NEPAL Poverty Alleviation Fund Project Report No. 116947 JUNE 28, 2017 © 2017 International Bank for Reconstruction This work is a product of the staff of The World RIGHTS AND PERMISSIONS and Development / The World Bank Bank with external contributions. The findings, The material in this work is subject to copyright. 1818 H Street NW interpretations, and conclusions expressed in Because The World Bank encourages Washington DC 20433 this work do not necessarily reflect the views of dissemination of its knowledge, this work may be Telephone: 202-473-1000 The World Bank, its Board of Executive reproduced, in whole or in part, for Internet: www.worldbank.org Directors, or the governments they represent. noncommercial purposes as long as full attribution to this work is given. Attribution—Please cite the work as follows: The World Bank does not guarantee the World Bank. 2017. Nepal—Poverty Alleviation accuracy of the data included in this work. The Any queries on rights and licenses, including Fund Project. Independent Evaluation Group, boundaries, colors, denominations, and other subsidiary rights, should be addressed to Project Performance Assessment Report information shown on any map in this work do World Bank Publications, The World Bank 116947. Washington, DC: World Bank. not imply any judgment on the part of The Group, 1818 H Street NW, Washington, DC World Bank concerning the legal status of any 20433, USA; fax: 202-522-2625; e-mail: territory or the endorsement or acceptance of pubrights@worldbank.org. such boundaries. Report No.: 116947 PROJECT PERFORMANCE ASSESSMENT REPORT NEPAL POVERTY ALLEVIATION FUND I (IDA-H091 IDA-H258) June 28, 2017 Financial, Private Sector, and Sustainable Development Independent Evaluation Group ii Currency Equivalents (annual averages) Currency Unit = Nepalese rupee (NPR) 2003 US$1.00 NPR 76.10 2004 US$1.00 NPR 73.70 2005 US$1.00 NPR 71.40 2006 US$1.00 NPR 72.80 2007 US$1.00 NPR 66.40 2008 US$1.00 NPR 69.80 All dollar amounts are U.S. dollars unless otherwise indicated. Abbreviations CAS country assistance strategy CDD community-driven development ICR Implementation Completion and Results Report IDA International Development Association IEG Independent Evaluation Group M&E monitoring and evaluation O&M operations and maintenance PAD project appraisal document PAF Poverty Alleviation Fund PPAR Project Performance Assessment Report PRSP Poverty Reduction Strategy Paper Fiscal Year Government: July 16–July 15 Director-General, Independent Evaluation: Ms. Caroline Heider Director, Financial, Private Sector, and Sustainable Development Mr. José C. Carbajo Martínez Manager, Sustainable Development: Ms. Midori Makino Task Manager: Ms. Elena Bardasi iii Contents Principal Ratings ................................................................................................................. v Key Staff Responsible......................................................................................................... v Preface............................................................................................................................... vii Summary .......................................................................................................................... viii 1. Background and Context................................................................................................. 1 Project Context................................................................................................................ 1 2. Objectives, Design, and Their Relevance ....................................................................... 3 Project Development Objectives..................................................................................... 3 Project Components ........................................................................................................ 4 Relevance of Objectives ................................................................................................. 5 3. Implementation ............................................................................................................... 6 Project Financing and Key Dates .................................................................................... 6 Implementation Experience ............................................................................................ 6 Safeguards Compliance .................................................................................................. 7 Financial Management and Procurement........................................................................ 8 4. Relevance of Design ....................................................................................................... 8 5. Achievement of the Objectives ..................................................................................... 10 Objective 1: Improving Access to Income-Generating Subprojects............................. 10 Outputs ...................................................................................................................... 10 Outcomes .................................................................................................................. 12 Objective 2: Improving Access to Community Infrastructure ...................................... 14 Outputs ...................................................................................................................... 14 Outcomes .................................................................................................................. 15 6. Efficiency ...................................................................................................................... 17 7. Ratings .......................................................................................................................... 19 Outcome ........................................................................................................................ 19 Risk to Development Outcome ..................................................................................... 20 Sustainability of Income-Generating Subprojects .................................................... 20 World Bank Performance ............................................................................................. 21 Borrower Performance .................................................................................................. 22 This report was prepared by Elena Bardasi and Javier Bronfman, who assessed the project in September 2016. The report was peer reviewed by Mr. Jack W. van Holst Pellekaan and panel reviewed by Christopher Nelson. Carla Fabiola Chalcatana, Dung Thi Kim Chu, and Richard Kraus provided administrative support. iv Monitoring and Evaluation ........................................................................................... 23 8. Lessons .......................................................................................................................... 23 References ......................................................................................................................... 25 Figure Figure 1.1. Nepal Poverty Statistics, 1995–2010 ................................................................ 2 Tables Table 1.1. Selected Human Development Indicators, 1990–2015...................................... 3 Table 5.1. Type of Income-Generating Activity Funded under PAF I ............................. 11 Table 5.2. Distribution of Livestock Activities Funded under PAF I ............................... 11 Table 5.3. Training Received for Income-Generating Activities ..................................... 12 Table 5.4. Changes in Food Security Status since Community Organization Formation ....13 Table 5.5. Changes in Food and Nonfood Consumption between 2011 and 2016 ........... 13 Table 5.6. Perceived Benefits from PAF by 2016 ............................................................ 17 Table 5.7. Women’s Role in Household Decision Making .............................................. 17 Appendixes Appendix A. Basic Data Sheet.......................................................................................... 27 Appendix B. 2016 IEG-Funded Survey and Focus Group Discussions ........................... 30 Appendix C. Methodology and List of Indicators Used to Rank Districts ....................... 32 Appendix D. List of Persons Met ..................................................................................... 34 Appendix E. Borrower Comments .................................................................................... 35 v Principal Ratings ICR* ICR Review* PPAR Outcome Satisfactory Moderately Satisfactory Satisfactory Risk to Moderate Moderate Moderate Development Outcome World Bank Moderately Satisfactory Moderately Satisfactory Moderately Satisfactory Performance Borrower Satisfactory Satisfactory Satisfactory Performance * The Implementation Completion and Results (ICR) report is a self-evaluation by the responsible World Bank global practice. The ICR Review is an intermediate IEG product that seeks to independently validate the findings of the ICR. Key Staff Responsible Task Manager or Division Chief or Project Leader Sector Director Country Director Appraisal Geeta Sethi Constance Kenichi Ohashi Bernard/Adolfo Brizzi Completion Geeta Sethi John H. Stein/Simeon K. Susan G. Goldmark Ehui vi IEG Mission: Improving World Bank Group development results through excellence in independent evaluation. About this Report The Independent Evaluation Group (IEG) assesses the programs and activities of the World Bank for two purposes: first, to ensure the integrity of the World Bank’s self-evaluation process and to verify that the World Bank’s work is producing the expected results, and second, to help develop improved directions, policies, and procedures through the dissemination of lessons drawn from experience. As part of this work, IEG annually assesses 20– 25 percent of the World Bank’s lending operations through fieldwork. In selecting operations for assessment, preference is given to those that are innovative, large, or complex; those that are relevant to upcoming studies or country evaluations; those for which Executive Directors or World Bank management have requested assessments; and those that are likely to generate important lessons. To prepare a Project Performance Assessment Report (PPAR), IEG staff examine project files and other documents, visit the borrowing country to discuss the operation with the government, and other in-country stakeholders, interview World Bank staff and other donor agency staff both at headquarters and in local offices as appropriate, and apply other evaluative methods as needed. Each PPAR is subject to technical peer review, internal IEG panel review, and management approval. Once cleared internally, the PPAR is commented on by the responsible World Bank country management unit. The PPAR is also sent to the borrower for review. IEG incorporates both World Bank and borrower comments as appropriate, and the borrowers’ comments are attached to the document that is sent to the World Bank’s Board of Executive Directors. After an assessment report has been sent to the Board, it is disclosed to the public. About the IEG Rating System for Public Sector Evaluations IEG’s use of multiple evaluation methods offers both rigor and a necessary level of flexibility to adapt to lending instrument, project design, or sectoral approach. IEG evaluators all apply the same basic method to arrive at their project ratings. Following is the definition and rating scale used for each evaluation criterion (additional information is available on the IEG website: http://ieg.worldbankgroup.org). Outcome: The extent to which the operation’s major relevant objectives were achieved, or are expected to be achieved, efficiently. The rating has three dimensions: relevance, efficacy, and efficiency. Relevance includes relevance of objectives and relevance of design. Relevance of objectives is the extent to which the project’s objectives are consistent with the country’s current development priorities and with current World Bank country and sectoral assistance strategies and corporate goals (expressed in poverty reduction strategy papers, Country Assistance Strategies, sector strategy papers, and operational policies). Relevance of design is the extent to which the project’s design is consistent with the stated objectives. Efficacy is the extent to which the project’s objectives were achieved, or are expected to be achieved, taking into account their relative importance. Efficiency is the extent to which the project achieved, or is expected to achieve, a return higher than the opportunity cost of capital and benefits at least cost compared with alternatives. The efficiency dimension is not applied to development policy operations, which provide general budget support. Possible ratings for Outcome: Highly Satisfactory, Satisfactory, Moderately Satisfactory, Moderately Unsatisfactory, Unsatisfactory, Highly Unsatisfactory. Risk to Development Outcome: The risk, at the time of evaluation, that development outcomes (or expected outcomes) will not be maintained (or realized). Possible ratings for Risk to Development Outcome: High, Significant, Moderate, Negligible to Low, and Not Evaluable. World Bank Performance: The extent to which services provided by the World Bank ensured quality at entry of the operation and supported effective implementation through appropriate supervision (including ensuring adequate transition arrangements for regular operation of supported activities after loan or credit closing, toward the achievement of development outcomes). The rating has two dimensions: quality at entry and quality of supervision. Possible Ratings for World Bank Performance: Highly Satisfactory, Satisfactory, Moderately Satisfactory, Moderately Unsatisfactory, Unsatisfactory, and Highly Unsatisfactory. Borrower Performance: The extent to which the borrower (including the government and implementing agency or agencies) ensured quality of preparation and implementation, and complied with covenants