82111 A Multi-Partner Evaluation of the Comprehensive Development Framework On the Long-Term Holistic Development Framework Principle of 1 the CDF: An Evaluation Ali Abdel Gadir Ali Arab Planning Institute Kuwait E-mail: aali@api.org.kw. Arne Disch Scanteam Oslo, Norway, E-mail: ardi@scanteam.no 1 We are grateful to P. Balogun, of DIFD, and J. Eriksson, of the World Bank, for encouragement and useful comments on the first draft of the paper. Remaining errors and inaccuracies are our responsibility. ii 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 The “ideas of economists and political philosophers, both when they are right and when they are wrong, are more powerful than is commonly understood. Indeed the world is ruled by nothing else. Practical men, who believe themselves to be quite exempt from any intellectual influences, are usually the slaves of some defunct economist,” J.M. Keynes, The General Theory of Employment, Interest and Money. iv Table of Contents 1. Introduction: ................................................................................................................1 2. An Analytical Framework: .........................................................................................4 3. Implementation of the “Long-Term Holistic Development Framework” Principle: 10 4. The Implementation of the CDF Principle: An Assessment:.................................18 Do Pilot Countries Have Long-Term Development Strategies?............................19 Do the Strategies Target Development Results? ...................................................19 Is the Content of the Development Strategy Holistic and Balanced and Well-Sequenced? ................................................................................................................................20 Is the Medium-Term Strategy Consistent with the Expenditure Framework?.......21 5. The CDF and the Behavior of Donors: ....................................................................23 CDF as a Development Framework ......................................................................24 CDF as Long-Term Holistic Framework...............................................................25 Long-Term Holistic Development Framework (LTHDF) and Impact on Donor Behavior ................................................................................................................................26 LTHDF and Impact on Partner Behavior..............................................................28 LTHDF and PRSPs ................................................................................................31 Summing Up...........................................................................................................32 6. Concluding Remarks and Exploratory Hypotheses: ..............................................34 Quality of Aid, Donor Behavior and LTHDF ........................................................35 Quality of Aid and Impact on Implementing LTHDF ............................................35 Impact of LTHDF on Processes and Outcomes.....................................................36 Selected Bibliography ............................................................................................37 Table 1: Development Indicators for the CDF Pilot Countries .........................................11 Table 2: Vietnam’s Development Goals............................................................................13 Table 3: Bolivia’s Development Goals..............................................................................14 Table 4: Ghana’s Poverty Reduction Monitoring Indicators 1999-2003...........................15 Table 5: Uganda’s Development Goals .............................................................................16 Appendix 1: Telephone Interview/Conversation Guide on the CDF's Long-Term Holistic Development Framework..............................................................................41 Appendix 2: Donors Views from Mozambique: Results from a Field Survey..................43 1 1. Introduction: 1.1 The Comprehensive Development Framework (CDF) is an initiative by the World Bank’s President James D. Wolfensohn (1999), to enhance the effectiveness of the partners of development of the developing countries in bringing about desired development outcomes. According to the CDF Secretariat (2000) the CDF is “an approach by which countries can achieve more effective poverty reduction. It emphasizes the interdependence of all elements of development—social, structural, human, governance, environmental, economic and financial.” The framework is articulated around four major principles: long-term, holistic development framework; country ownership of development programs and policies; country-led partnership among various stakeholders; and, results orientation. 1.2 A multi-stakeholder evaluation of the CDF is being undertaken “to ascertain the extent to which CDF principles have been practiced and whether their implementation has been oriented to pursuing overarching development objectives, such as the Millennium Development Goals (MDGs). The evaluation will deal with the relevance, efficacy, and efficiency of the overall development assistance system in selected countries, both on the ground and at the policy level, including linkages to the MDGs.” 1.3 The MDGs deal with poverty reduction under three broad categories of goals: goals for economic well-being; goals for social development; and goals for environmental sustainability and regeneration. The goal for well-being is formulated as reducing extreme poverty in the sense that the “proportion of people living in extreme poverty in developing countries should be reduced by at least one-half between 1990 and 2015.” The specification of this goal is in the tradition of the money metric approach to the measurement of poverty whereby extreme poverty is meant by the proportion of people living below an international poverty line of US$1 per person per day.2 1.4 Five broad goals for social development have been identified. These include: the attainment of “universal primary education in all countries by 2015”; elimination of “gender disparity in primary and secondary education by 2005”; “the death rates for infants and children under five years should be reduced in each developing country by two thirds between 1990 and 2015”; “the rate of maternal mortality should be reduced by three quarters between 1990 and 2015”; and, “access should be available through the primary healthcare system for all individuals of appropriate ages, no later than 2015.” The specification of these goals is in the tradition of the capability approach to the study of poverty.3 1.5 The environmental sustainability goal requires that “there should be a current national strategy for sustainable development, in the process of implementation, in every country by 2005, so as to ensure that the current trends in the loss of 2 For the concept of the “poverty line” see, for example, Ravallion (1998). 3 For the origins of the capability approach to the study of poverty see, for example, Sen (1999). 2 environmental resources are effectively reversed at both global and national levels by 2015.” 1.6 Under the multi-stakeholder evaluation of the CDF six country case studies, four thematic studies, one crosscutting econometric study, and a synthesis report are envisaged. One of the thematic studies is on the CDF principle of the “long-term holistic development framework.” The objectives of the evaluation have been formulated as (i) identification of the factors that have facilitated, or alternatively hindered, the implementation of the CDF principles; (ii) assessment of the extent to which CDF implementation has affected intermediate outcomes, and to the extent possible, long-term development outcomes; (iii) promotion of learning, and capacity building, in countries where the CDF principles are being implemented; and, (iv) creation of an on-going mechanism for periodic assessment, and feedback, of CDF processes. 1.7 Within the context of the above objectives, the thematic study on the “long- term holistic development framework” is envisaged to address the following broad issues: (a) undertake an analytical review of the emerging consensus on development and the role of the long-term holistic framework on the new development agenda; (b) explore the consequences of alternative ways of implementing the long-term holistic development framework on the relevant processes, intermediate outcomes and development goals; (c) assess the quality of development aid and the extent of the change in the behavior of donors as a result of the emerging consensus on CDF-like development strategies; (d) formulate relevant hypotheses on the potential impact of the quality of development aid on the capacity of aid-receiving countries implementing the “long-term holistic development framework”; (e) identify, to the extent possible, cross-country indicators on: the extent of implementation of the “long-term holistic development framework”; the quality of policies and institutions as a proxy for processes; and, assess the appropriateness of the identified indicators for hypothesis testing. In addition, undertake basic cross-country analysis with a focus on the “long-term holistic development framework”; (f) formulate, to the extent possible, relevant hypotheses on the potential impact of CDF implementation principles on processes and outcomes and examine the possible trade-offs due to the “long-term holistic development framework”; (g) drawing from country experience, refine the cross-country indicators already identified; and, (h) use the case of a particular country to assess the role of the “long-term holistic development framework” in the development discourse of the chosen country on how they affect development goals. 1.8 The remainder of this paper is organized in five sections. Section (2) offers an analytical framework suitable for the formulation of a holistic, long-term poverty reduction strategy. The framework is used as a benchmark against which the implementation of the CDF principle on the long-term holistic development framework (LTHDF) is evaluated. Section (3) provides a cursory and highly selective reading of the implementation of the CDF long-term holistic development framework in the six pilot countries. In this section it is assumed that the Poverty Reduction Strategy Papers provide the embodiment of the CDF principle irrespective of whether countries state this explicitly or not. Section (4) provides an evaluation of the 3 implementation of the CDF principle while section (5) provides an evaluation of the response of donors to the CDF principle on the long-term holistic development framework. This section is based on a survey instrument that has been designed to elicit these responses. Section 6 offers a few concluding remarks and proposes a number of hypotheses that can be tested in future evaluation of the CDF. 4 2. An Analytical Framework: 2.1 At the risk of simplification, the advent of the CDF can be taken as signifying a return to the application of the basic concepts of development economics, as distinct from the application of narrow neoclassical economic theory propositions, to developing countries.4 This, it is suggested, represents a major shift in emphasis in the conduct of development business especially among the multilateral development finance institutions. Thus, in evaluating the CDF from the perspective of its “long- term holistic development framework” principle it is perhaps instructive to ponder a recent evaluation of the state of development economics thinking that drew a distinction between a first generation of development economists (roughly dated 1950-1975) and a second generation (roughly dated 1975 to the present).5 2.2 The first generation of development economists, it is noted, formulated grand, and visionary, models of development strategy that aimed at effecting structural transformation with a central role assigned to the government in planning and programming development. The policy content of these models was informed by the observation that “a less-developed economy was characterized by pervasive market failures. To correct or avoid market failure, they advocated central coordination and allocation of resources. The newly expanding subject of welfare economics also provided considerable rationale for government action to correct market failure.”6 In addition to pervasive market failures, the role of the government was justified on the belief that the supply of entrepreneurs was limited in these countries, and that major structural changes, rather than marginal adjustments, were needed to effect development. Thus, the “government of a developmental state was to promote capital accumulation, utilize reserves of surplus labor, undertake policies of deliberate industrialization, relax the foreign exchange constraint through import substitution, and coordinate the allocation of resources through programming and planning.”7 4 For a recent survey of the state of development economics thought see, among others, Adelman (2001), Yusuf and Stiglitz (2001), and Hoff and Stiglitz (2001). Adelman identifies three fallacies in development theory and draws their policy implications. Yusuf and Stiglitz identify seven development issues as having been settled: sources of growth, importance of macroeconomic stability, trade liberalization, the importance of property rights, approaches to reduce poverty, protection of the environment, and the role of the government. Based on present world trends they also identify a number of new development issues that need to be resolved in the future. These issues are grouped in two broad categories: issues related to multilevel governance and regulation and issues related to the management of human, capital and natural resources. Hoff and Stiglitz present a critical overview of neoclassical economic theory from a development perspective and concentrate on the influence of institutions, the distribution of wealth, history, and the types of behavior that have spillover effects. All these influences lead to multiple equilibria, which poses serious problems for development policy design. 5 Meier (2001:13). 6 Meier (2001:14). 7 Meier (2001: 14-15). Classical examples of visionary models of development include Nurkse’s “vicious circle of poverty,” Rosenstein-Rodan’s “big push,” Leibenstein’s “critical minimum effort” and “low-level equilibrium trap,” Lewis’s “dual economy model,” Rostow’s “take-off stage,” and Chenery’s “two-gap model.” 5 2.3 In contrast to the visionary first generation of development economists, the second generation is seen as having been “almost moralistic, dedicated to a somber realism grounded on fundamental principles of neoclassical economics.”8 Neoclassical economics is seen by this generation as being good for the governments of the developing countries. “Governments were admonished not only to remove price distortions but also to ‘get all policies right.’ Not differences in initial conditions but differences in policies were now thought to explain the disparate performances of developing countries. A country was not poor because of the vicious circle of poverty but because of poor policies. Markets, prices, and incentives should be of central concern in policymaking.”9 The focus of development analysis moved from grand visionary models of the process of development to the investigation of particular features of underdevelopment using studies at the micro-level facilitated by the greater availability of micro-level data. According to the second generation of development economists the “correct policies were to move from inward-looking strategies toward liberalization of foreign trade regime and export promotion; to submit to stabilization programs; to privatize state-owned enterprises; and to follow the dictates of the market price system. Through its guidance toward the correct policies, neoclassical economics were believed to be the safeguard against policy- induced distortions and nonmarket failures.”10 2.4 Despite the dramatically opposed policy orientation of the two generations, a non-intended link between the two was unearthed by the second generation in the form of new market failures. “The recognition of the existence of imperfect and costly information, incomplete markets, and transaction costs and of the absence of future markets extended the range of market failures beyond the earlier attention to public goods and externalities that required only selective government intervention. Correction of the new market failures provided a basis for a potential role for more pervasive government intervention.”11 Important contributions in this area of analysis were made in the theory of rural organization using information, risk and contract frameworks. “The recognition of information constraints and transaction costs helped explain how rural institutions are a response to missing markets and also clarified the situations in which the potential benefits of government intervention are greatest.”12 2.5 Thus, the CDF could be looked at as an attempt to incorporate the lessons learned in development economics. The requirement of establishing a long-term holistic framework to guide thinking about the future and to help design policies and programs for effecting development is a major recognition of the complexity of the development process and its long-term nature. This is a completely different orientation from the various approaches that looked at the possibility of effecting development, through enhancing economic growth, over short periods of time by 8 Meier (2001:17). 9 Meier (2001:17). 10 Meier (2001:19). 11 Meier (2001:21). 12 Meier (2001:22). It is by now common knowledge that the major contributors to the role of information in economic theory have been awarded the 2002 Noble Prize for Economic Science. 6 targeting short run movements in aggregate economic indicators. In a sense, therefore, the CDF can be deemed consistent with the view that considers development as a “process of expanding the real freedoms that people enjoy. Focusing on human freedoms contrasts with narrower views of development, such as identifying development with the growth of gross national product, or with the rise in personal income, or with industrialization, or with technological advance or with modernization.”13 This view of development as a process of expanding the real freedoms that people enjoy is based on the capability approach to the study of poverty and deprivation alluded to above. The Human Development Index (HDI) of the UNDP is an attempt to consolidate some aspects of capability in a single index.14 2.6 On the basis of the above observations on the nature and time duration of the development process, and in view of the specification of the MDGs on money metric poverty as well as capability poverty, a reasonable “long-term holistic development framework” should be expected to be based on an explicit analysis of the “poverty reduction” objective. The objective, it will be recalled, is formulated as requiring the reduction by half by the year 2015 of the proportion of people living below an international threshold of $1 per person per day. In the money metric approach to the measurement of poverty the $1 per person per day is known as the poverty line while the proportion of people living below this poverty line is known as the headcount ratio.15 The headcount ratio is indeed the most widely used, and understood, money metric measure of poverty, despite its limitations. The headcount ratio measures the spread, or incidence, of poverty in a given society.16 2.7 In the context of this MDG on poverty a long-term holistic framework can be developed around the headcount ratio by noting that any poverty measure (in this case the headcount ratio) could be expressed as depending on mean consumption in society and on a measure of the underlying inequality in the distribution of 13 Sen (1999:3). See also Bruton (2001). 14 From an analytical point of view Ravallion (1998) argued that if capabilities are assumed to be a function of the standard of living as measured by per capita consumption expenditure, then the money metric approach could still be used as an informative poverty analysis framework. Moreover, a careful reading of the massive information provided by the participatory poverty assessment studies indicates that most of the expressions of “deprivations” were made with reference to the standard of living (see, e.g., Narayan [2000], Narayan et al [2000a], Narayan et al [2000b], and Narayan et al [1999]). 15 Sen (1976) pioneered the theory of poverty measurement by identifying a set of axioms that need to be satisfied by poverty measures. The literature that followed is indeed extensive as reviewed by Zheng (1997 and 2000). In Zheng (2000) seventeen axioms and sixteen poverty measures are identified. Of the sixteen poverty measures four are found to satisfy all seventeen axioms (famous among them is the Foster-Greer-Thorbecke [1984], measure); and, two are found to satisfy sixteen out of the seventeen axioms. At the other extreme, the headcount ratio is found to satisfy eight axioms while the poverty-gap ratio is found to satisfy eleven axioms. 16 Another useful measure is the poverty-gap measure, which is defined as the relative gap between the income of the poor and the poverty line. The poverty-gap index measures the depth of poverty. Both the head-count ratio and poverty-gap measures are special cases of the Foster-Greer-Thorbecke measure. 7 consumption expenditure, usually the Gini coefficient.17 In this general formulation of the headcount ratio it is expected that as per capita consumption increases (poverty line declines), other things remaining the same, poverty will decline. Similarly, as inequality in the distribution of consumption declines, other things remaining the same, poverty declines.18 2.8 For the purposes of developing the long-term holistic development framework, it should be noted that the change in poverty over time has two components. A growth component, given by the change in poverty as a result of changes in per capita consumption expenditure; and a distribution component, given by the change in poverty as a result of changes in the distribution of expenditure. The MDG on poverty is formulated in terms of a constant poverty line in real terms. The percentage change in poverty over time is given by a growth component capturing the effect of the increase in per capita income on poverty, and a distribution component capturing the effect of changes in the degree of inequality on poverty.19 The various magnitudes involved in these components can be calculated or estimated for any country where data is available.20 2.9 If it is believed that the distribution of consumption depends on per capita consumption expenditure in the society, then a powerful, yet simple, relationship between poverty reduction and economic growth can be established as the guiding analytical framework for the design of the “holistic development framework.” Thus, under this assumption the headcount ratio depends only on per capita consumption expenditure and the poverty line and changes in poverty over time can thus be calculated as a product of the elasticity of poverty with respect to per capita consumption and the rate of change in mean consumption.21 17 In technical terms the head-count ratio, say H, can be formulated as a general function of the following form: H = H(µ/z, θ), where µ is mean consumption expenditure, z is the poverty line and θ is a measure of the inequality in the distribution of consumption expenditure usually taken as the Gini coefficient. 18 These theoretical restrictions are usually expressed in technical terms as requiring the partial derivatives of the function H to obey the following: ∂H/∂µ < 0; ∂H/∂z > and ∂H/∂ θ >0. 19 In technical terms the change in poverty over time is given by the following: (2) G(H) = (dH/dt)(1/H) = η G(µ) + ν G(θ) Where η is the elasticity of the head-count ratio with respect to per capita consumption expenditure, which is negative; G(µ) is the growth rate of per capita consumption expenditure; ν is the elasticity of the head-count ratio with respect to the Gini coefficient and G(θ) is the rate of change in the Gini coefficient. 20 More generally and allowing for the poverty line to be a function of the standard of living the change in poverty over time is given by: G(H) = (dH/dt)(1/H) = η (1 – ε) G(µ) + ν G(θ), where ε is the elasticity of the poverty line with respect to consumption expenditure. 21 In technical terms the change in poverty over time is given by: (3) G(H) = βG(µ) 8 2.10 The MDG on poverty requires the headcount ratio to be reduced by half by the year 2015. Thus for a country drafting its comprehensive development framework in the year 2000, it has a time horizon of 15 years to achieve the MDG on poverty and as such, would require to reduce the headcount ratio by an annual rate of 4.52 percent. A direct application of this procedure would show that the growth objective of the country is given by the ratio of the desired rate of decline in the headcount ratio (i.e. – 0.0452) and the growth elasticity of the headcount ratio.22 2.11 Appropriately applying the above formula, together with information on projected population growth rates, the spatial distribution of population, the structure of the economy, and other relevant information, the country can formulate its broad growth objectives in a manner consistent with the MDG on poverty.23 2.12 An alternative use of the above framework is the determination of the time horizon over which a given country can attain the MDG on poverty. Thus, for example, if a country has sufficient information about its growth potential to enable it to appropriately decide on a growth rate for the future, and given the growth elasticity of the headcount ratio, the country can easily estimate the time required to reduce poverty by half. In this case, the expected annual rate of poverty reduction will be given by the headcount elasticity of growth multiplied by the expected real per capita consumption growth rate, and the time duration for poverty reduction by half can be calculated in a direct fashion.24 2.13 The above framework can be used as an organizing framework to estimate the required resources to achieve the MDG on poverty or to decide on the length of run over which the goal can be achieved. The estimated resource requirements can then be used for dialogue with development partners to mobilize their support for the development program to be built around the poverty reduction goal. An illustration of the use of such an organizing framework follows. 2.14 For a sample of 19 Sub-Saharan African countries the absolute value of the elasticity of the headcount ratio ranges from a low of 0.43 to a high of 2.2 with an average of 1.0254 and a standard deviation of 0.4879. Thus, achieving the MDG of poverty reduction by half by year 2015 would require mean consumption expenditure The elasticity involved, β, is called the “growth elasticity of poverty” and it can be estimated or calculated. In equation (3) the elasticity in question is the “growth elasticity of the headcount ratio.” 22 In technical terms G*(µ), the desired growth rate per capita consumption expenditure, is given by: (4) G*(µ) = [-0.0452/β]; where it will be recalled that the growth elasticity of the headcount ratio is negative. 23 For a possible use of such a framework in addressing the CDF objective of enhancing the effectiveness of the aid regime see Collier and Dollar (2001). 24 In technical terms, the expected annual rate of poverty reduction, Ge(H), will be given by the head- count elasticity of growth multiplied by the expected real per capita consumption growth rate, Ge(H) = βGe(µ) and the time duration for poverty reduction by half, Te, will be given by: (5) Te = 0.6932/[Ln (1+ Ge(H)] 9 to grow by an annual rate of 4.41 percent (0.0542/1.0254). With an average population growth rate of 2.7 percent, the achievement of the IDG of poverty reduction would require a GDP growth rate of about 7 percent per annum.25 To further appreciate the use of the framework in exploring the requirements for achieving the MDG of poverty reduction note that the incremental capital-output ratio for Sub-Saharan Africa for the period 1995-1999 averaged 5.47. This implies that to achieve the MDG goal of poverty reduction African countries would need an annual investment rate of about 39% of GDP (7.11x 5.47).26 2.15 An illustration of the use of the framework in equation (6) is to ask how long it would take the average Sub-Saharan African country, growing at the average rate for the second half of the 1990s, to achieve the MDG on poverty reduction. During the second half of the 1990s the annual real growth rate of GDP for Sub-Saharan Africa averaged 3.28 percent (ranging from a high of 4.8 percent for 1996 to a low of 2.1 percent in 1998). With an average population growth rate of 2.7 percent this means a real GDP per capita growth rate of 0.58 percent per annum. This implies a head-count poverty reduction rate of about 0.6 percent per annum. From this it follows that it will take an average Sub-Saharan African country, meaning a country growing at currently observed average growth rate for the region, about 116 years to reduce poverty by half (0.6932/0.005965=116 years). This is a time horizon that is about eight folds the identified time horizon in the context of the MDG for poverty reduction. 2.16 The above organizing framework, it should be noted, can be used in a flexible fashion to investigate various scenarios for the articulation of the long-term holistic framework. The usual “low,” “medium,” and “high” scenarios can be defined in terms of expected future growth prospects of the country. Moreover, given that the headcount ratio for the country can be calculated as the weighted average of the headcount ratios for various sections of the population, the long-term holistic development framework can be formulated by taking into account the contribution of various population sections to overall poverty. 25 See the World Bank (2001) where real GDP growth rates for Sub-Saharan Africa are recorded as 2.2 percent for the period 1975-1984, 2.3 percent for 1985-1989, and 2.1 percent for 1990-1999. 26 Note that Collier and Dollar (2001) use a growth elasticity of headcount ratio equal to two in absolute value for all the countries in their sample. With such an elasticity, the required per capita consumption growth will be 2.71 percent and the required GDP growth for African countries will be 5.42 percent and the required investment rate will be 29.65% of GDP. 10 3. Implementation of the “Long-Term Holistic Development Framework” Principle: 3.1 As noted above the initiative by the president of the World Bank regarding the imperative of conducting the business of development lending in the context of a Comprehensive Development Framework (CDF) reflects a realization that effecting meaningful development is a much more complex undertaking than managing economies for short run macroeconomic stability purposes. Consistent with this approach to development, and for the purposes of deciding on the provision of concessional lending to countries, a Poverty Reduction Strategy Paper (PRSP) process was established in December 1999 to guide World Bank and IMF boards in their lending decisions. 