Reviewing Some Early Poverty Reduction Strategy Papers in Africa

Contributor Notes

Author’s E-Mail Address: crobb@imf.org and ascott@imf.org

This paper reviews some early interim and full PRSPs for countries with which the authors worked during 1999-2000 (Uganda, Burkina Faso, Tanzania, Mozambique, Mali and The Gambia). The purpose of the review is to compare and contrast how the PRSP process was established there. It finds that rapid progress was made in implementing the initiative in all the countries, increasing commitment to poverty reduction amongst government and donors and encouraging broader participation in the policy dialogue. However, there was considerable variation between the cases, reflecting different local contexts and capacities.

Abstract

This paper reviews some early interim and full PRSPs for countries with which the authors worked during 1999-2000 (Uganda, Burkina Faso, Tanzania, Mozambique, Mali and The Gambia). The purpose of the review is to compare and contrast how the PRSP process was established there. It finds that rapid progress was made in implementing the initiative in all the countries, increasing commitment to poverty reduction amongst government and donors and encouraging broader participation in the policy dialogue. However, there was considerable variation between the cases, reflecting different local contexts and capacities.

I. Summary

In September 1999, the IMF and World Bank endorsed a new framework for poverty reduction that involves the development of nationally owned, participatory poverty reduction strategies (PRSs) - embodied in poverty reduction strategy papers (PRSPs).2 The PRSPs will provide the basis for debt relief under the enhanced Heavily Indebted Poor Countries (HIPC) Initiative, for Bank and Fund concessional lending, and, eventually, for all donor assistance.

This paper reviews some early PRSPs in order to highlight relevant analytical and methodological issues, emerging good practice, and operational constraints. It focuses on five aspects: poverty analysis; indicators and monitoring; the macroeconomic framework and pro-poor policies; costings, the budget, and the medium-term expenditure framework (MTEF); and participatory processes. The paper presents a brief description of good practice at the beginning of each section.3 However, it is not expected that all countries will achieve this level of good practice; it is the incremental progress from the starting point that matters most. The analysis concentrates on six countries with which the authors were directly involved during 1999-2000, and is based on PRSP documents including those that were presented to the Boards of the Bank and the Fund4 and on mission experience during that period. It includes three full PRSPs (Uganda,5 Burkina Faso, and Tanzania) and three interim PRSPs (The Gambia,6 Mozambique, and Mali).7 Of these, Tanzania is the only country to have completed both an interim PRSP and a full PRSP.8 More detailed examples of country experiences with poverty analysis, monitoring and participation are provided in appendices I and II.

A. General Points

Four general conclusions can be drawn from the analysis:

  • The six countries have made rapid progress in formulating the interim and full PRSP due, in part, to strong government ownership and leadership, the prior existence of national poverty reduction strategies, donor support, and the importance that governments attached to obtaining rapid HIPC Initiative relief.

  • Through the PRSP process, governments are working toward placing poverty reduction at the top of the development agenda, stressing the importance of achieving sustainable poverty reduction through broad-based growth, as well as policies aimed directly at tackling poverty. They are also setting up systems to monitor poverty outcomes, thereby creating a demand for timely and accurate poverty data; recognizing the importance of governance as part of a poverty reduction strategy; and encouraging broader participation and a more diverse dialogue around poverty issues. Donors are also giving these issues increased emphasis.

  • There were significant differences between the full and interim PRSPs in terms of data availability, quality of data analysis and policy formulation, development of medium-term budgeting procedures, and participatory processes. There were also differences within both groups. The diversity is the result of differences in: country starting points; political, economic, and social contexts; donor support; and country capacity. Such diversity should be welcomed as it indicates that a blueprint approach has been avoided. It also suggests areas where capacity building may be necessary.

  • The PRSP concept is ambitious in its attempt to establish a framework that fully integrates - perhaps for the first time in some countries - poverty analysis, public policy, budgetary processes and monitoring systems; and to do this in a participatory way. A start has been made in the countries reviewed, but inevitably it will take some time for a fully integrated framework to be developed. To improve the content of future PRSPs, governments are working to gather better information on: the conditions and causes of poverty; the impact on poverty of growth, macroeconomic and structural policies, and governance; and the relationship between public expenditure and poverty outcomes. Coordinated donor support and south-south exchanges would also help towards strengthening future PRSP processes in these respects.

B. Specific Points

In addition, five specific points can be made:

  • Poverty analysis. All cases presented an analysis of the current poverty situation, although the scope has varied. In general, the analysis could have been strengthened by integrating and utilizing existing data more fully. Some countries faced capacity bottlenecks in implementing such analysis, and access to technical support from donors has varied. Much of the poverty analysis was based upon the traditional poverty indicators of economic well-being and human development (education and health). However, some of the countries (Uganda, Tanzania, and The Gambia) have tried to include a preliminary analysis of other dimensions, such as security, vulnerability, and powerlessness. These countries have also made use of qualitative and participatory surveys, such as Participatory Poverty Assessments (PPAs), which means that the poor’s perceptions of poverty and their priorities influenced the development of the strategies.

  • Indicators and monitoring. While poverty monitoring is weak in most of the sample countries, with the possible exception of Uganda, all of them have recognized the need to improve this and plan to establish better integrated information systems. All countries are beginning to identify goals, indicators and targets. The next step will be to strengthen the links between them and to distinguish more clearly between intermediate and outcome indicators. All the full PRSPs and The Gambian I-PRSP refer to the International Development Goals. There is considerable variation in the number, breadth, and type of indicators, and they are not always clearly prioritized. Civil society has participated in the selection of indicators in the full PRSP cases (Uganda, Burkina Faso, and Tanzania).9 In the I-PRSPs, the indicators have generally been derived from existing government programs. As countries move to full PRSPs, these indicators may be modified, if necessary, to reflect the poverty analysis and poor people’s priorities.

