Sri Lanka - Poverty Alleviation Project

Document of The World Bank FOR OFFICIAL USE ONLY Report No. 9316-CE STAFF APPRAISAL REPORT SRI LANKA POVERTY ALLEVIATION PROJECT APRIL 3, 1991 Population and Human Resources Division Country Department I Asia Region This document has a restricted distribution and may be used by recipients only in the performance of | their official duties. Its contents may not otherwise be disclosed without World Bank authorization. J  CURRENCY EQUIVALENTS Currency Unit - Sri Lanka Rupee (SL Rs) US$1.00 = SL RS 40.30 (November 1990) FISCAL YEAR January 1 - December 31 PRINCIPAL ABBREVIATIONS AND ACRONYMS USED AGA - Assistant Government Agent ARTEP - Asia Regional Team for Employment Promotion CFS - Consumer Finance Survey EDP - Entrepreneurship Development Program EPPU - Employment and Poverty Policy Unit FSP - Food Stamp Program CBSL - Central Bank of Sri Lanka CRBs - Cooperative Rural Banks CF - Credit Fund GDP - Gross Domestic Product GOSL - Government of Sri Lanka GRO - Grassroots Organization HRDF - Human Resource Development Fund IDA - International Development Association IFAD - International Fund for Agricultural Development ILO - International Labor Organization IRDP - Integratel Rural Development Program JSP - Janasaviya Program (People's Power Program) KFW - Kreditanstalt fur Wiederafbau MIS - Management Information System MTRID - Ministry of Textiles and Rural Industrial Development MPCS - Multi-Purpose Cooperative Societies MPPI - Ministry of Policy Planning and Implementation MYAS - Ministry of Youth Affairs and Sports NF - Nutrition Fund NGO - Non-Governmenital Organization NSB - National Savings Bank NYSCO - National Youth Services Cooperati.ve PNN - Praja Naya Niyamaka ("Loan Agents") PO - Partner Organization RDA - Recommended Daily Allowance RDS - Rural Development Society REDS - Rural Enterprise Development Service RRDB - Regional Rural Development Banks RWF - Rural Works Fund SAL - Structural Adjustment Lending SEEDS - Sarvodaya Economic Enterprise Development Service SK - Sahaya Kandayama (JSP "Support Team") TCCS - Thrift and Credit Cooperative Society Trust - Janasaviya Trust Fund UNDP - United Nations Development Program UNICEF - United Nations Children's Fund USAID - United States Agency for International Development  SRI LANKA FOR OFFICIAL USE ONLY POVERTY ALLEVIATION PROJECT Credit and Prolect Summary Borrowers The Democratic Socialist Republic of Sri Lanka Beneficiarys The Janasaviya Trust Fund Amounts SDR 40.6 million (US$<7.5 million equivalent) Termst Standard with 40 years maturity Onlending Termst The Goversment of Sri Lanka (GOSL) would onlend US$20 million to the Janasaviya Trust Fund for its credit operations, at 32 interest per annum with a 25-year repayment period, including five years' grace period for principal and interest. US$36.2 million would be passed on to the Janasaviya Trust Fund as a grant for the implementation of non-credit activities including human resource development, rural works and nutrition interventions, and US$1.3 million to the Ministry of Policy Planning and Implementation for creating policy research and program formulation capacity on poverty and employment issues. Proiect Description: The project will assist the newly created Janasaviya Trust Fund (the Trust), with a governing board having representatives from the Government, Non-Government Organization (NGOs), private sector and the academia, in financing credit operations, human resource and infrastructure development, and nutrition intervention activities of NGOs and government agencies. The Trust will manage four funds: (a) a Credit Fund to lend to partner organizations (POs) which will on-lend to the poor in a manner prescribed by the Trust at interest rates which will make the credit fund operations self-supporting; (b) a Human Resources Development Fund for promoting the productive use of credit and for developing the lending capacity of POs; (c) a Rural Works Fund for building economically viable infrastructure and creating wage employment; and (d) a Nutrition Fund for reducing the proportion of wasting and stunting in children and reducing the incidence of low birth weight and the prevalence of maternal malnutrition. In addition, the project will support an Employment and Poverty Policy Unit in the MaiLstry of Policy Planning and Implementation. Benefits and Risks: Benefits under the proposed project are expected fromt (a) enhancing the capacity of participating NGO8 and This document has a restricted distribution and may be used by recipients only in the performance of their official duties. Its contents may not otherwise be disclosed without World Bank authorization.  - Li - government agencies to increase their human resource development efforts and to provide access to credit for productive activities to the poor, about half of whom would be women; (b) creating some 10 million workdays of wage employment for the poor; (c) improving the targeting and containing the cost of welfare programs; (d) improving nutrition among children and women; and (e) creating capacity to analyze employment and poverty Impact of macro-economic and sectoral policies. Possible risks include political interforence in the working of the Trust and poor decision-making and sub-loan supervision by the participating organizations. To minimize these risks the project will: (a) ensure that the Trust's professionalism is maintained by using transparent eligibility criteria, high calibre staff and Board members and an NGO advisory committee to the Board; (b) assist participating organizations in observing strict credit discipline by helping build their credit mechanisms and capacity of ultimate borrowers to make effective use of credit; and (c) deny project resources to P0O which do not meet performance criteria. Pro1ect Costs: Local Foreign Total Project Components --- (ss million)------ Credit Activities 35.0 - 35.0 Human Resource Development Fund 13.1 1.4 14.5 Rural Works Ftmd 14.7 1.6 16.3 Nutrition Fund 13.0 1.2 14.2 Employment Policy Planning Unit 0.2 1.2 1.4 Project Management 3.6 - 3.6 Total Project Cost 1 85 Financing Plan: Government 17.5 - 17.5 Federal Republic of Germany 10.0 - 10.0 IDA 52.1 5.4 57.5 Total Estimated IDA Disbursements: FY 92 93 94 95 96 -------------(USS million)------------ Annual 6.4 9.4 12.1 13.9 15.7 Cumulative 6.4 15.8 27.9 41.8 57.5 Nlps IBRD 20879  - iii - SRI LANR POVERTY ALLEVIATION PROJECT Table of Contents Page No. Credit and Project Summary . . . . . . . . . . . . . . . . . . . , BASIC DATA . . .. . . . . . . . . . * * * . * * . . . . . vi I. INTRODUCTION . . . . . . . . . . . . . . . . . . . . . . I II. THE SECTORAL CONTEXT . . . . . . . . . . . . . . . . . . . . 2 A. Background . . . . . . . . . . . . . . . . . . . . . . 2 B. Dimensions of Poverty . . . . . . . . . . . . . . . . . 2 III. GOVERNMENT'S POVERTY AND EMPLOYMENT STRATEGY . . . . . . . . 5 A. Social Welfare Programs . . . . . . . . . . . . . ... 5 B. Integrated Rural Development Programs . . . . . . . . . 7 C. Approaches to Self-Employment and Microenterprise Development . . . . . . . . . . . . . . . . . . . . . 8 D. Access to Credit . . . .. . . . . . . . . .... . 11 IV. IDA'S POVERTY ALLEVIATION STRATEGY IN SRI LANKA . . . . . . 1S A. General . . . . . . . . . . . . . . . . . . . . . . . . iS B. Rationale for IDA Involvement . . . . . . . . . . . . . 17 C. Project Strategy . . . . . . . . . . . . . . . . . . . . 17 This report is based on the findings of an appraisal mission to Sri Lanka in October/November 1990. The appraisal mission members included Messrs./Mme. Om Nijhavan (Mission Leader), Ana M. Jeria (ASlPH), James Greene (ASTPH), Peter Bowden, and V. Rahardjo (Consultants). Mr. Wijaya Wickrema (ASICA) participated in the preappraisal mission and assisted with financial analysis. This report has been endorsed by Messrs. David Groves (Acting Director, ASlDR), and Martin Karcher (Chief, ASlPH) Peer Reviewers of the Project are Messrs. V. Raghavan (ASTIF) and T. Tsui who focused on the Credit Fund and Project Implementation and Management, respectively.  - iv - V. THE PROJECT . . . . . . . . . . . . . . . . . . . . . . . 18 A. Project Formulation . . . . . . . . . . . . . . . . . . 18 B. Project Beneficiaries and Relationship with Janasaviya Program . . . . . . . . . .. e . . e . . 18 C. Project Objectives and Scope . . . . . . . . . . . . . . 20 D. Project Description ......... ......... 20 E. Credit Fund Component . . . . . . . . . . . . . . . . . 20 F. The Human Resource Development Fund Component . . . . . 24 G. The Rural Works Fund Component . 25 H. Nutritio.a Fund Component . . . . . . . . . . . . . . . . 27 I. Technical Assistance Component . . . . . . . . . . . . . 28 VI. PROJECT COSTS AND FINANCING . . . . . . . . . . . .... 29 A. Cofinancing . . . . . . . . . . . . . . . . . . . . 31 VII. PROJECT MANAGEMENT AND IMPLEMENTATION . . . . . . . . . . . 31 A. Institutional Responsibilities . . . . * . . . * . . . . 31 B. The Janasaviya Trust Fund . . * . . . . e X . e . . . 31 C. Partner Organizations . . . . . . . . . . . . . . . . . 34 D. Administrative Procedures . .... . .. . .. ... 35 VIII. BENEFITS AND RISKS . . . . . . . . . . . . . . .. . . .... 38 A. Benefits . . . o . . . . . . . . . . . . . . . . . . . 38 B. Risks . . . . . . . . . . . . . . . . . 39 C. Programs of Special Emphasis . . . . . . . . . . . . . 40 IX. AGREEMENTS REACHED AND RECOMMENDATION . . . . . . . . . . . 40 TABLES IN TEXT 5.1 The Credit fund: Lending Estimates . . . . . . . . . . . . . 21 5.2 The Credit Fund: Financial Projections . . . . . . . . . . . 24 5.3 The Human Resource Development Funds Grant Estimates . . . . 25 5.4 The Rural Works Funds Grant Estimates . . . . . . . . . . . . 26 5.5 The Nutrition Fund: Grant Estimates . . . . . . . . . . . . . 27 6.1 Project Cost Summary . . . . . . . . . . . . . . . . . . . . . 30 6.2 Project Financing Plan . . . . . . . . . . . . . . . . . . . . 30 7.1 Procurement Arrangements . . . . . . . . . . . . . . . . . . . 36 7.2 Allocation of Disbursement of the IDA Credit . . . . . . . . . 37 ANNEXES 1. Operation of the Credit and Grant ?unds . . . . . . . . . . . 44 2. Eligibility Criteria for Sub-Loans to Beneficiaries from Partner Organizations . . . . . . . . . . 52  3. Deposit and Lending Rates . . . . . . . . . . . . . . . . . . 54 4. tinancial Projections for the Credit Fund Operations . . . . . 55 5. Technical Assistance to the Janasaviya Trust Fund . . . . . . 69 6. Employment and Poverty Policy Unit: Cost Estimates . . . . . 