3.2 A PRSP is to be prepared, in collaboration with external partners if the need arises, and owned by countries. The core elements of a PRSP include (i) documentation of the participatory process invoked by the country to solidify the ownership of the development program; (ii) detailed diagnosis of the state of poverty in the country including both money metric dimensions, broader capability deprivation dimensions, and dimensions gleaned from participatory poverty assessments; (iii) rigorous identification and setting of medium- and long-term goals for poverty reduction with relevant, and realistic, indicators of progress inclusive of annual and medium-term targets; and, (iv) a clear specification of appropriate and feasible priorities for public actions. Central to these actions are the stability of the macroeconomic framework; the appropriate choice of fiscal policies and the adequacy and credibility of the financing plan of the development program; the suitability of the structural and sectoral policies and policies for social inclusion and equity; and, the directions of improvements in governance and public sector management. 3.3 Due to the time intensity of preparing development programs, on the one hand, and the need for countries to access concessional lending on the other hand, countries are availed the opportunity of approaching the World Bank and the IMF on the basis of interim PRSPs (IPRSPs). According to the current (as of end of March 2002) PRSP Web site, 41 IPRSPs and 10 full PRSPs have been prepared. Of the 41 IPRSPs 23 were prepared by African countries and of the 10 full PRSPs 6 were for African countries. Of the 10 full PRSPs 3 were prepared for countries included in the multi-stakeholder evaluation of the CDF (Bolivia and Uganda as CDF pilot countries and Burkina Faso as non-CDF pilot country). Ghana and Vietnam, two countries in the evaluation, had IPRSPs while Romania does not have a PRSP. 3.4 As noted by the IMF and the World Bank (2002: 6) the PRSP process embodies CDF principles. As such, evaluating the implementation of the CDF principle of the “holistic long-term development framework” can be undertaken by looking at the content of the PRSP from this perspective. In the case of Romania, which is not required to have a PRSP, the evaluation is undertaken on the basis of the World Bank’s (2001) Country Assistance Strategy document, which alludes to the CDF process undertaken by the country. Relevant development indicators for the CDF pilot countries are presented in table (1). 11 Table 1: Development Indicators for the CDF Pilot Countries Detail Romania Bolivia Vietnam Ghana Uganda Burkina Faso Population 22 8 79 19 22 11 (2000: million) Per Capita 6380 2380 2030 1940 1230 1020 GNI (US$ PPP Per Capita 1.7 0.2 5.4 1.8 2.2 3.1 GDP Growth Rate (1999- 2000: %) Life 69 62 69 58 42 45 Expectancy at Birth (1999: years) Under-5 24 83 42 109 162 210 Mortality Rate (1999: per thousand) Adult 2 15 7 30 34 77 Illiteracy Rate (1999: % of population 15 and above) HDI (1999) 0.772 0.642 0.682 0.542 0.435 0.320 Headcount 2.8 15.2 - 38.8 - 61.2 Ratio (US$1 (1994) (1997) (1998) (1994) per person per day) Headcount 27.5 51.4 - 74.6 - 85.8 Ratio (US$2 (1994) (1997) (1998) (1994) per person per day) Gini 28.2 58.9 36.1 39.6 37.4 48.2 Coefficient (1994) (1997) (1998) (1998) (1996) (1994) Source: World Bank (2002: 232-235) and UNDP (2002). 3.5 While it is admitted that the PRSP process is in its initial stages, preliminary assessments of the PRSP documents indicate wide variability among countries as far as adhering to the core components. An important observation regarding IDGs is that some IPRSPs used the attainment of the goals by the year 2015 as a reference framework while many abstained from doing so. The joint World Bank and IMF staff assessment of IPRSPs noted a number of areas where substantial work was needed to get to the stage of full PRSP documents. The areas identified include, among others: 12 the need for current and comprehensive information on poverty. The World Bank and IMF boards also noted that for IPRSPs to be upgraded to full PRSPs there is a need to fully address issues relating to measures for the enhancement of the growth process and of integrating into the economic mainstream the rural population. Moreover, it is recommended that growth scenarios need to be realistic given the initial conditions of various countries. Special attention was drawn to the importance of broadening the treatment of gender issues to go beyond girls’ education and women’s access to health services to such concerns as the gender disaggregation of poverty analysis and income generation opportunities for women. 3.6 Romania: as noted above Romania is not required to have a PRSP, but it is a CDF pilot country. The inclusion as a CDF pilot country was based on the recognition of the difficulties Romania faced in the 1990s in reaching a consensus within the country on medium- and long-term development priorities (World Bank (2001:7). As such, the CDF process that started in Romania in 1999 was aimed at building consensus on a development strategy. 3.7 According to the World Bank (2001: annex G) “CDF consultations were held in Bucharest as well as various locations around the country in 1999 and 2000. There were more than 500 participants: leaders of industry, labor, academia, religion, the media, other parts of civil society, and local and national government officials, as well as members of the political opposition.” Two overarching objectives of development are agreed upon: the reduction of poverty and accession to the European Union. 3.8 According to reported results Romania does not seem to suffer from capability poverty as measured by aggregative indicators on health and education. Thus, for example, life expectancy at birth is reported as 69 years; infant mortality of 21 per 1000 live births; illiteracy rate of 2 percent of the adult population over 15 years of age; and, gross primary enrolment rate of 104 percent of school-aged population (104 percent for males and 103 percent for females). By contrast consumption-based poverty in Romania increased significantly over the second half of the 1990s on account of the failure of the growth process. 3.9 While the Gini coefficient remained constant at 0.3, the headcount ratio, calculated on the basis of a poverty line that is 0.6 of average household consumption expenditure per equivalent adult, increased from 23.5% in 1995 to 44% in 2000 recording an annual rate of increase of 11.7 percent.27 During the same period real GDP per capita is estimated to have declined by an average annual rate of 5.8 percent and real private consumption per capita declined by an average annual rate of 4.8 percent. These imply relatively moderate sensitivity of consumption-based poverty to economic growth as captured by the implied absolute elasticity of the headcount ratio 27 Absolute poverty, defined on a locally constructed basic needs poverty line, increased from 25.2% of the population in 1995 to 30.6% in 2000 recording an annual rate of increase of 3.96 percent. Of course the implied growth elasticity of the headcount ratio is less than unity in this case (see, CDF Evaluation Secretariat [2002a]). 13 of 2.02 with respect to per capita GDP and of 2.44 with respect to per capita private consumption. 3.10 Vietnam: according to Vietnam’s Development Report 2002 the country is now in possession of a well articulated long-term holistic development framework backed by a highly coherent, and competent, institutional set-up. The broad development goals are documented in a ten-year socioeconomic strategy (2000-2010) and in a series of sectoral plans and a draft five-year plan for the period 2001-2005. Ten development goals, 23 targets and 48 indicators have been identified. A summary of these goals is given in Table (2). Table 2: Vietnam’s Development Goals Development Goal Number of Number of Targets Indicators Eradication of Poverty 3 7 Better Education for All 2 10 Achieve Gender Equality and Empower Women 4 4 Eradicate Poverty and Preserve the Culture and 3 3 Diversity of Ethnic Minorities Reduce Child Mortality 3 3 Improve Maternal Health 2 3 Combat HIV/AIDS 1 1 Ensure Environmental Sustainability 2 4 Provide Essential Infrastructure Services to the 2 8 Specially Disadvantaged Poor Ensure Good Governance for Poverty Reduction 1 5 Source: Vietnam Development Report 2002. 3.11 An inter-ministerial committee under the leadership of the Ministry of Planning and Investment is preparing a Comprehensive Poverty Reduction and Growth Strategy with the objective of formulating required structural reforms, public spending profiles and sectoral policies as well as clarifying outcome targets and intermediate indicators. In this highly disciplined planning process a Government- donor-NGO Poverty Task Force was involved in poverty analysis and strategic planning since 1999. The poverty task force is composed of 16 ministries, 6 donors, 4 international NGOs and 4 local NGOs. 3.12 Bolivia: is considered as one of the poorest Latin American countries both in terms of capability poverty and consumption-based poverty. In terms of consumption- based poverty it is estimated that 63 percent of the population lived below the poverty line in the late 1990s. Over the period 1985-1998 Bolivia was busy implementing economic policies aimed at stabilizing its economy. A fundamental achievement of these policies was the reduction in the inflation rate from 23,500 percent in 1985 to 4.5 percent in 1998. 14 3.13 In 1997 Bolivia launched a participatory process with the objective of mobilizing a consensus around the nature of development challenges facing the country and the possible approaches to their resolution, known as the “National Dialogue—Bolivia Towards the 21st Century.” According to the World Bank (2001:1) the dialogue “contributed significantly to the development of Bolivia’s Poverty Reduction Strategy, which was presented to civil society for discussion. With a view of fine tuning the Strategy, a workshop entitled ‘The Government Listens’ was organized and the changes suggested by civil society were incorporated.” 3.14 The Poverty Reduction Strategy comprises 4 broad strategic goals, 15 sub- goals and 30 targets. Table (3) summarizes these planning parameters for Bolivia. Table 3: Bolivia’s Development Goals Strategic Goal Number of Sub-Goals Number of Targets Expansion of Employment and Income Opportunities 6 12 Developing the Productive Capacities of the Poor 3 11 Increasing Security and Protection of the Poor 4 5 Increasing Participation and Social Integration 2 2 Source: Compiled from World Bank (2001: 6-12) 3.15 In addition to the strategic goals, the PRS includes two crosscutting issues relating to gender equity and the sustainable utilization of natural resources and environmental preservation. Within the context of the strategic goals, and the cross cutting issues, the principal goals for poverty reduction have been identified as (a) reducing the headcount ratio by 22 percent from about 63% in the year 2000 to about 41% in 2015; (b) reducing extreme poverty from a headcount ratio of 37% in the year 2000 to 17% in 2015; (c) increasing life expectancy from 62 years in the year 2000 to 69 years in 2015; and, (d) increasing the proportion of the population that has had at least eight years of schooling from 51 percent in the year 2000 to 67 percent in 2015. 3.16 The PRS recognizes the current weak institutional capacity of the state and as such includes specific institutional reforms aiming at improving the system of decentralized administration; establishing transparent mechanisms for the allocation of resources; strengthening the capacity of institutions charged with the implementation of the PRS; and, fighting corruption. 3.17 Ghana: according to the Ghana’s Ministry of Finance (GMF: 2000:1) the overriding aim of the government’s “economic development program is the reduction of poverty and general improvement in the welfare of Ghanaians. Developed in 1995, the strategy for poverty reduction lays emphasis on economic growth, integrated rural development, the expansion of employment opportunities, and improved access by the rural and urban poor to public services such as education, health care, water and sanitation, and family planning services.” A long-term vision for the year 2020 envisages that “Ghana will have achieved a balanced economy and a middle income status and standard of living. Specifically, the vision of Ghana in the year 2020 is of a country in which: long, healthy and productive life for all individuals is the norm, 15 with access to an enlarged range of choices for employment, shelter and leisure; the benefits of development are equitably distributed and gross deprivation and hard-core poverty are eliminated; national income is growing by at least 8% per annum; population growth is reduced from 3% to 2% per annum; solutions to sociocultural and economic problems of the individual, the community or the nation are recognized and sought within the domain of science and technology” (GMF: 1, box 1). 3.18 According to GMF (2000: 10-11) the Development Strategy for Poverty Reduction has been informed by the CDF principles as well as by the MDGs in their earlier version as the Copenhagen Declaration. For the medium term 1999-2003 six core indicators and 10 supplementary indicators have been identified to monitor the achievements of the development goals. Table (4) provides the information on the core monitoring indicators. Table 4: Ghana’s Poverty Reduction Monitoring Indicators 1999-2003 Core Indicators Actual 1999 Provisional Targets 2003 Extreme Poverty Headcount Ratio 29 25 (%) Extreme Poverty Gap (%)28 31 27 GDP Growth Rate (%) 4.7 6 Female Literacy (%) 36 45 Rural Child Malnutrition (%) 28 24 Infant Mortality Rate (per 1000) 57 52 Source: GMF (2000: 13, box 3). 3.19 Since 1996 Ghana established an institutional setup to coordinate poverty reduction initiatives. At the helm of the institutional arrangements is an Inter- Ministerial Committee on Poverty Reduction (IMCPR), chaired by the Minister of Finance. IMCPR comprises all ministers responsible for the social sectors, and those responsible for agriculture and infrastructure in addition to the heads of the National Council for Women and Development and the National Development Planning Commission (NDC). An interagency and multi-sectoral Technical Committee on Poverty (TCOP), with the Director-General of NDC as the chair and the Poverty Reduction Unit (PRU) as a secretariat, provides the technical support for IMCPR. Similar structures exist at the district and sector levels (for details see GMF [2000:7]). In addition to the above, the development planning and coordination arrangements in Ghana feature mini Consultative Group meetings, quarterly meetings of the government with its development partners, and the National Economic Forum. 28 Note that this indicator may be referring to the “income gap ratio” as per the discussion in GMF (2000: 3) where it is noted that the income gap ratio is the “proportion by which the average consumption level of the poor households falls below the poverty line” and where this ratio was “around 30 percent below the upper poverty line.. and the corresponding shortfall for extreme poor is 35 percent.” The extreme poverty gap ratio proper will be 0.35x0.29 = 10.15% for 1999 and 0.35x0.25=8.75% for 2003. 16 3.20 Uganda: according to the Uganda’s Ministry of Finance, Planning and Economic Development (hereinafter UMFPED: 2001: 2) the “strategy to eradicate poverty has its roots in two important government initiatives. Vision 2025 described the long-term national aspirations, and the 1997 Poverty Eradication Action Plan (PEAP) served as a national planning framework to guide medium-term sector plans, district plans, and the budget process. The PEAP was revised in early 2000, and a summary of the revised PEAP served as Uganda’s Poverty Reduction Strategy Paper PRSP.” The revised PEAP is Uganda’s Comprehensive Development Framework. The PEAP identified four broad development goals, 39 sub-goals, 23 targets, and 99 monitoring indicators. Table (5) gives a summary of Uganda’s CDF goals. Table 5: Uganda’s Development Goals Broad Goals Number of Number of Number of Sub-goals Targets Monitoring Indicators Creating a Framework for Economic 10 7 28 Growth and Transformation Good Governance and Security 8 1 26 Actions that Directly Increase the Ability of 11 5 22 the Poor to Raise their Incomes Quality of Life 10 10 23 Source: UMFPED (2000: annex table 1). 