  • Links to policy and the macroeconomic framework. In most of the sample cases the macroeconomic framework and three-year policy matrix has drawn at least partly on the current Fund-supported PRGF program.10 All strategies recognized that macroeconomic instability, structural weaknesses, inadequate social policies, and weak governance could impede the extent to which the poor will benefit from growth. However, so far, it seems to have been difficult to demonstrate these links in the full PRSPs because there was a need to develop more fully the analysis of the poverty impacts of these policies. In addition, in most of these early cases, there was limited discussion of policy options, trade-offs and sequencing. There has been little civil society participation in macroeconomic and structural policy formulation because of the limited capacity in this area, and some civil society groups have requested assistance in order to better understand the economic debates. Appropriate national institutions should therefore be encouraged, with donor support, to provide training for civil society organizations in financial programming, the poverty impacts of macro reforms, and the analysis of policy trade-offs.

  • Budget, costings and financing the PRSP. In the full PRSPs, the MTEF is an effective tool for policy planning and increasing the policy debate within the PRS framework, both within government and more widely. However, MTEFs are not yet established in most of the I-PRSP sample countries and it would be difficult to make any new ones operational before these countries complete their full PRSPs. Moreover, it has proved difficult to prepare and complete costings of priority public policies in most cases. Even the full PRSPs in the sample countries have only provided detailed costings for a few of their policy programs. Assessment of the resource implications for financing the poverty reduction strategy is one of the weakest points (with the possible exception of Uganda), and urgently requires donor-funded capacity building. Experiences in increasing citizen participation in the budget process have proved to be an effective way of prioritizing expenditures and increasing transparency.

  • Participation. A major achievement of the PRSP initiative has been the increase in government ownership of the process. In most sample countries, the process of developing the PRSP has been managed at high levels of political authority (although, in some cases, it has been restricted to a small group within government). As a result, there is a greater commitment to broader participation in policymaking and a more diverse dialogue on poverty issues through direct consultations with stakeholders, including poor communities. The three full PRSP countries differed in the scope of their participatory processes; and the more comprehensive ones have built on existing democratic and participatory processes (involving parliaments and local levels of authority, both traditional and formal). The extent of consultation in the preparation of the I-PRSPs was larger than expected, again reflecting countries’ ongoing processes; although plans for deepening and extending this participation for the full PRSP needed to be fully developed. There has been some confusion as to whether countries could use their own names for the PRSP. To increase ownership, it should be clarified that countries can name their strategies as they see appropriate, as was done in the case of Uganda, Mozambique and The Gambia.

Table 1 provides an overview of the extent of coverage of the key aspects analyzed in the six PRSPs.

Table 1.

Overall Review of Full PRSPs and Interim PRSPs1

article image

The criteria listed here are based on the good practice pointers outlined at the beginning of each section of the paper.

II. Poverty Analysis

A. What is Required for Poverty Analysis?

Effective policies for poverty reduction need to be grounded in a good-quality poverty analysis. Ideally, this should include a description of the key features and determinants of poverty, an indication of how these will be affected by proposed policy actions and an analysis of the impacts of past policies on poverty.11 The description of poverty should be based on a multidimensional perspective that encompasses the following:

  • economic well-being (as reflected in income, consumption, employment, and access to productive assets);

  • human development (such as health, education, and nutrition);

  • insecurity (including lack of assets for a secure livelihood, exposure to income shocks, personal violence, and civil conflict); and

  • powerlessness (the lack of human rights and weak capacity of poor people to influence policy).

Key determinants of poverty may include broad economic and political factors, such as weak or unstable growth, poor market integration, lack of access to technology or credit, insufficient or inefficient public spending, weak institutions, poor governance, or inequality.12 They may also include individual or social factors, such as age, disability, gender, ethnicity, caste, and social exclusion.

Comparisons of poverty over time require a quantitative comparator for measuring changes (usually based on household consumption). This information is used to calculate the poverty line, which is a reference point for estimates of the incidence, depth, and intensity of poverty.13 Sources for quantitative poverty analysis include household surveys,14 censuses, and administrative data from line ministries. Rapid surveys (such as the Core Welfare Indicators Questionnaire (CWIQ)) and consumer satisfaction surveys are increasingly used for monitoring the poor’s access to, use of and satisfaction with services.

Direct consultations with the poor through Participatory Poverty Assessment (PPAs) and specialized surveys using other qualitative methods have shown that poor people often emphasize multiple dimensions of poverty, such as vulnerability, physical and social isolation, lack of security and self-respect, a distrust of state institutions, powerlessness, and lack of dignity. PPA data can describe the priorities of the poor, their access to assets and services, and processes of impoverishment. Although PPAs use a selected sample of communities and are not as extensive (or representative) as household surveys, they are more rapid, less costly, and produce in-depth analyses of the perspectives of the poor. For PPA data to be credible, they should be combined with other data sources, including larger, representative household surveys, where the findings of one source may confirm, refute, enrich or explain the other.

Ideally, a good poverty analysis in a PRSP would use multiple data sources and include the following: a) measures of the incidence, depth, and intensity of poverty in relation to the poverty line; b) disaggregated analysis of who is poor, what are their priorities, and where they are located (by gender, ethnic group, socioeconomic group, income quintile and geographic region); c) changes over time, and processes by which people move in and out of poverty; d) identification of the determinants of poverty and obstacles to poverty reduction; e) impact of various policies on the poor; f) links between growth, inequality, and poverty reduction; assessments of the risks and insecurity of livelihoods faced by poor households; and g) analysis of the constraints preventing the poor from influencing state institutions and service providers.

