Document of The World Bank Report No: ICR00001760 IMPLEMENTATION COMPLETION AND RESULTS REPORT (IDA-H0780) ON A GRANT IN THE AMOUNT OF SDR 34.2 MILLION (US$ 51 MILLION EQUIVALENT) TO THE DEMOCRATIC SOCIALIST REPUBLIC OF SRI LANKA FOR A COMMUNITY DEVELOPMENT AND LIVELIHOOD IMPROVEMENT “GEMI DIRIYA� PROJECT IN SUPPORT OF THE FIRST PHASE OF THE COMMUNITY DEVELOPMENT AND LIVELIHOOD IMPROVEMENT “GEMI DIRIYA� PROGRAM December 31, 2010 Sustainable Development Department Agriculture and Rural Development Unit South Asia Region CURRENCY EQUIVALENTS (Exchange Rate Effective August 23, 2010) Currency Unit = Sri Lankan Rupees (Rs) Rs. 1.00 = US$ 0.01 US$ 1.00 = Rs. 112.58 FISCAL YEAR [January 1 – December 31] ABBREVIATIONS AND ACRONYMS ADB Asian Development Bank APL Adaptable Program Loan CAS Country Assistance Strategy CBO Community Based Organization CDD Community-Driven Development CDLIP Community Development and Livelihood Improvement Project CLPTC Community Learning and Professional Training Center COM Community Operations Manual CP Community Professional EA Environmental Assessment EMF Environmental Management Framework FAO Food and Agriculture Organization FMR Financial Management Report FMS Financial Management Statement GAAP Governance Accountability and Action Plan GDF Gemi Diriya Foundation GEMI Strength of A Village in local language DIRIYA GND Gram Niladhari Division GOSL Government of Sri Lanka GTZ Deutsche Gesellschaft fur Technische Zusammenarbeit HRM Human Resources Management IAS Impact Assessment Study of 2010 ICB International Competitive Bidding ICRR Implementation Completion Results Report IDA International Development Association IFAD International Fund for Agricultural Development JBIC Japan Bank for International Cooperation JSDF Japan Social Development Fund KfW Kreditanstalt fur Wiederafbau KPI Key Performance Indicator MIS Management Information System ML&E Monitoring, Learning and Evaluation MRRP Mahaweli Restructuring and Rehabilitation Project MS Maha Sabha MTR Mid-term Review NCB National Competitive Bidding NGO Non-governmental Organization O&M Operations and Maintenance PAD Project Appraisal Document PCN Project Concept Note PDO Project Development Objective PRA Participatory Rural Appraisal PRSP Poverty Reduction Support Paper PS Pradeshiya Sabha SO Service Organization SPO Senior Program Officer UBNI Unmet Basic Needs Index UNDP United Nations Development Program VO Village Organization VDP Village Development Plan VSCO Village Savings and Credit Organization VSHLI Village Self-Help Learning Initiative Vice President: Isabel M. Guerrero Acting Country Director: Amit Dar Sector Manager: Simeon Kacou Ehui Project Team Leader: Meena M. Munshi ICR Team Leader: Jim Hancock, FAO SRI LANKA COMMUNITY DEVELOPMENT AND LIVELIHOOD IMPROVEMENT “GEMI DIRIYA� PROJECT Table of Contents 1. Project Context, Development Objectives and Design ................................................... 1  2. Key Factors Affecting Implementation and Outcomes ................................................. 6  3. Assessment of Outcomes .............................................................................................. 12  4. Assessment of Risk to Development Outcome ............................................................. 20  5. Assessment of Bank and Borrower Performance ......................................................... 21  6. Lessons Learned............................................................................................................ 23  7. Comments on Issues Raised by Borrower/Implementing Agencies/Partners............... 25  Annex 1. Project Costs and Financing .............................................................................. 26  Annex 2. Outputs by Component...................................................................................... 27  Annex 3. Economic and Financial Analysis ..................................................................... 42  Annex 4. Bank Lending and Implementation Support/Supervision Processes................ 54  Annex 5. Beneficiary Survey Results ............................................................................... 56  Annex 6. Stakeholder Workshop Report and Results ....................................................... 57  Annex 7. Summary of Borrower's Completion Report..................................................... 64  Annex 8. Comments of Cofinanciers and Other Partners/Stakeholders ........................... 80  Annex 9. List of Supporting Documents .......................................................................... 81  MAP (IBRD 32943) .......................................................................................................... 82  A. Basic Information Community Development and Country: Sri Lanka Project Name: Livelihood Improvement "Gemi Diriya" Project Project ID: P074872 L/C/TF Number(s): IDA-H0780 ICR Date: 01/04/2011 ICR Type: Core ICR GOVERNMENT OF Lending Instrument: APL Borrower: SRI LANKA (GOSL) Original Total XDR 34.20M Disbursed Amount: XDR 34.20M Commitment: Revised Amount: XDR 34.20M Environmental Category: B Implementing Agencies: Gemi Diriya Foundation Cofinanciers and Other External Partners: B. Key Dates Revised / Actual Process Date Process Original Date Date(s) Concept Review: 02/25/2003 Effectiveness: 10/28/2004 10/28/2004 Appraisal: 12/04/2003 Restructuring(s): Approval: 03/30/2004 Mid-term Review: 05/14/2007 06/15/2007 Closing: 03/31/2009 03/31/2010 C. Ratings Summary C.1 Performance Rating by ICR Outcomes: Satisfactory Risk to Development Outcome: High Bank Performance: Satisfactory Borrower Performance: Satisfactory C.2 Detailed Ratings of Bank and Borrower Performance (by ICR) Bank Ratings Borrower Ratings Quality at Entry: Satisfactory Government: Satisfactory Implementing Quality of Supervision: Satisfactory Satisfactory Agency/Agencies: Overall Bank Overall Borrower Satisfactory Satisfactory Performance: Performance: i C.3 Quality at Entry and Implementation Performance Indicators Implementation QAG Assessments Indicators Rating Performance (if any) Potential Problem Project Quality at Entry No None at any time (Yes/No): (QEA): Problem Project at any Quality of No None time (Yes/No): Supervision (QSA): DO rating before Satisfactory Closing/Inactive status: D. Sector and Theme Codes Original Actual Sector Code (as % of total Bank financing) General agriculture, fishing and forestry sector 25 25 General industry and trade sector 15 15 Other social services 30 30 Sub-national government administration 10 10 Water supply 20 20 Theme Code (as % of total Bank financing) Gender 14 14 Participation and civic engagement 29 29 Rural non-farm income generation 28 28 Rural services and infrastructure 29 29 E. Bank Staff Positions At ICR At Approval Vice President: Isabel M. Guerrero Praful C. Patel Country Director: Amit Dar Peter C. Harrold Sector Manager: Simeon Kacou Ehui Constance A. Bernard Project Team Leader: Meena M. Munshi Meena M. Munshi ICR Team Leader: Meena M. Munshi ICR Primary Author: Jim Hancock ii F. Results Framework Analysis Project Development Objectives (from Project Appraisal Document) The development objective of the 12-year Gemi Diriya APL was to enable the rural poor to improve their livelihood and quality of life. The objective of the Project, which supported the first of the three phases of the APL, was to enable the communities of Uva and Southern Provinces to build accountable and self-governing institutions and to manage sustainable investments by: (i) devolving decision-making power and resources to community organizations; (ii) strengthening selected local governments which demonstrate responsiveness and accountability to rural communities; and (iii) working with federations of village organizations, the private sector and Non-Governmental Organizations (NGOs) on economic empowerment to increase the size and diversity of livelihood options. Revised Project Development Objectives (as approved by original approving authority) (a) PDO Indicator(s) Original Target Formally Actual Value Values (from Revised Achieved at Indicator Baseline Value approval Target Completion or documents) Values Target Years Indicator 1 : % of population earning less than $1 per day 40% (implies 10% 27% reduction from reduction in the 71% to 44% for Value share of the total project areas, quantitative or 50% N/A population living compared to 14% Qualitative) on less than $1 per reduction in control day) area (65% to 51%) Date achieved 12/05/2005 03/31/2009 03/31/2010 03/31/2010 Source: Impact Assessment Study, 2010. Comments The assumed poverty level in the PAD was 50%, but the actual baseline survey (incl. % which became available only at MTR showed the poverty starting point to have achievement) been much higher (71% in Gemi Diriya areas and 65% in contro % incremental income increase (including employment) generated from sub- Indicator 2 : project investments. 76% increase for households receiving project Value VSCO loans quantitative or No target set N/A compared to Qualitative) average 35% increase for control area poor households. Date achieved 03/31/2009 03/31/2010 03/31/2010 iii Source: Impact Assessment Study, 2010. Comments For VSCO loan households in project areas, and including only poor households (incl. % for both project and control sample. Batches 0 and 1 are same households in achievement) MTR and IAS, thus directly comparable, as were the controls Indicator 3 : % of households benefitting from increased income (in project area). Value quantitative or 0 30% N/A 52% Qualitative) Date achieved 01/05/2005 03/31/2009 03/31/2010 03/31/2010 Source: Impact Assessment Study, 2010. Comments Target assumed to be 30% of 150,000 beneficiary households (45,000). (incl. % According to IAS 210 Section 5.3, 43% of households surveyed (i.e. 79,000 out achievement) of all households surveyed) stated their incomes had increased due Indicator 4 : % of households benefitting from incremental infrastructure. Value quantitative or 0 70% N/A 81% Qualitative) Date achieved 01/05/2005 03/31/2009 03/31/2010 03/31/2010 Source: Impact Assessment Study, 2010. Comments This is based on number of beneficiary households to total number of households (incl. % in VO. The achievement) target is assumed to be 105,000 households (70% of 150,000 beneficiaries). As of IAS, 121,000 households (115%) have ben (b) Intermediate Outcome Indicator(s) Original Target Actual Value Formally Values (from Achieved at Indicator Baseline Value Revised approval Completion or Target Values documents) Target Years % of VOs formally registered, receiving funds and effectively implementing Indicator 1 : village priorities. 100% of 1,034 VOs Value 65% (650 of 1,000 actually formed, or (quantitative 0 VOs expected to N/A 103% of original or Qualitative) be formed) VO target of 1,000 Date achieved 01/05/2005 03/31/2009 03/31/2010 03/31/2010 Source: Gemi Diriya Foundation MIS Comments PAD text described this indicator as the number of village organizations (VOs) (incl. % formed and percentage of these functioning effectively in priority setting and achievement) decision-making on resource allocation. Results were 159% c % of completed sub-projects (and ratified by the Maha Sabha) operated and Indicator 2 : maintained by VOs. Value (quantitative 0 60% N/A 141% or Qualitative) Date achieved 01/05/2005 03/31/2009 03/31/2010 03/31/2010 Comments Source: Impact Assessment Study, 2010. iv (incl. % The Maha Sabha is the village assembly that makes decisions concerning village achievement) development and other issues. The original target was based on an assumption of one sub-project per village (600/1,000). The actual % of VOs with viable savings and credit organizations that are covering Indicator 3 : operating costs 100% of 1,034 VOs 40% (400/1,000 of actually formed (or Value VOs expected to 103% of original (quantitative 0 N/A be formed have VO target of 1,000) or Qualitative) viable VSCOs) have viable VSCOs. Date achieved 01/05/2005 03/31/2009 03/31/2010 03/31/2010 Source: Impact Assessment Study, 2010. Comments All 1,034 VSCOs are covering operating costs, i.e., 256% achieved compared to (incl. % target. achievement) Indicator achieved. Indicator 4 : % of VOs working in partnership with private sector organizations. 43% of 1,034 VOs actually formed (or Value 48% of 1,000 VOs (quantitative 0 30% N/A originally expected or Qualitative) to be formed) have increased private sector investment. Date achieved 01/05/2005 03/31/2009 03/31/2010 03/31/2010 Source: Impact Assessment Study, 2010. Comments PAD text described this indicator as an increase in private sector investment in (incl. % the communities. Results were 148% compared to the original target. achievement) Indicator achieved. G. Ratings of Project Performance in ISRs Actual Date ISR No. DO IP Disbursements Archived (USD millions) 1 06/22/2004 Satisfactory Satisfactory 0.00 2 12/27/2004 Satisfactory Satisfactory 0.00 3 05/24/2005 Satisfactory Satisfactory 1.37 4 12/13/2005 Satisfactory Satisfactory 2.70 5 06/23/2006 Satisfactory Satisfactory 6.61 6 12/19/2006 Satisfactory Satisfactory 10.76 7 06/13/2007 Satisfactory Satisfactory 15.32 8 12/19/2007 Satisfactory Satisfactory 27.50 9 06/28/2008 Satisfactory Satisfactory 33.34 10 12/28/2008 Satisfactory Satisfactory 42.97 11 05/26/2009 Satisfactory Satisfactory 48.42 v 12 11/28/2009 Satisfactory Satisfactory 51.04 13 05/23/2010 Satisfactory Satisfactory 51.04 H. Restructuring (if any) Not Applicable I. Disbursement Profile vi 1. Project Context, Development Objectives and Design 1.1 Context at Appraisal The PAD recognized regional disparities in poverty in Sri Lanka: in 2000, 90% of the country’s poor were living in rural areas; poverty was deeper in the south and in upland and drier areas; and the incidence of poverty was greater in more remote rural areas. Uva Province had the highest incidence of poverty (37%) compared to Western Province (12%). Most of the country’s poor were employed in agriculture. In remote rural areas there were long-term issues of limited access to critical infrastructure, and poor flow of information on markets and other livelihoods opportunities. There was also a generalized lack of voice by the rural poor in decision-making on policies, programs and resource allocation relevant to their well-being. At the time of project identification and preparation, Sri Lanka’s Poverty Reduction Strategy Paper (PRSP) was largely growth oriented, focusing on building opportunities for employment and income generation. Concurrently, a number of reviews pointed to weak targeting and inefficiencies in most ongoing and previous national poverty alleviation programs. IDA’s 2003 Country Assistance Strategy (CAS) for Sri Lanka had the goals of growth and equity, which were to be promoted through well- targeted anti-poverty interventions, especially in poorer regions. The Project was designed to be one such intervention, building on a successful experience piloted under an earlier IDA operation. Specifically, the Village Self-Help Learning Initiative (VSHLI), conceived in 1998 and piloted in Polonnaruwa district under the Mahaweli Restructuring and Rehabilitation Project of 2001, was yielding promising early results. IDA and the Government (GOSL) therefore decided to strengthen and scale up this experience into a much larger and longer-term national rural community-driven development (CDD) program, to be known as ‘Gemi Diriya’ (strength of a village). Gemi Diriya aimed to help GOSL improve livelihoods and directly address income generation opportunities in poorer areas, leverage additional funds for rural communities, and provide a demonstration effect for other national programs. The program would take place in three phases over 12 years; each phase would demonstrate, refine and expand geographically a highly participatory, beneficiary-led model of local institution building, support for income generation, and decentralized budgeting and transfers. The phases would each last about four/five years, with the first focusing on community mobilization, building and strengthening community level organization, empowerment, and developing institutional mechanisms for direct funding of sub-project investments at the community level. The second phase would expand horizontally to two new provinces and deepen the achievements of Phase 1 by focusing on building second generation community institutions by networking them into federations, and on linkages and partnerships with local governments, the private sector and across federations, as well as starting the integration process with other government programs. The third phase would focus on further scaling up and policy impact at the national level. There would be increasing emphasis on institutionalization and building capacity of local government to take over program functions. Due to conflict conditions, it was not possible for Gemi Diriya to work in some parts of Sri Lanka at the time the program was designed, but it was understood that if peace could be achieved during the 12 years, the program might eventually extend into the affected areas. 1 The Project, which is the subject of this Implementation Completion Results Report (ICRR), supported the first of the three phases of Gemi Diriya. It was financed by an IDA post conflict grant, using the adaptable program lending instrument (APL). There were performance triggers for moving from each phase to the next (see Section 1.2 below). This was achieved and therefore Phase 2 was approved on September 9, 2009 (IDA Credit: 4613-LK), and is presently under implementation. Phase 2 implementation started well soon after its effectiveness. However, following the national elections of May 2010, the Gemi Diriya Foundation (GDF) has been transferred to the Ministry of Economic Development. While the new Ministry basically replaced the previous Ministry of Nation Building and while there were assurances provided to the Bank that there will be no change in the project scope, components and modality of implementation of Phase 2 as well as continuity of implementation arrangements, the expressed needs of accelerating the integration of GDF into the regular government apparatus have created uncertainties in terms of the timing and process how this will be achieved and affected institutional effectiveness of GDF. Therefore, there are serious risks of delay in achieving targets and losing momentum and quality of the program with potential impacts affecting outcomes of Phase 2of the APL. 1.2 Original Program Development Objectives (PDO) and Key Indicators The PDO was to enable poor communities to build accountable and self-governing local institutions and to manage sustainable investments by: (i) devolving decision-making power and resources to community organizations; (ii) strengthening selected local governments that demonstrate responsiveness and accountability to rural communities; and (iii) working with federations of VOs, private sector and Non-Governmental Organizations (NGOs) on economic empowerment to increase size and diversity of livelihood. PDO results were to be measured by the following indicators (see PAD B.3.2 and Annex 3, and ICRR Section F (a)):  % incremental income increase (including employment) generated from sub-project investments  % of households benefiting from increased incomes  % households benefiting from community infrastructure  Number of VOs formed and percentage of these functioning effectively in priority setting and decision making on resource allocation  % of sub-projects that are implemented and maintained by the communities  % increase in household incomes of the poor  % of VOs with sustainable savings and credit system at the end of the project  % of VOs experiencing an increase in private sector investment in the communities. 2 In addition, as Phase 1 of an APL, the project had five performance triggers that needed to be met in order to move on to Phase 2 (Section 3.3 below)  At least 60% of VOs in the first two years of Phase 1 accessed Funds and completed at least one sub- project  At least 30% of women participated in decision making by holding management positions in the first two years of Phase 1  At least 50% of community members contributed 20% or more toward the capital cost of community infrastructure activities  At least 50% of members of project villages covered during the first two years of Phase 1 benefitted from project interventions; and at least 60% of these belonged to the poorest households  At least 25% of the Divisions covered in the first two years established participatory sub-committees at the local government level with 30% elected representatives and 70% representatives of community organizations. 1.3 Revised PDO (as approved by original approving authority) and Key Indicators, and reasons/justification. N/A 1.4 Main Beneficiaries The Project was to be implemented in about 1,000 village communities in 510 Gram Niladhari Divisions (GNDs) in five districts (Galle, Matara, Hambantota, Badulla, and Moneragala) of Uva and Southern provinces. In practice this number also covered additional areas under Component 5 on Piloting of the VSHLI model, and the District of Rathnapura in Sabaragamuwa Province, which was added in late 2004 (see Section 1.7 below). The project was to cover approximately 150,000 households and approximately 700,000 people, equivalent to about 20% of the population of the two primary provinces (PAD Annex 4). Targeting to reach poor and disadvantaged groups would take place at three levels. First, poverty and other criteria would be used to prioritize Divisions and GNDs (this targeting is further discussed in section 3.2); second, prioritized village communities would then have the option of joining the project through a self-selection process; and third, criteria would be applied for selection of households within the communities that had joined Gemi Diriya. Some 60% of participating households were to be from poor categories; at least 30% of leader positions were to be held by women; and there were livelihoods sub-components specifically directed towards youth (skills development and employment linkage), as well as provision for one time grants to the most vulnerable in the community, e.g. the disabled, widows without any income, single women with small children and without income, other destitute, etc. The Project would also have an indirect impact on other stakeholders – such as communities in the neighboring villages, and developmental agencies and government departments benefitting from Project investments although they were not direct participants; and individuals and organizations benefitting from Project institutional development activities at the local level. While these are noted, this ICRR does not count such indirect beneficiaries in assessing achievement of PDO targets. 1.5 Original Components (as approved) The original Project design had the following five components: 3 Component A: Village Development (US56.6 million). This included two sub-components: (1) Development and Strengthening of Village Organizations; and (2) Funding of Community Sub- projects. The first supported the development of self-reliant and self-managed people’s organizations for community development and livelihood improvement in rural areas. The second supported capacity building, social and community infrastructure, and livelihood activities (one-time grants to the poorest, loans for youth skills development, and to village savings and loan funds), and an incentive fund for well performing villages. Component B: Institutional Strengthening (USD6.0 million). This component supported capacity-building for local and national agencies and support organizations (SOs) to facilitate program implementation; broaden their skills, especially in participatory development; create a monitoring and learning system to capture results on the ground; and respond to project-related needs and demands of the community. Inter alia, the component would help to introduce social and public accountability mechanisms, like participatory budgeting, community report cards, social audit and expenditure tracking. It also included a Policy Dialogue activity to help policy makers develop policies, rules, systems and institutional arrangements that would allow the government to transfer funds and technical support to communities on a demand-driven basis. Component C: Innovation Seed Fund (USD1.5 million). This component supported the piloting of innovative ideas with potential to contribute to the overall PDO, but which needed experimentation, learning and incubation before possible scaling up and replication. Some of the pilots were to include technology and product development related to agricultural processing and value addition and information technology. Component D: Project Management (USD4.7 million). The component supported overall coordination and management of the Project at national, provincial and divisional levels. The component would finance goods, works, services and recurrent costs. Component E: Village Self-Help Learning Initiative (VSHLI) Pilot (USD1.0 million). The component would complete implementation of the then ongoing VSHLI pilot in Polonnaruwa district, deepening the focus on livelihoods development activities and testing a second stage of institutional development involving federation of village organizations (VOs) and partnerships with the private sector. 1.6 Revised Components There were no major modifications to the original five Project components, but at the MTR it was decided that since all of the main activities under Component E (VSHLI pilot) had been completed, as of September 2007 the pilot would be mainstreamed into the Village Development Component 1 (MTR Aide Memoire, June 2007). Thereafter, further support to the VSHLI villages would be channeled through the GDF structure and funding system, instead of the Mahaweli Authority. Also during the Project, there were two additions to the Project Description in Schedule 3 of the Development Grant Agreement (DGA): Part Z: Tsunami Response. In March 2005, SDR 5.5 million (USD 8.29 million equivalent at the time) of the IDA grant for the Project was reallocated to a new Part Z to help finance IDA’s response to the 4 impact of the December 2004 Tsunami on Sri Lanka. This ICRR does not assess the results of Part Z, which is taken up in the completion review for the overall Tsunami response.1 Component F: Gama Neguma Inter-connectivity Pilot. In July 2008, the DGA was amended to include the piloting of closer cooperation and possible integration of Gemi Diriya with the Government’s Gama Neguma Program. Specifically, the new component would: (a) test an institutional and financial model at the PSr level for implementing zonal/regional inter-village connectivity needs; and (b) provide technical assistance to enable the Ministry to scale up Gama Neguma. In effect, this brought forward in the timetable of the APL the mainstreaming of lessons from Gemi Diriya into other public programs and linking of communities to the local governments. 1.7 Other significant changes Expansion of geographic coverage. In late 2004, project activities also started in Sabaragumawa Province, which is one of the poorest districts and satisfied the eligibility criteria.2 Provision of support to complete some Phase 1 activity under Phase 2. At the time of the Tsunami amendment (see Section 1.6 above), it was agreed that some activities under Phase 1 would need to be covered by Phase 2 funds. Therefore, the Phase 2 project includes USD 15 million, mainly to compensate for the Tsunami reallocation, but also to offset shortfalls due to the additional Gama Neguma inter-connectivity pilot and the larger number of Phase I households than originally planned. For the purposes of this ICRR, only activities financed and fully completed under Phase 1 are included in the assessment of project results, but it is important to note that this is a conservative assessment because additional work started under Phase 1 and will soon be completed under Phase 2. Shift of GOSL ministerial responsibility for Gemi Diriya. Following the 2007 national elections, Project responsibility was transferred from the Ministry of Samurdhi and Poverty Alleviation, to the Ministry of Nation Building and Estate Infrastructure Development. This did not imply any change in the role of GDF, which continued to operate as an autonomous institution. However, it brought the Gemi Diriya program within the ambit of a more powerful Ministry (the President of Sri Lanka was also the Minister in charge , making it possible to strengthen the mainstreaming/integration of some of Gemi Diriya’s principles and service delivery strategy into other national poverty programs, such as Gama Neguma and Samurdhi. Reallocation of funds. After the MTR the DCA was amended in 2008 with an increased allocation of US$4.5m to Project management while reducing Village Development from US$40.5m to US$34.3m. The increase in management cost was due to the government salary structure in Sri Lanka having two 1 See World Bank Response to the Tsunami Disaster,� SecM2005-0035. This ICRR does not assess the results of Part Z, which is taken up in the completion review for the overall Tsunami response. See ICR 0001105 for a Tsunami Emergency Recovery Program for the Democratic Republic of Sri Lanka, dated June 30, 2009. 2 While the PAD referred to Uva and Southern Provinces, the Development Grant Agreement also referred to Sabaragamuwa Province and the VSHLI pilot was in the North Central Province. The Project Agreement generally provided for flexibility by permitting changes to the Project Implementation Plan and other basic project documents that, in IDA’s opinion, would not adversely affect implementation. 5 major revisions during 2005 to 2008, which were not anticipated during preparation and thus needed to be revised accordingly. Village Development Funds were not reduced per village, but some villages starting activities in 2007 had sub-projects costs covered under Phase II (as their fund allocations had been previously transferred to the Tsunami emergency – see above). Extension of IDA grant closing date. In 2008, the grant closing date was extended by one year to March 31, 2010 to ensure continuity on the ground between Phases 1 and 2. 