74 7. The Janasaviya Trust Funds Organization Structure and Responsibilities . . . . . . . . . . . . . . . . . . . . 75 8. The Janasaviya Trust Fund: Hedium-Term Implementation Plan . 87 9. Financial Simulation of a Partner Organization . . . . . . . . 90 10. Estimated Disbursement Schedule . . . . . . . . . . . . . . . 97 11. The Janasaviya Trust Funds Honitoring and Evaluation . . . . 98 12. Bank Supervision Input into Key Activities . . . . . . . . . 102 13. Assistance to Women . . . . ... . . . . . . . . . . 104 14. Selected Documents and Data Available in the Project File . . 108 MAP: IBRD 20879  SRI LANKA POVERTY ALLEVIATION PROJECT Basic Data AREA 65,606 sq.km. POPULATION a/ 16.6 million (mid-1988) Rate of growth: 1.4(1988) DENSITY 253 per sq.km. (1988) HEALTH 763 per sq.km. of cultivable land (1977) Population per physician (1988) 7,168 Population per nurse (1984) 1,290 POPULATION CHARACTERISTICS (1987) LIFE EXPECTANCY AT BIRTH (1988) Crude Birth Rate (per '000) 21.9 Male (Years) 68 Crude Death Rate (per '000) 5.9 Female (Years) 73 Infant Mortality (per '000 live births) 30.0 b/ ACCESS TO ELECTRICITY (1981) EDUCATION (1981) % of population - urban 48 Adult literacy rate 86.1% % of population - rural n.a. Primary School enroUment 98.0% ACCESS TO PIPED WATER (1980) INCOME DISTRIBUTION (1985/86) % of population - urban 47 % of household Income highest 20% 43.0 % of population - rural 40 % of household Income lowest 20% 4.8 UNEMPLOYMENT RATES BY SECTOR LABOR FORCE PARTICIPArON RATE 1978/79 1981/82 1985/86 1986/87 Male Female Island 14.7 11.7 14.1 18.0 C ensus of Population (1981) 49.8 19.4 Urban 20.7 14.2 19.5 17.2 LFS Survey 1980/81 53.1 21.2 Rural 14.6 12.0 13.2 15.9 LFS Survey 1985/86 52.7 25.4 Estate 5.6 5.0 7.8 9.6 ACCESS TO BANK CREDIT BY SIZE OF FIRM UNEMPLOYED POPULATION BY EDUCATION (1981/82) Number % of firms No Schooling lliterate 2.5 of with access No Schooling Literate 0.2 Emvlovees to credit Primary 14.3 1-5 6% Secondary 40.1 6-10 50% (CE (OL) 32.8 11/30 63% GCE (AL) 8.9 31 or+ 86% Undergraduates 0.4 Graduates 0.8 INCIDENCE OF POVERTY c/ Number of persons in poverty 5.16 miRion Number of households in poverty 896,000 (289%) % of households in poverty (rural) 32.8% % of households in poverty (urban) 12.0% % of households in poverty (estate) 13.5% a/ Registrar enerad's Department. b/ 1981. c/ The poor are deflned as those unable to allocate a level of food expenditures that would ensuro a calorie consumption level in the neighborhood of that recommended by WHO/PAO. To suit Sri Ianka conditions conditions and based on the structure of its population, average calorles per capita are calculated at 2,200.  -vii- Basic Data PROFILES OF POOR HOUSEHOLDS d/ HOUSEHOLD SIZE SEX OF MAIN INCOME EARNER Male FEmaol % distribu-don of % of total 81.20% 18.80% % of total Poor households (I to 4 members) 38% 24% (5 to 8 members) 55% 66% (9 to 104 members) 7% 10% EDUCATIONAL ATTAINMENT OF MAN INCOME- EARNE a a % within the % of totat povottv aroup DEPENDENCY RATIO e/ No schooling 10.3 14.8 The aveage dependency rato is 74% in Sri Lanka. Primay 37.2 49.1 The higher the incidence of poverty, the higher Secondary 48.5 35.5 the dWendency ratio. Passed Grade 12 2.4 0.4 Graduate 1.4 0.1 d/ Based on CFS 1986/87. e/ Ratio of children bdow 14 years to adul aged 15 to S9.  SRI LANKA POVERTY ALLEVIATION PROJECT I. INTRODUCTION 1.1 Sri Lanka is currently confronted by serious srcio-political and economic problems with high rates of unemployment and politically unacceptable poverty levels. Despite commendable achievements ln social indicators, e.g., high rates of life expectancy and literacy and low rates of infant mortality and population growth, malnutrition:. particularly among children, is out of line with other social indicators. Economic growth has slowed down in the past few years against the backdrop of unprecedented civil strife, exacerbating the poverty problem. The country faces a difficult task of generating productive employment and income earning opportunities for a work force which will continue for some timo to grow faster than the total population. Although the recently initiated adjustment process is expected to accelerate growth to en average of 5.2Z p.a. over the next three years, the poor are unlikely to benefit quickly enough from the adjustment process. 1.2 The government's poverty alleviation policies and programs which include food stamps, school mid-day meals, assistance to the d;sabled and moat rerently the Janasaviya Program (JSP), are essentially consumption-or'6tn:d welfare programs. They are financially and economically unsustainable and not well targeted. Production-oriented programs, such as micro-enterprise development and credit are fragmented and insignificant. Rural infrastructure is in a state of disrepair and is inadequate. Policy ard program responses therefore need to be broadened and better targeted to enable the poor to participate in the productive process by providing them access to resources for self-employment, designing productive rural infrastructure programs, and nutrition intervention. 