B. Country Experiences with Poverty Analysis15

Data availability. The PRSP process is creating a demand for improved poverty analysis, utilizing multiple data sources. In most of the sample countries, the poverty analysis was somewhat constrained by a lack of data or limited utilization and integration of existing sources, although new data are becoming available. In the countries covered here, data availability and reliability differed greatly (see Table 2). Uganda, Burkina Faso, The Gambia, and Mozambique all benefited from more up-to-date data and were able to estimate the extent of poverty nationally. Tanzania and Mali will not have a baseline from which to track the effects of the PRSP on poverty until next year. In the absence of a national poverty line, they had to rely on basic needs data, drawn from the demographic and health surveys and administrative statistics. Uganda and The Gambia have made effective use of participatory surveys, such as PPAs.

Table 2.

Data Sources Referred to in the Full PRSPs and Interim PRSPs

article image

Poverty analysis. Many of the sample countries provided a sound analysis of poverty, benefiting from inputs from donors and international researchers. Others were constrained by limited resources and weak analytical capacity. Uganda’s PRSP/PEAP and The Gambia’s I-PRSP (I-SPAII) presented a multidimensional definition of poverty, and the poor were directly consulted. In Uganda and Burkina Faso, a detailed causal analysis was undertaken and the links between growth and poverty were analyzed. Uganda’s good quality poverty analysis was also the result of high-level government commitment to poverty reduction, donor support, and the allocation of resources to poverty analysis.

Much of the poverty analysis was based upon the traditional indicators of poverty - economic well-being and human development (see Table 3). Ideally, future analyses would also cover security, vulnerability, and powerlessness. This could include people’s access to assets and security of livelihood; gender; ethnicity; and the informal institutions and coping strategies that mitigate poverty. Looking ahead, in most of the sample countries, fuller analyses of urban poverty, employment, migration and HIV/AIDS would be useful.

Table 3.

Examples of the Content of Poverty Analysis in the PRSPs

article image

Improved analysis of the links between growth and poverty reduction, and the poverty impact of past policies, conflict, external shocks, and regional, institutional and political factors could also be useful. However, these are complex issues that will require efforts to develop countries’ analytical capacity, with coordinated donor support. Ideally, the poverty analysis should link forward to the policy design as was done in the Uganda PRSP/PEAP. This example is summarized in Table 4 below, which was based on the PEAP.

Table 4.

Linking Poverty Data to Policy Responses in Uganda2

article image

Adapted from Bella and Margaret Kakande, “The Uganda Participatory Assessment Process” in A. Norton, 2001, A Rough Guide to PPAs (London: U.K. Department for International Development).

III. Indicators and Monitoring

A. Establishing a Poverty Monitoring System

To assess whether the PRSP is having the expected impact on poverty requires an integrated monitoring system with goals, indicators, and targets. Poverty monitoring also promotes public accountability and helps to ensure that information about outcomes influences policies and budget allocations. Goals, indicators, and targets help to establish priorities, focus internal resources, and mobilize external resources. They provide benchmarks to monitor progress and adjust policies and programs.

Goals are the general objectives that a country wants to achieve, indicators are the variables used to measure progress toward the goals; and targets are the quantifiable outcomes that are expected to be achieved.16 Indicators can be broadly classified into two types, outcome and intermediate. “Outcome,” or “impact,” indicators refer to the final effect of an intervention on an individual or group. “Intermediate” indicators measure factors that contribute to the outcome and, depending on the stage in the process, could either be “input” or “output” indicators. Input indicators generally refer to the resources used to produce public services, and output indicators may refer to the services provided, for example, classrooms built (see Table 5 below). Indicators tend to vary according to who was involved in their selection, the accuracy of the indicator, what is achievable, and the timeliness of data collection.

Table 5.

Example of Goals and Indicators

article image

The PRSPs can be guided by the International Development Goals (IDGs) or the Millennium Development Goals (MDGs). The IDGs were established in a series of UN summits during the 1990s to promote social development and social justice. They focus on economic well-being, social development, and environmental sustainability.17 The MDGs were established at the 2000 Millennium UN Heads of State Summit and differ from the IDGs only in the exclusion of reproductive health and infant mortality goals and the addition of a goal to reduce major communicable diseases, including HIV/AIDS. At the country level, the international goals would need to be modified to reflect national economic, social, cultural, and historical circumstances, as well as the participatory process through which policy objectives and indicators were agreed.

A good monitoring system would include indicators that specifically target improved outcomes in terms of economic well-being, human development, security, and empowerment of the poor. It would also include indicators that will improve poverty indirectly, for example, by promoting growth and structural reform and by enhancing governance and public sector performance. Ideally, the indicators should be limited in number, prioritized, focused on the key poverty problems identified, and able to measure change. In order to reflect diversity amongst the poor indicators should be disaggregated by age, gender, socioeconomic group, and income quintile. The geographical dimension is also important for decentralized budget administration and service delivery.

Civil society groups, including the poor, may participate in the selection of goals, indicators and targets, and in monitoring the delivery of public services. Participation in poverty monitoring promotes transparency and accountability and increases ownership and acceptance of the findings.18 Participatory monitoring can enhance people’s awareness of their rights and strengthen their claims. In Africa, communities often have strong, local institutions that can help to monitor programs and policies, which can be particularly useful in countries where public institutions are weak. People’s own perception of their well-being can be added as an outcome indicator.