2. Key Factors Affecting Implementation and Outcomes 2.1 Project Preparation, Design and Quality at Entry The design of Gemi Diriya drew heavily on the Village Self Help Learning Initiative (VSHLI) pilot in Polonnaruwa District in the North Central Province, under the Mahaweli Rehabilitation and Restructuring Project (MRRP, Cr. 3058-CE)). The idea of VSHLI had emerged during a Village Immersion Program for World Bank Staff in Sri Lanka villages in 1997-98. The pilot under MRRP aimed to test a community-driven development (CDD) institutional model that empowers village communities and builds partnership at the local level. This was in response to the perceived weaknesses of earlier development programs in Sri Lanka, including lack of ownership at the local level, exclusion of the poorest, insufficient sustainability of investments, and general absence of a strategy for economic improvement of the poor. A Japan Social Development Fund (JSDF) Grant linked to MRRP supported the pilot in 3 villages, and then expanded to 12 villages. Although MRRP closed in December 2003, the JSDF grant was extended until 2005. A seminar was organized in June 2002 to disseminate key lessons from the pilot to national level policy makers, and a series of interactive workshops were organized to design the Gemi Diriya program. VHSLI was incorporated as a separate component of the first phase of Gemi Diriya. The pilot demonstrated the feasibility of a direct funding model leading to village-wide planning, community contribution (cash and labor) for infrastructure and their involvement in implementation, operation and maintenance. It also successfully demonstrated that, within a participatory framework, communities could articulate their needs and priorities and take responsibility for implementing and monitoring village development. Finally, the model tested and demonstrated the village company model as being the most appropriate institutional arrangements, due to its flexibility with regard to fostering a range of business options, and its independence from political influence and bureaucracy. The design of both VSHLI and Gemi Diriya itself also drew on a range of World Bank and other agencies’ CDD experiences and good practices, with regards to: community ownership and contributions, direct financing to communities, accountability and transparency mechanisms, and intensive supervision. Under VSHLI, in 2001, local level and Ministry officials went on an exposure visits to Indonesia and Andhra Pradesh in India to learn from best practice CDD experiences. During Project preparation, 30 pilot villages were selected in considerably different geographical settings compared to the irrigated areas of Polonnaruwa where VSHLI was being implemented. These villages in the uplands of Badulla and drier areas such as Moneragala engaged in intense monitoring and learning activities to fine tune the Gemi Diriya model, and they sustained their piloting role throughout the life of the Project. This helped to reduce risks identified in the PAD, such as potential weakness in 6 VOs, low women’s participation and low community willingness to contribute to costs, and also build capacity of the staff and communities to scale up under the Project. The objectives and component structure of the Project were straightforward and the organizational structure simple, with the Project implemented wholly through one independent Gemi Diriya Foundation (GDF) set up specifically to implement it. It was considered critical to maintain political and operational independence, flexibility in responding to community needs and demands, and continuity in the piloting and scaling-up of the Project. Government commitment was high at design stage, with adequate resources allocated to the Project and support to the set up of the independent implementing institution. The Project was consistent with GOSL’s strategy of pro-poor growth and of improving access by the poor to basic economic and social infrastructure and productive opportunities. The IDA CAS for Sri Lanka, which focused on peace, growth and equity, recommended targeted anti-poverty programs to enhance the human and physical capital of the poor to improve social conditions and to allow communities to make better use of economic opportunities. During preparation, IDA had a constant dialogue with other donors including ADB, JBIC, UNDP, GTZ, KfW and IFAD. All of these elements contributed to a sound project design based on adequate technical, economic and financial preparation. 2.2 Implementation On the whole, there was good momentum throughout the Project period. With a one-year extension of the closing date, grant funds were fully disbursed and most major outcome and output targets achieved (see Annex 2). The number of households covered was almost a third more than expected at appraisal (see Section 3.2). The Project has supported the establishment of Village Organizations (VOs) and development and implementation of village development plans (VDPs) in every Project village, through a highly participatory, inclusive and community-owned process. Communities in 1,034 villages contributed labor and cash to implement over 3,500 small rural infrastructure sub-projects, of which 1,415 have completed construction and in operation. Livelihoods Support Funds managed by communities have been released as planned for one-time grants to the poorest, as soft loans for youth employment and skills development, and as revolving funds to Village Savings and Credit Organizations (VSCOs) operating in every village, most of which are now covering their costs, and re-lending their funds to a wide range of livelihoods activities. VSCOs are a banking arm of the Village Organizations that are legal entities registered under the Company’s Act. The village organizations (VSCOs and VOs) have been aggregated at the district level; and five District VO/VSCO Federations and 5 District Community Professional Training Centers were established and are providing services to member VOs. Phase II consolidates gains of Phase I through supporting community institutions beyond village level and aggregating VOs and VSCOs into federations and linkages with markets, and financial sector linkages to ensure that the community level investments and institutions are sustainable in the longer run and promote sustainable livelihoods. A range of innovations have been adopted, for example on the use information communication technology (the Community Professional Learning and Training Centers and the Community Radio pilot.) As a result of the additional Gama Neguma Interconnectivity component, more complex sub-projects are under construction with direct community supervision and labor inputs. 7 A very large number of community infrastructure sub-projects were identified under 1,034 VDPs, have been approved in Maha Sabha (4,387, of which 3,528 were approved in sessions with over 80% of households in the villages attending) and initiated, although in some cases some installments will be covered under Phase 2. This is because some batches of villages started only in 2007, and, for the reasons explained above (allocation to Tsunami, etc.), the resources available could not cover all activities included in the VDPs. Foremost among the factors contributing to this strong performance was the soundness of the Project village development model with its strong buy-in from the community. This was confirmed consistently by feedback workshops and discussions with communities, interviews with local officials, local and national politicians, and the Impact Assessment Study of 2010. Core features of the model were:  Inclusive involvement and ownership of the whole village in decision-making, especially the poor and other targeted disadvantaged groups.  Building of practical capacity “community to community� to manage the process in the village, among small groups, youth, Community Professionals (CPs), and up to the VOs and federations.  Transfer of resources directly to the community, with community setting up funds and building up savings, for commitment and ownership. The community and its committees also control such key processes as procurement. This financial oversight by the community considerably contributed to cost savings (see Section 3.3).  A village company model that ensures benefits go to all members and the organization can focus on economic activities.  A Village Savings and Credit Organization (VSCO) microfinance model that is owned by the community, has very large coverage in the village, and is run almost completely by women, with arrangements that include several controls and safeguards against risks.  Clear consistent rules and principles for accountability and transparency that are integrated into Project-wide procedures and refined for each major activity in the community operations manuals (COMs). On the management side, the relatively simple GDF structure (with National, District, Hub, and later also regional offices) allowed for relatively coordinated and consistent implementation across project areas. A very hands-on top management approach and high staff commitment resulted in close monitoring, troubleshooting and continuous refinement of activities. After some early recruiting delays, a very thorough human resources management (HRM) system was developed, for which GDF received a national Human Resources award for results-focused and performance-based management in 2010. The Project trained a large number of staff from line agencies, PSs, NGO Service organizations (SOs) and the private sector, as well as Community Professionals, to support community mobilization, planning and technical needs, where these capacities were not previously available. This contributed both to the successful implementation of Phase 1 of Gemi Diriya, as well as to the internalization of its principles by some local governments. 8 IDA supervision missions and the MTR recognized adjustments that needed to take place, such as accelerating the mainstreaming of the project through the additional Gama Neguma interconnectivity component introduced in 2008. This particular change considerably increased engagement by local officials and politicians especially at the Pradeshiya sabha (PS) level, as discussed below. It is also worth noting that that the Gemi Diriya model employed an innovative youth strategy that tapped the energy and ideas of young people. ICT as it evolved from a small pilot in 2004 to a much scaled up version by 2009 helped tremendously in disseminating messages and linking with the private sector and markets. In recognition, in 2008 Gemi Diriya received a South Asia Manthon award for best e-enterprise and livelihood development. Finally, Gemi Diriya also developed capacity of about 20,000 women leaders who are now helping to manage and maintain community assets in their villages. One factor that caused some delays (mainly for infrastructure sub-projects under Component 1) was the difficulty of raising community contribution in some cases in a timely manner– especially those in cash, for some poorer households in the villages, and/or in years of drought in drier locations such as Moneragala. However, even with occasional delays, the 30% contribution level was consistently achieved for all completed intra-village infrastructure. With regard to larger investments under the pilot inter-village connectivity component, the expansion of other government infrastructure programs into some areas, without requiring cash contributions, reduced the incentive to contribute under Gemi Diriya. Here also, with investments involving up to 8 villages in a ‘zone’ of inter-village connectivity, it was more difficult to gather contributions from individual households, some of whose members also had less clear immediate benefits, and may also have needed to sacrifice considerable time to reach work sites. Another factor causing slightly slower start up and variable quality during earlier stages of the Project concerned the weaker than expected performance of Support Organizations (SOs), such as NGOs that were contracted for mobilization, facilitation of participatory rural appraisal (PRA) and community planning. With time, the development of an experienced and skilled cadre of community professionals (CPs) has provided cheaper and more indigenous support to member organizations, which were contracted by GDF to provide services previously handled by SOs. By the last stage of the Project, SOs was hiring CPs as village facilitators, and Phase II engages only CPs. Finally, the availability of technical services providers was sometimes a constraint. This was particularly true for water related infrastructure, and for activities requiring specialist livelihoods support to link communities to private businesses. Overall, it was noted during the ICR field mission workshops and interviews that more concerted convergence activities – to facilitate linkages with government and private agencies – could have accelerated implementation earlier in the Project. 2.3 Monitoring and Evaluation (M&E) Design, Implementation and Utilization The Project’s ML&E design was comprehensive and included participatory monitoring, an MIS, baseline survey, MTR and end-of-project impact assessment, six-monthly independent reviews of processes and procedures, and mechanisms to capture and incorporate process and learning aspects. The main elements of the ML&E system were implemented, with some exceptions noted below. The ML&E system evolved as project implementation moved forward. Project staff developed and refined these systems as they learned from experience. One example is Gemi Diriya’s development of a 9 village monitoring matrix (discussed below) to capture community level information in a systematic way. The learning system played an important part in making needed adjustments and fine tuning rules and guidelines. The learning and monitoring for a while were held on a monthly basis at the district and hub level. Finally, as noted above, over time the ICT pilot was scaled up and used for monitoring at the village level. Sub-project tracking and village organization status were monitored initially using a spreadsheet system that provided generally adequate to good data on Project progress and outputs. Close field supervision by management, monthly learning and monitoring for compliance checking, provided the main bases for addressing management, coordination and implementation issues, and adaptation of processes. The system had important participatory features that added value in terms of strengthening performance, particularly the use of Social Accountability tools, Village Monitoring Matrix, Score or Report Card by the VO members, who assessed the support from office bearers, hub office, service providers etc. As the volume of Project activities increased, however, large amounts of quantitative information and heavy staff workloads stretched the capacity of a basic paper and spreadsheet system. Since Gemi Diriya is an APL with two phases to come after the initial project would be completed, the task team and counterparts initiated the design of a web-based management information system (MIS) that would incorporate information down to the individual household level. The process of designing what became a very information intensive system took longer than expected. At the time of the ICR mission, staff and trained community youth were inputting the very substantive amount of information on hand, but had not yet produced any specific outputs from the MIS. The monitoring and learning system and MIS are being further fine-tuned and strengthened under Phase II. While there was considerable emphasis during Phase 1 on monitoring quantitative indicators of village organizational development and tracking sub-project implementation, with satisfactory results, the monitoring of livelihoods developments, qualitative factors and the identification of good practices for replication proved more difficult. Identifying capable agencies to contract for key ML&E products has been challenging. The baseline done at MTR, which also faced contracting problems, produced a proper survey. Process monitoring (to identify weaknesses, good practices for replication, etc., to help continually strengthen implementation), which was meant to be contracted out, did not proceed as expected because of the shortage of qualified agencies. On the other hand, CPs who were involved as part of IDA supervision missions, did provide excellent process feedback to peers from other Project villages. To assess key results and outcomes GDF commissioned an independent Impact Assessment Study of 2010 (IAS 2010) which was well implemented and drew on baseline samples comparable to end of project beneficiary households, and had a reasonable comparison to controls. The findings have been central to the assessment of the achievement of PDO (section 3.2 and elsewhere in ICR). Findings of the end-of-project IAS 2010 will also contribute to Phase 2 implementation. The Project log frame indicators largely captured the essence of the Project objectives, although there were some differences within PAD texts and tables. Greater emphasis on livelihoods indicators and rapid appraisal and communication of impacts might facilitate the learning of lessons and communication of important messages to policy makers. Measuring these types of impacts rigorously, in the context of projects with a mix of empowerment, social infrastructure, capacity development and economic activities is notoriously difficult, but a challenge that must be met. 10 2.4 Safeguard and Fiduciary Compliance Fiduciary. Financial management and procurement have been satisfactory throughout the Project. Systems and capacity for sound financial management and procurement were largely in place from an early stage, and were supported by transparent and clear procedures that were rigorously implemented and monitored by the communities themselves. VOs as companies have had to do annual audits and most of them have done this. Although Governance and Accountability Action Plans (GAAPs) had not yet been introduced by IDA at the time when the Project was being prepared, in the design of the Gemi Diriya program the task team paid very close attention to the issues that GAAPs would later address. A GAAP review in 2008 concluded that “good governance principles have been infused in its [Gemi Diriya’s] core values and embedded in project policies and procedures�. A formal GAAP action plan validated in October 2008 further enhanced this. Safeguards. As intra-village sub-projects were not large and were carefully located, and implementation involved heavy community participation and close supervision, no major safeguard issues were anticipated. The Project put in place an Environmental and Social Management Framework (ESMF) and a system of screening and checklists, and capacity building. With regard to environmental aspects, during supervision GDF and IDA detected some issues under infrastructure and agricultural livelihoods sub-projects (e.g. lack of drainage, minor erosion). For several reasons, only about 60% of EMPs were being formally implemented. One of the main reasons was the lack of capacity to guide the VOs at the field level, largely attributable to increasing staff workloads and therefore reduced follow-up monitoring (IAS 2010 report). This workload has been especially heavy due to the addition of the Gama Neguma inter-village component and the start-up of Phase 2 sub-projects. An innovative decision was to start building the capacity of a specially trained group of community professionals (CPs) to address this, but efforts will need to be intensified because CPs have not yet been mobilized in all areas. Generally speaking, the larger inter-village activities (under the Gama Neguma component) were being well handled under Phase I, but these inter-village sub-projects and the number of intra-village investments will be expanding rapidly under Phase II and therefore the mobilization of the trained CPs becomes extremely crucial to ensure adequate level of safeguards. Regarding social safeguards, the project did not resort to involuntary land acquisition and hence OP 4.12 was not triggered. Analysis of lands acquired by the project indicated that there has been no involuntary land acquisition by the project. Government, line departments, Pradeshiya Sabhas and temples, have provided or donated most of the lands. Less than a quarter of the land has come from private owners, mostly large landowners, individual land donations were very small in any case, and no complaints were noted. Small areas for improvement included: speeding up the process of release of Government and temple lands, and better maintenance of a land acquisition data base. Social assessment concluded that OP 4.10 was not triggered due to scarce presence of tribal communities in the project areas. Implementation arrangements included screening for the presence of such groups. Implementation experience confirmed that there were no groups of traditional tribals (Veddas) in the Phase I project villages. While no safeguard is triggered by the presence of ethnic groups like Tamils in plantation estate areas, the Project nevertheless put special effort to adapt manuals to support these communities, with significant positive effects (see 3.5.a). This targeting and support has been developed more strategically in Phase II. 11 2.5 Post-completion Operation/Next Phase Phase II of the APL was approved in 2009, purposefully overlapping implementation with Phase I, as originally planned. The main components, most of Project districts (although with expansion to new villages) and the implementing agency have remained the same, which has assured essential continuity for Phase I areas in terms of activity monitoring, follow-up support, operation and maintenance of sub- projects. The Phase II Project ML&E and MIS system covers the Phase I program areas, villages and organizations, with indicators only slightly modified vis-à-vis those of Phase I. Phase II also supports the graduation of VOs as members of federations and experienced CPs into skilled and employable facilitators. This is meant to ensure that the VOs continue to operate their infrastructure investments, and that VSCOs further revolve and build up their capital, and expand membership. Phase II is continuing funding to some of Phase I areas to finish some activities of Village Development Plans and the Interconnectivity pilot. Phase II is working towards the gradual mainstreaming of the Project model into Government programs like Samurdhi and Gama Neguma, through larger scale and more focused pilot activities. Phase II is expected to expand horizontally (geographically) to two new provinces, and vertically consolidate gains of Phase I through further linking VO and VSCO federations with private and public sector agencies and local governments. This is expected to develop more the technical capacity and market linkages of federations to support to their members’ livelihoods. Infrastructure sub-projects were mostly designed with good attention to technical suitability and quality, albeit with some issues concerning the occasional shortages of technical services. They were planned in a participatory manner such that ownership and buy-in meant that beneficiary communities contributed to the implementation and oversight, resulting in cost savings and quality control. This has meant resources have been allocated for operations, and means for cost recovery instituted, often to a degree not possible in public projects, such as through the installation of water meters for each household under water supply sub-projects. Nevertheless, while operational costs have been generally provided as needed, there is some concern that reserve funds have not been put in place, which could become an issue if major repairs should be necessary; especially for water related projects (see also under Sustainability). 3. Assessment of Outcomes 3.1 Relevance of Objectives, Design and Implementation The Project objectives continue to have high strategic relevance for Sri Lanka and remain well aligned with IDA’s country assistance strategy. Poverty remains pervasive in rural areas, and addressing this is critical both for growth and equity objectives, as well as for consolidation of peace. At the time of the ICR mission there were considerable political pressures to consolidate and direct donor funding mainly towards larger scale economic infrastructure. However, long cross-country experience has demonstrated that while such endeavors are extremely important, the benefits do not automatically ‘trickle down’ unless they are complemented by policies and investments that facilitate efficient livelihoods development at the grassroots level through critical ‘last-mile’, small scale infrastructure services, and empowering poorer communities to create their own opportunities for 12 growth. During the ICR mission interviews and workshops, communities, local officials and politicians - while suggesting small modifications, - attested to the strength of the Gemi Diriya approach and principles. The mainstreaming of Project elements - for example through the interconnectivity component (for Gama Neguma) in Phase I, the restructured IDA-funded Reawakening Project in 2008 and into the Samurdhi program in Phase II - also reflect the Program’s relevance for strengthening transparency, efficiency and good governance. 3.2 Achievement of Project Development Objectives The Project achieved and some cases considerably exceeded the PDO targets. The objective of Phase 1 was to improve livelihoods and quality of life of the poor and disadvantaged in the Uva and Southern provinces (and Sabaragumawa). The Project operated in Districts with the highest poverty headcount outside the North and Eastern conflict provinces. The Unmet Basic Needs Index (UBNI) was used to develop lists for Divisional and GND selection, and this method of targeting poorer villages was relatively successful (IAS 2010). The incidence of political influence appears to have been low. A visual comparison of poverty maps of selected divisions 3 tends to confirm that selected villages had higher poverty headcounts and were in less accessible areas. The initial Project target was to reach 150,000 direct beneficiaries in about 1,000 villages. By closing in early 2010, the Project was being implemented in 1,034 poorer villages,, with 183,898 4 households participating in VOs (equivalent to more than 90% of the 202,000 households in those villages). By the end of project 121,313 households were benefiting from 1,415 completed infrastructure works, and a total of 191,600 households are expected to benefit when all infrastructure sub-projects initiated under Phase 1 are completed. Further 96,853 households had accessed Livelihoods Fund loans from VSCOs; 12,427 youth had taken loans for skills development; and 10,645 poorest of the poor had received one- time grants. There is some overlap among the above beneficiary numbers, as many households with loans or grants are also be infrastructure beneficiaries. It is also important to note that number of beneficiaries of the various Project funds is increasing, as the latter revolve (VSCOs are revolving their funds to more members, and graduating the size of loans for repaying members, and youth loans are being recycled, with over half of skills loans have been repaid). With regard to the PDO indicators, based on the impact assessment and MIS data, the Project has achieved and exceeded its income, poverty alleviation and outreach targets (Table F (a)). a) Percent incremental increase in income (including from employment) to be generated from sub- project investments. While no specific target was set for this indicator, the IAS 2010 showed an incremental increase of 41% for poor project households taking VSCO loans, which is considerably better than the income changes experienced by the poor in control areas. This finding is based on households that had been sampled both at baseline (MTR) and in 2010. Incomes increased for the 3 World Bank and DCS (Department of Census and Statistics). 2005. “A Poverty Map for Sri Lanka: Lessons and Findings.� Policy Note 35605, Poverty Reduction and Economic Management, South Asia Region, World Bank, Washington, DC. 4 Gemi Diriya at a Glance, June 2010. 13 Project poor in Batches 0 and 1 (from Rs 2,727 at MTR to Rs 4,802 in 2010, equivalent to about 76%). In control areas, the average incomes of the poor rose from Rs 2,926 to Rs 3,949 (by 35%). Poor households overall in Project villages, i.e. not just loan beneficiary households, saw an increase in income of 27% more than did comparable households in control households. This was supported by the findings during the ICR mission, particularly the financial analysis of sub-project models (see Section 3.3) that showed incremental income increases that were 20-66% greater than ‘business as usual.’ This was confirmed in nearly all beneficiary workshops and interviews. Based on IAS 2010 survey data, close to 50% of youths who had taken loans got wage employment, contributing to 80% of them in turn having repaid their loans. It was estimated by the ICR team that those employed were earning on average Rp 13-19,000 per month, enough to support a family of 4-5 people. It was further estimated by the team that, given the number of youths who had gotten jobs through Project support, there had been an annual decrease in youth unemployment rate of 4.2% in Project areas compared to 2.2 nationally (which includes urban areas with more opportunities). b) The percent of households earning less than 1 USD a day was expected to be reduced by 10% (from 50% to 40%). The baseline at MTR indicated that the initial extent of poverty was greater than assumed in the PAD (71% in Project villages and 65% in control areas v. 50% assumed at appraisal). According to the IAS 2010, by closing share of households earning less than 1 USD had decreased to 44% in project villages and 51% in control village. The population earning less than 1 USD per day in Project areas was thus reduced more than 13% compared to controls. Also, the percent of poor non-loan beneficiaries in Project areas has decreased by over 24%, 10% more than in control areas. For Batches 0 and 1 beneficiaries only, who have had longest time to see benefits, the overall reduction is greater at 34%. c) The percent of households benefiting from increased income directly as a result of the Project was greater (52%) than the target (30%), based on estimates in the IAS 2010. This is a conservative estimate, as the IAS focused only on tangible benefits from livelihoods activities, but did not include other indirect benefits (e.g. from time savings). d) The percent of households benefiting from community infrastructure was expected to be 70% of 150,000 households (105,000). Some 121,000 households are already benefiting from completed and operating infrastructure (15% above, the original target). With 191,000 households expected to benefit once all the infrastructure is complete, close to double to original target would be achieved. The PAD also states that the Phase I Project would enable the rural poor to build accountable and self- governing local institutions and to manage sustainable investments. In summary the Project considerably exceeded targets of other intermediate outcome oriented key performance indicators included under PDO text in the PAD (summarized in Table F (b) and further described in Annex 2), regarding the functioning of village organizations, VSCOs operating in every village and covering operating costs, community led implementation and key aspects of managing economic activities. The broader institutional changes are described in Section 3.5.b. The outcome of Component 2 was to develop responsive and accountable local and national agencies and SOs (PAD p. 36 para 18). There were no clearly associated targets for this. Nevertheless, despite some delays in engaging PSs, decreased dependency on SOs, and at times limited availability of 14 technical service providers, the Project achieved considerable institutional capacity building, by (a) assisting the development of community owned and managed VO federations and Community Professional Centers providing services to their members, including some that had previously been provided by SOs; (b) engaging the PS officials to approach rural development in a more participatory manner; (c ) beginning to mainstream the Project in national programs such as Gama Neguma, and (d) introducing social and public accountability mechanisms like participatory budgeting, community reports cards, social audit and expenditure tracking, increasing responsiveness and accountability among government institutions. 