1.3 After considerable dialogue with IDA, the Government of Sri Lanka (GOSL) has realized that tradltional welfare approaches to poverty are not sustainable and that there is an urgent need to reorient and expand poverty alleviation programs towards productive activities. GOSL has agreed with IDA that the total expenditure on welfare programs would be contained within 3% of GDP, and requested IDA's assistance in designing and financing a production- oriented poverty alleviation project. 1.4 This project is aimed at increasing income earning opportunities among the poor. It adopts a dual strategy -- a rural works program, designed to increase the availability of wage employment on economically viable projects, and a concurrent provision of credit and support services to those in the target group willing to engage themselves in self-employment and micro-enterprise development. Experience in other countries and evaluation of current activities in Sri Lanka indicate that, when effective mobilization and support for the poor is linked to credlt, self-employment and micro-enterprise development can make a noticeable impact on rural poverty. In addition, the project includes selected nutrition interventions to help reduce the incidence of low birth weight babies and child malnutrition, and technical assistance for creating policy research and program formulation capacity to take greater account of employment and poverty issues in overall growth policies and public investment projects.  II. THE SECTORAL CONTEXT A. Background 2.1 Sri Lanka is predominantly rural; of its 17 million people, about 70: live in rural areas -- 10.7 million living in some 30,000 villages and ond million in the estate sector. Agricu.ture continues to be the dominant activity, contributing 25% to GDP, 402 to export earnings and 52X to total employment. Of the 3.4 million acres of cropped area, 2 million acres are under major export crops of tea, rubber and coconut, and 1.4 miulion acres under paddy and other minor crops. While the importance of the three major export crcps has declined, that of the paddy sector has grown. 2.2 Structurally, there has been some transformation of the economy over the past 20 years, shifting from traditional to non-traditional sources of income in the key sectors. Most notable have been the large increase in paddy production, emergence of tourism and garments and shrimp exports. Nevertheless, the largest sector of the economy continues to be the service sector, reflecting the continuing dominance of the public sector (which grew from 472 of GDP in 1970 to 561 in 1984) with agriculture declining in relative importance from 281 to 23Z. Over the same period unemployment has crept up to over one million, with the growth of the labor-force outpacing that of employment opportunities. 2.3 Sri Lanka is unique in having achieved ar. impressive record in social and human resource development despite a low per capita income. By the mid- 1970s, Sri Lanka had achieved a literacy rate of 851, second only to Japan in Asia, and infant mortality and death rates lower than in many middle-income countries. This was, however, not matched by similar performance in economic growth. Conseque..tly, a change in development strategy since 1977, shifting emphasis from welfare orientation (in 1977 consumer subsidies accounted for over 202 of the government's recurrent expenditure) to growth with *reduced redistribution" produced an overall growth rate of 5-61 during 1977-87 which was twice that experienced during the previous decade. While this strategy was successful in increasing the growth rate and reducing unemployment somewhat, the benefits of this initial growth were not distributed broadly enough, contributing to ethnic, social and political strife. After 1986 growth started to decline to reach 22 in 1989 partly due to the uncompleted liberalization of 1977, the continuing dominant role of the public sector and civil conflict. This created serious internal and external imbalances, and exacerbated the employment and poverty problems. B. Dimensions of Poverty 2.4 Estimates of the incidence of poverty using the 1986/87 Consumer Finance Survey (CFS) data,3J based on food consumption related poverty line1t l/ The survey excludes zone 3, i.e., the north-east districts of Jaffna, Mannar, Vavuniya, Trincomalee and Batticaloa.  -3- indicate that nearly 282 of the households (900,000 households or nearly 5.2 miliion persons) appear to be in t-3?erty. This incidence drops to 24S when the food consumption poverty line is defined to allow calorie consumption between 802!' and 3002 of the norm. The percentage of the ultra-poor defined as those consuming less than 80Z of the recommended level although they allocate more than 80% of their total expenditure for food consumption is 4Z. 2.5 The magnitude of poverty varies across regions depending upon resource endowments, technological developirent, infrastructure facilities, market linkages, asset ownership, tenancy arrangements and institutional framework for resource support. Poverty is disproportionately concentrated in the rural areas. Eighty eight percent of the poor households are rural; 32.82 of the rural houaeholds, 13.52 in the estate sector and 122 in urban areas are in poverty. 