Poverty monitoring requires the establishment of an integrated information system, in which different types of data are pooled and made available to the public in accessible format. Frequent problems in poverty monitoring are: a) overemphasis on large-scale quantitative data and neglect of small-scale participatory or qualitative data; b) poor coordination between data collection agencies and their sponsors; c) underestimation of the costs of monitoring; d) proliferation of indicators; e) a poor understanding of the needs of data users; f) a lack of civil society participation; and, g) limited resources for analyzing and publishing the results. In Africa, few survey programs have been institutionalized to monitor the extent and nature of poverty and the effectiveness of specific antipoverty interventions. Therefore, their relevance may be short-lived as conditions change, and the investments in new survey skills, policy analysis, and program design may not be fully exploited.

A major data source for poverty monitoring is administrative data, collected by statistical units in the line ministries. General survey and census data are usually collected by national statistical offices or specialized line ministries (e.g., agricultural and employment surveys). Small-scale surveys are often conducted by academic institutions, donors, and/or nongovernmental organizations (NGOs). PPAs are often organized by governments in collaboration with NGOs and academic institutions. Close coordination is required among all these institutions. However, often there is a lack of information about who has what data, and access to it is difficult. This can lead to duplication, noncomparability, and underutilization of data. Ideally, PRSPs should set out clearly defined institutional roles regarding data collection, analysis, and publication. A sustainable, financed and well coordinated information system can facilitate a more effective planning of activities and more efficient allocation of resources.

B. Country Experiences with Indicators and Monitoring19

In the sample countries, most full PRSPs and interim PRSPs presented a set of goals, indicators and quantified targets.20 As was to be expected at this early stage, the indicators often reflected existing government programs. In some countries, a multitude of indicators and targets existed at the national level, often set by donors. However, in Uganda and Burkina Faso, indicators were being streamlined and prioritized. Uganda’s PRSP/PEAP had the most comprehensive set of indicators, reflecting access to more detailed data, the wider scope of the poverty analysis, the outcome of the participatory process (in which issues of security, corruption, and empowerment were emphasized), and the inclusion of politically sensitive issues such as corruption (see Table 6). Civil society participation in the selection of indicators had led to increased commitment and ownership. Most of the sample countries had not focused on indicators for measuring social exclusion, social cohesion/fragmentation, trust, and powerlessness. This may be due to the fact that such indicators are relatively new and countries do not have the capacity or methodologies to define and monitor them.

Table 6 shows the variability among PRSPs in the number and types of indicators. Most sample countries had indicators on economic well-being and human development, but few had them for security and empowerment.

Table 6.

Indicators in the Full PRSPs and Interim PRSPs

article image

The PRSP focus on linking indicators to outcomes will create a greater demand for poverty data and analysis. The countries analyzed have been able to respond to this through their existing poverty units. In Burkina Faso and Mali, “Observatories” had already been established to analyze poverty data for the annual human development reports. Uganda, The Gambia and Mozambique have poverty-monitoring units in the Ministry of Finance. In Tanzania, the Vice President’s Office has the overall responsibility for monitoring the implementation of the PRSP. All strategies recognized the need to develop more coordinated information systems for monitoring poverty.

IV. Pro-poor policies and the Macroeconomic Framework

A. What are Pro-Poor Policies?

The characteristics of pro-poor policies can be identified as follows:

  • Pro-poor policies may benefit the poor directly or indirectly. It is useful to distinguish between general, sectoral/cross-sectoral, and targeted policies. General policies are those with economy-wide effects, and are designed to influence the general economic and institutional environment in order to promote growth from which the poor will benefit (e.g., macroeconomic policies and structural reforms). Sectoral and cross-sectoral policies are more specific in focus, and benefit both the poor and the non-poor (e.g., universal access to primary education). Targeted policies refer to specific targeted programs, such as social safety nets.

  • Pro-poor policies aim to promote the necessary conditions for growth, for example, macroeconomic stability, competitiveness, and an enabling institutional framework that will attract private investment and generate employment. The poverty impact of such policies may sometimes be difficult to quantify, but should be analyzed to the extent possible, to clarify the links between macroeconomic policies and poverty reduction.

  • Sectoral or cross-sectoral programs (e.g. HIV/AIDS) may be oriented to the general population, including the poor. Such programs are not confined to the social sectors (health and education) but also include, for example, rural development, land reform, small enterprise development and judicial reform.

  • Given resource constraints, the policy package should analyze the trade-offs implied in setting goals and priorities. An assessment of the effectiveness of past policies should inform the policy debate. For the full PRSP, decisions about goals, policy priorities, and public actions should be influenced by discussions with a broad range of civil society stakeholders.

The full PRSP should be consistent with a sustainable framework of macroeconomic stability. Pro-poor budgets would be undermined if fiscal sustainability was compromised, and they did not take account of absorptive capacity constraints. A PRSP should assess the scope for domestic resource mobilization; increasing the proportion of pro-poor spending in the budget; external financing that reflects debt-servicing capacity; and domestic borrowing consistent with sustainable economic growth and low inflation. Policymakers would need to go through an iterative process of adapting the spending plans in the PRSP to external and domestic financing constraints, in order to ensure that the macroeconomic framework is consistent with the objectives of rapid economic growth, low and stable inflation, and poverty alleviation.

Structural reforms are necessary if countries are to move to a higher path of sustainable growth. Weaknesses or distortions in land ownership, financial markets, labor markets, the civil service, and the foreign trade regime may impede the extent to which the poor can share the benefits of growth. While such issues are not expected to be fully addressed in I-PRSPs,21 it is important for countries to begin to identify them as early as possible, so that they can be discussed in the context of the participatory process leading up to the full PRSP. In this regard, questions about the appropriate sequencing of structural reforms, the short- and long-term poverty effects of reforms, and actions needed to offset any adverse effects on the poor would need to be addressed. Public debate on these issues is becoming more frequent and effective in many countries.