and agreements, toward the achievement of development outcomes. The rating has two dimensions: government performance and implementing agency(ies) performance. Possible ratings for Borrower Performance: Highly Satisfactory, Satisfactory, Moderately Satisfactory, Moderately Unsatisfactory, Unsatisfactory, and Highly Unsatisfactory. vii Preface This is the Project Performance Assessment Report (PPAR) by the Independent Evaluation Group (IEG) of the World Bank Group on the Poverty Alleviation Fund I project in Nepal. This report presents findings based largely on a two-week mission to Nepal from September 5, 2016, to September 17, 2016, by IEG. Additionally, in 2016, IEG collected data through focus group discussions and a household survey (appendix B). The findings presented in the report are based on an assessment of (i) World Bank project files; (ii) project-related reporting documents and evaluations, including data collected by the World Bank and Poverty Alleviation Fund staff; (iii) studies produced by government, the Center for Economic Development and Administration at Tribhuvan University, one impact evaluation (Parajuli et al. 2012); and (iv) the analysis of the focus group discussion and survey data findings. In Kathmandu, the mission met authorities from the Ministry of Poverty Alleviation and Cooperatives, the Ministry of Agriculture, the Office of the Prime Minister, with past and present functionaries of the Poverty Alleviation Fund (the implementing agency) including the chief and the monitoring and evaluation officer. The mission also visited several Poverty Alleviation Fund I beneficiary villages and partner organizations in three regions (Rasuwa, Ramechhap, and Mahottari). The assistance and contributions of all stakeholders, including World Bank staff in Washington, D.C., and Kathmandu, and the translator are gratefully acknowledged. The list of persons met is in appendix D. Following standard IEG procedures, the draft PPAR will be shared with relevant government officials and agencies for their review and comments, which will be presented in appendix E. viii Summary This is the Project Performance Assessment Report for the Poverty Alleviation Fund I (PAF I; P081968; including additional financing) in Nepal. PAF I was approved by the World Bank’s Board of Executive Directors on June 1, 2004. The total project cost at appraisal was $16.60 million ($15 million in International Development Association (IDA) grant, and the rest to be funded by contributions of the Government of Nepal and community organizations). The actual project cost at project closing was $48.88 million, inclusive of additional financing of $25 million in IDA grant, approved on November 14, 2006. The project’s development objective was “to carry out demand- and community-driven approach to improve access to income-generating subprojects and community infrastructure, benefiting groups excluded by reasons of gender, ethnicity and caste, as well as the poorest groups in rural communities” (World Bank 2004b, 13). The relevance of the project development objective is rated high. Reducing poverty in rural areas, especially among communities historically excluded and disenfranchised, was a top priority of the Government of Nepal at the beginning and during the life of the project. The objectives were well justified by the country’s postconflict conditions and aligned with Nepal’s Poverty Reduction Strategy, which focused on the social inclusion of marginalized groups such as women, Dalit, and ethnic minorities. The development objectives were also consistent with the World Bank’s 2003 country assistance strategy that aimed at addressing social exclusion and rural poverty in Nepal. The relevance of the project design is rated substantial. Building a new independent institution to work in parallel with the government and using a community-driven development (CDD) design was justified by the need to transfer resources to those areas that the government was unable to reach (because of weak local administrative structures, resource limitations, and high levels of violence). The CDD approach was well conceived and implemented; communities were able to prioritize beneficiaries and select projects under a participatory and inclusive scheme. The project’s design was consistent with the development objectives, and the theory of change (reflected in the project’s activities) was logical although the project appraisal document lacked an explicit results framework. At appraisal stage, it was expected that 70 percent of the total funds would be used for community infrastructure projects (for example, classrooms, health centers, irrigation, and water and sanitation subprojects) and the rest for income-generating activities (for example, livestock, trade, and agriculture and forest activities), Though such expectations were not met, the project reached the target population, including traditionally excluded groups, and provided resources and services to implement subprojects chosen by the community. The successful outreach of the project was largely owing to the choice of nongovernmental organizations as the partner organizations to reach communities. Moreover, funds were not disbursed through government channels but delivered directly to the community organizations, which lowered transaction costs between the resource origination and its destination, with positive results for efficacy and efficiency. Another strength of PAF I’s initial design was its phased approach. A successful initial pilot conducted in six districts was scaled up ix after additional financing to another 19 districts became available, permitting the development of PAF institutional capacity and the gathering of lessons from the pilot phase. Efficacy of the first objective (improved access to income-generating subprojects) is rated substantial. The project delivered many income-generating subprojects (especially livestock subprojects), twice as many as the target established at the time of additional financing. Moreover, based on the findings of an impact evaluation prepared in 2012, the subprojects generated positive impacts on the well-being of the targeted communities in terms of food security, consumption expenditure, employment, and child education outcomes. A drawback the project experienced was insufficient support to income- generating activities related to training and facilitating access to markets. Efficacy of the second objective (improved access to community infrastructure) is also rated substantial. The number of infrastructure subprojects was well below the initial expectations as a reflection of lower-than-expected demand from communities; however, all requests for infrastructure subprojects were met. The impact evaluation documents reflect the positive impact on enrollment towing to education infrastructure; but not much is known of impacts on health, earnings, and productivity. Both objectives (improving access to income-generating activities and improving access to community infrastructure) were achieved also in terms of the project outreach. PAF I was able to benefit about 15 percent of the population in the 25 poorer districts of Nepal. The project exceeded its targets for participation of women and disenfranchised groups (Dalit and Janajati) in project activities. Lower-caste groups also benefited from improved consumption and food security. Efficiency is rated substantial. The empirical evidence, albeit limited, shows that a sample of income-generating projects produced high financial rates of return, and that the costs of community infrastructure were lower than the costs of comparable government- financed infrastructure. The impact evaluation showed that the benefits— increased consumption levels and food security achieved by the project—were larger than the costs incurred to generate them. However, there was no comprehensive estimation of the return on investments, which could have helped toward better understanding the project’s overall economic impact and provided an indication of the success of income-generating activities. Finally, that project management costs were only 3.6 percent of total project costs is a measure of the project’s overall cost effectiveness. The development outcome rating is satisfactory, based on high relevance of the project objectives, substantial relevance of the project design, substantial efficacy of both the first and the second objective, and substantial efficiency. The risk to development outcome is rated moderate. The survival of PAF I over 12 years is evidence of its sustainability as an independent institution, and evidence of the project’s sustainability. The project was well established at closure and supported by the government and communities. This initial project was followed by a much larger PAF II and additional financing that supported program outcomes and expanded it to reach x almost all districts in Nepal. PAF II is currently coming to closure, and there are discussions about a potential PAF III. Lessons The CDD approach can provide a powerful organizing framework for improving the welfare of communities. In the PAF, community organizations were provided with opportunities to make competitive proposals for broad CDD programs. When provided with grants, these community organizations were entrusted with managing the allocation of funds either among members for the best subprojects on a participative basis or for community infrastructure. As this PPAR has concluded, the CDD framework generated considerable social capital in poor and excluded communities. The sustainability of CDD programs hinges on finding a wise balance between program “agility”—achieved through independence of the implementing agency— and government ownership and coordination with line ministries. Originally, PAF was created as an independent agency because this was the best (or only) option to maximize project outreach and minimize leakages of funds. At the time, the government was unable to reach more remote areas—owing to weak local administrative structures, human resource limitations, and the high level of violence in some areas. In a more stable political environment, however, the independence of the implementing agency may undermine the role and coordination function of line ministries, which need to ensure that the program works in synergy with other development interventions. Though the project needs to achieve more direct collaboration and coordination with line ministries and local governments, its efficacy and efficiency must be maintained too. A sound assessment of community needs is essential for CDD programs. Even though the PAF design benefited from lessons of preexisting Asian CDD experiences, the design of the project did not initially reflect the actual demand of the recipient population. The target communities prioritized income-generating activities over community infrastructure as a more direct way to improve family consumption and food security. Explicitly acknowledging and addressing diversity within communities can help achieve substantial results for marginalized groups. PAF recognized the importance of reaching out to women and lower castes; many partner organizations used targets, quotas, and incentives to ensure participation of these traditionally excluded groups. As a result, a consensus among beneficiary communities increasing female participation was one of the main benefits of PAF. Moreover, Dalit and Janajati who were part of community organizations were able to achieve food security at the same rate as higher castes. Income-generating subprojects require complementary activities, such as specific training and help in accessing markets, which need to be embedded in project design to ensure success. The data collected by IEG for this PPAR indicate that a large proportion of beneficiaries of income-generating subprojects lamented lack of training and poor access to markets as drawbacks for the success of their activities. xi A carefully planned and implemented pilot phase for a project provides essential learning for a successful scaling up. Piloting of the PAF in six districts initially, then scaling up to the poorest 25 districts, helped to fine-tune some key features of the project, such as moving from providing grants for subprojects to the creation of revolving funds, and identifying weaknesses and potential improvements in delivering resources to communities. Robust data collection and effective analysis of available data are crucial to modifying the project design and to identifying progress and results. PAF conducted several household surveys and collected valuable data on outputs and outcomes during the life cycle of the project. Nonetheless, some variables were not collected, and those available were underutilized to produce evidence for decision making on the project’s design. The project could have planned for more extensive monitoring and evaluation on some aspects such as the project’s efficiency, and its impact on employment, health, earnings, and productivity. Mr. José Cándido Carbajo Martínez Director, Financial, Private Sector, and Sustainable Development Evaluation Independent Evaluation Group 1 1. Background and Context 1.1 This is a Project Performance Assessment Report (PPAR) of the Poverty Alleviation Fund (PAF) Project approved by the World Bank’s Board of Executive Directors on June 1, 2004. Total estimated project costs at appraisal were $16.60 million, comprising $15 million in International Development Association (IDA) grant, and $1.6 million to be funded by contributions from the Government of Nepal and community organizations. Additional financing was approved on November 11, 2006, for $25 million of IDA grant. Actual project costs at project closing were $48.88 million, including the two IDA grants of $15.22 million and $26.22 million, and $7.44 million in contributions from the government and community organizations (World Bank 2009, 34). The project closed on February 1, 2009. 