A case study from Reducing Poverty, Sustaining Growth--What Works, What Doesn't, and Why A Global Exchange for Scaling Up Success Scaling Up Poverty Reduction: A Global Learning Process and Conference Shanghai, May 25­27, 2004 The Yemen Social Fund for Development Willem Struben 10047 El Capitan Real Road El Cajon, CA 92021 phone: 619-749-7372 fax: 619=749-9332 email: willem.struben@cox.net The findings, interpretations, and conclusions expressed here are those of the author(s) and do not necessarily reflect the views of the Board of Executive Directors of the World Bank or the governments they represent. The World Bank cannot guarantee the accuracy of the data included in this work. Copyright © 2004. The International Bank for Reconstruction and Development / THE WORLD BANK All rights reserved. The material in this work is copyrighted. No part of this work may be reproduced or transmitted in any form or by any means, electronic or mechanical, including photocopying, recording, or inclusion in any information storage and retrieval system, without the prior written permission of the World Bank. The World Bank encourages dissemination of its work and will normally grant permission promptly. THE YEMEN SOCIAL FUND FOR DEVELOPMENT Executive Summary With annual per capita income of barely $450 and with 42 percent of its 18 million population estimated to be living in poverty, Yemen is among the poorest countries in the world. Its social indicators reflect its income poverty. Life expectancy at birth is 56 years; the infant mortality rate is 76 per 1,000 live births; 46 percent of children under five are malnourished; and 52 percent of the population aged 15 and older are illiterate. Poverty is pervasive in rural areas. Children and women living in rural areas without access to education and health services rank highest among the vulnerable. Public expenditures in social sectors (education and health) are mildly pro-poor, but they do not adequately address the magnitude of rural-urban and gender gaps. Almost all social programs are urban-oriented and tend to benefit the better-off. Benefit-incidence analysis of safety nets shows that (i) their coverage is limited; (ii) they fail to address short-term downturns and vulnerability of the able- bodied poor; and (iii) they fail to reach the poorest and most needy, especially children. The extent and distribution of poverty has also been affected by such events as the return of almost one million Yemenis who worked in Saudi Arabia before the Gulf war, the civil conflict in 1994, and the decline in foreign assistance following the end of the Cold War. The Social Fund for Development (SFD) is a major tool being used by the government of Yemen to alleviate poverty. It was established in 1997, in the wake of a period of macroeconomic difficulties, to provide support for long-term development opportunities for the poor, and to encourage innovative and participatory approaches to delivering demand-driven social services. Objectives and approach The SFD's objectives include mobilizing communities to help themselves; building the capacity of communities, nongovernmental organizations (NGOs), and local governments to take key roles in local development; and supporting income-generating and microenterprise activities. Even when financing public works in connection with the community development program, the SFD first mobilizes communities to prioritize investments and to make contributions to investments and recurrent costs. In essence, its focus is more on helping the poor help themselves, rather than simply providing assistance. It was designed to broaden the base of beneficiaries from economic change, on the understanding that growth alone will not reduce poverty, unless the more disadvantaged sections of the population have the tools to access the benefits of growth. The SFD's approach to poverty reduction relies on three principal components: · Community development projects for small scale labor-intensive infrastructure works (water supply, sanitation, and feeder roads) and delivery of basic social services (education and health). Project selection is transparent and based on efficient use of resources and sustainability, measured as the ability to recover the costs of operation and maintenance costs. Intermediary organizations, including NGOs and local governments, are to assist communities in preparing 1 CASE STUDIES IN SCALING UP POVERTY REDUCTION and implementing projects identified by those poor communities as priorities. SFD's autonomy from the government bureaucracy has allowed it considerable flexibility in disbursement and operational procedures. · Microenterprise development through technical assistance, training, and access to credit. Initially, the focus was on capacity building of intermediaries to deliver credit, savings, and other services to the economically active poor. The microenterprise activities deliver credit, technical assistance, and training to microenterprises through NGOs. · Capacity building to assist NGOs, local communities, and the private sector in identifying, implementing, and operating SFD projects. Capacity building helps local associations and groups develop and execute projects; enhances the ability to execute sustainable microenterprise program; and provides technical assistance to small contractors and consulting firms on engineering and construction standards, bids, and small works projects. It also provides for nationwide poverty monitoring. Impact The above-mentioned activities have had a major impact. In a very difficult work environment, the first two years of the SFD were a period of intensive learning and experimenting, which set the stage for a highly productive operation. The new demand-driven concept proved successful. The community-developed infrastructure is of good quality and is provided cheaper and faster than similar works provided by line agencies. The SFD has been functioning as a source of innovation in Yemen, with successful new approaches in different sectors being applied on a larger scale by the line agencies. SFD continuously adjusts its coverage and approach based on changes in demand and lessons learned. Moreover, the demonstration impact of the SFD is remarkable--it offers a good model of an efficient and cost-effective public agency that works entirely with Yemeni staff. Income generation has been improved through additional temporary and permanent employment provided by community works projects, as well as through microenterprise development activities channeled through NGOs. The autonomous setup and a flexible, demand-driven approach to service delivery also allowed the SFD to successfully scale up its operations, as financing increased from $80 million for the first phase to $175 million in the second phase, and to about $340 million in the third phase, starting in 2004. In scaling up operations, the SFD built on experience acquired through the first phase by weeding out unsuccessful components (such as the small-enterprise promotion program), modifying ineffective practices (such as building new clinics that could not be maintained), introducing new activities in the investment menu (such as special programs for vulnerable groups and cluster-type projects) and expanding coverage of activities prioritized by local communities--particularly education, water supply, and health. The evidence points to SFD's success in reaching poor and disadvantaged groups. With almost 57 percent of total investments allocated to education projects (and representing almost 20 percent of the country's total investment expenditures in education), girls' enrollment rates and 2 THE YEMEN SOCIAL FUND FOR DEVELOPMENT net enrollment rates in primary education increased substantially. Investments in health facilities led to an increase in access to, and availability of, health services. A total of 284 water supply projects led to an increase in the proportion of households with in-house tap water and a rise in per capita consumption and supply of water, while feeder roads projects considerably reduced travel times and costs to local markets. In terms of microcredit, the SFD was instrumental in developing the foundations for an otherwise nonexistent microenterprise industry, with the United Nations Capital Development Fund selecting it as a partner for the implementation of its renowned Micro Start program in Yemen. Finally, capacity-building activities helped disseminate the community-based approach of the SFD and strengthen community participation and ownership of proposed projects. Capacity building extended to SFD staff as well, ensuring a gender focus in SFD projects and improving its poverty targeting and resource allocation criteria. Success factors Several factors have contributed to SFD's success: · (1) Political commitment of the president and senior leadership, which ensured SFD's autonomy and enabled it to recruit qualified and experienced staff · Learning and experimentation, building on the experience and best practice of other social funds and the ability to experiment with new approaches and adjust existing practices · Institutional innovation, including the use of a demand-driven approach to service delivery, the contracting of well-paid staff on fixed appointments, the use of a variety of intermediaries (especially NGOs and local small-scale contractors), and the involvement of all stakeholders in policy and operational decisions · Funding and technical assistance from well-coordinated external donors. The SFD is playing a critically important role in helping alleviate Yemen's major social problems. Through its demand-driven approach, it has introduced new concepts in community participation--among them the involvement of women. It has contributed much to improving access to priority social and economic infrastructure and microcredit, and used NGOs and other intermediaries to the maximum extent. In the process, it has helped build up the implementation capacity of the poor beneficiary communities, NGOs, contractors, and government agencies. With recent developments in decentralization in the country, SFD will now have to also ensure the participation of local governments and councils in the process of local development. The SFD has rapidly adjusted to new needs in the past, and its highly experienced and professional management and staff can be expected to continue to do that. It has proved to be a useful adjunct to existing line agencies in a range of poverty-relevant areas. Indeed, some of the approaches to service delivery that it has pioneered in the Yemeni context could have positive spillover or demonstration effects on other government agencies. 