3.3 Achievement of Triggers for Phase II The APL program had envisaged that Phases II and III will be initiated independently of the termination dates of the previous phase and each Phase will begin when the readiness criteria for expansion are satisfied. Moving from Phase I to Phase II was dependent on compliance with five performance triggers. The government has met or exceeded the triggers: S.No. Triggers Status 1. At least 60 percent of Village Organizations (VOs) in the first two years of Achieved Phase 2. At least 30 percent of women participate in decision making by holding Achieved management positions in the first two years of Phase 1 3. At least 50 percent of community members have contributed 20 percent or more Achieved toward capital cost of community infrastructure activities 4. At least 50 percent of members of project villages covered during the first two Achieved years of Phase 1 have benefited from project interventions; and at least 60 percent of these beneficiaries belong to the poorest households 5. At least 25 percent of the Divisions covered in the first two years of Phase 1 Achieved have established participatory sub-committees at the local government level with 30 percent elected representatives and 70 percent representatives of community organizations 3.4 Efficiency The substantive outcomes and impacts of the Project were achieved with considerable efficiency. Estimates indicate that the overall Project financial rate of return is over 30% (over full life of Project). This was based on conservative assumptions including all Phase 1 Project costs (including mobilization and management), only livelihoods income generation benefits at 2010 (end of Project) levels of intensity, and for infrastructure only using direct cost savings (not those saved by community as result lower input prices for example). The number and size of livelihood activities were expected to increase with time and infrastructure benefits would also be expected to accelerate. High value crops accounted for 55% of the VSCO loans and generated Rs 22,460 as average incremental benefit per beneficiary. Average incremental benefits for livelihood investments ranged from Rs 14,840 (cereal crops) to Rs 26,660 (animal husbandry) per beneficiary. Overall the project 15 economic rate of return is 30.5% and project financial rate of return is 32% (over full life of project, including all Phase 1 Project costs-including mobilization and management). This is based on conservative assumptions, i.e. limiting livelihoods income generation at 2010 (end of Project) levels of intensity, and infrastructure benefits to only cost savings. The number and size of livelihood activities is expected to expand and diversify with time and infrastructure led benefits would also be expected to accelerate. The savings in infrastructure costs realized by using the Project approach were significant. Considering a range of infrastructure types in different areas, and comparing them to equivalent government contracted works, the ICR mission calculations showed an average savings of 36% under Gemi Diriya. This conclusion was echoed by local officials, politicians and community members who consistently estimated savings of between 30-50%. Savings to government would additionally increase when taking into account the communities’ 30% contribution in cash and labor. The quality of infrastructure was also good – local officials estimated that it is likely that completed works would have 25-30% longer lifespan compared to equivalent government constructed works, and improved beneficiary incomes would better position them to be able to contribute to sub-project O&M needs. Several of the CDD features of Gemi Diriya contributed to these positive findings on efficiency. Through the livelihoods support process and easy access credit system, people started with activities they were familiar with and then took larger loans; the good repayment rates meant loan funds were recycled for other people to use; and this was sustained by a rigorous village microfinance system. Community led investments in skill development achieved cost efficient job placements for unemployed youth in project villages: (a) average cost of job creation is only Rs 12,550 per job, and 58% of the skill development loans have already been re-paid, which will generate more jobs at reducing costs ;and (b) more jobs were created for rural youths in project village, and this helped to reduce the annual youth unemployment rate by 4.2% in the project villages as compared to 2.2% for the country as a whole. The infrastructure savings arise as community selection and contributions meant greater ownership of prioritized works, there were no or very low overheads to pay, the community contributed labor (see above), and the community supervised implementation in detail, reducing waste. Overall, strict adherence to rules, accountability, community decision-making and inclusion, reduced leakage and wastage. Also communities and federations were increasingly taking over facilitation, management, capacity building and supervision functions, which yielded further savings in the support costs. 3.5 Justification of Overall Outcome Rating Rating: Satisfactory Based on ICR mission findings the Project has achieved a Satisfactory overall outcome, reaching most of its objectives with continued relevance and satisfactory efficiency. It has also had satisfactory outcomes in terms of institutional change with strong community institutions and contributing to changes in government programs. However it has yet to mainstream the capacity development of government agencies that might eventually take over the development tasks, although this is the aim of the overall 12-year program and not just of Phase I by itself. 16 3.6 Overarching Themes, Other Outcomes and Impacts (a) Poverty Impacts, Gender Aspects, and Social Development The assessment of the achievement of the PDO above notes the considerable success of the Project in reaching out and successfully supporting a broad range of beneficiaries in poorer Project villages. From the results on impacts (3.2) it is estimated by the ICR team that the number of people who have thus rise above 1 USD per day the as roughly equivalent to 60,000 households (or about 300,000 people) Specifically the Project reached out consistently to the poorest of the poor through one-time grants, to youth from poorer families, and women, as well as workers in the estates– all sectors which government programs - even poverty targeted ones like Samurdhi – rarely reached. . The Project developed a model for working in traditionally poor plantation estate communities that was tested successfully in about 60 villages. The estate sector communities interviewed under the ICR mission noted with great appreciation the sensitive, effective and comprehensive approach to the Project’s support. They also noted the opening up of new livelihoods, which could combine their wage earning with small but significant agricultural activities. Sub-project identification and design were highly inclusive of the poor, with a criterion of approval that 80% of the poor should benefit, and special prioritization should be given to the poorest of the poor – and identification of the needs of remoter, poorer hamlets. Participatory wealth ranking was based on identifying poor and poorest of the poor, by the communities themselves. The VSCOs provided an unprecedented opportunity for poor to save and borrow compared to the situation in control villages, with 63% increase in household saving. The one time grant support was also seen as direct tool to graduate the poorest of poor – disabled, widows, etc, with a support on using the grant for livelihoods activities, so reducing dependency on handouts and letting them participate in the VSCOs’ regular lending. Cases of very poor Samurdhi beneficiaries graduating out of the Samurdhi system were reported during ICR field interviews. Infrastructure sub-projects identified and implemented by communities met acute needs typical of poorer members of rural communities. Clean water supply and sanitation constituted 11% of projects being implemented. The Project had notable positive gender results: A large proportion of VO leaders (66% Board of Directors) were women, and their positive influence on the Project implementation overall was noted frequently in ICR mission interviews and workshops, as well as highlighted by local politicians. The VSCOs' membership was primarily of women (71%) and their leadership dominated by women. As a consequence loans were taken mainly by women. About 20,000 women leaders have emerged as a result of the project. During the ICR mission, with great consistency communities and other stakeholders mentioned the much increased community unity and joint action, greater respect and confidence they had as a result of the Project – especially the ability of women, youth and the poorest to stand up and effectively voice their views and choices. This applied equally in traditionally marginalized communities such as in the estate sector, and in addressing long running issues of for example alcoholism in some communities. (b) Institutional Change/Strengthening 17 The project has had significant results in terms of institutional change. The sustainability and engagement of the community institutions was central to the Project and has been satisfactorily achieved (see intermediate outcomes in Section 3.2). The organizations show strong implementation and self- sustaining capacities that have been built up through refined and well-tested process of community mobilization, strengthening, and accountability. Specific efforts have been made to support and measure the community organizations sustainability. For VOs a regenerative capacity index shows that 72% of VOs have generated 50% or more of the original Project funds invested in them and 22% of them more than 100%. Just under half of VOs now form part of federations, and the federations are operating business services to members, as well as enterprises that generate incomes to the village organizations. Some Federations of village credit organizations are now addressing Phase 1 villages’ service needs, such as further capacity building on bookkeeping. A notable feature is the rigor applied in sustaining village microfinance organizations (VSCOs): all 1,034 of them are covering own operational cost, and have in place innovative security and welfare funds, to cover risks. The loan recovery rate is (92%), greater than for local microfinance institutions, even when VSCOs are reaching poor beneficiaries not generally clients of MFIs. However while all can cover operational costs, there is some concern that portfolio at risk (PAR) is greater than 10% for 60 days, for over 30% of VSCOs, which is below financial sector levels for sustainability. This is partly because VSCOs are not writing off loans, despite droughts. What will need to be worked out soon under Phase II is the legal relationship between VOs and VSCOs – particular the financial dependency of VOs on VSCOs for maintaining operational costs, and the respective contractual and debt obligations should either of them fail. A range of new important skills and capabilities have been built up in the Project communities: from business management to managing information technology. With this wide range of community skills, poor villagers are now taking over service provision that was previously done by external service organizations. Community professionals (CPs) are beginning to provide experienced services to other areas and agencies, as contracted service providers. CPLTCs and CPs from 787 VOs are providing paid services to the Project and to other projects, such as CARE. CPLTCs are covering their costs and also paying VOs for services rendered by CPs. Overall Gemi Diriya Foundation (GDF) as an institution has been fundamental in the implementation oversight and evolution of the Project’s core community development model, especially its adaptation to scale. Despite early difficulties in recruiting and maintaining staff, the Project developed an innovative and award winning Performance and Personal Development Appraisal (PIPDA) results based system. Part of GDF’s HRM strengths has been increasingly target oriented planning, of devolution of responsibilities to Regions and District teams, and downward accountability scorecards, to ensure strong service oriented orientation. Two key challenges for GDF at the end of Phase I and beginning of Phase II were to develop its means of mainstreaming management systems into key government agencies by doing more capacity building and lessons learning role while maintaining community participation and technical quality at scale. All this, while under considerable political pressure to accelerate handover of Project management to Government agencies. With regard to local government institutions, by design in Phase 1 Gemi Diriya was relatively disengaged, with plans to increase linkages and cooperation progressively during Phases 2 and 3. This was noted in the MTR and, given Government’s desire to advance more towards integration among its 18 major poverty interventions, by end of Phase 1 this changed. With infrastructure interconnectivity projects linking several villages, Pradeshiya Sabhas (PS) were fully involved and local politicians very supportive of Gemi Diriya – a point not always fully recognized at Ministry levels. The interconnectivity projects gave District and PS staff an enhanced role in engaging with communities and in performing their technical functions. The PS officials also described instances in the application of Gemi Diriya principles (for example contributions and community participation) in other local projects, with consequent improved quality and savings – even to the point of being able to purchase further trucks and bulldozers for infrastructure works. By engaging local staff to management positions at District level close involvement in District coordination and planning meetings, the Project could further build linkages for Project communities and raising awareness of the Project with local governments. The integration of the Gemi Diriya approach into other government programs such as Gama Neguma and Samurdhi has been an important aim, accelerated in Phase II. . Early lessons from the Phase I Project have led to key features have been adopted in these major poverty alleviation programs: GOSL has allocated Rs.26 billion (approximately USD217million) for financial year 2009, for Gama Neguma to respond to village level priorities using some of the principles of the GD approach. A new Village planning system is to be applied through in all the country’s villages from 2011 in preparation for the country’s next long term plan. All Gama Neguma sub-projects outside Phase 1 and 2, are also meant to raise community contributions. The Samurdhi Program, which has had serious leakage problems, has adopted a more comprehensive targeting methodology, influenced by Gemi Diriya. (c) Other Unintended Outcomes and Impacts (positive or negative) The Project also resulted in unplanned positive households impacts. Some examples, by mid 2010, were: (a) Inter-connectivity pilot activities were being completed and were to provide mainly road linkages between villages and to larger roads and benefiting 8,000 households, half of which are non- Gemi Diriya. As a consequence it was noted that land prices are going up in adjacent areas. Some of the roads would also be providing shortened routes between Provinces greatly reducing time and cost of transporting products to markets. (b) Certified paddy seeds produced by Gemi Diriya farmers are estimated to supply 14,000 ha of grain paddy, resulting in over 500 million Rs incremental incomes. Communities and other stakeholders reported numerous impacts from roads and water: on health and schooling, and overall a large percent of beneficiaries could identify improved housing and assets, including amongst poorest of the poor, as a result of the Project and (c) Further, the Phase I Project has contributed to the peace process by extending the core methodology to the IDA-supported Community Livelihoods in Conflict Affected Areas Project (also called Re-Awakening; P086747). 3.7 Summary of Findings of Beneficiary Survey and/or Stakeholder Workshops The ICR mission got direct information from representatives, of around 200 (20%), of project villages through direct interviews and workshops covering all aspects of the project. Main findings are detailed in Annex 6, incorporated in ICR findings and summarized below:  The formation of the village organization (VOs) and the requirement wide participation in decision-making, brought unity and inclusion (especially of youth and women).  The rules of participation in the project cycle and investments (with direct funding to communities and counterparts) resulted in community ownership and cost efficiency. 19  The VSCO has provided access to loans by people who had not previously had any such access, at interest rates they could afford, allowing them to escape from dependency syndrome.  A range of household impacts from improved loans and livelihoods activities.  Infrastructure greatly improved living conditions. For example, irrigation investments allowed up to three crops a year and reduced conflict over use of scarce water resources.  A greater confidence and sense of empowerment to talk to government agencies. Suggestions of things that could have been better, included ways to facilitate the 30% community contribution, legal advice and getting insurance for crops and livestock. 4. Assessment of Risk to Development Outcome Rating: High Important and successful mitigation measures implemented were: (i) ensuring that GDF would have a competent CEO and independence to use well defined technical criteria and a transparent process for the selections of beneficiaries, GNDs and Villages; (ii) intensive information and communication campaigns; (iii) resources for capacity building of Village Organizations; and (iv) emphasis on community participation in the identification, preparation, and implementation of investments. The development outcomes achieved by the end of the Project are likely to be sustained. The beneficiaries who have taken loans are likely to sustain their income generating activities, as these have been shown to produce good returns at household level. Nonetheless, several issues/risk factors elaborated below will need attention during Phases 2 and 3 of Gemi Diriya.  Contributions for community infrastructure sub-projects were in some cases difficult to collect, and funds for long-term maintenance and repairs are not always in place.  While the Gemi Diriya model is robust, it is evolving and becoming more complex. Therefore it does need continuing monitoring and capacity building, especially to strengthen the relationships between VO federations, VSCOs, CPLTC and other agencies. Sustainability of VSCOs will need further attention. VSCOs have shown already many strong features, but their PAR needs to be assured for them to be become truly viable for an increasing number of beneficiaries and larger sizes of loans. Finally, at the time of the ICR, and following the national elections of May 2010, the GDF has been transferred to the Ministry of Economic Development. While the new Ministry basically replaced the previous Ministry of Nation Building, with assurances that there will be no change in the scope components and modality of implementation of phase 2, the needs of accelerating the integration of GDF into the regular state apparatus, without consideration to lessons learned during Phase 1, have created huge uncertainties in the terms of the timing and process with potential impacts affecting the implementation of Phase 2. This is the main reason why the risk to development outcome has been given a rating of High instead of Low. 20 5. Assessment of Bank and Borrower Performance 5.1 Bank Performance (a) Bank Performance in Ensuring Quality at Entry Rating: Satisfactory The Bank’s role in ensuring quality at entry was satisfactory. Preparations were adequate and could draw on considerably detailed lessons from the VSHLI pilot project, directly scaling up the lessons with well-developed procedures, as well as further developing the model. The implementation and institutional arrangements put in place were on the whole appropriate to the context at appraisal, i.e. the need for independent foundation to demonstrate the effective implementation of a participatory and multi-sectoral poverty reduction process. Safeguards put in place and risk assessment were satisfactory. (b) Quality of Supervision Rating: Satisfactory The quality of supervision by IDA was satisfactory. The task team, through frequent supervision, hands on workshop type support and facilitation of cross-country visits to good practices, helped GDF to develop the capacity and procedures to make certain core Project features were in place and operating smoothly throughout the Project. IDA missions had an adequate skill mix; they were proactive in the identification of potential issues and got involved in discussions with relevant levels of government. The good relationship and complementarity between staff located in the IDA’s country office and headquarters ensured consistent follow-up of agreed actions, as it did the fact that IDA was able to ensure continuity of the project’s team leadership. At MTR, the need for some adjustments were noted and subsequently implemented, including the addition of a component to pilot Gemi Diriya processes within the Government’s Gama Neguma program, and initiating adequate transition arrangements through the timely initiation of preparations for Phase 2. The task team received a Bank internal recognition for its supervision. An area where the task team could have put greater attention was in some aspects of M&E. By the end of the Project the impact outcomes were satisfactory, but the tracking of emerging PDO level outcomes could have been monitored on a more continuous basis throughout the Project. This would have helped to sustain more policy and politically relevant messages at later stages of the Project, where economic rather than process related results would have greater influence. There could have been greater clarification and revision of some indicators, with supervision focusing gradually more on fewer monitoring indicators at the outcome level, and less on every detail of process. On a minor point: counterparts noted that there were quite a number of IDA consultant missions, many of which made detailed information requests, and some findings were either not disseminated or led to considerable complexity (e.g., the case of the MIS development) which posed burdens on Project staff and communities. 21 (c) Justification of Rating for Overall Bank Performance Rating: Satisfactory The overall quality of Bank supervision was rated as satisfactory. In the context of achieving the project development objectives, it must be highlighted that the borrower widely acknowledged the Bank’s support and collaboration during supervision as a strong contributors to the success of the project and to the Gemi Diriya Project, and the approach a model for other development programs in Sri Lanka. 5.2 Borrower Performance (a) Government Performance Rating: Satisfactory In the design and initiation of the Project, the GOSL showed considerable commitment to the community driven development approach, as proven during the implementation of the VSHLI pilot, and during implementation by supporting the Gemi Diriya Development Foundation, by respecting its autonomy, by providing adequate and timely counterpart funds, and by ensuring compliance with all credit covenants. Fiduciary and financial management aspects have been well maintained throughout the Project. The Government showed additional commitment to the emerging model, by introducing some key aspects of the Gemi Diriya model into key government programs (see section 3.5(b)), as well as fully supporting the development of Phase 2. There were some initial delays in putting in place staffing and resources in the beginning of the Project, leading to about a 6-12 month delay, but otherwise the Government did made sure that staff was selected according to GDF technical criteria for staff selection, and ensured continuity both of GDF staff and management through the life of the project. The role and functioning of the Board of GDF did not result in many substantial inputs in terms of policy, largely paying attention to administrative aspects. Nevertheless with access to the Minister (who also was the president of the board) the foundation was provided with the support it needed to effectively implement the project. (b) Implementing Agency Performance Rating: Satisfactory There were only minor shortcomings in implementing agency’s performance. The GDF implemented and scaled up effectively a community development model, combining institutional development, rural infrastructure, microfinance, livelihoods, while maintaining high community participation. It did so in an accountable manner, achieving or exceeding its main targets. The implementation agency’s interactions with community beneficiaries, putting them at the forefront of the development process was exceptional. Issues raised during supervision missions were usually addressed promptly. While at times there were minor shortfalls in the capacity and clarity of functions of staff roles, these were rectified, and systems were put in place to address them. While M&E had some weaknesses (see section 2.3), it was adequate to the day-to-day function of managing the Project. Streamlining information flows between different levels of the implementing agency was a recurring issue, but a natural challenge in a 22 Project that is growing in scale and complexity. Time pressures on staff to achieve outputs, also meant that the Project did not clearly identify and implement village specific exit strategies consistently. A minor shortcoming was that clear mechanisms for greater linkage and convergence with line department programs and local development coordination could have been developed earlier. Similarly engaging local politicians, at least in policy discussions and to experience good examples of the village development process was not done enough till the latter half of the Project. Similarly more emphasis could have been placed on policy oriented studies to accelerate political support and engagement. (c) Justification of Rating for Overall Borrower Performance Rating: Satisfactory Overall borrowers’ performance takes into consideration both the government and the implementing agency’s performance during preparation and implementation. On the basis of the justification provided above, the borrower’s overall performance was rated as satisfactory. At the time of project closing, the borrower had complied with the covenants and agreements needed to achieve development outcomes. 6. Lessons Learned During the course of the Gemi Diriya Phase I several important lessons were learned in terms of delivering inclusive and efficient rural poverty alleviation. It has been clearly been demonstrated that in Sri Lanka, through the participatory approach (CDD), investments and services can be delivered in an efficient, accountable and sustainable manner - with considerable impact, and at scale. Most of the lessons derived have already been incorporated in to Phase II. Key lessons are that:  Longer term continuity in development support that preserves community decision-making and participatory institutions, and helps them graduate, is critical for sustaining and expanding local efforts. Foremost among the factors contributing to the strong performance is the soundness of the Project village development model with its strong buy-in from the community.  Involvement by the whole community, especially the poor and marginalized, is essential to ensure that ownership and commitment to development efforts is maintained. The inclusion is necessary to focus efforts more on people who need to benefit most, and to strengthen the quality of outcomes. This process has to be supported by simple and clear principles (participation, inclusion, transparency, accountability and cost sharing), which are agreed up-front with communities.  Leadership by experienced community members in the development of procedures and transfer of knowledge to other communities is a highly effective and efficient method for scaling-up. A process of using regular leader workshops to update and disseminate the Community Operational Manuals, created profound ownership and consistency – based on openness and fair rules. Further, through the use of community professionals and the development of their training centers, not only reinforced the communities’ sense that the Project process is theirs, but helped to reduce costs and scale-up faster. Local CPs as facilitators, also understand better the situation of local households and communities.  Participation of women and youths in village organization decision-making positions, has given them voice, empowerment and improved their own and the community’s socio-economic situation. The involvement of women was noted as a key factor underpinning the integrity of implementation. This is 23 especially so in the community micro-financing institutions that is primarily managed by women, and strengthen women’s critical role in household livelihoods systems. Similarly youth groups have been especially successful in fostering greater ethnic integration and identification of work opportunities. Development of employable skills and work linkages for youths also results in the release of major potential within households for generating additional incomes.  With households scaling-up or graduating from individual livelihoods activities, it becomes even more important to develop synergy and convergence between infrastructure and livelihoods investments, on an increasing geographic scale. Here strong federations will need full support to come together with local governments, work with enhanced District coordination, draw on local political support, private sector advisors and investors, and financing institutions, to ensure that their communities broader needs are still met in an accountable and efficient manner. Such efforts require a careful institutional analysis, some of which has been initiated in the preparations of Phase II.  Learning by doing and piloting. Experimentation during project preparation is critical for reducing risks and building confidence and capacity of the staff and communities to scale up under the Project.  A programmatic approach, with adequate flexibility to incorporate innovations and sufficient technical assistance for piloting and testing at each stage are key success factors, which help both in building the confidence of staff and communities, fine tuning the model and enhancing prospects for achieving development outcomes. At the level of the Borrower and World Bank the long-term 12 year commitment (in the form of an APL), and consistency in team leadership, have also without doubt been critical in being able to provide detailed technical and goal oriented support for programmatic learning. Such commitment and support have been shown globally to be essential for successful scaling-up.  