Information about Sub-Saharan Africa África subsahariana

2 Impact of the PPA

Caroline Kende-Robb
Published Date:
January 2002
  • ShareShare
Information about Sub-Saharan Africa África subsahariana
Show Summary Details

Including the poor in policy dialogue has great potential for creating better poverty reduction policies. The original rationale of the participatory poverty assessments (PPAs) was to influence the policy dialogue by collecting information on the poor’s perceptions of poverty. Most PPAs have achieved this objective to some degree, but with substantial variation in the level of impact. The PPAs with the greatest impact tended to be those that implicitly or explicitly had more ambitious objectives. It is useful to assess impact in relation to three objectives:

  • Deepening the understanding of poverty: Through the incorporation of the results of participatory techniques into a diagnosis of the nature and causes of poverty
  • Influencing attitudes and policy: Through the use of the PPA process within a broader participatory process that engages policymakers
  • Strengthening the policy delivery framework: Through the creation of a new institutional alignment that increases policy impact for effective, sustainable poverty reduction.

Although the principal objectives of PPAs have been to diagnose the causes and nature of poverty and to influence policy, some PPAs have been successful in fostering dialogue with and building the capacity of credible poverty reduction institutions, which then create links between traditional and formal institutions. These links have created room for a more coordinated approach to poverty reduction among various stakeholders, including donors, with ongoing and increasing interaction between policy change and stakeholder dialogue. This process is long, slow, and continuous and requires the redefinition of stakeholder relationships, including relationships with the World Bank. Ideally, governments should lead the process, or lead in partnership with other institutions, and development partners should offer support and advice. This policy change and institutional strengthening at all levels are part of a wider process of establishing linkages between the poor and those in power.

This section uses case examples to explore the diverse array of observed impacts of PPAs in the context of the three categories mentioned above: the extent to which PPAs have deepened the understanding of poverty; their impact on attitudes and consequent policy change; and the extent to which frameworks for policy implementation have been strengthened. These impacts are summarized in table 5, which also links the various levels of impact to the different PPA approaches and the required shifts in the thinking of policymakers.

Table 5.Range of PPA Impacts
PPA impactTypical approach

and inputs
Potential outcomeRequired shift

in the thinking

of policymakers
1. Deepening the understanding of poverty

Participatory data incorporated in the analysis of poverty.
  • Rapid appraisals in the field (e.g., 3 weeks). Information extraction. Limited feedback and action at the community level.
  • Policymakers not necessarily included in the process.
  • Prescriptive and often top-down in nature.
  • Isolated exercises with limited impact on the wider development process.
  • Changes in policy documents reflecting views of the poor.
  • The information might be accurate and interesting, but there is limited room for government ownership or for changing attitudes of policymakers.
  • The poor are given a voice, but mere is limited commitment from the top to ensure that the poor’s concerns remain on the agenda.
  • Issues such as power, decentralization, and gender are considered but not always included in ongoing debate.
  • Policymakers might feel threatened.
  • Poverty viewed as a multidimensional phenomenon, the character of which is defined by the community.
  • Policymakers understand the value of participatory research and of including the perceptions of the poor.
2. Influencing policy

Policies realigned toward poverty on a long-term basis by focusing on changing attitudes.
  • Feedback and follow-up to field appraisals (e.g., the poor validate information). Ownership of the information at the community level Development of action plans and follow-up. Longer process (e.g., 1 year).
  • Government involvement from the beginning. Administrators and those who implement policy are included in the debate.
  • Redefinition of the relationships among stakeholders. Emphasis on building partnerships and trust. Increasing coordination and conflict resolution through consensus building.
  • Attitudinal shifts of key stakeholders are reflected in policy changes. Policies are refocused toward poverty.
  • More politically sensitive issues, such as power, decentralization, and gender, are put on the agenda for continuous negotiation.
  • Government ownership and commitment are high.
  • Policymakers are seen as partners who should be included from the beginning of the planning process.
  • World Bank is seen as one of many stakeholders.
  • Policy change is viewed as part of a wider social process.
3. Linking formulation with policy implementation