3 CASE STUDIES IN SCALING UP POVERTY REDUCTION WITH an annual per capita income of barely $450 and with 42 percent of its 18 million population estimated to be living in poverty, Yemen is among the poorest countries in the world. Its social indicators also reflect its income poverty. Life expectancy at birth is 56 years, the infant mortality is 76 per 1,000 live births, child malnutrition is 46 percent of children under five, and illiteracy is 52 percent of the population of fifteen and older. The challenge of reducing poverty in Yemen is enormous. A sense of this is provided by the following observations: · Poverty is widespread nationally and pervasive in rural areas. · Factors that affect the risk of being poor in Yemen are (i) lack of education, (b) large household size with a large number of children, (iii) geographical location, and (iv) lack of worker remittances from abroad. · Children and women living in rural areas without access to education and health services rank highest among the vulnerable. · Public expenditures in social sectors (education and health) are mildly pro-poor, but they do not address the magnitude of rural-urban and gender gaps. Many social programs are urban- biased and tend to benefit the better-off. · Benefit-incidence analysis of the safety nets shows that (i) their coverage is limited, (ii) they fail to address short-term downturns and vulnerability of the able-bodied poor, and (iii) they fail to reach the poorest and most-needy, especially children. · The extent and distribution of poverty has also been affected by such events as the return of almost one million Yemenis who worked in Saudi Arabia before the Gulf war, the civil conflict in 1994, the decline in foreign assistance following the end of the Cold War, and the fall-out of 9/11. The Social Fund for Development (SFD) In the early 1990s, Yemen witnessed a macroeconomic crisis that featured high inflation and required a drastic devaluation. To deal with this crisis, with support from the IMF and the World Bank, the Government embarked in the mid-1990s on a program of economic stabilization followed by structural adjustment. These efforts did have positive results. However, they also led to sharp reductions in food subsidies, while the persistence of high inflation also significantly reduced the purchasing power of the poor. Therefore, the Government, with the support of donors, launched three programs aimed at providing a cushion for the poor. These were (i) the Social Welfare Fund (SWF) to provide cash transfers to those unable to work and to members of the poorest sections of the Yemeni society, (ii) the Public Works Project (PWP) designed to create short-term employment opportunities for those able to work, while also upgrading the social infrastructure base in Yemen and (iii) the Social Fund for Development to provide support for long-term development opportunities for the poor, and establish new innovative and participative approaches in delivering social services. 4 THE YEMEN SOCIAL FUND FOR DEVELOPMENT The SWF has been modest in terms of its reach out and ability to target the most disadvantaged groups. It faced issues related to identification of beneficiaries and delivery of cash. During the past six years, the SFD has provided technical support in these areas to the SWF. The intervention had positive impacts on the performance of the SWF. The PWP developed to be a cost effective and efficient mechanism for providing social infrastructure and contributed to increasing the access of rural population to social services. It has also been successful in generating income for the poor in the areas in has been active. A second phase of the program was supported, and a third phase is starting in 2004 parallel to the SFD's third phase. The SFD had the broadest objectives among the three programs, including mobilizing communities to help themselves, building the capacity of communities, local NGOs, and local governments to take key roles in local development, and supporting income-generating and micro-enterprise activities. Even while financing public works in connection with the community development program, the SFD first mobilizes communities to prioritize investments and to make contributions to investments and recurrent costs. In essence, its focus is more on helping the poor help themselves, rather than simply providing assistance. It was designed to broaden the base of beneficiaries from economic change, on the understanding that growth alone will not reduce poverty, unless the more disadvantaged sections of the population have the tools to access the benefits of growth. The SFD was established under Law No. 10 of 1997, which provided for a great degree of autonomy. This in turn allowed for the recruitment of highly-qualified and experienced managers and staff, being paid well above civil service salaries. It also allowed for the establishment of a special account in a commercial bank, providing the SFD with ready cash for the fast payment of its contractors and consultants. The Board was chaired by the Prime Minister, and also included the Ministers of the main social sector line agencies, as well as representatives of NGOs. After substantial preparation, the SFD started operations in 1997, and quickly established itself as an innovative, flexible organization, managed like a private firm. Approach and Content of the SFD The SFD's poverty reduction efforts focus on (a) building up social and economic infrastructure and services, especially in rural areas, as well as the generation of employment and income, all aimed at reducing the vulnerability of the poor in the depressed economic environment, and (b) strengthening communities and NGOs capacities to take chare of their local development. To achieve this, the SFD implements three kinds of activities: · Community development projects (mainly in education, health, water supply, sanitation, and feeder roads) for small-scale labor-intensive works and the delivery of community services. Project selection is transparent and based on efficient use of resources and sustainability, measured as the ability to recover operations and maintenance costs. Intermediary 5 CASE STUDIES IN SCALING UP POVERTY REDUCTION organizations, including NGOs and local governments are to assist communities in preparing and implementing projects identified by those poor communities as priorities; · Micro-enterprise development through technical assistance, training, and access to credit. Initially, the focus was on capacity building of intermediaries to deliver credit, savings, and other services to the economically active poor. The micro-enterprise activities are to deliver credit, technical assistance, and training to micro-enterprises through NGOs; and · Capacity building is to assist stakeholders such as communities, NGOs, the private sector, and the SFD in improving their effectiveness in identifying, implementing, and operating projects. It helps local associations and groups develop and execute projects, enhances the ability to execute sustainable micro-enterprise programs, and provides technical assistance to small contractors and consulting firms on engineering and construction standards, bids, and the undertaking of small works. It also provides for nation-wide poverty monitoring. Much has been achieved since the start of these activities. As of July 2003, the SFD had committed a total of US$213 million for 3,035 projects, of which 1,686 projects amounting to US$94 million had been completed, 1,232 projects amounting to US$111 million were under implementation, and 117 projects amounting to US$8 million had been approved pending implementation. Details of the completed projects are shown in the table below: Table 1. Number and Cost of Completed SFD Projects Cost Average Cost Sector Number (US$'000) (US$) Education 888 53,452 60,200 Water Supply 284 16,743 58,900 Health 177 9,360 52,900 Feeder Roads 32 2,984 93,200 Special Needs 41 2,416 58,900 Environment 41 4,273 60,300 Cultural Heritage 8 654 81,700 Micro Credit 26 1,417 54,500 Organ. Support 86 1,234 14,300 Training 96 1,166 12,100 Integrated Interventions 5 111 22,200 Small Enterprises 2 14 7,000 Total 1,686 93,827 55,600 The table shows that the SFD's main commitment, in response to the priorities expressed by the communities, has been in education with 53 percent of the projects and 57 percent of total investment, followed by water supply with 17 percent of the projects and 18 percent of total investment. Some sector-specific observations follow: Education: This sector constitutes over half of the SFD financial resources and this is consistent with the overall country's priorities and demand of local communities. Although the main activity in this sector was the construction & rehabilitation of school buildings, it was 6 THE YEMEN