3.21 Without getting involved in details it is worth noting that Uganda has evolved a fairly coherent institutional planning setup where the country’s “overarching planning document is the PEAP, signaling poverty eradication as the fundamental goal of the government (and providing) a framework for the development of detailed sector plans and investment programs.” The PEAP is seen as a flexible framework that needs to be revised every two years to take into account implementation results and the dynamics of poverty in the country. With the PEAP providing the overall framework the planning process features sector planning (detailing sector priorities subject to hard budget constraints) and district planning (for the implementation of sector priorities based on local needs and priorities). Sector and district planning and implementation are guided by a medium-term expenditure framework (MTEF) and by district MTEF. The MTEF is a 3-year rolling planning instrument setting out the medium-term expenditure priorities and hard budget constraints. Annual budgets, both at the national and district levels, are seen as implementation instruments of the MTEF. 3.22 Within the context of the planning setup a participatory process is incorporated where the donors, the non-governmental organizations and the private sector, contribute ideas and information regarding sector plans and budgets. In addition, bottom-up participation of districts in the planning and monitoring process is allowed to provide feedback on progress towards poverty eradication goals (for details see CDF Evaluation Secretariat [2202e]). 17 3.23 Burkina Faso: is a very poor Sub-Saharan African country. In terms of capability poverty as measured by the UNDP’s HDI it ranked as the poorest country in the world with an HDI of 0.32. In terms of consumption-based poverty it is estimated that about 45 percent of the population was living below a national poverty line. 3.24 According to Burkina Faso’s Ministry of Economy and Finance (BFMEF; 2000) the country’s PRSP is based on the development commitments made by the government in a 1995 Letter of Intent on Sustainable Human Development that articulated the development vision of the country. “This vision is based on the concept of human security, which entails guaranteeing that every Burkinabe national will enjoy economic security, meaning access to training and employment, health security, food security, environmental security, and individual and political security.” Key quantitative objectives deriving from this vision, and to which the government was committed, included increasing per capita GDP by at least 3 percent per annum over the period 2000-2002 and by 4-5 percent starting 2003; reducing the incidence of poverty from 45% of the population to 30% of the population by the year 2015; and increasing life expectancy by at least ten years by 2015. 3.25 Four strategic objectives of the PRSP are identified as the acceleration of growth based on equity; guaranteeing access by the poor to basic social services; expanding the opportunities for employment and income-generating activities for the poor; and, the promotion of good governance. For the medium term, 1999-2003, monitoring indicators have been identified for the basic education sector, the health sector and water sub-sector. Other indicators are planned once information based on survey work become available. 3.26 At the institutional level a PRSP monitoring system was established in March 2001, eight months after the adoption of the poverty reduction strategy document. The monitoring system comprises three levels: a ministerial oversight committee as a decisionmaking body that reports to the Prime Minister; an inter-ministerial technical committee as a technical arm to the oversight committee that is chaired by the Secretary-General of the Ministry of Economy and Finance; and, sectoral monitoring groups organized at the ministerial department levels. In the context of this PRSP monitoring system, outreach activities have been organized to involve wider sections of the society in the matters relating to the implementation of the strategy. 18 4. The Implementation of the CDF Principle: An Assessment 4.1 In the context of its report on progress made in the implementation of the CDF principles the CDF Secretariat (2001) designed a qualitative tracking methodology. The methodology relies on asking a number of questions about each CDF principle and ranking the progress made by countries on the basis of qualitative scores in answering the questions. Five qualitative scores are assigned as answers ranging from low progress indicated by “little or no action” to high progress indicated by the virtual completion of all required action to implement the principle indicated by the qualitative score “substantially” in place. 4.2 In what follows, use is made of the questions identified by the CDF Secretariat for the Long-Term Holistic Development Framework principle without using the qualitative scores. At this stage, however, it may be important to flag the fact that the “semantics” surrounding the use of the word “vision” in the context of the CDF may prove to be problematic in the long run. Given the complexity of the development process including its time duration, and given the vast amount of ignorance regarding the various ways and means by which the transformation of a given society, from “traditional” to “modern” ways of thinking and acting, can be effected, asking poor countries to articulate visions that extend several years into the future, which is itself uncertain, does not seem to be a very useful exercise.29 Large modern corporations, and modern institutions like the World Bank, can afford to articulate precise visions simply because their nature of work is very precise and focused. Thus, for example, the precise statement of the vision of the World Bank as a “world free of poverty” can have meaning for the nature of the work the World Bank does. A similar articulation for a poor country in terms of a “country free of poverty” does not have the same force as guiding governments in poor countries in their efforts to effect development. If it does, then it stands to reason to argue that the vision of each poor country should be taken as a “country free of poverty,” irrespective of whether this vision is stated or not. Asking questions about whether such a vision exists or not thus becomes redundant. 4.3 By contrast, asking a question about whether a long-term holistic development framework exists does make sense. Indeed, the major strength of the CDF is its emphasis on the existence of such a framework as a guide to development policy formulation and a reference point for the evaluation of the contribution of various parties in the achievement of desired development results. Frameworks, by their nature, are broad based constructs on the basis of which strategies for the future can be designed. As noted earlier there is a need to differentiate between a development strategy and a development plan or program. The notion of a development strategy consistent with the CDF is that of a flexible document that “needs to set forth how it is to be created, revised, and adopted, the process of participation, the means by 29 See, for example, Bruton (2001: 201) for the effects of ignorance on thinking about the nature of development. According to Bruton, development in the sense of enhancement of well-being means building from the institutions, values, and technology—preferences that define a society seeking to develop. Due to ignorance about how these characteristics can be changed with a view to enhancing well-being, these societies will be involved in a continuous process of searching and learning. 19 which ownership and consensus is to be obtained, and how the details will be fleshed out” (Stiglitz [1998:16]). 4.4 On the above understanding, an evaluation on the basis of the four questions posed by the CDF Secretariat regarding the Holistic Long-Term Development Framework is provided below. At this stage of the thematic study on the Holistic Long-Term Development Framework it should be noted that the evaluation is based on the selective reading of the implementation experience of the six pilot countries presented in section (3) above supplemented by the results of the country case studies from the six pilot countries. Do Pilot Countries Have Long-Term Development Strategies? 4.5 From the above rather cursory survey of the implementation experience of the CDF principles, as well as from the country case studies, it is perhaps clear that each of the six CDF pilot countries has developed a long-term holistic development framework. There is, of course, variability among countries in terms of the details of the articulated strategy. This is expected in view of the vast differences among them in terms of initial conditions inclusive of the stage of their development, as reflected by real per capita GDP, the spread and depth of consumption-based poverty, the severity of capability poverty and the stage of evolution of institutions. One significant observation is that work on the articulation of a poverty-reduction driven development framework predated the CDF and the PRSP processes. Thus, for example, work on Ghana’s framework was started in 1995; that on Uganda’s started in 1997; that on Burkina Faso’s in 1995; and, that on Bolivia’s in 1997. Vietnam’s case is not very different from that of the above pilot countries. For Vietnam the poverty reduction strategy is embedded in a ten-year socioeconomic development strategy covering the period 2000-2010. The development strategy was approved in April 2001 after the advent of the PRSP and the CDF processes, but the Vietnamese planning process must have been at work much earlier (see, for example, JBIC [nd]). 4.6 It is perhaps important to note that for a country to have a “long-term development strategy” at any point in time is a political variable. The length of the “long-term” and the content of the “development strategy” could change with the change in governments in a democratic governance regime. The case study of Ghana amply demonstrates this. In this respect it is noted that the New Patriotic Party that won the December 2000 elections came 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” (World Bank 2003). Do the Strategies Target Development Results? 4.7 In almost all poor countries in the evaluation exercise the identification of development goals has been informed by the MDGs. The overarching development goal in all articulated visions was the reduction of poverty or its eradication (e.g. Uganda). While countries differ in terms of the number of goals and sub-goals to be achieved in the medium- and long-term monitoring targets, and indicators are 20 identified by almost all countries in accordance with MDG targets especially as they relate to the headcount ratio measure of consumption-based poverty, education enrollment ratios, and health indicators. 4.8 Due to obvious reasons relating to the availability of detailed information, however, the analysis of poverty varies a great deal between countries. Rather surprisingly, the poverty analysis in the PRSPs of the three African countries is much more detailed, despite the remaining gaps. Thus, for example, the Ghana IPRSP gives in summary form what must have been the results of an extensive analysis of money- metric poverty. The summary provides estimates of the headcount ratio for 1991/1992 and 1998/99 by geographic location and by main economic activity. Similarly, the Uganda PRSP includes an analysis of poverty trends over the period 1992-2000 where results on the incidence of poverty, as measured by the headcount ratio, are presented for sectors (rural-urban), geographic regions, and socioeconomic groups. Moreover, results on poor people’s perceptions as well as the major determinants of poverty are also presented. 4.9 In terms of the MDGs as a guide to targeting development results perhaps the most impressive, and systematic utilization, is that provided by the case of Vietnam. The formulation of the targets reflects the depth of the planning process that informed their formulation. Thus, for example, under the goal for the eradication of poverty and hunger the formulation clearly distinguishes between the targets expressed in terms of the international poverty line and the corresponding magnitude in terms of the national poverty line. Given that the poverty reduction strategy is embedded in the overall ten-year development strategy the targets are clearly stated in the planned time frame. Is the Content of the Development Strategy Holistic and Balanced and Well- Sequenced? 4.10 As per the requirements for the content of an acceptable PRSP the development strategy for each country is drafted in such a way as to include major sections on poverty analysis, a macroeconomic framework, sectoral plans, medium- term public expenditure framework, and institutional arrangements for the implementation of the strategy. As an example, the PRSP document for Burkina Faso contains six sections as follows: an introduction providing the justification for the formulation of the PRSP; section 2 on the “characteristics of poverty in Burkina Faso: background and causes”; section 3 on “development objectives and national priorities” providing information on the development vision and the medium and long-term development objectives; section 4 on the “poverty reduction strategy”; section 5 on “costing and financing the strategy” providing information on the “medium-term programs under way in priority sectors,” the “cost of complementary actions to be financed on HIPC resources,” the “poverty reduction strategy implementation tools,” and “risk assessment”; and, section 6 on “poverty monitoring and assessment strategy.” 4.11 In view of this it will be fair to conclude that the content of the development strategy of each pilot country was formulated in a holistic manner (see, for example, 21 the PRSP for Uganda and IPRSP for Ghana). Judging a balanced content of a strategy by the attention given to the major components, it is also fair to conclude that the strategies were formulated in a balanced fashion paying due attention to maintaining the stability of the macroeconomic framework to provide the requisite economic environment for decisionmaking by economic agents; exploring the major investments that need to be made to support the process of growth in the medium term by relaxing constraints on production in various sectors and sub-sectors in the economy; programming public sector expenditure in such a way as to emphasize poverty-reducing expenditure allocation; and, identifying sociopolitical arrangements that stifle private initiatives with a view of overcoming them over a reasonable time frame. 4.12 Despite the above rather positive assessment, and despite the centrality of the goal of poverty reduction in all visions, there does not exist evidence that the building blocks of the PRSPs are related in a coherent analytical framework of poverty reduction. On the contrary, in almost all documents following a section on poverty analysis, brief or elaborate, depending on availability of data and results, the documents usually lapse into detailed discussion of macroeconomic frameworks, sector plans and institutional reforms that may or may not have been related to an underlying poverty-reduction framework. Moreover, despite the fact that changes in poverty over time are known to be influenced by growth and distribution, most documents are silent about issues of distribution. Needless to note that such issues represent major risk factors to the sustainability of the future development process especially in ethnically diverse societies. As is clear from Table 1 compared to Romania, for example, all the remaining CDF pilot countries can be considered as medium- to high-income inequality countries.30 Obviously policies aiming at enhancing the role of the distribution component in reducing poverty will have implications for the medium-term public expenditure framework and for the sequencing of public actions.31 Is the Medium-Term Strategy Consistent with the Expenditure Framework? 