Policy delivery framework
  • A continuous process of cross-checking and dialogue.
  • Identification of credible institutions for capacity building. Strengthening relationships between formal and informal institutions. Awareness of traditional management practices.
  • Organizational development and institutional change.
  • Ongoing participatory monitoring and evaluation of poverty.
  • Strengthening the policy delivery framework by building the capacity of appropriate institutions, both formal and traditional.
  • Those who implement policy are not just included in the debate but their capacity is also increased.
  • Cross-stakeholder ownership and commitment.
  • Increasing transparency and accountability.
  • Building institutions at microeconomic level contributes to decentralization.
  • Beginning to challenge existing power relations (control by elites, patronage, exclusion of the poor).
  • Policy agenda setting, decision-making, and implementation are interrelated processes. Recognition that policy change does not automatically mean policy implementation and that there is a difference between discourse and outcomes.
  • Participation is seen as more than an addon or a component It is viewed as an approach within which an overall framework is created for more effective policy formulation and implementation.
  • The process of policy change is part of a wider process of establishing linkages between the poor and those in power.

Deepening the Understanding of Poverty and Combining Data Sources

PPAs are deepening our understanding of poverty and contributing to a more in-depth analysis of this complex problem. PPAs are beginning to provide insights into dimensions of poverty, the causes and dynamics of poverty, priorities of the poor, and different levels of analysis.

Dimensions of Poverty

It is well known that poverty has many dimensions beyond income and consumption. However, policy dialogue has focused primarily on income and consumption measures of poverty, while other dimensions highlighted in the PPAs have been underemphasized in the policy debate (see box 5).

Box 5.Enriching the Diagnosis of the Nature of Poverty

The strength of many PPAs has been to highlight the diversity of poor people’s experience of poverty. Definitions of well-being and ill-being vary dramatically depending on many aspects beyond income: a lack of self-respect and dignity; poor health; lack of skills, education, or information; lack of access to assets for a secure livelihood; lack of time; insecurity; lack of freedom of choice; helplessness; exclusion; loneliness; and isolation. These aspects are often interconnected and lead to the disempowerment of poor people.

State institutions and powerlessness

In many PPAs, poor people expressed a desire to have better access to information and to be able to influence decisionmakers. But many PPAs have also highlighted the fact that poor people do not trust state institutions. In The Gambia and Uganda, the poor expressed frustration about their lack of influence on government policies. Ugandans also expressed concern about government corruption and a distrust of state institutions, especially the police and the judiciary. In Vietnam’s PPA, people said they lacked information about their entitlements and rights, and about the activities of local government.


The Zambian PPA was able to distinguish different kinds of female-headed households. “Women without support,” as opposed to female-headed households, were identified as the poorer group. In northern Mexico, the PPA found that it was easier for women than for men to obtain jobs. This was challenging the traditional gender roles, as many men found themselves out of work. Conflict within the household had become a major issue. Similarly, in Mexico City, the urban poor, especially the men, felt excluded from job opportunities. Some men were turning to alcohol. Women were left with the double burden of income earning and child rearing, which put pressure on traditional gender roles and fueled the increase in domestic violence. In Mongolia, alcoholism, crime, and domestic violence were seen as symptoms of poverty, particularly by women. Some PPAs have focused on the informal sector and nonremunerated activities, which in many cases, represent a major part of women’s lives.

Social ill-being and exclusion

Many PPAs have shed light on social ill-being, which includes isolation and exclusion. Excluded groups are not always identified in household surveys. Neither is the fact that their access to productive resources might be constrained by political, cultural, and social factors. In Zambia, the PPA highlighted the fact that children were increasingly going into prostitution and that child-headed households were becoming more common. In Guatemala, the PPA data showed that alcoholism was a major problem for men in the indigenous areas. This had not previously been acknowledged. Most of these men were unemployed or underemployed and felt excluded from the limited opportunities for employment. In Armenia, single pensioners were consistently ranked by communities as the poorest—not because they had the least income but because they were isolated and socially excluded. In Eastern Europe, the PPAs’ analysis of social connections revealed that the poor tend to be connected horizontally—that is, within their own networks—for survival and to reduce vulnerability. As a result, poor households tend to remain excluded. In contrast, the better-off households tend to be connected both horizontally and vertically—that is, to better-off networks—which enabled them to improve their situation. In Mongolia, the PPAs highlighted the fact that there was a weakening of kinship networks but that the most vulnerable were those excluded from kinship and other social networks.