SOCIAL FUND FOR DEVELOPMENT carefully designed to achieve the desired impact and at the same time to ensure sustainable operation. Coordination with the Ministry of Education (MOE) is achieved at two levels: (a) at the central level, where the Minister of Education is a member of the SFD Board of Directors, which is responsible for approving annual plans as well as reviewing the performance at the level of each sector; and (b) at the level of MOE offices, where each office approves every individual project. The project agreement is signed by the community representative, the Ministry of Education office and the Social Fund for Development. In that agreement, the MOE office agrees to provide the staff, supply and supervision. Once the project is completed, it is hand-over to the MOE office. The SFD continues to have a special focus on girls' education. The interventions in girls' education do not only cover construction of schools, but also increase the demand for schooling. For attaining this goal, the SFD is carefully preparing a special program that deals with different factors affecting enrollments such as roads, water facilities and availability of female teachers. Box 1. The SFD Focus on Girls' In addition, a number of projects Education sponsored by NGOs providing literacy classes were also supported. The SFD has also started A significant amount of work is usually involved to engage in pilot programs to improve the during the preparation of school projects. Office and quality of education and explore interventions field screening is carried out. The SFD benefited in preschool education, as well as support the from the data collected at the national level and capacity of the Ministry of Education in the linked to SFD Geographical Information System. decentralization process and the development The SFD is targeting areas where there is a demand for education, and the process starts out under a tree and execution of education policies. or in a mosque with a budget allocated by the Health: The 177 health projects were Ministry of Education. As the impact evaluation initially mainly for the establishment of new study concluded, providing the proper structure has health services in deprived areas, but later also increased the enrolment of children in schools and included expansion and improvement of improved net enrolment rates (age-specific grade educational training facilities. Bad experience enrolment), especially for girls. with operation and maintenance of completed works caused the SFD to focus on ensuring the availability of health workers, especially women in remote rural areas. This also meant the rehabilitation and equipment of existing health facilities, rather than building new ones where operational needs are not yet ascertained. Interventions in new areas also are phased; medical services start in temporary premises with SFD providing minimal equipment, and only start construction after those services are functioning. Health committee members are officially responsible for operation and maintenance of completed facilities. Water Supply: The 284 water supply projects are mainly in rural areas, using surface water resources, rather than ground water resources, to avoid depletion of the latter. After a few years, the SFD also shifted from mechanized water supply systems to non-mechanized systems (mechanized systems proved expensive to operate, were risky in the absence of regulations on the 7 CASE STUDIES IN SCALING UP POVERTY REDUCTION digging of boreholes, and might not be sustainable). In addition, the SFD designed a special program for water harvesting systems to expand the rehabilitation of traditional cisterns and the construction of new ones to increase water supply in drought-affected areas. At the same time, water quality was improved through fencing of cisterns, and the introduction of slow sand filters and hand pumps. The awareness component is becoming part of water projects to provide communities with knowledge in how to deal with water borne diseases. The establishment of water user committees is an official requirement. A recent study found that 72 percent of the installed pumps are still in good condition. Payment for water is arranged for in 42 percent of newly completed works. Where established, payments are timely in slightly less than two-thirds of the sub-projects. More than four-fifths of the sub-projects have community water committees. Box 2. Building a School in a Tribal Area The Qaflat Uthr district, located in the northwestern part of the Amran governorate, faces major hardships, including violence, tribal wars, and high illiteracy. Moreover, the Al-Ma'aresh region, the area selected fro an SFD school project, is highly tribal. It has been in continuous conflict with other tribes, and is considered very complicated from a tribal and social viewpoint. Because the community did not trust any other entities, at first it was not responsive to SFD initiatives. Thus, the SFD decided to delay its interventions in the area. But, recently, after the community observed the SFD's credibility in implementing projects in a nearby area, community leaders visited the SFD office in Amran, and requested a school project as a priority for their area. It was agreed that four classrooms, along with supporting facilities, would be constructed. Following approval by the SFD Project Committee, the assigned team, along with a project officer, visited the project site to form community committees, and to raise awareness about the importance of community contributions. The community agreed to the contributions it would have to make, including preparing the site, providing some funding for the project, and installing a water tank at the project site. Just one week after the community was informed about the required contributions, students' parents started distributing lists of the amounts needed, and collecting the money. Because community members traditionally own weapons, many parents who could not afford to pay cash offered their rifles to the project's accountant and manager as "pledge for fulfillment". After a week, the equivalent of US$2,300 was collected. In recognition, a list of the contributors and the amounts they paid was put in the project site. Feeder Roads: The 32 feeder road and street pavement projects are mainly in poor rural areas. They use labor-intensive methods, and require the establishment of a road user committee, as well as a maintenance plan. Feeder road projects cause immediate positive impacts in terms of mountainous communities access to basic commodities and services. A recent study identified communities where basic food commodity prices fell by 50 percent after the feeder road was built. Since feeder roads now are part of the Local Council jurisdiction, close cooperation with the Councils is being sought on operation and maintenance. Environment: The 41 environmental protection projects include sanitation projects targeting poor urban and semi-urban areas, awareness activities, garbage collection, and terrace rehabilitation. Simple and natural treatment technologies are adopted for treating wastewater. In some urban areas with sufficient treatment capacity, new networks are connected to the main sewerage network, in cooperation with local authorities such as the National Water and Sanitation Authority. 8 THE YEMEN SOCIAL FUND FOR DEVELOPMENT Box 3. Pilot Project for Fog Water Collection Fog is a natural phenomenon in some regions of Yemen, such as the governorates of Hajjah, Al -Mahweet and Al-Maharah and on Socotra Island. Fog forms in these regions more frequently between December and March, which is also when droughts occur. Developments in meteorology and in the manufacture of high- quality materials capable of collecting fog water drops have helped make fog a major source of water in such areas. A fog water collector consists of a frame made of aluminum or iron and a polypropylene net sewn into the frame. The collector is installed facing the direction of the wind, and gathers fog water in a channel which is then drained into a collection tank. In December 2002, an SFD pilot project installed 30 standard collectors in different parts of Hajjah, in addition to a large collector. The local community was involved in the entire process, and trained in collector installation and data gathering. The daily gathered data included the quantity of collected water and the wind velocity and direction. A collector's productivity depends on several factors, including its height above sea level and the local temperature, wind velocity and humidity. Production can reach 30 liters per day per standard collector, especially if fog is accompanied by rain. A site is considered feasible for fog collection if a standard collector can gather a minimum of 5 liters per day. Special Needs Groups: As of 2000, the SFD has also completed 41 projects aimed at improving facilities and services for the disabled, children at risk (orphans, children in conflict with the law, street children, children of imprisoned women), women at risk (women in prison or coming out of prison), and the socially-marginalized (mentally ill, elderly, returnees, shanty dwellers). Interventions are decided based on the capacity of the implementing organization, the priority of the targeted population group, and the sustainability prospects. Box 4. Safe Childhood Center in Sana'a The Safe Childhood Center in Sana'a , established in January 2000 by the Ministry of Social Affairs and Labor, is a home and rehabilitation center for male street children under 15. The SFD has provided the center with furniture and equipment, developed its administrative and financial systems, and provided technical support for its activities. The SFD is