution of poverty. Poverty levels are highest in the three northern regions of the country ranging between 69% and 88%. Moreover, the incidence of poverty increased in the 1990s in three regions –Central, Northern, and Upper East, the latter two being among the poorest in the country (Table 1). Table 1. Incidence of Poverty by Region and Location in the 1990s Proportion Below the Lower Poverty Proportion Below the Upper Poverty Region Line, i.e. Extreme Poverty Line 1991/92 1998/99 1991/92 1998/99 Western 0.42 0.14 0.60 0.27 Central 0.24 0.31 0.44 0.48 Greater Accra 0.13 0.02 0.26 0.05 Eastern 0.35 0.30 0.48 0.44 Volta 0.42 0.20 0.57 0.38 Ashanti 0.25 0.16 0.41 0.28 Brong-Ahafo 0.46 0.19 0.65 0.36 Northern 0.54 0.57 0.63 0.70 Upper West 0.74 0.68 0.88 0.84 Upper East 0.53 0.80 0.67 0.88 Urban 15.1 11.6 27.7 19.4 Rural 47.2 34.4 63.6 49.5 Total 36.5 26.8 51.7 39.5 Source: Ghana Statistical Services (2000): The Pattern of Poverty in the 1990s, Accra. 3. The Ghana Statistical Service in its analysis of poverty in Ghana defines two nutrition-based poverty lines. The lower poverty line is the quantum of consumption expenditure needed to achieve minimum nutritional needs. Households falling below this lower poverty line live in extreme poverty and cannot meet their minimum nutritional needs even if their entire expenditure is used to purchase food.  4 2.3 Nutrition-based measures of poverty show similar regional patterns. Although the incidence of stunting and wasting has declined nationwide in the ten-year period since 1988, the incidence of wasting increased in three regions (Table 2). Chronic malnutrition, i.e. stunting, increased in five regions between 1993 and 1998. Table 2: Child Nutrition Status Region Height for Age (Stunting) % Weight for Height (Wasting) % 19881 1993 1998 1988 1993 1998 Western 30.3 33.1 29.4 11.2 13.0 9.2 Central 40.4 23.0 26.8 11.1 11.5 10.3 Greater Accra 22.1 15.7 11.3 4.6 7.9 5.5 Eastern 30.2 19.8 25.1 5.1 10.4 15.2 Volta 15.5 25.0 23.6 6.6 5.9 8.7 Ashanti 26.7 27.9 27.6 9.0 8.4 9.2 Brong-Ahafo 26.8 24.5 17.8 7.5 13.0 8.1 Northern 48.8 35.9 39.6 10.4 19.0 12.7 Upper East 33.3 34.6 20.6 7.1 Upper West 26.0 35.9 14.5 8.2 Total 30.0 26.0 25.9 8.0 11.4 9.5 Source: Ghana Statistical Service: Demographic and Health Surveys, 1989, 1993 and 1998, Accra. Notes: 1. In 1988 separate data were not collected for the three Northern regions. The figure for the Northern region applies to all three Northern regions. 2.4 Poverty in Ghana, as is the case in many other African countries, is not linked to unemployment. Poverty is highest among food crop farmers, among whom almost 60% fall below the poverty line. This section of the labor force experienced the smallest decline in poverty in the 1990s. This contrasts with the export farmer group that registered the largest percentage point decline in the headcount index. 2.5 Some groups and locations have not improved, or even deteriorated in terms of well-being. There are also biases in the distribution of social and economic infrastructure that facilitate the capacity to take advantage of opportunities as they arise. The spatial pattern of poverty very closely reflects the spatial patterns of school enrollment rates, literacy rates, access to safe water, health facilities, roads, and markets.4 Political Setting 2.6 Following the ousting in 1966 of Kwame Nkrumah, Ghana’s first independence leader, six shifting military and civil governments—generally characterized as both very corrupt and incompetent—ruled until The Provisional National Defense Council (PNDC) took over the helm of government in 1981 through a coup d’etat. For the next decade there was a ban on political parties. Pressures for a shift to a democratic form of government emerged in the late 1980s. One response to these demands was the establishment of the District Assemblies, formally composed of non-partisan representatives. In 1990 a series of 4. GRPS, Appendix A.  5 regional consultations was embarked upon to elicit views on a new political system. A new Constitution was drafted in 1991 and approved in a referendum in early 1992. 2.7 Elections were held in November 1992, and again in 1996. The incumbent government, i.e. PNDC that had transformed itself into the National Democratic Congress (NDC), won both. In the interim period between the two elections political liberalization advanced further with the emergence of the private media. The 1992 Constitution does not permit an incumbent to hold the Presidency for more than two consecutive terms, and this was respected by the NDC, which presented a new Presidential candidate for the 2000 elections. The election was successfully held in December that year with an opposition party, the National Patriotic Party winning the Presidential vote after a run-off. The peaceful transition of power is one of the few instances in African history where an incumbent president has stepped down or accepted a defeat in democratic elections. Box 1: Milestones in Ghana’s Political Economy 1981 Coup d’etat that brought in the Provisional National Defense Council (PNDC), headed by J.J. Rawlings 1983 Introduction of the Economic Recovery Programme (ERP) 1987 Education Sector Reforms Decision taken to create the National Development Planning Commission (NDPC) 1988 Program of Actions to Mitigate the Social Costs of Adjustment (PAMSCAD) First Ghana Living Standards Survey Completed Creation of 110 District Assemblies 1989 Second Ghana Living Standards Survey Completed 1990 Establishment of the NDPC 1992 Promulgation of 1992 Constitution Elections held in November Third Ghana Living Standards Survey Completed 1993 Act 455 Establishing the District Assemblies Common Fund 1994 Start of the National Institutional Renewal Program (NIRP) 1995 Introduction and Withdrawal of VAT Ghana Vision 2020 The First Step 1996 Start of implementation of Public Financial Management Reform Program (PUFMARP). Elections held in December 1997 Structural Adjustment Participatory Review Initiative (SAPRI) launched National Economic Forum Start of Mini-CG meetings held quarterly in Accra 1998 Re-introduction of VAT at lower rate of 10% 1999 First CG meeting held in Ghana Fourth Ghana Living Standards Survey Completed Medium Term Expenditure Framework (MTEF) budgeting process used for the national budget 2000 Elections held in December Preparation of the Ghana Poverty Reduction Strategy Paper begins 2001 Government takes decision to apply for HIPC relief National Economic Dialogue held in May 2002 Heavily Indebted Poor Country (HIPC) Decision Point reached in February Draft GPRS presented to CG-meeting in April National Economic Dialogue held in May  6 2.8 Complementary to the victory of representative democracy, consensus-building events and consultations have been on the increase during the last five years, though it is still too early to characterize them as institutionalized. A National Economic Forum, organized by the National Development Planning Commission (NDPC) was first held in 1997 with the objective of achieving a consensus on policy measures for accelerated growth. Participation at the National Economic Forum covered a wide range of actors within the different arms of government, the private sector and civil society. Proposals were made by participants on a wide range of issues, for example value added tax (VAT) and the budget. However implementation of these proposals has been limited. 2.9 The new NPP Government in its first year in office also held a National Economic Dialogue. An oversight committee was set up to ensure that implementation of recommendations did take place.5 A year later in 2002 a second dialogue was held, as promised, to assess what progress had been made during the year in reaching the targets that had been set the year before. The forum found that substantial progress had been made in implementing macroeconomic recommendations. The National Economic Dialogue may thus become a permanent feature of Ghana’s political life. The specific consultative process related to the GPRS is discussed in this section. The Economic Reforms 2.10 Prior to the implementation in 1983 of the Economic Recovery Program (ERP), as the economic reforms were known in Ghana at the time, the Ghanaian economy was in a desperate state. Real GNP per capita had contracted considerably (an estimated 37% compared to 1970), negative growth rates had been recorded for three consecutive years, investment rates were a fraction of the 1960 rates, import and export volumes had declined precipitously and inflation was accelerating to three digits. A parallel economy emerged, eroding the tax base of government. Tax revenues and government expenditures declined as a ratio of GDP and in real terms. Physical infrastructure was in a state of disrepair and the provision of social services had deteriorated. In a World Bank study on price distortions and growth, the Ghanaian economy was ranked as having the highest degree of price distortions among a group of 31 Asian, Latin American and East European countries.6 2.11 It was against this backdrop that the ERP was negotiated between the government of Ghana and the IMF. The negotiations were kept secret and there was limited if any consultation with major stakeholders about the contents of the program. Even staff of the Ministry of Finance and Economic Planning complained that they were not privy to the contents of the program. 2.12 The main objectives of the ERP were to increase production, particularly of food, industrial raw materials and exports through an improvement in the structure of incentives; 5. Government of Ghana: National Economic Dialogue. Action Plan, May 2001. 6. Agarwala, R. (1979), Price Distortions and Growth in Developing Countries, World Bank Staff Working Papers, No. 573, Washington D.C. The study examined distortions in exchange rate pricing, factor pricing and product pricing.  7 increase the availability of consumer goods; rehabilitate the physical infrastructure; increase the overall availability of foreign exchange and improve its allocation; and lower the rate of inflation through the pursuit of prudent macroeconomic policies. 2.13 The focus of the first three years of the program was mainly on stabilization. Substantial improvements were recorded in most important macroeconomic indicators in the first years of the program (Table 3). Table 3: Macroeconomic Indicators 1983-2000 – Annual Averages 1983 1984-86 1987-89 1990-92 1993-95 1996-98 1999 2000 GDP growth -4.3 6.3 5.2 4.1 4.3 4.5 4.4 3.7 rate (%) Gross Fixed 3.7 8.7 14.4 15.2 22.5 22.01 n.a. n.a. Investment Rate (%) Average 123 25 32 22 36 371 n.a. n.a. annual inflation rate (%) Exports 439 649 838 960 1244 1717 2006 1937 (f.o.b.) US$ Million Share of 89 89 86 82 86 80 72 68 traditional exports (%)2 1 Annual average for 96-97 2 Cocoa and cocoa products, minerals and timber Source: Ghana Statistical Services Quarterly Digest of Statistics, various issues, Accra. International Monetary Fund International Finance Statistics Yearbook (1994), Washington D.C. Bank of Ghana Statistical Bulletin various issues, Accra. 2.14 The second phase of the ERP began in 1987. The ambit of the reform was widened to address structural and institutional issues. A number of reforms were started in that year in education, the financial sector, state enterprises and the civil service. Macroeconomic aggregates remained favorable relative to other African countries, and exports doubled in value from 83-88. The anticipated transformation of the export base had, though, not been achieved. Despite the growth of the economy and exports, concerns were expressed about the limited impact of the reforms on employment creation, and the economy was still dependent on external inflows. 2.15 In the 1990s growth has been stable, but never exceeding 5% annually. Gross investment rates have been fairly stable, lying in the range of 20-23% (table 3), and exports continued to grow much faster than GDP, with decreasing weight of the traditional exports. The expected focus on accelerating growth has been undermined by the need to implement stabilization measures to deal with the fiscal imbalances that have characterized most of the period since political liberalization. Among others, fiscal expansionism in election years has been a feature of all the national elections that have been held since 1992.  8 2.16 1999 and 2000 were less positive years for the Ghanaian economy. The combination of the decline in cocoa and gold prices, rising oil prices and a shortfall in expected external inflows affected the economy. On the domestic front the government had been running large budgetary deficits beginning from the second half of 1998. The central bank had instituted a policy where it did not automatically honor government checks. This led to the build up of arrears estimated to be about 5% of GDP at the end of 1999.7 The 1999 fiscal crisis illustrated that the reforms had not yet succeeded in diversifying the structure of the economy, and that the economy continued to depend on external inflows. 2.17 The focus of the new NPP Government in its first year in office was to stabilize the economy. It succeeded in reversing the strongly negative inflationary tendencies through prudent fiscal management and a tight monetary policy, thereby reducing inflation from a peak 42% when it took office to 21% by the end of 2001, and with a likely year-end scenario of 10% for 2002. The reduction in aggregate demand pressures contributed to decreasing the demand for foreign exchange. In addition, there was an improved inflow of official capital transfers and an agreement to defer repayment of Paris Club and other bilateral debts. The cedi-dollar rate depreciated by less than 5% in 2001. With an average inflation rate of 32.9% during the period the net effect was a considerable appreciation of the real exchange rate in 2001, thus hampering exports. 2.18 One of the early controversial actions of the NPP Government was the decision to apply to be included in the Heavily Indebted Poor Country (HIPC) debt relief initiative. In February 2002 Ghana reached the decision point and should benefit from US$96 million in debt relief in 2002 (out of total external repayments budgeted at US$375 million) increasing to about US$200 million a year until 2011. It is planned that 20% of the HIPC debt relief will be used to reduce the domestic debt and the remainder is to be used for poverty reducing activities that have not yet been identified. 2.19 Even though the stabilization of the economy is being applauded, many observers, both within and outside of government, express concerns about the implications of the fiscal restraint for investment, employment generation and the implementation of programs and projects, particularly in the social sector. The continuing challenge of economic reforms in Ghana is to effect the transition from stabilization to long-term, equitable growth at rates beyond the 4-5% that have been achieved in recent decades. Visions and Policy Frameworks 2.20 The GPRS, which will be discussed in details in chapter 3, is by no means Ghana's first attempt to formulate policies for the reduction of poverty, or its first attempt to produce a broad and comprehensive development vision, as evidenced in Box 2 below. 7. CEPA, Macroeconomic Review and Outlook, 2000, Accra.  9 Box 2: Visions and Poverty Reduction Strategies in Ghana Dates Visions, Poverty Reduction Comments, Other Events, Donor Strategies Participation 1983 Economic Recovery Program Negotiated with IMF 1988 Program to Mitigate and Alleviate Response to concerns raised by UNICEF and the Social Costs of Adjustment other donors (PAMSCAD) 1995 Ghana Vision 2020, the First Step Ministries, Departments, and Agencies (1996 – 2000), with support from (MDAs) were asked to identify and monitor UNDP poverty reduction indicators. Inter-Ministerial Committee on Poverty Reduction (IMCRP) formed following CG-meeting. 1996 Policy Focus on Poverty Led to formulation of various parallel donor Reduction Document published, funded programs (National Poverty Reduction providing policy framework for Program [UNDP], Social Investment Fund medium term [AfDB], Village Investment Program [WB]. 1997 Medium Term Development Plan produced 1998 United Nations Development Training on Medium Term Expenditure Assistance Framework (UNDAF) Framework given for all ministries; 1999 produced budget produced in MTEF form 11 / 1999 First Draft Comprehensive Tenth Consultative Group meeting held in Development Framework towards Accra; draft CDF document reviewed. Ghana Vision 2020, also known as “Reducing Poverty Through Improved Accelerated Poverty Reduction Agriculture” document also presented to CG- Programme. meeting 1 / 2000 Ghana Vision 2020, Second Step The process, lead by the Vice-president and (2001-2005) preparation launched; the National Development Planning Cross-Sectoral Planning Groups Commission, quickly disintegrated. created Simultaneously, PRSP Task Force begins work on developing Interim PRSP, based on draft CDF document 3 / 2000 "Development Strategy for CAS and CDF processes converge, but Poverty Reduction" presented as strategy prepared under pressure to meet Part I of Country Assistance timetables, and strategy not released to the Strategy (CAS) to Board of World public. Bank 7 / 2000 Ghana Poverty Reduction Strategy process launched; Core Teams established 8 / 2000 Interim PRSP endorsed by Boards Building on the “Development Strategy for of IMF and World Bank Poverty,” but a “Policy Matrix for poverty Reduction 2002-2002” was added, building on the Policy Framework Paper for the Enhanced Structural Adjustment Facility 1999-2001. The I-PRSP was not released to the public. 4 / 2002 GPRS – volume one Presented to the 11th CG meeting Source: Amended from World Bank, PRSP Secretariat: Ghana Process Case Study, 2001.  10 2.21 Ghana was one of the first countries to introduce safety-net provisions in connection with the structural adjustment reforms of the 1980s. Although the 1988 Program to Mitigate and Alleviate the Social Costs of Adjustment (PAMSCAD) scheme was far from successful, it led to the subsequent promulgation of a Policy Focus for Poverty Reduction (1996). Subsequently, an Accelerated Poverty Reduction Program (1999) presented a costed set of poverty-focused activities for 1999-2001 and, in the same year, the government submitted to the CG meeting a document entitled Reducing Poverty Through Improved Agriculture. In 2000, this was followed by Development Strategy for Poverty Reduction, which mainly served for the World Bank’s Country Assistance Strategy (CAS) preparation. In July of the same year, an interim PRSP was presented to the boards of the WB and the IMF as part of the negotiation of credits. 2.22 Various coordination mechanisms were formed during the years, and “Poverty Desk Officers” were appointed in Ministries, Departments, and Agencies (MDAs) to serve as focal points and links to the Poverty Reduction Unit of the NDPC. 2.23 Ghana Vision 2020 aimed at Ghana achieving middle-income status by the year 2020. In the medium term the objective was to build upon the gains of the ERP and create the conditions for accelerated growth. The Vision provided the framework within which the constitutionally required Medium Term Development Plan was produced. Preparation of this document involved a more participatory approach than previously with the NDPC providing guidelines to the Districts and sectors for the preparation of District and sector plans. 2.24 By general consent among all informants interviewed, the various strategies and programs produced over the last decade were never implemented to any significant degree and none of them made a discernible difference in government actions at ground level. Many Ghanaians would argue that Vision 2020 had long lost whatever practical meaning it might ever have possessed, having never been integrated with budgetary processes, and matched by implementation capacity and monitoring. 2.25 The preparation of the different strategies was in all cases linked to some donor- related event or process (CG-meetings, CAS preparation, negotiation with the IFIs). This does not necessarily imply that donors had a strong say on content (though several documents clearly were inputs in the negotiation process with the IFIs) but it does suggest that the basic incentive for producing the documents was relations with international partners, rather than dynamics created by Ghanaian political processes. Social Sector Reforms 2.26 A large public health and education system had been created during the post- independence period. Free tuition public education was introduced in the 1960s. Although there was private provision of education and health services, the public sector was the sole provider of formal education and health in many communities. The decline in government’s capacity to spend during the 1970s took its toll on the health and education sectors. The real earnings of personnel in both sectors had declined  11 substantially in the period preceding 1983, and there was a large exodus of teachers and qualified health personnel from the country because of poor conditions of service. The health centers and educational institutions suffered shortages in necessary materials and supplies. Table 4: Social Sector Indicators (percent, except last 2 rows) 1991/921 1998/992 Net primary school enrollment ratio 74.1 83.4 Net secondary school enrollment ratio 37.5 40.7 Individuals seeking modern medical care in case of illness, in rural 45.3 39.8 areas Natural sources as main source of drinking water – rural areas 48.9 35.4 Infant Mortality Rates (per 1000) 74.7 61.2 Child Mortality Rates (per 1000) 62.8 52.4 Source: Ghana Statistical Service (2000) Poverty Trends in Ghana in the 1990s, Accra, Ghana Statistical Service Demographic and Health Survey, 1993, 1998 Accra. Notes: 1. 1993 for Mortality Rates. 2. 1998 for Mortality Rates. 2.27 Education reforms began in 1987, aiming at equity of access and improvement of quality, efficiency and relevance of education. Tuition remains free in the public school system but book user fees have been introduced at the basic education level. Boarding fees have been introduced and a policy of establishing non-residential secondary schools is being implemented. At the tertiary level academic user fees and residential fees were introduced with some resistance. 