1.2 Nepal is a small landlocked country in South Asia, with a population of approximately 28.5 million. In 2008, after a long conflict, the first Nepali Constituent Assembly ruled in favor of abolishing the monarchy. Since then, Nepal has become a federal multiparty representative democratic republic. The new political structure encountered significant and continuing challenges. The new constitution was promulgated with a resounding majority by the Constituent Assembly on September 20, 2015, but it fueled agitation and violence among aggrieved population groups, most notably the Madhesis, during the last months of 2015 until the spring of 2016. 1.3 Nepal’s PAF is one of several longstanding World Bank–assisted community-driven development (CDD) projects. The project was prepared in 2004, when Nepal was facing serious political turmoil ensuing from violent conflict between the government and the Maoist insurgency. The historic context under which the PAF was conceived influenced its design. In a situation of state fragility and ongoing conflict, the PAF was conceived as an independent body that could reach rural communities more easily than the government administration, through partnership with nongovernmental organizations active in remote areas.1 The assumption was that improving the situation of rural areas could help deflate the escalating insurgency. 1.4 Although caste-based discrimination was abolished officially in Nepal in 1963 and democracy was established in 1990, discrimination and lack of inclusion were key elements of concern at project approval and are still pervasive to this day (World Bank and DFID 2004). Women, the formerly “untouchable” castes now called Dalit, the ethnic groups or Janajati, the Muslims, and the plains dwellers or Madhesis, were recognized as facing multiple deprivations and social biases and therefore in need of direct intervention. Project Context 1.5 At the time of appraisal, Nepal was one of the poorest countries in the world, and the poorest in South Asia, with a 46.2 percent poverty headcount (figure 1.1) and a per capita gross national income of $260 in 2003.2 World Bank diagnostic work (World Bank and DFID 2004) revealed that Nepal’s most important development challenges were (i) high levels of poverty; (ii) slow economic growth; (iii) social exclusion; (iv) increasing inequality; and (v) poor governance. 2 1.6 Poverty in Nepal is essentially a rural phenomenon. At appraisal about 85 percent of the total poor lived in rural areas. The poor in these rural and remote areas depended predominantly on agricultural production, making them particularly vulnerable to climate shocks and dependent on market access, thus limiting their opportunities for income generation and access to social services. It is therefore not surprising that development indicators reveal marked differences across geographic regions in access to education and basic welfare services. 1.7 Though still high, according to the latest data poverty and extreme poverty have decreased significantly in Nepal. Still, in 2010 poverty was as high as 50 percent based on the $3.10 per capita per day international poverty line; extreme poverty was 15 percent (figure 1.1). Figure 1.1. Nepal Poverty Statistics, 1995–2010 (percent) 90 85.5 80 73.8 70 Poverty Headcount ratio at $3.10 60 dollars a day per 61.9 person (2011 PPP) 50 48.4 40 46.1 30 Poverty Headcount ratio at $1.90 20 dollars a day per person (2011 PPP) 10 15.0 0 1995 2003 2010 Source: World Bank 2016. Note: PPP = purchasing power parity. 1.8 As poverty decreased, other human development indicators show some improvements over the past few decades (table 1.1). Primary school enrollment almost reached 100 percent for both female and male students by 2012. Child and infant mortality rates decreased substantially, surpassing the Millennium Development Goal target number 4 of Reducing Child Mortality by two-thirds between 1990 and 2015. Access to improved water sources increased in rural but decreased in urban areas (probably because of in-country migrations). There was some progress in reducing gender disparities—especially in life expectancy, literacy levels, and school completion rates. 3 Table 1.1. Selected Human Development Indicators, 1990–2015 Source: World Bank 2016. 1.9 Both the World Bank’s country assistance strategy (CAS) 2004–07 and the 10th Poverty Reduction Strategy Paper (PRSP) 2002–07 envisaged the PAF as the vehicle to address rural poverty and reduce gender, ethnic, and caste related disparities, by ensuring equity of access to resources.3 The CAS highlighted the progress in reducing gender disparities, but acknowledged that additional efforts were required for Nepali women to be included in the development process.4 Similarly, the PRSP identified the high level of population stratification as a crucial challenge, and highlighted the discrimination and multiple dimensions of economic hardships faced by women and groups such as Dalit and Janajati living in remote areas. 1.10 Facing the challenges discussed above, the proposed solution was the creation of the PAF, a CDD project based on a bottom-up approach that would place the most vulnerable segments of the population in a leading role in setting their own development priorities, managing resources, and implementing their selected subprojects. 1.11 In the PRSP, the government called for direct and quick interventions targeted specifically to help these marginalized groups. Including them in the development process was identified as important not only to fill their multiple development gaps, but also to help cut down the escalating insurgency in rural areas, as it was stressed in several interviews. 2. Objectives, Design, and Their Relevance Project Development Objectives 2.1 The PAF project is a World Bank–supported project established in 2004. The project aimed to bring traditionally excluded communities into the mainstream of development, by providing poor and disadvantaged groups with direct access to resources. 2.2 Approved on June 1, 2004, the project appraisal document (PAD) stated the development objectives as follows: “The Project’s development objective is ‘to support His 4 Majesty’s Government of Nepal in implementing a new, targeted instrument—the Poverty Alleviation Fund—for reaching poor and excluded communities.’” It aims to improve access to income-generating projects and community infrastructure for the groups that have tended to be excluded by reasons of gender, ethnicity, and caste, as well as for the poorest groups in rural communities.” The PAD also states that “Unlike the past targeted interventions by [His Majesty’s Government of Nepal], the PAF will be based on a demand-driven and community-centered approach, implemented by an autonomous body in close collaboration with other nongovernmental groups as well as local bodies.” (World Bank 2004a, 2). 2.3 The development grant agreement states the development objectives more succinctly: “The objective of the Project is to carry out demand- and community-driven approach to improve access to income-generating subprojects and community infrastructure, benefiting groups excluded by reasons of gender, ethnicity and caste, as well as the poorest groups in rural communities” (World Bank 2004b, 13). 2.4 The 2006 additional financing paper that provided an additional financing of $25 million did not change the development objectives. 2.5 The PAD and the development grant agreement objectives are essentially the same: improving access to income-generating activities and community infrastructure for socially excluded groups by using a demand-driven approach. The PPAR will be centered on the development grant agreement objectives in its assessment of the project’s achievements. Project Components 2.6 As described in the PAD, PAF I originally included five components: (i) income generation subprojects targeted to the poorest and excluded groups; (ii) small-scale village and community infrastructure; (iii) innovation and special programs; (iv) capacity building; and (v) administration of the PAF. 2.7 Component 1: Income generation subprojects targeted to the poorest and excluded groups (Project costs: appraisal $4.27 million; actual $25.18 million). This component was to provide grants for income-generating activities, selected by the community, to groups of poor and excluded people. Beneficiaries would be expected to contribute 10 percent of the subproject cost in cash. A proportion of the grant was reserved for technical assistance. Under this component, partner organizations would be contracted to assist the communities in establishing participatory processes, forming groups, and developing sound subproject proposals to get funded. 2.8 Component 2: Small-scale village and community infrastructure (Project costs: appraisal $9.28 million; actual $9.09 million). This component was meant to provide grants for small-scale infrastructure subprojects, selected by the community from an open menu, to groups of poor and excluded people. Partner organizations were to facilitate the process of preparation of subprojects. Communities were expected to contribute 10 percent of the subproject cost and prepare and commit to a credible Operation and Maintenance (O&M) plan. Priority was to be given to subprojects already included in Village Development 5 Committee plans and/or where the District or Village Development Committees would provide matching funds. 2.9 Component 3: Innovation and special programs (Project costs: appraisal $0.84 million; actual $11.02 million, of which $8.65 million financed income-generating activities). This component was meant to support innovative proposals that might have arisen during implementation, to address poverty, reach target groups, and finance subprojects outside the six pilot districts. 2.10 Component 4: Capacity building (Project costs: appraisal $1.41 million; actual $1.82 million). This component was meant to support capacity building and social mobilization in five areas: (i) capacity of partner organizations to mobilize target beneficiaries and develop subprojects; (ii) capacity building for Village Development Committees in financial management, budgeting, social assessments, and evaluation of impacts, and to enable local bodies to help targeted groups to access PAF resources; (iii) building target groups’ capacity to implement their subprojects; (iv) support to microfinance intermediaries to increase their lending capacity to target communities; and (v) support to information, monitoring, and evaluation of progress, processes, and results of the program. 2.11 Component 5: Administration of the PAF (Project costs: appraisal $0.80 million; actual: $1.77 million). The final component was meant to finance PAF administration and covers various operational costs and human resource development activities for PAF staff. Relevance of Objectives 2.12 The relevance of the PAF’s objectives to the government and the World Bank strategies was high. Through a community-driven approach, the project aimed at increasing opportunities for income generation and developing infrastructure in poor communities, a response to country conditions and the development challenges faced by Nepal both at appraisal and at closing of the project. 2.13 The government’s 10th development plan, and its resulting PRSP, 2002–07, advocated breaking out of longstanding combined states of poverty, slow growth, weak governance, social exclusion, and great inequalities (Nepal 2003). The PRSP articulated four key pillars: (i) achieve and sustain high and broad-based economic growth, focusing particularly on the rural economy; (ii) accelerate human development through a renewed emphasis on effective delivery of basic social services and economic infrastructure; (iii) ensure social and economic inclusion of the poor, marginalized groups and less developed regions, by using targeted programs where appropriate; and (iv) pursue good governance as a means of achieving better development results, and ensuring social and economic justice (Nepal 2003, 22). Under the plan, the PAF project was conceived as a vehicle to deliver resource directly to the rural poor and socially excluded; thus, it was highly consistent with several of the key pillars. 2.14 The project objectives were also consistent with the World Bank’s 2003 CAS, which stressed the need to address social exclusion and rural poverty in Nepal. Additionally, the PAF’s objectives also aimed at supporting the decentralization process. At appraisal, it was 6 envisioned that the PAF would play an important role in decentralizing resource management, a goal articulated in a 2002 World Bank Policy Note, the 2003 CAS, and the 2003 PRSP. 2.15 Both the government and World Bank strategies were consistent with the development needs highlighted in the country diagnostic work. Bringing development resources to foster income-generating activities and access to community infrastructure for the poor and the excluded was vitally important to the government and the World Bank. It was especially important given the fragility conditions caused by the insurgency and civil unrest, which made it hard to deliver resources to remote areas. Under these conditions the PAF project objectives—and, crucially, the project’s delivery modality—assumed particular relevance at appraisal. 2.16 The objectives remained vital to the government and the World Bank at project closure. The PAF was endorsed by the World Bank’s 2007 Interim Strategy Note (World Bank 2007a). Attesting to the relevance of the program, a larger $100.00-million PAF II project was approved on December 6, 2007, by the Board of Executive Directors of the World Bank. 3. Implementation Project Financing and Key Dates 3.1 PAF I was approved by the World Bank’s Board of Executive Directors on June 1, 2004. The total project cost at appraisal was $16.60 million ($15 million in IDA grant, the rest to be funded by contributions of the Government of Nepal and community organizations). In 2006, the PAF was about to commit all available resources (two years ahead of original schedule) when the newly formed government decided to utilize CDD projects to establish its credibility and agreed with the World Bank to an additional financing for the PAF. The additional financing of $25 million in IDA grant was approved on November 14, 2006. The project closed as planned on February 1, 2009, and the total cost at closing was $48.88 million (inclusive of additional financing). Reallocations of the original grant were made twice (in June 2006 and May 2008) and once for the additional grant (May 2008). The goal of the reallocations was to increase financing of income-generating activities and decrease financing of infrastructure subprojects. Implementation Experience 3.2 Though many projects were forced to stop operation because of the conflict situation, quite remarkably PAF implementation was never disrupted. The exception was Pyuthan District, where implementation was suspended for about one year in 2006/07 because of opposition from Maoist district cadres. 