B. Country Experiences with the Design of Pro-Poor Policies

Most of the I-PRSPs and PRSPs provided a good justification for the policy framework proposed, drawing on policies and strategies already in place. However, the links between macro and structural policies and poverty could be strengthened (sec Table 7).22 Almost all PRSPs recognized that weaknesses or distortions in land ownership, financial markets, labor markets, the civil service, and the foreign trade regime can impede economic growth. However, the poverty effects were only analyzed in some cases (Table 8). To support the analysis of the poverty impact of macroeconomic policies and structural reforms, there is an urgent need for donors and governments to work together, on an ongoing basis, using a variety of methodologies. In the sample countries, there was limited discussion of policy options, trade-offs and sequencing of reforms, and the actions needed to offset any temporary adverse effects on the poor. Most countries have not included detailed policies for social protection or an analysis of the environmental impacts of reforms.23 In all the sample countries, except Uganda, there was limited civil society participation in policy formulation.

Table 7.

Macroeconomic and Poverty Linkages

article image
Table 8.

Macroeconomic and Poverty Linkages

article image

Uganda’s PRSP/PEAP provides an exception to the comment above. Its analysis of the impacts of macroeconomic and structural policies on the poor is outlined Box 1 below.

Uganda’s PRSP/PEAP: Macroeconomic, Structural and Poverty Linkages

Links between growth and poverty reduction: The strategy acknowledges the importance of income distribution and growth. The growth of incomes and investment over the next 20 years is projected. The PRSP also analyzes the impact of the growth in incomes on absolute poverty and of the sectoral pattern of growth on poverty reduction. The PRSP lays out a strategy to promote pro-poor growth by increasing the accumulated assets of the poorest, distributing an equalization grant to the poorest areas, and increasing food production (through agricultural modernization and market access). The strategy recognizes that the impact of HIV/AIDS is yet to be factored into the growth projection and, in addition, that clearer analysis of the distributional considerations, especially with regards to women and children, is required.

Macroeconomic framework. The PRSP aims to provide a stable macroeconomic environment by containing inflation (which can directly reduce the poor’s ability to accumulate wealth by saving); maintaining reduced debt ratios; and ensuring that inflows of grants are growth enhancing and poverty reducing. The PRSP proposes a strategy to keep the real exchange rate competitive that will benefit farmers in the export crop sector. However, the strategy recognizes that not all farmers will benefit since the poorest are in the food crop sector.

Impacts of the taxation system on the poor. On the revenue side, the PRSP states that research undertaken shows that recent tax reforms, including the introduction of the value-added-tax (VAT), have made the incidence of taxes more progressive. The strategy adds that local taxation (for example, the graduated tax) is not progressive and the tax on kerosene may have a damaging environmental impact by encouraging the use of fuel wood. The PRSP adds that taxes on capital assets used by the poor will be avoided.

Medium-term expenditure framework. Budget reform in which the government is focusing on pro-poor budgeting is central to the PRSP. The Poverty Action Fund reallocates expenditures to directly poverty-reducing services, and the budget is being decentralized to districts. Local communities and political leaders are involved in monitoring budget delivery.

Governance/institutional factors affecting poverty. The PRSP focuses on addressing corruption, judicial reform, and human rights issues. Measures to increase political accountability are now being considered for funding from HIPC Initiative savings. The PRSP states that transparency will be increased through enhanced flows of information to the public. Commercial Justice is being reformed in order to promote the private sector, along with a new regulatory structure for procurement. A sector-wide approach is proposed in the PRSP by which reforms for the criminal justice sector will be costed.

Liberalization. The PRSP includes an analysis of the positive impact of the liberalization of the coffee sector on the poor. Competition in the coffee sector has meant that farmers are now paid promptly for their crops. However, the strategy recognizes that the main beneficiaries of liberalization are cash crop farmers and therefore proposes that more farmers need to enter the cash crop market. A strategy to increase the physical access to markets is detailed. The PRSP states that more data are required on the impact of liberalization on women since they control the income only from food crops, not cash crops. In some parts of Uganda, liberalization has led to a shift from food crops to cash crops, leaving women with fewer resources to feed their families and pay school fees. Furthermore, the commercialization of food crops may have reduced the availability of food provided to old people in rural areas.

Financial sector reform: The PRSP states that the government is establishing a new regulatory and supervisory structure for microfinance in order to increase poor people’s access to financial services. The government has withdrawn from the provision of capital through credit but will still support capacity building. The PRSP aims to reduce the problems the private sector has in accessing finance.

Privatization. The PRSP states that the government will analyze how to prioritize the privatization of the remaining commercial enterprises and recognizes that corruption and lack of broad participation have adversely affected the privatization process in the past.

The PRSP/PEAP analyzed the effects of the removal of tax distortions and the regressiveness of the taxation system, as well as the impact of liberalization in the coffee sector, on poverty. It also recognized that strengthening the financial institutions and the legal framework is important for increasing growth, although the steps required to achieve this were not clearly defined. The PRSP/PEAP also presented a redefinition of government priorities toward more pro-poor policies. For example, safe water has received significantly more resources; actions to improve security, governance, and public service delivery to the poor are central features; budget flexibility has increased to allow lower-level governments to respond to local priorities and political accountability; and sector policies have become increasingly pro-poor. Findings on powerlessness, corruption, and restricted information flows to communities have led to a strong policy recommendation to improve public information, and findings on poor leadership and a lack of accountability of public service delivery agents have led to a recommendation to strengthen local political accountability.