2.28 Gender and regional variations in access to education persist. Recent data show an increase in the number of girls attending school while there is a drop in enrollment among boys in some regions. Primary enrollment rates have increased (Table 5). At the secondary level there was a drop in enrollment rates among the very poor.8 The quality of education remains a major problem. The learning outcomes for children in rural and public schools are very low. 2.29 Health sector reforms begun in 1986 have involved the introduction of user fees. There is also a policy to refocus expenditure toward preventive health care and away from curative health care. A third objective was to decentralize the health care system. 2.30 With the introduction of the user fees there has been an improvement in the supply of drugs, etc. in the public hospitals. At the same time it is also one of the most unpopular reforms because the general public now has to pay for services which – though maybe irregularly available – were previously free. There is an exemption policy, but it is not effectively implemented. The NPP Government intends to abolish this “cash and carry system” and introduce a national insurance scheme. 2.31 The Living Standards Survey reveals that the proportion of people who consult a health worker when they fall ill declined between 1991/92 and 1998/99 (Table 5). The cost 8. Ghana Statistical Service (2000) Poverty Trends in the 1990s, Accra.  12 of accessing health care in terms of the fees, cost of travel and time spent at the formal health facility have been found to be important in explaining the reduced demand for formal health care services. Infant and child mortality rates have declined in the 1990s (Table 5). Again the average masks regional patterns. Child mortality rates increased in the 1990s in five regions in the southern part of the country. The three northern regions still have rates higher than the national average, although the overall trend has been downwards. Agriculture 2.32 Ghana’s rural population represents 63% of the country’s total population, and agriculture represents 38% of GDP. Poverty in rural areas affects 52% compared to 23% in urban areas.9 Regional disparities should be added to this picture. The growth rates in agriculture during the 1990s have been below overall economic growth rates, and poverty has actually increased in several rural areas. 2.33 In the light of these facts, and in support of Vision 2020, the Ministry of Food and Agriculture (MoFA) began the development of new strategy to accelerate agricultural growth and development in 1995. After a series of consultations of stakeholders across the country MoFA prepared the ‘Draft Accelerated Agricultural Growth and Development Strategy’ (AAGDS) document. The policy reforms, institutional reforms and investment programs designed to achieve the objectives of the strategy are based on five major elements: • Improving access to markets and promoting the production and export of selected commodities • Facilitating access to agricultural technology • Facilitating and increasing access to rural finance • Providing rural infrastructure and utilities, and • Building institutional capacity 2.34 The AAGDS has proven fairly stable as an overall policy framework for the sector, also after the change of government. However, major policy differences have evolved between the government and various donors, as well as between donors, regarding if and how different service delivery functions to farmers (extension service, animal health care) should be fully or partly privatized. 2.35 As in other sectors, implementation of reforms and policies in agriculture is closely dependent on the capacity of the public sector institutions. This central theme is discussed below. Public Sector Capacity and Public Sector Reforms 2.36 Several institutional reforms have been implemented since the mid-1980s. In many instances the reforms may be seen as attempts by the administration to improve 9. World Bank: Project Appraisal Document, Agricultural Services Sub-sector Investment Project, 2000.  13 upon the management and implementation of the economic reforms. Below, focus is limited to decentralization, civil service reform and economic governance reform, which are the areas chosen for closer scrutiny in the evaluation. 2.37 Decentralization: In 1988, 110 district, municipal and metropolitan assemblies came into being with the enactment of the local government law. Thirty percent of the membership of the district assemblies are appointed and the remaining 70% elected, all formally on a non-partisan basis. The district chief executive is appointed by the president. The district assembly is responsible for district level planning, service delivery and revenue generation. In 1993 the enactment of Act 455 created the District Assemblies Common Fund (DACF). At least 5% of national tax revenues are to be allocated to the districts. The monies are disbursed among the 110 districts using a formula weighing equality, need, responsiveness, and service pressure. In the last four years there have been arrears in payments into the DACF from central government. The DACF mechanism has also been plagued by corruption. 2.38 A number of sector ministries are deconcentrated (including health, education, forestry), where district level agencies report to the sector ministries rather than to the district administration. These sectors have so far opposed moves from deconcentration to devolution of any significant authority to local levels. There are no structures to ensure effective coordination between the deconcentrated agencies and the district administration. Decentralization was expected to encourage participation and local ownership, but this has been difficult to attain in some districts because of problems of lack of accountability of Assemblymen and women to their constituents. Capacity constraints exist both at the level of the districts and at subdistrict level, where capacity is virtually non-existent in many districts. 2.39 Donors have been very supportive of decentralization, and have also raised their concerns about apparent lack of progress.10 The NPP government acknowledges that progress has been modest in recent years, and that implementation has been fragmented, without a clear leadership of the process on behalf of the government. 2.40 National Institutional Renewal Program (NIRP): The National Institutional Renewal Program was initiated in 1994, covering three main elements: Public Sector Management Reform Program (PSMRP), Public Finance Management Reform Program (PUFMARP) and Civil Service Performance Improvement Program (CSPIP). It was acknowledged that a transparent, accountable and cost-effective administration at the national, regional and district levels was needed to achieve the targets set for economic growth. 2.41 Public Financial Management Reforms: Public expenditure reviews (PERs) conducted by the World Bank between 1985 and 1990 revealed several problems in public financial management. Some of these were a poor budgeting system, lack of expenditure reporting and low skill level of professional accounting personnel. Several 10. The key donors participating in the CDF Thematic Group on Decentralization prepared a joint, fairly comprehensive joint statement to the CG-meeting in April 2002.  14 programs were initiated to address these problems, but they did not take a comprehensive approach but addressed the problems in a piece-meal fashion. The 1993 and 1994 public expenditure reviews conducted by the Government of Ghana found that there was no rationale or basis for the allocation of resources in the budget. The Public Financial Management Reform Program (PUFMARP), launched in 1996, aims at providing an integrated approach to dealing with the problems of public finance management. The objective of the program is to enhance efficiency, accountability and transparency of the financial management functions of government, so as to enable the maintenance of macroeconomic stability. PUFMARP includes 7 components: Medium Term Expenditure Framework (MTEF), Budget and Public Expenditure Monitoring System (BPEMS), Revenue Management, Aid and Debt Management, Cash Management, Audit and Fiscal Decentralization. 2.42 Civil Service Reform: Ghana has had a reputation for having a relatively well- functioning civil service. However, a de-motivated and often corrupt public sector was the legacy of the decade of economic and social decline in the 1970s. Public sector positions – though poorly paid – were used as rewards to clients, and payroll fraud – adding the so-called “ghost workers” to the payroll – is reportedly still frequent.11 Reform initiatives commenced in the late 80s, including attempts of downsizing of the civil service. Success was limited, and a new phase of civil service reform began under NIRP in 1995. 2.43 Both PUFMARP and civil service reform projects counted on heavy financial and technical backing from donors. The controversial and politically sensitive issue of downsizing and pay reform in the civil service was addressed, but not effectively, in 1995 by the government itself (see box). Recently, the new government has reactivated attempts to formulate a reform package, but expectations are that serious reform attempts would only materialize in a possible second term of office of President J.A. Kufuor. 11. Joe Amoako-Tuffour: Ghost Names, Shadow Workers, and the Public Sector Wage Bill, Institute of Economic Affairs, Accra, 2002.  15 Box 3: The Nettle of Public Sector Pay Reform in Ghana The dismal service conditions in the public service have been obvious for years. Surprisingly little has been done to remedy the situation, pointing to the potential high political costs involved. In 1995 government decided to develop a uniform salary structure for the public service. Consultants carried out a job evaluation and re-classification exercise, and proposed a rationalization of the 130 salary steps and the plethora of special allowances (70 different allowances), which to a significant extent provided discretionary compensation for the completely inflation-eroded official salaries. The consultant study provided recommendations on the public sector job structure, the development of a credible, viable salary relativity chart for the public sector job structure, and the determination of the cost and timing of rationalizing the pay for comparable jobs across the public sector. There was, however, little implementation of the recommendations in the report. The contribution of development partners has been quite limited. Pay reform was intended to be an element of the 1999 World Bank-funded Public Sector Reform Management Program. The government requested that the component be removed from the program because it had its own strategy. In the CAS for 2000-2003, the Bank notes, apparently with some regret, that it “has no instrument to address directly the size of the bloated public service.” Reportedly, bilateral donors are cautious about the subject because they are concerned about ending up as scapegoats when the political heat increases, which is very likely since the “right-sizing” of the public sector associated with pay reform will require layoffs. The NPP Government has again broached the issue of pay reform. A pay policy study has been commissioned. Consultants are in the process of identifying alternative strategies. A public sector census is being conducted before the end of 2002; a presentation on alternative strategy options was made to the government’s Economic Management Team in June 2002. 2.44 Though both comprehensive and expensive—since 1990 more than US$175 million has been provided from donors—the public sector reform initiatives have only had limited impact, as frankly recognized in the GPRS (see box 4). The potential political costs of reform are high: both cleaning the ghost worker payroll and laying off regular staff will have repercussions, and the NPP government may risk being accused of trying to restore dominance of regional/ethnic groups. 2.45 The new government has also met the public administration cadre with apparent suspicion. The president, in his State of the Nation Address 2002, found reason to say that: “The [public] service has become so politicized and left to its own devices for so long that radical restructuring and refocusing of its purpose are a necessity. And government is determined to do this. Government inherited a big program of reform and renewal of the public services. A lot of money has already been used in the exercise. Unfortunately, the program does not seem to have made the requisite impact. Unless the program is revisited, it might become yet another layer of bureaucracy.”  16 Box 4: Factors Contributing to the Failed Attempts of Public Sector Management Reforms “[The NIRP] represents a comprehensive and impressive package. Despite a number of significant improvements empirical observation suggests that the public sector overall has no more capacity and is no more effective than five years ago. Four policy objectives were set out in 1999. These were the achievement of good governance, accelerated growth and equitable social development, a core compact public service and enhanced public private partnership. Contributory factors to the failure to achieve these objectives appear to include: - A lack of sustained political commitment to the reforms and the institutions responsible for carrying out the reforms - Lack of strong leadership in the public service - Failure to enforce rules, regulations and instructions - Lack of delegation of responsibility to Director level and below - Dysfunctional relationship between political and public service decision- makers - Failure to enforce the effective implementation of agreed policies and programs - Lack of clear, unequivocal institutional structures, roles and procedures in the public sector - Lack of observance of existing rules and regulations - Apparent lack of institutional ownership of reforms - Inadequate resources and provision of funds in a timely fashion - Absence of MDA procedural and operational manuals - Lack of performance related to terms of reference - Unacceptably poor conditions of service It would appear that the totality of the public sector reform program might be beyond the capacity of available human and financial resources to plan and implement.” Source: GPRS, Analysis and Policy Statement, p. 108f, February 20, 2002. 2.46 The GPRS, in addition to the extraordinarily frank analysis of the multiple capacity constraints reflected above, notes that “poverty reduction, sustained economic growth and government’s public sector reform program are dependent upon a public service of high morale, commitment and capacity underwritten by the political leadership. Thirty years of attrition have left their mark. Returning the service to the high standards of the 1950s and 60s requires a total commitment to public sector reform...” (ibid., p.116). Elsewhere it is said that without civil service reform, including improvement of the “unacceptably poor conditions of service.... government reform programs and the GPRS are unlikely to succeed” (ibid, p.107). 2.47 The capacity constraints are not unrelated to past and present donor or partnership behavior, as will be discussed later. They are evidently also central to the application of any of the other CDF principles, and we will revert to the theme of public sector capacity throughout the report.  17 Aid Dependency 2.48 One other important contextual factor is Ghana’s heavy dependence on external aid as a source of financing for imports, public investment and other state spending. Ghana’s foreign aid per capita, which stood at around $18 in 1980, has sharply risen since, particularly following massive influxes of donor assistance in the 1990s in support of economic and structural reforms, reaching a peak of $38 in 1998. Recently, net foreign aid per capita has fallen down to $32 (1999 & 2000); however, net aid as percentage of gross national income, which had remained in the neighborhood of 10% throughout the 1990s, has risen to 13% (2000). Table 5: Comparative Aid Dependency Ratios (2000) Ghana sub-Saharan Africa Net aid as % of gross national income 13 4 Net aid as % of gross capital formation 52 23 Net aid as % of imports 18 11 Net aid per capita ($) 32 20 2.49 Net aid as a percentage of investment and imports stand at 52% and 18% (2000), respectively – lower compared to the levels of the 1990s, during which time Ghana averaged around 50% and 30%, respectively, but considerably higher than the averages for sub-Saharan Africa. External aid represented more than 50% of government expenditure in the early 1990s, and has fluctuated between 40-50% ever since. The fiscal crisis following the terms of trade losses in 1999-2000 and election spending has further increased the country’s reliance in foreign aid, with grants and loans financing about 46% of government spending in FY 2001. Aid inflows (grants and loans) are expected in FY 2002 to be equivalent to more than half of domestic revenues and to finance two-fifths of total public expenditures (about 90% of public investment expenditures). 2.50 External aid inflows to Ghana have fluctuated from year to year. Aid disbursements averaged around $500 million per annum between 1990-1994, peaking in the neighborhood of $800 million dollars in 1995 and 1996, and declining thereafter until 2001.12 The sharpest declines were in 1999 and 2000, due to budget constraints in the wake of the East Asia financial crisis and donor concern about high-level off-budget government spending for the 2000 presidential election. The terms of trade crisis during this period exacerbating the impact of donor cutbacks on the country’s economy. 12. Data for Program and Project Aid disbursements (grants and loans) for the period 1992-2000 is based on actual data obtained by the evaluation team from the Aid and Debt Management Unit (ADMU) of the Ministry of Finance. The team has found discrepancies with the data provided to it by ADMU and data used elsewhere, including with data used in “Study of Aid Flows: Budget, Balance of Payments and Monetary Effects” (October 1999) prepared by the Ministry of Finance and Bank of Ghana. The numbers provided to the team are generally higher than the numbers reported in the October 1999 study.  18 Chart 1: Total Program & Project Aid Disbursements 1992-2001: Actuals; 2002: Projection 800 700 600 US$ millions 500 400 300 200 100 0 1992 1993 1994 1995 1996 1997 1998 1999 2000 2001 2002 Program Aid Year Project Aid 2.51 Major donors to Ghana include the World Bank (infrastructure, education, and health); the International Monetary Fund (structural adjustment); United Nations agencies; Japan (up until Ghana’s decision to participate in the HIPC initiative, the largest bilateral donor with programs in education, health, and agriculture); the European Union (transportation infrastructure); Great Britain (at present, Ghana’s largest bilateral donor with programs in public administration, health, education, rural infrastructure and agriculture); Denmark (health, energy, and water); the Netherlands; Germany; France; Canada; Italy; and Spain. 2.52 In terms of the composition of aid flows, total project aid (grants and loans) has been significantly higher than total program aid throughout much of the 1990s. And for both types of aid, the share of loans has been much higher than grants. This trend was however reversed in 2001 – program grant prose sharply to $148 million in 2001, compared to $44 million in 2000, and project grant stood at $71 million, compared to $62 million in 2000. The increase in program aid reflects a growing confidence and willingness by Ghana’s external partners to move toward more programmatic aid as Ghana continues to adhere to more prudent economic and financial policies and reforms.  19 Chart 2: Program Aid Chart 3: Project Aid 1992-2001: Actuals; 2002: Projection 1992-2001: Actuals; 2002: Projection 250 500 450 200 400 350 US$ millions US$ millions 150 300 250 100 200 150 50 100 50 0 0 1992 1993 1994 1995 1996 1997 1998 1999 2000 2001 2002 1992 1993 1994 1995 1996 1997 1998 1999 2000 2001 2002 Grants Loans Grants Loans 2.53 In terms sectors, commitments to social infrastructure and services (i.e., education, health, population, water and sanitation) have substantially increased in the latter part of the 1990s, edging out commitments to economic infrastructure and services (i.e., transportation, communications, energy, banking and financial services), which accounted for the largest share of aid commitments in the early 1990s. Commitments to the productive sector (i.e., agriculture, fisheries, industry, construction) have generally been lower throughout the 1990s, accounting for less than 15% of commitments. Table 6: Distribution/Purpose of aid commitments to Ghana Distribution/purpose 1991-1994 1997-2000 Social Infrastructure and Services 22% 38% Production Services 11% 7% Economic Infrastructure and Services 35% 27% Source: OECD/DAC Creditor Reporting System 2.54 The relatively large commitments of aid to social and economic infrastructure and services are in line with MDGs and general consensus in the international communities that infrastructure development is a prerequisite for socioeconomic development. This necessity notwithstanding, Ghana must also formulate appropriate strategies for using aid to boost its production sectors of the economy, which has a direct bearing on its export performance, and ultimately, its ability to curtail its over-dependence on foreign aid. Introduction of the CDF in Ghana 2.55 Ghana was designated by the World Bank, in agreement with Ghanaian government leaders, as a CDF pilot country in May 1999. The immediate effects in Ghana were: i) the formation of 14 government-led thematic “CDF-groups” for co- ordination and dialogue between the government and donors, ii) hosting of full CG  20 meetings in Accra, now chaired by the vice president of Ghana, and iii) monthly working lunches between donor representatives. The CDF was strongly promoted by the World Bank office in Accra and accepted by the previous government as a means to strengthen collaboration across all the development partners. The initiative enjoyed support from donors such as Denmark, the Netherlands and the United Kingdom, since the CDF principles matched their own thinking, while initial reactions from others such as the EU and US were considered lukewarm. 