Illegal activities, crime, and violence

Household surveys often are not able to access information on illegal activities because of the reluctance of the respondent to answer questions from an enumerator she or he does not trust. PPAs, however, have been able to shed light on the relationship between poverty and illegal activities. For example, the PPAs in Zambia and Jamaica revealed that prostitution, crime, and violence were major concerns among the poor. People were feeling increasingly scared, unsafe, and insecure as community coherence was threatened because of violence. In some communities, women and the elderly were reluctant to use public transport, particularly at night, because of safety concerns. In Mongolia, Thailand, and Cambodia, economic stress had forced some poor people into such degrading or illegal activities as begging or theft.


In many PPAs, seasonality analysis highlighted great differences in poverty, vulnerability, and coping strategies throughout the year. In South Africa, for example, the PPA revealed that payment of school fees coincided with a season of financial stress resulting from a high incidence of sickness and hard work combined with shortages of money and food. The household survey in Tanzania concluded that 22 percent of the poor had access to safe water from protected sources, indoor plumbing, standpipes, and covered wells with hand pumps. But the survey overlooked the seasonal dimension of access to safe water and therefore overestimated the access. The PPA, which collected information from the same villages, revealed that in two-thirds of the villages thought to have access to safe water, access was actually a major problem. In the dry season, as water tables fell, people were forced to walk farther for water or switch to such unsafe alternatives as uncovered dug wells, ponds, streams, and rivers

PPAs also portray the reality of poor people’s lives. In Equatorial Guinea the results of the PPA highlighted the feelings of hopelessness and despair many people felt after years of declining well-being and repression. Suicide—not generally considered an issue in Africa—was mentioned as a problem. The results of this PPA were described by one Bank economist as “terrifying.”

In the postconflict countries of Burundi, Rwanda, and Uganda, it was not possible to undertake household surveys. In these countries, the PPA proved to be a very useful tool in providing initial data on poverty and conflict impacts.

Explanatory Power of the PPAs

Many PPAs provide insights into the dynamics and processes of impoverishment (see box 6).

Priorities of the Poor

Although some problems highlighted by the PPAs were already known, the PPAs have made it clear that the poor have the capacity to analyze the causes of their vulnerability and rank their priorities. In PPAs carried out in Ghana, Mali, and Nigeria, for example, the poor said that physical isolation and lack of access to water were problems. In Costa Rica, the PPA highlighted linkages between home ownership and status in society.

Box 6.Examples of Explanations Provided by PPAs

Why some women were not working in Mexico City: The PPA in Mexico found that some women in Mexico City were unwilling to leave their houses and go to work. Because they did not have tenancy rights, they were afraid that their houses might become occupied in their absence.

Why the poor spend a “disproportionate” amount of money on clothing in Mali: The results of the quantitative survey showed that a disproportionate amount of money was spent on clothing. The PPA explained that clothing and cloth are investment items in addition to being status symbols and therefore play an insurance role.

Why migrants with money still lack access to land in Zambia: The PPA explained that the social status of certain groups sometimes determines their economic status. Migrant groups might lack access to high-value land not because they lack money but because they lack entitlement in the view of local social institutions that determine land ownership. Lack of social status therefore prevents migrants from gaining title because social institutions actively prevent transfers of land.

Why people were not using health facilities in Kenya and Pakistan: The PPAs explained that communities were discouraged from using health facilities because health staff were often rude and condescending.

Beyond the Household Unit: Combining Data Sources

Household surveys often interview only the head of household (usually a man). PPAs typically gather information on intrahousehold issues from more than one perspective, and also explore interhousehold and community-level social issues in addition to gathering household data. PPAs have focused on individual case studies of people, providing insights into the dynamics of poverty and survival strategies; intrahousehold dynamics, revealing both the unequal allocation of resources among household members and the impact of power relations on the poverty of women, men, and children within the household; interhousehold dynamics, illustrating, for example, the fact that female-headed households might rely on interhousehold transfers; household-level information; and a community perspective highlighting the diversity of social or cultural groups and their wide-ranging coping strategies.