also managing the financial support provided to the center by the Sharjah Charitable Society. The center's main goals are to protect its residents, provide them with social, educational, and vocational rehabilitation, and return them to their homes. It offers education, health, and social services, and helps the boys adjust their behavior and attitudes, so that they can integrate with mainstream society. The center also offers follow-up care, checking on the status of boys after they leave. This care can take the form of staff visits to boys' households and schools, provision of school uniforms, and monthly subsidies (cash, in kind, or both), or monitoring of how the boys are behaving and treated in the schools they attend. Mohamed A., a formed resident of the center, could not succeed in school. But he showed interest and talent in repairing watches. Therefore, the center funded and supervised a project designed for him in the Amran governorate, where he came from. Now he is working in watch repair, and supporting himself and his family. Ahmad H. A., in third grade, has become cheerful and social since moving into the center. He now refuses to beg "to preserve dignity". When asked what dignity means to him, hew instantly replied: "Dignity means that I respect you and you respect me, doesn't it?" At school, Ahmad recently received a yellow belt in karate. Cultural Heritage: This activity was added in 2000. The first eight completed projects are located in Sana'a, and include gardens and street pavements in the "old city", as well as the 9 CASE STUDIES IN SCALING UP POVERTY REDUCTION expansion and restoration of the National Museum. The projects emphasize employment creation for local artisans and the use of local materials and traditional techniques. They also aim at encouraging a new generation of artisans, who specialize in traditional techniques of rehabilitating old buildings, thus preserving the country's cultural assets. Integrated Interventions: The ongoing pilot program in four different geographic areas also started in 2000, and aims at testing a more integrated ("cluster") approach to increase the impact of the SFD interventions. This includes the establishment of development committees, intensive training of those committees in planning and implementation, the formulation of development plans by the committees, and the implementation of successive sub-projects according to the agreed priorities. Achieving genuine women's participation is being pursued actively in these clusters. Micro-Credit: When the project started, there was no experience with micro-finance in the country whatsoever. The SFD had not only to build capacity of micro-finance intermediaries, but also its own capacity. Eligible intermediaries received a SFD loan for on-lending (instilling financial discipline right from the start) and a grant for operational cost coverage until break- even. It is worth noting that borrowing is not common in Yemen, and because of traditional and religious reasons, there are no moneylenders and people depend mainly or family and their own savings. This put an additional barrier to the development of microfinance in Yemen as the SFD has to promote these new services. The SFD helped to implement several pilots to experiment with the provision of microfinance and income generating activities. The SFD learned from these experiences and is currently supporting several microfinance initiatives that are potential breakthrough programs. One of these potential breakthroughs is the program run by several NGOs in Aden. The programs were developed by the clients on their own. Women, who had participated in an earlier closed- down savings and credit program, formed their own NGOs, and continued forming groups and saving without SFD support. After they had done so with success for over a year, the SFD started supporting these newly created NGOs. Another program is the savings-and credit program that uses the ASA methodology (a well-known Bangladeshi micro-finance provider, which assisted the SFD). Currently, this program has five branches with over 4,200 savers and over 1,500 borrowers, all women. In addition to the potential microfinance winners, the SFD was able to promote best practices in microfinance and prove that microfinance could be done successfully in Yemen. The SFD is also playing an important role in developing local capacity that will help the industry grow in the future. The SFD has developed several training modules and tools that suit local conditions and is continuing this task to develop local capacity in microfinance. Capacity Building: These activities got off to a slow start. By now, however, the number and variety of training activities for a range of stakeholders is quite impressive. The approach to community capacity building was developed gradually. It started with providing training to SFD staff and consultants in participatory development and methods to 10 THE YEMEN SOCIAL FUND FOR DEVELOPMENT identify priorities in collaboration with the communities, the establishment of local community committees, and their training in sub-project implementation and operation, including basic finance and management. Community committees are established for most projects, and receive basic training. A great number receive additional training, including for the integrated pilot programs and on health and environmental issues for water harvesting schemes. The NGOs and CBOs, which receive training, are selected based on agreed criteria, giving priority to those that have been successful in securing SFD funding for their projects. In addition, priority is given to CBOs in the poorest areas, which would not otherwise be able to prepare project proposals. Training depends on the specific project needs; most projects receive training in bookkeeping and resource management. During the first phase of the SFD operation, priority was given to training of trainers, refinement of training methods, development of training modules, and the implementation of a select number of well-defined capacity building projects. In addition, organizational support was provided to increase the capacities of intermediary structures. NGOs were assisted through training and/or material support, and received training in a range of subjects (e.g., literacy programs, handicraft production, home economics, and basic accounting for agricultural cooperatives). Private consultants and contractors have received, and continue to receive, substantial training, including in water harvesting schemes and community participation techniques. Government agencies too have received, and continue to receive, SFD support. The Ministry of Education has been supported for selected activities within the planning department as well as literacy activities; the Ministry of Health through supporting the training of health personnel (150 traditional midwives and 160 scholarships for paramedics at the Health Training Institute) on introducing management and technical issues, as well as on pro-poor cost recovery systems in SFD supported projects with outreach in remote areas; and the Ministry of Social Affairs and Labor by a variety of support, in particular to its Social Welfare Fund (organization and management of cash transfer programs) and the National Program for Productive Families. The SFD Capacity Evolution and Scaling Up Over the last 7 years the SFD has been gradually building up its capacity and the number of sub- projects generated per year has been increasing. The following table compares the three phases of the SFD (with the third about to start) in terms of amount of funding, number of donors supporting, and numbers of projects supported (figures for phase III are estimates). Table 2: Evolution of SFD Activities, 1997-2008 Phase I (1997-2000) Phase II (2001-2003) Phase III (2004-2008) Funding (US$ Million) 80 145 400 Number of donors 4 7 10 Number of projects 1086 2000 4000 Community based operations of the type of the SFD are always found to be a challenge to scale up, given the extensive time and effort that are needed to mobilize communities and ensure 11 CASE STUDIES IN SCALING UP POVERTY REDUCTION their ownership to the project in question. The following three elements explain the SFD's ability to considerably expand its capacity to support local development projects and scale up its impact: Contracting out: to ensure that it continues to keep a lean structure and at the same time expand its capacity to support local development projects, the SFD has invested great deal of resources, time and effort in training local consultants to identify, appraise and evaluate local development projects. A large number of young local consultants were trained in areas like Participatory Rapid Appraisals, social assessments, and community mobilization during the first three years of the SFD operation. They also received training that familiarized them with the SFD's project cycle, and its operational manual. The SFD now draws on the capacity of these consultants to respond to the increasing number of project requests it receives. These consultants are categorized by sector and geographically and are easy to mobilize and, because of the training received, provide the SFD with good quality reports based on which funding decisions are taken. It should be noted that, local consultants capacity building was a gradual process. While an experienced consultant who has demonstrated good performance submits reports that are rarely verified, those who are starting to work with the SFD usually submit reports that are usually verified, in the field, by SFD staff (which is part of the capacity building of the local consultant). Decentralization: to regional offices of the SFD. While the SFD started