2.56 Though useful, neither the CDF-groups nor the quarterly mini-CGs (which pre- dates the CDF) can claim to have had broad or significant impact on the behavior and procedures of government and donors on the ground. Other sector working groups, pre- dating the CDF and linked to specific programs, have proven more effective than the CDF-groups, some of which never became effective. The much-cited sector wide approach (SWAp) in health pre-dated CDF (but fed into development of CDF thinking). The profile of the CDF has receded with the launch of the PRSP/GPRS. However, CDF thinking clearly influenced the global design of PRSPs, whose processes and activities were intended to embody the four CDF principles. In Ghana, there is a low awareness of the CDF concept outside donor and some government circles, and to avoid any identification between CDF and the previous government, the World Bank office has, since the change of government, downplayed the use of the term. 2.57 The most significant institutional effect of the launching of the CDF has been to change the orientation and image of the Accra World Bank office, which has very actively promoted the principles and sought to work in a more collaborative manner both with government and with other donors. The Bank has redefined the concept of the Country Assistance Strategy, as the Bank’s business plan of support to government projects and programs, in an attempt to move away from the notion of “World Bank projects.” The Bank has further decentralized responsibility and authority to the country office. The country director shifted most responsibility for task management to national staff, and reported significant improvements in the quality of the Bank’s portfolio as a result. 2.58 Initial enthusiasm about the CDF was considerable among Bank staff working on the Ghana program. But as events played out, and a new government took office, both the Country Assistance Evaluation from 200013 and the CAS 2000-200314 proved to be overly optimistic about the durability and importance of Vision 2020 as a holistic framework, the degree of country ownership it had, and how far partnership has deepened, compared to what this evaluation team found in place in 2002. 13. World Bank, OED: Ghana Country Assistance Evaluation, 2000, p. 9, p.17. 14. Government of Ghana and The World Bank: Country Assistance Strategy for Ghana 2000-2003.  21 Box 5: The “Work Out” In February 1999, the Bank held a “work out” attended by key actors from the government and Bank staff. The agenda was unstructured— the team was asked to answer only one question; “What would you do to improve the way business is done between the Bank and the government in Ghana?” The group sequestered itself and re-emerged two days later with their proposals. These included three main ideas: • The separation between the Board, the Bank and the government should be broken down. The government (rather than Bank staff) should present its development strategy to the Board as part of the CAS process. • The Board would approve a three-year lending envelope based on agreed results and overall program framework. The Bank Country team would be accountable for these results and staff salaries would be tied to achievement of associated indicators. • The Ghana country team (and a local Board of Directors) would then have authority to approve specific projects/programs within the three-year envelope in order to achieve agreed results. At roughly the same time, the Bolivia country team launched a “work out,” which yielded similar proposals for decentralization of authority and staff to the field. In March 1999, the Bank’s Board of Directors discussed the Bolivia proposals. Despite enthusiasm of staff, the proposals were not endorsed because it was felt that this “new way of doing business” implied a loss of Board control and a major shift in their oversight functions. As a result, the Ghana team had to abandon its original proposals to decentralize approval authority to the field and to tie staff salaries to results all had to be dropped. Instead, the Country Office focused on supporting CDF principles such as ownership and partnership in ways that did not require changing Bank rules and regulations or the functions and procedures of the Board. Summary 2.59 The deepening of democracy, more openness and involvement of a broader set of actors are strongly positive developments in Ghana’s recent history. Economic progress has also been made after almost twenty years of economic reforms, but the experience has been mixed. Growth rates have remained positive, but have never moved close to the 8% that was targeted in the Vision 2020. Overall poverty rates have fallen but within the context of increasing inequality. 2.60 The previous government was successful in introducing basic macroeconomic reform in the 1980s, but it did not complete the comprehensive “second generation” reform agenda. The privatization program has, in ten years, only covered 25% of public enterprises (by value). Public sector employment remains high, and civil service reform, including re-establishment of proper incentives to performance, has not advanced significantly.  22 2.61 The new government, which took office in January 2001, was elected on a liberal platform promoting a free enterprise economy as the guarantee for growth and prosperity. In addition to the unfinished reform agenda, the government also inherited a serious macroeconomic imbalance caused by pre-election overspending. 2.62 After successful stabilization of the economy, some initial steps have been taken to revitalize the reform agenda in key areas like divestiture, utility pricing, public financial management, and civil service reform. However, it is too early to see concrete results of these initiatives, and the capacity to implement reforms and the solidity of the government’s commitment in these highly sensitive areas are still untested. 2.63 Ghana continues to be highly dependent on foreign assistance, which finances roughly 90% of public investment expenditures. Even a strong political commitment to civil service and public financial management reform will not develop the capacity of the public sector in the short run although they are fundamental prerequisites to capacity building. Similarly, at a more fundamental level, the previously dominant patronage elements of the political system, though apparently weakened, will not disappear rapidly, and will continue to seek expressions in favoritism, ghost-workers in the public sector, and other expressions of corruption. 2.64 The application of the CDF principles will in any context imply often-large changes, many of which probably can only be achieved in the long-term. Even those donors and recipient governments most receptive to CDF concepts find they are constricted in practice by their internal regulations, managerial weaknesses or incentive systems. In the Ghanaian context of aid dependency, weak public sector capacity, and an overloaded reform agenda it is therefore especially important not to set unreasonably demanding success criteria for the initial period.  23 3. The CDF and Ghana's Poverty Reduction Strategy 3.1 Countries, like Ghana, which wish to secure HIPC debt relief and continuing access to soft-window credits from the IMF and World Bank are obliged to produce Poverty Reduction Strategy Papers that must eventually be endorsed by the Boards of both institutions. Ghana's Poverty Reduction Strategy (GPRS) is its response to this requirement, although preparations for it commenced long before Ghana decided to apply for HIPC relief. The GPRS has become the centerpiece process and product in Ghana that today embodies the CDF principles at the macro-level. It has attracted strong attention from all donors, and the draft GPRS was the key document in the April 2002 CG meeting in Accra. 3.2 A whole chapter of this report is devoted to the GPRS because it represents a major policy-formation exercise that can be used to assess the extent to which CDF principles have been complied with, and have influenced, the significant efforts that have gone into the preparation of the GPRS. The GPRS experience particularly serves to illuminate the principles of long-term, holistic framework and country ownership, thus focusing both on the content and the process of the GPRS. While partnership and results orientation aspects will be crucial, they are currently less visible. 3.3 The new NPP government entered with a long-term political vision expressed in the party manifesto “Agenda for Positive Change.” Notably, but not surprisingly, the manifesto makes no reference at all to Vision 2020, or to the GPRS process. However, President Kufuor’s new administration was quick to say that it would continue with the GPRS process, but that it wished to put its own stamp on it. Accordingly, the timetable was extended into 2002. Early in its term, the new government reversed previous policy by deciding to apply for HIPC debt relief - a development that increased the already considerable financial stakes attached to the GPRS. Having been forced initially to concentrate on restoring a rather chaotic macroeconomic situation, the Kufuor administration came eventually to see the GPRS as its first substantial opportunity to set out a more forward-looking and comprehensive statement of its economic and social priorities. As will be shown, this had a large bearing on subsequent developments. The Evolution of the GPRS 3.4 The process of preparing the GPRS started in July 2000. Responsibility was given to a special Task Force within the NDPC. Five “Core Teams” were established, with defined responsibilities for providing inputs into the GPRS. A mixture of information dissemination, collaboration, coordination, and consultation at national and local levels secured stakeholder participation in the process. 3.5 Consultants (mainly Ghanaian) were hired to synthesize the reports of the Core Teams, contributions by line ministries and other materials to produce a draft of the main volume of the full GPRS. This went through a number of iterations before finally being submitted for Cabinet discussion late in 2001. The deadline for submission was the Consultative Group scheduled for April 2002, at which time a second volume was to have been presented setting out costed priorities. This second volume was not ready in  24 time for the CG however, and was presented to the IMF and World Bank in July 2002. (though preliminary estimates were presented at the CG meeting). 3.6 At least until late 2001, most ministers had reportedly taken little interest in the GPRS process, regarding it as an activity of the Planning Commission with little bearing on their own work. Only in early 2002, when the Bank’s Country Director made it abundantly clear to the government that the GPRS would be the exclusive reference document on which donors would base their pledges, did the GPRS move into the mainstream of government policy formation. This change of status of the GPRS left a number of spending ministries (and a number of bilateral aid representatives) scrambling to make sure that their own programs were squeezed into the document. The result was that the GPRS lost coherence and credibility as a poverty-reducing strategy, as many only marginally relevant items were included. 3.7 Finally, the president and his advisers determined to place their own stamp on the product and to use it as a statement of the government's medium-term priorities (already foreshadowed in the President's State of the Nation address of January 2002) to implement the election manifesto on which the NPP had been elected. They did this by leaving the main text of the GPRS largely unchanged but prefacing it with an "Executive Summary" which, rather than summarizing the main text, asserted the government's general development priorities for the next three years. The document was also given a significant subtitle: An Agenda for Growth and Prosperity.15 It was this hybrid document that was submitted to the CG. Not covered by the document and not separately identified, were additional activities to be financed out of HIPC debt relief, which have yet to be determined. 3.8 The GPRS process thus generated an un-prioritized list costing far in excess of realistic resource possibilities. The government's presentation to the CG made it clear that, because the full costing of the total GPRS (then estimated at US$8.3 billion over 3 years) would far exceed any feasible resource availabilities, it would concentrate its efforts on securing financing for the medium-term priorities set out in the "Executive Summary" (then estimated at US$1.6 billion over 3 years). In practical budgetary terms it thus appears that the actual GPRS for 2002-04 will consist of: • the government’s medium-term priorities • any ongoing programs of line ministries and other MDAs that can be represented as poverty-reducing • additional MDA activities financed from HIPC revenues. 3.9 Box 3 summarizes the six largest of the medium-term priorities in budgetary terms (accounting for two-thirds of the total priority program16) and compares them with 15. Even more significantly, perhaps, the titling of the second GPRS volume, presenting costings, was reversed to read, An Agenda for Growth and Prosperity: Ghana's Poverty Reduction Strategy. 16. In the costings published by the GoG in July 2002, the total cost of the medium-term priorities over three years is put at $2,424 million. However, this includes a provision for (a HIPC-funded) reduction of domestic national debt and related items of $341 million. We have excluded this item, as being of a  25 the equivalent discussion in the main text of the GPRS. Two points emerge from this comparison: (a) the connection between the two components is weak; and (b) it would be difficult to present these priorities as constituting a poverty reduction strategy, or even a pro-poor growth strategy. Box 6 - The Medium-Term Priorities and the GPRS Compared Medium-term Priorities17 GPRS main text Energy: Only a brief, unspecific mention (p.75). Main Increase power output to boost industrial emphasis is on raising financial viability of growth & production. Includes investments utility corporations through improved pricing in a gas pipeline, and expansions of existing and cost recovery. Connection to poverty hydro and thermal power plants. ($450m.) reduction unstated. Major highways: There is reference (p.82) to completion of two Three major highways to be constructed to major roads, under the heading of 'Regional 'open up the country, introduce competition Trade Barriers.' No connection with poverty and create an enabling environment for the reduction suggested. Main discussion (pp.66- private sector.' ($318m.) 67) is on improved provisions for maintenance of feeder and trunk roads. Health Centers: Not mentioned. (Also not mentioned in Health Development of a model health center in all SWAp II Program of Work, 2002-06.) Main 110 Districts. ($220m.) stress in very brief discussion (p.98) is on reducing equity gaps, improving financing arrangements and raising efficiency of service delivery. Rural-urban link roads: Rehabilitation or development of one major Not mentioned. There is a general reference to road in every Region, linking rural road the targeting of 'some key trunk roads' (p.67) systems to urban centers. ($173m.) Girls' education: No special mention of girls' education, Not mentioned in February 2002 Executive although targets are set for improved female Summary but upgraded in July 2002 Vol. II: enrollment rates relative to boys. A package of measures to increase the enrollment and retention of girls, together with rehabilitation of classrooms. ($126m) HIV-AIDS: 'AIDS and poverty are intricately linked… Not mentioned in February 2002 Executive Resources will… be provided for care and Summary but upgraded in July 2002 Vol. II: support for those living with the disease…' 'care to be provided to people living with HIV/AIDs, and support provided for orphans, vulnerable people, the physically challenged and the aged.' ($110m.) different nature from the rest, leaving an adjusted total of $2,083 million. The six items shown in Box 4 total to $1,397 million, or 67% of the adjusted total. 17. Arranged in descending order of budgetary cost (provisional estimates).  26 3.10 Against this background, the remainder of this chapter examines the consistency of the GPRS processes with the four CDF principles. Long-term Holistic Development Framework 3.11 Questions under this heading include: (a) whether the GPRS is informed by a long-term vision; (b) whether it gives adequate attention to human aspects and strikes a reasonable balance between macro, structural and social dimensions; and (c) whether it covers all sectors and inter-linkages between them. The link between budget processes and the GPRS is discussed under results-orientation. 3.12 To address the first of these questions, it is essential to remember the political context in which the GPRS was prepared. The previous NDC government had, in fact, prepared a major document setting out a vision of the country's development from 1995 to 2020. However, this statement was viewed by the incoming NPP administration as too closely associated with the politics of their predecessors and they quickly announced that Vision 2020 had been abandoned. The pre-election spending-spree by the NDC dictated that the new government had no alternative but to first attend to an immediate macroeconomic crisis - which had been brought under reasonable control by 2002. With the GPRS document substantially completed (without the benefit of any formal long-term vision statement), the government's next task will be to comply with a constitutional requirement to present to Parliament 'a coordinated program of economic and social development policies…' within two years of taking office, i.e. by January 2003. This is widely interpreted to require preparation of a medium-term development plan but would also be consistent with a long-term vision statement. The government in fact intends to produce both and preparations have already begun. 3.13 Judged by the yardstick of the CDF, the actual sequence appears to have been precisely the opposite of that implied by the CDF: from short-term crisis management to preparation of a medium-term poverty strategy to the future development of a more comprehensive medium-term plan and of a long-term vision statement. This progression is, however, explicable in terms of the politics. Moreover, it can be argued that the new government does actually have a consistent, if only very broad, vision of where it wants to go with its economic and social policies - in favor of wealth creation led by development of the private sector.18 3.14 What about the issue of balance between structural, growth and social considerations? Without doubt, the GPRS has represented a much more substantial and serious attempt to place the poverty problem at the top of the policy agenda than has ever before been attempted in Ghana. It also demonstrates the need for wide-ranging change and initiatives on a wide front, if satisfactory poverty reduction is to be achieved. There is much in it of value from this point of view and concern to improve the welfare of the poor pervades the document. In that sense, it presents a better balance than obtained under the status quo ante. That said, even the main text of the GPRS is already strongly 18. Expressed in NPP’s election manifesto: Agenda for Positive Change.  27 influenced by the growth-maximizing agenda of the NPP. Thus, its analysis of the poverty situation takes just 5½ pages and contains virtually no discussion of the causes of deprivation, whereas its subsequent discussion of long-term economic growth is much more expansive (21 pages). Its treatment of the social sectors (significantly under the heading of 'human resource development') is also skimpy, with health issues (other than HIV/AIDS) claiming a mere three pages. 3.15 In the event, such criticisms have been made largely redundant by the super- imposition of the medium-term priorities. It is these that best express the government's priorities and which are most likely to be implemented. As can be judged from the left- hand column of Box 4, this is a growth strategy with weak claims to be considered pro- poor. This focus was made even clearer by the President's January 2002 State of the Nation address, which already set out most of the medium-term priorities but which (apart from a brief discussion of HIPC) made no reference to poverty at all. As President Kufuor put it, "…we were not hoping to base our economic recovery program on HIPC. Nor indeed on the GPRS alone… My government's avowed determination and economic program was, is and always will be to create wealth in this country and thereby combat poverty."19 No doubts about trickle-down there.20 It is not suggested that the government is in any sense hostile to the poverty reduction agenda. Simply, that is not its priority. 3.16 What about the coverage of all the sectors and the interlinkages between them? Even if we confine ourselves to the GPRS, the score here cannot be very high. One of the structural weaknesses of Ghana's GPRS process was that it did not fully incorporate the MDAs into the preparatory process. As already described, it was only at a late stage that the MDAs began to take much active interest, with the end-result that the final document is a patchwork of specific measures, some pro-poor, some with little obvious connection to poverty reduction, with little prioritization or overall coherence. 3.17 The cost of the full GPRS, far beyond resource availability, adds to the question about the nature of the GPRS – is it a policy, a strategy or a plan? Although it is called a strategy paper, it is hard to work out what the intended strategy is, except that it is growth-led. As the paper's authors frankly admit (p.177) in response to a criticism of this aspect, “The document requires a more cohesive framework…An attempt has been made within very, very limited resources, to improve the document in this regard.” There is no pretense at taking account of sectoral inter-linkages in any formal modeling sense. The insertion of the medium-term priorities, mostly only loosely related to the main text of the GPRS, increases the incoherence. 