4.13 There is evidence, based on the cursory reading of the available documents of the pilot countries, that most of the countries in question were able to ensue the consistency of the medium-term development strategy with the derived medium-term expenditure framework (MTEF). This should be expected in view of the fact that the medium-term public expenditure framework in a planning setting should be informed 30 Note that Bolivia’s Gini coefficient of 58.9% is based on the distribution of income while all other Gini coefficients are based on expenditure distribution. For purposes of comparisons these other Gini coefficients need to be adjusted upwards by 6.6 percentage points so that their values would become 42.7% for Vietnam; 46.3% for Ghana; 44% for Uganda and 54.8% for Burkina Faso (for the need to adjust Gini coefficients, see, for example, Dieninger and Squire [1996]). 31 See, for example, Ferreira et al (2000) on the imperative of establishing modern social safety nets to protect the poor during crises and on considering public spending on such safety nets as “investment” rather than current expenditure. Needless to note that such a “public action” will have major implications for the medium-term public expenditure framework and the cost of the poverty-reduction strategy. 22 by the medium-term goals of the strategy as well as the targets identified to monitor the achievement of the development results. 4.14 Perhaps the best example for this is the case of Uganda where experience with the medium-term public expenditure framework dates back to 1992/93. In 1998 Uganda created the Poverty Action Fund (PAF) within the context of its MTEF to “ensure that increased funds are allocated to poverty reducing activities specified in PEAP/PRSP. Initially, the PAF was created to ensue that resources saved from HIPC debt relief were spent on priority programs. Subsequently, the PAF has attracted additional donor funding and has in practice become a mechanism for ensuring the reallocation of incremental expenditures directly to poverty reducing public services” (UMFPED [2001:10]). 4.15 In assessing the consistency of the medium-term strategy with the MTEF it is noted that the overall performance has been satisfactory in terms of “containing aggregate spending within the resource envelope and delivering the desired composition of spending.” Thus macroeconomic budgetary management and expenditure control was achieved and hence macroeconomic stability was maintained. In addition, major shifts in public expenditure patterns towards the social sectors and infrastructure was attained. Despite these achievements it is also noted that further improvements to ensure consistency between the medium-term strategy and the MTEF are required (for details see UFPED [2001:11-12]). But of course this should be expected in the context of the “learning by doing” nature of the PRSP as an embodiment of the CDF principles. 4.16 At the design level the Ghana IPRSP paid special attention to ensure the required consistency between the goals of poverty reduction strategy for the medium term 2000-2002 and the MTEF. This is expressed in terms of the government commitment to allocate more resources to the social sector (defined to include basic education, primary health care, rural water, rural housing, feeder roads, rural electrification, poverty focused agriculture, social welfare and population management). Without getting into details it should be noted that as a whole “poverty targeted expenditure is expected to increase from 16.7 percent of total expenditure in 1999 to 21.2 percent in 2001” (GMF [2000:14]). 4.17 The consistency of the medium-term strategy with MTEF is judged as lacking in the case of Vietnam. According to the case study it is hoped that the Comprehensive Poverty Reduction and Growth Strategy will provide the link between national plans and public investments and as such ensure the desired consistency between the development strategy and public expenditure. 23 5. The CDF and the Behavior of Donors: 5.1 As noted in the introduction, a main concern of the evaluation exercise is to assess to what extent donors and other development agencies have changed their behavior sufficiently to enable the implementation of the CDF principles to have an impact at the country level. To explore the issues involved, a telephone interview guide and a questionnaire were designed as survey instruments to elicit primary information. The results reported in this section are based on the telephone interviews.32 5.2 Around 30 telephone interviews with representatives from a dozen agencies were carried out. Apart from the European Commission, this covered bilateral agencies in North America (CIDA and USAID) and Europe (Denmark, DFID/UK, France, Germany—both the Federal Ministry and GTZ—Italy, Netherlands, Norway, Sweden, and Spain). Among the 30 persons spoken with, seven were in field offices of two bilateral agencies (NORAD and USAID) in seven countries (Bangladesh, Bolivia, India, Mozambique, Sri Lanka, Tanzania and Vietnam). 5.3 There is thus a bias in the selection of agencies in that none of the Asian donors are represented (primarily Australia, Japan, New Zealand). It was furthermore difficult to find staff in a number of the Southern European countries who felt comfortable discussing the CDF. The conversations tended to be somewhat shorter and more general with higher-level officials. 5.4 The staff spoken with varied from senior managers in regional departments to technical staff in policy and regional (operational) departments. The intention was to get a mix of staff who were "best informed" such as CDF Focal Points, and others who might be expected to be aware of the CDF due to their position in the organization—if the CDF were a well-known concept. 5.5 The interviews were based on a conversation guide that was sent to each person beforehand, so that they could prepare themselves. This guide was modified twice during the process, partly because some of the answers became rather standard and no new information was gleaned from repeating those questions, but also because new issues or precisions arose as part of the "learning process" of carrying out the interviews. The final version of the conversation guide is attached as appendix. 5.6 The typical interview lasted 20-40 minutes, with a few lasting nearly one hour. Notes have been written out for each interview. These have only been made available to the team due to the confidentiality promise given to the persons interviewed. 32 Another largely closed form-survey questionnaire regarding the implementation and perceptions of the CDF as a long-term holistic framework was sent through the CDF Secretariat to donor respondents. In large part due to this questionnaire being sent during the summer holidays, only six forms were received three of which from the same donor. While some of the comments in the forms received were of interest, it was felt that the sample was too small and skewed such that it would not be appropriate to use to get further insight to those obtained through telephone interviews. 24 CDF as a Development Framework 5.7 A first comment from many of the respondents was about the CDF as a general development framework. To many in the bilateral agencies, the CDF represented nothing new in terms of analytical approach and the proposed matrix was seen as naïve in its attempt at mapping out the complexities of broad-based development. 5.8 On the other hand, the CDF was credited with being quickly disseminated around the globe. The four principles were furthermore a good summary of what many other documents had used a lot more space on establishing as central "best practice" principles for good development cooperation. The fact that it was the World Bank that was promoting the CDF also meant that it had credibility immediately. For a number of the bilateral donors, however, the most important aspect was that it had an impact on the Bank itself – its view of development (going beyond economics), its understanding of its own role and its relations to other actors, primarily to partner countries, but also to other donors and to civil society. So while the CDF may not have represented much new to many bilateral agencies, it clearly changed the World Bank's own cosmology and hence its operational procedures and relations. 5.9 For some bilateral donors, this was a critical step in that it opened up a new way of working together in the field. The Bank was seen as becoming more listening and thus more "partnering" in its approach. The controversies surrounding conditionalities and other forms of impositions were considerably reduced as new ways of establishing ownership to development programs and policies and country- led partnerships were sought. 5.10 But the CDF, with its ambition of establishing an overarching view of a country's development process and mapping it out in matrix form, never became an important operational concept except where it was able to latch onto national processes already in place (Bolivia, Uganda, etc.). There were a couple of CDF-like exercises that were carried out in a couple of countries: Zambia and Mozambique were mentioned. In Zambia, the two donors that mentioned it said that their impression was that the exercise was not taken very seriously, but as another piece of paper that had to be filled out. In Mozambique, the exercise was carried out by local consultants and where the impression was that the Bank and perhaps the government used the results to rethink their own programs and funding priorities, but it did not play any role for overall joint discussions on more optimal resource allocations or changes in priorities. 5.11 Most donors, however, were not able to recall any form of matrix mapping. One bilateral noted that some of their field offices had used the matrix try to get an overview of the current donor situation. Most, however, felt that such exercises were very similar to the kinds of Cooperation Development Reports (CDRs) that the UNDP offices in many countries had for years tried to put together, and which never became very useful because (i) data were incomplete, inaccurate, non-compatible, (ii) tended to be historical rather than forward-looking and hence of little use for coordination and planning purposes, (iii) took enormous resources without providing 25 clear benefits, and (iv) suffered in general from considerable lack of cooperation from both authorities but in particular donors about providing complete information on planned or foreseen activities. 5.12 The principles of the CDF were well known and applied, however, not because they were CDF so much as that they represented what was by then accepted "best practice" standards – the CDF "reflected the spirit of the times very well," was the kind of comment several respondents made. Again, the succinct formulation into four principles were seen as helpful, and thus having more consequences than some of the respondents were aware of since they at least were known, accepted and applied. 5.13 Following less than a year after the introduction of the CDF, the Poverty Reduction Strategy Papers (PRSPs) that were proposed in connection with the HIPC program were immediately seen as the operational distillation of the CDF principles (see section "LTHF and PRSPs"). CDF as Long-Term Holistic Framework 5.14 The first principle of the CDF is at the same time the most obvious and perhaps most difficult one: Long-Term Holistic Framework. While all development partners would claim that this is "of course" the starting point for their work, in practice, few actors can point to much systematic and long-term thinking behind many of the activities that are being funded. This is not because they do not understand the importance of it, but because it is extremely difficult and costly, and contains so much uncertainty that it is unclear how to make it operationally significant. 5.15 The Bank's insistence on this principle is becoming more appreciated as the focus on sector programs and poverty reduction strategies have – in many of the "first generation" programs and strategies – been fairly short-term and immediate in focus and often not well linked either across sectors nor over time. 5.16 Pairing the LTHDF with Results-orientation, a number of donors felt was a powerful combination: forcing actors to look ahead and try to grasp the larger picture while focusing on which results were most important in terms of sustainable poverty- reduction. Since no strategy in practice can be completely "holistic" in terms of all aspects being covered, the need to think through alternatives and consider other dimensions in addition to those with immediate use-value to the poor is important. Developing long-term growth strategies with reasonable targets coupled with monitorable indicators – often stated as country-specific MDG-compatible – was thus the logical outcome of this coupling. 5.17 For this reason, the LTHDF was considered of great importance. Several of those interviewed felt that this was perhaps the principle that was "sinned against" the most. It was also noted, though, that the appreciation of its importance was growing with the realization that without a good long-term vision, short-term timeframes can 26 lead to inefficient choices and thus a suboptimal developmental pattern and path, including for the poor. Long-Term Holistic Development Framework (LTHDF) and Impact on Donor Behavior 5.18 The impact of the CDF on bilateral donors varied. In some agencies, such as CIDA, management strongly supported the CDF as representing principles that should lead to changes in corporate behavior. Other agencies, particularly USAID and the northern European ones, felt the CDF principles were "old hat" by the time the Bank came around to adopting them, but that the CDF made them sharper, more operational. The political leadership in the various countries also generally supported the CDF, in part as a means of opening contacts to and strengthening the dialogue with the Bank. In terms of actual behavioral change, most donors felt the greatest impact had been on the Bank itself. It became less "economistic," more open to process issues, and not so domineering. The LTHDF, however, was a dimension that had always been a key characteristic of the Bank compared with many of the bilateral donors, and this comparative strength they largely maintained. 5.19 The LTHDF dimension is hence one that other donors are now looking at more carefully. More and more, donors are moving away from project support to sector programs and general budget support. This is in part a reflection of the acceptance that fragmented aid has caused enormous costs and problems to partner countries. This has not just been in the form of high direct transaction costs but also through segmenting vertical empires with discrete donor funding inside various public sector entities, whereas the general trend is towards more open and program oriented structures. As partner countries strengthen their own institutions, processes, and procedures and there is more insight and control into resources use, donors also have more confidence in the local instruments and information flows and are thus willing to test out more partner-based mechanisms for cooperation. The "holism" in donor behavior is thus clearly increasing. But this is more heavily subscribed to by the northern Atlantic donors whereas southern European donors acknowledge that they still work a lot through projects. While some see this as remaining the more efficient way of providing assistance in environments with weak implementing capacity, others see this as a "lag" issue but one where they see that discussions within their own organizations are becoming more open to more program-based cooperation (see the section on Mozambique below). 