with its central office in Sana'a, over the years it established seven regional offices and invested in the training and capacity building of the staff of these offices. Once the main institutional tools, like the administrative, finance and management information systems were developed, responsibilities have gradually been shifted from the center to the offices. The SFD did not follow a blue print in going though this decentralization process. Those offices that showed increased capacity, were delegated more responsibilities. Now most identification, appraisal and implementation issues, including recruitment of local consultants take place at the regional office levels. Financial transaction have been kept at the center. Focus on institutional capacity: related to the two issues mentioned above, the SFD would not have been able to contract out, and decentralized if it had not first focused on developing systems and building the capacity at the center. This included the development of the state-of-the art MIS that has facilitated the decentralization to regional offices, and reporting to the donors which increased the SFD's credibility. It also included the development of a strong financial system that was able to provide the complex balance of fiduciary control with flexibility and speed. Finally the continued investment in staff training ensure that staff's capacity to handle project increase gradually until it reaches its threshold. Impact of the SFD The above-mentioned activities have had a major impact. In a very difficult work environment, the first two years of the SFD were a period of intensive learning and experimenting, which set the stage for a highly productive operation. The new demand-driven concept proved successful, and the community development infrastructure is of good quality, and is provided cheaper and faster than similar works provided by line agencies. The SFD has been functioning as a source of 12 THE YEMEN SOCIAL FUND FOR DEVELOPMENT innovation in Yemen, with new approaches in different sectors applied on a larger scale by the line agencies when successful, and continuously adjusting its coverage and approach based on changes in demand and lessons learned. Moreover, the demonstration impact of the SFD is remarkable. It gives a good model of an efficient and cost effective public sector agency that works entirely with Yemeni staff. Income generation has been improved through additional temporary and permanent employment provided by the community works construction activities, as well as through micro- enterprise development activities channeled through NGOs. Although the SFD aims at labor- intensive methods, those are not used as selection criteria for contractors. Still, several factors ensure labor intensity, including the prohibition of machine-cut stones, ready-mix concrete, and heavy machinery. Evidence as to the impact of the Community Development and Capacity Building activities is provided in the December 2002 consultant report on Findings from the 1999 National Poverty Survey and the September 2003 consultant report on Key Findings-Yemen SFD Impact Evaluation. The Small and Micro-Enterprise Component was assessed and described in the November 2000 consultant report on MSE Baseline Survey-Yemen as well as in the September 2003 report. These reports show the following: Poverty Targeting. No agreed formula existed during the first year, but target areas were prioritized in 1998 by using an index of relative poverty based on data from the 1994 Census and the 1998 Household Budget Survey. Officially, a two-stage procedure was designed by SFD for its resource allocation. First, an allocation by governorate, based on population (75 percent) and poverty index (25 percent), and second by district level, based on the same criteria. In practice, there was no strict allocation rule that generated entitlements to resources by the local governments. The SFD's management had considerable discretion in approving projects, using a sophisticated data system on available services at the community level. Initially, this resulted in a regressive allocation of resources. The 1999 data show that "SFD's funding is distributed very closely in proportion to the total population of each governorate, but it is not proportional to the burden of poverty carried by each governorate." The first study also carried out a household level analysis of targeting, and found that SFD's resource allocation was more progressive than indicated by the district level data; each of the two poorest deciles received more than their pro- rata share of resources. However, the same applied to the two richest deciles, and this confirmed that the SFD had "no clear overall tendency to favor the poor over the less poor in its resource allocation". The first study added that, "the findings should be interpreted in the context of Yemen's very high level of poverty and relatively flat income distribution", making "leakage" to the relatively better-off less problematic than in countries with a skewed income distribution. Poverty targeting improved substantially with the SFD's decision in 2000 to change the targeting formula: 70 percent was to continue to be allocated to governorates and districts (based on number of households living in poverty both in terms of consumption and access to services) and 30 percent was to be allocated to supply-driven special programs and for socially- marginalized groups. 13 CASE STUDIES IN SCALING UP POVERTY REDUCTION As shown in table 3, the data from the second (2003) study show a much improved picture, with a strongly progressive allocation of resources by level of poverty of the households. The data reflect the outcome of 198 sub-projects in education, health, water supply, rural roads, and micro-credit, and show that a high proportion of SFD resources are benefiting the poorest households, with 17 percent going to the poorest decile, 31 percent to the poorest quintile, and 44 percent to the poorest three deciles. This is much better than in other social funds where this analysis was carried out. And it is especially impressive in Yemen, because of the earlier regressive allocation, and because the high level of poverty across the country might have been expected to make it more difficult to target relatively poor households. It also should be noted that this was consistent across sub-project types, except for micro-credit (however, the latter is rarely aimed at the poorest. In addition, it had a small weight in the total distribution). It is always difficult to show the impact of microfinance on the poor. However, the SFD was able to develop the capacity of several NGOs that provide services to the poor. Initially the SFD focused on the provision of credit only. Later the SFD promoted savings and credit which attracted more women and currently the majority of microfinance clients are actually women. Taking the average loan size as an indicator of client poverty, we find that the SFD financed programs whose average loan sizes decreased over time, indicating better targeting of the poor. Table 3. Household Level Targeting Outcomes Poverty Decile Population (percent) Resources (percent) Population (cumul.) Resources (cumul.) 1 10.0 16.6 10.0 16.6 2 10.0 14.5 20.0 31.1 3 10.0 12.6 30.0 43.7 4 10.0 10.7 40.0 54.4 5 10.0 11.4 50.0 65.8 6 10.0 7.4 60.0 73.2 7 10.0 8.2 70.0 81.4 8 10.0 8.3 80.0 89.7 9 10.0 6.0 90.0 95.7 10 10.0 4.3 100.0 100.0 Source: Key Findings: Yemen SFD Impact Evaluation. Sector Impact. As mentioned before, the SFD supported a great number of projects, mainly in education, health, water supply, and rural roads, with a substantial sector impact. In 2001, the SFD's share of the country's total investment expenditures in education was 19.5 percent and in health 8.4 percent, up from 3.5 percent and 1.3 percent, respectively, in 1998. The earlier-mentioned studies provided relevant information on this subject, including that these four sectors also had the most beneficiaries; together they generated 79 percent of all direct beneficiaries, a figure which goes to 97 percent if environment sub-projects are included. Indirect beneficiaries only add 19 percent to the number of direct beneficiaries. Gender distribution showed more or less equal shares. Sector-specific findings are listed below: Education: The 1999 data show that, by the end of 1999, the program already had a clearly positive impact on girls' enrollment and grade-for-age attainment in primary education. In 14 THE YEMEN SOCIAL FUND FOR DEVELOPMENT the beneficiary communities, girls' enrollment rates were almost 20 percent higher than in the comparison group, up from 41 percent to 48 percent. The 2003 data confirm that, as a result of the SFD investments, the number of students increased by 29 percent, with the highest increase (38 percent) for female students. The increase in number of classrooms was 12 percent (29 percent for female student classrooms). The average class size rose by 38 percent (only 8 percent for female students). And the proportion of girls enrolled increased from 42 percent in 1999 to 56 percent in 2003. Over the same period, male student enrollment increased from 76 percent to 78 percent. As a result, for all students, enrollment went up from 60 percent to 68 percent, confirming the major impact of the SFD on this important point. Health: In the early stages, health sub-projects focused on building health and primary care facilities. However, it proved difficult to ensure continued support of these facilities in terms of staff and supplies. The objective of this component was subsequently modified from construction of health facilities into training of health personnel and promotion of efforts to improve utilization of existing