2.6 Child Malnutrition. There is considerable evidence of malnutrition among children in Sri Lanka. A nutrition survey conducted by the Hinistry of Policy Planning and Implementation (MPPI) during 1980-82 indicated that 36Z of all pre-school children suffered from chronic malnutrition,!/ while 122 were found to be suffering from acute malnutrition. The survey also showed that low income households had relatively large proportions of malnourished children. Preliminary results from a recent (1988-89) nutrition survey seem to indicate that there has been an increase in acute malnutrition to 16.7S and a decrease in chronic malnutrition to 322. These results however, are not strictly comparable as the North and Eastern Provinces and some areas in the estate sector were not included in the latter survey. 2.7 Women and Poverty. Among the poor, women's income is crucia.,. Dry zone village studies suggest that the poorer the household, the greater the relative contribution of women to family income and expenditure. Among the landless and near landless households, the main source of income both for males and females is wage labor supplemented by own account production involving home gardening, some livestock rearing and household production of crafts. Women in these households tend to work longer hours than men because of time spent on 2/ The poor are defined as those who are unable to allocate a level of food expenditures that would ensure a calorie consumption level in the neighborhood of ..ie recommended energy intake for an active and healthy life. The daily calorie allowance recommended by WHO/FAO to suit Sri Lanka conditions is 2,200 calories on a per capita basis. The poverty line is the average consumption expenditure of those low-income households which obtain calorie consumption levels that fall within the adequacy range (defined as 80-1OOS of the recommended daily allowance (RDA). 3/ Eighty percent of the recommended level approximately equals 1.5 times the Basic Metabolic Rate and is considered the minimum critical energy requirement for a moderately active person. 4/ Chronic malnutrition results from low nutritional quality foods and affects children most during 3 months and 18 months of age.  domestic chores. They also tend to receive lower rates of remuneration, have even more restricted credit options than men and have almost no contact with extension and training systems. Data from micro-level case studies indicate that female-headed households which accounted for 172 of all households, particularly those with no adult child,en are amcng the poorest of the pior. Finally, women tend to earn their livelihood in the weaker sectors of the economy, which have been most adversely effected by liberalization. This is particularly true of cottage industry activities like handloom production and coir processing. 2.8 Poverty and Unemployment Overlap. An analysis of the 1986/87 CFS daza shows that the unemployed are not necessarily poor and the employed are not necessarily not poor. Nearly 902 of the poor are 'working poor". In 1986/87 252 of all unemployed lived in poor households; 852 of these lived in the poor households in the rural sector, 112 in urban areas and 42 in the estate sector. The unemployed in the poor households are generally less educated (primary education or high school drop-outs). The poor generally engage themselves in multiple activities of marginal productivity, and only earn meager incomes from irregular work opportunities. 2.9 Causes of Poverty. The poor tend to be *resource poor" with little or no ownership of productive assets such as land and equipment. Seventy percent of the country's 2.1 million rural households live in the wet zone which accounts for only 30Z of the cultivable area. Operational land holdings are extremely small; 382 of the tenants farm less than one-half acre and 712 less than one acre. Nearly 112 of the households are absolutely landless. Settlement schemes in the dry zone were intended to relieve this pressure but have proven inadequate. 2.10 The poor remain largely excluded from the development process because they lack control over productive resources and have limited access to credit, extension systems and other sources of information and assistance. Their survival strategies largely preclude them from organizing, raising capital and taking risks. The formal village level organizations tend to be led by the political elite of the village. Although formal leadership may on occasions nominally rest with the poor and the landless, the actual functioning of the rural organizations is controlled by the landed with close connections to politicians and influential government personnel. It is this group which controls decision-making and resource flows. Available rural resources are thus generally diverted to the better-off members of the community. 2.11 The web of social, political and economic constraints such as insecure tenancy arrangements, control over resources, indebtedness to the traders and money lenders, lack of organization, and the entrenched power structure forces the poor to engage in activities of marginal productivity and income, and prevents them from using their energies and abilities to engage in micro- enterprise development.  -5- III. GOVERNMENT' S POVERTY AND EMPLOYMENT STRATEGY 3.1 GOSL policies have tradicionally reflected a strong concern for the poor. Specific policy and program responses to poverty have essentially been welfare programs. These includes (i) food stamp schemes, public assistance to special categories (aged, disabled); (ii) social services (education, health, housing); and most recently (iii) the mid-day meal program for school children and the Janasaviya Program (JSP). For employment generation Government policies are primarily directed towards tourism development and large scale infrastructure development such as the Mahaweli development. In addition, the Government has also promoted Integrated Rural Development Programs (IRDPs), micro-enterprise and credit programs, which are not particularly targeted towards the poor. A. Social Welfare Programs 3.2 Food Stamp Program (MSP). In the 19609 a food (rice) subsidy was provided to the entire population, which became a drain on the national budget as the population grew and rice prices increased. As a result, in 1972 policy changed by withdrawing the rice entitlement from taxpayers and dependents (some 10? of population) and in 1978 by instituting an income criterion which reduced eligibility to 50X of the population. The scheme was further modified in 1979 by switching to food stamps (denominated in monetary terms) and bringing some food item prices more into line with world prices. 3.3 The late 1970s reforms attempted to reduce government intervention in the economy and gave the mar.et a larger role in determining prices and allocating resources. The change in the subsidy program was intended to protect low-income households from the effects of the removal of price subsidies. Only households whose declared income was less than a specified level were to receive food stamps, which they could use to buy basic foods made available at authorized shops at nonsubsidized prices. Although targeting attempts under the food stamp scheme restricted the transfers to only half of all households, not all the beneficiaries were in the lower half of the income range. Per capita calorie consumption of the bottom 20? declined about 8X from an already low 1,490 calories during 1978/79 to 1,368 calories during 1981/82. It appears that these households were not able to take advantage of the new income-earning opportunities created by the economic reforms during this period and that the food stamp scheme (hit by inflation) was not an effective safety net for protecting the most vulnerable households. 3.4 Social Services Programs have also been an important source of assistance, particularly in the areac of education, health and housing. However, a smaller proportion of resources reached the poorest 40? of households - given high drop-out rates from the education system and their limited access to housing loan schemes. 3.5 Janasaviya Program (JSP). The most recent Government initiative at poverty alleviation is the JSP which attempts to apply objective criteria for identification of the poor. The JSP was designed in 1989 as a crash program for poverty alleviation. It aimed at covering all households enrolled in the food stamp program, whose benefits had fallen in real terms over the years due to inflation. Under JSP each poor household is to receive a monthly grant of Rs  -6- 2,500 for a period of 24 months. The grant is in two parts - Rs 1,458 for the purchase of a specified basket of mainly foodstuffs, of which Rs 458 could be spent or saved with the National Savings Bank (NSB) according to individual household wishes. The balance of Rs 1,042 per month is to be deposited with NSB until it had accumulated after two years to the sum of Rs 25,000, at which point it should be made available as capital or as collateral for a loan to be invested in an income generating activity. 3.6 The first Round of JSP started in September 1989 in 28 of the country's 278 Assistant Government Agent (AGA) divisions. Originally in these divisions there were 224,000 families entitled to food stamps, from whom 164,000 were selected as JSP beneficiaries after two rounds of screening. This screening was conducted by the JSP support teams -- comprising four volunteers (Sahaya Kandayama - SK -, also called "internal change agent,") selected by the village and one publicly-paid village worker (the Grama Sevaka Nilandari, also called the "external change agent"). This support team is meant to organize the Saragam (community work), under which JSP recipient families are expected to provide 24 days of work per month, and mobilize JSP families to organize income-earning activities. 3.7 Originally conceived as a nation-wide "crash" program, JSP was unaffordable. Ai a result, it was phased out for implementation over 11 consecutive rounds. The second Round started in December 1990 in another 30 Divisions, and may include nearly 120,000 families drawn from families entitled to food stamps with an income of less than Rs 700 per month. In addition to it being unaffordable, JSP is a strong disincentive to work, providing -- as it does -- a monthly consumption grant of Rs 1,458 in a rural economy which pays an unskilled wage of some Rs 1,000 a month (when work is available). It also creates inequality between JSP recipients and those originally above the income cut-off-line, and between the early JSP beneficiaries and the later ones (assuming no indexing for inflation). 