5.20 But whether they are becoming more long-term in their behavior and perspectives or not, donors see two somewhat contradictory trends. Sector programs and strategic frameworks such as Memoranda of Understanding (MoU) or country strategies are becoming more common and often have a slightly longer time horizon. Whereas three-year programs were often the norm, donors see that there are now more five-year programs. And more of these are built around government programs, so that an MoU may have as its point of departure a country's own long-term development plan and a medium-term expenditure framework, for example. Concerning even longer time frames, one example that was noted by several is DFID's indicative ten-year program with Rwanda. There are to be annual reviews that 27 will decide whether the program will continue or not, and at what level. But what it does do is provide the government with a clear signal that as long as mutually accepted targets are reached and the government makes documentable best efforts at achieving them, they should be able to count on DFID support. The Danes also now are willing to make long-term commitments in principle to a sector, and can talk of 10 years or even longer as being realistic for their support. So donors are looking further ahead at sector and at overall macro/national levels. 5.21 At the same time, donors seem also more willing to make drastic changes at short notice, especially when it comes to program aid. One thing is that most see their poorest partners – many of the African countries, for example – as representing high- volatility situations. Tying themselves too tightly to a particular country or regime is thus seen as potentially counterproductive. The experience with Zimbabwe over the last several years has been sobering for a number of donors, and the extreme disruption that internal conflicts have caused has of course also drawn attention to the difficulties of designing long-term plans. At the level of principles there is increasing consensus that untied forms of aid such as budget support needs to become more predictable and thus long-term in nature. But at the same time, the abrupt cut in previous long-term relations that have occurred in for example Malawi and Zimbabwe over the last year has also sent a clear signal that whereas donors may be swayed to become more long-term in principle, donor politicians also have no qualms about taking drastic decisions if they feel that key performance criteria are not met. Important donors such as Denmark, the Netherlands and Norway have for example changed the status of their relationships with a number of countries or important programs (such as Denmark's Environment, Peace and Stability Facility) quite substantially at very short notice and driven by internal factors rather than through any dialogue with the partner countries involved. 5.22 There is in fact an apparent contradiction between project and program budgets when it comes to long-term commitment. Whereas donors seldom will cut project budgets if things are more or less on track within the project, donors have made dramatic cuts or shifts in program aid, sometimes for reasons that have little to do with the actual program agreement. Program aid, which is often paid out as large lump-sum transfers, is easy to cut when there is a budget squeeze, but also easy to push unspent resources through at the end of the fiscal year when disbursement pressures mount. Neither reaction is helpful in terms of their long-term vision or impact, and both are detrimental to quality, impact and sustainability. The shift from project to program aid is by some of the informants thus seen as often being contradictory: it may be more holistic, but also more subject to abrupt changes and non-transparent decisions. It may thus be less long-term in its vision than forms of project aid. 5.23 This dual nature of program aid is seen as important and an area that requires more discussion to ensure that the long-term nature of it is strengthened. 28 LTHDF and Impact on Partner Behavior 5.24 Most informants could identify little difference in the CDF pilots pre- and post-CDF. That is, those countries that had embarked on longer-term and more holistic planning and were developing instruments to incorporate this, were strengthened in this work and the legitimacy of it by declaring themselves CDF pilots. But the adoption of such principles was seen—as with most important changes in the political and economic developments of these countries—to be driven by internal forces and agendas rather than external instruments like the CDF. 5.25 In non-CDF countries, progress in adopting CDF principles were said to sometimes be faster. The cases of Kenya and Mozambique were given as countries that went from very centralized policy decisionmaking systems to quite open and participatory PRSP processes. One view believed this was due to "latecomers' learning advantage": they could learn from the experience gathered elsewhere. But it also seems clear that the external pressures to design some form of consultation process and try to include longer-term issues had an impact in how the PRSP both was designed as a process and as a final product. But this again could hardly be attributed to the CDF as such, but more to the PRSP, and the fact that countries were beginning to look at each other and adapt and implement those approaches that were seen as helpful. The learning across borders also went beyond the public sector: NGOs and private sector representatives, through their own international networks, were also becoming more informed and engaged, and demanding that national processes open up and allow them a role. 5.26 A first major change was often seen in the field of aid coordination. With the CDF, host governments had a powerful tool to use in trying to get the donors to come together around a more common program. To the extent that the governments were able to infuse this national program with a more holistic and long-term dimension, the CDF lent legitimacy to it. 5.27 Ghana is mentioned as one of the countries that benefited from the introduction of the CDF, not only in terms of rhetoric, but also by applying CDF principles for structuring the dialogue with the donor community better. CDF working groups were set up for the different sectors, and these were to be brought together into the national planning process. In this way both a more holistic and long- term system was to be established. In practice the process was not a simple one, in part because the outcome of the processes were not always obvious, neither to national nor to international participants. While the intention was to produce gap analyses, for example using the matrix, and hence try to steer donor resources to high- priority underfunded areas, reality was that this kind of joint resource planning without some "fresh" money on the table to work with, was slow going and very frustrating. The great advantage when the PRSP came along was exactly that the parties then had some additional resources at the core of the discussions, something that the endless CDF meetings seldom attained. For this reason also, in countries that had not adopted CDF principles, most donors were unclear as to what impact the CDF might have had. A key reason for this was that in poor countries that were not CDF pilots but could be included in the HIPC process, the PRSPs were introduced 29 almost immediately, so that any incipient CDF work would have been overtaken by the more operational PRSP. 5.28 The most successful use of the matrix was probably Bolivia, where a systematic process led by the Vice President's office further developed work already begun before the advent of the CDF. The process has gone through some ups and downs, among other things, due to the costs of trying to use the matrix as a dynamic tool—that is, that requires constant updating. It has probably had its greatest value for the central ministries, both in terms of identifying total resource levels and potential gaps, and because it has brought a more structured dialogue between the parties. 5.29 One comment that a number of informants made, however, was that while the CDF may not have left many concrete marks, it clearly had initiated processes that were of great importance to the PRSP. Several dimensions were mentioned. The first was the concept of dialogue with civil society—opening up the policy debate and mobilizing participation around longer-term processes. The other was exactly for the government to take a more proactive role in the dialogue with donors and address them more as a group of partners rather than as individual funding agencies. A third area mentioned by some was that the CDF helped the internal coordination and discussion. Central institutions, in particular ministries of finance/planning, used the CDF approach to try to get unwieldy line ministries with their myriad projects to respect increased fiscal discipline and organization through demanding adherence to key planning instruments. The fact that donors in principle supported this and pledged that their own resources would increasingly move towards supporting such strategic and longer-term instruments strengthened the legitimacy of this demand. 5.30 But again the problem of attribution is not easy to disentangle. That is, the idea of consulting and bringing other stakeholders into the national policy debate had been promoted by the UN system for considerable time using mechanisms such as National Long-Term Perspective Studies, and later on numerous "2020 Vision" or similar long-term general policy processes. Once the World Bank came onto the scene with its CDF, however, the entire process was often energized by the fact that it was the Bank and not a bilateral or a UN agency that promoted it. Whereas the UN system could have great legitimacy in the policy arena, it often lacked credibility. It had too often not delivered the goods, or with insufficient quality, and almost never had sufficient own resources to either fund the planning much less the implementation of such vision processes. 5.31 The Bank, on the other hand, had lots of financial credibility—so it often did not have to worry about lacking political legitimacy. The fact that the CDF itself had come about as a personal document by the president of the World Bank was both its strength and weakness. The strength lay in the fact that the president put a lot of personal prestige into it and used his position to ensure that the Bank as an organization promoted it, and his network to convince major actors to come on- board—which they did. The weakness was of course that the CDF had no consultation process behind it. It was in fact itself a contradiction—a document that promoted country-based partnership and local ownership yet was an externally concocted, driven and to some extent imposed concept. 30 5.32 Particularly once the PRSPs were established, where the consulting process itself had to be documented and was demanded as an integral part of the document, governments that otherwise had not much considered popular participation had somehow to organize these events. And here the result is an interesting one of process imposing itself on product. What often began as largely symbolic, superficial and in some cases even manipulated events were often over time transformed into more genuine consultative processes that began having an impact on decisions and policy development. Actors who had been excluded from having their voices heard, found to their surprise that an organization that had more often than not been considered to be on the side of the elites was now the guarantor of more open and competitive exchange of ideas. While the nature of consultations is still often more formal and perhaps closer to dissemination rather than discussion, there is also little doubt that many see this as a process that can be opened up, which will provide an arena for more genuine debate. 5.33 To what extent does the more extensive consultation strengthen long-term and holistic planning? With more stakeholder groups entering the fray, there are always more competing agendas and conflicts coming to the fore. But at the same time it is making politics more "real"—having to confront the preferences and forces of stakeholder groups. Most donors, however, see the consultation process as being largely very positive, and one that has strengthened the ability to discuss the longer- term goals of public action. Coming together around a longer-term poverty reduction agenda seems to be working. Influential civic organizations such as labor unions, employers' and professional associations, religious organizations, and various forms of community based and non-governmental organizations are beginning to mobilize themselves as organizations and their membership and networks to participate, and in a constructive fashion. Many of the international conferences of the 1990's are known among different segments of civil society—the women's conferences among women, the Rio conference among the environmental community, etc. Having some of their conclusions and main recommendations essentially distilled into the Millennium Development Goals has thus lent an operational dimension to the national policy debate that is long-term yet concrete enough to allow for meaningful debate based on a common core or vision, so that discussions on priorities and how to achieve the goals can be purposeful. 5.34 The LTHDF principle also helps define the frameworks for other operational tools such as Public Expenditure Reviews and key change processes such as Public Sector Reforms and Public Finance Management Reforms. 5.35 While the CDF as such is thus not used or even mentioned anymore, the LTHDF principle is seen by donors as remaining important both as a concept and in practice. On the one hand national actors are constantly demanding that a more "holistic" or "longer-term view" is employed as justification for choices in government programs and proposals. On the other, governments are strengthening in particular their medium-term instruments, such as Medium-Term Expenditure Frameworks (MTEF) or Medium-Term Fiscal Framework (MTFF) so as to ensure comprehensive coverage, balanced choices, inter-sectoral links, and more and more based on more solid long-term economic models. 31 5.36 Several donors noted rather cynically that whether the CDF had ever been invented or not, these kinds of developments would undoubtedly have occurred in any event. The net value-added of the CDF was in their opinion zero, and in some cases negative, because it had been imposed and taken a lot of time and led to considerable debate that was not on partner countries' own premises and hence not of a constructive nature. 5.37 The other side of the coin, however, is that the CDF brought these principles together in a package and pushed them very hard and systematically. While there is no doubt that there has been a lot of cross-border learning and that these kinds of instruments would in one form or another evolve, it also seems clear that having the Bank push the CDF worldwide gave an impetus and dynamic to implementation that comparative initiatives—such as the OECD's "Shaping the 21st Century"—never attained. The LTHDF was strongly embedded in earlier instruments like National Long-Term Perspective Studies (NLTPSs), but few NLTPSs had much impact: they were not able to mobilize the same critical mass of principles, donors and funds the way the World Bank was able to around the CDF. LTHDF and PRSPs 5.38 Almost all informants believed that the PRSPs are what have given the CDF principles life and legitimacy in the long run. In countries where national planning and consultation processes already were in place, the CDF was adopted, incorporated and used to improve and deepen instruments that already had been decided upon politically. In countries where there was no indigenous CDF process in place, little progress was made before the PRSPs were introduced. One reason was that a number of countries that came onboard the CDF bandwagon clearly had expectations that this would lead to easier access to donor resources. When this did not occur, disillusionment set it coupled with uncertainty about what the purpose was. 5.39 The PRSPs are seen as "the legitimate son of CDF" as the CDF principles are to guide the development of the PRSPs. This is to ensure that the "lessons learned" regarding good development cooperation, which was the original justification for Wolfensohn's document to be produced, will be institutionalized. 