facilities. This demonstrates the flexibility of SFD management in adapting program components in response to local needs and conditions, ultimately enhancing the effectiveness and outreach of its service delivery. The 2003 Impact Evaluation Study showed that, in those health facilities that were completed, the proportion of male and female patients who managed to get health care services for their illness rose from 55 percent to 70 percent, which is evidence of the impact of SFD investment in health facilities. The data also showed that the presence of physicians and/or midwives was associated with a 10-30 percent increase in the availability of specific services. Drug supplies too had a positive impact on patient attendance, but clinics had, on average, only about 55 percent of the standard supplies. Surprisingly, the lack of medicine did not seem to be a major concern of the patients. Water Supply: The SFD 2003 Impact Evaluation Study suggested a very positive impact from coverage of water systems. The projects, on average, benefited 9.8 communities with an increase in the number of people covered from 1,251 to 3,251, and a corresponding increase in the number of households covered from 109 to 328. The SFD interventions also led to an increase in per capita consumption of water and frequency of supply and to lower travel times to get water in the dry season for beneficiary communities than for the comparison group. However, there is no measurable impact on diarrhea incidence. Feeder Roads: The 2003 data showed that projects, on average, served 19.5 communities and about 17,000 people, and had a quite positive impact on travel times, also resulting in a cut in travel costs; both declined by on average 40 percent. This has also led to a sharp increase in the number of trips per day, which is up, on average, by 180 percent, much of it reflecting additional visits to market local products in town. Micro-Credit: Before the SFD started operating, there was no formal micro-finance sector in Yemen. Informal micro-finance was limited to supplier-credit. Moneylenders were and are virtually non-existent, due to religious reasons. The SFD helped establish the foundation of an emerging micro-finance industry, and create awareness among policy makers on issues such as interest rates, sustainability, and best practices. In the early years, the SFD mainly focused on experimenting with different micro-finance models, creating a knowledge base of what works and 15 CASE STUDIES IN SCALING UP POVERTY REDUCTION what does not work. The key role of the SFD in developing the foundation of a micro-finance industry was also recognized by other donors; recently, the UN Capital Development Fund (UNCDF) selected the SFD as its partner in the implementation of the Micro Start program, which has been highly successful around the world in supporting so-called "break-through" micro-finance organizations. GTZ and KfW have also started discussing possible cooperation with the SFD. Box 5. Benefits of a Microfinance In all, it should be noted that (i) the Program in Hays SFD did create awareness and understanding among policy makers of the importance of Salamah Hali, a woman in Al-Qataba village in the micro-finance and best practices, (ii) the coastal region of Al-Khawkhah, shows the benefits SFD did develop micro-finance models that micro-finance can deliver to small entrepreneurs. Salamah has five children and a husband who cannot appropriate for the country, (iii) progress work. She used to charge small fees to teach her was made with sustainable intermediaries, neighbors how to read and write, but she did not earn but more time and capacity-building efforts enough to support her family. are required, and (iv) over 13,000 borrowers Recognizing the opportunities offered by financial services, in 1999, Salamah opened a savings account and 2,500 voluntary savers were reached. with a savings and credit program in Hays. After six months, she received a loan of 6,000 Rial (about Capacity Building: As mentioned US$35), which she used to buy hand decorating before, the capacity building efforts had a supplies (used by women). After paying off the first slow start, mainly because the demand- loan, Salamah received a second one of 20,000 Rial driven, community-based concept was new (about US$120), which she used to buy more hand decorating supplies, as well as incense making to the country, and initially little or no supplies. She sold the incense for twice what she paid experience could be found in SFD, nor in for the supplies, and paid back the second loan. the country's NGOs. However, increasing Salamah recently expanded her business by selling clothes, and she plans to open a small store in the area. training of stakeholders took place, with capacity building of the beneficiary communities, as well as of participating NGOs, contractors, and central government officials. Knowledge on this subject kept evolving. For example, SFD staff and consultants learned on-the- job about the best ways to include women and tribal communities, and thus to ensure a more sustainable process. Apart from community training in project identification and community responsibilities, training did also cover the needs of operation and maintenance of completed works, and its handling by project committees (e.g., road and water user committees). Institutional development of micro-finance intermediaries is a time-consuming process, and CGAP standards show that it takes three to five years to reach operational sustainability and another two years to reach financial and institutional sustainability. In Yemen, it could well take much longer for micro-finance intermediaries to become financially and institutionally sustainable, as the absorption capacity is weak. For instance, CGAP standards on issues like financial analysis and ratio analysis were introduced right from the start, but most of the emerging micro-finance programs seem to latch on only now to the implementation of these standards. Still, at least four of the SFD's micro-finance programs are likely to become sustainable in the next phase. 16 THE YEMEN SOCIAL FUND FOR DEVELOPMENT The small local contracting industry and consultants have improved considerably. Contractors have been exposed to the SFD's strict rules on design standards, procurement, and funding, and have gained much experience. Consultants too were associated with various studies, and gained additional experience and knowledge. The SFD has developed into a major instrument to reduce poverty in the country, with the community-based approach, its private sector, performance-based mechanism, and its transparency. In the process, it also has had a tangible impact on the line agencies. Sector Policies and Operations: As mentioned before, the SFD is an integral part of the Government's social sector policy and poverty reduction efforts. It works in close association with relevant sector agencies, and has helped to introduce new concepts of community participation and project operation and maintenance. In addition, the SFD started some innovative, NGO-based social program activities, dealing with the needs of the most vulnerable groups. Main Operational Features of the SFD Various implementation issues are worth mentioning: Community Participation and Contribution: Although the concept of self help traditionally existed in the country, it had died over time. The SFD reintroduced and promoted the spirit of self initiatives through substantial involvement of the communities, and all projects are now supposed to have community or user committees, to be trained by the SFD. In more recent years, the SFD also introduced, on a pilot basis, the contracting of projects by the communities themselves, thus giving them responsibility for the full project cycle. However, this effort is still ongoing, and not yet replicated on a large scale. The 2003 impact evaluation also confirmed substantial community contributions in money, labor, or construction materials. The contribution of labor was highest for rural roads. Community contributions as a percentage of completed sub-project costs were estimated at 6.9 percent for education, 7.3 percent for health, 16.5 percent for water supply, and 11.3 percent for rural roads, well above the SFD minimum requirements. Gender: The impact evaluation also confirmed the major share of women beneficiaries, with the number of female beneficiaries almost equal to that of males. Cooperation with NGOs: The SFD has recognized the weak capacity of NGOs and cooperatives and provided for major training and procurement support to ensure adequate involvement of these organizations. The new Law on NGOs, designed in part with Bank and SFD assistance and passed in 2001, very much facilitates NGO operations. 