3.8 The implementation of the JSP is aimed at achieving self-sustaining development among the poor that is initiated and supported through their own efforts. However, an evaluation of the first Round of the JSP after nine months of operation of Round I reveals that despite huge expenditure and involvement of the entire machinery of the Government, the implementation of the JSP strategy still keeps 90% of the beneficiaries in poverty with very little chance of getting out of it, and in addition more exploited than before. Based on this experience the JSP is now being made production-oriented (see paras. 5.3-5.8). 3.9 The Mid-Day Meal Program was launched in the beginning of 1989 with a budgetary allocation of Rs 2 billion (US$50 million) to provide one meal a day to all children enrolled in primary and secondary schools in the country. In addition to being costly the program cannot achieve the nutritional improvements it aimed at since it fails to reach the population most nutritionally at risk, i.e., pre-school children. The program was eventually reduced to Rs 1 billion in 1990. 3.10 Nutrition Intervention. Current nutrition programs specifically directed at mothers and children focus on supplementary feeding with thriposha, combined with nutrition education, through the Ministry of Health and Women's Affairs. Thriposha is a formulated food produced locally at an MHWA factory with  sufficient capacity to more than meet the supplementation needs of Sri Lanka's vulnerable groups. However, Thriposha is distributed by Family Health Workers on a take-home basis with scant control over diversion to ineligible beneficiaries. Nutrition education through the health system is spotty and generally ineffective. 3.11 Under IDA's Economic Restructuring Credit (Cr. 2128-CE), GOSL is to review the entirety of its welfare programs with a view to improving targeting and reducing the budgetary cost to no more than 3Z of the GDP. In addition, the number of families in FSP is to be reduced to 1 million but with increased benefits, and the JSP is to be reformulated to improve its targeting and to make it more production-oriented. Beginning with Round II all JSP beneficiaries would be enrolled in production-oriented work program which includes rural works, land development, skills and entrepreneurship training for self-employment and micro- enterprise development to enable them to receive Rs .,458 per month JSP entitlement (see para. 5.3). B. Integrated Rural Development Programs (IRDPs) 3.12 During the past 10 years, GOSL has committed about US$200 millio-. to IRDPs covering 12 of the country's 25 districts as part of its strategy to decentralize development as a means for responding more effectively to the needs of the rural population and also reducing economic disparities among regions and groups within those regions. 3.13 IRDPs (which have been supported by a number of donors) have been effective in executing a number of small and geographically dispersed high quality investments in both directly productive and infrastructure development activities. The IRDPs have also promoted the formation of consultative and coordinating groups of beneficiaries down to the village level. However, an almost exclusive emphasis on agricultural production primarily through intensification, excluded the participation of the landless poor. While intensification of agricultural production could lead to increased labor absorption, increasing mechanization financed by subsidized credit has undermined this possibility. Both credit and extension services have focused on relatively large landowners. On the other hand, infrastructure development was able to provide employment of short duration to a very small proportion of the landless. In many cases, contractors hire outside permanent labor for rural works to work at different sites. Most of the benefits have thus accrued to relatively large landowners, creating larger disparities. The traditional rural power structure has continued to dominate and to influence the allocation of resources favoring the better off. 3.14 Rural works are an important means to provide employment opportunities and improve incomes of the poor. Mobilizing the poor into labor-intensive activities for the construction, rehabilitation and maintenance of infrastructural assets also contributes to production and development goals. To be durable solutions to seasonal unemployment or underemployment, these programs need to lay the ground for structural changes in resource utilization, implying substantial modification to administrative systems and procedures and integration of the projects undertaken under rural works schemes with local level and state- wide plans so that follow-up use and maintenance are ensured.  