3.18 It can further be asked whether there is enough human capacity available within Ghana for it to be reasonable to apply the cross-sectoral criteria for assessing the GPRS. Without question and by general consent, the NDPC is weak. A dedicated and able group 19. At this time the GPRS was thus mainly seen as an instrument relevant for the HIPC process. 20. In line with this assessment, Canada presented a paper in the CG meeting stating that the GPRS is “a country-wide growth strategy with an implied trickle-down positive impact on poverty reduction...However, it is not clear how the Strategy will ameliorate the lot of the people in the most poverty stricken regions.”  28 of people prepared the GPRS but it was a serious flaw in the initial conceptualization of the GPRS to entrust the task to an ad hoc Task Force, operating outside the normal workings of the civil service and located within a weak agency not only starved of resources but semi-detached from regular policy-formation processes. Against this, the present planning minister is held in high regard and he has moved to strengthen the NDPC by drawing upon the large and well-qualified Ghanaian diaspora to assist him as advisers. There is also a substantial capacity within Ghana, working in think tanks, the universities, and NGOs that can be mobilized for special tasks. Admittedly, the availability of such resources is no substitute for having a strong, well-motivated civil service. But the answer to the question is affirmative: the capacity does exist in the country, combining forces, to achieve the holistic, balanced and integrated outcomes sought within the CDF philosophy, if enough priority is given to doing so. In fact, Ghana has a reputation for producing technically sound plans and policy documents. The main problems start when it comes to implementation. 3.19 Lastly, the GPRS case raises questions about the appropriateness of the CDF's stress on the importance of a country having a long-term holistic development framework. There is no doubt that it is desirable for a country's leadership to have a clear sense of direction, and for it to be able to carry the population with it, and no doubt a country must balance macroeconomic and social policy concerns. But the team is more doubtful about requiring anything more formal and worked-out than that, at least in the context of Ghana's political and democratic traditions. Given a four-year electoral cycle, the feasibility of a long-term (i.e. at least 10-year) vision statement necessitates cross- party agreement and, therefore, a consensus-building political culture and institutions. 3.20 These conditions do not readily apply in Ghana. Democratic traditions are still too new, the influence of clientelist systems of reward remains too strong, the base of educational achievement is still too limited, information flows too poor, and local-level participation still very incomplete. A top-down tradition still dominates and decentralization has far to go. There are too many uncertainties, compounded by heavy external dependence and weak implementation capacity, for long-range planning to be either feasible or sensible. No doubt, it could have been done better, but it is certainly no surprise that Vision 2020 had little practical effect, nor that it was abandoned when a different political party came to office. The GPRS may fare no better fate. Country Ownership 3.21 The CDF ideal of ownership, as reflected in the background information provided for those taking part in the CDF evaluation, specifies that the country should be in control and that, for this condition to be satisfied, the government of the day should build a consensus, within itself and other levels of government and with civil society, private sector and other domestic stakeholders. This, in turn, requires extensive consultative processes. There is a clear sense in which the GPRS, in its final form, will be Ghana- owned, but, as the next pages will explore, the concept of national 'ownership' is elusive. We will look first at the processes leading up to the submission of a draft of the main volume of the GPRS to the Cabinet at the end of 2001, concentrating particularly on  29 processes after the election of the Kufuor administration, and then examine subsequent developments. (a) Preparing the GPRS 3.22 Particularly after the election of the NPP administration, with its more open style of government, there is no doubt that a major consultative effort was made. Donors were, quite properly, treated as legitimate stakeholders in the process but this evaluation team found no evidence that donors - and particularly the IFIs - tried to take control of the process. The local representatives of both the Fund and the Bank made a conscious decision to remain at arms-length and were diligent in maintaining this posture. Bilateral donors too were generally content to play a supportive role, although there are complaints that, as time wore on, their restraint became eroded and that some of them actively lobbied for inclusion of their own projects. 3.23 There were complaints, perhaps inevitably, from within civil society that there was a degree of formalism in all this: that it was consultation rather than participation, that the NDPC saw the various fora of consultation chiefly as opportunities to explain what they were doing and that they were reluctant to engage with the more radical critics of the mainstream political end economic model. However, the GPRS authors could point to many changes, set out in a rather impressive appendix of the report, in response to comments on earlier drafts. 3.24 There were other respects in which the consensus-building processes could be faulted. Parliament was little involved at any stage, although this was partly an unavoidable reflection of the occurrence of an election in the middle of the process. In the initial period there was very low public awareness of the GPRS and, although press coverage greatly increased as a result of the controversial decision to apply for HIPC relief and subsequent publicity given to the medium-term priorities, the general public still has a rather low awareness of the GPRS. There was criticism of the opaque way in which the Core Teams were put together, with the process firmly in the hands of the government. Observers found that CSO team members were chosen on the basis of their technical abilities rather than because they came from organizations with a broad social base (e.g. churches and trade unions). They were also largely Accra-based. The subject areas covered by the teams scarcely corresponded to the structure of government. As a result, key ministries and other agencies, such as agriculture, health and education were given little prominence in the teams and some of them sent relatively junior representatives. One consequence of this was the weak buy-in of MDAs in the early stages of the process. Another twist to this was provided by the ongoing tension between the NDPC and Ministry of Finance. Indeed, at this stage of the process, the commitment of the whole government was questionable, with the GPRS in danger in being viewed within government as simply an NDPC product. 3.25 Two other qualifying factors merit mention. The first was the debilitating effect of time pressure on the quality of the consensus-building process. This was exacerbated by the decision to apply for HIPC debt relief but was present from the earliest days. In retrospect, with the process still not quite completed nearly two years after it commenced,  30 this may not look like a rushed process. But it was intended to be completed much more quickly and working with a succession of deadline extensions is not at all the same as working from the beginning with a two-year schedule. 3.26 Lastly, Ghana’s heavy aid dependence qualifies any story of national ownership. While the NDPC was able to stand firm against the lobbying of some donors, the key original motivation for preparation of the GPRS was to negotiate credits with the World Bank and the IMF. The subsequent decision to apply for HIPC debt relief required Ghana to prepare a PRSP, as conditionality for HIPC access. We received complaints that there was a good deal of self-censorship by the NDPC and its advisers, so as to make sure their final product would be broadly acceptable to creditors and donors. It is also important to bear in mind here that the macroeconomic section of the GPRS is directly derived from the current IMF-approved program. 3.27 Despite these qualifications, the draft GPRS had a stronger claim to be a nationally-owned and consensual document than probably any other major statement of policy since independence in 1957. It did, though, suggest that the consultative process might have constrained the taking of hard choices (e.g. there was little discussion of trade-offs between market-led growth and poverty reduction or how to address inter- personal inequalities). (b) Asserting the Medium-Term Priorities 3.28 The decision to superimpose the government's medium-term priorities on the GPRS completely changed the above story. On the positive side, it removed doubts about the commitment of the NPP government to the resulting document. The president is now strongly and publicly committed. Line ministry commitment, of a sort, has been secured through their efforts to make sure that the process adequately safeguards their own interests. Another important sense in which assertion of the medium-term priorities could be seen as strengthening Ghanaian ownership (although it is not clear how deliberate this was) was in the extent to which these priorities demonstrated a willingness to depart from the PRSP agenda set down under HIPC conditionality and the anti-poverty preferences of major donors. 3.29 In other respects, however, the move was a clear departure from the CDF precepts of ownership cited earlier. The rather elaborate mechanisms of consultation described earlier were effectively set aside, for there was nothing participatory at all about determination of the medium-term priorities. Indeed, the process of preparing the "Executive Summary" and, especially, the second volume of the GPRS, were entirely closed-door affairs, to which neither civil society nor bilateral donors were privy. If the suggestion earlier that it is the medium-term priorities, plus yet-to-be-identified (again by non-participatory means) HIPC expenditures, that are the aspects of the final GPRS most likely to be implemented then it has virtually no claims to be a consensual document. 3.30 This may provide the creditor-donor community (and especially the World Bank) with a dilemma. The NPP government has a strong defense to offer. It is recently and freely elected with a decisive majority and can reasonably claim that the thrust of the medium-  31 term priorities is in line with the policies put forward in its election manifesto. It can further argue that Ghana is not one of the countries most seriously affected by poverty and that other policy objectives are important too. It gives priority to the maximization of growth and takes a sanguine view of the beneficial effects of general growth on the welfare of the poor. It can therefore argue that its priorities are poverty reducing. Moreover, this is undoubtedly a government, which donors like to deal with, in part because of the priority it gives to fiscal prudence, and its pursuit of a development path led by market forces and private enterprise. Donors clearly expressed that they find their relationships with the government easier than they did under successive Rawlings administrations. 3.31 All these factors may predispose donors to accept what the government presents to them as its PRSP, though the final document—with the insertion of the medium-term priorities—does not possess many of the characteristics normally looked for in a PRSP. It is not broadly consensual. Rather it is the embodiment of a party platform. Its medium- term priorities pursue a growth process that —with almost half of all planned investment devoted to major power supply, and highway construction projects, and health centers for all Districts (regardless of poverty status)—cannot plausibly be portrayed as giving first priority to poverty reduction. Its planned actions are not reasonably linked to a diagnosis of the country's poverty problem. And underlying these factors is a clash of goals, between the normal poverty-reduction priorities of donors and the growth priority of the Kufuor administration. 3.32 Combining the holistic vision and country ownership principles, the GPRS story reveals a rather sharp conflict between the political mandate of a popularly and recently elected government, and the outcomes of consultative GPRS processes. Some degree of tension between the two was probably inevitable—and has been observed in other countries—but the situation could have been made much easier had the Kufuor administration realized at an earlier stage in its life that the GPRS was important, not least because of its likely influence on future aid flows. Had the political leadership sought soon after their accession to power to ensure that the GPRS consultation processes also gave due weight to the government's own political priorities, the conflict between the two could have been better managed and the tensions minimized. As things now stand, the dilemma for donors is to decide to what degree they will stick to the outcomes of policy processes they themselves favor (i.e. a broad-based and consensual approach, albeit resulting in an unprioritized GPRS) and to what degree they will accept that a democratically elected government with clear commitment to a growth policy which is not notably pro-poor should have it its way. Partnership and Donor Behavior 3.33 The general issue of the nature of government-donor relationships and of trends in donor policies and practices is dealt with elsewhere in this report. A study of the GPRS does not add very much under this heading. Just a few points may be made: 3.34 As reported above, most donors, particularly the IFIs, maintained an arms-length posture during the preparation of the GPRS, while standing ready to provide support when asked, e.g. by way of provision of expertise or money. At least some of the bilateral  32 donors were heavily engaged in the consultation process described above but not in the sense of seeking to control it. However, this rather exemplary behavior came under some stress in the latter stages, when some donors were reportedly active in trying to make sure their own favored projects/programs were included in the GPRS, in some cases to the extent of threatening that aid volumes might otherwise be reduced. 3.35 Mixed signals have so far come from bilateral donors about their likely responses to the GPRS. A fairly wide range of criticisms were voiced at the April 2002 CG but also quite a lot of support. A total of about $1 billion was promised for FY2002. This was regarded as a vote of confidence, although past experience suggests that actual receipts are likely to be well below this promised level. More generally, donors appeared still to be in the process of considering their response to the GPRS at the time of the present study. 3.36 Not too many changes should be expected from the bilaterals.21 Most are increasingly committed, in principle and, in a few sectors, in action, to moving away from stand-alone projects toward more programmatic aid and SWAp approaches. With the exception of one, all other donors participate to a limited extent in programs that are jointly funded with other donors and are implemented by government. Encouraging, there is a declared intention to move toward a higher proportion of general budgetary support on the part of a number of the more sizable bilateral donors - a development which can be regarded as within the CDF spirit. 3.37 However, most donors interviewed conditioned final decisions for increased budget support on improvements in financial management and procurement, which are unlikely to materialize in the short term. Despite the positive trends, many donors thus remain deeply skeptical and are moving at a measured pace due to deep-rooted institutional weaknesses in most sector ministries—lack of transparency, weak financial accountability, etc.—that require more than the GPRS to fix. 3.38 Donors do not expect the adoption of the GPRS to make more than a marginal difference to the composition and size of their aid, as between program and project support, although it might well affect decisions about the sectors in which aid would be located and projects selected. Aid commitments by donors have over the last decade shifted toward the social sectors and poverty reducing activities (health, education, water & sanitation, rural infrastructure), and the introduction of the GPRS, along with increased availability of resources through the HIPC debt relief, is likely to accelerate the trend. Most donors have, however, adopted a ‘wait and see’ attitude to what the potential impacts of the GPRS would be. 3.39 It is too early to tell the impact of the GPRS on the design of donor’s country aid strategy papers. Most donors stated, however, that although it was too early to be sure, there was growing acceptance by both donors and sector ministries that the GPRS is 21. As part of this Country Case study, a survey of 13 major donors in Ghana was conducted (responses attached in Annex 1). Comments on donor experiences and perceptions are drawn primarily from these surveys and follow-up interviews. Donors responding to the survey include Denmark, Canada, EU, FAO, Germany, France, Japan, Netherlands, UNDP, UK, UNICEF, US, and the World Bank.  33 Ghana’s main development strategy. However, this related to the draft full GPRS, prior to the confusion caused by the assertion of the government's medium-term priorities. What the present attitude may be is unclear. Among 13 major donors consulted by the team, almost half expressed that there was better alignment of approaches with the country’s development strategy, although strongly emphasizing that there was more alignment with the health sector than others. Some donors, not all, have also already been moving down more consultative, consensual strategy preparation paths and the advent of the GPRS may not necessarily constitute fundamental differences in their approaches. Results Orientation 3.40 The CDF approach requires that application of its other principles should be linked to concrete results and monitorable performance indicators. How does the GPRS score by that criterion? It has been mentioned earlier that previous policy statements on the reduction of poverty had not been implemented and hence had had no impact. Is the GPRS likely to suffer the same fate? 3.41 On the face of it, the institutional base is unfavorable. The NDPC is held in low regard within the civil service and elsewhere, even though members of the GPRS Task Force are well respected. By comparison with the MoF, with which to some extent it is in a competitive position, the NDPC is seen as distinctly the weaker. The Task Force responsible for the GPRS only consisted of five professionals. Moreover, they worked without support: no accounting officer, no administrative assistant or secretary. One of the NDPC’s specific weaknesses, potentially serious for implementation of the GPRS, is a nearly zero capability in the area of monitoring and evaluation. In the districts, where the results benefiting the poor should materialize, both implementation and monitoring capacity is limited, both in the District Administrations and in district-level units of sector ministries. 3.42 Substantial efforts are under way to strengthen the capabilities of the NDPC. The GPRS itself is replete with a large number of specific, potentially monitorable targets. An appendix to Volume I sets out specific plans for monitoring the implementation of the GPRS. Expert help has been drafted in and more is in the pipeline,22 with the appendix proposals regarded as a sound starting point for strengthening monitoring capabilities. The NDPC is also planning to produce an annual 'Book of Government Business,' which will set out specifically what each MDA is expected to do during the year. There are also pilot plans for using the GPRS as a basis for the preparation of District-level plans. 3.43 One of the most puzzling weaknesses of the GPRS preparation process was that it effectively excluded the Ghana Statistical Service. Nevertheless, the GSS has set up a new poverty-monitoring desk and has active plans for the administration in 2002 of a fairly large scale CWIQ survey (with a sample of 48,000 households), intended to provide rapid production of indicators, disaggregated (for the first time) to the District level. The design of this questionnaire will be modified for the purposes of the GPRS. In 22. A draft inception report, “Ghana Poverty Reduction Strategy Monitoring and Evaluation Plan,” was published in July 2002.  34 2003 the GSS further plans to undertake a major living standard survey, covering a smaller sample but in greater depth than the CWIQ. Unfortunately, the next planned activity thereafter is for a follow-up CWIQ in 2006-07, meaning that it will not be phased to permit testing of the GPRS targets (which relate to 2004). 