3.3 In the second year (2005/06) activities started in districts outside the six pilot districts, under the innovation component 3. In July 2005, the PAF board meeting had decided to extend the project to 19 additional districts. However, because additional financing was only approved in December 2006, Component 3 supported the implementation of some 7 subprojects in these 19 districts. Component 3 also financed winning proposals of the W o r l d Bank’s Nepal Development Marketplace, as was consistent with the innovation objectives of that component. However, the Implementation Completion and Results Report (ICR) recognizes that “little attempt has been made to widely disseminate or scale up any of the innovations—a shortcoming, given the lesson-learning intent of the component” (World Bank 2009, 17). 3.4 Components number 1 and 2 funded income-generating and infrastructure subprojects. Funds were reaching beneficiaries, organized in community organizations, through partner organizations (189 partner organizations were selected in 25 districts), which provided technical and managerial assistance to members of community organizations. 3.5 Under component 4, no capacity building activities were undertaken for local governments or microfinance intermediaries, because the political situation in the country did not allow decentralization to proceed (World Bank 2009, 17). Safeguards Compliance 3.6 Compliance with the environmental and indigenous people safeguards was incorporated in the operation manuals. The subproject negative list of eligible subprojects reflected environment and indigenous populations safeguards criteria. During the supervision missions the World Bank monitored these criteria and no significant issues were detected. 3.7 During preparation, several stakeholder workshops were held to develop the Environmental Management Framework. Government, community representatives, civil society, PAF Board members, and IDA staff discussed the roles of different stakeholders, and the operational rules of PAF implementation, and identified the environmental guidelines of the project. These discussions led to the adoption of an Environmental Management Framework containing clear operational guidelines for environmental assessment including checklists, a negative list, mitigation measures, and guidelines for monitoring (World Bank 2007b, 22). 3.8 Given that the objective of the project was to reach indigenous groups, foster their inclusion, and promote alleviation of their poverty alleviation, the project was expected to have no negative impacts on these groups; furthermore, it was anticipated that indigenous groups would receive both direct and indirect benefits. Based on other Vulnerable Communities Development Plans the operations manual made sure that key documents were available in the indigenous languages, and that the communication strategy for the project included specific efforts to reach indigenous groups. 3.9 Even though indigenous peoples were part of the targeted group, and thus served by the PAF, culturally appropriate support could have been devised to better comply with Operational Policy 4.10 (World Bank 2011). 8 Financial Management and Procurement 3.10 The PAF complied with World Bank fiduciary rules and regulations. The ICR states that some delays in presenting reports and audits took place (World Bank 2009, 20). Financial management manuals and software were developed and implemented in an early stage of the project with satisfactory working results. No fiduciary issues were raised. 4. Relevance of Design 4.1 Relevance of the project’s design to its objectives was rated substantial. The project’s design was consistent with the development objectives, and the results chain reflected in the project’s activities is logical although a results framework was not explicitly presented in the PAD. 4.2 The choice of institutional design was discussed at length with the government, starting in 2000. Although the government wanted to have considerable control over the project, the World Bank team pushed for the PAF’s autonomy as an essential institutional feature to ensure efficiency, based on previous experiences in Asia. Given the government’s inability to reach the poor and excluded at that time (end of 2002) an agreement was reached to give the PAF project full autonomy to implement the program according to the implementing agency’s best judgment. The government set broad objectives and established the PAF’s accountability against agreed targets. 4.3 The PAF was wisely designed in phases. The initial phase supported project implementation in six pilot districts. In a second phase, the project was to be scaled up as experience was gained and the first results achieved. The pilot districts were selected from among the 25 poorest ones according to a development ranking created by the National Planning Commission. 4.4 Improving access to income-generating activities and community infrastructure for socially excluded groups and the poor in rural areas by using a demand-driven approach was arguably not ambitious enough. Aiming at improving access, the project stopped short of explicitly setting the more meaningful goal of improving the well-being of groups and communities. Moreover, without specific targets for income generation and community infrastructure the development objective of improving access would not break longstanding poverty, social exclusion, and inadequate community infrastructure. 4.5 However, the PAF was designed with a coherent theoretical link between the project inputs and the intended outcomes implied in the project’s design summary table (World Bank 2004a, 26). The PAD made it clear that the project-development objective intended access to be a precondition for achieving the desired outcomes of better livelihood and poverty reduction. The component structure is simple and straightforward; by providing direct grants for income-generating activities and community infrastructure projects beneficiaries could take advantage of newly created work opportunities as well as access to communal assets leading to better livelihoods and poverty alleviation. 9 4.6 The CDD approach provided an adequate institutional framework for the project design. This approach aimed at increasing participation, improving social capital within the communities, and fostering inclusion. The mandatory inclusion of the groups targeted by the project in community organizations, and the requirement of weekly meetings to decide how project resources had to be allocated, were meant to promote participation and give voice and decision-making power to the target groups. 4.7 Several design features were meant to ensure the inclusion of target groups in the project’s activities and their sharing of the project’s benefits. Community organizations were to be established at the village level and to comprise the target population. The inclusion of women, Dalit, Janajati, and the poor was addressed in the project design by requiring their presence as mandatory in community organizations, although no specific quota was established. Moreover, the project required that eligibility for income-generating subprojects be based on criteria including caste, ethnicity, and gender, in addition to poverty (World Bank 2004a, 9). For infrastructure subprojects, eligibility was based on the expected impact on those groups, among other criteria. 4.8 The project design also involved the creation of working partnerships with existing local nongovernmental organizations, called partner organizations. This partnership was designed to tap into their knowledge of the local context and make it possible for the PAF to remain a lean institution. They were to become the delivery mechanism in the communities, bridging the gap between the government and the target population. Partner organizations were to be selected based on their presence and experience working in the districts. 4.9 The PAF was designed as an autonomous institution for two main reasons. First, delivering resources directly to communities minimized the probability that subprojects would be damaged (because insurgency groups tended to spare community-owned infrastructure from attack). Second, transferring money directly to the community organization’s account also minimized the risk of resource leakage. The design was therefore able to circumvent the poorly functioning local governments and reach communities in remote districts. 4.10 Elements were included in the design to increase ownership and sustainability. By design, beneficiaries (for income-generating activities) and communities (for infrastructure subprojects) were required to contribute 10 percent of the selected subproject’s cost. This design feature aimed at providing the right incentives for beneficiaries to choose and invest in sustainable activities that could contribute to community organization members’ welfare. In the same way, the O&M plans were meant to ensure that infrastructure subprojects be sustainable over time. 4.11 Both the institutional arrangements and the project components were designed to complement each other in attaining the project development outcomes. The first three components provided money to be invested in income-generating activities and community infrastructure with assistance from the partner organizations for the preparation of subproject proposals and O&M plans. Initially the money was delivered as direct grants to beneficiaries, but over time resources for income-generating activities turned into revolving funds owned and managed by the community organizations. This formula was not foreseen at appraisal, 10 but it was supported by the World Bank and the PAF and later introduced under component 1. Rules requiring weekly meetings and proper bookkeeping were established to ensure sustainability. 4.12 The fourth component was designed to build the partner organizations’ capacity to manage the PAF project activities and provide capacity building to community organizations, PAF project beneficiaries, and local governments, with the aim of strengthening Village Development Committee and District Development Committee government functions. 4.13 Because of the relatively limited aspiration of the project development objectives in terms of producing an impact on the well-being of the target communities, the relevance of objectives is rated substantial instead of high. 5. Achievement of the Objectives 5.1 This section of the PPAR evaluates the extent to which the following objectives of the project were achieved: (i) improved access to income-generating subprojects and (ii) improved access to community infrastructure. This section will also assess the extent to which the achievements of these two objectives have benefited the excluded and the poorest groups in communities targeted by the project. An additional perspective on the project’s efficacy will be based on an assessment of the results achieved by the CDD approach in fostering participation and developing social capital in the participating communities. 5.2 The analysis in this section uses information from the ICR, PAF project reports based on household surveys,5 and the impact evaluation by Parajuli et al. (2012), based on the same survey data. Additionally, the section draws on information acquired during the PPAR mission, the household survey funded by the Independent Evaluation Group (IEG),6 and the focus group discussions carried out in 2016. For additional information about the 2016 IEG- financed survey and focus group discussions see appendix B. Objective 1: Improving Access to Income-Generating Subprojects 5.3 Achievement of the first objective was rated substantial. OUTPUTS 5.4 The number of income-generating subprojects supported by the PAF I and its additional financing was above expectations. About 64 percent of the total project funds went to finance income-generating activities. These funds resulted in output levels well beyond that originally envisioned. At appraisal, there was no specific target regarding income- generating subprojects, but at additional financing a target of 2,500 subprojects was adopted (including income-generating activities community infrastructure, and special programs). At closure, 5,099 income-generating block grants had been supported in the 25 PAF I districts (World Bank 2009, 53).7 5.5 While the number of community organizations that received block grants for income- generating subprojects was very high, little information exists on the distribution of these 11 grants across households within that organization. About 92 percent of the 5,530 community organizations participating in PAF I received income-generating block grants. These organizations comprised 159,777 households (representing 13 percent of the households in the 25 PAF I districts).8 The allocation among individual households within the community organization was decided by the members and recorded by the partner organization, but because the criteria were community organization–specific, it is difficult to assess adherence to these criteria. 5.6 The IEG field visit found no evidence of elite capture. Interviews and focus group discussions indicated that resources were allocated among community organization members according to beneficiary needs and the soundness of their business plans. 