Local people’s understanding of their poverty can be increased if the PPA—especially if it includes a PRA—involves the community in the analysis. In Zambia, one participant in the yearly participatory poverty monitoring stated that the research had enabled the people in the community to get together to discuss their problems and reflect on their situation, that of their neighbors, and the community as a whole. Owen (1997) adds that by using PRAs, the PPA in Mozambique encouraged communities to become conscious of their life conditions, opportunities, strengths, and limitations. This, he says, is particularly important where government has limited capacity to assist people in many areas of the country.

In summary, PPAs are deepening the understanding and providing a dynamic picture of poverty. For example, all of the following insights have emerged from Zambia’s PPA: child-headed households, child labor, crime, violence, and prostitution as coping strategies; increased feelings of insecurity and lack of safety as an outcome of these strategies; seasonal fluctuations in sickness, rates of work, and access to food as triggers of greater vulnerability; and the impact of these new dimensions on people’s behavior as individuals, as household members, and as part of a community.

Attitudes and Policy Change

Formulation of more appropriate and poverty-focused policies can be constrained when Bank staff and government officials involved in the policy dialogue have different attitudes. Some governments have little immediate political incentive to help the poor because the poor are often not organized, have a weak voice, and are difficult to reach. And most Bank staff have little direct experience with poverty.

Some PPAs have helped to change the attitudes of both Bank staff and senior government policymakers, thereby contributing to policy formulation. It is rarely possible to establish clear causality between the PPA and policy change because policymaking is part of a wider social process. In addition, it is usually difficult to separate the impact of the PPA from that of the poverty assessment. However, some indicative evidence is presented below.

World Bank

Insights arising from the PPAs are contributing to the broader debates within the World Bank on how to measure and monitor poverty, integrate social dimensions into policy and project work, and increase the impact of the Bank’s operations by adopting participatory approaches. There is a growing realization of the value of integrating quantitative and qualitative data in the analysis of poverty, in order to produce better measurement, better analysis, and, through more appropriate policy recommendations, better action (Carvalho and White 1997; Chung 2000).

Influencing the World Bank lending program

In certain instances, PPAs have successfully contributed to a shift of policy emphasis. In Nigeria, for example, the World Bank had been focusing on health and education, yet the PPA highlighted that the poor viewed water and roads as the priorities. There is now a greater focus on water and roads. In Ghana, the PPA contributed to a shift of emphasis within the Bank to rural infrastructure and the quality and accessibility of education and health care (see Norton 1996), which was subsequently followed by the preparation of the Village Infrastructure Project.

In Ecuador, the PPA highlighted the fact that women were reluctant to work away from home because of street crime and violence. The poverty assessment identified the provision of street lighting and guarded public buses in the evening as effective ways to address this problem. In Zambia, the World Bank’s Social Fund supported some of the priorities identified by the communities in the PPA, and a health project now includes cost recovery conditions as identified in the PPA. In Niger, the PPA influenced the design of the proposed Infrastructure Project to be more poverty focused and include pilot rural operations. And in Burundi, the Bank designed a community-based poverty project, which used the recommendations of the PPA.

In other cases, the PPA impact has been less evident. The poverty assessment in Kenya reflected the major findings of the PPA, but the results have not been extensively incorporated into other country reports. In Costa Rica, delays in the analysis and dissemination of the findings have limited the impact of the PPA.

Rapid assessments using PPA approaches were used to better understand the social impacts of the financial crisis in East Asia. Initial surveys were undertaken in Thailand, the Philippines, Indonesia, Cambodia, and Laos.1 The objective was to consult with a cross-section of organizations—including community groups, local and international NGO networks, academic institutions, labor unions, professional associations, other donors, and government departments—to determine shifting patterns of vulnerability. Focus groups, rapid assessment techniques, and participatory exercises were used. These initial assessments contributed to creating the framework to begin a dialogue with governments and other donors and jointly formulate strategies for action (see appendix B).