5.40 In practice, the first generation PRSPs often exhibit a number of weaknesses—naturally enough. The most pronounced is often the lack of a genuine LTHDF. While there has been some form of consultation and governments clearly have taken an ownership to the documents, the longer-term vision is often somewhat blurred, and the holism often replaced by a focus on public provision of social services of particular concern to the poor. That is, a number of PRSPs have become efficient vehicles for mobilizing donor resources for key public sector programs rather than a broad-based policy and national resources mobilization program for long-term sustainable poverty reduction. 5.41 The need to go back and re-think some of the PRSPs in terms of a more explicit LTHDF is thus clear to many of the respondents. At the same time, this is not seen as weakening or undermining the PRSP concept—it is simply "teething 32 problems" for a new instrument which most donors consider to be of strategic importance. Whereas no donor used the CDF for any kind of operational or allocative purposes—being a CDF pilot country never meant anything for any donor covered in this phone interview regarding allocation of funds—most donors say they now look carefully at PRSPs and want to use an increasing share of funding to support these plans as they mature, become realistic and are seen to have a solid grounding in what key stakeholder groups themselves say are priorities. But donors see that as a genuine LTHF aspect of a PRSP is improved, this will lend much more weight and substance to a government's claim on external funding. 5.42 The challenge is thus largely seen to be to ensure that the PRSPs are realistic with regards to the longer-term sustainability—both fiscal but particularly economic—and that it also has a broad enough vision that it makes intelligent choices when it focuses its resources. Summing Up 5.43 The telephone interviews provided the following main conclusions: • The CDF as a development framework had little to offer in terms of concepts or understanding of development processes or partnerships. But the CDF was a well-focused summary of established "best practice" ideas. • The largest impact in terms of donor practices was on the World Bank itself, as it shifted focus away from purely economic dimensions to broader concerns; became more process oriented and less conditionality-based; and opened up to more genuine dialogue with both partner countries, other donors, and civil society actors in the countries themselves. In this way, the "holism" of the LTHDF was strengthened inside the Bank itself. • The CDF also quickly focused the debate in the field around these key issues, and brought a weight and urgency to these matters that pushed principles such as LTHDF much faster forward than they otherwise would have. • The CDF as such did not have much impact on development practice within most of the bilateral agencies as such. The matrix never was used or accepted as a useful tool, and the concepts were for the most part already part of their policies and practices. But CDF principles were followed in the field, and provided strong signal effects to all donors regarding the need for better country- led coordination. This promoted again the more holistic nature of the development process as all donors were now being brought more together. The donor community was often quite fragmented, and the CDF lent a legitimacy not only to the government, but also to those donors that had adopted partnership- concepts and practices. They were now able to argue more forcefully with colleagues who were not yet convinced that they also had to change their approaches and join up in the more holistic approaches to development cooperation. • Donors have been following LTHDF principles through strengthening their budget and sector program support, and also through signing more long-term 33 indicative framework agreements. At the same time, however, program aid remains more vulnerable to sudden shifts in donor decisionmaking, so that it is difficult to say if the vulnerability of the overall partnership really has decreased or not. • In partner countries, where CDF-like processes did not exist before, the CDF per se did not have much direct impact. But the CDF introduced new ways of working—consultation, partnerships with donors, improved internal coordination—that facilitated these approaches when the PRSPs were introduced. The CDF thus laid constructive foundations for more efficient and effective PRSPs. • Central ministries are the ones that have applied the LTHDF concept most assiduously, among other things because it assists in getting away from the project aid and the bilateralization of relations between line ministries and external support. But it has also helped national authorities to get line ministries to think in more long-term and broader perspectives, getting a more realistic and healthier internal debate on trade-offs and priorities. • Perhaps the greatest weakness of first-generation PRSPs is exactly their lack of a genuine LTHF. This is seen as a major challenge in making the PRSPs better developmental tools, and one where donors also need to become more engaged. • The bottom line, however, is that the CDF as such is no longer required. The four principles are important, and are now known and largely embedded in the key instruments of PRSPs, MTEFs/MTFFs and perhaps in some of the longer-term "vision" documents. But there is a strong feeling in the donor community that there are too many instruments, concepts and processes around. There is a need to focus on a few, make them better and interlinked—and in that perspective the CDF has outplayed its role and should, in the words of several, "be given a decent burial.” 34 6. Concluding Remarks and Exploratory Hypotheses: 6.1 There is evidence that poverty in the majority of the less developed countries is a predominantly rural phenomenon (see, for example, IFAD [2002] and World Bank [2001]). The reduction of such poverty requires effecting deep structural changes not only in the rural areas but also at the national level. Structural changes are by their very nature long-term and less developed countries have accepted to embark on the business of devising their development programs and strategies in a long-term holistic development framework. 6.2 There is evidence also that in the design of policies and programs with the aim of reducing poverty, the developing countries are cognizant of the fact that they need to respect the requirements for maintaining macroeconomic stability. Based on experience in various developing countries there is now a “broad consensus that low and stable rates of inflation were desirable goals for developing countries, even though moderate rates of inflation—up to 40 percent per year—were not necessarily injurious to growth. Research has also shown that macrostability requires flexible labor markets and a prudent and coordinated mix of fiscal, monetary, and exchange rate policies—a mix now enshrined among the fundamentals of macroeconomic policies.”33 In the context of the PRSP and CDF processes poor countries aspiring for concessional assistance from the donor community may have a bigger room to respect the fundamentals of macroeconomic stability in view of the envisaged reduced conditionality on such assistance.34 6.3 In addition to respecting macroeconomic stabilization fundamentals the logic of PRSP and CDF processes also requires respecting other developmental fundamentals. Among other things, and for the purposes of poverty reduction, broadly defined development fundamentals include “proactive supply-side production fundamentals; and sociopolitical fundamentals.”35 Given the nature of money-metric poverty in poor countries, central to the remaining development fundamentals are investment policies, and institutions. The two together will define an enabling environment in the context of which poverty-reducing development can take place. On the basis of these two development fundamentals a number of hypotheses could be advanced for appropriate monitoring of the future implementation of the CDF principle on the long-term holistic development framework. 33 Yusuf and Stiglitz (2001: 229). 34 Needless to note that relaxing the impact of the external debt constraint is not a policy option for poor developing countries but one for the donor community. Over the past few years the issue of debt cancellation has become a political one where the non-governmental organizations, especially those based in the advanced countries, have played the leading role. A critical evaluation of the “debt cancellation” proposal for poverty reduction has recently been undertaken by Easterly (2001). Easterly’s (2001: 136) ultimate conclusion is that for debt cancellation to contribute to poverty reduction it needs to be granted only once to a “responsible government.” 35 Mkandawire and Soludo (1999: 94). See also ECA (2001). 35 6.4 At such an aggregative level, one such hypothesis would be that the implementation of the LTHDF principle would lead to a noticeable, and sustained, increase in the investment rate in developing countries over a reasonable period of time. Another hypothesis would be that the implementation of the LTHDF principle will facilitate the emergence of a developmental state appropriately defined as one that has a long-term vision. 6.5 Be the above as it may, and using the results in sections (3)-(5), in the following sub-sections a number of observations and specific hypotheses on the CDF are formulated for appropriate perusal. Quality of Aid, Donor Behavior and LTHDF 6.6 Donors have moved towards more long-term and holistic ways of planning and disbursing aid. Sector program and budget support is increasing in importance, and more and longer-term framework agreements lay behind much of bilateral cooperation. This is rationalizing collaboration considerably, reducing various forms of direct and indirect transactions costs, such as administration, information and uncertainty. It is also allowing governments to develop their national planning, budget, execution, monitoring and evaluation instruments, often in partnership with the donor community and more and more also with national stakeholders involved. All of these trends allow governments to also devote more energy and be more confident in the development of Long-Term Holistic Framework instruments and processes: PRSPs, PERs, MTFFs/MTEFs, etc. Quality of Aid and Impact on Implementing LTHDF 6.7 From what has been said above, the improved quality of aid should make several contributions to implementing Long-Term Holistic Frameworks in partner countries, and these can be formulated in the form of testable hypotheses: (H.1) The average transaction cost per dollar of aid is lower for assistance that is given under conditions of LTHDF frameworks, such as sector program and budget support aid than for project aid. (H.2) The share of donor resources that is being transferred through LTHDF mechanisms is increasing as a share of total aid, and should thus further contribute to lowering total administrative costs of development cooperation. (H.3) The increased adoption and implementation of LTHDF is both a result of and contributes to improving trust between the parties. This increased level of trust provides a more solid basis for further developing the quality and coverage of LTHDF-compatible instruments, thus further strengthening partner governments' ability to assume ownership and leadership of their development process and results. 36 Impact of LTHDF on Processes and Outcomes (H.4) A focus on Long-Term Holistic Development Frameworks will help PRSPs, the most important joint instrument for poverty reduction, become more realistic and ensure that they become better embedded in medium-term frames such as MTEFs and MTFFs. (H.5) Formalizing LTHDF in the form of various kinds of documents (PRSPs, MTFFS etc.) is key to strengthening institutional memory. This is particularly important regarding the donor agencies, since they tend to have high turnover among their staff, particularly in the poorer countries where such frameworks are needed the most. (H.6) Increased focus on LTHDF instruments can help reduce total risk of development cooperation by allowing the parties to openly discuss trade-offs between long-term political commitments (framework agreements) and financial commitments (disbursement profiles). LTHDF can thus become important instruments for reducing total system vulnerability due to the kinds of risk exposures donors feel when they have to commit both funds and policies around recipient government instruments. 37 Selected Bibliography Acemoglu, D., S. Johnson and J. Robinson. 2000. “The Colonial Origins of Comparative Development: An Empirical Investigation.” Working Paper no. 7771, NBER, Cambridge, MA. Adelman, I. 2001. “Fallacies in Development Theory and Their Implications for Policy.” in Meier and Stiglitz (2001). Bruton, H. 2001. On the Search for Well-Being. University of Michigan Press, Ann Arbor. BFMEF (Burkina Faso Ministry of Economy and Finance). 2000. Burkina Faso Poverty Reduction Strategy Paper. www.imf.org. CDF Evaluation Secretariat. 2002a. Implementation of CDF Principles in a Transition Economy: A Case Study of Romanian Experience. Washington, D.C., World Bank. ———. 2002b. Evaluation of CDF: Bolivia Case Study. Washington, D.C., World Bank. ———. 2002c. Vietnam Country Study: Executive Summary., Washington, D.C., World Bank. ———. 2002d. Ghana Case Study: Preliminary Executive Summary. Washington, D.C., World Bank. ———. 2002e. CDF Evaluation: Uganda Case Study: Summary. Washington, D.C., World Bank. CDF Secretariat. 2001. Comprehensive Development Framework: Implementation Experience in Low-and Middle-Income Countries: Progress Report. www.worldbank.org/cdf. ———. 2000. Overview and Background of the Comprehensive Development Framework. www.worldbank.org/cdf. Center for Economic Research. 1998. “Donor Coordination in Mozambique.” memorandum no. 45/98; Oslo. Collier, P., and D. Dollar. 2001. “Can the World Cut Poverty in Half? How Policy Reform and Effective Aid Can Meet International Development Goals.” World Development, vol. 29, no. 11. Easterly, W. 2001. The Elusive Quest for Growth: Economists’ Adventures and Misadventures in the Tropics. MIT Press, Massachusetts and London. 38 Ferreira, F., G. Prennushi, and M. Ravallion. 1999. “Protecting the Poor from Macroeconomic Shocks: An Agenda for Action in a Crisis and Beyond.” Washington, D.C., World Bank. GMF (Ghana Ministry of Finance). 2000. Republic of Ghana: Interim Poverty Reduction Strategy Paper: 2000-2002. Accra, Ghana. www.worldbank.org. Helleiner, G. 2000. “Towards Balance in Aid Relationships: Donor Performance Monitoring in Low-Income Developing Countries.” Munk Center for International Studies, University of Toronto. Helleiner, G., T. Killick, N. Lipumbab, B. Ndulu, and, K.E. Svendsen. 1995. Report of the Group of Independent Advisors on Development Cooperation Issues between Tanzania and its Aid Donors. Ministry of Foreign Affairs, Danida, Copenhagen. Hoff, K. and J. Stiglitz. 