A relatively autonomous implementing agency with flexibility is critical for testing of new models, quick results, compliance to the rules and maintaining political and operational independence for piloting and scaling-up of the Project. The challenge will be to ensure the continued evolution of this role to focus on institutionalizing the model, by building capacity of regular Government agencies and programs, and addressing key policy and administrative issues.  A very hands-on top management approach and high staff commitment resulted in close monitoring, troubleshooting and continuous refinement of activities. An incentive and performance based recruitment system is critical for showing results and maintaining quality of the project. The challenges that emerged during the implementation of the Project also provided lessons for implementation aspects that could be improved. These are above all significant with increasing scale and diversity of project components: While preserving the basic principle of generating ownership and local resources it is important to explore flexibility in how contributions are planned and raised, so as to reduce the time that communities (particularly by the poor) need to start investments. Part of this is to a great extent about how much can be raised reasonably by which community groups, in what timescale. More contributions should be collected early, when there is greater interest. Perhaps smaller labor contributions are needed for infrastructure projects connecting several more difficult to define beneficiary communities. Modified contributions could be graduated according to ability to pay. 24 Operation and especially maintenance is as always a continuing concern for sustainability and planning and support into the medium and long term needs a concerted effort, especially for potential major repairs. This will require setting up dedicated funds and/or local government financing. With increasing scale of project, and evolving lessons and activities, there is a real danger in M&E information flows becoming overwhelming. An intensive effort needs to be made to clearly allocate ownership and responsibilities for information collection and processing, tied in as closely as possible with the relevant management responsibilities and interests, from the community level up. How such information and processing will be used and maintained in the long term through developing and changing institutions (federations, local governments, government agencies) has to be also considered from the start. 7. Comments on Issues Raised by Borrower/Implementing Agencies/Partners In their electronic letter dated December 29, 2010, the Implementing Agency commented as follows: “The draft Assessment report was prepared by the FAO/WB Team independently and the performance of the borrower and the Bank rated “Satisfactory�. After carefully scrutinizing the Bank ICR we have observed that the Project Development Objectives have been achieved and majority of the indicators targets have been exceeded. The Bank report under many indicators has shown that PDOs have been achieved over 100% in most. One might argue that in that case why it is not rated as “excellent�. Also any outsider might also see and ask whether the targets set at the beginning were rather low. In order to avoid these negative aspects it is better if reported as targets have exceeded by percentage, which would be more palatable for the reader. The GDF still believes the project performance as “excellent�. The VSHLI Pilots at the appraisal stage (beginning) were only with 12 villages and it has subsequently increased to 47 villages. Hence when evaluated at the end the project, performance indicators show very high achievements. These subsequent changes need to be considered when working out the achievements rather than indicating achievements of over 500% under some indicators. The convergence aspect has been of moderately successful and need to strengthen further during Phase II of the project. The Samurdhi Villages model and the Demonstration village programmes have been initiated during the latter stages of the project and hence they need more time for consolidation through capacity building of the state sector line agency staff. The scaling up through the Pradeshiya Sabha was also brought in the latter stages of Phase I and these took some time to speed up. Since this been a novel experiment with the local political leaders and the community initially resistance was observed. However it gradually improved and many local politicians backed this model by now. Thus the project has been able to scale up the CDD model from the village level to the next, closest local government level and will be expanded further with more local authorities joining within the province to see its impact. Community Driven Development in the long run will evolve as a self governing, community managed, less bureaucratic and politically free approach which will reduce the cost of delivery in development to the government. Initial draw backs, resistance and criticism are normal under the prevailing situation where the top- down approach has been the practice for several decades. However eventually all levels will realize the development benefits of this approach but it needs time. The Adoptable Program Loan (APL), the flexibility of the programme and constant technical input and supervision by the Bank team has enabled the programme to be grounded in the country�. 25 Annex 1. Project Costs and Financing (a) Project Cost by Component (in US$ million equivalent) Appraisal Re-allocated Estimate Percentage of Appraisal Actual /Latest Estimate Estimate at MTR Components (US$ million) (US$ million) IDA Share IDA IDA Share A. Village Development 40.5 34.3 35.3 87% B. Institutional Strengthening 3.5 3.0 2.07 59% C. Innovation Seed Fund 1.00 1.0 0.05 4% D. Project Management 2.50 4.5 4.70 188% E. Village Self-Help Learning 1.00 1.0 1.00 100% F. Unallocated 4.0 0.0 G. Gamaneguma Pradeshiya Sabha 1.0 1.35 Interconnectivity H. Tsunami 8.0 7.7 Total IDA Cost 51.0 51.0 52.17 83% Total Project Financing Required 69.80 58.11 83% (b) Financing Appraisal Actual/Latest Percentage of Source of Funds Type of Financing Estimate Estimate Appraisal (US$ million) (US$ million) Borrower Budget provisions 11.00 5.35 50% Local Communities Cash, labor & 7.80 3.66 46%* materials IDA Grant 51.00 52.17 100% Total **Includes only contributions to infrastructure projects. 5 The borrower financing was less than estimated at the appraisal time due to two factors: (a) Tsunami – given the priorities of the government, the counterpart funding was reduced for on-going operations; and (b) Gamaneguma funds -some of the counterpart funding was directly released to some communities under Gamaneguma funds and is not included in the above borrower costs. 6 The actual community contribution is much higher than US$3.6 million. The above community contribution includes only 30% of community contribution of 1415 completed infrastructure sub-projects. However, the total number of infrastructure subprojects in much higher and would be completed under Phase II. Secondly, the above amount of US3.6 million does not include the time that the communities have devoted, as implementing agencies, to plan, manage, and implement their village development activities; and their contribution to livelihood promotion. The community contribution would be much higher if those factors are costed. 26 Annex 2. Outputs by Component Component A - Village Development This component had two subcomponents: (1) Development and Strengthening of VOs; and (2) Funding of Community Sub-projects, forming the core of the project activities, and the main allocation of the resources. Sub-component A-1: Development and Strengthening of VOs. The Project aimed to develop self- reliant, self-managed, inclusive and sustainable village organizations (VOs) under a company model. With technical services and training by support organizations (SOs) and GDF district and hub teams, the VOs would gain the capacity and be accountable to their membership for planning, implementing and managing village development plans (VDPs). The Project would also promote partnerships between the VOs and the private sector to expand livelihood options for the poor. During VSHLI and Gemi Diriya piloting, one of the lessons that emerged concerned the need for strengthening of the ‘last mile� of support to communities – hence the need for field teams to be competent, empowered and accountable for results. The project therefore adopted a recruitment strategy that focused on aptitude and attitude of the staff, their commitment and willingness to serve in remote areas. There was focus on both staff training and incentives to encourage good performance and results on the ground. These incentives included simply recognizing the teams, promotions, annual salary increases, etc. SO support was needed during Phase 1, as there was not much staff capacity in the beginning and the community professionals were only emerging. Key lessons were learned from that experience as well during the hiring of the four batches of SOs. Each batch improved and their contracts were performance based. Most SOs, including the more well established ones, told the ICR mission that although there was not much money for them in the contracts with Gemi Diriya, there were huge opportunities to learn and gain knowledge which helped them with their own programs. The Community Professional model also proved far more successful than initially anticipated. CPs started helping soon after Batch 1 – and with every year the numbers increased and their thematic reach widened. The CP center was formally established in 2007, and by completion of Phase 1, CPs had taken on most roles originally planned for SOs. Finally, although initially GDF staff wanted to hire technical service providers, the decision was made to use community procurement so that the TSPs would be accountable to the communities and the latter would have options from which to choose. This brought in a new culture in Sri Lanka, making TSPs accountable to the people, at reduced cost and with improved quality. It also forced the project to think through strategies for addressing issues concerning availability TSPs. One of the strategies used is the training of CPs on infrastructure topics through the Institute for Construction Training and Development (ICTAD, reputed engineering training institution). A final lesson learned was that staffing and support needs change as communities graduate from one level to the next and the project needed to adjust to these changes during implementation -- e.g. the evolving need for staff and resource persons on producer groups/business orientation, and skills for 2nd generation institutions, federations, and inter-village infrastructure. 27 GDF led the initial information and education campaign to raise awareness and understanding of the principles of the Project in 1,0347 villages. This was done first in pilot villages during preparation and then in four batches of villages during implementation (2005, 2006 and two in 2007). The Project engaged over 55 SOs, with village facilitators (VFs) for community mobilization and training. It emerged relatively early on that several of the SOs, and most of the VFs had little direct experience with CDD type approaches and thus themselves needed considerable follow-up training and support from GDF. GDF assisted communities in identifying technical service providers for infrastructure sub-project design and construction, and potential employers for youths taking skill development loans (see next sub-component). In 2009, the Project also facilitated ‘topical’ PRA with village transects and other methods to assist communities identify livelihoods opportunities more strategically (4th Quarter Progress Report 2009). Based on the experiences under the VSHLI project, federations were set up in all districts, primarily facilitated by GDF, and about 48% of VOs form part of these federations. Many of them have been assisted to form business linkages, run trading activities to generate income, and some are providing support services to their VOs. Again based on the VSHLI experience, GDF and VOs identified and trained 1,473 community professionals (CPs), of whom 73% are women and 30% are poor. A majority of VOs (787) have CPs, who has a range of areas of expertise which they are using to support other Project communities, and thereby are increasingly earning an income for themselves, as well as resources for their VOs. The CPs has earned a total of Rs 41.6 million in professional fees, or about Rs 26,000 per person. Over 40% (411) of VOs are earning income from CPs, against a target of 25%. With the relative lower cost of CPs, and their increased skills, they are beginning to assume VF functions: in batch 4 villages all VF positions under SOs are held by CPs. Sub-component A-2. Community Sub-projects. The bulk of Project funds were to be transferred directly to the VOs to finance their capacity building and priority investments. The budget was allocated depending on the population of the village communities. The communities decided the final proportion of funds between different activities. A-2-i. Capacity Building (CB) Fund. About 10-12% of the village development funds were earmarked for capacity building. The objective of the CB Fund was to build institutional and human resource competencies and skills of VO office bearers and community members, to empower them to plan, implement and manage village development activities. A key outcome of this support would be to reinforce subcomponent 1(i) to ensure sustainable village level institutions were built, and skills base developed, through a demand driven process. 7 The total was 1036, but only two villages were dropped out (less than 0.1%), so the figure of 1034 has been as Project villages in the report. 28 All communities drew on the first installment of the capacity building fund, and nearly all drew the second (4th Quarter Progress report 2009). The resources were used to set up VO offices and basic equipment, meet office management costs, obtain training on legal aspects of village companies, training for the different VO committees, exposure visits to other VOs, technical assistance for the design and supervision of sub-project implementation, and specific training for youths and women. Particular attention was paid to training of VSCO small group treasurers, key persons from the standpoint of accountability of the community institutions. Training was provided by GDF, SOs and others. The MTR report flagged that the sheer volume of training might be causing some loss of enthusiasm by communities, especially when attending training events entailed travel and expenses. Nevertheless, overall VO members appreciated the training. They considered exposure visits and visual communication tools particularly effective the latter helping them to disseminate their messages quickly and monitor progress. Secondly, CB funds were used for Community Information and Monitoring Centers – the conditions for these being that VOs could demonstrate their ability maintain and operate them on their own. Third, the communities invested in developing CPs, which inter alia held the promise of generating some future income for the VOs (5% of CP fees revert to the village). A large number of women (about 20,000) have been trained specifically for the VSCOs. These are not CPs, but rather female and male leaders in the villages with capacity to implement other programs as well. The Project in its design adopted a radical approach in that it allocated resources directly to communities to be used by them to access technical service providers. As noted in the main text, overall this worked and community infrastructure sub-project quality has generally been of high standard, albeit with considerable assistance from GDF technical staff. Nevertheless, there were at times constraints in identifying available technical staff, mostly from line agencies, to provide the necessary services and supervision, especially for water related infrastructure. A-2-ii. Community and Social Infrastructure Activities and Services. The objective of this sub-component was to empower communities to manage the execution of infrastructure activities and expand the scope of social services accessible to them. The beneficiaries were expected to contribute at least 30% of the cost of such activities, 10% of which would be in cash. The communities identified over 4,387 sub-projects in their VDPs, these have all been approved by the Maha Sabha, though in a few cases priorities were changed when constraints arose under design. Some 3,528 of these have been initiated, of which 1,414 are completed, and 42% were at a stage of 70% completion or more, with most sub-projects with high likelihood of being finished within 1-2 years, although 33% of sub-projects were not yet completed in the earliest pilot and batch 1 villages balance to be covered by Phase II funds. There were no major differences in the completion rate between different types of projects, although percent of water supply projects not yet complete was slightly lower (25%). Communities contributed a total of Rs 407 million (USD 3.6 million) - Rs 169 million in cash and the equivalent of Rs 238 million in labor (IAS 2010), equivalent to 26% of total 29 infrastructure sub-project costs (Rs 1,596 million) 8 . The total amount of contribution will increase when all the projects will be completed. The percent of beneficiaries contributing to agreed cost sharing was 87%, considerably over the project plan estimate of 60%, showing a high acceptance and understanding of this condition. Communities interviewed during the ICR mission were very satisfied with the targeting, appropriateness, quality and impact of infrastructure works, and in several cases had set up mechanisms to cover operational costs, such as water meters for individual houses. The significant impacts and cost-effectiveness of the infrastructure sub-projects are discussed in Annex 3: for roads, which were the largest share of sub-projects, impacts were in the form of reduced travel time, lower transport costs, lower input prices and higher sales prices, in addition to social benefits such as quicker and more reliable access to hospitals and schools. As indicated previously, while the number of sub-projects completed as of the March 31, 2010 grant closing date exceeds Phase 1 project targets, there are additional sub-projects which are in various stages of completion. Project benefits figures in this ICRR include only those deriving from the completed investments, but there will also be significant post-completion benefits attributable to Phase I. The reasons why some sub-projects were still in implementation at the closing date are that: (i) Batch 3 and 4 VOs began only in 2007; (ii) some sub-projects, especially water supply for which land was needed, required numerous clearances by the administration to transfer public lands to beneficiaries; (iii) some communities took more time to raise contributions, especially cash, due to droughts or the de-motivating presence of other similar government programs which did not require contributions; (iv) some VOs faced constraints in identifying necessary technical service providers; and (v) a few sub-projects had to be redesigned due to complex technical issues, such as gradual reduction in quality of groundwater. One such case in Polonnaruwa highlighted also the Project’s flexibility and capacity to motivate cooperative action across villages, where several communities agreed to join funds to connect to more distant mains water sources. As noted in Section 4, some concerns were noted in the IAS 2010 for the long-term maintenance of larger works, especially with regards to funds for potential major repairs, which was also a result of stretched resources for technical facilitation. A-2-iii. Livelihood Support Fund. The objective of this Fund was to improve the livelihoods of community members, especially poor women and unemployed youths by linking them to employment and income generating opportunities. The size of the LS Fund could vary from village to village, and VOs would have the flexibility to earmark up to half of the eligible VDP funds for Livelihood Support activities. The LS Fund could finance a savings and loan fund managed by the VSCOs, grants to the poorest and skill development training, as follow. 8 The IAS 2010 notes some inconsistencies in how this has been calculated, but overall community contributions, including planning time and other efforts, suggest that the achievement of the 30% contribution is highly likely. 30  Savings and Loan Fund: The objective of the component was to build a sustainable village- based savings and credit system that will expand opportunities for income generation for the community. This sub-component was to assist setting up of a Village Savings and Credit Organization (VSCO) consisting of Small Groups (SGs) and Cluster Committees (CCs). One of the main successes of the Project has been the formation of well functioning and effective VSCOs through a bottom-up process, following well developed and tested microfinance methodologies. VSCOs were established in each of the 1,034 Project villages. Some 22,640 SGs formed into 4,278 clusters, based on a membership of 150,119 villagers, 73% of who were women. Members have built up Rs 106 million of compulsory savings and Rs 166 million in voluntary savings. The resources from compulsory savings were used for small-scale internal lending (Rs 185 million had been re-lent), often for household consumption, but also some income generation activities, contributing to building SG confidence and management experience. All VSCOs had accessed the Savings and Loan Fund for livelihoods loans. Over 115,000 people had taken loans; 71% of the recipients were women and 68% were poor or very poor (VO survey under IAS 2010). A small proportion of people had taken two or more loans. The system provided a completely new credit mechanism for a very large proportion of village households, with 80% of households in the Gemi Diriya area having become members of small groups (HH survey in IAS 2010) and 62% receiving VSCO loans. With time the average size of loans has increased from less than Rs 5,000 to over 20,000 per loan. Aside from the flexibility and tailoring the loans to needs, the VSCO loan system was particularly attractive to SG members, as average transaction costs were 50% less than those of other sources (most of which were in any case not accessible to Gemi Diriya households. The VSCO organizational model involved officers and members to handle the portfolio, accounting and loan management, in a manner that would allow them to cover operational costs. GDF graded the VSCOs management in a participatory manner, and 62% were deemed satisfactory (Grade A & B). The VSCOs were also able to cover staff, office and allocations for loan security and loan loss provisions (see below). The operational self- sufficiency (OSS) ratio was on average 271%, well over the Sri Lanka microfinance averages, and the 100% breakeven. While this indicates considerable progress towards financial self- sufficiency, the portfolio at risk (PAR) of the VSCOs, an indicator reflecting risk of loan delinquency, indicated some cause for concern. From the VO survey (IAS 2010), 53% of VSCOs have less than 10% PAR at 30 days, and 69% at 60 days, which is below industry standards. The IAS 2010 noted that some VOs put less attention on loan collection and portfolio management; however it also notes that the VSCOs provide a large portion of the loans to agriculture, which is avoided by most MFIs due to the weather-related and other risks involved. Also the VSCOs avoided re-negotiating loan terms or writing off outstanding agricultural loans during droughts, typical of other financial institutions, and therefore had a relatively larger number of delayed loan repayments. Given the large share of agricultural loans in the portfolio, a longer term (PAR 90 or 180 days) would provide a better perspective on risks, but the data was not available. In addition, there are other reasons that can affect loan repayments e.g. wrong leaders, political influence, and wrong messages. One lesson is that VSCOs need strategies and capacity building and strengthening of the small groups as a basic unit to manage higher risk loan portfolios, and this is an issue which should receive more attention during the Phase 2 project. 31 One important risk management strategy which the Gemi Diriya savings and credit system has employed is the following: VSCOs pay 2-3%, and members 1-2% of value of each loan made into a Loan Security Fund, which presently have accumulated RS 90 million. The Fund covers the VSCO from loans that cannot be repaid due to the death of the borrower. For the families of VSCO members who die, this means they are not burdened with debt and can also cover funeral and related expenses. VSCOs also pay a small percent per loan into a special account to provision against defaults. This has not been applied delayed agricultural loans, as the practice is not to write them off but to work with members to try to collect.  Grants to the Poorest: VOs could provide a one-time grant to the poorest and most vulnerable in the community (e.g., the disabled, widows without any income, single women with small children, other destitute people, etc.) to start very small-scale income generating activities. The rationale for this activity was that the extreme poor would be unable to benefit initially from VSCOs, as these required savings from day one. The one-time grant was intended to serve as a bridge for such people to be able to participate in VSCOs and eventually other microfinance opportunities. The Project was highly effective in strengthening VOs to identify over 11,900 of the poorest of the poor and vulnerable though the participatory wealth ranking, and for the VOs to allocate a total of Rs 86 million in grants (of about Rs 8,000 per household) to 10,648 households (89% of all identified). Most of those identified were widows and disabled members of the VO. The IAS 2010 survey show that 86% of the grantees began to use the grants for livelihood activities, half for agriculture but also micro-enterprises and trading; 88% of the grantees had joined small groups; and over 4,458 (42%) of the grantees had already graduated to using VSCO loans.  Skills Development Training: VOs could provide loans to eligible community members, especially to youths, to acquire skills, such as, computer literacy, financial accounting, business development, etc. The rationale was similar to that of the one-time grants discussed above. Unemployed young people could not save initially (hence could not obtain VSCO loans), and the soft loans for skills development addressed this issue. The training was expected to lead to employment or to the starting up of small businesses. Over 12,000 youths between 18 and 35 years of age were identified by the VOs as eligible for skills loans. VOs provided Rs 145.77 million in loans to 11,792 youths (around Rs 12,360 each), of which 92% received employment related training by specialized training institutes or private sector employers. Some 69% of men and 43% women found employment in the field in which they had received training, men mostly as wage earners and women mostly self-employed (largely due to women being more reluctant to relocate away from their communities). Average earnings of those in wage employment were Rs. 15,000 per month. As a result, 33% of males and 30% of female youths had already repaid their loans, and some have gone on to become VSCO members and benefit from VSCO loans. A-3.Incentive Fund: Finally, about 10% of the total budget envelope for the village development sub-component was allocated to an Incentive Fund, as a special reward for those VOs that achieved agreed performance milestones after an external evaluation. The main objective was to encourage VOs to adopt good governance measures. A number of stringent criteria were set up to access the Incentive Fund, including a report card for good governance. VOs were eligible only after 3 years, which meant that batch 3 and 4 villages were not yet eligible in 2010. By March 31, 2010, resources from the Incentive Fund had been disbursed to 26 32 VOs; however, many more VOs of Phase 1 are eligible and will be receiving disbursements by December 31, 2010. This has been a learning process as this was the first time such a fund had been attempted (there was no similar provision in the VISHLI pilot). Overall, the Incentive Fund has helped in maintaining VO quality and enthusiasm; based on lessons learned, the Fund guidelines have been revised and are on an annual basis. Overall, the results of Component A of have exceeded all key indicators, as summarized in Table A.2.1 below. By Project completion, VOs had been set up in all villages, and all were registered as companies, with a range of functioning committees and a high degree of inclusion of women, the poor and youths, both as members and in leadership positions. All VOs had prepared VDPs, approved by the Maha Sabha; all had initiated subprojects and were managing VDFs; the majority had completed at least one infrastructure sub-project, many of them two, or even three; and all had created village savings and credit organizations (VSCOs) which had their own management structures and were covering their operational costs. Several important design and implementation features in the above component systems contributed to the strong achievements in reaching the project PDO outcome indicators:  Loans were low interest compared to moneylenders, and easily accessible compared to banks (although effectively charging a 12% real interest), supported by a village savings and credit organizations system which was building up capital, and increasing numbers of borrowers all the time, with rules set by the community;  Since repayment schedules where tailored by the VSCOs to agricultural production needs, people could sell their harvest when prices were higher, rather than to moneylenders immediately upon harvest, when prices were low;  Individuals could apply credit to the livelihood type and level that suited those best; and borrowers were also increasing the loan size, so graduating and diversifying. Average loan size per borrower increased from Rs 1,790 (2004) to 21,770 (2010), 26% of the borrowers have gone for a second cycle of loans to expand/diversify livelihood activities. The livelihood portfolio of the project beneficiaries is diversified with agriculture (foodgrains-21%, horticulture and commercial crops-55%), self-employment (20%) and animal husbandry (4%).  In addition, the combination of livelihoods and complementary infrastructure provided a strong synergistic effect: with improved access, road transport costs were reduced and there were less post-harvest losses.  With drinking water systems in place, there was time savings reported, so households could spend more time on farm and in wage work. Drinking water supply accounted for about 9% of completed infrastructure sub-projects. Benefited households reported time savings equivalent to 34 person days per year per family due to improved access to quality water provided by the project. About 79% of the VSCO beneficiaries are women and the saved time is utilized in livelihood income generating opportunities. 