3.44 So real efforts are being made, even though the institutional and resource base remains weak (e.g. no funds were available to produce and disseminate popular vernacular versions of the GPRS). But the question needs to be asked, what meaning do the GPRS targets have and, therefore, what is the likely usefulness of monitoring them? There are two reasons for raising this question: (a) For a target to be meaningful there needs to be some established and quantifiable relationship between the targeted variable and the policy instruments utilized to influence that variable. We doubt whether that condition has been satisfied for more than a very small proportion of the targets in the GPRS. For example, a wide range of highly specific targets is set for educational achievements by 2004 (GPRS: 87). However, the brief textual discussion of educational policies is not related to these targets, so the basis for these is unclear. The same weakness affects most other targets in the document. (b) However, even if we assume all the printed targets to be analytically well grounded, what meaning do they have now that the main text of the GPRS has been overtaken by the medium-term priorities? As explained above, what is likely to be implemented through these priorities is only a modest part of the GPRS total and, for the most part, is only loosely related to it. For example, the GPRS targets indicate a falling share of total spending on secondary education and a sharply rising share for primary schooling. But the medium-term spending priorities concentrate on measures to improve girls' education and an extension of pre- school provision for all children, neither of which is to be found in the main text of the GPRS at all. How in this case should we view its educational targets?  35 Box 7: The GPRS and the Millennium Development Goals To what extent are the targets in the GPRS consistent with the global MDGs? The fact that the NDPC is currently working with UNDP on the reconciliation of these two sets of targets suggests that the GPRS was not much influenced by the MDGs. However, it does contain a mention of them which is worth quoting in full: "The GPRS attempts to outperform the [MDGs] cast in terms of reduction of extreme poverty, social and human development, environmental sustainability and regeneration as well as democratic accountability, protection of human rights and the rule of law. These targets presuppose attainment of an overall GDP growth averaging at least 6 percent per annum and implementation of public policies for drastic spatial reorganization of investment for economic activity and social protection." The same question can be posed here as in the main text, about the realism of the GPRS targets. The Medium-term priorities provide no basis for expecting the "drastic spatial reorganization" of economic activity. The growth target of at least 6 percent should be compared with an estimated annual growth rate in 1996-2001 of a little under 4 percent. 3.45 A basic condition for the GPRS to be given any practical reality is that its provisions should be incorporated in the annual budget. Here again the story is one of good intentions but mixed outcomes. Take first the budget for 2002.23 During most of the budget preparation phase the GPRS process, itself still under way, was practically ignored, the situation not being helped by tensions between the Ministers of Finance and Planning and their respective agencies. MoF officials were reportedly told to 'concentrate on 2002' rather than taking a medium-term view. The situation changed after the Cabinet meeting at which the president swung his support behind the modified GPRS, which incorporated the medium-term priorities. There was a large amount of last minute to-ing and fro-ing in order to incorporate as much as possible of the modified GPRS into a budget that had been nearly completed. 3.46 Given this last-minute rush, quite a lot was achieved in reshaping the budget. A study by Armah (2002) shows that, in broad terms, changes in expenditures were in line with GPRS intentions, with the shares of the social services and other priority sectors rising (except infrastructure and agriculture) and a substantial decline in the non-priority category of 'administration.' However, MoF evidence shows that budget allocations may not have much bearing on eventual outcomes, with very large differences between the composition of budgeted and actual spending. Indeed, comparing the 2002 budget with preliminary estimates of actuals for 2001 reveals a different pattern altogether, with the share of the social sectors actually going down a little and administration reduced only marginally. There is also some question about the extent to which the budget 23. Ghana's fiscal year coincides with the calendar year.  36 reallocations away from administration were 'real', in that they took the form of reduced provision for a vote, which had previously acted as a type of contingency fund. With this fund now much reduced for 2002, the unresolved question is what happens when new contingency requirements occur? If these needs are funded, what then gives way? 3.47 Looking beyond 2002, the best prospect for progressive integration of the GPRS into budgetary processes is through the Medium Term Expenditure Framework (MTEF) mechanism. On paper, this provides a ready-made mechanism. Like the GPRS, it adopts a three-year time frame, it seeks to relate budgetary provisions to statements of MDA objectives and missions, which can readily be adapted to an overall poverty reduction goal, and tries to establish links between budgetary provisions and service deliveries. But, as already noted above and further discussed below, the MTEF and the broader PUFMARP has yet to exert influence on everyday fiscal realities. 3.48 Lastly, we should mention the expenditures associated with HIPC debt relief. A fifth of these are to be devoted to the retirement of domestically held public debt but the remaining 80% is to be devoted to various poverty-reducing lines of expenditure to be agreed between the government (presumably the MoF) and the IFIs on behalf of the creditors. Such expenditures will be folded into the GPRS. A monitoring system, over and above normal budgetary procedures, is already being put into place. MDAs have been asked to identify additional lines of expenditure that can be proposed as suitable uses for HIPC money. Given the amount of international scrutiny to which agreed programs will be subjected, it is reasonable to expect that these, at least, will be quite effectively monitored. 3.49 All in all, then, the GPRS marks a step forward in the direction of an increased results orientation. Despite the institutional weaknesses and ongoing problems, serious efforts are being made. At the same time, though, these weaknesses and problems are large and deep-seated, so it would be unrealistic to expect a quantum leap forward. If the effort is sustained, the most realistic hope is that an incremental trend of improvement will be set in motion. That would be no small achievement. 3.50 Considering the history and context of the GPRS, the study team finds that neither donors nor national actors should focus exclusively on the GPRS and make unreasonable demands for its inclusiveness and resilience, nor build up special purpose monitoring schemes narrowly focusing on the GPRS. By doing so, they risk overlooking other important policy processes and broader institutional capacity development processes, and the GPRS needs to be seen as part of a wider scheme.  37 4. Long-term Holistic Development Framework 4.1 While the GPRS and other macro-level strategies and plans intend to provide an overall framework, efforts have also been made to establish holistic development frameworks—or comprehensive policies—in various sectors and cross-cutting areas, including agriculture, health, decentralization and economic governance. Especially in health, the successive five-year sector strategies have served as a framework for both government and (most) donor actions, as shown in Box 6. In agriculture, the policies have been fairly stable, but implementation has been delayed. In education, there is no comprehensive policy framework. Below, the main focus of the discussion is on health and economic governance. Box 8: The Sector Wide Approach (SWAp) in Health – an Overview of Achievements The Health SWAp is the most advanced such arrangement in Ghana (and one of the most advanced in Africa). Although it predated the formal development of CDF principles, it fed into the development of the CDF concept and has become the most serious articulation of its principles in Ghana. The Health SWAp has a strong reputation among donors and others, and is seen as having enabled: - The development of two successive five-year sector strategies built on broad consensus between the Ministry of Health and donors. - Consolidation of the funding of most major donors to the Ministry of Health into a common basket, although a substantial number of other donors remains outside (including other health related areas like water and sanitation). Donors use common management arrangements to govern operations and a joint appraisal process to evaluate progress. - Translation of the 5-year plans into annual work programs, which are reviewed through a joint appraisal process at major annual summits attended by both donors and GoG. Substantive issues are debated at length and the forthcoming year’s work plan agreed. - There are also quarterly and monthly MoH-donor meetings. Donor-partners occasionally meet by themselves but keep that to a minimum. - No donor is supposed to operate outside of the sector plan, although it happens in practice. In a few important cases, donors have exerted peer pressures to force would- be backsliders into line. - More predictable financing through quarterly releases. Agreed procurement and financial systems have been built up over time to satisfy donor standards. - Standardized reporting. Through the joint appraisal process, all donors receive the same annual report and financial statement (although many also require supplementary information for themselves). Previously there were separate reports and review missions for each of the roughly 30 donors in the sector. - On occasion, donors and MOH staff have acted collectively on certain policy disputes and have lobbied the MOF to ensure steady financial flows.  38 4.2 In health, the first 5-year Program of Work (SWAp I) was set in the context of the then-extant Vision 2020 but that has since been abandoned and has not yet been replaced. So, SWAp II (2002-06) stands alone for those five years. Moreover, SWAp II was completed before the GPRS. 4.3 Is the SWAp II consistent with known wider national aspirations? If these are taken to include poverty reduction, there seems to be good congruence between this and the main text of the GPRS. While SWAp II does not contain much explicit discussion of poverty reduction, its general thrust, and many of its particular provisions, are pro-poor – and more so than SWAp I. However, the provision in the ‘Medium-Term Priorities’ of the GPRS for ‘model health centers’ in every district is not included in SWAp II and seems likely to compete for resources with the priorities set out there. It is not yet clear how this apparent conflict will be resolved and what impact it may have on the credibility of the SWAp. 4.4 Ghana's Health SWAp offers an excellent example of the benefits that can be derived from local leadership and ownership. It was born out of a strong mutual agreement on the need for a radical break with past practices, seen by the Ministry of Health and donors alike as failing to deliver results at grassroots levels, and it was made possible by an enlightened, skillful and patient MoH leadership that gradually built a coalition in favor of what at the time was seen as a radically new way of doing business. As described in Box 7, a consensus-building approach was adopted to the formulation and execution of health programs, under MoH leadership, leading to a more coordinated, less impossibly demanding process, and moving away from a large number of individual, typically donor-initiated and extra-budgetary, projects toward a pooling of resources within the health budget, linked to more highly articulated processes of dialogue and monitoring. 4.5 The health case under review thus points up some of the limitations of single- sector SWAps. What can be achieved in health depends on a lot of factors, which are outside the purview of the SWAp and MoH, (such as the fact that Ghana’s total expenditure on health is only US$7.50 per person per annum). Also, there are complaints that SWAp II is too narrowly focused on public sector delivery, to the exclusion, e.g., of church, NGO, and private provisions. 4.6 Also, SWAp II calls for better integration of its own efforts into the wider context of GoG governance reforms. This is particularly the case in relation to decentralization, where there is a clear clash between the vision developed for health – which contemplates a nationwide health service governed and managed by the Ministry of Health and other central level institutions – and the vision for decentralization, which sees delivery of health services at the district level integrated under the district assemblies. 4.7 The existence of incongruent and—as in the case of health and decentralization— openly conflicting visions—is by no means an anomaly in democratic societies, since holistic visions embody differing institutional, professional or socioeconomic interests of individuals and groups. The key issue is rather how the conflicts and frictions develop. In  39 the case of the conflicts between health and decentralization authorities, the impression of the study team is that both sides have become entrenched by their own partial visions: on the one hand, the decentralization concept applied fails to acknowledge that some sectors, because of the nature of their services and clients, need mechanisms of coordination and division of authority that are inappropriate for other sectors. One concept is not suitable for all. Health professionals, on the other hand, display strong reluctance, not least rooted in the ethos of the medical professions, to accept that non-health authorities could play an important role related to health – even though health impact, as is well known, is a cross sectoral issue. 4.8 As already described in chapter 2, the area of public sector management reform and, specifically, economic governance reform, has been in focus for quite an extended period, but the desired or expected impact has not materialized. 4.9 The disappointing results cannot be explained with reference to the absence of a broad vision or development framework as such. There is little doubt – and, on the surface, little disagreement – that Ghana needs a slim, “right-sized” and honest civil service, and that prudent and transparent financial management should take place in a rolling 3-year budgetary framework linked to objectives and activities, with strong correlation between budgets, transfers and actual expenditures. 4.10 A number of more specific development frameworks – the National Institutional Reform Progam (NIRP), the Public Financial Management Reform Program (PUFMARP) have also been developed, detailing a comprehensive set of specific actions, which are organized in separate components. At this level, however, it is more difficult to discern a clear-shared vision. Different components were designed in separate processes, depending on which donor they were supported by. Critics have further pointed out that the PUFMARP program has had an excessive focus on information management and information systems, seeing grand scheme computerization as the solution to problems that are linked to definition of roles and responsibilities, vested institutional interests, and to weak incentives for adherence to rules and regulations. 4.11 Efforts to enhance comprehensiveness and coherence of the specific reform processes related to economic governance are thus urgently needed but yet to emerge. In recognition of the need for greater coherence, government and aid partners have recently taken initial steps to intensify analytical work and dialogue through the formation of a special task force related to economic governance. Similar processes are underway in relation to other themes and sectors (agriculture, decentralization). 4.12 A key challenge in these renewed processes of analysis and dialogue appears to be the need for “right-sizing” the holistic visions or the development frameworks to the available capacity for change. The notion in the GPRS that the public sector reform program might be beyond the capacity of available human and financial resources to plan and implement, points to the need for careful consideration of sequencing and complexity levels of reforms and specific programs.  40 4.13 In summary, and including the GPRS, there is progress toward a long-term holistic development framework in Ghana, but it is still limited. This is true whether a framework is understood as a policy/strategy framework or as an operational matrix-like framework for concrete actions. 4.14 This limited progress invites caution about how to interpret and operationalize the holistic framework-principle in a situation where both national policy formulation and implementation capacity is limited, and where uncertainties are many and dependency on external aid considerable. While a clear sense of direction is without doubt desirable, it may simply not be feasible or sensible to engage in more binding longer-term planning.  41 5. Country Ownership 5.1 This section focuses on two aspects of country ownership: First, the strengths and limitations of national participatory and democratic processes and their ability to foster broader consensus, and secondly, the degree of government ownership in such key sectors as public sector reform and health. 5.2 Most fundamentally, Ghanaian democracy has reached an important milestone. The free and fair election of a new government is a primary assertion of country ownership. Though elections and democracy are not often linked directly to the CDF principle of ownership and not widely discussed in this context, the improvements in the quality of the democratic processes is maybe the most significant advance toward broader country ownership.24 5.3 Donors have generally been supportive of a shift toward country ownership at policy level (GPRS, the Health SWAp, the agricultural policy, support for CSOs & dialogue processes). There is a clear appreciation at conceptual level among donors that country ownership is important for sustainability, and that country ownership implies a broader group of stakeholders than the government. Notably, donors have sought to build capacity in think-tank type NGOs to stimulate national policy dialogue, and bilateral donors, often through northern NGOs, support several CSOs with claims of broader representation. 5.4 However, by general agreement, the quality of the domestic dialogue leaves room for improvement. On the one hand, the organization of dialogue events is often poor, with inaccessible documentation, which often arrives late, absence of material in vernacular, and not enough time for consideration. On the other hand, there is a tendency to formalism and to read ‘consultation’ as dissemination and validation of pre-set government intentions, and some issues are perceived as non-negotiable by CSOs. The broader CSOs themselves often lack the capacity and resources to engage seriously on policy issues that require expertise and research. The country’s think tanks has an apparent potential for being tapped more for this purpose, as well as to augment capacities in government and parliament. 5.5 The machinery of consultation/dialogue still tends to be ad hoc, and is not institutionally embedded, partly because civil society representatives are themselves divided on what they want. Parliamentary Committees and District Assemblies are not systematically included in the processes, thereby fostering a potentially dangerous divide between consultative processes and the representative democracy. Parliamentarians and District Assembly representatives may conduct their own consultative processes in their 24. In an interesting analysis of aid and reform in Ghana, Tsikata argues that democracy in Ghana implied less reform capability, because decisionmaking could no longer be reserved to the group of technocrats who had ably managed the economy since 1983. However, though both electoral cycles and democratic features like labor unions may have made macroeconomic management more difficult, second generation reforms – which were the key agenda when Ghana became a democracy – can typically not be implemented by small elite groups. Yvonne M. Tsikata: Aid and Reform in Ghana. 1999. The World Bank.  42 constituencies, though many District Assembly members reportedly do not perform this function on a regular basis. The time and cost of transport may be one good reason for this. 5.6 In both districts visited by the study team, the District Chief Executives had meeting with traditional authorities (chiefs) just before or immediately after meeting the study team, and both confirmed that such meetings were regular. This indicates that consultative processes – albeit not organized – are quite frequent. They involve – as in most societies – those with a voice, i.e. with some potential power or influence. This seems, in Ghana, to be a quite small group, linked to an urban-based, well-educated middle- and upper-class, and a rural traditional authority structure. 5.7 While the realities of high aid dependence (with 40% of the budget coming from ODA) and limited local capacities inevitably condition how much ownership it is possible to achieve, there are possibilities of enhancing national ownership through the further deepening within the political culture of democratic values, institutions and accountability. A way forward would particularly be further opening up of policy processes, with a rapid transition from ‘validation’ to ‘consultation’ to greater ‘participation.’ Such openings should be carefully integrated into the democratic system, to avoid that participation become an ad hoc event for those already privileged by a strong voice, often backed by donor-funds. 