V. Costings, Budget and the Medium-Term Expenditure Framework

A. Making the Budget System Pro-Poor

For budgets to be more pro-poor it would be useful to consider the following aspects:

  • Developing a medium-term framework. The PRSP provides a medium to long-term perspective, and the scale of future public spending is affected by macroeconomic prospects. Annual budgets traditionally do not account well for the multiyear objectives necessary for effective poverty reduction strategies. Developing multiyear budgets can assist countries in preparing realistic strategic plans within the context of resource constraints, and help donors to make their plans and commitments more predictable. The medium-term expenditure framework (MTEF) is one vehicle for introducing a multiyear context to budgeting. In countries with MTEFs, it is easier to track fund disbursal and delivery because budgets are organized to supply this information. Also, funds are allocated in line with specified poverty reduction objectives rather than according to historical patterns. The development of an MTEF can also assist countries in assessing strategic policy choices with respect to budget constraints and implementation capacity.

  • Linking the budget to poverty priorities and costings. An objective of the PRSP is to link the budget to poverty outcomes. This would include analyzing the degree to which public expenditures are oriented towards the poor as well as assessing the efficiency and equity implications of these expenditures.

  • Focusing on management, tracking, and performance/delivery. In taking on the responsibility for tracking their poverty-related public spending (including the use of HIPC Initiative resources), countries have had to address the weaknesses of their public expenditure management systems. In this context, improvements would be helpful across the full range of financial management, from budget formulation to execution, monitoring, reporting and auditing.

  • Increasing transparency and participation in the budget-making process. Medium-term public spending plans also need to be transparent and accountable through the PRSP process. Increasingly, elected assemblies and civil society groups (including the poor) are becoming involved in discussing budget allocations and monitoring budget delivery, thereby increasing transparency and accountability. Publishing information on the budget increases the quality of public debate and the transparency of institutions. Budgetary expenditures should be made public and opinions about the quality of services should be sought from direct users, as well as from local government officials, politicians, representatives of local communities, and the media.

B. Country Experiences with Pro-Poor Budgeting

Developing a MTEF. In Uganda, Burkina Faso, and Tanzania, the PRSP has built on existing MTEF processes and used the budget as a key tool for policymaking. Similarities in all three countries may be identified. First, sectoral components of the PRSP were linked to the MTEF sectors, and line ministries were given budgetary ceilings for their allocations. Second, medium-term budget priorities were consistent with poverty priorities and medium-term macroeconomic stability. Third, the implementation of an MTEF was monitored through public expenditure reviews (PERs), with public consultations. In Uganda, some sector budget allocations are now more closely linked to anticipated outputs, and detailed sector plans are providing a framework for clear prioritization of public actions. To better reflect district priorities and to bring local governments into the medium-term expenditure process, local government officials also prepared medium-term expenditure plans. This process fed into the budget framework paper and annual budgets. An MTEF has been under development in Mozambique for some time, but it has only recently become operational and its links to the annual budgetary process were weak. For Mali and The Gambia, which have relied on an annual budgeting system, it would be useful to develop an MTEF.

Linking the budget to poverty priorities and program costings. Most sample countries have made some progress in costing some of their programs, although at this early stage, it has proved to be difficult because of the weakness of public expenditure management systems. Only a few countries had detailed financing plans for their sector programs, and these were usually prepared with donor support in the context of sector wide approaches (SWAPs). In most of the sample countries, costed programs, sector budget frameworks, and outputs were weakly linked; most of them had only a description of budgetary allocations. In Burkina Faso, the PRSP provided estimated costs for some priority poverty programs and indicated how additional donor assistance and HIPC Initiative resources were integrated into the national budget. In Uganda, the PRSP estimated the costs of some programs, but the accuracy of these estimates is unclear. Furthermore, not all interventions were costed because of limited information, and the sources of funding are unclear. In both Uganda and Burkina Faso, poverty reduction targets and indicators could be more strongly linked to the MTEF. The lack of costings and links to the MTEF makes it hard to determine whether a poverty reduction strategy is consistent with a country’s fiscal framework, or to make a case for changing the framework to accommodate a poverty reduction strategy. This is clearly an area for capacity building, with donor support.

Focusing on management, tracking and performance/delivery. The PRSP process is highlighting the need for improved public expenditure management systems, with Uganda and Burkina Faso having the more developed systems. Burkina Faso has a computerized expenditure management system that delivers data on commitments and expenditures for every line item. In addition, the budget presents a breakdown of proposed donor projects (by grant, loan, and subsidy). In six ministries, there has been a shift to performance budgeting. Finally, budget monitoring indicators were defined through the Pilot Test for Conditionality. However, although public expenditure on health and education increased, it is unclear to what extent the budget reflected the priorities of the poor, and the delivery of the budget at the community level could still be improved. Both Uganda and Tanzania have undertaken some analysis of the gender equity and efficiency implications of budget allocations.

In Uganda, accountability for implementing the poverty reduction strategy was strengthened through the MTEF, with limits placed on supplementary budgeting and conditions imposed on the release of district-level funds to ensure that they are spent for poverty reduction purposes. In addition, the budget captures donor flows, and donors are consulted on priorities. A virtual Poverty Action Fund (PAF) was established in 1998 to prioritize budget allocations that directly benefit the poor, to enhance transparency, and to track the use of donor resources. Critical features of the PAF are that (a) it is fully integrated in the budget; (b) it involves civil society in the selection of programs and projects; and (c) it allocates resources to monitor the impact of PAF outlays, with the involvement of both civil society and government.

Many of the sample countries are currently upgrading their public expenditure management systems, although this work is not detailed in the PRSPs. Most countries have had some PERs, but generally the results need to be better integrated into the PRSP through a discussion of the efficiency and poverty incidence of public spending.