33 Table A.2.1 : Results Achieved for Component A Indicators Component A Indicators PAD At end of Achievement Target Phase I compared to PAD target Organizational:  65% of VOs are formally registered, 650 1,034 159% receiving funds and effectively implementing (65% of (103%) village priorities 1,000)  60% of sub-projects (ratified by the Maha 600 Sabha) are completed, operated and (60% of 1,414 236% maintained by VOs 1,000) (141%) Inclusion:  50% of women and youths are in decision- 3,500 4,620 132% making positions (e.g., Chairperson or 50% of (66%) Treasurer of various sub-committees), at 7,000 village level 172%  60% VDPs have at least one sub-project 600 activity completed that benefits at least 80% (60% of 1,034 of the poorest community members (as 1,000) (103%) identified by Maha Sabha) Component B: Institutional Strengthening The objective of this component was to encourage the establishment of “In other programs we highly sensitized, pro poor and responsive Project teams, Government give Rs. 100,000 and of agencies, Pradesiya Sabhas (PS), Support Organizations (SOs) and the that Rs. 10,000 reaches In Gemi private sector, with capacities to promote CDD. At the beginning of the the people. give Rs. Diriya, we program, there was little or no understanding of CDD in Sri Lanka – now 100,000 and they get many people at the division, district and provincial level knows about the benefits worth Rs. Gemi Diriya model and the CDD approach. During its meetings with 100,000.� officials at these various levels, the ICR mission found attitudes to be very Ex-Minister Pyitra Wanirachi positive towards the model and understanding of how it works was quite good. Politicians at the PS, provincial and national level were also quite familiar with, and supportive of, the Gemi Diriya program, including opposition members. The set of people who benefited from the program’s capacity building activities included many NGOs, TSPs and project staff. The market value of staff and NGOs who worked with the project is very high – in most cases their salaries have increased from 50 to 100%. Finally, the program is having an influence on many other departments, line agencies and senior officials at the national level, including staff of Gamaneguma, Samurdhi, Water Board, Health Ministry, etc. 34 Sub-component B-1: Capacity Building of National, District and Divisional Project Teams. This sub-component aimed to develop a shared vision and build required competencies among project teams at national, district and divisional levels to enable them to perform their respective roles. This was to be achieved by exposure to best practices and learning by doing experiences, and training on a range of community development methods. The line departments would receive training on participatory strategies and on improving the quality of services to the rural poor. The PSs were to be facilitated to participate actively in the program and their capacities were to be built to replace gradually the project teams and take over management of the program in Phase 2. A 5-6 week induction training, which included in-house training and field immersion, was provided for all staff to understand Gemi Diriya concepts, principles and practices. Throughout the Project period GDF staff at all levels and key partners were engaged in workshops, to reinforce knowledge, attitudes and skills of all project stakeholders gained through training sessions. A series of training programs and workshops have been conducted to develop stakeholder’s skills on CDD approaches, though at MTR it was noted that capacity building was still relatively ad-hoc. Various activities by GDF and outside training agencies and consultants were carried out to build capacity of the management team, District and Hub Facilitators, SOs, government institutions, NGOs, and private service providers. Project teams received intensive training on social mobilization, community participation, social capital, community- based organizations, micro-finance institutions, village development planning, environment management, and ML&E. The Project put considerable emphasis on learning by doing and reflection through planning workshops. A series of workshops were conducted for senior government officers and private sector entities to promote business linkages with Gemi Diriya activities at district level. Over 100 participants were involved in exposure visits to related livelihoods and rural infrastructure projects in India, to learn about practical and management issues related to facilitation and accountability mechanisms, value chain approaches for village communities. Exposure to the very large scale activities of Society for Elimination of Rural Poverty (SERP) in Andhra Pradesh provided the stakeholders opportunity to interact with local Panchayat members. There were opportunities to see how SERP managed institution building and social capital development, credit and finance, livelihood promotion through partnerships, social action issues, including judicial services for the poor. The visiting participants in this program included national and local political leaders, Chairpersons of VOs, Divisional Secretaries of the respective areas, District and Hub level Facilitators, and VO members and management staff. Other exposure visits took place to Philippines, Brazil, Bangladesh and Indonesia. Sub-component B-2. Capacity Building of Support Organizations (SOs) and Service Providers: The objective of this sub-component was to facilitate the active participation of SOs and the private sector. Over 55 SOs were contracted by GDF to assist in the basic mobilization, community planning facilitation and basic organizational development process. The project provided basic orientation on the assumption the SOs would recruit competent village facilitators. Possibly due to the scale of recruitment needed the quality of VFs and PRA officers was not always as required, and there was also considerable turnover of staff. The Project assisted with further training. According to the evaluation done on ‘People’s perception on overall SO performance’ by an independent evaluation of SOs (Evaluation of Performance of Supporting Organizations of Gemi Diriya, Paltra Guarantee Limited, 2009), about 70% of VO members who participated in the evaluation expressed satisfaction with the work carried of the SOs, but noted that there were weaknesses in the depth of facilitation and support. On balance, SOs 35 learned quite a lot about CDD strategies and techniques through their participation in the project, and their performance did improve over the Phase 1 implementation period. At the same time, for reasons of sustainability and costs, the Project also increasingly involved community professionals in some roles initially played by SOs, and this has proven quite successful. Sub-component B-3. Monitoring and Learning: The objective of this sub-component was to monitor performance and progress of the project and to disseminate it among all stakeholders to enable them to effectively address project implementation issues. Most expected ML&E activities were completed and outputs delivered, as discussed in the main text. The system has evolved considerably during Phase 1 as experience was gained in participatory monitoring. Some of the more important developments included the following: (i) a village monitoring matrix was designed and institutionalized at the village level, and each village engaged in its own results; (ii) the Hub would invite all VOs in its jurisdiction on a monthly basis to review progress and results and resolve issues; (iii) at the division level, the Divisional Secretary would chair monthly meetings – this participatory forum would take place in one of the villages in the division and other village chairpersons would be invited; (iv) at district and GDF levels, there are monthly and quarterly fora to review progress, share information and resolve issues; (v) Social Audit Committees engage in monitoring at the village level; (vi) as a result of the Project ICT initiatives, a lot of monitoring and learning now happens through Information and Monitoring centers; and (vii) a number of thematic reviews and assessments were carried out by GDF management, some in cooperation with IDA missions. All indicators for achievement of Component B results have been exceeded, as summarized below: Table A.2.2: Results Achieved for Component B. Key Performance Indicators Component 2 Indicators PAD At end of Achievement Target Phase I against PAD target 70% of VOs receive support for VDP formation and 700 (70% of 1,034 143% implementation in accordance with agreed service 1,000) standards 40% of appraisal sub-committees at PS level are The Other formal functioning with satisfactory service delivery 40% out of original committees standards 54 PS trigger of 4 were not set PSs for up but first two operated years was informally. met. 25% of VOs to earn income from providing para- 250 (25% of 411 164% professional services to project and non-project 1,000) communities 30% of VOs are working in partnership with private 300 (30% of 450 150% sector organizations 1000) 36 25% VOs are federated, merged and functioning as 250 (25% of 498 199% business enterprises 1000) Component C: Innovation Seed Fund The objective of this component was to pilot innovative ideas that would have potential for scaling-up and replication. Proposals for would be selected through a competitive screening process and, as this was a testing/piloting program, no ex ante targets were set. The Project did not find the method of advertising for innovation to be very useful, because too many proposals were received and most did not qualify as innovation. The Project therefore chose another strategy, of asking VOs, GDG staff and others to showcase their ideas and suggestions, out of which came a number of innovations –e.g., community radio, environmental proposals, CP centers, youth festivals and cultural groups where there was competition among various districts and 100 youths were selected and trained. Innovations for technologies such as maize moisture meters and specialized tea carriers were also supported. Most significant was the establishment of the ITSHED Commercial Web Portal. Specifically resourced under this component, ITSHED was set up and operated with the support of GDF, the Federations and a technical service provider. Village youth groups in 175 VOs were mobilized to provide commercial resource information to the web portal displayed under each linked village. This helped Project villages to access market information, display their products to attract business and receive valuable information to make investment decisions and improve livelihoods (www.itshed.net). Further information technology systems with internet connectivity were set up in over 450 villages, with material support through a range of national and other agency programs. Relevant IT skills were developed in parallel, mostly with poor youths. Another important innovation was the establishment of Community Professional Learning and Training Centers, where the early development was under this component, but implemented under component A. This has now been scaled up in all Project areas, but even regionally with models in India and Bangladesh. Results in achieving Component C performance indicators are summarized below: Table A.2.3: Results Achieved for Component C Key Performance Indicators Component C %PAD At end of Achievement Remarks Indicators Target Phase I 20% proposals identified 200 146 proposals received were for Seed Funds (20% of 0 - rejected. However, there were 1,000) a number of initiatives/proposals received within VOs and from sources which were found innovative; and those were approved, tested and scaled up. 20% households adopting 200 175 88% 175 VOs have joined ITSHED innovative practices (20% of web portal, and 500 are (assumed to be solely 1,000) expected to join within one 37 from competitive system) year of the closing date. However, a range of other innovations emerging were also funded e.g. community radio and other innovations are also being scaled up. Component D: Project Management The main objective of this component was to facilitate overall co-ordination and management of the project at national and divisional levels. The considerable achievements of the Project can be largely attributed to considerable implementation capacity, supervision, coordination, staff commitment of the public but autonomous Gemi Diriya Foundation (GDF). A key feature of the management was the rigorous and participatory process of updating 14 community operations manuals, with five revisions through the project period. Some 87% of district and divisional facilitation teams received positive scores from communities through Community Score Cards system, itself an innovation. GDF operates at the national level, with 6 District and 35 Hub level offices, with over 400 staff by 2010. Staff were mostly men at higher levels, though there was a larger percentage at hub level (largely related to less willingness by women to spend long time in the field). Later in the Project period there was an addition of 2 regional level offices, decentralizing some of the national level technical staff and functions. Appraisal teams reviewed all projects, and on the whole the processing of community applications for Project funds was adequate. Financial, procurement and safeguard and management support were considered satisfactory throughout the project, although staff capacity was stretched as the project scaled up. The GD Board consisted of an impressive and creative group of people representing civil society, the private sector and the bureaucracy. Both ex-Chairs encouraged GDF and gave it space for various technical and managerial innovations. Placing qualified and well-trained people at the field level, supporting them and making them accountable to the communities were also a critical success feature of Gemi Diriya’s project management arrangements. In recognition of these and other human resource management features of Gemi Diriya, the program won a national award for excellence in performance management. Table A.2.4: Results Achieved for Component D Key Performance Indicators Component D Indicators PAD Target At end of Achievement Phase I 70% of VOs have access to funds in less than 2 weeks 700 610 87% from appraisal/certification (70% of of milestones 1,000) 38 60% of district and 60 % of 6 All District divisional facilitation teams District teams teams have Target exceeded receive positive scores from (4)- and 60% pf satisfactory communities through 35 Hub teams scores, and Community Score Cards (21) 31 (88%) Hub teams 4 six-monthly COM 5 annual ones Target achieved revisions based on feedback 4 were done from the field and from (125%) Independent Audits Component E: Village Self-Help Learning Initiative The VSHLI pilot was initiated in year 2000 under the IDA-supported Mahaweli Restructuring and Rehabilitation Project in three villages (see main text, Section 2.1), expanding to 12 villages under a US$ 750,000 JSDF grant which was active until 2005. VSHLI provided the model for the Gemi Diriya Project, and VSHLI became a component of CDLIP from March 2005 onwards, to fund completion of the remaining VDPs in the pilot villages and further strengthen institutional arrangements at the local level, by developing the community professionals and village federation model. Beginning in 2005, the 12 villages were increased to 27 under Gemi Diriya and then after MTR to 47. Authority over management of the VSHLI was assumed by GDF in 2007. All the targets for Component E have been considerably exceeded, especially taking into account that it covered an additional 35 villages. 47 VSHLI batch villages have fulfilled their VDPs with sub- projects completed and being operated and maintained (compared with an original target of 75% of 12 villages). About 83% of the households in VSHLI sub-projects indicated that their household income had increased due to activities of the Project and the VOs. The Project put particular effort on strengthening livelihoods support activities in VSHLI villages under Gemi Diriya. All VOs are providing CP services to Gemi Diriya and other villages. About 50% of VOs are federated and 40% of VOs are functioning as business enterprises. Finally, the innovative establishment of the CPLTCs under Gemi Diriya was a result of VSHLI. Table A.2.5: Results Achieved for Component E Key Performance Indicators Component E Indicators PAD At end of Achievement (VSHLI areas only) Target Phase 1 75% of VDPs are completed, 9 (75% 47 520% operated and maintained by VOs of 12) 25% increase in household 25% No Based on interviews income from project activities specific increases similar to across data on the project, but for much income larger number of VSHLI increases beneficiaries. under 39 VSHLI 50% VOs Target considerably 25% VSHLI VOs federated, 3 federated exceeded. Also all 47 VOs merged and functioning as (25% of 40% VOs providing CP services business enterprises 12) (over 18 (covering their costs) VOs) functionin g as business enterprises Component F. Gama Neguma – Interconnectivity Program This component was added in 2008 to test an institutional and financial model at the Pradeshiya Sabha level for employing Gemi Diriya principles to identify, mobilize communities, and implement Gama Neguma’s larger scale inter-village infrastructure and connectivity investments at the zonal/regional level. In essence this implied advancing into Phase 1 the mainstreaming of Gemi Diriya into other programs and activities (originally planned for Phases 2 and 3), at the Government’s request. Designing and implementing this component presented several challenges. The aim was to maintain the Gemi Diriya principles, particularly participation and accountability, at inter-village scale and engage the Pradesiya Sabhas much more intensively. It took some time initially to overcome community skepticism that they would be given responsibility over such larger amounts of money (compared with Gemi Diriya’s smaller intra-sub-projects) on the one hand, and from local officials on the other that communities would be able to handle this responsibility. A considerable number of training events were conducted for PSs and workshop meetings held with communities to develop a specific operational Manual and implementation plan. By mid-2010, feasibility studies and detailed plans for 8 sub-projects had been completed in 4 PSs in 4 Districts involving 60 villages organized into zonal committees - about half project and half non-project. The investment proposals identified and planned by communities, together with technical service providers they chose, were all examined by a PS sub-committee and passed by the PS general body. All projects involved key connecting roads or road-related investments (one sub-project was for a causeway). Estimated total cost was Rs 302 million (with a further 30 million to be contributed by the community in cash and labor). The total length of roads planned was about 65km, which would cover villages with a population of 34,700 (1st Quarter Progress Report 2010). As of mid-2010, road construction had started in 3 of of the PSs, with 20-30% physical progress completed. The communities were playing a major role in supervision of construction (ICR mission field interviews and meeting with PS staff and politicians, community representatives and Project team in Passara). A number of lessons have been learned from this inter-connectivity pilot, and local officials and politicians -including from opposition parties - were highly positive about the process, and on the nature of possible impacts. PS officials note that they had developed planning skills that they had not previously possessed, and newfound sense of fulfillment in performing their technical roles. Overall the sub-projects also provided an important additional boost in the resources available to the local governments. Villagers also developed a confidence they had previously lacked in terms of approaching 40 and directly engaging local officials. Based on interviews and feedback during the Passara meeting, stakeholders broadly agreed that the expected benefits were highly likely to materialize, and they estimated that about half of the local population in the villages would benefit directly from reduced travel time, lower costs and losses, especially for tea, fruit and vegetables. Transport costs were expected to be reduced by up to two-thirds. There were indications that land prices were increasing, with some of the roads possibly becoming important inter-provincial through routes. Numerous other benefits were specified, including the fact that villagers’ children would have a much easier and safer time getting to and from school. One issue, which arose during implementation, was the lower level of community contributions under Gama Neguma, compared with that under Gemi Diriya sub-projects. Villagers living closer to the road building were more willing to contribute, but others living many kilometers away perceived fewer benefits and incurred greater time and costs in getting to the construction areas. Thus labor contributions were about 50% of those expected in some cases. This was also partly due to the fact that the contribution requirement was not clearly conveyed upfront, but rather was communicated to the PS and communities at a later stage. A second point to be addressed as Gemi Diriya moves into Phase 2 is the fact that while small-scale, regular upkeep is planned by communities, there is no direct reassurance of how major damage or repairs would be funded in the absence of an allocation in the PS budget. 41 Annex 3. Economic and Financial Analysis The Project Development Objective of Gemi Diriya (CDLIP) is to improve the livelihood and quality of life through sustainable development of the village communities in the project area. The Project originally had five components namely; village development, institutional strengthening, innovation seed fund, project management and village self-help learning initiative. A sixth component to support the Gama Neguma interconnectivity program was introduced in 2008. The project has supported 1,034 VOs through which the project interventions are planned and implemented during 2004-10. No ex ante estimation of rate of return to the project investments was done in PAD. The stated reasons were; (i) the demand driven nature of community-led investment design of the project and (ii) the fact that empowerment and strengthening of communities and institutions targeted by the project was not quantifiable for cost-benefit analysis purposes. Based on similar CDD projects in the South Asia region, the PAD expected the project investments to be cost-efficient. A. ICRR Estimation methodology 1. The project has supported 139,223 livelihood investments and initiated 3,528 community infrastructure sub-projects in 1,034 VOs. VSCO loans, totaling Rs 2434 M, are availed by 110,758 borrowers to start their livelihood activities (Tables A-3.A1). Additionally, VSCO Loan beneficiaries have contributed one third of the livelihood project cost through equity. 79% of the borrowers are female. 80% of the borrowers belong to poor and poorest category. 1414 community infrastructure investments are completed benefiting 121,313 HHs.9 2. Benefits are quantified for four major groups of livelihood investment activities (food grains, horticulture and commercial crops, animal husbandry and self-employment) using major dominating activities in each group and then aggregated for livelihood support interventions. About 40% of one- time grantees (4,458) have also availed livelihood support loans from the VSCOs. Employment generated through skill development loans is assessed. Cost efficiency analysis was done for the roads/bridges/culverts/causeways accounting for 75% of the community infrastructure investments. Impacts of convergence in livelihood and/or infrastructure investments are quantified based on case studies representing major project interventions. 3. For cost benefit analysis, data from multiple sources were compiled and used. They include: the Project MIS database up to first quarter of 2010, Impact Assessment Study by SLBDC and mission field visits. For cost efficiency assessment, data compiled by the Infrastructure unit of GDF and cross- verified in the field by the Impact Assessment Team for sample comparable roads, drawn from three project districts, representing Gemi Diriya, PRDA, and Samurdhi programs are used. These data are supplemented by secondary data sources, the ICR Mission’s field visits, IDA supervision reports and interactions with the project implementing agencies for quantifying the physical benefits and assessing the project costs. All project costs are included in the analysis at constant 2010 prices. Total project cost 9 The beneficiary data provided by the project team for the community infrastructure investments suffer from some inconsistencies; overlapping of beneficiaries between different infrastructure categories as well as lack of clarity in defining direct and indirect beneficiary HHs and therefore not dependable. Thus the estimates and methods provided in Tables A-3.A1. 42 at constant 2010 prices is Rs. 7,429 million. Detailed project cost-break up by components is provided in A-2. Economic analysis was done by using parity prices for internationally tradable inputs and outputs and a SCF of 0.9 was used for internationally non-tradable inputs and outputs and for converting the project costs to economic costs. Project Benefits 1. Livelihood Sub-Project Benefits Cereal crops: About 21% of the livelihood investment loans are availed by the project beneficiaries to cultivate the cereal crops like paddy and maize. Average sub-project cost for cultivating cereal crops is Rs 19067, including 22% coming through beneficiary contribution. Average cropped area under cereals is 0.6 ha per loan. About 85% of the food grain crop loans were taken for cultivating grain paddy. The project- led initiative resulted in 3% of the VSCO loans under cereals going for seed paddy production. The mission observed during the field visits that seed paddy produced by the farmers with VSCO loans is getting utilized by other VSCO grain paddy loan beneficiaries, besides meeting the needs of other farmers from outside project villages. Many VSCOs have started distributing part of the VSCO loan for paddy through kind, in the form of good quality seed. Because of this, about 26% of the VSCO loans are used for producing grain paddy by using the good quality seeds. This convergence of producing quality seed and supporting the grain paddy producers both within and outside project villages through VSCO loans has generated positive externalities (Box-1). Paddy productivity impacts: VSCO loan T�1 Livelihood investment impacts on paddy growers   beneficiaries recorded 12% increase in paddy WP   WP  yields (weighted across maha and yala seasons) (farmer  (certified  due to timely application of inputs like farm  Impacts  WOP  seed)  seed)  machinery, plant protection chemicals and farm Grain yield (kg/ha)  5270  5918  6461  labor (T-1). But, in case of paddy farmers who Gross margin (Rs/ha)  47744  66800  80945  Cost of production (Rs/kg)  21.1  18.9  17.7  have used either own seeds or purchased locally, poor seed quality has restricted the yield advantage. Poor germination and susceptibility to pests and diseases are the problems highlighted by the farmers, which has increased either the cultivation costs and/or depressed the potential yield. When certified quality seeds are used by the beneficiaries, the paddy yields have gone up by 23% across seasons. 43 Box.1 Project led seed paddy production initiatives and positive externalities (A�4)    Project  initiative:  The  project  facilitated  the  production  of  seed  paddy  in  project  V.Os  by  utilizing  VSCO  revolving funds. During 2008�10, seed paddy production was taken up during both maha and yala seasons.  About 566 farmers1 from 47 V.Os participated from four project districts by allocating 309 ha for seed paddy  production. Through this initiative, 1370 MT of certified seed paddy has been produced and sold both within  and  outside  the  project  villages.  Total  grain  paddy  area  supported  by  the  certified  seed  paddy  from  gemidiriya  V.Os  is  13704  ha.  Within  the  project  villages,  certified  seed  paddy  is  provided  by  VSCO  to  the  farmers, by tying it up with VSCO loans for grain paddy cultivation to ensure better seed replacement rate  and higher paddy yields for the grain paddy cultivators.     Average  weighted  WOP  irrigated  paddy  yield  in  the  project  villages  was  5.2  MT/ha  across  maha  and  yala  seasons2. Poor seed replacement ratio, high fertilizer price and high interest rate were the major production  constraints faced by the paddy farmers. GoSL is heavily subsidizing the fertilizer supply to the paddy farmers  since 2005. During 2008/09, cost to GoSL due to subsidized fertilizer supply to the paddy farmers is Rs 65  billion3. A total of 505 kg of fertilizer per ha per season is provided to the paddy farmers at a subsidized rate  of Rs 7 per kg4. Expected paddy yield with this recommended fertilizer dose is above 7 MT per ha. To realize  this,  it  is  critical  to  have  certified  seed  paddy5  and  easy  access  to  credit  at  competitive  interest  rate  for  ensuring timely cultural practices. Gemidiriya V.Os provided both the quality seed as well as easy access to  credit  through  VSCO  loans  for  paddy  cultivation.  The  project  V.Os  produced  and  sold  1370  MT  of  certified  seed paddy, which is adequate to cover 13704 ha of grain paddy area in and outside the project villages. In  addition  to  this,  about  25770  VSCO  loans  (Rs  400  M)  are  given  for  paddy  cultivation  during  the  project  period.     While  the  government’s  fertilizer  subsidy  policy  for  paddy  cultivation  helped  in  substantially  reducing  the  financial  prices  for  the  fertilizers,  VSCO  provided  timely  credit6,  which  resulted  in  better  production  management including the balanced application of fertilizers and plant protection chemicals as reported by  the  paddy  farmers.  VSCO  loan  for  paddy  is  utilized  by  the  project  beneficiaries  for  hiring  farm  machinery  (28%),  buying  quality  seed  (8%),  applying  agrochemicals  (12%)  and  hiring  farm  labor  (52%).  Multiple  impacts  reported  by  the  farmers  are  intensification  in  paddy  cultivation,  increased  yields  and  enhanced  efficiency in the use of inputs as reflected in the reduction in the unit cost of production.  _____________________________________________________  1 The paddy federation under Gemidiriya project has linked selected farmers with Department of agriculture and Agricultural research  institutes  to  provide  necessary  training  and  skill  development  in  seed  paddy  cultivation  due  to  which  these  farmers  have  become  registered paddy seed growers to produce certified seed paddy.  