5.8 The discussion in chapter 3 on ownership in relation to the GPRS concluded that ownership is an elusive concept. It is often – also in Ghana – used to judge short-term processes (who are doing what – who are drafting, who convene meetings, take initiatives), or even as argument for abstaining from action (“because the government did not ask”). A supplementary line of analysis will be pursued which focuses on the outcome of these short-term processes, particularly in terms of commitment of important stakeholders at any point in time. Health 5.9 In terms of its genesis, strong MoH ownership of the SWAp was not given from the first day: initially, an alliance between a group of reformers among MOH managers forged an alliance with a limited number of donors. When a new minister was appointed in 1993, this alliance—which shared deep dissatisfaction with the poor record of stand- alone and dispersed donor-funded projects—got the political support it needed. The minister appointed 14 working groups consisting of public and private sector health professionals to give inputs to a Medium Term Health Strategy, which lead to the first 5- year SWAp in 1997.25 5.10 SWAp II was jointly developed by MoH and donors and is strongly influenced by the concerns of the latter (poverty reduction, gender, HIV-AIDS). Donors also influence priorities via their earmarked interventions. SWAp II espouses a partnership approach with a variety of domestic stakeholders, as well as with donors, but there are NGO 25. Edward Addai and Liz Gaere. Capacity Building and Systems Development for SWAps: The Experience of the Ghana Health Sector, DFID, 2001.  43 complaints of official reluctance to engage on some issues of health policy (e.g. role of private insurance). In practice, the most active dialogue is with donors. 5.11 The degree of MoH ownership of SWAp II should also be qualified by reference to reported limited capacities within the ministry (especially after the hiving-off of professionals into Health Service), and by the reported dominant role of a donor-provided consultant in drafting it. Questions have also been raised about the level of commitment of the NPP government to the arrangements, although it appears that this is a misunderstanding arising from a change of minister after only a few months and uncertainties by him and his advisers about how they can best fit into a predetermined set of arrangements and policies. These elements of an otherwise positive story underline that commitment rises and falls and needs to be continuously asserted. 5.12 Public Sector Reform: Strong country ownership and subsequent commitment in the area of public sector reform and economic governance reform have been more doubtful. In a review of the Medium Term Expenditure Framework (MTEF), a component of PUFMARP, in December 2000, the review team was “struck by the strength of the basic commitment to the MTEF principles on the part of all stakeholders.”26 Interviews conducted for the CDF evaluation two years later suggest that this basic commitment to the principles is not matched by a commitment to their effective implementation. 5.13 A concern expressed by some GOG officials to the CDF evaluation team is that some important stakeholders have still not bought into the reforms. Some ministers are not well attuned with the public financial sector reforms. This may be explained by the newness of the ministers, although many of them are now at least fifteen months in office.27 The potential users of the systems that are being delivered are not participating effectively in the process. The Controller and Accountant General’s Department (CAGD) were singled out more than once as an organization that should be more interested in the progress being made by PUFMARP. 5.14 Another critical area is the District Administrations. The Fiscal Decentralization project has been added to PUFMARP to ensure coherence in the systems and approaches to financial management capacity development. However, this project – with close links to the Ministry of Local Government and Regional Development, and therefore more responsive to District Administrations’ point of view – is not well coordinated with the other elements of PUFMARP, which are the result of a Ministry of Finance-perspective. 5.15 Five factors illuminate the rather poor level of participation of key stakeholders of the PUFMARP. First, there is broad consensus among informants that PUFMARP is 26. Government of Ghana and Department for International Development (2000) Joint Review of UK Support to PUFMARP Draft Report. p.7. 27. At the mini-CG meeting held in February 2002, the Minister of Finance acknowledged that the change in government had slowed down PUFMARP mainly because the new government had to go through a learning period.  44 “donor driven.” Second, consultants drive the implementation.28 Many of these are former staff of the Ministry of Finance, and have an excellent background for the work. But they are paid significantly higher salaries than the staff of the Ministries they are supposed to work with. This undermines morale, commitment, and ownership, and creates outright resistance against collaboration. The third factor is reform fatigue among those relatively few civil servants who have been involved in many different reforms, leading to faltering attendance at meetings and workshops – unless handsome attendance allowances are paid. This reportedly even drags out processes (since more meetings imply more allowances). The fourth factor is significant delays in implementation, especially linked to procurement of information technology and software.29 Finally, donor’s ever stronger focus on expenditure control implies that MoF officials focus on this and can devote even less time to reform work. 5.16 These factors can be interpreted as causes for weak national ownership, but they may as well be interpreted as effects. The broad perception that PUFMARP is donor driven merely signals that government ownership has been or is weak, and that donors have decided – most likely unwisely – to continue reform attempts anyway. 5.17 Looking for explanatory factors behind the immediately observable weaknesses in ownership, two key aspects appear important: the level of commitment from politicians, civil servants and donors to reform, and the capability to implement reform. 5.18 The commitment to reform of public financial management among key stakeholders in the public sector and among politicians should from the outset be expected to be low in the Ghanaian context. Such reforms aim very clearly at limiting the scope of discretionary decisions and discretionary expenditures – whether within or outside legal boundaries. Patronage is still an important element of the political culture, and though anti-corruption sentiments appear to be on the increase among voters in Ghana, the impetus for public financial management reform is probably stronger from donors than from domestic stakeholders. Though the public administration is in a poor state, there is no immediate crisis making the costs of remaining largely with status quo higher than the cost of reform. 5.19 The other aspect is the limited capacity – even if strong commitment is available – to implement reforms. PUFMARP comes under the Ministry of Finance, but its implementation requires cooperation of all the MDAs and its success depends on broader organizational reforms in the public sector. The public financial reform program is only one of several, simultaneously ongoing programs under the umbrella of the NIRP. 5.20 The vicious circle commences when a weak civil service cannot deliver reforms because incentives and qualifications to do so are insufficient. Consultants – often the best qualified civil servants – are entrusted the torch of reform, thereby alienating civil servants 28. The Fiscal Decentralization project, funded by CIDA, is managed by government staff and supported by a Canadian consultant. 29. The PUFMARP coordinator and the World Bank staff in Accra responsible for the Bank’s support to PUFMARP gave totally opposite accounts of the causes for procurement delay.  45 further, eroding broader based ownership, and undermining the very capacity building that was the whole objective of the exercise. Based on this analysis, some observers found that country ownership in Ghana is diminishing, and that increased donor dominance is appearing as a new system of relations where the capacity of aid agencies and donor funded consultants is increasing while ordinary public sector capacity is eroding. 5.21 As long as the sensitive and complex nettle of pay reform and downsizing of the public sector is not addressed —which is estimated by the involved consultants to be a 5- 10 year exercise —it is difficult to see widespread substantial and sustainable progress in economic governance reform, or indeed in other comprehensive public sector reform areas. The conditions for creating the sufficient commitment – or, in CDF language, country ownership – to push the reforms ahead is simply not likely to be there. 5.22 The new government has expressed its strong commitment to reform in the area of economic governance, and indicated that it wishes to review PUFMARP. Under the leadership of the Deputy Minister of Finance, a task force has been constituted to prepare an action plan that would bring the reforms back on track. The solidity of the commitment to reform has yet to be tested when the hard, potentially costly decisions have to be taken. 5.23 In conclusion, the key progress toward country ownership is the consolidation of democracy. In the narrower context of government-donor relations, the government is still sometimes too passive vis-à-vis the donors. For the government to take effective charge would require a clear, convincing program around which coordination can occur. On the other hand, donors may also be passive, using references to the respect for (absent) national ownership as an argument for not insisting on more or more focused attention to critical areas as e.g. civil service reform. Opportunities for gradual, joint construction of country ownership and joint commitment between all stakeholders – government, civil society, and development partners – which have proven successful in other areas, may thereby be lost.  47 6. Country-led partnership 6.1 The focus of the discussion in this chapter is on the relations between the recipients and providers of financial and/or technical development assistance in the form of concessional loans or grants. Country-led Strategic Coordination and Alignment of Donor Assistance 6.2 There have been positive moves in desirable (CDF) directions. CG meetings are now held in Accra every two years and co-chaired by government, complemented with quarterly “mini-CGs.” Both meetings are open to a wide range of stakeholders, including civil society, and are well attended by donors. Donors, however, had mixed feeling about the usefulness of the mini-CG meetings. According to the donor survey conducted as a part of this case study30 – most agencies indicated that the meetings do not result in in-depth discussions of serious development issues given the large number of participants, the relatively short time-span (three hours), and the various competing agendas people attempt to get across. Nevertheless, the meetings have proved to be opportunities for both government and donors to raise important topics such as pay reform, harmonization of pay structure by donors, corruption and alignment of conditionalities. 6.3 Among donors, there is clearly greater acceptance of the need for coordination, and of group processes leading to peer pressures for change and selectivity. Donor representatives have monthly meetings, and in general, there is more frequent contact between donors. All donor representatives able to make the assessment found that the frequency of donor coordination activities had increased over the last five years. The World Bank representative has played a very important role in this positive development, and has succeeded changing the traditional self-contained image of the Bank. Donors expressed that relations between the new government and donors are better than the relations in the last years of the previous government. 6.4 Donor coordination activities are largely focused on sector issues, with more than half of thirteen donors consulted actively involved in health, agriculture, education, and governance sectors issues. The health sector, which has a well functioning SWAp in place, draws by far the highest number of donors together, with ten of the thirteen agencies consulted actively present in the sector. Private Sector Development, decentralization, water, roads, and rural water project activities also attract a considerable number of donors. 6.5 Other coordination activities include PUFMARP, urban development, land management, National Institutional Renewal Program (NIRP), CDF thematic/sector groups, and support for the first round and run-off of the 2000 elections. 6.6 The efficiency and effectiveness of donor coordination activities have been mixed. The majority of donors surveyed (9 out of 13) find that donor coordination efficiency has 30. As part of this Country Case study, a survey of 13 major donors in Ghana was conducted (responses attached in Annex 1). Comments on donor experiences and perceptions are drawn primarily from these surveys and follow up interviews. Donors responding to the survey include Denmark, Canada, EU, FAO, Germany, France, Japan, Netherlands, UNDP, UK, UNICEF, US, and the World Bank.  48 improved over the last five years, while a few donors maintained that it has been mixed – good in the health sector, but stagnant, even deteriorating, in most other sectors. The majority felt that, in spite of the improvements, donor coordination efficiency was still unsatisfactory. The majority of donors (8 out of 13) also find that the results (effectiveness) of donor coordination have improved over the last five years, while some see it as a mixed picture. Donor coordination has resulted in better flows of information but have had limited success in shifting government policy and strategy. Nine out of 13 donors felt that the effectiveness of donor coordination was unsatisfactory. 6.7 The overall commitment of the government to policy dialogue with donors appears strong, and was forcefully expressed in the recent CG-meeting, though some government ministers aired concern about the potential downside of stronger coordination, e.g. that it would lead to donors ‘ganging-up.’ Several steps have recently been taken to revitalize hitherto dormant fora (agriculture, decentralization, economic governance). The capacity of the government to lead overall donor coordination is, however, severely limited, and meetings are often too formal, with most time spent on one-way provision of information rather than effective dialogue. Only a very thin layer of top-level executives – and more often ministers, vice-ministers or special advisers than regular civil servants – are providing the leadership that may make coordination effective. The capacity constraints are clearly demonstrated by the fact the government has had to request donor-funded consultant support to arrange the quarterly mini-CGs and take minutes from the meetings, and that the World Bank managed all practical preparatory work of the April 2002 CG meeting. 6.8 Central aid and debt management is formally the responsibility of the Ministry of Finance (MoF), but systems, structures, and leadership in this area are very weak. There are no well functioning links between donor desks, budget desks, and the sector ministries. Timeliness and usefulness of donor reporting to GoG is improving but still seriously incomplete, partly because of donor inability (and in some cases unwillingness) to provide information, partly because of a fragmented, unsystematic ad hoc-approach of the MoF when seeking information. Except for debt management information, which is up-to-date and regularly used, other information is incomplete31 and stored in (fairly obsolete) data systems. According to staff in charge of data registration, the information is not regularly used, and donors are neither getting nor requesting information about other agencies’ activities. The latest formal report summarizing and analyzing aid flows dates back to 1998, and it is very difficult to get any clear picture of distribution of inflows and disbursements of aid. 6.9 At sector and district level, both strategic and operative coordination capacity varies. At the strategic level, the sector or district policy platforms and plans have – with health as the notable exception – often not been strong, specific, or pervasive enough to guide a demand-led process of aid acquisition. In some cases (education, agriculture), donors have not agreed to the policies – but continued to offer support pointing in other directions. In the districts, the frameworks for determining whether an aid offer is relevant or not are virtually non-existent. 6.10 As evident from the discussion on the GPRS in section 3, it is at this stage difficult to affirm if donor’s assistance is aligned to the priorities of government as expressed in the 31. Staff in MOF estimated that about 80% of information about non-lending assistance is captured.  49 medium-term priorities. Given the limited information base, the weaknesses in operational coordination and the plethora of distorting elements in donor practices (discussed below), the answer to the question of alignment is that it is close to impossible to provide a meaningful answer across the board. Coordination of Concrete Assistance, Multi-Donor Approaches 6.11 Coordination on the ground of specific donor activities is much less advanced than the overall mechanisms for dialogue. The real challenges for government and donors are to get coordination out of the meeting rooms in Accra, and transformed into joint funding schemes, uniform approaches, harmonization of procedures and lowering of transaction costs. Health 6.12 The partnership between the Ministry of Health and the main donors to the health system sector has already been mentioned as the primary example of a functioning multi- donor approach. Some qualifications must, however, be added. While the common basket arrangement includes most major health donors, a good number of others have remained outside. Moreover, all 'basket' donors reportedly insist on also retaining some earmarked activities outside the basket arrangements. Box 9: Why the SWAp Worked…A Donor’s View. “…. the Ministry had a long term vision, it had the people, the right donors, world opinion was changing and the sector was in crisis. Those with strong commitment pulled the others along. They built consensus all around the country for two years to create a mid term strategy 1997- 2002. Donors were involved and agreed on the funding envelope. The Government was fully in control and convinced the Bank and others to go with a common management arrangement and basket funding. They also made provisions for those that couldn’t pool funds. The World Bank had to build up the country’s procurement capacity but it also had to change some of its own internal policies. …………” Interview with “Basket Donor” 6.13 There is thus a three-tier donor structure: (i) basket donors, (ii) donors with earmarked funds that go through the health budget and (iii) donors that directly finance earmarked activities. There must be a strong presumption that (by comparison with a fully inclusive SWAp) the effect of this structure increases the efficient allocation of scarce resources and lowers transactions costs, although it would take a major research effort to unravel the facts. 6.14 All donors participate in health sector dialogue; receive draft documents, joint appraisal reports, etc. The main penalty on donors that remain outside the basket is exclusion from the small group of (basket) donors who have an inside track on information and policy discussions. There are some peer pressures on non-basket donors to join but these are not heavy. The Common Management Arrangements associated with SWAp II sets out an ambitious target for 90% of all donor funds to be channeled through the basket by 2006, although this is predicated on further improvements in planning, budgeting and procurement.  50 6.15 Some assert that there have been major reductions in transactions costs because of the SWAp, e.g. through joint appraisal missions of basket donors, but others dispute this. It seems that costs are still high and the volume of missions and reporting requirements remains very demanding. Though the introduction of the annual “summit” has streamlined the multiple monitoring and review missions into a single event one time per year for all donors, there is not yet much synchronization of bilateral appraisal missions. There is some trend toward a growing membership of the basket arrangements but there are also reports of increasing individual demands by basket donors, e.g. for donor-specific reporting requirements. There are also reports of increasing competition among donors. The overall record illustrates the difficulties of inducing changes in donor behavior, despite the high reputation of the SWAp, and the slowness of any resulting improvements. Box 10: Insider’s View of the Health SWAp “…….we desperately needed reform. We had so many vertical programs addressing the same issues; some areas were over funded while vital areas were neglected. And overall Government funds were falling. There was one man who was the visionary in the Ministry and the whole reform process was something he had been working on for a long time…” The first five years were exciting times. We convinced 6 big donors--DFID, Danida, WB, Dutch, EU, Nordic Development Fund--to provide “basket funding” for a common program. We put a lot of energy into strengthening our financial management and procurement capacity so donors could rely on Ghanaian systems. We began with an annual work program with an agreed budget of support from donors. Each year we held a Summit in the districts to review progress and agree on the next year’s program. When we started the Summits we had 100 people coming—now we have 150 senior-level people, many coming in from their HQs. In the old days, staff would spend the whole year being visited by accountants and auditors from each of the 30 donors. Now we only have the one review mission taking up three weeks a year. The rest of the time, we can do our normal work. We have equipment standardized and all donors have to comply. Donors each get the same financial statement and one annual report. We do procurement ourselves. The policy is our own. Some donors took time to adjust administratively—one of our largest donors, for example, could not let go of certain procedures such as insisting on original receipts and this took time to change. There has been some backtracking. Once the euphoria subsided, donors realized they were losing territory, and when they saw us fumble in areas they had lost, there was some attempt to recolonize these spaces. We were doing well in coordinating the donors but they have their own rivalries. There is a danger that these territorial claims will increase as Ghana increasingly becomes a donor darling. While there is an efficiency effect of important donors being in one funding basket, there are still a large number of donors outside the basket and running parallel systems can be very cumbersome…” Interview with Senior Staff from Ministry of Health. 