Increasing transparency and participation in the budget-making process. Experiences in opening up the budget process have proved to be an effective way of increasing transparency. In Uganda, a participatory budget preparation and monitoring process has been developed, with multiple channels for accountability (involving local authorities, press, community groups, NGOs, and donors). Civil society is involved in the dialogue on priorities and spending commitments, as well as on expenditure monitoring. Budget allocations are published, both nationally and at the local level. For example, advertisements are placed in the press indicating amounts disbursed to each district. In the education sector, budget allocations for schools are posted on school notice boards—although compliance is not perfect. Civil society groups meet quarterly with central government officials to discuss delivery against budget allocations.

In Burkina Faso, PERs were undertaken with the participation of local consultants to build local capacity. In addition, the budget is prepared in consultation with civil society groups. However, a system still needs to be developed whereby communities can report back when the budget is not being delivered. In Tanzania, the budget process is being opened up, with quarterly review meetings conducted with civil society groups to monitor public expenditure, and annual participatory PERs undertaken. In 1997, the Tanzania Gender Networking Program launched a Gender Budget Initiative (GBI), with the involvement of over 20 NGOs (although details of this were not reflected in the PRSP). In The Gambia, the Ministry of Finance is working with The Association of Non-Governmental Organizations (TANGO) and Action Aid, to find ways to increase the participation of the poor in budget allocations, expenditure tracking, and service delivery. So far, there has been relatively limited participation of civil society in budgetary processes in Mozambique and Mali.

VI. Participatory Process

A. Participation in Policymaking

Participation is a process through which groups in society can influence development initiatives and public policy. Box 2 below outlines why participation is important in policymaking.

Why Is Participation Important in Policymaking?

Participation can strengthen policymaking in a number of ways:

  • Participation processes build ownership of policies and actions, with political and administrative commitment, by which governments are held accountable to their people.

  • Participation contributes to the establishment of a more accountable government. Transparency is increased when the public better understands government processes; the accountability of institutions is strengthened: the poor are better informed of governments’ commitments to tackling key poverty issues; and governments arc held accountable to their domestic constituencies for actual performance. Building the role and capacity of nonstatc institutions can counterbalance the powers of the state.

  • Civil society organizations (CSOs1/) can provide specialist and local knowledge to improve the quality of policymaking. Although not all are representative and some are politically aligned, others genuinely represent the views of the poorer citizens and other interest groups, such as the business community.

  • The poor are empowered by bringing their analysis, priorities and voice into the decision-making process, thereby making the policy framework more relevant and responsive to their needs.

  • Equity and social integration are promoted through appreciation of varied needs within society (by geographic region, gender, ethnic group, etc). Participation can lead to the creation of durable, inclusive political systems (which are particularly important in ethnically diverse societies).

  • Partnerships are built between governments and business and civil society groups, leading to joint actions to achieve poverty reduction goals.

1/ CSOs include NGOs, labor unions, business and professional organizations, religious bodies, and other citizens groups.

Participation is not a uniform process. Rather, it involves a continuum of approaches ranging from information dissemination, consultation, and collaboration to shared decision making. Of these four approaches, (a) information dissemination is a one-way, top-down flow of information to the public, with little opportunity for feedback; (b) consultation is a two-way flow of information between the government and the public, where feedback from the latter may influence a policy or product that has been initiated by the government; (c) collaboration involves shared design of the policy or product; and (d) shared decision making involves direct participation and control over policy decisions, including policy formulation, implementation, monitoring, and evaluation. As is widely recognized, it is not appropriate to have rigid guidelines for participation since country contexts and political systems vary.

Consultations are not required to be undertaken for the preparation of the I-PRSP, although many countries have been able to build on some prior experience. As such, the tension between speed and participation is not relevant since extensive participation is not expected. However, in order for the public to have a better understanding of the boundaries of participation and, therefore, more realistic expectations, the I-PRSP should ideally detail a participation action plan that sets out the nature of participation in formulating the full PRSP; equally, the full PRSP should include a participation action plan for implementing it (see Table 9). The action plan promotes discussion and consensus around who should be involved, provides a basis for monitoring the participation process itself, and indicates when and how to build on existing democratic processes. It will vary from country to country depending on the political context and the extent of participation that the government feels is appropriate. In some countries, narrow participation (even limited to within the government) may be more appropriate at the beginning of the process. Different groups, such as parliament, local representative bodies, civil society and private sector groups, and the poor, can be gradually included at a later stage when politically feasible.

Table 9.

Participation in Full PRSPs and Interim PRSPs

article image

The participatory process refers to all phases of policy formulation and implementation and should not be limited to the preparation of the PRSP document. There are four main entry points for participation in the PRSP process:

  • participation in the preparation of the I-PRSP/PRSP document;

  • participation in poverty analysis to understand the multidimensional aspects of poverty (see Section II above);

  • participation in the policy dialogue (macroeconomic, structural and sector policies), including budget allocation, expenditure tracking, and service delivery (see Sections IV and V above); and

  • participation in monitoring and evaluation to promote transparency and accountability, and increase ownership and acceptance of the findings (see Section III above).

A first step in the PRSP process is to identify and build on existing democratic and participatory processes (e.g., consultations on existing poverty strategies, civil society organizations (CSO) networks on policy advocacy, parliamentary and local government structures, and other elected officials, donor collaboration on sector-wide approaches (SWAPs), and participation in the budget process). A stakeholder analysis can be conducted to identify stakeholders, evaluate their interests and capacities, and determine what type of participatory process will be most appropriate. Transparent decisions as to who participates, and when and how they participate, will add legitimacy to the participatory process. Ideally, poor communities would be involved in the consultations.

Throughout the PRSP process, governments are encouraged to consult on such issues as a) poverty reduction goals; b) priority public actions to achieve those goals; and c) the selection of indicators and monitoring of the implementation of the PRS. Where possible, the full PRSP would include discussions on such policy issues as the links between growth and poverty, structural reforms, public expenditure allocations, and policy trade-offs. The outcome of the consultation process, including differences of opinion, would be documented in the interim and full PRSP.