In all these countries, there was a time lag in obtaining reliable statistical data. The advantage of the initial rapid assessments was to quickly produce a series of hypotheses about the potential impacts of the financial crisis on the poor. The participatory surveys highlighted the new social dimensions of the crisis beyond unemployment price changes, access to services, and the credit crunch, to include intrahousehold dynamics, coping strategies, and social capital. These data now form the basis for further ongoing problem identification with the objective of providing a baseline and defining the issues for more detailed, systematic, and representative participatory surveys. For example, this next step was taken in Thailand, where a national participatory assessment, using the PPA approach, was designed as part of the Bank’s Social Investment Project. The objectives of this assessment were to, first, increase the Bank’s and the country’s understanding of the shifting patterns of vulnerability as the impacts of the crisis deepen; second, inform Bank and government policymakers and therefore influence policy; and third, strengthen the capacity in country to undertake participatory surveys and to analyze results, particularly by consolidating the results of participatory and traditional surveys. In addition, the ADB worked with the governments of Lao and the Philippines to undertake more detailed participatory assessments on the impacts of the crisis.

Links to World Bank policy documents

The results of some PPAs have been reflected in Bank policy documents. An example is the Bank’s “Taking Action for Poverty Reduction in Sub-Saharan Africa” (World Bank 1996i), the product of the Bank’s Africa Region Poverty Task Force that resulted from extensive dialogue with development partners. The task force was established to assess the Bank’s operations in the Africa Region, and the report is now being used as a basis for the Bank’s strategy in Africa. In addition, a series of PPAs in Zambia, Mali, Ghana, and Nigeria identified both physical isolation and lack of access to water as major concerns. As a consequence, it was recognized that rural water and roads infrastructure had been neglected areas of investment for the Bank. The report recommended that these should be priority areas in the future. In Gabon, the results of the PPA influenced the Bank’s decisions to undertake a Public Expenditure Review in the health and education sectors.

Links to the country assistance strategy (CAS) are difficult to determine at this stage. However, five examples in which the CAS was clearly influenced by the PPA and the poverty assessment are Armenia, Zambia, Niger, Ethiopia, and Rwanda. In Armenia, the PPA highlighted the importance people place on health and education. The CAS emphasizes protecting access to these sectors. In Zambia, the PPA highlighted the limited access the poor have to public services. The CAS has made this a central theme. In Niger, the value added of the participatory process for the poverty assessment and the PPA was recognized and was adopted for the Niger CAS. In Ethiopia, the results of the quantitative survey were delayed because of data problems. The CAS for Ethiopia therefore drew extensively on the data from the PPA. In Rwanda, the PPA results fed into the CAS, the agriculture strategy note, and an agricultural learning and innovation loan.

The importance of using poverty assessment to focus the CAS on the Bank’s overarching objective of poverty reduction is now widely recognized in the Bank. Thus, in order to more effectively focus the CAS and country programs on poverty, the Bank is shifting from mandatory onetime poverty assessments to long-term strategic poverty monitoring that combines periodic household surveys with periodic participatory research.

The impacts of some PPAs on poverty reduction strategies are beginning to emerge. For example, PPAs are being used to consult with the poor as part of the process of participation, in addition to using the results of the PPA to diagnose poverty Chapter 4 discusses this in more detail.

National Level

Attitudinal change starts with appreciating the value of how the poor perceive their situation. In Tanzania, the government was initially cautious about the PPA exercise but became more receptive when the PPA highlighted the capacity of local people to analyze their own problems. Policymakers began to understand the value of including the poor in the policy dialogue. Similarly, in Benin the PPA strengthened the interest of the Ministry of Planning in consulting the poor through a participatory assessment.

In Zambia, the government was influenced by the priorities expressed by the poor in a ranking exercise. The Ministry of Health has been using the results of the PPA and the poverty assessment to develop policy. In other countries, the poverty assessment and PPA have opened up the policy debate, enabling discussions of once highly sensitive issues. In Swaziland, the workshop convened in February 1997 to discuss the results of the PPA was the first government-sponsored workshop on national poverty. Key insights from the PPA on such issues as women’s rights, land tenure, and the role of traditional authority were given higher priority in the policy agenda as a result of this workshop and the dialogue surrounding the PPA. In Lesotho, three key themes emerged from the PPA that were not identified in the household surveys: alcoholism, political factors, and corruption. Through the government’s action plan, these issues were placed on the policy agenda. As the process has developed in Lesotho, government ownership has increased, and the topic of corruption has now appeared in speeches and policy discussion documents.