17 CASE STUDIES IN SCALING UP POVERTY REDUCTION Box 6. NGO Cooperation in Dhi'Ataa Role of Local Governments: The SFD has started to develop programs to Village strengthen local governments, following the Yemen faces great challenges in mobilizing and introduction of the Local Authorities Law in supporting health workers, particularly in rural 2000, and the subsequent election of regions, though some health projects in those regions have been able to sustain service provision. The health Governorate and District Local Councils. unit in Dhi'Ataa village, established by the SFD, is The SFD's modus operandi (transparent one of the earliest for which the SFD provided procedures, participatory process, support. The contract was signed with the participation of the Al-Fallah Agricultural institutional arrangements, and close Cooperative. Dhi'Ataa is 40 kilometers from the city collaboration with all stakeholders) has of Yarim, which had been the nearest location for already assisted in laying important building health services. The cooperative was able to ensure the availability of health workers by using money blocks to advance the decentralization from health service fees, thus ensuring the process. sustainability of the unit and the provision of primary health care for the village and surrounding regions. Technical Quality and The project exemplifies the partnership between the Sustainability: The 2003 impact evaluation SFD, NGOs, and the Ministry of Health. confirmed that projects were generally of reasonable quality, and well operated and maintained, although this varied by sector. Operation and maintenance is benefiting considerably from Yemen's tradition of strong community organizational structures and major community participation. The project agreement between SFD, the line agency, and the community has explicit statements on community rights regarding participation and responsibilities regarding contributions, and the SFD is nowadays providing adequate training of NGOs and community project committees in these subjects. Community ownership will be further promoted by the above-mentioned gradual introduction of fully community-managed sub-projects, including procurement, and this should further strengthen the sense of community ownership and, thus, overcome ongoing line agency weaknesses in following up on their operation and maintenance commitments. It should also be noted that attention has been paid to appropriate design standards (e.g., for schools, designs differ by region; for feeder roads, design standards were developed; and for water supply, mechanical components have been kept to a minimum to reduce the maintenance burden) and the provision of substantial training in operation and maintenance, especially for water supply projects. Completed projects are all visited after handing-over, to check that all components are working according to plan, and that corrective action is taken to overcome remaining organizational or technical problems. In the case of water supply and rural road projects, the 2003 impact evaluation showed some interesting results. For water, the upgrading of the piped systems has led to an increase in their operating costs by about 200 percent on average, due to the use and maintenance of electromechanical pumps, but average revenue generation has quintupled. As a result, there is still a shortfall of, on average, 35 percent of the total cost, but this was earlier almost 70 percent. This points to the increasing likelihood of being able to establish sustainable systems. Still, as mentioned above, the SFD decided years ago to shift to water harvesting schemes because of a concern about sustainability. This issue may require further study. 18 THE YEMEN SOCIAL FUND FOR DEVELOPMENT In addition, the study found that, for rural roads, 86 percent of the beneficiary communities had signed an agreement on road maintenance of completed projects. In most communities, maintenance follow-up committees had indeed been established, but in only one- fifth of the communities did the population contribute funding. Other communities have non-fund based arrangements for routine maintenance. Training by the SFD only took place for one project. The only somewhat weak link is the line agency responsibilities, in particular in health, where the ministry is to ensure sufficient staffing and recurrent budgets. Although each proposed project receives the go-ahead of the Ministry of Health, staffing and funding has been a problem, the reason why the SFD redirected its involvement in this sector. In micro-finance, the SFD is aiming at "best possible practices" in the sector, including careful monitoring of both institutional and financial sustainability. Project Cycle: The 2003 impact evaluation showed that the overall duration of the project production process was 576 days, including 221 days from community request to committee approval, 16 days from committee approval to contract with community, 102 days from contract with community to construction contract, and 237 days from construction contract to completion. There were some sector differences, with shorter periods for water supply and health sub-projects and longer periods for feeder roads. Even more interesting was that this time period has increased substantially over time, from 282 days in 1997 to more than 600 days in 2000 and 2002. Although this no doubt reflects the increasing number of fiduciary issues the SFD has to deal with, as well as the increasing focus on up-front community training, and still is not bad by international standards, it is important to note that SFD accepts new requests whether or not there is availability of funding at the time. Administrative Costs: Studies confirm that the SFD is a quite efficient delivery mechanism when compared with other agencies, with lower unit costs for each type of infrastructure. The data also show that a relatively small number of staff (128 by June 30, 2003) have been able to process a substantial program, leading to an administrative overhead of less than 10 percent, a percentage far below the overheads shown by most social funds. Success Factors Several factors have contributed to the success of the SFD operation to date: Leadership and Political Commitment: The SFD has benefited from great support from the President and senior leadership in Yemen, starting with Law Number 10 of 1997, establishing the organization. It is true that it took some government officials a while to accept the autonomy of the SFD and the new concept of community participation, and that this initially led to poor government counterpart funding, but the SFD's increasing effectiveness and efficiency rapidly overcame most bureaucratic concerns, and has helped provide for regular and increasing government budget support. A crucial part of the autonomy was the authority granted to the SFD to contract staff at rates sufficient to recruit highly qualified, experienced and motivated people. In addition, the SFD was allowed to establish a special account in a commercial bank, allowing it 19 CASE STUDIES IN SCALING UP POVERTY REDUCTION in turn to provide for fast payment of contractors and consultants, over time contributing much to lowering the cost of its projects. Not least important has been the strong daily leadership provided by the Managing Director, who has helped establish a productive and well-motivated organization. Learning and Experimentation: From the start, the SFD management has benefited from the experience of other social funds and developed its systems and procedures in house to be adapted to the local context. Throughout implementation, the SFD has been open to advice, and has aimed at "best practices". A variety of beneficiary assessments and evaluations have taken place, and has led the SFD management to adjust its organization and operational policies and practices in line with those studies' recommendations. For example, those evaluations led the SFD to introduce new items in its investment menu, including special programs for vulnerable groups, cultural heritage projects, and cluster-type programs. They also led the SFD to change its approach in health, steering away from new clinics, and instead focusing on health capacity building. The current Operational Manual reflects the latest in social fund thinking. Institutional Innovation: The SFD introduced a host of innovations, including the demand-driven approach, the contracting of well-paid staff on fixed appointments, the use of a variety of intermediaries, and the involvement of all stakeholders in its policy and operational decisions. In addition, it has been providing ample support to those stakeholders to build up their capacity, including community user committees, NGOs, contractors, and government agencies. More detailed information was discussed earlier. External Catalysts: Since the beginning of the SFD, donor support, both funding and technical assistance, has been of crucial importance. The Japanese and World Bank provided assistance with project preparation, and the Arab Fund for Economic and Social Development and World Bank have assisted implementation since 1997. The European Commission, the Dutch, the OPEC Fund for International Development, and the USA joined in 1998. The Islamic Development Bank joined more recently. Donor support is well-coordinated, and facilitated through conferences and study tours. Proper Preparation: SFD preparation activities were funded with the help of a Japanese Grant and an advance on the IDA Credit provided by the World Bank. This funding allowed for the contracting of substantial technical assistance to give the SFD a "flying start". Since then, the SFD has used various pilots to try out new approaches, to be "scaled up" if successful. Lessons Learned Design Poverty Targeting and Monitoring: At the time of the SFD start, the country's poverty data were far from reliable, and, although the SFD had an agreed system of regional targeting as of 1998 based on the number of population in each governorate and districts, its management preferred to use data on lack of services in each client community. However, this did not help establish a progressive allocation to the governorates and districts in accordance with their level of poverty. 