6.16 Economic governance: Instead of a regular SWAp approach with joint implementation arrangement, the assistance in the area of public sector financial management has been “sliced” in parts taken by different donors, each supporting according to their preferred modalities. Implementation of the components is characterized as donor-driven, with upward accountability to donors rather than to the government, and with considerable duplication.  51 6.17 A sub-group on economic governance has at some points been functioning, but has also deteriorated to become a too large group where an unproductive mixture of project specific details and overall strategies has been discussed, without a clear mandate of the group. After a longer period where PUFMARP was virtually inactive, the government has recently taken initiative to reactivate the group. According to some donor representatives, one of the key problems impeding more effective coordination is the conceptual weakness of the current approach, which does not link up to a consistent broader vision. Limited specialist on- site competency among the involved donors is also mentioned as a constraining factor in the ongoing dialogue. Box 11: The Agriculture Sector Wide Approach Attempt Though all stakeholders have made considerable efforts, it has been hard to move forward to a joint Agricultural Services Sector Investment Program (AGSSIP). Preparatory work on a sector policy and a comprehensive program started in September 95. Two years later, the Accelerated Agricultural Growth and Development Strategy (AAGDS) document was launched, as well as a concept paper on the AgSSIP, which was envisaged by the government to be the main instrument for the implementation of the AAGDS. The preparation of the AgSSIP was envisaged to take 9-10 months. A national Task Force, mainly with non-government members, was appointed to guide and coordinate the comprehensive preparation process. After two years, in autumn 1999, a joint appraisal was made by 8 teams in a process led by the World Bank. At this stage, some elements of the AAGDS (dealing with infrastructure and rural finance) were unilaterally taken out by the Bank, in a maneuver that created strong resentment among other donors. As a result, the AgSSIP ended up being a stand-alone World Bank-project (which only recently has become effective). It still has many virtues (country- wide in coverage, flexible and in principle demand-driven). Other donors meanwhile continue their projects. Parallel implementation structures proliferate, topping up schemes and poaching of staff are widespread, and transaction costs have not been reduced. Recently, fresh initiatives have been taken and new hopes for improved co-ordination are raised. Five groups have been formed to discuss very specific issues. Based on positive DFID-experience from Malawi with a similar approach, it is also considered to donor-fund a coordination unit to assist the Ministry. 6.18 Decentralization: The Ministry of Local Government and Regional Development (MLGRD) registers no less than 23 different programs and projects in support of decentralization32 not including the fund-type projects mentioned in Box 6 below. The government affirms that “many of these donor assisted programs have been implemented through various Ministries, Departments and Agencies, using structures and approaches that have not always been supportive of the decentralization policy to devolve responsibility, power competence and resources to the local level, since they tend to retain decision making power in the center. This difficulty has arisen from the absence of an internally coherent framework (a phased implementation program) or even a clear-cut strategy for coordinating 32. MLGRD: Decentralisation in Ghana – Implementation Status and Proposed Future Directions, March 2002.  52 locally initiated and donor-supported programs at various levels of the decentralized system.”33 6.19 Of the 23 projects listed, 16 support in one way or another capacity building of district administrations or communities. A very considerable part of this capacity building support consists of training and workshops. 6.20 To promote a joint approach, the Institute for Local Government Studies has since 2001 been tasked with the coordination of approaches to local capacity development. The Institute, whose operations have not yet been given a formal legal mandate, tries to: i) act as a clearance house and network node for activities to ensure a minimum level of information, ii) become a depository for training materials, iii) maintain a database of ongoing training activities, and iv) develop a local government training strategy. 6.21 Results have been disappointing. Though the need for coordination is acknowledged, the commitment to pursue joint approaches appears limited. There are duplication of efforts, diversity in approaches, apparent very limited mutual communication of best practices, and proliferation of workshops. Capacity building efforts focus excessively on training and workshops, rather than addressing structural constraints (incentives to performance). Several sources further indicated that quite a significant proportion of trained local government staff seeks greener pastures with NGOs operating in the districts. 33. MLGRD: Status of Decentralization Implementation in Eleventh Consultative Group Meeting, Volume I, Accra, April 2002.  53 Box 12: Delivery Mechanisms to Rural Communities A key challenge for both government and donors in Ghana is to reach the rural poor with relevant service-delivery. The government mechanisms pass either through deconcentrated sector ministry outlets (e.g. the Ministry of Agriculture delivering extension service, MOH delivering health service, etc.), or through the District Assemblies, especially through the District Assemblies Common Fund (DACF) described in section 2. In principle, the DACF should enable a demand-driven response to community needs. The weak capacity of district administrations and the many corruption charges associated with the DACF have made donors unwilling to use the DACF and the corresponding government procedures. Unfortunately, a coordinated alternative has not emerged – instead, each donor acts according to their own procedures, not only in terms of financial management procedures, but also in terms of criteria for funding. A community may thus have to contribute 20% of the worth of a project if the source is one program, and only 5% if it is another. The main “multi-purpose” delivery programs are the World Bank-funded Village Investment Project (VIP), the UNDP and ADB-funded Social Investment Fund, and the EU Microprojects Program, each working according to own procedures and with different degrees of coordination with district authorities. In addition, several bilateral donors fund quite comprehensive investment and service delivery projects in specific districts and communities, as do NGOs. In one district visited, the District Chief Executive estimated that he used 20% of his time on different donor-funded projects. An additional significant problem with the many disperse activities is that they only give arbitrary and disperse coverage of Ghana’s communities, largely based on non-transparent criteria. Though most donor-funded activities may be located in poorer communities, the activities are not based on equitable distribution criteria, and donors may in this manner unwillingly become part of and strengthen the patronage system. A government representative noted in this vein that the parallel delivery systems risk undermining democracy. Earlier attempts to coordinate approaches (including a CDF Thematic Workgroup on Poverty) have failed. Recently, the Minister of Planning has taken initiative to revitalize attempts to develop a common approach, initially between SIF and VIP covering 12 districts. A note explaining the “new, common approach to poverty reduction” frankly admits that past attempts have led to duplication of efforts and dissipation of scarce resources, having little or no impact. Implementation Modalities, Project Implementation Units, Incentives 6.22 The response to the weak capacity in the public sector has with very few exceptions been to establish parallel mechanisms enabling certain levels of short-term implementation of donor-funded activities. A plethora of different modalities are applied. They include traditional temporary implementation units only loosely linked to “host” institutions, topping- up in a number of more or less imaginative ways, and not least implementation driven by consultants, who often turn out to be former civil servants who in many cases end up doing what they could, in principle, have done as civil servants. Eight out of 13 donors consulted provide salary incentives to national staff involved in implementation of projects/programs, and the trend in doing so has remained the same over the last five years.  54 6.23 Similarly, 69% of donors, including the Bank, use project/program implementation units (PIUs). Most plan to continue using PIUs for the foreseeable future, particularly in complex projects, but the majority of these indicated their intention to gradually mainstream PIUs into the public service structure once the capacity has been built. 6.24 Some donors continue to rely heavily on technical assistance from their own country. In addition, most projects provide easily accessible funds for logistics, computers, stationary and other essentials that are in deplorable short supply in the “ordinary” civil service. As only few projects can get anywhere without some interaction with those who remain in the public service, so-called “attendance allowances” are offered to civil servants to incentive their participation in meetings and workshops.34 Civil servants interviewed expressed clear frustrations and reluctance to provide information to the far better remunerated consultants. 6.25 Because of these practices, a ‘parallel civil service’ is emerging in several areas, funded by donors, and taking over strategic functions of government. The new government, who has met the civil service with some reservations as described in section 2, is also largely drawing on special advisers outside the line organizations. 6.26 The strongly distorting and capacity-destructive effects of these different parallel mechanisms were acutely present in the accounts of civil servants. They were acknowledged, but much less forthcoming, by donors. In fact, most donors appear to consider these conditions as given or as necessary evils. The only initiative taken in this area that the study team heard of did not aim at abandoning the different distorting practices, but to harmonize employment conditions, salary levels, and allowances.35 A more thorough analysis of this particular aspect of partnership is offered at the end of this section. Box 13: The Challenge of Change One of the most serious translations of the CDF principles is a well-managed SWAp. But SWAps require a common set of procurement and financial management arrangements that all partners can accept. This is where the Bank’s internal policies and CDF goals clashed. The Bank convinced Denmark and the EU to join a common pool funding arrangement for the health SWAp. But then it found that its own regulations and bylaws prevented it from funding anything without first specifying exactly what the funds would be used for. There was significant internal opposition to basket funding within the Bank, because this transfers too much authority and responsibility to partners before the proper financial controls are in place. It took three years before the Bank could disburse into a common fund, and even then, it had to be on a “pilot” basis. Bank Country team staff interviewed reflected that despite the Bank’s commitment to CDF principles, it faces real operational challenges in finding out how to disburse in a “CDF way” and in internalizing CDF norms in a corporate culture where control is so deep. 34. The study team was advised to offer such an incentive (roughly 10-15% of a monthly salary in the higher end of the pay scale) to ensure attendance to the final half-day workshop. 35. F.H. Ashiabor. Survey of Compensation Arrangements for Contract Staff on World Bank Financed Projects in Ghana, 2000.  55 Transaction Costs 6.27 The transactions cost of delivering and administering aid in Ghana remains fairly high, but some positive and tangible shifts can be observed in donors’ behavior that, if maintained, can reduce procedural and administrative burdens for both government and donors. Multitudes of donor visits and reporting requirements continue to stretch the capacity of most line ministries. Among the 13 donors consulted, 3 had undertaken more than 10 missions over the last 12 months, 5 between 3-10 missions, and 4 less than 3 missions. One agency had over 20 missions in one year. 6.28 A more encouraging trend however emerges in the number of joint missions donors are undertaking – eight donors report undertaking joint missions, some as high as five joint missions. The trend has been increasing over the last five years for the majority (7 out of 13) of donors. 6.29 While donors increasingly appear more willing to come together in the early stages of project/program formulation, the desire to maintain a bilateral role is evident in the few number of joint monitoring and evaluation missions donors undertake. The majority reports that the trend has remained flat over the last five years. 6.30 The majority of donors demand four or more reports per project assistance per year from government, but the type and timing of reports demanded are not standardized across donors and/or sectors. Donors are increasingly accepting common reporting formats, but this has largely been limited to the health sector. Seven of the 13 donors consulted have adopted joint reporting formats endorsed by government, but only three of these donors report an increase in the use of joint formats for their operation in Ghana over the last five years. Almost half of the donors have no joint reporting format with the government. 6.31 Only three donors have in some cases adopted the procurement rules of other donors, in two of the cases by adoption of the procurement rules of the World Bank. None of the other donors report the use of other development partner’s procurement rules. The majority of donors (7 out of 13) use the procurement rules of government for some of their operations, most frequently in health. Only one donor reports the use of the procurement rules of government for all its projects. Decentralization of Donor Agencies 6.32 Several donor agencies have strengthened their presence in Ghana, thus increasing their capacity to dialogue with national stakeholders and accompany implementation processes. The World Bank decentralization attempts are much appreciated by Ghanaians, as is the shift from a controlling and supervisory role to the role of a partner supporting implementation. The conceptual shift promoted by the Bank team is that the Bank does not have projects, but it is supporting projects of the Ghanaian Government and as such assumes co-responsibility for the timely and efficient implementation of those projects.36 36. Similar concepts have been explicitly adopted by some bilateral donors since the mid-nineties and thus pre- dates CDF.  56 6.33 There are three distinct advantages of increased, qualified presence of donors: firstly, that donors can learn about local conditions in detail, including political conditions and processes, thereby adapting headquarters defined approaches to ground realities. Secondly, presence and knowledge about political processes also enable donors to contribute more substantially and constructively to policy discussions with local stakeholders. Thirdly, local representatives equipped with sufficient authority can introduce a healthy dose of flexibility in the management of aid portfolios. There is progress on both accounts among donors in Ghana, including increasing abilities to provide “just-in-time” technical assistance through flexible trust fund and inter-agency collaboration mechanisms. But though some decentralization has occurred, most agencies—including the World Bank—continue to draw significantly on headquarter-based staff. Causes, Ways Forward and Challenges 6.34 It is evident from the preceding paragraphs that although some progress has occurred, the partnership principle is far from adhered to. In key areas, donors continue to apply modalities that evidently undermine sustainable capacity building. The puzzling question is why these practices continue, when it is both easy and obvious to see that the impact of so many donor-supported projects is limited – and in some important aspects sometimes negative. 6.35 On the donor side, the answer lies in the combination of disbursement imperatives, demand for quick results, and to some degree visibility. Add to this donor’s risk avoidance profile, dictated by real domestic policy concerns about possible “scandals.” The Health SWAp has gone to extraordinary lengths to ensure sound financial management of donor funds. 6.36 Unfortunately, the government shares a large part of the incentives to opt for short- term efficiency at the cost of long-term sustainability. A four-year election period is a strong trigger to focus on short-term solutions. So are aid dependency and the need to be able to produce adequate policy statements, reform plans, and justifications for continued and increased donor assistance. High-caliber consultants and special advisers deliver this kind of product much more efficiently than an underpaid and demotivated civil service. 6.37 This may explain why “the parallel civil service” thrives. It is simply doing government’s work. But it does not explain why the government does not take a firmer stand when donors offer assistance to projects with objectives and implementation modalities that are inconsistent with government priorities. Several informants from the civil service found that the government should take a firmer stance and reject project offers not in line with key priorities. 6.38 There are, however, very evident reasons why few government officials would reject a project offer, even if the objectives were fairly wide off the mark of government priorities. The funds available in the national budget for operations of the MDAs – gasoline, cars, paper, photocopiers, computers etc. – are far below what is required to keep staff meaningfully busy. And salaries are not motivating. In the districts visited by the team, both health and agricultural workers are staying more behind their desks than necessary because they have no means of getting to their clients. In such a situation a project – which typically includes allowances or even consultant position opportunities, computers, transportation and some  57 operational funds that can be quickly released – end up solving problems unrelated to the formal objectives of the project. 6.39 The solution to what has in effect become a vicious circle is multifaceted and long term. But it must include a civil service reform that restores sound incentives to performance and which systematically phases out the distorting practices. Donors have so far remained surprisingly passive in this area. If adherence to CDF principles shall increase, this appears a key area for action, however painful it will be. The government has taken the first step to re- vitalize reform attempts – donors should, in a concerted effort, back this up loudly and effectively, both politically and funding wise. 6.40 Another, complementary part of the solution would be for donors to move toward increased program or budget support. Donors—are under the influence of CDF-principles— moving in this direction, and the most important partners are for example discussing to move to a “SWAp” in balance of payment support, with a joint set of triggers based on ex-post assessment of performance rather than ex-ante conditionalities. The key constraints for more comprehensive moves in the direction of budget support are, however, precisely the weak financial management capacity and the limited implementation capacity of the public sector. Stalling of key reforms, such as the national procurement reform, is putting in doubt government’s commitment to push through fundamental reforms in public financial management that ultimately impacts donor behavior. 6.41 As also reflected in the preceding paragraphs, several other initiatives have recently been taken to address problems of poor coordination (agriculture, decentralization, economic governance). Key government ministers and officials, after an initial familiarization period, are showing will and commitment to address some of the weaknesses in the partnership relations to donors. Positive as this is, it is not the first time that fresh initiatives of this kind have been taken. The years before and around the launch of the CDF-principles saw a similar wave of sector working groups and coordination initiatives, and an optimism about it, which has proved unfounded. A key question for all partners to Ghana’s development is therefore what it takes to ensure that coordination will fare a better fate this time.  59 7. Results Orientation 7.1 Despite efforts, there is not much progress in relation to this CDF-principle. This is linked to the overall capacity constraints, and to a generally weak monitoring situation. But some efforts are being made. 7.2 As discussed in chapter 3, there are active plans to strengthen (previously near- zero) monitoring capabilities of NDPC, especially in connection with GPRS (but staffing and budget of NDPC remains quite inadequate to the tasks of monitoring plan and GPRS). Health 7.3 The Health SWAp arrangements have increased the attention paid to performance, although there has been particular stress on financial management and aspects of input performance, perhaps rather less on outputs and impact. This appears to have brought positive outcomes on various management and input indicators but rather more mixed results as regards outputs and health indicators. 