Opening up the policy debate can be a conflict-ridden process. The results of both the PPA and poverty assessment were a shock to people in Cameroon inside and outside the government, as poverty had not previously been acknowledged as a serious problem. Ownership had not been developed among key policymakers during the PPA process, because the central government was not strongly committed to poverty reduction. As a consequence, there was little initial acceptance or use of the results. The poverty assessment and PPA did seem to have an impact at the local government level, where officials expressed a great deal of interest in replicating the methodology of the PPA.

Strengthening the Capacity to Deliver Policy

The impact of a PPA on strengthening the capacity to deliver poverty-focused policies can be assessed by identifying new institutional alignments and partnerships that arise as a result of the PPA. Increased dialogue and consequent partnerships can also contribute to widening the constituency for reform, increasing ownership, and strengthening the commitment to poverty reduction.

World Bank

The extent to which PPAs have had an impact on the Bank’s capacity to fulfill its poverty reduction mandate is difficult to determine at this early stage. As stated above, the links between the PPA and poverty assessment, the lending program, and the CAS are evident in only a few countries. The Bank is now developing an interdisciplinary approach to the diagnosis of poverty and the analysis of how all types of institutions affect the poor. This approach will lead to a better understanding of the problem of poverty and increase the Bank’s capacity to work with the relevant institutions. In addition, the Bank is in the process of ensuring that the findings of the WDR 2000/2001 influence its policies and projects, with the objective of increasing empowerment and security for the poor.

National Level

In some countries the process of compiling the PPA has helped to create a dialogue and partnerships for policy delivery. One of the strengths of the Mozambique PPA was the diversity of the involved institutions (university, government, NGOs) and researchers. The multi-institutional approach has strengthened relationships between the various participating institutions (Owen 1997). In Argentina, increased coordination between government agencies and programs has been developed. In addition, dissemination of the results of the PPA has validated the methodology and contributed to the development of an integrated (qualitative and quantitative) approach to monitoring and evaluating social programs undertaken by different organizations.

In other countries, the PPAs have increased the capacity of certain civil society institutions as well as government. In Cameroon, the Bank manager of the PPA stated in an internal communication to team members that “involving local institutions and holding workshops with both government and civil society are mechanisms for expanding ownership of the poverty problem and in-country capacity to analyze and address it.” In some cases, the researchers and intermediary institutions that undertook the PPA were empowered by the process. In South Africa, for example, the local researchers later adopted an activist role. In Ghana, the capacity of local organizations to undertake credible participatory research has been developed, with the local NGO, Centre for Development of People, benefiting from extensive training and institutional linkages created by the PPA process. In Uganda, the PPA impact has even extended to the budget (see box 7).

Through dialogue at the community level, communities that are no longer passive recipients of a policy might become more committed to policy delivery. In some communities, PRAs resulted in local people identifying their priorities, which were later followed up in the form of projects supported by various agencies. In South Africa, for example, the PRA work became a catalyst for communities to initiate a project to benefit the poor. The impact of PPAs has been limited where follow-up has not been extensive, leading many to question the value to the poor of such work. Indeed, at the workshop in Mozambique organized for this study, participants explained that many communities had “respondent fatigue—fadiga dos informantes.” The workshop concluded that many communities, especially those accessible from major cities, are the subject of excessive research, and “agreed…that before initiating any study, a review be undertaken of existing data and material pertaining to the area” (Owen 1997).

Determinants of the Level of Impact


Of the 43 PPAs reviewed for this study, 21 (the ones with sufficient data) were analyzed in more detail to quantify the level of impact and take the first step in exploring the effects of a variety of possible explanatory variables.2 From the data in appendixes A and B, a list of key impact variables was identified. For each variable, a rating of high, medium, low, or zero was assigned to each country PPA on the basis of desk work, discussions with participants, and field research in the five countries the author visited for this study. The ratings are largely subjective; they are not objectively measured indexes of a PPA’s success. In the future, more empirical research will be required. The results of the analysis are summarized in table 6.

Table 6.Summary of PPA Impacts
Impact variableNumber of PPAs

(out of 21 analyzed)
Percentage of PPAs,

by impact
Percentage of

PPAs with


1. Deepening the understanding of poverty0610528482471
2. Influencing policy at the World Bank874233191029
3. Influencing policy at the national level964228191029
4. Increasing country’s capacity to implement policy1333214141024

Significant impact = medium or high rating.

Significant impact = medium or high rating.