20 THE YEMEN SOCIAL FUND FOR DEVELOPMENT This was corrected in 2000. The methodology was changed and distribution of resources became based on the number of households living in poverty in each area, thus channeling more resources to poorer areas. The SFD had generated baseline data on the situation in areas of its interventions in 1999 which helped conduct its impact evaluation in early 2003. Impact Evaluation Design: The "control group" problem: The SFD's Monitoring and Evaluation System seeks to determine the impact of its programs' interventions. This involves the specification and measurement of a "without project" scenario that in fact, does not exist. The specification of what would have happened without the intervention is known as the "counterfactual" statement. Ideally, an impact evaluation design would specify a fully-fledged "control group" for which both baseline and ex-post observations are taken in parallel with the observations on the "intervention group" (communities and households that have benefited from the project's intervention). However, in the case of the SFD, this was difficult to implement for two reasons. Firstly, there might be problems of "control group contamination" i.e. the possibility that other programs (such as the Public Works Program, Basic Education Expansion Program and others) might intervene in ways similar to the SFD's in communities that do not get SFD support. Secondly, there are also political and ethical difficulties with deciding to exclude some otherwise similar communities from SFD support, simply in order to provide a control group for impact evaluation purposes. This would directly run counter to two basic principles of the SFD's operations, namely: (a) that the SFD is a national program whose resources are distributed geographically according to number of households living poverty in each area, so that targeting is equitable (i.e. similar communities should get similar resource assignments); and (b) that communities should be able to demand the type of investment that is most important from their point of view. To provide meaningful points of comparison for evaluating the SFD's impact without attempting to define a fully-fledged control group, the Impact Monitoring and Evaluation System generates two sorts of observations: · Baseline data for a group of sub projects that have been programmed by the SFD, but not yet completed. · Ex-post data for a group of finished projects for which baseline data were collected previously; and recall data on ex-ante conditions for variables for which baseline data are not available. The exercise permits a systematic comparison of conditions relevant to the SFD projects before and after the intervention takes place, through two sets of comparisons: (a) the conditions in beneficiary communities before the SFD intervention, and approximately two years after the intervention; and (b) the conditions in SFD beneficiary communities that have already had the investment, compared with those in the program "pipeline" that are slated to benefit from similar investments. If it is found that the SFD's interventions improve conditions in the beneficiary communities on the before and after comparison; and that at any point in time the conditions in 21 CASE STUDIES IN SCALING UP POVERTY REDUCTION pipeline communities are inferior to those in recent beneficiary communities, it becomes reasonable to conclude that SFD is making a difference Credit Operations: The preparation and appraisal of the small and micro-enterprise activities did not foresee the difficulties in implementing the program including the underdeveloped banking services and high inflation (in 1998 when SFD started operations). In addition, the SFD encountered difficulties in convincing commercial banks to finance small enterprises. Later a baseline survey of small and micro enterprises sponsored by the SFD indicated that finance was not on top of the list of problems of small enterprise owners. So the decision to postpone the implementation of this component was well justified. The SFD continues to give priority to providing women with financial services and was able to learn from experience that savings and group methodologies are more successful in increasing the participation of women. Operation: Community Participation and Contracting: The SFD has gradually built its approach towards working with community. This was achieved through: provision of training to SFD staff and consultants in participatory development and methods to identify priorities in collaboration with communities; establishment of local committees representing the community in the implementation; and operation of the projects and provision of training to local committees in basic financial and management skills and in operation and maintenance. In each rural community development projects beneficiaries' committees were established. These committees received on-the-job training during the implementation of the projects, in the formulation of project requests and management of projects. In-depth capacity building was given to the integrated interventions, the school community contracting interventions and to direct implementations in road and water. Several lessons are learned such as: (a) a long-term commitment is needed to achieve a lasting impact on community development; (b) community contracting is more time consuming and close supervision is needed; and (c) it leads to better quality of works and lower unit costs and it is expected that the maintenance of these projects will be guaranteed much easier. In a conservative society such as Yemen, time, advocacy efforts and role model are still needed to achieve women genuine involvement. Operation and Maintenance: One has to differentiate between projects operated by the communities such as the water harvesting, and those operated by the line ministries such as health facilities and schools. For the first type of projects, the SFD focuses on providing training to the water and road users committees in operation and maintenance. For the second type of projects, the SFD signs agreements with ministries of education and health in order to provide the staffing and the needed inputs. Furthermore, completed subprojects are revisited after handing-over by SFD staff and consultants to check their level of operation and take corrective action to overcome organizational or technical problems. Government Agency Cooperation: The success of support provided by the SFD to central government agencies and officials depends primarily on their interest and cooperation. It worked 22 THE YEMEN SOCIAL FUND FOR DEVELOPMENT reasonably well with the Ministry of Education and Social Affairs but less so in the case of the Ministry of Health. Unit Costs: The introduction of proper technical design and procurement requirements by the SFD led to increasing competition among contractors and lower unit costs for basic infrastructure, such as school construction. Organization and Management: Gender Sensitization: The initial consideration of a separate gender office in SFD, to be copied from the Egypt Social Fund, was rejected in favor of "mainstreaming" the gender issue. A gender sensitization workshop for the SFD was held by an international consultant, a substantial number of female managers was appointed, and a specialist was assigned in the SFD to screen all sub- project proposals on this issue. In addition, each regional office had to recruit female staff to ensure that this aspect was taken care of. Still, at the regional offices and at the community level, this subject could have benefited from some more systematic and intensive attention. Autonomy: Because of its ability to function with considerable autonomy (Law Number 10 of 1997), the SFD was able to quickly establish its credibility as a neutral and professional outfit. This was possibly the most important point in ensuring the project's successful outcome. Management and Staffing: The Government's willingness to exempt the SFD from standard government requirements also allowed for its recruitment of highly qualified and well- paid managers and staff, a critically important condition for any well-functioning organization. Progression and Flexibility: The SFD management's willingness to pursue the latest international knowledge on social fund policies and practices helped establish a progressive and flexible mechanism. Bank Staff Resources: The Bank assigned staff to its field office; which provided the SFD with quick and ongoing access to Bank supervision and technical assistance. Technical Assistance: The availability of Bank-managed technical assistance, funded by Trust Funds from the Dutch and Japanese, allowed for easy and fast access when and where needed. Conclusion The SFD has played, and is playing, a unique and critically important role in helping alleviate Yemen's major social problems. Through its demand-driven approach, it has introduced new concepts concerning community participation, including ensuring the involvement of women. It has contributed much to improving access to priority social and economic infrastructure and micro-credit, and used NGOs and other intermediaries to the maximum extent. In the process, it has helped build up the implementation capacity of the poor beneficiary communities, NGOs, contractors, and government agencies. With the latest developments on decentralization in the country, it will now have to also ensure the participation of local governments and councils in the 23 CASE STUDIES IN SCALING UP POVERTY REDUCTION process of local development. The SFD is also helping in building the capacities of these councils by supporting a number of capacity building activities. The SFD has rapidly adjusted to new needs in the past, and its highly experienced and professional management and staff can be expected to continue to do that. As long as the responsible line agencies are not able to meet their sector obligations and the SFD can do this efficiently and effectively, there is ample justification for maintaining this crucial organization. The SFD will continue as a house of innovation and experimentation that transfers the models to other government agencies. 24