5.7 The project exceeded the targets for reaching the excluded groups under the first objective (targets of 30 percent of Dalit and 30 percent of Janajati as members of the community organizations). Of the total beneficiaries of income-generating subprojects under the first component, 41 percent were Dalit and 27 percent Janajati. Similarly, of the total beneficiaries of income-generating subprojects under the third component, 26 percent were Dalit and 66 percent Janajati. 5.8 According to the management information system data recorded in the ICR, most of the resources allocated to improve access to income-generating activities went to fund livestock projects (World Bank 2009). Within livestock activities, meat production was the most frequent choice among community organization members. Tables 5.1 and 5.2 show the distribution of resources allocated to income-generating subprojects and the livestock choices made. Similar results were found in the 2016 IEG-financed survey, where about 60 percent of income-generating activities concerned animal raising for meat production. Table 5.1. Type of Income-Generating Activity Funded under PAF I (percent) Type of Income-Generating Activity Funds Allocated Livestock activities 72 Trade activities 12 Agriculture and forest activities 10 Services sector activities 3 Manufacture activities 3 Source: World Bank 2009, 27. Table 5.2. Distribution of Livestock Activities Funded under PAF I (percent) Type of Livestock Subprojects Funds Allocated Cow and buffalo raising (for milk) 30 Meat production (goats, pig and poultry) 64 Ox and buffalo raising for farm traction or breading 4 Animals for transportation purposes 2 Source: World Bank 2009, 27. 5.9 The 2016 IEG survey reports that during implementation of income-generating subprojects community organization members experienced different types of problems. Among those engaged with income-generating activities, issues associated with the management of family matters and the lack of technical assistance were the most frequently 12 cited problems (29 percent reported experiencing these problems). Less frequently, survey respondents reported problems in accessing markets. 5.10 Only 37 percent of income-generating beneficiaries received skill development training. In addition to the financial resources, the program also planned to deliver skill development training to improve the results of the subprojects supported. However, training was not delivered systematically. Goat raising was the most common training received, consistent with the higher demand for this type of activity. Table 5.3 reports the results of the 2016 IEG-financed survey on skill training for income-generating activities.9 Table 5.3. Training Received for Income-Generating Activities Households Type of Training Received Number Percent Animal husbandry 5 2.3 Buffalo raising 29 13.6 Goat raising 103 48.4 Cow raising 51 23.9 Retail business 7 3.3 Vegetable farming 5 2.3 Others (animal health training, tailoring, computer training, etc.) 24 11.3 All those who received training 213 37 Those who did not receive training 362 63 Source: 2016 IEG household survey data. Note: Percent total may exceed 100 because of multiple responses. OUTCOMES 5.11 The evidence gathered for this PPAR suggests that the project achieved positive outcomes well beyond achieving access to income-generating subprojects. Indeed, food security, consumption expenditure, and employment in income-generating subprojects all increased among beneficiaries. 5.12 Beneficiaries reported improvements in both total food consumption and expenditure—hence, improved food security. According to the 2009 survey data, the number of beneficiaries suffering from severe food insecurity (defined as less than 3 months of food covered by their yearly income) decreased by 15 percent, and the number of beneficiaries who could secure 10–12 months or more of food increased by 9 percent thanks to the program. At the same time, the number of households with less than three months of food decreased by 3.2 percent in the project area as opposed to an increase of 5 percent in the control area (World Bank 2009, 39). 5.13 The same data also showed that per capita consumption increased in both project and control areas, but 50 percent more in project areas than in control areas. Findings from the survey indicate that household expenditure increased by 29 percent and income by 19 percent because of the project (World Bank 2009, 39). Thus, increased access also led to higher levels of income and consumption. 13 5.14 In line with the previous results, the 2016 IEG survey data show that 48 percent of PAF I beneficiaries surveyed became food-sufficient after participating in the program. Moreover, 70 percent of beneficiary households reported an increase of food consumption since 2011, although this was not enough to bring all of them to yearly food security. Similarly, 68 percent of beneficiary households experienced an increase in non-food consumption during the same period (tables 5.4. and 5.5).10 5.15 Even though minorities and excluded groups experienced a substantial increase in food security, the improvement for these groups was below the average of all beneficiaries. According to the 2016 IEG survey, 41 percent of Janajati and 39 percent of Dalit beneficiaries became food-sufficient, as opposed to 65 percent of the other beneficiaries. Table 5.4. Changes in Food Security Status since Community Organization Formation (percent) Food Security Status at the 2016 Food Security Status, by Caste Time of CO Other Food Security Group Formation Janajati Dalit caste Total Sufficient for <=3 months 59.6 8.4 13.9 3.0 8.2 Sufficient for 4–6 months 30.9 32.0 28.3 13.6 24.7 Sufficient for 7–11 months 9.2 18.7 19.1 18.2 19.1 Sufficient for all 12 months 0.3 40.9 38.7 65.2 48.0 Total 100.0 100.0 100.0 100.0 100.0 Source: 2016 IEG household survey data. Note: Altogether, 30 treated community organizations were included in the sample, but comparable data on food security at the start of the PAF program was available for only 24 because this question required a recall of food security status five years before the survey. CO = community organization. Table 5.5. Changes in Food and Nonfood Consumption between 2011 and 2016 Improvement in Food Consumption Improvement in Nonfood Consumption Other Other Factor Janajati Dalit caste Total Janajati Dalit caste Total Percent 74.7 64.7 67.9 69.5 74.7 63.6 67.4 69.0 claiming improvement Number 225 173 202 600 225 173 202 600 Source: 2016 IEG household survey data. Note: The table reports the percent of households who reported increases in food and non-food consumption since 2011. 5.16 The number of working hours per week increased in project areas (+18 percent). However, control areas recorded an even higher increase, +27 percent (World Bank 2009 39). Higher increases in income and consumption increased could indicate improvements in productivity owing to time savings in collecting fodder, fuel wood, and drinking water. 5.17 Evidence from the 2016 focus group discussions also indicated improvements in children’s education, because women could earn more money and spend more on their children’s education. This was a frequent message among women in the 2016 focus group discussions when asked about the benefits from the PAF. “In the past we did not have 14 anything to sell. But PAF supported us and now we started selling goats and increased our income. Thus, we are able to send our children to school and spend money on their education” (Dalit women in a focus group discussion). Objective 2: Improving Access to Community Infrastructure 5.18 Achievement of the second objective was rated substantial. OUTPUTS 5.19 Even though the initial demand for community infrastructure was low, all infrastructure subprojects that communities applied for were supported. A total of 1,216 infrastructure subprojects were financed and built in 23 districts under the PAF I project. A total of $9.1 million were devoted to these subprojects, corresponding to 18.6 percent of the project cost (World Bank 2009, 33). This represents half of the monetary resources allocated to support community infrastructure at appraisal. 5.20 Although the PAD estimated a larger number of community infrastructure subprojects to be supported by PAF I, the project did not specify nor enforce any specific target, because the choice was to be made by the community organizations. Therefore, the lower-than- expected amount of community infrastructure supported by PAF I cannot be seen as a failure but as a result of the demand-driven nature of the project. 5.21 The main reason behind the lack of interest in community infrastructure and its low uptake was the high incidence and depth of poverty that community members were experiencing when the PAF was introduced. Income-generating activities were seen as a more direct and faster way of improving household food security. Also, under an imminent threat of violence that could displace people from their communities, beneficiaries preferred access to resources that they could take with them in case of need. Furthermore, the state of fragility and the fear of having the infrastructure destroyed by the insurgency groups led to lower demand for these types of subprojects. 5.22 Infrastructure subprojects also benefited households that were not community organization members. According to the ICR, 34,783 nonmember households also benefited from infrastructure subprojects, which brings the total number of direct and indirect beneficiaries from PAF I subprojects to 194,560 (corresponding to 122 percent of the target of beneficiary households set at additional financing; World Bank 2009, 22). 5.23 The results of the 2016 IEG survey showed that of the 600 community organization members in the sample only 34 percent applied for community infrastructure subprojects, while 96 percent applied for income-generating activities. All applications for community infrastructure were approved, as reported in the survey. 5.24 Overall, most beneficiaries of community infrastructure supported under component 2 were not Dalit or Janajati. Only 16 percent of the direct beneficiaries were Dalit and 11 percent were Janajati. However, of all direct beneficiaries of infrastructure subprojects supported under component 3 72 percent were Janajati and 14 percent Dalit. 15 5.25 The quality of the community infrastructure supported was good and sustained over time by appropriate maintenance. The community contributions and the O&M plans provided the right incentives and fostered a sense of ownership that helped sustain the infrastructure over time. At closure of PAF I no major quality issues were identified (World Bank 2009, 27). However, when the ICR mission was conducted most infrastructure subprojects had been recently built. OUTCOMES 5.26 Education infrastructure supported by PAF I improved education outputs and outcomes. According to Parajuli et al. (2012), the project helped increase school enrollment among those 6–15 years old. The impact evaluation estimated a 14 percentage points increase in school enrollment for all children in the given age group. Girls appeared to have benefited even more, with 21 percentage points net increase (Parajuli et al. 2012, 10). Furthermore, the estimated impacts were as strong among children from disadvantaged caste/ethnic groups. 5.27 PAF infrastructure subprojects have contributed to improved health and children’s education, as reported by beneficiaries in the 2016 IEG survey. About 55 percent of community organization members involved in infrastructure subprojects reported perceived benefits to children’s education and health. 5.28 Significant improvements in children’s education were also reported as one of the benefits from PAF in the focus group discussions conducted in 2016. These results are in line with what the impact evaluation by Parajuli et al. (2012) found in education. However, that study did not find any statistically significant effect on child nutrition. 5.29 IEG visited two community infrastructure subprojects during the PPAR mission (an irrigation project in Rasuwa district and one water point in Ramechhap district). Both were in good working order. The beneficiaries claimed that these subprojects have improved their lives by allowing them to cultivate their crops closer to home and by facilitating access to potable water. Participation, Social Mobilization, and Social Capital Improvements 5.30 Beside the achievement of the two project objectives examined in the previous two sections, the project produced other direct and indirect benefits, including improvements in participation and social capital. This subsection focuses on these additional achievements from PAF I, based on the evidence available in program documents and the data collected for this PPAR. 5.31 Decision making has become more participative over time in PAF beneficiary households. According to the ICR (and based on the 2006 and 2008/9 survey data), before the PAF, 40.2 percent of household decisions were made by the head of household; after participating in the program this percentage went down to 26.7 percent (World Bank 2009). Similarly, decisions jointly made by all members of the household increased from 16.4 percent to 33.5 percent after the program. 16 5.32 Participation of women and minorities in community organizations exceeded the target. Women represented 67 percent of community organization participants, surpassing the 40 percent target set at additional financing. Similarly, Dalit and Janajati represented 37 percent and 44 percent of the total community organization participants respectively, surpassing the 30 percent target set at additional financing. 