Box 7.Can the Poor Influence the Budget? Case Example from Uganda

What is the Medium-Term Expenditure Framework (MTEF)?

  • The Ugandan government adopted an MTEF under which medium-term budget priorities are formulated, consistent with Uganda’s Poverty Eradication Action Plan and medium-term financial stability.
  • Under the MTEF, line ministries are provided global budgetary ceilings on which to base their sector allocations. Sector working groups comprising the Finance Ministry, line ministries, technical advisers, and some civil society representatives develop sector priorities within these limits.
  • Given this resource envelope, a comprehensive review of ail sector spending proposals is undertaken. This requires an analysis of tradeoffs of various funding decisions among and within sectors on the basis of affordability and intersectoral priorities. This new process forces ministries to prioritize early in the budgeting process.
  • This review feeds into the Budget Framework Paper (BFP) and the annual Background to the Budget.

Participation at the central level

  • Since 1998–99, civil society organizations have been involved in the dialogue on priorities and spending commitments, and public debate in the media has been encouraged. However, officials recognize that there is still a long way to go to open up civil society and engage the media.
  • In 1998–99, the government implemented the Uganda Participatory Poverty Assessment Project (the UPPAP), for which the poor in rural and urban areas were directly consulted. The results of the UPPAP influenced budget allocations. For example, a higher weighting was given to provision of safe water supply in central and district-level budgets as a result of communities identifying access to clean water as a priority Findings from the UPPAP were also included in the Background to the Budget 1999–2000.
  • Involvement of civil society is also encouraged by publication of (a) an abbreviated version of the BFP (the version that goes to the cabinet before expenditure allocations are approved); (b) an annual Background to the Budget; and (c) a detailed summary of the composition of expenditure for all sectors for the three-year MTEF, as an appendix table in the Budget Speech document.
  • To broaden consultation in the budget process, donors are invited to join sector working groups and participate in the public expenditure review, where discussions focus on sector priorities of government expenditure and on the consistency of government assumptions regarding external financing and donor financing plans.
  • In the 2000–01 budget process, the Poverty Eradication Working Group was established, comprising Ministry of Finance officials and donor and civil society representatives. The group reviews each sector working paper for a poverty focus. It then makes recommendations on inter and intrasectoral allocations of resources and on poverty-focused output indicators, based on the latest poverty analysis.
  • The budget process also provides an opportunity for parliament to take a more strategic look at the government’s expenditure plans and examine how the government is performing in the implementation of its overall budget strategy.
  • A strategy to increase public awareness and transparency to further the budget process is currently being developed. This involvement will enhance the evolving partnership between civil society and government

Participation at the district level

  • Responsibility for the provision of a large number of services was devolved to district and urban authorities. This increased people’s participation in the decisionmaking process, and made decisions more transparent and public officers more accountable. The long-term aim is to integrate central and local government budgetary processes and devolve spending decisions to local governments to enable them to respond to district poverty priorities.
  • Recent UPPAP findings demonstrate major differences in the poverty profile among districts. This resulted in policymakers’ recognizing the need for flexibility in the use of central government conditional grants to districts.
  • In partnership with the Ministry of Local Government, the UPPAP will work directly with nine districts, initially to strengthen their capacity to consult poor communities for the purposes of district planning and budgeting.

Monitoring budget expenditures

  • The government established a Poverty Action Fund (PAF) to enhance transparency and monitoring of HIPC and other donor resources going to expenditure programs focused on poverty. The PAF involves both civil society and government in monitoring the impact of PAF outlays, and the government holds quarterly meetings to discuss delivery against budget allocations.
  • The budget process is being further developed to open up multiple channels of accountability. For example, to increase transparency in decentralized management of resources, advertisements are placed in the press indicating amounts disbursed to each district by sector. In the education sector, budget allocations for schools are posted on school notice boards.
  • In addition, the Poverty Monitoring Unit integrates annual household surveys, conducted by the Uganda Bureau of Statistics, with other data sources (e.g., participatory analysis, sector surveys, line ministry data sources) to ensure that impacts of policy are understood and policy development is informed by poverty data and the perceptions of the poor.