2 Compiled during the mission’s field visits and focus group discussions and cross checked with the published secondary data sources  (Ref: Cost of cultivation of agricultural crops 2004/06 Maha and 2005/06 Yala, Socio�Economics and Planning Centre, Department of  Agriculture, Peradeniya)  3  In 2008/09, the fertilizer subsidy accounted for 3% of total government expenditure and 0.6% of GDP. For paddy cultivation, 0.6  million MT fertilizer was used at the subsidized rate.   4  Subsidized fertilizer is supplied through 557 district centres distributed island wide.   5 Supply of quality seed paddy is not adequate to meet the potential demand to maintain the desired seed replacement rate of once in  three  years.  For  instance,  in  Trincomalee  district,  it  is  estimated  that  available  certified  seed  is  sufficient  to  meet  only  41%  of  the  potential demand.  6 VSCO loan is available at a monthly interest rate of 2%, as compared to the minimum interest rate of 5% per month under WOP  situation. The beneficiaries reported less than two weeks as the time taken by VSCO for the processing and delivery of loan amount.  44 Income impacts: Financial gross margin from paddy has increased by Rs 19,056 per ha across seasons among the loan beneficiaries (T-1). With certified seeds, gross margin from paddy has further increased by Rs 14,145 per ha. Certified seed paddy farmers are trained in the cultivation of paddy for seed production, which is cultivated in both maha and yala seasons10. On an average one ha produces 5.4 to 6.5 MT of seed paddy as reported by the beneficiaries. Weighted certified seed yield is 5.9 MT and gross margin is Rs 128,700 per ha from certified seed paddy cultivation. As compared to WOP situation, certified seed paddy farmers have realized incremental gross margin of Rs 80956 per ha. Efficiency impacts: Using these certified seeds helped in realizing increased gross margin by 70% mainly driven by reduced unit cost of production (T-2). Unit cost of grain paddy production has come down by 10% to 16% with VSCO loans due to the usage of certified paddy seeds and timely use of critical inputs facilitated by easy access to community managed cheaper credit. Equity impacts: In all the seven project districts, over 4/5th of paddy farmers are smallholders, cultivating on an average 0.5 ha of paddy, which is heavily dependent on rainfall. The income earned from paddy is inadequate to meet basic needs of a family and poverty is prevalent among small scale paddy farmers11 . Among the 25,770 VSCO loans availed by the project beneficiaries; 83% of the borrowers are female and 89% of the paddy loans are availed by poor and poorest households. Thus, incremental gross margin generated by the VSCO loans for grain paddy cultivation has benefited the women and poor and poorest families most. Maize impacts: About 12% of the VSCO loans for T�3 Livelihood investment impacts on maize growers   cereals cultivation were used for cultivating maize.  Impacts  WOP  WP    Average loan provided by VSCO is used to cultivate 0.6 Grain yield (kg/ha)  3300  5000  ha of maize cultivation. The participating farmers are Gross margin (Rs/ha)  25246  46046  trained in the hybrid cultivation, harvesting and storing Cost of production (Rs/kg)  21.1  18.9  practices. Moisture meters are provided to the VOs while training the maize producing farmers in partnership mode in hybrid maize cultivation and quality improvement for marketing. Maize is predominantly cultivated under rain fed condition by the resource poor farmers. Average productivity (WOP) is 3.3 MT per ha and beneficiary farmers with VSCO loan are cultivating hybrid seed and getting an average yield of 5 MT per ha (T-3). Farmers, in a partnership mode, are joining together to pool their stored produce and entering into agreement with seed company for getting a better price for their produce (Box.2). Technical assistance is provided to the farmers for establishing business linkages. But, currently the village organizations are federated instead of federating the maize producer groups. 10 GDF  identified  the  immediate  requirement  of  paddy  farmers  and  initiated  production  of  certified  seed  paddy  with  paddy  farmers in Hambantota, Monaragala, Badulla and Polonnaruwa districts. Technical support has been obtained through the Dept. of  Agriculture  and  District  Federation  has  been  involved  with  marketing  of  seed  paddy.  GDF  support  extended  to  supply  moisture  meters,  bagging  machines  and  packaging  to  VOs/Federations.  District  Federation  purchased  paddy  from  farmers  and  sold  seed  paddy under own brand name to Gemi Diriya and non Gemi Diriya villagers and to the Department of Agrarian Services.    11 Jeevika Weerahewa, Impacts of Trade Liberalization and Market Reforms on the Paddy/Rice Sector in Sri Lanka, MTID  Discussion  Paper No. 70, Markets, Trade, and Institutions Division, International Food Policy Research Institute, Washington, D.C. May 2004  45 Hybrid maize farmers with VSCO loan recorded incremental gross margin of Rs 21,000 per ha. While cultivation cost has gone up by 40%, gross margin has increased by 84% mainly due to 50% increase in the productivity of maize with hybrid seed and improved method of cultivation. Additionally, the farmers in a partnership mode are able to store and collectively bargain for a better price for their produce. If maize is stored for about four months after harvest, producer price increases by 20%. VSCO supported maize sub-projects have contributed to import substitution and foreign exchange savings. Increased resource use efficiency: Maize is a rain fed crop grown in high lands. Maize farmers have brought uncultivated areas under maize cultivation. And in many VOs, maize is newly introduced in yala season and thereby improving the cropping intensity. This has led to better utilization of both land and labor in the project villages. Productivity of land and labor has increased with the introduction of hybrid maize with increased use efficiency of these resources. High Value Crops: Horticulture and commercial crop cultivation accounted for respectively 27% and 28% of the VSCO loans availed by the beneficiaries. Average cultivated area under these high value crops (potato, chillies, beans, tomato, brinjal, bell pepper, capsicum, tea etc.) is 0.45 ha, supported by an average loan size of Rs 27105, including 38% coming through beneficiary contribution. Weighted incremental gross margin across major high value crops is Rs 99,840 per ha, an increase of 153% over the WOP situation. The mission observed efficient recycling of VSCO loans by the beneficiaries by cultivating vegetable crops in two seasons during the loan repayment period, which maximizes the returns to the livelihood loans. In many V.Os like Girambe VO for example, VSCO members have collectively decided to use the funds for vegetable cultivation. Agriculture loans accounted for 78% and the rest 22% for self-employment activities in this V.O. In agriculture, potato and beans alone accounted for 90% of the loans disbursed. And most of the farmers have taken VSCO loan once and used it effectively to first cultivate irrigated potato and recycle the profit for cultivating beans within the stipulated six months grace period and the entire loan gets repaid at the end of the beans harvest. The potato farmers are brought under partnership mode and linked with Colombo based company for bulk purchase at fair price to the producers. Such innovative and efficient use of the VSCO credit generates multiplier effect on income generated. The livelihood fund has almost doubled in this VO due to quick and productive turnover. The village has added six hand tractors, 4 power tillers and seven irrigation pumps since the flow of VSCO loans underlining the multiplier effects of making credit available to the needy farmers both timely and at an affordable rate. Animal Husbandry: About 4% of the VSCO loans are availed by the beneficiaries to generate livelihood income by rearing dairy animals. Average loan provided by VSCO is Rs 31,751. The beneficiary contribution was 38% of the total cost of the animal, which was also partly financed by the bank in some cases to supplement the VSCO loan. The dairy animals yield 7 liters/day for a lactation period of 250 days and the milk is sold through organized milk collectors linked to milk route at the rate of Rs 30/liter, which fluctuates between Rs 27 to 31/liter. Maintenance cost for the animal is Rs 30,000 per year and incremental gross margin from one animal is Rs 13,330 per lactation season. Fresh fodder supplement and concentrate feed account for 70% of the cost and the rest is shared between labor, insurance, medical and other expenses. With improved health care and green fodder support, milk yield can be increased by 25%. 46 Incremental gross margin per animal is Rs 13,330 per lactation period. Milk producers should be federated and encouraged to go in for creating storage facilities and complementary sub-projects for milk processing in the milk catchments in order to optimize the income for the dairy beneficiaries. This will also encourage evening milking as against one time milking practiced now. Sale of calves will generate more income in subsequent years and with increase in the size of the milk animal stock, income for the dairy beneficiaries will further improve. Self-Employment/ non-farm activities: About 20% of the VSCO loans are used by the beneficiaries for starting self employed livelihood income generating activities like retail shops, handicraft, cane basket making (bamboo), grocery shop, floor mats (linen), tailoring, sweets making, mobile recharge shops, stationery shops, compost making, carpentry, plastic items procurement and sale. There is wide variation in the performance of these projects and on an average for a livelihood investment of Rs 29,349, annual incremental gross margin of Rs 20,125 is generated per loan beneficiary. B. Rate of Return Analysis T�5  Project analysis summary    Average loan size per borrower increased ICR Estimation (2010 Prices)  PAD  from Rs 1,790 (2004) to 21,770 (2010), Project Analysis  PVC  PVB  NPV  IRR  IRR  26% of the borrowers have gone for the Financial analysis  4,453  10,571  6,118  31.7%  Not  second cycle of loans for Economic analysis  4,052  9,048  4,996  30.2%  Estimated  expanding/diversifying the livelihood activities, livelihood portfolio of the project T�4  Livelihood investment impacts     beneficiaries is diversified with agriculture Impacts  Unit  WOP  WP  Change  (foodgrains-21%, horticulture and Yield              commercial crops-55%), self employment Paddy  Kg/ha  5270  6033  14%  (20%) and animal husbandry (4%). High Maize  kg/ha  3300  5000  52%  value crops accounted for 55% of the VSCO Dairy  l/d/animal  3.5  5.5  57%  loans and generated Rs 22,460 as average Gross Margin              incremental benefit per beneficiary. Cereals  Rs/ha  45117  68814  53%  Average incremental benefits for livelihood High value crops@  Rs/ha  65163  165000  153%  investments ranged from Rs 14,840 (food Animal husbandry  Rs/ha  8983  22313  148%  grains) to Rs 26,660 (animal husbandry) per Self employment  Rs/year 60000  80125  34%  beneficiary. @Horticulture and commercial crops combined'  Overall project economic rate of return is Fig.1 Average income impacts of Livelihood Loans 30.5% and project financial rate of return is 30000 26661 32% (over full life of project, including all 25000 22463 Phase 1 Project costs-including mobilization 20125 20000 and management). This is based on 14838 conservative assumptions, like limiting 15000 livelihoods income generation at 2010 (end 10000 of Project) levels of intensity, and 5000 infrastructure benefits to only cost savings. Anima l High Va lue Se lf Food Gra ins The number and size of livelihood activities Hus ba ndry Crops Em ployme nt 47 were expected to expand and diversify with time and infrastructure led benefits would also be expected to accelerate. C. Skill development and youth employment: The project has identified 35,555 unemployed youth in the project villages, out of which 17,842 unemployed youth are selected for skill development. Up to end-project, 12,133 unemployed youth have availed skill development loans to the tune of Rs 152.4 M for upgrading their skills. 45% of the youth opted for driving and computer operating skills matching with the job needs in the market. So far, 5951 trained youth are employed; 59% self-employed and the rest in paid wage employment. Average monthly income varies from Rs 13,120 for self-employed categories to Rs 19,260 for the paid wage employment categories (T-6). Annual decrease in youth unemployment during 2005/09 in the Gemi Diriya villages is 4.2% as compared to 2.2% for the country as a whole during 2001/09. Employability development and job placement for unemployed rural youth Skill development through project interventions is cost efficient due to (i) creation of more jobs for rural unemployed youth in the project villages, (ii) average cost of job creation is only Rs 12,550 per job, and (iii) 58% of the skill development loans have already been repaid which will generate more and more jobs. T-6 Gemi Diriya Project: Youth Skill Development and Employment Generation Impacts Gemi Diriya V.Os#  Number  1,034 Status of Youth    Unemployed youth identified  Number  35,555 Youth identified for skill development  Number  17,842 Youth training for skill development  Number  12,133 Skill development loans disbursed  Rs (m)  152.36 Skill development loans recovered  Rs (m)  88.4 Youth employed  5,951 Paid wage employment  Number  2426 Self-employment  Number  3525 Average cost of Job creation (Revolving fund)  Rs/Job  12557 Average monthly Income (Rs)    Paid wage employment  Rs  19,260 Self-employment  Rs  13,120 Annual decrease in youth unemployment (2005-09)   %  4.2 Sri Lanka @    Annual decrease in youth unemployment (2001-09)   %  2.2 # GDF Youth and Gender Unit for Gemi Diriya project level data @ Ref: Estimated for comparable youth age groups from Economic and Social Statistics of Sri Lanka 2010, Statistics Department, Central Bank of Sri Lanka) 48 D. Convergence in resource allocation for maximizing benefits The investment portfolio management of Gemi Diriya village Wellawela VO is good example12 for designing complementary infrastructure and livelihood investments to optimize the agriculture benefits from the available resources within the village community (Fig-2). In this VO, priority for allocating infrastructure fund of Rs 1.9 M was given first to extend irrigation water to 40 acre of rainfed lands benefiting 41 rainfed farms (inclusiveness) and next to provide connectivity to main road for 80 acre of lands benefiting another 82 farms (equity). Carefully designed infrastructure investments improved the resource base of the village community with more irrigation coverage (equity) and better access to input and output markets (efficiency). Variable user fee, depending on irrigation water used, is collected to maintain the infrastructure created by the community (sustainability). Further, to optimize the benefits from such infrastructure investments, VSCO disbursed 2/3rd of the livelihood funds for high value crops like beans, cabbage, tomato, potato, brinjal and capsicum benefiting 120 farmers (Convergence). Fig. 2 Convergence Model of Wellawela V.O in Badullah District Infra Investment Convergence Equity & Efficiency Wellawela V.O, Badullah Dt Total Crop Land 252 Acres Paddy Low Land High Land 80 Acres 172 Acres Rainfed (RF) Land 40 Acres Irrigated Land HHs 41 ; Population 218; Women 26; Men 13; Youth 2 132 Acres With Project Poor 32; Most Poor 2 Tank Irrigated Land With Project 40 Acres TEA Plantations Road Connectivity to 20 Acres 80 Acres The village community is now reaping the benefits of agriculture expansion and intensification: increased irrigated area (by 40 acres), maximum irrigated cropping intensity (250%), mechanization of farm operations and crop diversification (five high value vegetable crops); reduced transportation cost (by 9%); reduced post harvest losses in transportation and handling (by 6%); and better price for the 12 Similar convergence impacts are also observed in four other V.Os namely Kudapolwattha, Welikumbura, Kempanawattha and Uda Alupodhaniya in Rathnapura and Hambanthota districts. In the absence of V.O level livelihood loan distribution by major sub-projects, the impacts of convergence in investments could not be scaled up for the whole project. 49 produce (7.5%). This V.O invested a total cost of Rs 12 M in constant 2010 prices for village development. Beneficiaries, directly benefited by the infrastructure and livelihood investments increased their income by 178% aided by the convergence of irrigation and roads investments with the allocation of livelihood funds for high value crops. On an average, each direct beneficiary has realized incremental income of Rs 110,500. In the village, 123 families are benefited by these specific interventions. This V.O has generated an ERR of 42% for the community investments. E. Efficiency T�7 Cost Efficiency of Funds Delivery in Gemi Diriya CDLIP The outcomes and impacts of the project are   realized with considerable efficiency. Some of Project Costs  Unit  Amount  the key efficiency elements of Gemi Diriya Project management costs   Rs M  469  Mobilization costs   Rs M  1122  project are: Community Investment funds  Rs M  3142  Cost of Delivering Community Funds (1 Rs)        � CDD led approach delivered community      With recycling of RF as on 2010 $  Rs  0.26  funds more efficiently (T-7). Cost of      With projected recycling of RF @  Rs  0.22  delivering one rupee of community funds is   Rs 0.22, which will further decline with $ RF recycling ratio is l.58 as observed by end�project (2010)   @ RF ratio is projected at 2.0 based on pilot and earlier batch experiences   continuous recycling of revolving funds by the community. � Community led recycling of livelihood funds resulted in efficient use of funds and coverage of project targeted HHs (T-8). Weighted recycling ratio is 1.6 in 2010, varied from 1.15 to 3.6 across batches, projected to reach an overall ratio of 2.0, based on batch-1,2 and pilot batch experiences. Weighted loan amount is Rs 17480 per VSCO loan beneficiary, increased annually by 12% (2009/10). Livelihood loans covered 74% of the VSCO members. � VSCO led livelihood investments promoted T�8   VSCO Performance by Batches    efficient use of heavily subsidized fertilizer and VSCO Funds  VSCO Loan  other scarce resources in paddy (accounting for Batches  Rs M  Disbursement Rs M  Ratio  1/5th of VSCO loans). Community opted for Pilot Batch  89  304  3.41  production of 1370 MT of seeds to support First Batch  194  316  1.63  grain paddy farmers, Easy access to community Second Batch  544  891  1.64  managed cheaper credit aided timely use of Third Batch  439  612  1.39  inputs like seed, machinery, chemicals and Fourth Batch  270  311  1.15  labor to reduce unit cost of production of paddy  Total  1,536  2,434  1.58  by 18% (Box-1) T�9 Gemi Diriya: Cost efficiency in Infrastructure investments � CDD led infrastructure investments   achieved cost efficiency, as compared to public Concrete Road  Unit Cost  Expected Life  funded infrastructure works, estimated at 20% Roads  Length, m  (Rs/m)  (years)  due to savings in unit investment cost of Gemi Diriya  4418  1863  20  infrastructure investments, and 43% including GoSL  Depts  1073  3295  15  community contribution through cash and labor Comparable sample roads are drawn from Ratnapura, Badulla and Matara  districts,  representing  Gemi  Diriya,  Provincial  Roads  Development  (T-9 and A-5) Authority, and Samurdhi. Data compiled by Infrastructure unit of GDF and  cross verified in the field by the Impact Assessment Team 50 � Community led investments in skill development achieved cost efficient job placements for unemployed youth in project villages: Average cost of job creation is only Rs 12,550 per job, and 58% of the skill development loans have already been repaid which will generate more jobs at reducing cost. More jobs created for rural youth in project villages- annually, youth unemployment rate came down by 4.2% in the project villages as compared to 2.2% for the country as a whole (T-6). � The project has demonstrated that with convergence in investments both within and among the livelihood and infrastructure activities, returns to V.O investment funds will be maximized (ERR at 42%). Annex 3. A-1 Estimation of Project Beneficiary Households VO VSCO Sub- Total Member Member VSCO Beneficiary   V Os  Projects  HHs  HHs  HHs  Borrowers  HHs  Livelihood Investments 1/  1,034  139,223  202,498  183,349  149,392  110,758  110,758  Infrastructure Investments 2/                Sub-projects initiated  1,009  3,518  191,618        191,618  Sub-projects completed  648  1,415  121,313        121,313  Total Beneficiary HHs   3/               Livelihood only  386            41,347  Livelihood & Infrastructure  648            69,411  Infrastructure only  648            51,902  Total Project Beneficiaries from Livelihood and Infrastructure Investments  1034            162,660                80.3%  References for Table A-1:                1/ Gemi Diriya 1st Quarter 2010 Progress Report.pdf, Fin Phy Progress 31st-Mar 2010_PI.pdf and eGemi Diriya at a Glance_March 2010.JPG, Gemi Diriya Foundation; and analysis of sample HUBs from all project districts by GDF Micro finance unit, covering 67% of the total VSCO loans issued by the project.  2/ Infrastructure database 2010MAR28 (Version 1)_1.xlsx and E-mail communication from Gemi Diriya Project Technical Unit.  3/ 80% of the V.Os have more than one infrastructure sub-project. Roads, multipurpose buildings, bridges, culverts and causeways which will potentially benefit entire village community account for more than 75% of the infrastructure sub-projects. Even in case of other infrastructure projects like drinking water, sanitation and other projects, about 50% of the V.Os has more than one infrastructure sub-project. Hence all HHs of the V.O are considered as beneficiaries if infrastructure sub-projects are completed.  51 Annex 3. A-2 Project Costs by Component (in US$ million equivalent) (NB unlike Annex 1 total per component costs are included) Appraisal Estimate Actual /Latest Estimate Percentage of Components (US$ million) (US$ million) Appraisal A. Village Development 49.50 40.78 82% B. Institutional Strengthening 5.70 2.7 49% C. Innovation Seed Fund 1.30 0.05 4% D. Project Management 4.40 4.70 107% E. Village Self-Help Learning 1.00 1.00 100% Total Baseline Cost 61.90 F. Gamaneguma 2.38 Z. Tsunami 8.29 Physical Contingencies 2.30 0% Price Contingencies 5.60 0% Total Project Costs 69.80 58.11 83% Total Financing Required 69.80 58.11 83% Annex 3. A-3 Gemi Diriya Project: Livelihood investment activities by major categories@ VSCO Loans Equity Average Livelihood sub- VSCO VSCO Disbursed Contribution project project categories  Borrowers Loans  (Rs M)  (Rs M)  cost (Rs)  Food grains  23,524 29,177 437 120  19,067  Horticulture Crops  33,518 39,154 652 528  30,161  Commercial Crops  31,545 37,181 638 250  23,888  Animal Husbandry  3,827 5,457 106 67  31,751  Self Employment  25,203 28,254 601 229  29,349  Total  117,617 $ 139,223 2,434 1194  26,058  @ Estimated for the project, based on the analysis of sample HUBs from all districts, Covering 67% of the total VSCO loans. The sample analysis was done by GDF Microfinance Unit. $ 79% of the borrowers are female. 80% of the borrowers belong to poor and poorest category. 52 Annex 3. A­4 Gemi Diriya Project:  Seed paddy impacts$    Seed Paddy farmers  Number  566  Project V.Os  Number  47  Area under seed paddy  Hectares  309  Seed Paddy Production  MT  1,370  Grain paddy area under certified paddy  Hectares  13,704  Annual Incremental gross margin@      Seed paddy production Rs/HH  44,526  Grain paddy production Rs/HH  18,260  Fertilizer Subsidy     Fertilizer provided kg/ha  505  Financial price Rs/kg  7  Economic price Rs/kg  140  Fertilizer subsidy Rs/HH  33,583  Fertilizer subsidy, Total Rs M  920  Incremental grain yield      With certified seed Kg/HH  600  Incremental paddy Production  MT  16,444    With project interventions (certified quality paddy seed and credit) through livelihood loan support,  public funds (fertilizer subsidy) are efficiently utilized.    $ Project level data compiled from GDF livelihood unit    Annex 3. A-5 Community Infrastructure Projects Cost of Community completed Cost Share of Infrastructure Started Completed projects (SLR Completed Cost per sub- projects  (No)  (No) M) projects  project (SLR) Drinking Water  385  129 147 27% 114,1042 Road Development  1,941  859 254 46% 295,493 Building  562  133 82 15% 616,549 Irrigation  63  28 8 1% 270,867 Sanitation  229  110 25 5% 231,142 Bridges, Culverts & Causeways  166  98 23 4% 238,183 Other Projects  172  58 9 2% 151,835 Total  3,518  1,415 548 100% 387,408 53 Annex 4. Bank Lending and Implementation Support/Supervision Processes (a) Task Team Members Responsibility/ Names Title Unit Specialty Lending Meena Munshi Task Team Leader SASRD Terrence Abeysekera Sr. Economist/Co-Task Leader SASRD Community Anne Ritchie Consultant Finance/micro finance Samantha De Silva Sr. Social Protection Specialist SASED Community development Surendra Govinda Joshi Sr. Transport Specialist SASDT Rural Infra. Spec. Community Organization Wahida Huq Sr. Operations Officers SASDA and Operational Aspects Barbara Verardo Social Development Specialist SASRD Social, gender and youth R.R. Mohan Sr. Social Development Specialist SASDS Social safeguards Sumith Pilapitiya Lead Environmental Specialist SASDI Environmental Aspects Yoshiko Ishihara FAO, Social/Gender Aspects Gender Aspects Ghazali Raheem FAO, M&E Aspects MIS Deborah Ricks Program Assistant SASRD Team Assistant Sector Manager (until Gajan Pathmanathan Sector Manager SASRD 2007) Adolfo Brizi Sector Manager SASRD Sector Manager 2008-09 Harini Wijesundara Program Assistant Jiwanka B. Wickramasinghe Financial Management Specialist SARFM Financial Management Manvinder Mamak Sr. Financial Specialist SARFM Financial Management CDD/Community Report Parmesh Shah Participatory/CDD Aspects Card Suzanne F. Morris LOAGS Disbursement Tashi Tenzing Sr. Sanitation Engineer Infrastructure Vikram Raghavan Counsel LEG Legal Business Linkages/Private Vinayak Narayan Ghatate Consultant SASDA Sector Sector Manager Gajan Pathmanathan Sector Manager SASDA Supervision/ICR Terrence Abeysekera Consultant SASDA Economic Aspects Henry K Bagazonzya Sr Financial Sector Spec. SASFP Community Financing Vijaya Bharathi Consultant Social Mobilization Dhimant Jayendraray Baxi Sr Procurement Spec. SARPS Procurement Vinay K. Bhargava Consultant SDV Governance Darshani De Silva Environmental Spec. SASDI Environment Vinayak Narayan Ghatate Consultant SASDA Partnerships Vichitrani Liyana Gunawardene Rural Development Specialist SASDA Livelihood Aspects Priyantha Jayasuriya Arachchi Team Assistant SASDO Team Assistant Business Promotion and Sitaramachandra Machiraju E T Consultant SASDA Microenterprise Manvinder Mamak Sr Financial Management Specialist SARFM Financial Management Seenithamby Manoharan Senior Rural Development Specialist SASDA Infrastructure 54 Ramachandran R. Mohan Senior Social Development Spec SASDS Social Safeguards Sumith Pilapitiya Lead Environmental Specialist SASDI Environment Ghazali Raheem Consultant SASHD MIS Eashwary Ramachandran Operations Analyst SASDI Environment Local Governments, and CS Renjit Consultant SASDA CDD aspects Anne Ritchie Consultant Community Finance Brenda Lee Scott Information Assistant SASDO Team Assistant Parmesh Shah Lead Rural Development Special SASDA Community Organizations Meera Shenoy Consultant SASDA Employment and Youth Economic and Mio Takada Rural Development Specialist SASDA Operational Tashi Tenzing Sr Sanitary Engineer SASDU Infrastructure Inclusion, Gender and Barbara Verardo Senior Rural Development Specialist SASDA Youth and Operations Jiwanka B. Wickramasinghe Sr Financial Management Specialist SARFM Financial Management Enoka Wijegunawardene Financial Management Specialist SARFM Financial Management Miriam Witana Procurement Specialist EAPPR Procurement Samantha de Silva Sr Social Protection Specialist SASED CDD Economic and Financial S. Selvarajan Economist/Consultant FAO Analysis/ICR Team Luis Coirolo Lead Sector Specialist/Consultant CDD Advisor/ICR Team Lead ICR Author/Team Jim Hancock Monitoring and Evaluation/Livelihood FAO Leader b) Staff Time and Cost Staff Time and Cost (Bank Budget Only) Stage of Project Cycle USD Thousands (including No. of staff weeks travel and consultant costs) Lending FY03 48 125.86 FY04 78 262.45 FY05 0.00 FY06 0.00 FY07 0.00 FY08 0.00 Total: 126 388.31 Supervision/ICR FY03 0.00 FY04 10 12.61 FY05 65 145.47 FY06 84 238.59 FY07 57 165.66 FY08 56 220.66 FY09 17 171.34 FY11 7.65 59.20 Total: 289 954.33 55 Annex 5. Beneficiary Survey Results The results of the beneficiary survey that was done as part of the Impact Assessment 2010 was also covered in the Borrower’s report (annex 7). 56 Annex 6. Stakeholder Workshop Report and Results Gemi Diriya ICR Beneficiary Workshop, Badulla, 16 August 2010 A beneficiary workshop was held with 115 participants, mostly leaders, from village organizations from 73 villages across Badulla. The VOs represented were from all 10 Phase I Divisions across Badulla (Ella, Haldumulla, Kandeketiya, Mahiyanganaya, Lunugulla, Passara, Rideemaliyadda, Soranthota, Uva Paranagama, and Welimada). The primary purpose of the workshop was to get feedback from the project participants on their views on of concrete achievements and lessons learned. For most of the workshop the participants were divided into thematic groups, based on the main activities and benefits they had been involved in, where on their own initiative they facilitated the recounting of testimonials from all participants, and establishing a group consensus to be presented. In some groups they used the testimonials to identify average benefits across the group. MAIN ACHIEVEMENTS OF THE PROJECT Beneficiaries of 1-time grants for the poorest of the poor (10 component beneficiary participants)  Had used the grant for animal husbandry, farming, and small shops. The grantees had all graduated to VSCO loans.  Impact: Initially their houses were very small, and with few amenities. Now 60% had good housing, and all had all facilities: electricity and water.  Before no land ownership, now 30% of them had land, and one was cultivating an acre of tea.  They have been able to uplift their children’s education  They noted an ‘80% improvement in living standards’  Access to loans has increased by 80%, and they can even access loans from outside  Income has doubled. From total income saving 20% for hh.  Some of the poorest of the poor are now bookkeepers and members of the Board of the VSCOs.  Before they did not have savings or savings capacity, now even their children have passbooks. One has saved Rs 5,000. Livelihoods (25 component beneficiary participants)  All had benefited from VSCO loans.  Assets: Average housing improvement was ‘15%, electricity 5%, water 15%, land 5%’ (units not specified).  25% improvement in nutrition, health and education.  Incomes have increased by 80%, for a few of them it has more than doubled. 57  1 person is now ready to take Rs 300,000 bank loan.  Group practices and unity are the secret behind the success.  1 small company owned for dairy.  