7.4 Positive results include: • Improved financial management, reporting and accountability, including greater decentralization of financial management and spending to district level. • Reported improvements in buildings, a rising number of district-level outreach clinics; better drug availability and logistics; improved planning at central and sub-national levels. • Improvements in some health outcomes: infant mortality is reportedly down; immunization coverage is up. • Receipts from donors are claimed to be more predictable and timely, although many donor commitments remain outside the budgeting process. 7.5 Other indicators point to an absence of improvement or to actual deterioration: • A reported increase in maternal mortality. • Slight declines in per capita outpatient and in-patient facility use. • Mixed indicators on family planning/contraceptive use; a constant less than 1% of children using impregnated bed nets; unchanging overall hospital bed occupancy rates (fluctuating around 75%). • Slower than intended progress in reallocating resources to priority disadvantaged geographical areas. 7.6 Moreover, despite efforts to improve these, data on performance remain weak and patchy. There are surprisingly few indicators for which time series of even four or five years are available (reflected in the choice of items just listed above) and indicator tables contain more 'nil returns' than positive entries.  60 7.7 Furthermore, the reported improvements in financial management should be kept in context. A recent pilot tracking study of non-salary expenditures (Ye and Canagarajah, 2002) found low standards of record keeping and large discrepancies between MoH estimates of resource allocations and estimates of receipts by sub-district health clinics/centres/posts. Ye and Canagarajah's, admittedly only preliminary, estimates suggest that four-fifths (79%) of resources allocated in Accra do not find their way to the sub-district delivery points. There is a consequently greater reliance by the latter on own- generated funds, which account for more than half of all financial resources, against MoH expectations of only 13%. It is likely that this greater reliance on fee revenues has adverse implications for equity and for service take-up. 7.8 Local constraints on better MoH results include: the disruptive effect of a recent poorly-handled reorganization, creating a professional Health Service separate from the MoH; associated losses of personnel, aggravated by substantial losses abroad of qualified doctors and nurses (aided by the active recruitment efforts of at least one of the major donors) and consequential serious shortages of staff; insufficient political will to address the problem of regional inequalities of access to health services by putting in place an adequate incentive system; and weaknesses of monitoring and supervision arising from decentralization of functions. Economic Governance 7.9 Attempts to build the MTEF into budgetary processes are also a positive factor, although its effectiveness is constrained. The basic machinery required to foster results- orientation is still quite weak, largely because of the endemic problems of public financial management in Ghana, i.e. weak monitoring and evaluation and poor data generation and dissemination. These problems have to be addressed if the results orientation principle is to be effectively implemented. 7.10 Mechanisms have been put in place in the last year to track expenditures better and an improvement has been reported. There is better tracking of expenditures now than there was 2-3 years ago. An Economic Policy Coordinating Committee has been established comprising officials of the Ministry of Finance, Controller and Accountant General’s Department (CAGD), the Bank of Ghana and the revenue agencies. The Committee oversees the forecasting and monitoring of expenditures commitments and cash transactions month by month. It recommends a set of quarterly ceilings based on cash flow forecasts and monitors performance of the MDAs against the ceilings. The CAGD has shortcomings in technical skills to perform the function of expenditure reporting. It lacks the needed technical skills. 7.11 The MTEF budgeting process introduces a shift in emphasis from inputs to a focus on outputs and outcomes. But effective implementation of the MTEF has been constrained by various problems. The resource envelope is not credible or predictable, and the internally generated funds by MDAs and earmarked donor funds are not all  61 captured in the framework. Further, real options for prioritization are limited because the major part of funds are statutory or, in the case of salaries, non-negotiable.37 7.12 Implementation of the Budget and Public Expenditure Management System (BPEMS) is critical to the establishment of an effective and efficient expenditure monitoring system. Its implementation has been delayed for several years by problems in different areas, as discussed in Chapter 5. BPEMS is not expected to come on stream until 2004. 7.13 Even so, it is not clear how far the initiatives under PUFMARP will be effective in improving accountability toward the public and the electorate, and to what degree they only focus on technical and bureaucratic aspects of public financial management. Incentives fostering political accountability, which in the end can be considered the key to foster results-orientation, are still weak, and both Ghana’s past political history, and factors like low literacy, point at serious challenges to advance in this area. Agriculture 7.14 Government and donors – and not least donors internally – have had clearly divergent viewpoint on how best to organize and finance a demand-driven supply of sustainable and effective extension services to farmers. The contentious issue has been whether and how to privatize the service. According to informants, disagreements between donors in this area were one of the key reasons why the Agricultural SWAp did not materialize. But, as far as the study team was informed, the dialogue in this area was not supported by evidence from impact studies of current or previous extension efforts. The office of the Ministry of Agriculture in the districts visited did no regular monitoring of the quality and relevance of their service. 7.15 Donors have been strongly supportive to government initiatives for improved monitoring. They are also, rightly, insisting that monitoring systems must be kept simple, commensurate with the capacity limitations of government. Again, worthwhile as specific monitoring mechanism may be for learning purposes and accountability to donors, the fundamental key to increased results-orientation may well be the gradual strengthening of the political accountability of the government to citizens at all levels. Different CSOs can play an enormously important role in this aspect, not least by ensuring that such accountability mechanisms are developed also outside the circles of Accra’s well- educated elite. 37. Bartholomew Armah. The Medium Term Expenditure Framework: A Case Study of Ghana in Legislative Alert, Vol 01, Accra, 2001.  63 8. Conclusions, Key Issues and Challenges 8.1 Summary of findings: The direct impact in Ghana of the launching of the CDF- principles in 1999 was limited. There is a low awareness of it outside donor and some government circles; it has not become institutionalized; and its specific initiatives have had only mixed impact. But it has changed the orientation of the World Bank office, which has actively promoted the principles and has decentralized its operations to a certain degree. 8.2 The mixed record thus far with regard to actual implementation of CDF principles demonstrates that the principles demand comprehensive changes of behavior and institutional practices and it would therefore be unrealistic to expect major breakthroughs. On the other hand, the NPP Government has evinced a resolve to undertake difficult reforms, including of the public service, and genuine multi-party democracy has been firmly launched. But the way forward is likely to be incremental, with steps back as well as forward. 8.3 The GPRS story is, as far as it is yet told, mixed. The recent CG meeting underlined that the GPRS document—much more than the GPRS process—is now becoming the key element defining relations between government and donors. But it is not linked to a holistic long-term vision. Ownership of the medium-term priorities is strong, but also very limited to the top political leadership. Ghanaians author the GPRS, and the (lack of) involvement of international partners during the process has certainly positive elements – but this is only a small part of the unfinished story. And it is yet too early to judge how results-orientation will turn out in relation to the GPRS. 8.4 Overall, the health SWAp is a clear improvement on the status quo and in some respects serves as a model for aid management in other sectors. Its main contributions have been in the areas of improved channels for donor-MoH dialogue, strengthening of financial management, better donor coordination, and reduction in MoH time spent servicing donors. Considered as an application of CDF principles, the Health SWAp has demonstrated their value. There has been, though, backtracking by some donors - attempts to recolonize areas they had previously vacated, especially where donors have viewed MoH as performing poorly. 8.5 The benefits on the ground for the Ghanaian users of the health service have so far been less obvious but that is partly because of the influence of factors outside the direct control of the health sector - demographic trends, the inexorable spread of the HIV-AIDS pandemic, the government's overall fiscal constraints (US$7.50 per capita expenditure on health per year), the problems of the civil service - and the inevitable slowness of translating improved input practices into felt improvements in mortality and morbidity. 8.6 To a substantial extent, especially donors could view the SWAp’s remaining deficiencies as the result of the incomplete application of the CDF principles. In particular, future progress could especially benefit from a more rigorous application of the 'partnership' (i.e. donor behavior) and 'results orientation' aspects of the CDF model.  64 8.7 In the other areas reviewed (agriculture, decentralization, economic governance), there is less progress. There have been positive developments with regard to each of the four principles. Attempts to produce holistic sector visions, assert country leadership, and develop coordination mechanisms have been made in all areas. However, the negatives outweigh these positive elements. The visions—except in agriculture—have not passed the tests of durability and consistence. Broad and strong commitment from national and international stakeholders to the visions has not been the constant that could ensure that they were carried through. The partnership approaches displayed in specific donor practices (parallel units, distorting incentives) are generally far from standards of the CDF principles. 8.8 The frank recognition of problems by government, but to a certain degree also donors, gives some reason for comfort. The government has taken initiatives in many areas to address the problems. But, in many areas, such initiatives have been taken previously, without major impact. The immense challenges ahead need to be explicitly recognized followed by a careful reflection on why things have not worked before and why and how they will work better now. 8.9 Sequencing and relations between the CDF principles: The story of CDF in Ghana demonstrates the intimate relation between the four principles: A long-term holistic development framework is of limited value if consensus among stakeholders— including donor partners—is not broad enough to make it fairly robust over time. Consensus and ownership at “framework-level” is of little significance if operational behavior remains uncoordinated and if donors, seeking to produce visible results and to protect their particular investment, apply implementation practices that are destructive to sustainable capacity building. Finally, results orientation requires not only country monitoring capacity but, fundamentally, broad implementation and accountability capacity, including key areas linked to budgeting/planning, execution of budgets and financial and political accountability. 8.10 Application of the CDF is intimately linked to the capacity issue: There is no evidence that promotion of one of the CDF principles prior to the pursuit of another would enhance development assistance efficiency and effectiveness. Further progress in the application of the principles requires multiple efforts in a number of arenas, or, differently put, broad-based growth of domestic capacity. Capacity parameters—including leadership—define how CDF shall or can be applied. Though top political leadership is certainly exerted in Ghana, it is not linked to broader capacity in the public sector. Building this capacity will be a long-term effort, and it will demand that donors radically change their approaches to implementation, which will also be a long-term endeavor. Looking back, the willingness to do so – and to push and pull for the necessary civil service reform – has been quite low. It is not easy to see the changes in the incentive structures for either government or donors that would put the capacity issue on the top of the agenda. 8.11 The risk of aid dependency and Ghana again becoming a donor darling. Perversely, Ghana’s successful economic adjustment in the eighties led to massive donor flows and with it greater dependency. After some set back in the nineties, it may appear that Ghana’s star is again rising. And as well deserved as that may be from geopolitical,  65 democratic and economic perspectives, it certainly poses new risks of increasing or maintaining the already high level of aid dependency of the government. Rather than promoting the application of the CDF principles, indications are already visible that this may lead to increased competition and crowding in some areas favored by donors.38 It will be hard for the government to refuse the donor offers in these areas, not least because of the considerable fringe benefits that any project brings to a resource starved and underpaid public administration. 8.12 A way forward could include the negotiation between the government and the donors of a “Ghana’s Strategy for External Assistance.” It could include five important elements, which could be monitored by peers from other aid receiving countries and donors and NGOs not active in Ghana: A commitment of the government to a plan for gradually decreasing overall levels of aid. A clear set of preferences concerning aid modalities, balancing government preferences for budget support with donor’s domestic accountability requirements, and balancing the need for short-term results with long-term capacity development objectives. A framework for strategic and operative coordination of external assistance. A clear commitment of donors to a plan for phasing out distorting incentives. An internal peer pressure mechanism among donors to avoid inappropriate assistance offers, linked to a code of conduct building on the CDF principles, but spelling out much more specific norms for tolerable behavior. 8.13 The CDF-principles and reforms: There is a dilemma in Ghana: Reforms are needed which, for their successful implementation, require broader capacity in many institutions and at different levels. Simultaneously, the development of the capacity required for reform presupposes - reform. How do the CDF-principles and their application provide guidance to this chicken and egg situation? 8.14 It is illustrative to look at the factors that made a SWAp arrangement possible in health: • There was general and acute dissatisfaction with the pre-existing situation. This had led to various reform initiatives that, however, needed to be brought together. • Visionary MoH leadership identified talent from districts and built capacity up over time, and talented, committed and risk-willing individuals on the GoG and donor sides pulled the others along. There was an initially low staff turnover - key actors stayed in place for several years. • Benefits were achieved quite quickly, with reductions in transactions costs, greater coherence, etc. 38. A rise in competition for “good projects,” and tendencies for crowding in “attractive sectors” like e.g. support to private sector development, and HIV/AIDS, were mentioned by several donor informants.  66 • Donors adjusted their internal requirements (although it sometimes took a long time) and MoH built up its financial systems to satisfy donors. 8.15 These factors are not far from textbook prescriptions for what is needed for reforms or change to happen. Translated into CDF-principles, they also confirm that vision; leadership and commitment: dedicated, flexible, and resilient stakeholders and quick results are good ingredients in reform processes. 8.16 While the validity of the CDF-principles certainly appear confirmed in this manner, it is also clear that mechanical interpretations of the principles will not give the answers to the two fundamental questions which both government, national stakeholders and donors in Ghana must ask themselves: (1) how can the most important additional reforms that will increase capacity and decrease aid-dependency, get off the ground; and (2) how can their implementation be sustained when the initial impetus meets hard resistance from vested interests in the country and entrenched cultures and/or rigid assistance schemes among donors? 8.17 Change toward such flexible CDF-like behavior from both national and international stakeholders to Ghana’s development is in itself a reform, and is likely to be only gradual, and not consistent. The most important contributing factor at country level in Ghana will be the further deepening of democratic values, processes and institutions relative to the forces of clientelism, including the growth of civil society, with its associated momentum in favor of greater accountability. 8.18 For the government, the key challenge will be to bring the “second generation reforms” forward, which can gradually – and combined with the deepening of democracy – address the constraints to the long-term restoration of an efficient civil service. 8.19 For the donor community, the key challenge is to patiently align policy dialogue methodologies, aid negotiations, size of aid and implementation modalities to this complex national process. It points in the direction of further moves from project to program aid, to agency decentralization and to the redeployment of agency staff from headquarters to local offices so as to be able to participate more fully in dialogue with the government and the wider society.  67 Annex 1: List of People Met Name Organization Gov’t. of Ghana Officials Ministry of Finance **Ms Helena Allotey MOF, World Bank Desk, R.422 Mr. Quist Therson Dr. Tony Osei Special Adviser to Minister of Fin. **Mrs Agatha Nketia-Gaisie Pufmarp Coordinator Alex Tetteh Aid and Debt Mgmt Unit, MOFA *Hudu Sita MOF, Project Monitoring *David M.A. Quist MOF, World Bank Desk CG Secretariat Phillip Owusu CG Secretariat Bartholomew Armah Institute of Economic Affairs Ministry of Health Jennifer Aryee Brown Ministry of Health Sam Adjei Ministry of Health Dr. Akoh Ministry of Health NDPC Dr. Gyan-Baffour NDPC Lanto Harding MEPR **Angela Brown Farhat NDPC *Patrick Yireyelleh Min. of Education Kofi Asante Frimpong NDPC/UNDP National Pov Reduction Pro *Mr. Kofi Dadzie NDPC Dr. Paa Kwesi Nduom Minister, MEPRC Seth Asante MEPRC Min. of Local Gov’t. Rural Mr. Addai Kyereme Min. of Local Gov’t. Rural Devt. Mr. Andan Chief Dir., Min. of Local Govt. Rural Devt. Decentralization Ama Servwaa Dapaah Exec. Director Social Investment Fund Cofie- Agama MOF, Fiscal Decentralization Mr. Ono IHDP, Japan  68 Economic Governance **Mr. A.B. Klu MOF, MTEF **Mr Seth Anipa MOF, MTEF **Mr Sam Daisie MOF, MTEF Mr. S.B. Nyantakyi MOF, Director of Budget Min. of Food and Agric. Malam Seidu MOFA, AGSSIP Coordinator 1 *Emmanuel Aggrey-Fynn Director, MOFA Dr. Francis Ofori Chief Director, MOFA Civil Society **Greg Ramm Save the Children Fund *Charles Abugre ISODEC *Yao Graham Third World Network Alban Bagbin Minority Leader, Parliament Gyimah-Boadi, Prof Centre for Democracy & Development Emmanuel Akuetteh IDEG K. Adu-Amankwah TUC, Head Akoto Ampaw Jubilee 2000 Esther Ofei-Aboagye Institute of Local Government Dr. Aboagye-Mensah Christian Council of Ghana Kofi Adu GAPVOD The Districts A. D. Oddom DCE Twifo-Praso Hemang Lower t District Agric. Officer Twifo-Praso Hemang Lower Denkyira District Ken Brifu Deputy District Health Director Twifo. Mr. Wilson DCE, Mfantseman District Mr. Kpoh Dist. Agric. Officer, Mfantseman Dist. James S.K. Akpablie District Health Director, Mfanstseman Vicky Cooper Consultant, CoEn Donors *Elsebeth Tarp Danish Embassy * George Adjei-Laryea UNICEF Aliu Bello UNICEF * E. de la Piedra IMF Rep Evelyn Awittor World Bank Peter Harrold World Bank Smile Kwakume World Bank Nduong Mba FAO Brigitte Heuel-Rolf GTZ, Director Johann Hesse, Dr. MOFA GTZ  69 *Robert Foote CIDA Governance Programme Lindsay Jones EU, Economic Adviser Liz Gaere Dfid Heather Cruden Counsellor, CIDA Frank I Young Director, USAID Giancarlo Izzo Ambassador, Italy David Kuijper First Secretary, Netherlands Emb. Stefan Frowein Ambassador, EU Alfred Sallia Fawundu UNDP Res. Coord. Ramesh M. Shrestha Res. rep, UNICEF Alvaro Kirkpatrick Emb. of Spain Alicia Varela Emb. of Spain/Lagos Laurent Duriez French Aid Agency Notes * Contacts made during scoping mission ** Contacts made during April and May