The analysis of the 21 PPAs suggests a significant influence on the diagnosis of poverty in 71 percent of the cases examined. Twenty-nine percent of the PPAs had a significant impact on policy formulation, both in the Bank and in country, while in only 24 percent of the cases did the PPA have a significant effect on the country’s capacity to deliver some policies.

The analysis involved classifying the PPAs in a 3 × 3 matrix based on the composite impact index (CII) and a variety of possible explanatory variables. The CII used the ratings assigned to four key impact variables. These variables and the weights assigned to them were as follows:

  • DUP—deepening the understanding of poverty (1)
  • IWP—influencing World Bank policy (2)
  • ICP—influencing country policy (2)
  • ICD—increasing capacity to deliver policy (4)

The four rating levels were assigned a score as follows: none = 0; low = 1; medium = 2; high = 3.

For each of the 21 country PPAs, the CII was calculated as follows:

The maximum attainable score was therefore 27, calculated as follows:

The next step was to define a series of independent variables that had potential to explain the level of PPA impact as measured by the CII. For each independent variable, subjective ratings of low, medium, and high were assigned for each country PPA. These results are set out in table 7.

Table 7.Summary of Impact Variables
Impact variableNumber of PPAs (out of 21)



World Bank
1. Ownership within the World Bank (by staff, departments)01065
2. Bank management support01155
3. Links to poverty assessment2856
4. Links to country assistance strategy01902
5. Team work7563
Country level
1. Involvement of policymakers0768
2. Involvement of other stakeholders0894
3. Ownership by government0867
4. Dissemination at the national level0948
Community level
1. Skills of researchers0786
2. Dissemination to communities13620
3. Length of time in field0498
4. Cost0777
5. Follow-up; action with communities9660

Pairs of impact variables were then chosen and charted against the CII. Only high and low ratings were included to highlight the more marked differences observed among the PPAs.

Main Findings

When the CII was plotted against the subjective estimates of the quality of the PPA research team, there was a clear positive correlation (see figure 2). PPAs judged to have high-quality teams averaged more than 15 out of a possible 27 on the CII. Those judged to have medium-quality teams averaged 7, and those with low-quality research teams averaged only 2.

Figure 2.PPA Impact by Quality of Research Team

The relationship between the CII, the quality of the PPA manager, and Bank management support revealed a high level of interaction between the latter two variables (see figure 3). Where both Bank management support and PPA manager quality were high, average CII was high (17 out of a possible 27). However, even high-quality PPA managers were unlikely to produce high-impact PPAs without strong Bank management support—the PPAs in that category scored only 7 out of a possible 27. There were no PPAs with high Bank management support and low manager quality; hence the zero CII score in the lower left corner of figure 3.

Figure 3.FPA Impact by Quality of PPA Manager and Level of Bank Management Support

When the CII was plotted against links to the poverty assessment, there was a clear positive correlation (see figure 4). PPAs judged to have a greater link to the poverty assessment averaged a CII of more than 19 out of a possible 27. Those judged to have medium links averaged 5, and those with limited links averaged only 2. To have a significant impact, PPAs need to be linked to the poverty assessment.

Figure 4.PPA Impact by Link to Poverty Assessment

The relationship between CII, ownership by the government, and ownership within the World Bank revealed a high level of interaction between the latter two explanatory variables (see figure 5). Where there was a high degree of ownership in the Bank and by government, the CII reached 20 out of a possible 27. Ownership by both the Bank and government is important to achieve a high-impact PPA. There was no instance of a PPA with high World Bank ownership and low government ownership; hence the zero CII score in the top right corner of figure 5.

Figure 5.PPA Impact by World Bank and Government Ownership

When the CII was plotted against the extent to which policymakers were involved, there was a clear positive correlation (see figure 6). Where there was a high level of policymaker involvement, the CII was more than 16 out of a possible 27. Those PPAs judged to have medium involvement averaged 7 and those with limited involvement averaged only 3. The level of PPA impact depends to some extent on the level of policymaker involvement.

Figure 6.PFA Impact by Level of Policymaker Involvement

The following chapter builds on this analysis of the key variables in more detail, analyzing case examples to elucidate recommendations for good practice.

1.See Robb (1998) for a summary of these initial surveys.
2.The section titled “Determinants of the Level of impact” was compiled with assistance from James Edgerton of the World Bank’s Social Development Department.

    Other Resources Citing This Publication