Disabled boy in household recovering and supported through all the benefits. Infrastructure (28 component beneficiary participants) – split into further subgroups Multipurpose building  Neutral meeting place where none before, for officers, also for local government, and training centre.  Also economic centre (for sales or collection of goods), and IT centre. Water supply  Saving time in collecting water.  Mental pressures are reduced (from previously having to worry and spend time on this).  Water quality available with health improvements – kidney problems reduced.  Youth have become caretakers and earning from this.  Have developed knowledge on procurement.  They put labour and cash contribution into infrastructure, which has created a real sense of ownership. Irrigation  Previously had only 1 season cultivation and dependent on rain. Now can have even 2-3 cropped seasons per year.  Before used to labor for one season, now own the farm!  Reduced time consumption away from farm.  Previous conflict over water now solved.  Previously paid more for water, now can spend it on children.  With secure irrigation water here has been an expansion of agricultural land.  Cost effectively completed project.  Income increased 50%.  Feel ownership due to the contributions they made in cash and labor. Roads  Has made time savings for a number of community activities.  Transport costs reduced  Agricultural products transportation difficulties reduced.  It has reduced post harvest losses.  Better price for agricultural products – as now not a monopoly by one trader.  With illness and disease can get treatment fast.  Contribution has created ownership, and willingness to sustain the structures.  Has minimized environmental issues. 58  Impact has also been in generally better linkages by the community to other communities and urban areas: better connectedness and access to facilities (don’t feel left out).  Land values have increased by 50-80%. Playground  Place for youth to get together.  Improves unity and, and has reduced mental unrest of children. Sanitation  Has reduced pollution in the environment.  Has reduced waterborne diseases. And hence also reduced expenditure on diseases.  It is a cost effective process: Under Gemi Diriya Rs 0.5 m constructs 9ft wide, 5 inch thick concrete road that is 482 feet long. With a contractor, with same specifications it is 322 feet long (project road is 50% longer). This is even though there is also 20% labor contribution on other projects (i.e. close supervision under the Project further enhances quality).  For buildings construction costs are Rs 1250 per square foot for government projects, Rs 900 under Gemi Diriya (almost 30% less) Youth groups and IT trainees (15 component beneficiary participants) Youth circles  Career guidance has been provided.  Get chance to express ideas, provided chance to reveal hidden competencies, and build leadership among the youth. Youth can now make important decisions.  Youth have been important to build interaction between villagers and other organizations.  Youth groups help to create a sense of overall unity in the community, and have helped to minimize antisocial behavior in the community.  Organize action to deal with hazards and issues, such as dengue outbreaks IT trainees  IT centers, and e-villages have been established. And knowledge has been developed on wider IT services - for example in the use of range of software packages.  This has helped to develop access to agricultural knowledge and information, and identify job opportunities. They can now identify markets through the Internet.  They have helped to buildup information technology system (MIS) of the Project.  Other IT services are provided in the community including photocopying and internet. Skills development for youth  The process has identified youth needs, and job opportunities. 59  The Project has provided financial facilities for vocational training, which has also created innovative capacities.  This has increased profitability of existing businesses and resulted in income generation. Earlier they had very low or zero income (no employment), now from IT centre earn Rs 15000 per month. The discussion also examined impacts of institutions at different levels: VO Federations (18 leaders as participants)  Through the federations business links have been developed: with Cargill’s, Luck Yoghurt, for supplying raw materials.  They have provided coordination for job opportunities for community professionals: with for example Cargills, and Plan International.  They have ensured proper linkages with other institutions on agricultural production, inputs and marketing, and also strengthening Bank network.  They have helped to strengthen weak villages and help to startup new ones. Accountability group (14 committee members as participants)  Have been able to minimize malpractices, and do it in accost-effective way. They have been able to establish procurement system and put in place proper accounting systems to reduce project cost.  There is a clear legal understanding by BOD members on company activities.  They do self assessments to establish organizational status, and establish problem solving system within the village.  They have contributed to improved linkages between villages (due to peer exchange). Small groups and VSCOs (19 SG members and VSCO leaders as participants)  Institutions have been established to access loans more easily and at lower interest rates. They also have also access to instant loans. Members have used compulsory and voluntary build up of savings, and both are continuing. Have increased capacity, through savings, and have been able to repay loans with interest.  As a result they have escaped from dependency to moneylenders. For example in Badulla 50% of people are in the estate sector, people who have to pay 20% interest rates per month to money lenders, now this is 1.5% from VSCOs.  Due to a proper monitoring and evaluation system the group can screen for suitable activities. All this has contributed to household economy.  Members have been able to increase income directly though this experience of project support, by 5000-6000 per month. Some, with milking cows, now get up to 18,000 per month. Estate salary is Rs6000 per month, now through all benefits this is doubled.  Due to household income increases they can provide better education for their children and people are diversifying into new types of livelihoods activities.  Their prominence in VSCOs has built up acceptance of women as key players in the process. The sense of decision-making power is very much with the group. 60  They do not have fears and risks due to the presence of the loan security fund.  Also due to second-generation institutions (VSCO, DSCO), they have been able to keep strengthening their capacity.  8 VSCOs form a Zonal SCO, and 5% of profits (from VSCOs) are put into this. They will have a social welfare fund for disabled, children and widows.  Other benefits include the reduction in the production and consumption of illegal alcohol. Additional benefits not noted during the group discussions but participants wanted to add:  Especially in the estate communities, monthly earnings are Rs 6000, but only in particular season when work was available. Now have extra income generating activities. Off-season they may earn Rs 3000 per month.  Community members have also been empowered in the sense that there has been recognition for people with talent, not just for paper qualifications.  There has been also a sense of empowerment in that they now deal with officials up to the District level.  Also the there was a considerable degree of ethnic integration trough the village wide and inclusive VOs.  Before activities were implemented by NGOs, but now by community professionals, who provide very good support, to build up skills and to earn money.  The participants were also asked of there had been any major significant negative impacts. Again because of the inclusiveness and breadth of interventions, they felt there were no major negative impacts. LESSONS LEARNED Poorest of the poor with 1 time grants and Livelihoods Beneficiaries (35 participants) Why did it work well?  80% of the community agrees to the plan.  Planned prepared with the village.  Implementation is with ongoing support of the village.  The small group concept means that there is savings buildup, technology transfer, new development vision and direct income generation. This it is not just about loans. The small group level capacities, serves as building blocks for the village level organization. Suggestions for improvement:  More convergence with government agencies, to get their support through specific programs.  But otherwise all the rest is done in a ‘perfect manner’! Infrastructure group (28 participants) 61 Why did it work well?  Selection, planning, implementation and maintenance of development activities are done by the whole community - this creates ownership.  There is direct finance to the village, with no intermediary barrier. This means they can do procurement correctly and therefore trust the quality of the infrastructure. Suggestions for improvement:  Emerging community professional leaders, give them professional qualifications. Provide career opportunities, locally and abroad.  Get more involvement from outside organizations.  Arrange to do all infrastructures the Gemi Diriya way!  Community infrastructure contributions are now 20% labor, and 10% in cash. Suggest reducing cash contributions to 5%. Youth and skills development group (15 Participants) Why did it work well?  Youth have the ability to make the decisions themselves. A youth circle member cannot have a decision made by their father who is may be a VO Chair.  The amount of funds allocated to youth is mentioned very clearly in rules in the COM, as are the number of youth who are to be on every committee. Suggestion for improvement:  Most youth are being involved, but provide more recognition, for example for skills developed through Project, by some form of certification from the District Officials. VOs and Federations, and Accountability groups (29 participants) Why did it work well?  They have financial management authority as well as decision making power.  There is a focus on all key categories of people in the village: women, youth, poor, and very poor.  There are good governance and accountability rules and principles.  On the Village Organization model as company: it provides VO with legal entity that can have sales and make commercial agreements. Benefits go to each and every member of the company. They also have decision-making power over all activities. Suggestions for improvement:  It is the best model for development; it does not need any changes, just has to be scaled up. VSCO group (19 Participants) 62 Why did it work well?  99% participation by women and where caste or age does not matter, and benefits go to each and every member.  The use of the 10 Golden Rules.  It is important that they have allocated to a livelihoods development fund which is dedicated to livelihoods, and have in place a mechanism to focus on building this up, so more people can use it.  Credit accessibility before was low, with many round trips to bank, now the community groups can issue certification. Suggestions for improvement:  Need more insurance for crops and livestock. Some village has this already, but there is a need to spread it out further.  Sometimes they have to take legal actions against member, so they suggest to recruit legal officer at District level to advice on such cases. SUMMING UP OF THE DAY’S WORKSHOP FINDINGS – AGREED WITH PARTICIPANTS Benefits:  A wide range of specific benefits were noted.  No major significant negative impacts were noted by this group.  The model should not be changed. Lessons:  Universal involvement is critical.  Decisions are made locally.  There is direct ownership of finances and implementation.  There are clear universal rules. Recommendations  Would like to see certification of skills developed.  There should be more flexibility on cash contributions.  There should be more convergence with other agencies. There should be specific services to spread out or add, like insurance and legal advice. 63 Annex 7. Summary of Borrower's Completion Report November 2010 Evolution of the project The village development efforts in Sri Lanka dates back to almost two millennia. During King Panduwasadeva the “Grama� (village) concept began and evolved to its peak during King Pandukabhaya’s regime. The village development concept at that time was multi- faceted and holistic incorporating social, ecological and environmental in addition to economic and political considerations. More labor oriented with limited funding from the kings treasury. Coming back to the last century we should not forget that we had been ruled by three European states for about five centuries and our systems were disrupted and alien thinking and processes were planted among us. The village expansion programmes, colonization schemes, and more recently Janasaviya, Samurdhi, and several other state sponsored programmes and projects were centrally planned and peoples’ participation was mostly passive but active only on certain activity in the process and often never heeded the local community aspirations adequately. Supply driven processes under the control of the state bureaucracy and the political leadership made rural communities dependent upon the government service delivery for development which no government could guarantee fully given the human and financial constraints encountered. The Village Self Help Learning Initiative (VSHLI) tested the hypothesis that if communities are given information, decision making power and supplementary resource assistance they would manage their resources better. VSHLI was implemented in three villages in Polonnaruwa at the beginning under the Mahaweli Restructuring & Rehabilitation Project (MRRP) in 1999. The successful results prompted the government to scale-up this community driven development approach in seven districts and “Gemidiriya� Community Development and Livelihood Improvement Project was launched in 2004 as a result with a grant of US$ 51 million from the World Bank as the first phase of the 12 year adaptable programme of three phases with assistance of US$ 181 million on successful results. The second phase of the Gemidiriya Project was launched with the government’s acceptance of the design and approach of the project and the granted title of Gama Neguma Community Driven Development (GNCDD), the Second Community Development and Livelihood Improvement Project to practice CDD with the graduation from village to the local government level with the Pradeshiya Sabha inter village connectivity development with a sub regional level development intervention). Project Coverage The project covered Southern and Uva provinces and added Rathnapura District on a request made by the government at the time of launching the project in late 2004 with the justification that Rathnapura district was the third poorest district after Badulla and Monaragala being the first and the second among the districts covered in the South. Thus the inclusion of Sabaragamuwa and North Central expanded the project coverage to 64 Hambantota, Galle-Matara, Monaragala, Badulla Polonnaruwa and Rathnapura districts, covering 1034 villages in 54 Divisional secretariats in 851 Grama Niladhari Divisions. The total population in the project area is 876,033 and the number of households is 202,498. PAD indicates 276 as mean HH/GND. For targeted 510 GNDs the expected HH would then be estimated as 510 x 276 = 140,760. The population of those 510 would be estimated as 140,760 x 4.4 (average family size) = 619,344. PAD estimate is 150,000 HH and 700,000 people as beneficiaries. Also PAD estimated about 1000 village communities in 510 GNDs. The 26 villages of Polonnaruwa VSHLI were included in the US$ 1 million of the US$ 51 million grant. These villages continued to get support after the US$ 1 million was utilized. Those villages are included in the 1034 villages. The PAD estimate of the number of beneficiaries was 700,000 people in 150,000 HH where as the project at the end of the Phase I yielded benefits to 876,033 people in 202,498 HH, with an investment minus US$ 8 million from the original US$ 51 million moved to Tsunami relief. The additional funds required for all the villages were estimated as US$ 15 million which has been provided under the second phase. Overall achievements The development objective of the project is to improve the livelihood and quality of life of the rural poor. The main outcome indicators are the percentage incremental income increase, percentage households benefiting from increased income and the percentage households benefiting from community infrastructure. The percentage of households earning less than 1 $ a day is expected to be reduced from bench mark level 50% to 40% at the end of the phase one. The implementation of phase one has followed a batch-wise approach and the four-year period has been completed by Pilot (59 VOs initiated in 2000), batch-1 (123 VOs initiated in 2004) and batch-2 (350 VOs initiated in 2006). By the end of March 2010, batch-3 (299 VOs initiated in February 2007) and batch-4 (203 VOs and initiated in Oct-2007) have been completed 3 years and 2 years respectively. The project cycle of the last two batches will end in March 2011 and December 2011 respectively. Poverty: The income based poverty indicators show that 43% of households live below poverty line in GD villages while it is 48% for control villages based on the impact assessment study. The depth and severity of poverty appears to be higher in control villages than GD villages. The expenditure approach to the poverty measurements indicates similar results as the income approach, except in depth of poverty. The proportion of households with per capita income of less than a dollar per day in treatment villages is 44%. This overall figure is little less than the indicative target (40%) but the initial batches have reached this indicative target (37%). When compared with the MTR estimates (71%), the present level (44%) indicates a substantial reduction in poverty level in GD villages. Similar changes could be observed in control villages but, the poverty level remains at a higher level (51%) than that of GD villages (44%). The changes of the relative poverty levels in control villages and GD villages indicate the positive impact of the GD programme. The estimated figures are much higher than the national estimate of 65 HC (15.2%). The GD villages represent the poorest villages and the estimates are not directly comparable with the national figure. Change in Income: Household income increase after GDF interventions was taken for calculation by considering the household survey carried out in 2007 for Mid Term Review (MTR) and the household survey carried out in 2010 at end of Phase I for the Impact Assessment Survey (IAS). The treatment population of the project villages includes those who have received livelihood support through the VSCO loans. Therefore, the analysis of income change includes only beneficiaries of VSCO loans. The sample households of pilot and batch-1 in MTR and IAS are the same and thus, the income changes of these two groups are directly comparable. The results indicate that poor households in GD villages have increased their household income by 54%. The overall increase in income of poor GD households at the end of the phase one is 42%. In case of control villages, the poor households have increased their income by 17%. The poor households in GD villages have increased their income more than the poor households in control villages and the poverty levels. The project interventions are largely directed at infrastructure development and livelihood improvement activities. The provision of low cost credit to invest on income generating activities, one time grant to poorest of the poor to initiate income avenues, and a special loan scheme to unemployed youths for skill development programmes also contribute to household income. The production and productivity improvement technologies (new varieties, better quality seeds), capacity building programmes (farm extension), and exposure visits provide direct and indirect effects on household income. When compared with the results of the MTR, the majority of households have been benefited from increased income. The majority of households have increased income from agriculture related activities and the majority of beneficiaries have taken VSCO loans to initiate new agricultural activities and to develop existing agricultural activities. Horticultural crop activities have generated the highest increment in income (43%) The income estimations include the direct income changes due to the project interventions. Indirect income changes due to time savings, quality of inputs, cost reductions, etc are not included in the analysis. Analysis includes only short-term gains. Certain investments such as perennial crops (tea, coconuts etc), which brings higher returns after a certain period of time, are not included. Even with these shortcomings of analysis, a higher proportion of households have increased income due to the interventions of the GDF. Infrastructure Development: The project targeted to cover 150,000 households (or approximately 700,000 people). Infrastructure development such as, rural roads, drinking water, irrigation systems, etc. have been promoted and the poor is expected to be the main beneficiary of these infrastructure developments. The households benefiting from community infrastructure was calculated as a ratio of number of beneficiary households to total number of households in community. The results indicate that 95% of the households have been benefitted from the infrastructure development activities and it indicates the distribution of infrastructure development expenditure to improve the 66 welfare of the majority. Community contribution amounted to Rs 407 million and it accounts for 26% of the total value of construction cost. This Rs. 407 million is from all the completed and ongoing subprojects in 1007 VOs. In addition the community has donated lands worth Rs. 148 million. If all these are added up, the total community contribution accounts for 38% of the total value of construction cost. The majority of households have indicated that funds have been utilized prudently with the consent of more than 80% of members of the VOs. The quality of construction and cost efficiencies are higher than the similar projects conducted by other agencies. Long-term maintenance of infrastructure has to be considered and development of a maintenance fund and linkages with line agencies would facilitate such maintenance activities and obtain necessary technical support to ensure the long term sustainability. When the 3518 sub projects are fully completed 191,698 HHs will benefit. However by 31st March 2010 only 1415 subprojects are being fully completed and 121,313 HHs are the direct beneficiaries. This gives a figure of 60% of the HHs in the project villages. However the sub projects still not fully completed too provides some benefits to the projects HHs particularly the road sub projects where some sections that have been completed. The key performance indicators to measure the achievements of the project activities as indicated in the PAD are (B.3.2, page 4), given in the table below. 67 Achievement of PDO against key performance indicators End of project Achieve- Key Performance Indicator Remarks Target ment (PAD) 1 No of VOs formed & percentage Criterion for this KPI is that VOs have of these functioning effectively in 65% 100% accessed the CB 1st installment; priority setting & decision making Source: ML&E data base GDF on resource allocation 2 Percentage of sub projects that are The communities have completed and implemented and maintained by 60% 141% maintained 1415 sub projects; many the communities VOs have done more than one sub project; Source: Infra data base, GDF 3 Percentage increase in household At MTR stage it was 17%. IAS-2010 incomes of the poor 30% 42% recorded 42% in GD villages and corresponding control villages shows only 17%.; Source MTR-2007 & IAS- 2010 4 Percentage of VO (VSCO) with IAS survey revealed that in the sample sustainable savings & credit 40% 87% it is 100% (72 VOs); the microfinance systems at the end of the project database indicates the overall figure for the project; Source: MF database, GDF 5 An increase in private sector 25% 45% The number of private sector investments in the communities; partnerships is counted here; no direct qualified as partnerships investments yet; Source: Livelihood Database, GDF Most of the targets set out in the indicators have been achieved or exceeded. Despite several VOs have made partnerships with the private sector agencies these agencies have not invested substantially. It is evident that more vigorous lobbying for private sector investment to boost the value addition of agricultural products, technology transfer and cost sharing with the VOs and VO federations is needed. Component wise achievements The process adopted to achieve a totally independent, self reliant and politically independent environment for project implementation by Gemidiriya at village level is innovative and commendable. Strategic changes in the current very independent approach of GDF in project implementation may be needed to provide opportunity for participation of line agencies, local governments and administrative authorities in supporting the project implementation without damaging or diluting the basic principles of GDF and self-initiatives of target communities. This may be vital for long term sustainability of the project in the current operations mode. Safeguards were adopted. Implementation of a large number of projects reduces the availability of technical experts for proper guidance. No major environmental issues have arisen due to infrastructure development. Principles of gender equity and inclusion of youth need at all levels of infrastructure development for the benefit of the entire community has created an enabling environment for women and youth’s participation. Inbuilt financial controls, thrift measures, transparency measures and community 68 participation and code of conduct, etc. have lead to higher quality, environmentally friendly and highly cost efficient infrastructure development. The demand driven approaches is quite commendable and will be a positive factor for sustainability. Community Professionals / Facilitators play an important role in community mobilization and imparting basic practical technical knowledge to VOs in implementing infrastructure projects. It is recommended to take measures to upgrade the theoretical technical knowledge of CPs and to arrange them to get nationally accredited qualifications in collaboration with universities, technical colleges and vocational training institutions. More attention could be given on promoting green and clean villages and training more CPs to develop environmental friendly sustainable livelihoods. Component I: Village Development – All VOs (1034) have been formally registered, receiving funds and effectively implementing village priorities. The Information Communication and Education (IEC) programs, project management, and support of provided by regional level public agencies have highly contributed to achieve the project targets of formation of VOs (1,034) and development of these VOs as legal entities. The success of formation of VOs and initiation of activities indicate both the effectiveness of project design and implementation. The experiences and interactions within and among the membership of VOs have largely contributed to the development of social capital in project villages. All VOs have initiated at least one subproject and 61% VOs have completed at least one sub-project. The community contributions (cash and labor) to infrastructure projects are at satisfactory levels and 87% of beneficiaries have contributed to cost of infrastructure sub projects. The village infrastructure sub projects show a higher quality and are more cost efficient than similar sub projects provided by other agencies. The inclusion of the majority of the population and in particular the inclusion of the poor and the poorest of the poor, of women and youth in project activities as well as in decision making positions of various sub committees has been achieved to the expected level. The selection of the target groups have been conducted with the participation of village community and a high proportion of the poorest of the poor and poor groups has included in project activities. About 89% of the poorest of the poor have received one time grant and 72% of the recipients of the skill development loans represents the poor youth. All VOs have initiated or completed at least one sub project that benefit 80% of the poorest members. The close interaction between VOs and hub/district/national teams and utilization of own resources (community professionals) and very clear guidelines, internal and external monitoring learning and evaluation process have supported the development of strong village level institutions. Release of village development fund is bound by timeframes and milestones. The higher level of progress of release of initial installments indicates the effectiveness of VOs and supporting institutions. 