2001. “Modern Economic Theory and Development.” Meier and Stiglitz. Hoff, K., A. Braverman, and J. Stiglitz, eds. 1993. The Economics of Rural Organization: Theory, Practice and Policy. Oxford University Press, Oxford. IDA and IMF (International Development Association and International Monetary Fund). 2002. Review of the Poverty Reduction Strategy Paper (PRSP) Approach: Main Findings. www.worldbank.org. IFAD (International Fund for Agricultural Development). 2001. Rural Poverty Report 2001: The Challenge of Ending Rural Poverty. Oxford University Press, Oxford. IMF and IDA (International Monetary Fund and International Development Association). 2001. Poverty Reduction Strategy Papers- Progress in Implementation. PRSP website, www.worldbank.org or www.imf.org. JIBC (Japan Bank for International Cooperation) nd. 2002. “JBIC’s Comments to the World Bank/IMF PRSP Review: Reflections from our Field Experience.” www.worldbank.org. Maxwell, S. and T. Conway. 2000. “New Approaches to Planning.” OED working paper no. 14, World Bank, Washington, D.C. Meier, G. 2001. “The Old Generation of Development Economists and the New.” Meier and Stiglitz. Meier, G. and J. Stiglitz, eds. 2001. Frontiers of Development Economics: The Future in Perspective. Oxford University Press, Oxford. Narayan, D. 2000. “Poverty is Powerlessness and Voicelessness.” Finance and Development, vol. 37, no. 4. 39 Narayan, D., R. Chambers, M.K. Shah, and P. Petesch. 2000a. Voices of the Poor: Crying Out for Change. Oxford University Press, Oxford. ———. 1999. Global Synthesis: Consultations with the Poor. Washington, D.C., World Bank. Narayan, D., R. Patel, K. Schafft, A. Rademacher, and S. Koch-Schulte. 2000b. Voices of the Poor: Can Anyone Hear Us?. Oxford University Press, Oxford. Norad. 1998. “Mottakeransvar og Praktisering av Giverrollen.” [Recipient Responsibility and the Implementation of the Donor Role]; Oslo. OECD (Organisation for Economic Co-operation and Development). 1996. Shaping the 21st Century: The Contribution of Development Cooperation. Paris. ———. 1992. DAC Principles for Effective Aid: Development Assistance Manual. Paris. Pearson, L. et al. 1969. Partners in Development: Report of the Commission on International Development. Praeger, New York. Ravallion, M. 1998. “Poverty Lines in Theory and Practice”; LSMS Working Paper no. 133, Washington, D.C., World Bank. Rodrik, D. 2000. “Growth Versus Poverty Reduction: A Hollow Debate.” Finance and Development, vol. 37, no. 4. Sen, A. K. 1999. Development as Freedom. Anchor Books, New York. ———. 1981. Poverty and Famines: An Essay on Entitlement and Deprivation. Clarendon Press, Oxford. ———. 1976. “Poverty: An Ordinal Approach to Measurement.” Econometrica, vol. 44. Swedish Ministry of Foreign Affairs. 1998. Africa on the Move: Revitalizing Swedish Policy Towards Africa for the 21st Century. Stockholm. ———. 1997. Partnership with Africa: Proposals for a New Swedish Policy towards Sub- Saharan Africa. Stockholm. Stiglitz, J. 1998. “Towards a New Paradigm for Development: Strategies, Policies and Processes.” UNCTAD, Geneva. UMFPED (Uganda Ministry of Finance, Planning and Economic Development). 2000. Poverty Reduction Strategy Paper: Uganda’s Poverty Eradication Action Plan: Summary and Main Objectives. www.worldbank.org and www.imf.org. UK (United Kingdom). 1997. Eliminating World Poverty: A Challenge for the 21st Century. Whit Paper on International Development, London. 40 UN (United Nations). 2001. Road Map Towards the Implementation of the United Nations Millennium Declaration: Report of the Secretary General. A/56/326; www.un.org. UNDP (The United Nations Development Program). 2002. Human Development Report 2001. www.undp.org. World Bank. 2002. World Development Report 2002: Building Institutions for Markets. Oxford University Press, Oxford. World Bank. 2001. “Memorandum on a Country Assistance Strategy of the World Bank Group for Romania.” Report no. 22180-RO; www.worldbank.org. World Bank and IMF. 2002. Review of PRSP Experience: An Issues Paper for the January 2002 Conference. www.worldbank.org/poverty/strategies/review. Yusuf, S., and J. Stiglitz. 2001. “Development Issues: Settled and Open.” Meier and Stiglitz. 41 Appendix 1: Telephone Interview/Conversation Guide on the CDF's Long-Term Holistic Development Framework The World Bank in collaboration with a number of donor agencies is carrying out an evaluation of the Comprehensive Development Framework, CDF. One thematic study is to look at the CDF as a Long-term Holistic Framework from both donor and partner country perspectives. This conversation guide is meant to help record donor views. Normally such a conversation can be done in less than 30 minutes. Donor Perceptions of the CDF a. The CDF was introduced in 1999. Is the CDF well known in your organization? Are the main principles (long-term holistic framework; country-led partnership; country ownership; results-orientation) accepted/endorsed by your organization? b. What do you see as the strengths and weaknesses of the CDF? What is new about the CDF compared with for example the OECD's 1996 policy document "Shaping the 21st Century," or partner countries' "2020" visions and similar exercises that took place during the 1990s? c. One operational aspect of the CDF was the concept of a matrix to map out the most important dimensions of socioeconomic development. This is particularly relevant for the CDF as a holistic framework. What is your organization's impression of the matrix? Is this a tool that your organization either has used or has seen used? What are the best examples of the matrix that you are aware of? CDF Impact on Donor Behavior/Practices a. If your organization supports the CDF, has this led to changes to your organization's policies, procedures? In particular, how is the principle of Long- term holistic framework being pursued? Do you support partner countries putting more emphasis on this principle, and if so, in what ways? Are there any concrete instances you are aware of where the principle of Long-term holistic framework has been supported and applied? Donor Views of Partner Country Use/Impact of CDF a. Do you see CDF pilot countries as having changed their attitudes or behavior due to adopting the CDF? In particular do you see any more comprehensive or long- term thinking/planning behind their resource allocation decisions? Does the CDF seem to improve host country capacity, focus, macro-management of the economy? Has the CDF been used to strengthen more holistic planning? Linkages between economic and structural dimensions? Links between expenditure frameworks and medium-term strategies? Aid and donor coordination? b. Do you believe that the CDF may have contributed to improved resource allocations decisions, better national "vision" documents? Are such longer-term documents qualitatively better, for example with respect to monitoring results? Does the CDF help bring coherence between national priorities and the international goals as defined in the Millennium Development Goals (MDGs)? Does the CDF help national authorities "get into the driver's seat," support 42 national consultative processes, strengthen legitimacy of national planning, budgeting, monitoring instruments? Summing Up a. What is your impression of your own organization's overall assessment of the CDF as a long-term holistic development framework? What is your assessment of other donor and multilateral agencies' views of the CDF? What is your assessment of partner countries' views of the CDF? b. Bottom line: Given the number of initiatives and instruments that the development community has been presented with over the last years, what is the role of the CDF? Where do you think its net value-added is the greatest? What would you like to see happen to the CDF and its role as a long-term holistic development framework? 43 Appendix 2: Donors’ Views from Mozambique: Results from a Field Survey 1. The interviews that were done in Mozambique took place during the period April 12-23, and were of two kinds. First, a series of interviews of donor staff in a dozen bilateral and multilateral agencies were carried out, primarily looking at their views on budget support and the donor coordination taking place in this context. As part of the interview, issues related to the CDF were introduced and recorded. 2. The other set of interviews covered staff from NORAD and SIDA. The Norwegian and Swedish embassies in Mozambique had organized a two-day workshop for staff from six countries in the region on public finance management, budget support and results management. The relationship to the Bretton Woods institutions and their instruments, such as the CDF and PRSPs, were included. In all, about 20 staff from the two organizations, including the head offices in Oslo and Stockholm, were present. Donor Group in Mozambique 3. The donor group in Mozambique has for years been very active in its dialogue with the authorities. The relationship has for many years been seen as very good, with the government largely credited with pursuing good policies and being committed to joint objectives. Events over the last couple of years—banking scandals, high-level corruption, non-prosecution of two high-profile murders—have dented this image somewhat, but the overarching mood is still one of partnership and progress. 4. One of the results of this close relationship is that a group of 10 donors is now providing budget support through a common mechanism known as the Joint Program (JP). About USD$100 million a year is now being provided directly to the budget, where the only conditionalities are those identified in the PRGF. For the rest, the JP relies on government policy documents, including the local PRSP known as the PARPA, the Portuguese acronym. There are in addition sector programs in health, education, agriculture, and similar arrangements for rural water, and roads and coastal shipping. There are thus a large number of quite complex partnership arrangements in place. These have undoubtedly improved the long-term thinking within the sectors where such joint programs exist. They are also helping the ministry of planning and finance to get a better picture of what is happening, though a lot remains: it is estimated that only about 40% of total public resources are recorded properly in the budget and state accounts. The off-budget items include both donor resources and a large number of public fees. This is clearly limiting the ability of the authorities to get a total or holistic overview, and also hampers the long-term planning: the real resources picture is not clear. 5. One step taken by the donor community some years ago was to see if their own practices could be improved in terms of more consistent planning and programming. An analysis done identified exactly different donor practices as the key stumbling block, and that the donor community instead should build on government 44 policies and help strengthen and interlink these, rather than try to have more consensus amongst the donors per se (ECON 1998). This was hence an analysis that promoted key aspects of the CDF well before the CDF was announced. It was also largely in line with a similar study that had been done a couple of years earlier in Tanzania, but where the scope of the study was wider: it was to look at both donor and government practices, and come up with a fairly strong critique of both sets of actors (Helleiner et. al. [1995], known as the "Helleiner report"). 6. The need for a longer-term vision and a more comprehensive view of development needs were echoed by many. There is a lot of effort being put into strengthening the country's MTFF, and a consultancy underway is looking at how the PARPA can be better embedded in the MTFF. A Ugandan is carrying out this consultancy exactly to allow transfer of lessons learned. 7. The PARPA itself is being criticized for being too narrowly focused on public sector service delivery and of not having a genuine growth strategy that will ensure sustainability. It was also noted that in a sector like education, a disproportionate amount of attention had been given to primary education. When challenged on this, education officials acknowledged that a long-term view of a poverty-reduction education system had to include more secondary and tertiary level training and education, but that they had interpreted donor signals regarding PRSPs to mean that only services of immediate value to the poor should be included. This distortion away from LTHDF thinking was thus seen as an implicit imposition by the donors, and not something the authorities themselves wished for. 8. In a country as poor and aid dependent as Mozambique, the need for a clear LTHDF as the point of departure for dialoguing is acknowledged by both sides. The CDF itself never comes up as an issue or concept, but the need to ensure realism and good choices from the LTHDF standpoint is accepted as necessary, and a weak point in the current situation. 9. One interesting way in which this had come to the fore, was the reactions to the banking crisis in the country in 2001. Two of the biggest banks that had been privatized a few years back went bankrupt, largely due to poor banking practices, corruption and poor supervision. Some of the bilateral donors reacted stronger than the Bretton Woods institutions to this because they felt the long-term impact on the ability to address poverty was undermined by the public sector refinancing their shares in the banks through issuing government bonds. One of the key discussion points was exactly what this would do to the long-term viability of the PARPA, and what kinds of long-term visions the government could put in place to assure the donors that commitments made would not be undone. 45 Scandinavian Agencies 10. The two Scandinavian countries have been early adopters of partnership concepts. Sweden published its policy paper "Partnership with Africa. Proposals for a New Swedish Policy towards Sub-Saharan Africa" in 1997, while NORAD had promulgated its "recipient responsibility" policy some years earlier (codified in a public paper later—NORAD 1998). But while the partnership concept was strongly embedded, the long-term vision and holistic dimension was not so strong. In particular the role of a long-term and coherent growth strategy was less pronounced than the equity, gender and poverty-oriented foundations for their development cooperation. 11. Most staff echoed what was said during the telephone interviews: nobody talks about the CDF anymore, partly because it never was very strong as an instrument and thus never really took "hold" any place. But the LTHDF is an area that many accept is a weak point in the traditional approaches of these two agencies, and where changes have been underway for some time. It was not clear to what extent therefore the discussions about the CDF and the need for long-term holistic frameworks had had an impact on the agencies. While most of those spoken with denied that the CDF or LTHDF as such had been an issue, it was clear that "the trends of the time" had pushed these kinds of concerns much higher on the agenda. Whether the agencies themselves recognized this as CDF-principles or not was less important than the fact that most agreed that they are much more concerned about LTHDF issues than they were some years ago. 12. The fact that both agencies are now paying more attention to public finance management than they did a few years ago (this is particularly true for NORAD) is also contributing to their focus on LTHDF. Ensuring that the public sector is becoming more fiscally prudent and sustainable is making them look harder at macro- economic issues than they used to. This has led them to become more actively involved in processes such as Public Expenditure Reviews, Country Financial Accountability Assessments, and appraisal on monitoring missions for large-scale credits, such as Uganda's PRSC or Public Sector Reform credits. Thus, there is a felt need for understanding the big picture—of sharing in a more genuine LTHDF. And this was a change that was being felt across the region and being supported from the two head offices.