5.33 Quotas for women in community organizations that adopted them played an important role in fostering women’s participation and helped strengthen their voice. “Now, women also participate because there is a mandatory provision of the 33 per cent of women participation in [community organizations]” (woman’s statement in a focus group discussion). Participants saw the quota as a positive feature of the PAF, because it helped integrate women in the community and helped make their voices heard.11 5.34 Increased female participation in community organization and community decision making was the most important benefit reported in the 2016 IEG survey. Eighty-six percent of respondents reported that female participation increased because of the PAF (table 5.6). Moreover, 74 percent of survey respondents responded that PAF has helped improve social capital and the “spirit and sense of social togetherness.” This finding was also supported by the focus group discussions and interviews in the field. 5.35 Group meetings and the CDD model followed by PAF were seen by beneficiaries as instrumental in fostering women’s participation and decision-making power. Focus group discussions revealed that, over time and because of the community meetings, husband-wife joint decision making had become more prevalent than in the past. The IEG survey confirmed that women in the PAF communities have the power to make decisions within the household for the most part. Most women can decide by themselves on “minor household expenses” and participate jointly with other household members in most of the other decisions (table 5.6). Several interviewees stated that the PAF has created an atmosphere where people can put forward their ideas without being discriminated against, and has helped increase women’s decision-making power over time. 5.36 Although the PAF has contributed to positive changes, gender roles still prevent women from fully participating in decision making. Close to one-fifth of the women who responded to the IEG survey could not make decisions affecting their life (table 5.6.). Similarly, though more than 90 percent of women said that they would be consulted when buying/selling household property, very few of them indicated that they can sell their own property (land, house, jewelry, and livestock) independently. Focus group discussion participants also expressed concerns about the prevailing social norms that still discriminate against women. “Women are not involved or called in to meetings probably because of their poor education, silence habit, and the assumption that women’s role is to do household tasks,” “Due to lack of education we are not called or involved in meetings” (Janajati women in focus group discussion), “The in-laws are responsible for making decisions if a house has in-laws; after them, the husbands are responsible as head of the family” (woman in focus group discussion); similarly, another female participant said, “We have to obey the decisions made by our in-laws.” 17 5.37 Though elite capture did not appear to be an issue, community members indicated that “village elites” are especially active in the village decision-making process. However, community members stated that, in the PAF context, those elites (for example, committee members, village elders, and local-level political leaders) made decisions after consulting and receiving support from people in the village. Most of the participants said that they were happy with the decisions made in the meetings. Table 5.6. Perceived Benefits from PAF by 2016 (percent) Perceived Benefit Households Improvement in living standard 78.3 Improvement in social harmony/community relation 78.7 Increase in income and saving 69.3 Increase in access to market 41.0 Increase in female participation 85.8 Increase in spirit of social togetherness 73.8 Improvement in practice of collective decision making in family 64.8 Promotion in skill development 32.5 Improvement in children’s education 51.7 Increase in family assets (physical) 39.8 Improvement in health conditions 41.2 Others 1.8 Number of households 600 Source: 2016 IEG household survey data. Table 5.7. Women’s Role in Household Decision Making (percent) Decision Maker Decision Making Area Self Self + other Other Minor household expenses 63.6 19.6 16.8 What to do if you are sick 29.9 48.6 21.4 How to protect yourself from violence 31.3 49.0 19.7 Whether and how to exercise religion 31.8 48.8 19.4 What kinds of tasks to do 33.2 48.3 18.5 Children’s education 27.6 56.4 16.0 Use and nonuse of contraceptives 9.5 75.0 15.5 How many children to bear 9.0 75.1 15.9 Total 30.0 52.1 18.0 Source: 2016 IEG household survey data. 5.38 Networks created around similar activities led to the formation of cooperatives improving the sense of collectivism, inclusion, and social capital. The benefits of revolving funds and regular meetings in communities receiving income-generating activities were associated with the development of networks, according to community organization members interviewed in the field. 6. Efficiency The efficiency of achieving the development objectives is substantial. 18 6.1 Despite limited evidence on the efficiency of the project, this PPAR rates efficiency substantial. The PPAR acknowledges that calculating economic and financial rates of return for a project that contains a wide array of subprojects is a difficult task. Yet, the PAF (especially PAF II) has made an effort to collect data on efficiency; the available evidence suggests that the program was cost-effective. The evidence for this conclusion is as follows. 6.2 The ICR shows that a select sample of subprojects were financially and economically viable. Based on a small sample of subprojects (both income-generating and community infrastructure) the ICR presents an assessment of incremental annual income and rates of return (World Bank 2009, annex 3).12 This analysis showed that the subprojects evaluated generated financial rates of return ranging from 21 percent for water storage pond to 847 percent for a grocery shop.13 Additionally, the analysis showed that household incremental annual incomes ranged from NPR 4,500 (equivalent to about $65) for goat fattening to NPR 25,650 (equivalent to about $370) for milking buffalo. However, these positive results are based on a very small sample, which limits their generalizability. 6.3 Based on available information, infrastructure subprojects supported by the PAF were built at lower cost than similar projects implemented by other agencies. The ICR presents a cost comparison between government infrastructure and community-implemented subprojects (World Bank 2009, 39). The results showed costs to be 13 percent to 48 percent lower for community-based subprojects, depending on the type of subproject (drinking water, irrigation, and roads). Even though these numbers suggest higher efficiency of PAF infrastructure subprojects, there is no clarity about how many and which projects are included in the ICR calculations and to what extent the full cost of labor was considered for the community projects. 6.4 Several knowledgeable interviewees provided anecdotal evidence of PAF “being efficient” in delivering results, given the particularly difficult political environment. According to these sources, the PAF was less prone to be captured by corruption and inefficient practices prevailing in Nepal at the time of implementation. Many of these interviewees advised IEG that the PAF was able to circumvent the existing resource delivery structure, avoiding money leakages in the process. However, though plausible, none of these claims were corroborated empirically. 6.5 In terms of the broader objective of providing access to income-generating activities and community infrastructure to excluded groups, PAF administration costs were low when compared with similar government investments. According to the project’s expenditure report presented in the ICR, the direct administrative costs for the PAF project amounted to only 3.6 percent of the total project costs (World Bank 2009, 33). Most of the project budget went to subprojects (51.5 percent to income-generating subprojects, 18.6 percent to community infrastructure, 22.5 percent to special projects, and 3.7 percent to capacity building). During the PPAR mission several knowledgeable interviewees pointed out how much leaner the PAF was compared with the Government of Nepal as a project delivery mechanism. All of them explained that PAF was more cost-effective than the government because of its design and its partnership with local organizations. 19 6.6 As mentioned earlier in the text, Parajuli et al. (2012) estimate the impact of PAF on several outcomes. Using different methodologies (difference in difference and instrumental variables estimators) the evaluation found positive results on per capita consumption, food security, and school enrollment for those 6–15 years old. The 19 percent increase in per capita consumption translates to about additional $40. This meant that the PAF provided a more than 130 percent return for an initial per capita investment of $30 in just over two years (Parajuli et al. 2012, 9). 7. Ratings Outcome 7.1 The rating of development outcome is satisfactory. 7.2 The relevance of the objectives is high because they were fully aligned with the country conditions, and with the government’s and the World Bank’s strategies. The relevance of design is rated substantial. In stressing access to income-generating activities and infrastructure (and not also an increase in community well-being), the project objectives were relatively unambitious. Yet, the project design included many elements to support the achievement of its objectives and to ensure inclusion, ownership, and sustainability. The CDD approach and the creation of the PAF as an autonomous body helped overcome the difficulties of reaching the target groups given the specific circumstances of the country at the time – characterized by fragility and conflict. 7.3 The efficacy of objective 1 (improved access to income-generating subprojects) is rated substantial because the project delivered far more income-generating subprojects way than were originally planned (and far more than the target set at the time of additional financing). Moreover, these subprojects generated positive impacts on the well-being of the targeted communities, in food security, consumption expenditure, and employment. The efficacy of objective 2 (improved access to community infrastructure) is rated substantial. Even if the number of infrastructure subprojects was well below the initial expectations, all the requests from communities were fulfilled; simply, there was little demand for infrastructure. The impact evaluation documents positive impact on enrollment effected by education infrastructure. Both objectives were achieved also in including the poor and the excluded, who shared in the benefits brought by income-generating and infrastructure subprojects. 7.4 Efficiency is rated substantial. The empirical evidence, albeit limited, showed that a sample of income-generating projects generated high financial rates of return and that the community infrastructure was cost-effective. The impact evaluation showed that the benefits, in consumption levels and food security achieved thanks to the project, were larger than the costs sustained to generate them. In addition, project management costs were only 3.6 percent of total costs. 20 Risk to Development Outcome 7.5 The overall risk to development outcomes is rated moderate. The PAF has survived over time as an independent body and has so far proven to be “institutionally” sustainable. The initial project was followed by a much larger PAF II and additional financing that supported program outcomes and expanded its outreach to almost all districts in Nepal. PAF II is currently coming to closure and there are ongoing discussions about a potential PAF III. Several people interviewed during the IEG mission stressed that a new PAF project would need to be redesigned to adapt to the new political and economic situation. The creation of a new Ministry for Poverty Alleviation and Cooperatives may also require the PAF to change, depending on how the competencies of the Ministry are going to be defined. Key informants interviewed for this evaluation stressed that focusing on livelihood, productivity improvements, and access to markets could be welcome modifications of PAF activities. 7.6 The IEG-financed survey showed that more than two-thirds of income-generating activities funded by PAF I have been sustained over time, seven years after the project closed. Participants declared being highly involved in planning, implementing and managing their subprojects. Moreover, IEG survey data indicate that more than 82 percent of community organization members engaged in community infrastructure subprojects have set up O&M rules and close to 70 percent of them regularly collect money for maintenance. 7.7 Lack of integration with other programs is a risk that the PAF faces. Many knowledgeable people interviewed during field visits stressed that it needs a systematic link with other projects, especially infrastructure projects and productivity-related interventions (such as the Rural Water Supply and Sanitation Improvement Project and the Urban Governance and Development program). Poverty alleviation will become more complex because, after addressing basic deprivation, poverty will be increasingly recognized as a multidimensional and multifaceted phenomenon. To support higher productivity gains the PAF will need to connect with programs expanding infrastructure and access to services. SUSTAINABILITY OF INCOME-GENERATING SUBPROJECTS 7.8 Most income-generating activities funded under PAF I were related to animal husbandry,14 which involves several risks. Lack of capacity and access to veterinary care have been experienced by several community organization members. Animal illness could be extremely damaging to beneficiaries, leaving them in debt unless they have access to insurance or veterinary services. The 2016 IEG survey data show that about 18 percent of PAF I beneficiaries who engaged in income-generating subprojects have terminated their activities because of losses or low profits. 