69 Structurally vulnerable communities having less coping capacities eventually depend on natural resources for their livelihood. Agriculture, fishing and related wage labouring are the principle livelihoods of such population. Having identified the importance of increasing size and diversity of livelihood in increasing income of the poor and the vulnerable, from its design of the program, sizable funds were allocated to develop livelihood activities of the selected villages (Rs. 2,340 million). Particularly, the project has targeted the resource poor, unemployed, socially excluded households and household members of the poor villages. Through the livelihood support fund, Village Savings and Credit Organization (VSCO) have been instrumental in disbursing project loans as livelihood support. In particular, several income generating activities were supported through these loans either to start an income generating activity or to improve the existing income generating activity. The main crops and the livestock activities supported by the project were maize, tea, cocoyam, potato, vegetables, paddy cultivation and dairy production. Loans that have taken to start or to develop an IGA in non-farming sector are relatively less (22%). Rice Processing, cinnamon and other spices processing, dehydration of vegetables, lime production, copra and coconut oil production, handicrafts are the main IGAs supported by the GDF. The smaller size of the loan and the larger number of recipients indicate the presence of other resource constraints on investments agricultural activities. It also indicates the need for diversification of livelihood activities. The one-time grant concept shows higher success rate and it provides a means of addressing the poverty alleviation of the poorest of the poor. The skill development loans show similar success but the lower repayments indicate a sign of poor monitoring of skill development loans. The VSCOs has contributed to improve savings habits of Gemidiriya community. Higher proportion of households borrows money from VSCO (85%) than from other financial institutes (9%) and Samurdhi bank (6%). All VSCOs have the ability to cover their operational costs while some of them are showing the movement towards financial self sufficiency. Depth and outreach of the VSCO loans have been remain at very satisfactory levels. The majority of poor (82%) heavily depends on VSCOs for their credit needs. The relatively lower portfolio at risk and achievement of financial self- sufficiency and operational self-sufficiency (100%) are important characteristic of the VSCO. However, the portfolio at risk (PAR) is higher than the standards industry level and it needs particular attention. However, PAR is not directly comparable with other micro finance institutions due to absence of writing off policies in VSCOs. The majority of VSCOs are able to cover their administrative costs. Communities’ capability to save has substantially increased with the VSCO. When the VSCO loan beneficiaries are concerned, 62% of them have increased their income by either improving their existing income generation activity or introducing a new enterprise. The outcome indicators related to village development have surpassed the project targets. The other positive transformation of Gemidiriya village include changes in perception, increase in confidence, good governance, thrift in spending, improved participation in 70 decision making, transparency in funds utilization and grater equity in distribution of benefits of projects, higher quality and cost efficiency of infrastructure development projects. Women and youth empowerment and development of social capital are the direct benefits to the Gemidiriya villages. Component II: Institutional Strengthening - According to the VO survey, 90% of VOs claimed that they have received adequate support for formation of VDP while the MIS information indicates that all VOs have received adequate support for formation of VDP. Appraisals at PS level subcommittees have been initiated in 4 PSs and they are functioning with satisfactory service delivery standards. It was observed that strong linkage between VO and the PS has provided a synergy to the village development process. CP services have provided an employment opportunity for poor youth and 41% of VOs are receiving income from CPs. CP services are well recognized by VOs and they are performing better than the SOs. However, functioning of some VOs have severely affected due to placement of CPs in other GD villages. Establishment of partnership with the private sector organizations is expected to promote investments, employment and diversify livelihood activities. About 45% of VOs have established such partnership with private sector organizations. Among these partnerships, marketing arrangements for agricultural produce and input supply are the dominant types. About 50% VOs have been federated and functioning as business enterprises. The federations provide wider opportunities for village development and in order to reap of more benefits from the integration federations need further capacity building on potential enterprises and development of supply chains as a group. Though, the Gemidiriya project has shown some promise in empowering the communities with new vista in development, building public private and community partnership to assist in livelihood development was paid little attention. It has lost the opportunities opened within the dynamism in development in blending public sector resources, private sector capacity in obtaining new capital, technologies and market links and community commitment to promote a common goal in participatory development in rural sector. It was observed that most VOs operated in project areas were not directed with a clearly targeted plan for livelihood development by opening options for area based development with participation of communities. Without developing the required livelihood development plans, VOs were faced with an uphill task to operationalize with clear cut targets for area based development through empowering communities. Component III: Innovative Seed Fund - The innovative seed fund has mainly utilized for implementing innovative ideas such as establishment of IT centers in 175 VOs, establishment of CPLTCs, community radio, ICT initiatives of the project. About 17.5% of VOs (175 VOs) have been benefited directly from IT sheds. Households have access 71 to the IT centers established at the multi-purpose buildings and thus the members’ IT knowledge has improved. The internet facilities are being utilized to update data bases and as a medium of communication. The CPLTCs have provided an effective means of sharing the skills gained from different aspects of rural development and the idea has successfully scaled up to cover the all GD districts. The service of CPs has immensely contributed to improve the IEC and social mobilization programs in latter batches and also rectify the deficiencies of VOs of initial batches who were supported by SOs and also lagging behind. Component IV: Project Management - The facilitating team supports the release of funds for village development activities within two weeks and 61% of VOs have received funds within two weeks. About 90% of district and divisional facilitating teams have received positive scores from community score cards. The result indicates the provision of efficient service by the staff of the GDF. The management of the GDF also adopted the set of guiding principles, “Golden Rules�, which had helped in ensuring good governance, accountability, and transparency. The human resources consist of competent and committed workforce. In some districts and regions, the GDF has weaker linkages with public agencies and local governments. The communication gaps at the district development meetings were observed and this may have provided an incorrect image of the project to local and district level local governments. However, in some regions where such linkages are strong, the VOs in the area have highly benefited from a new synergy During the second phase of the project, GDF has to develop close link with local governments and project related public agencies, Non-Governmental and private sector organizations. The staff strength of GDF has gradually grown up in numbers as the activities of the project had expanded. These demand expertise of other areas particularly in the banking and finance, value chain development, brand management, research and development and development of business linkages, etc. System of budgetary control and planning is in place, expenditures are tracked against the budget and variances properly reported. An efficient appraisal and monitoring system has been implemented. Requests for funds by the village organizations are expeditiously processed; discussions held with the VOs indicate that they have been able to access funds within the stated time frames A significant sense of dedication and commitment prevails in most village level organizations visited, particularly in regard to record keeping and display of information regarding community activities. However, it is noted from ground realities that further accounting, book keeping and financial management support may need to be provided to village level organizations in order to strengthen and further fortify their abilities before the protective hand of Gemidiriya is withdrawn. 72 Component V: VSHLI - The VOs of VSHLI have federated and generate own resources for future development through CPs and business enterprises. About 50% of VOs are federated and 40% of VOs are functioning as business enterprises. About 83% of the HH in VSHLI projects indicated that their household’s income has increased due to activities of the project and the VOs. All VOs in VSHLI are providing CPs services to GD villages as well as to other villages. The innovative idea of establishment of CPLTCs has given much needed institution for GD approach. The experience of the VSHLI shows that strong linkages with other agencies promote the village development and federations offer more scope for livelihood development. Sustainability of VOs: The sustainability of VOs were assessed using regenerative capacity of funds, operational status of VSCOs, presence of sustainability plans, accountability of VOs, growth of membership, etc. All these indicators show high score indicating the sustainability of institutions. Communities have no major constraints in accepting the conditions imposed by the GDF on inclusiveness, community contributions, etc. The sustainability of VOs depend highly on how these community based organizations are capable of enlarging the opportunities of the VOs to develop linkages with the institutions outside the village/regions and to capture more opportunities to shape their members livelihood development. The longer term sustainability of such linkages, however, requires adequate support from public institutions and strong relationships with the local governments. Further, present exist arrangements are not clear and proper guidelines are required to ensure the sustainability of these institutions. Lessons learnt During the phase I of the project numerous lessons have been learnt and below we produce both positive and negative lessons. Positive Lessons 1. Core principles and elements of Project design 2. VO as a center for economic resurgence 3. Gender mainstreaming & principles of inclusion 4. Demand driven approach in identification of sub projects 5. Elimination of constraints at rural settings 6. Governance & participatory democracy 7. Accountability & financial discipline 8. Empowering the rural poor 9. Community savings as social capital 10. Emergence of rural leadership and competent technocrats Negative Lessons 1. Delays in completion of Infrastructure Sub projects 2. Convergence with other state agencies 3. Funding for Balance of VDP and other activity 4. Dearth of Community Professionals for services within & outside their villages Borrowers rating on Performance 73 Having considered all performance indicators reflecting the acceptance of the reform, practically demonstrated by exceeding the targets, utilizing the funds allocated 100%, winning local and international awards, and acceptance to scale up country wide by full ownership of the project under the government’s flagship name Gama Neguma village upliftment. The project has been already awarded, having received the admiration of performance by the Secretary to the President of the Country. It has already demonstrated ability to assisting projects in the countries in the south Asia region on innovative institutional development and governance practices and winning the hearts of the beneficiaries of over thousand villages with shared appreciation by their political representatives without political division. Based on this, the project justifiably and sincerely claims the performance rating “Excellent�. Banks’ Performance The performance of the Bank is reflected in the event of the task team winning the award for the best practice for supervision and implementation support among 270 bank funded projects of the Agriculture and Social Development Programme of the World Bank in 2010. This event marks a culmination of a process of a high degree of collaboration between the Bank team and the GOSL partners which created a series of significant events. It was observed that the high degree of collaboration resulting from making joint decisions and sharing responsibility of implementation from the Bank team caused a cohesive team effort to use the tools for supervision and implementation support effectively. The effective use of continuous and rigorous monitoring of agreed actions, the use of community professionals to work with bank missions to dig into information and probe in a transparent and trustworthy manner, allowed covering larger number of remote villages. Well represented gatherings of community members and staff were used to design manuals, change procedures and review project performance. The continued leadership of the Task Team with a very strong interest to test the model, sharing the project goal of poverty reduction from the very inception of the VSHLI pilot, and throughout the implementation process with prudent choice of team members from the best available had a strong bearing on the performance. The relationship of the three key agencies i.e. the Bank headquarters, the Country office and the Local Project office with its Governing Board as a triangle, is important to protect the project agreements, concepts, smooth implementation and compliance. If one corner weakens it will affect adversely in bringing about the necessary effects of the economic and social reforms envisaged by the project. Recommendations Village Development Plans: The Village Development Plan (VDP) preparation in phase I did not capture much of the livelihood activities and were not comprehensive. It should include and cover the total development of the village. The VDPs’ prioritized activities could be initially funded by the CDLIP, and the remaining through the decentralized GOSL provisions that go through the Divisional Secretariat, Pradeshiya Sabha, District Secretariat and the Provincial Council. A copy of the plan should be made available to the Divisional Secretary who would look for possible funding of various projects in the village. 74 Distribution of Funds for different types of activities in the village: It has been pointed out that funds were deposited in banks and not effectively used for the development activity. It is said that the proportional distribution for livelihoods, one time grants, infrastructure, suggested by the GDF, despite being flexible, was not so efficient for the utilization of the funds. Depending on the VDP needs and the community should have more autonomy in deciding the fund distribution. Livelihood Development: The rural areas of Sri Lanka are predominantly agricultural and often expansion of land (state lands) for rain fed cultivation is main the practice. No intensive farming has been carried out except for private sector initiatives with large capital investments. The livelihood programme of Gemidiriya targets the poorest of the poor, and the poor categories, in the rural/estate household profile. Hence it should have a number of strategies, where, depending on the availability of family labor, land, capital and supply of water, the community members decide on the best option to them. The skill development fund utilization should be closely supervised and BOD members should be properly trained to screen the applications of prospective trainees and only such training should be allowed through the NAITA or VTEC accredited institutions available in the district or elsewhere. The youth employment division in collaboration with the communication and IT divisions should provide more internet facilities to the VOs or at least to the clusters (group of 6-8 VOs). The youth Employment division should provide all the details of the websites of the reputed organizations and even help in verifying those behalf of the VOs/Federations until they become familiar with the process. Every effort should be taken to assess the employment availability in the area of training obtained since the loan has to be repaid. Further tracking of loan repayment has to be inbuilt to reduce the default rates. The livelihood development and VSCOs have to move hand in hand. Once the livelihood development options are introduced, financial capital should be built among the VOs through VSCOs to fund for activities in livelihood development. However, what was observed that without having livelihood options for the community targeting poor, poorest of the poor (POP) and other communities in the villages, they were facing the problems of increasing community income and improving savings and their livelihood. In this respect, to improve the effectiveness of VSCOs in livelihood development (i.e. poverty alleviation), communities should be guided with a clear area based development plan, incorporating livelihood development options and strengthening the provision of financial capital through VSCOs with new and appropriate technologies. Raising awareness among the community members of VOs on potential areas for livelihood development through establishing VCM approaches are recommended. Though some supply driven options was operationalized, the effectiveness in improving household income was limited. In this context, it is required a twin track approach with blending demand driven and supply driven livelihood options to cater to the needs of the communities, considering opportunities opened through neo-liberal market mechanisms with effective private sector organizations. 75 It is required to strengthen the input delivery systems and output marketing of agricultural and other SME’s within the framework of area based development. Within each district of operation, federating VOs needs to be strengthened in order to facilitate sharing resources, information, knowledge, skills and experience within the district to empower the community to gain additional benefits as a group. Linking of VOs into national and international networks of market development is required. This area based networks are a fairly recent phenomena in the Gemidiriya project and without such networks, VOs may not be shifted to value chain management (VCM) projects in the long run. GDF has to carefully review their existing livelihood development programmes and develop a new livelihood development plan considering VCM based agriculture, SMEs and other service sectors to cater to the community needs in livelihood development. More follow-ups are needed for improvement of federations’ activities to promote new business avenues. Support for innovations, research and development and incubation are necessary to facilitate faster rural livelihood development. However, the project funding for these activities have been very low. Institutional strengthening is necessary, particularly the capacity of local and national agencies and support organizations, to facilitate implementation of the phase II and develop effective policies to support CDD/livelihoods approach of rural development on larger more integrated scale. Village Savings and Credit Organization: Asset quality is an important element of VSCO and VO sustainability therefore it is advisable to give more attention on this to protect credit discipline built in VOs to ensure a smooth operation in the future. Also it was observed that some VSCOs are inactive due to poor loan repayment which mainly leads to affect the other operations of VO as well. Weather and price related risks make a higher threat to the sustainability of micro lending programmes introduced by the VSCO. Introduction of micro insurance programmes for agriculture and diversification of livelihood activities would help communities to cope up with weather and price risks respectively. The interest rates of formal financial institutions and Samurdhi banks show a downward trend and revision of interest rate policy of VSCOs may be required to match with the current trend. Loan loss provision was adopted by all the VSCOs based of a fixed percentage amount from loans issued during the year and credited to loan loss provision account. Thereafter, all the VOs have growing loan loss provision accounts. It was observed that some long term overdue loans were in the portfolio which contains high risk. Therefore, it is advisable to review adequacy of loan loss provision to cover the risk with age of overdue loans to ensure the revolving of loan fund without shrinking. Although voluntary savings is an important product of VSCO, those details are not reflected in the final account statements. Institutional Development: The functioning of Maha Sabhas (MS) is at a satisfactory level. However, size of the village, nature of employment of members and location of households act as a barrier to obtaining the minimum 80% of attendance. Innovations are needed to cater to location specific needs that support the sustainability of VOs. The capacity of VOs is mainly limited to being service providers but their capacity to develop as a business unit or economic center, that increases the value of shares, would provide strong momentum for sustainability. A federation is an ideal concept for the rural sector 76 but it needs innovative ideas and finds avenues to coordinate production groups and ensures a win-win situation for the partners of lower parts of the value chain of the production process. The social accountability approaches introduced by GD play a major role in promoting sustainability of VOs and capacities of social accountability and procurement committees have to be further strengthened to deal with various socio- economic backgrounds of the communities. Further, training and capacity building of CPs are prime needs to strengthen their services to the VOs. The use of CPs can further be promoted in Phase II as it is very effective to community mobilization process. It is recommended to take measures to upgrade the theoretical technical knowledge of CPs and to arrange them to get nationally accredited qualifications in collaboration with universities, technical colleges, and vocational training institutions. The process adopted to achieve a totally independent, self reliant and politically neutral approach for project implementation by GDF at village level is innovative and commendable. However, strategic changes in the present approach may be needed to provide an opportunity for local governments and other public sector institutions to support project implementation without damaging the basic principles of GDF and self- initiatives of the target communities. Though the Gemidiriya project has shown some promise in empowering the communities with new vista in development, building public private and community partnership to assist in livelihood development needs to be developed further. Blending public sector resources, private sector capacity in obtaining new capital, technologies and market links and community commitment to promote a common goal in participatory development in rural sector, needs to be pursued actively. The social mobilization and capacity building components are costly exercises for private and public sector institutions in order to do extension and development works, organize and capacity building of producer groups and promote sustainable agribusiness in Sri Lanka. But the GD programme has mobilized a large group of farmers, which could provide mutual benefits between VOs and other agencies. The private and public sector institutions possess meager knowledge about such development at GD villages and awareness programmes have to be conducted. Further, GD management/ federations/VOs should seek to establish strong linkages with the other national development plans/projects that could benefits the rural communities. The current institutional setup could easily be utilized to support social marketing, niche marketing, etc approaches and develop brand images to win the hearts of consumers. Similarly, failures in supply of services could be overcome by developing strong linkages with the provincial and central government institutions and private sector agencies. Social Infrastructure Development: A high share of community contribution for community infrastructure development was observed, but some marginalized and poorest communities could not afford to pay the cash amount at the necessary stage. As a result, they may have dropped out in the project process or the completion of the projects could be delayed. Further, the reasons for delaying of community infrastructure development need to be investigated and appropriate measures could be incorporated in to the implementation arrangements. There seems to be inadequacies in services provided by some TSPs, (they are available only on part time basis, and some lack specific expertise and experience required) especially during project implementation stages which caused 77 delays in infrastructural development. In such cases GDF staff had filled the gap. This was a burden to most of the hubs since their approved technical cadre was not expected to cater to that gap. Technical capacity of selected CPs, if enhanced, would strengthen the availability of TSPs. Building up a maintenance fund, especially in water supply schemes, to ensure long term sustainability of facilities is recommended. Project Management: M&EL and indicators. The MIS is not fully automated yet, often due to technical difficulties in remote areas. Nevertheless, timely access to information has largely been improved with the introduction of IT technology to the VOs. The VOs are directly involved in entering the data into MIS and real time data availability ensures prompt actions whenever necessary. The feedback on assessments has to be delivered to stakeholders. The success stories are given prominence but a lot can be learnt from failures as well. Therefore, evaluation component has to be strengthened. The ML&E depends on various methods and integration of all these methods would reduce the excessive workload on the members of VOs. There is some development in ML&E systems but the systems need further improvement to avoid delays in reporting. Simple indicators of results frame work provide easy but approximate measurements in many outcomes. Redefinition of outcome indicators would be needed to give real meaning to outcomes which ensures the sustainability of institutions. Post evaluation of completed projects and livelihood activities is lacking and presence of sound evaluation system would provide more information for learning. Establishment of sustainable VOs/federations and development of strong partnerships with public and private sector institutions, within a framework of evolving institutions that empower rural communities is a prime objective the GD programme. Such key objectives are not reflected in indicators of result framework and methodology has to be developed to include such indicators in the result framework. Management systems and capacity building. The Financial Management system is being conducted without any major irregularities at all levels of the Gemidiriya Project. Further accounting, book keeping and financial management support may need to be provided to village level organizations in order to strengthen their capabilities further. The organization has transformed through different stages, and further strengthening of the professional staff is also necessary to face the dynamic changes of the GDF. The institutional changes taking place, such as development of federations, demands expertise in new areas: particularly in the banking and finance, value chain development, brand management, research and development and development of business linkages, etc. Further, trained and experienced staff could be promoted for positions of national, appraisal, regional, district and divisional teams. This would improve the efficiency of the system and reduce the cost of HRD. Frequent review of institutional set up and internal management of the GDF could be carried out as an independent and comprehensive management audit, identifying best practices as well as gaps, and to propose remedial measures for the improvement of organizational capacity and managerial effectiveness of the project in a more efficient and effective manner. The capacity building needs are mostly identified by capacity building agencies and management teams, but a more systematic approach using continuous participatory 78 assessments of needs is necessary. Development of training evaluation component to provide effective feed back to stakeholders is necessary. Sustainability of the institutional arrangements: The smooth application of GD principles and the longer term sustainability, however, require adequate instruments to transfer public resources to these organizations, and it depends on the strength of the linkages of VO and the local government administration. Therefore, the adoption of new tools is required to complement existing methodologies of project formulation, appraisal, implementation monitoring, and performance evaluation. Further, one of the major objectives of the GD programme is to establish sustainable VOs/federations and partnerships with public and private sector institutions, within a framework of evolving institutions that empower rural communities to shape their own socio-economic development. Such key objectives are not reflected in indicators of result framework and a methodology has to be developed to include such indicators in the result framework. The VOs/Federations survival depends on the capabilities these institutions that have developed during the project period. The present arrangement is not clear and it is based on the formation of regional grouping, project milestones, etc. Proper guidelines have to be formulated to ensure the development and sustainability of these institutions. Therefore, formulation of project exit arrangements, development of milestones that lead to exit arrangements, identification of necessary and sufficient conditions for project to exist are essential to ensure long-term sustainability. In this exercise identification of partner organizations (private and public sector and NGOs) that can link for long-term programmes is important. 79 Annex 8. Comments of Cofinanciers and Other Partners/Stakeholders 80 Annex 9. List of Supporting Documents 1. Project Concept Note, February 25, 2003 2. Project Appraisal Document, March 5, 2004 3. Project Implementation Plan 4. Community Operational Manual 5. Project Preparation/appraisal mission documents 6. Back-to-office and aide-memoires 7. Implementation status reports, June 2004 – May 2010 8. Midterm review, September 2007 9. Country Assistance Strategy, April 2003 10. Project Appraisal Document for Second Community and Livelihood Improvement Project, August 2009 11. Borrower’s Implementation Completion Report, September 2010 12. Impact Assessment Study, October 2010 81 80° 81° SRI LANKA JAFFNA COMMUNITY DEVELOPMENT AND LIVELIHOOD IMPROVEMENT “GEMI DIRIYA� KILLINOCHCHI PROJECT MULLAITIVU SELECTED AREAS FOR PROJECT IMPLEMENTATION DURING FIRST PHASE: NORTH SRI LANKA BATCH 1 9° & EAST 9° VAVUNIYA MANNAR BATCH 2 & 3 DIVISION BOUNDARIES TRINCOMALEE DISTRICT BOUNDARIES ANURADHAPURA PROVINCE BOUNDARIES Mahiyanganaya NORTHCENTRAL PUTTALAM PALONNARUWA Rideemaliyadda Madulla 8° 8° NORTH Meega- BATTICALOA WESTERN hakuvula 0 10 20 30 40 50 Soran- athota Lunugala Medagama NORTH KILOMETERS KURUNEGALA MATALE Uva Paranagama Passara Passara & EAST Siyambalanduwa This map was produced by the Map Design Unit of The World Bank. Welimada CENTRAL The boundaries, colors, denominations KANDY and any other information shown on this map do not imply, on the part of AMPARA Haldummulla The World Bank Group, any judgment KEGALLE GAMPAHA on the legal status of any territory, or SAB A Buttala WEST any endorsement or acceptance of NUWARA BADULLA such boundaries. 7° ELIYA 7° Wellawaya U VA RA COLOMBO Thanamalvila MONARAGALA GA Thissamaharama ERN See project area M Kotapola Lunugam- RATNAPURA vehera detail at left UW Sooriyawewa Pasgoda KALUTARA Katuwana A Angunakola- Pitabeddara pelessa Hambantota Weeraketiya Baddegama Mulatiyana Okewela Hikkaduwa Yakkalamulla HAMBANTOTA Hakmana Tangalle Akmeemana GALLE Kadawathsatara Beliatta S O U T H E R N FEBRUARY 2004 Habaraduwa 0 10 20 30 40 50 Welipitiya Devin- IBRD 32943 uwara MATARA Weligama KILOMETERS 6° 6° 80° 81° 82°