B. Country Participation Experiences in the Preparation of the PRSPs24

The levels of participation in preparing the I-PRSPs and PRSPs and the content of discussions have varied across the cases reviewed (Table 10). The level of participation was influenced by the political context, time constraints, existing processes, and World Bank/IMF advice. In all countries, the PRSPs have been managed and guided at very high levels of political authority, although the breadth of involvement at this level has been variable. The extent of participation in the I-PRSPs was greater than expected, reflecting countries’ existing processes. However, the full PRSP countries differed in the comprehensiveness of their participatory processes. Of all the countries, Uganda exhibited the greatest range in the types and levels of participation, with full participation by the poor in some policymaking processes, especially through the decentralization of decision making and resource allocation. The Gambia has also had participation at higher levels through extensive consultation and collaboration. Tanzania’s process included both information dissemination and consultation, while, in all the other countries, the participatory process was mainly limited to information dissemination. With the exception of Uganda, the outcome of the consultation process and its effects on policy changes were not clear. It would be useful for the full PRSPs now being developed to summarize the major discussions held during the consultation process and give details of what can and cannot be followed up on through public actions.

Table 10.

Levels of Participation at a Glance

article image

In all the sample countries, the capacity of civil society to participate in dialogue on complex policy issues, such as the macroeconomic framework, the sequencing of structural reforms, and policy trade-offs, was limited. This situation reflects the limited degree of consultation and public discussion of poverty and policy issues in the past, and the lack of accessible information on the impact of macroeconomic and structural reforms on poverty. In both The Gambia and Uganda, civil society groups requested training on assessing macro economic issues.

Full PRSPs. As can be seen in Table 11, the full PRSPs varied in the comprehensiveness of their participatory processes. In Uganda, existing democratic consultations were used as the basis for the PRSP, building on years of experience, and country ownership is high. Consultations with civil society in Burkina Faso were much more limited in spite of its tradition of participation, although discussions did build on the development during the last two years of sector strategies in health and education. The PRSP was drafted by a small group in government with assistance from a few donors, but the document was presented to both chambers of parliament (National Assembly and House of Representatives). Both the Bank and the Fund Boards discussed whether Burkina Faso’s PRSP should be considered as an interim rather than a full PRSP on the grounds of limited participation. In the event, both Boards endorsed the document as a full PRSP because of other strengths in the document.

Table 11.

Participation Inputs for the Full PRSPs

article image

Interim PRSPs. Table 12 shows that all three countries were able to describe and build on existing participatory processes. However, Mali’s and Mozambique’s I-PRSPs did not include a stakeholder analysis or a participation action plan for the formulation and disseminating of the full PRSP. In The Gambia, the I-PRSP (I-SPA II) built on the first Strategy for Poverty Alleviation I (SPA-I), which was formulated in 1994, and in addition, a clear participation plan is presented.

Table 12.

Participation Inputs for the Interim PRSPs

article image

Table 13 below provides a summary of information on who was consulted, as well as a comparison of the content of the consultation process, and further details are provided in Appendix II.

Table 13.

Process and Content of the PRSP Consultations

article image

VII. Conclusion

In just over a year since the PRSP initiative was launched, the six countries analyzed in this paper have produced coherent national poverty reduction strategies or have viable road maps for the production of one. This is a remarkable achievement, given capacity, time, and resource constraints. However, the PRSPs have been uneven, with both strengths and weaknesses in different areas.

A major accomplishment of the PRSP initiative has been the visible efforts to build government ownership of the process, with a commitment to broader participation in policymaking and a more diverse dialogue on poverty issues through direct consultations with stakeholders, including poor communities. However, the extent of participation in the PRSP process has been variable, reflecting different local contexts and degrees of experience with participatory approaches. In countries where the process has been more participatory, the increased dialogue has enhanced the quality of policymaking and encouraged greater transparency in government. Participation has had a greater impact where consultations have been continued as an ongoing process, beyond the completion of the PRSP document (for example, in Uganda and The Gambia). Participation has been less successful where it has not built on existing democratic or administrative processes or has attempted to duplicate existing ones.

The analytical content of the PRSPs has been of variable quality, reflecting differences in capacity and donor coordination. Some PRSPs were able to draw on good poverty data, while others needed to update their information. In most cases, there was also a need to deepen the analysis and to link it to policy decisions. Many of the strategies highlight the importance of broad-based growth and better governance for poverty reduction. However, in all cases studied, the lack of analysis of the poverty impacts of prior and current programs limited the discussion of policy options, trade-offs, and sequencing. These issues should be debated more widely within government and civil society. There is also a need for donors and governments to work together in developing capacity to analyze these issues, as part of routine poverty analysis.

The linkage between pro-poor policies and resource allocation was weak in almost all the sample cases. Only a few of the proposed policies and programs were costed, and budgetary ceilings were not linked to medium-term expenditure frameworks. However, in those countries that had them, MTEFs were proving to be an effective tool for policy planning, and these now need to be developed in the other countries.

The emphasis on poverty outcomes has encouraged countries to establish more systematic monitoring systems, which are beginning to create a greater demand for accurate poverty data, promote public accountability, and ensure that monitored outcomes inform policies and budget allocations.

The Uganda case stands out in comparison to the other countries; it has been the result of, inter alia, strong political commitment and leadership by the government and well-coordinated donor interventions over several years. It provides an example of how a participatory poverty strategy can be developed. However, it should not be used as a yardstick against which other countries (which are confronting different problems and operating with different resources and constraints) will be measured. For all countries, it is the incremental progress from the starting point that matters most.