7.9 Most community organizations reported high repayment rates and high user satisfaction with the revolving funds. In many cases members were observed saving money weekly to increase the revolving fund and access to more resources in the future. According to the interviews and the survey data, most community organizations meet regularly to discuss community matters and collect and allocate funds. Little or no evidence was found of depletion of the social funds or members not repaying the interest and capital. Even though interest rates are very low compared with moneylenders and microfinance institutions, social 21 funds have increased substantially since their inceptions. Many communities have created cooperatives to institutionalize these funds, thus improving their sustainability over time. 7.10 No major issues were observed in the quality of O&M of infrastructure subprojects, on which their sustainability depends (World Bank 2009, 27). During the PPAR mission, the few infrastructure projects visited evidenced proper maintenance and usage. Several beneficiaries stated that the community was in charge of maintaining them or raising funds to pay for their maintenance. Even though few observations cannot be generalized, the fact that the PAF requires an O&M committee and raising funds for maintenance as a requirement for community infrastructure approval foster subproject sustainability. That said, subprojects are still relatively new to properly assess their sustainability over a longer time-period. World Bank Performance 7.11 Quality at entry was moderately satisfactory. The project design was well aligned with the needs of the country, and the choice of a targeted CDD scheme proved to be a suitable way to reach the poor and the excluded groups under the economic and social conditions prevailing in Nepal at time of design. The World Bank team engaged in a long discussion about the institutional arrangements of the project before designing the final version of the PAF. 7.12 During appraisal, the project team assumed that income-generating activities would have met little demand, and that community infrastructure would have absorbed most of the project’s budget. Accordingly, the PAD allocated 70 percent of the budget to community infrastructure. During implementation, however, the communities’ preference for income- generating activities required a reallocation of funds; at closing the share of this type of subprojects was 70 percent—the opposite of what was estimated at appraisal. The inability to anticipate the desired mix of subprojects did not allow deploying the necessary training and building on the complementarity of income-generating and infrastructure subprojects. This was a clear shortcoming of the project. 7.13 Starting with an initial set of six pilot districts was appropriate, and proved to be beneficial for the project, given the prevailing fragility of the country and that a demand- driven approach had not been implemented before in Nepal. The phased approach allowed learning to take place and implementation issues to be addressed the project was scaled up to the additional 19 districts. 7.14 In addition, there was no capacity building in the local governments. Because of the political instability, this component of the project was not implemented. During the implementation period, elected local governments were dissolved and the expected decentralization process did not take place, undermining the PAF’s capacity-building intentions. The World Bank could have assessed the political environment more accurately so as to have a contingency plan in case the decentralization process stalled. 7.15 Similarly, capacity building for microfinance intermediaries, originally included in the project, did not prove feasible in the light of their very limited outreach in rural areas, an issue that might have been better foreseen at appraisal. 22 7.16 Quality of World Bank supervision was moderately satisfactory. The World Bank team was able to verify firsthand through a number of supervision missions the quality of the project, reaching remote and conflict-affected areas where the project was implemented. Additionally, short field visits complemented the supervision efforts. Additional missions to understand the project were conducted before additional financing and the preparation of the second phase of the project. 7.17 A midterm review was not undertaken; such an exercise could have highlighted some key issues in a timely fashion. However, 10 ISRs were prepared, one every six months on average during implementation; progress toward achievement of project development objective and overall implementation progress ratings were satisfactory throughout all Implementation Status and Results Reports, reflecting the smooth implementation of the project and its adequate fund disbursement. A stocktaking to prepare the Additional Funding was conducted in 2006, which explains why the project team opted not to carry out the midterm review planned for the following year. However, the midterm review could have benefited the project by, for example, highlighting the need to measure results beyond outputs and helping the PAF in the PAF’s transition to project in which income-generating activity was prevalent. The revolving fund manual was developed just in time for the second cycle. 7.18 Consistent collaborative work between the World Bank team and the PAF Secretariat helped to refine project operations and delivery mechanisms. This collaboration resulted in the successful scaling up from six pilot districts to the 25 poorest districts in the country. The close partnership developed by the task team and the PAF was highlighted during the PPAR mission in several meetings and confirmed by the PAF task team leader. Borrower Performance 7.19 Government performance was satisfactory. The government established the PAF as an independent body through an Act in 2006, reflecting a long-term commitment to the project. Throughout the first phase (PAF I and the additional financing), the Government of Nepal supported the PAF despite political instability and administration changes. The government respected the PAF’s autonomy and established an adequate relationship with its board of directors. Counterpart budget was allocated in time to ensure service delivery. 7.20 Better coordination with other policies and programs could have been beneficial. The government was unable to promote consistency between the PAF and the plans for decentralization. At the same time the government could have facilitated coordination between the PAF and other programs (for example, microenterprises, productivity, and access to markets) to exploit synergies and further catalyze the impact of the project. 7.21 The PAF’s performance as the implementing agency was satisfactory. The PAF implemented the program as designed, collaborating fluidly with the World Bank task team. As in the borrower’s own performance assessment, better coordination with line ministries could have helped the PAF achieve better results. Additionally, it could have developed the internal M&E system further to provide robust evidence of results, beyond outputs. That said, 23 PAF worked under very difficult circumstances, due to conflict and extreme institutional weakness, and yet managed to deliver satisfactory results. 7.22 During the PPAR mission PAF staff stated that after more than a decade of engagement, some partnership organizations have become less effective in helping the community organizations. High turnover of staff and social mobilizers have diminished the effectiveness of partnership organizations. Under PAF II a thorough review of partnership organizations analyzed several aspects of their engagement with community organizations. Several have progressively disengaged because of weak community organization results (for example, low ability to manage funds, and coordination problems with other development interventions). Monitoring and Evaluation 7.23 The PPAR rates overall monitoring and evaluation (M&E) of the project modest. 7.24 The M&E at design and implementation focused mostly on outputs, which resulted in a weak results framework. The project monitored and assessed indicators related to the number and type of subprojects funded and beneficiaries served, without paying much attention to the outcomes that these subprojects could have generated. Focusing on who received benefits helped ensure that funds reached those for whom the program was intended. 7.25 An important feature of the PAF has been the strong emphasis on data collection through household surveys. Baseline and follow-up surveys have been carried out during PAF I to better understand the program results. Thus, the PAF has a rich set of cross-section and panel data available to conduct evaluations (including impact evaluations). However, apart from the Parajuli et al. 2012 impact evaluation (that came after PAF I was closed) there was no other effort to evaluate the outcomes and impacts of PAF I. 7.26 The PAF and partner organizations regularly used participatory monitoring at the community level to assess program results and implementation. This process promoted beneficiary engagement, and fostered ownership by transparently identifying issues and taking corrective actions when needed. However, the revolving fund feature was not properly investigated because the PAF monitoring information system had not gathered information to assess the sustainability of these funds. 7.27 Better use of the M&E data to generate more evidence about the project outcomes and impacts could have taken place. The data collected was underutilized to produce evidence for decision making. 8. Lessons 8.1 The community-driven development approach can provide a powerful organizing framework for the improvement of the welfare of communities. In the PAF, community organizations were provided with opportunities to make competitive proposals for broad CDD programs, and when provided with grants were entrusted with managing the 24 allocation of funds, either among members for the best subprojects on a participative basis, or for community infrastructure. As this PPAR has concluded, the CDD framework generated considerable social capital in poor and excluded communities. 8.2 The sustainability of CDD programs hinges on finding a wise balance between program ‘agility’—achieved through independence of the implementing agency—and government ownership and coordination with line ministries. Originally, the PAF was created as an independent agency because this was the best (or only) option to maximize project outreach and minimize leakages of funds. This was at a time when the government was unable to reach more remote areas, because of weak local administrative structures, human resource limitations, and the high level of violence present in some areas. In a more stable political environment, however, the independence of the implementing agency may undermine the role and coordination function of line ministries, which need to ensure that the program works in synergy with other development interventions. Even as this issue must be addressed through more direct collaboration and coordination with line ministries and local governments, efficacy and efficiency of the project must be maintained too. 8.3 A better assessment of the communities’ needs would probably have anticipated their demand for subprojects and helped improve the design of the project. Even though the PAF design benefited from lessons learned in previous Asian CDD experiences, the design of the project did not initially reflect the actual demand of the recipient population. The target communities prioritized income-generating activities over community infrastructure as a more direct way to improve family consumption and food security. 8.4 Income-generating subprojects require complementary activities, such as specific training and support to accessing markets. These complementary activities should be embedded in the design to ensure full project success. The data collected by IEG indicate that many beneficiaries of income-generating subprojects cited lack of training and poor access to markets as potential risks for their activities. 8.5 A carefully planned and implemented pilot phase for a project provides essential learning for a successful scaling up. Piloting of the PAF in six initial districts and then scaling up to the poorest 25 districts helped fine-tune some key features of the project, such as moving from providing grants for subprojects to the creation of revolving funds, and identifying weaknesses and potential improvements in delivering resources to communities. 8.6 Robust data collection and effective analysis of available data are crucial for project corrections and to showcase positive results. The PAF conducted several household surveys, and collected valuable data on outputs and outcomes over the life cycle of the project. Nonetheless, some variables were not collected and those available were underutilized to produce evidence for decision making. The project could have planned for more extensive M&E on some aspects, such as its efficiency, and its impact on employment, health, earnings, and productivity. 25 References Nepal, Government of. 2003. The Tenth Plan: Poverty Reduction Strategy Paper 2002–2007. Kathmandu, Nepal: National Planning Commission. Parajuli, D., G. Acharya, N. Chaudhury, and B. Bahadur. 2012. “Impact of Social Fund on the Welfare of Rural Households: Evidence from the Nepal Poverty Alleviation Fund.” Policy Research Working Paper 6042, World Bank, Washington, DC. World Bank. 2003. Nepal—Country Assistance Strategy. World Bank, Washington DC. ———. 2004a. “Nepal— Poverty Alleviation Fund.” Project Appraisal Document Report 28606-NEP, World Bank, Washington, DC. ———. 2004b. “Nepal—Poverty Alleviation Fund.” Project Development Grant Agreement H091-NEP, World Bank, Washington, DC. ———. 2007a. “Nepal—Interim Strategy Note.” Report 38119-NEP, World Bank, Washington, DC. ———. 2007b. “Nepal—Poverty Alleviation Fund Project II.” Project Appraisal Document Report 41317–NP, World Bank, Washington, DC. ———. 2009. “Nepal—Poverty Alleviation Fund Project.” Implementation Completion and Results Report ICR1197, World Bank, Washington, DC. ———. 2011. “Nepal—Poverty Alleviation Fund Project I.” Independent Evaluation Group, Implementation and Completion Results Report Review ICRR13295, World Bank, Washington, DC. ———. 2016. World Development Indicators: 2016. Washington, DC: World Bank. World Bank and DFID (Department for International Development). 2004. Unequal Citizens: Gender, Caste and Ethnic Exclusion in Nepal. Washington, DC: World Bank and Department for International Development, Kathmandu, Nepal. 1 The Poverty Alleviation Fund was initially established by an ordinance, and it has been governed by its own act since 2006. It is an autonomous, independent and professional organization governed by law. 2 According to PovCalNet figures based on a poverty line of $1.90 dollar per capita a day. 3 It is worth noting that reducing gender, ethnic, and caste disparities is a long and complex process that takes much more than a project to be successfully addressed. The new Constitution (established in September 2015) improved women’s rights in several areas, but still denies women the right to pass their citizenship on to their children; only men can pass their citizenship on to their children. 4 The CAS linked discrimination to lack of access to property, obstacles in investing in health and education, and lack of voice in community decision making and national governance. 5 For evaluation purposes, under Poverty Alleviation Fund I several household surveys were conducted. A baseline survey in the initial six pilot districts was carried out in 2005/06 with a follow-up survey in 2007/08. Similarly, a baseline and follow-up surveys covering the additional 19 districts (six sample districts) were conducted in 2007 and in 2009/10, respectively. 6 It must be noted that the control group in the Independent Evaluation Group survey was not a credible comparator of the treatment group, and therefore the survey could not be used to measure the impact of the project. See appendix B for an explanation of this shortcoming. Another shortcoming of the Independent Evaluation Group survey was that it collected information on the whole Poverty Alleviation Fund (PAF) program with no ability to distinguish between PAF I and PAF II because of obvious recall issues. 7 The IG block grant was the total money given to the community organization financed by the Poverty Alleviation Fund project for income-generating activities. To receive this grant community organization 26 members (as a group) had to choose this option over the infrastructure grant. After the organization received the grant, the money was allocated among the members according to preferences expressed through a participatory process. 8 According to the 2011 Nepal National Population and Housing Census a total of 1,198,495 households lived in the 25 Poverty Alleviation Fund I districts. Considering that at appraisal the national poverty rate was about 70 percent, the PAF might have reached close to 20 percent of the poor households, assuming it only reached poor households. 9 The question was asked to members of households involved in income generating activities supported by Poverty Alleviation Fund. It is a recall question without a specific time frame: “Did you/any member of your family receive any training to run this program?” 10 These outcomes are attributable to the program in general, not necessarily to its initial phase (PAF I). 11 Partner organizations worked in tandem with communities, using participatory methods to identify and involve the poor and the excluded. In addition, the Poverty Alleviation Fund used targets and incentives to encourage community organizations to include women, Dalit, and Janajati as their members, office bearers, and beneficiaries. Specific quotas were not set at the central level, but some partner organizations made their own rules and introduced quotas to ensure participation of women from the target groups. 12 Subprojects assessed covered only two districts Kapilvastu and Pyuthan Districts. 13 Note that this is a very high percentage but of a very low base income. 14 According to the 2016 Independent Evaluation Group household survey, about 78.9 percent of Poverty Alleviation Fund I participants that chose IG activities opted for animal related activities. 27 Appendix A. Basic Data Sheet POVERTY ALLEVIATION FUND I (IDA-H091 IDA-H258) Key Project Data (amounts in US$, millions) Actual or Actual as % of Appraisal Current Appraisal Estimate Estimate Estimate Total project costs 44.71 48.88 109 Loan amount 40.00 41.44 104 Cofinancing 4.71 7.44 158 Cumulative Estimated and Actual Disbursements FY05 FY06 FY07 FY08 FY09 Appraisal estimate (US$, 1.6 4.5 14.3 29.3 40 millions) Actual (US$, millions) 0.75 4.62 13.32 33.44 41.54 Actual as % of appraisal 47% 103% 93% 114% 104% Date of final disbursement: Project Dates Original Actual 04/21/2003 and 02/09/2006 Concept Review (additional financing [AF]) Appraisal 02/06/2004 and 03/01/2006 (AF) Approval 06/01/2004 and 11/14/2006 (AF) Effectiveness 08/30/2004 12/13/2004 and 06/04/2007 (AF) Midterm Review 02/08/2007 02/08/2007 Closing date 02/01/2009 02/01/2009 28 Staff Time and Cost Staff Time and Cost (World Bank budget only) Staff weeks US$, thousands (number) (including travel and Stage of Project Cycle consultant costs) Lending FY03 10.20 68.01 FY04 39.11 233.84 FY05 0 0.00 FY06 0 0.00 FY07 0 0.00 FY08 0 0.00 Total: 49.31 301.85 Supervision/ICR FY03 0.00 FY04 0.00 FY05 56.46 277.99 FY06 40.12 245.75 FY07 31.06 207.29 FY08 13.25 105.39 FY08 8.94 82.49 Total: 149.83 918.91 29 Task Team Members Title (at time of appraisal and Name closure, respectively) Unit Responsibility/Specialty Lending Geeta Sethi Economist SASDA Task team leader Bigyan B. Pradhan Sr. Financial Management SARFM Fin. Management Ian Bannon Specialist AFTCS Soc.Dev; Conflict Kiran R. Baral Sector Manager, Social SARPS Procurement Luis O. Coirolo Development SASDA CDD Strategy Luis F. Constantino Sr. Procurement Off. MNSSD Rural Development Naima A Hasci Consultant AFTCS Social Development Yuka Makino Sector Manager Environment Nazmul Chaudhury Sr. Social Scientist M&E Kiran Gautam Environmental Specialist SACNP Paul Jonathan Martin Economist AFTEN Environment Yoshiko Masuyama Senior Executive Assistant SASDO Jeffrey S. Hammer Sr. Environment Spec. DECPO M&E Nawaf. A. Al Mahamel Program Assistant Legal Sarita Rana Consultant Counsel Project Assistant Supervision/ICR Geeta Sethi Economist SASDA Task Team leader Ian Bannon Sector Manager, Social AFTCS Soc.Dev; Conflict Kiran R. Baral Development SARPS Procurement Luis O. Coirolo Sr. Procurement Off. SASDA CDD Strategy Luis F. Constantino Consultant MNSSD Rural Development Kiran Gautam Sector Manager SACNP Drona Raj Ghimire Senior Executive Assistant SASDI Environment Jeffrey S. Hammer E T Consultant DECPO Naima A Hasci Consultant AFTCS Social Development Yoshiko Ishihara Sr. Social Scientist FAO ICR Natasha Hayward Consultant SASDA ICR Paul Jonathan Martin Sr. Rural Dev. Specialist AFTEN Environment Yoshiko Masuyama Sr. Environment Spec. SASDO Dilip Parajuli Program Assistant SASHD M&E Atul Pokharel Education Economist SASDI M&E Bigyan B. Pradhan Consultant SARFM Fin. Management Donna Thompson Sr. Financial Management CTRLP Fin. Management Specialist Sr. Financial Management Specialist 30 Appendix B. 2016 IEG-Funded Survey and Focus Group Discussions To understand the results of the PAF project and its benefits, the IEG team conducted a set of focus group discussions as well as collected quantitative data from a household survey to evaluate PAF independently (Focus group discussion and survey work was carried out by the Center for Economic Development and Administration from Tribhuvan University during 2016). Twelve hundred households were surveyed in three PAF I districts (Rolpa, Dailekh, and Mahottari). 2016 IEG-Financed Survey The multimodule quantitative survey was conducted in 1,200 households (600 in treatment areas and 600 in control areas). The 600 sample households in treatment areas were randomly selected from 30 community organizations where the program was implemented, and the 600 control households were randomly selected from 30 wards of village development committees and municipalities without PAF. 20 household surveys were conducted in each community organization and ward (see table B.1 for details on the sample). As for the control groups, unfortunately, they were chosen at random but from the poor population of households in control wards. This sampling system generated biases in the control villages and thus making it difficult to use them as a proper counterfactual. Table B.1. IEG-Funded Household Survey, Number of Community Organizations, Wards, and Households Sampled by District District Treatment Control Total Rolpa CO: 10 Ward: 10 CO+Ward: 20 HH: 200 HH: 200 HH: 400 Dailekh CO: 10 Ward: 10 CO+Ward: 20 HH: 200 HH: 200 HH: 400 Mahottari CO: 10 Ward: 10 CO+Ward: 20 HH: 200 HH: 200 HH:400 Total CO: 30 Ward: 30 CO+Ward: 60 HH: 600 HH: 600 HH: 1,200 Note: CO = community organization; HH = households. Focus Group Discussions Thirty-one focus group discussions were conducted in treatment areas of five PAF I districts (Ramechap, Rasuwa, Rolpa, Dailekh, and Mahottari). The main purpose of the focus group discussions was understanding how PAF has helped with community decision making, the role of women and their level of participation in decision making both at the community and household level Across the five districts 23 focus group discussions were conducted with women-only groups and eight with men-only groups, to better understand participation and gender dynamics, (see table B.2 for more information about the composition of the groups interviewed). 31 Table B.2. Number of Focus Groups by Sex, Ethnicity, and District Type of FGD FGDs District Women Only Men Only (no.) Ramechhap 6 (J = 1, D = 1, J = 2, B/C = 2) 2 (B/C = 1, J = 1) 8 Rasuwa 6 (B/C = 1, J = 4, D = 1) 2 (B/C = 1, J = 1) 8 Mahottari 4 (B/C = 1, J = 1, D = 1, M = 1) 1 (D = 1) 5 Rolpa 4 (B/C = 1, J = 2, D = 1) 1 (J = 1) 5 Dailekh 3 (B/C = 1, J = 1, D = 1) 2 (B/C = 1, D = 1) 5 Total 23 8 31 Note: B/C = Brahmin/Chhetri; FGD = focus group discussion; J = Janajati; D = Dalit; M = Muslim. 32 Appendix C. Methodology and List of Indicators Used to Rank Districts The National Planning Commission used a Composite Index technique to rank districts by development status. This index transforms different indicators into 0 to 1, regardless of their original unit of measurement. The smallest value 0 represents the worst development condition and the largest value 1 represents the highest level of development (known as a Zero-to-One Scoring Transformation). The computation formula for this transformation is as follow: Let, Xij denote the value of the jth indicator in ith district Maxj and Minj denote the corresponding maximum and minimum values of the jth indicator in all 75 districts, then the Composite Index for the jth indicator in the ith district, say Cij is computed by using the following expression: Cij = Xij –Minj / Maxj – Minj (for high values like literacy rate) Cij = Maxj –Xij / Maxj - Minj (for low values like proportion of households using solid fuel for cooking) Once all indicators are transformed into composite indexes, the overall composite index (OCI) is computed for the ranking of the districts. The OCI is calculated as follows: OCIi = Ai – Min A / Max A – Min A Where Ai corresponds to the average value of the composite indexes in the ith district, and Max A and Min A denote the corresponding maximum and minimum values of each composite index in all 75 districts. Table C.1. Indicators Used to Rank Districts Indicator at District Level 1 Access to improved source of drinking water 2 Access to toilet facility 3 Proportion of households having electricity facility 4 Proportion of households using solid fuels for cooking 5 Proportion of households having radio facility 6 Telephone lines per thousand population 7 Road density (length/km2) 8 Singulated mean age at marriage female 9 Child dependency ratio 10 Incidence of ARI per 1,000 children < 5 years 11 Incidence of diarrhea per 1,000 children < 5 years 33 12 Proportion of malnourished children under 3 years 13 Reported death per 1,000 population 14 Primary school net enrollment ratio 15 Ratio of girls to boys in primary education 16 Student-teacher ratio in secondary education 17 Literacy rate of population, age15–24 years 18 Ratio of literate female to literate male, age 15–24 years 19 Share of women in wage employment in nonagriculture sector 20 Employment to population of working age ratio 21 Proportion of children 10–14 who are working 22 Proportion of urban population 23 Yield of vegetables 24 Yield of fruits 25 Yield of cash crops 26 Yield of fisheries 27 Yield of cereal crops 28 Yield of pulses Source: Central Bureau of Statistics 2003. Note: ARI = acute respiratory infection. Reference Nepal, Central Bureau of Statistics. 2003. “District Level Indicators of NEPAL for Monitoring Overall Development” (Based on Selected Socio-Economic Indicators). Central Bureau of Statistics, Kathmandu, Nepal. 34 Appendix D. List of Persons Met Roshan Bajracharya Sr. Economist Takuya Kamata Country Manager Bigyan Pradhan Sr. Operations Officer Mr. Prem Upadhyay Under Secretary-MoF Mr. Kamran Akbar Sr. DRM Specialist Mr. Rajib Upadhya Sr. External Affairs Officer Mr. Raj Babu Shrestha Former PAF ED Mr. Tanka Mani Sharma Secretary Office of the Prime Minister Mr. Laxam Pandey PAF M&E Chief Mrs. Indira Pzneru PAF M&E officer Dr. Mohan Man Sainju Former Board Member of PAF Mr. Shreekrishna Upadhyay Executive Chairman SAPPROS Mr. Yogendra Kumar Karki Joint Secretary Ministry of Agricultural Development Mr. Suresh Pradhan Joint Secretary Ministry of Poverty Alleviation and Cooperatives Mr. Bhim Suwal CEDA Mr.Nahakul K.C. PAF ED Mr. Arun Kumar CEDA Mr. Santos Upadhyaya CEDA Mr. Yogendra Gurung CEDA Mio Takada Rural Development Specialist PAF II TTL Mr. Satya Narayan SODEP ED Mr. Chhabi Pokher PAF Deputy ED P Mr. Moti Prasad PAF Chief Program Coordinator and Communication Division Mr. Dilip Parajuli Sr. Economist World Bank Mr. Nazmul Chaudhury Sr. Economist World Bank Ms. Rajita Dhungana Translator