3.15 There is limited rural works implementation capacity outside of IRDPs and the line agencies. This capacity needs strengthening at the AGA divisional level, and additional capacity needs to be created within the NGOs and grassroot organizations. Experience shows that people are more willing to participate in implementation, operation and maintenance of infrastructure works, if they are given the opportunity to influence the earlier stages of identification and formulation. Local level organizations -- particularly NGOs and Grass Root Organizations GROs) -- need support to become more effective using specially trained change agents to organize people for this purpose. There are promising models of community participation in and control of rural works contracts -- largely in the urban sector -- but changes in contracting procedures are imperative. C. Approaches to Self-Employment and Mlcroenterprise Development 3.16 Profile of the Sector. In Sri Lanka micro-enterprises are defined as small income generating units employing less than five people (including self employed owrers). They include most household businesses and cottage industries. Such enterprises have little contact with formal training, credit or technology institutions. "Micro-enterprises" are largely non-farm enterprises. A large number of people earn income working in self-employment activities often as a supplement to agricultural wage labor. Estimates suggest that as much as 162 of manufacturing sector value added and 301 of industrial sector employment are created in the micro-enterprise sector. 3.17 A recent study1 of small enterprise programs found that 582 of the enterprise owners were unemployed prior to the commencement of the enterprises and another 312 were casual laborers. On average, each employed two persons per enterprise. The capital invested per job created was Re 1,800 in crop- agriculture (excluding the cost of land), Rs 3,200 for non-crop agriculture (primarily poultry and animal husbandry), Rs 5,000 for manufacturing and for service enterprises, and Rs 14,200 for trading. Fifty percent of enterprises earned incomes per worker equal to or higher than the equivalent wage labor rste. While some of the employment created is often low-paying, part-time and provides only supplemental household income, many trading, manufacturing, and some service projects have contributed siguificantly to family incomes. 3.18 Host of the production is sold in village and regional markets. Seventy nine percent of the enterprises reported some backward linkages and 171 forward linkages. USAID (1988)1' estimates on the basis of empirical investigations that a 1OZ increase in rural incomes would result in as much as a 211 increase in spending for locally produced goods and services. While households continue to be the majo' buyers of the products made by micro- enterprises, there has been a rel -e shift towards inter-industry and export 5/ ILO/Asian Regional Teams for Employment Promotion (ARTEP), "The Elusive Target" (1989). 6/ USAID, "Small Enterprise Promotion: Applyinr 'eneral Lessons to Sri Lanka" (1988).  demand. Thus many rural industries may be capable of responding to the changing pattern of demand associated with rising income levels. Subcontracting as a tool of enterprise development, however, is still in its infancy. 3.19 Entrepreneurship Development. In recent years, GOSL has given high priority to the problem of youth unemployment. Policy has repeatedly emphasized the need to create viable alternatives to wage employment through entrepreneurship development programs (EDPs). These have become the main thrust of the Government's micro-enterprise development strategy. Traditional approaches to micro-enterprise development have stressed the "enterprise' and its environments markets, raw materials, credit, technology, etc. In contrast, the EDP approach begins with the entrepreneur and prepares him or her for the challenge of going into self-employment or small business. 3.20 The best designed EDPs are based on practical training where trainees select an idea which is developed into a business plan. After the training, follow-up advisory services are provided by the implementing agency with some support extended to enhance access to credit institutions. The moot significant programs are those under the Ministry of Youth Affairs and Sports (MYAS) the Ministry of Textiles and Rural Industrial Development (MTRID), and the Ministry of Lands and Mahaweli Development. Recent ILO/UNDP a0sistance to the MYAS and the MTRID to launch the Integrated Entrepreneurship Development Program is an important step towards developing a systematic and coordinated approach to EDPs in Sri Lanka. The objective is to build up the government's capacity both to plan and implement integrated EDPs. Over the next three years an estimated 9,000 youth will receive four-day awareness training, out of whom 1,625 will be trained in a 3-5 week EDP. Folow-up extension services will be provided to an estimated 800 who are projected to start their own business. Government capacity for planning and implementing EDPs is however limited. 3.21 The MYAS program is implemented through its Small-Scale Enterprise Division, the National Youth Services Council and its cooperative banking affiliate, the National Youth Services Cooperative (NYSCO). The program aims to provide entrepreneurship development and business start up assistance (including access to bank credit) to unemployed, rural youth. Initially implemented through Youth Clubs, the program is being re-structured through the creation of youth enterprise clubs. Loans advanced to members to date total only US$750,000 by NYSCO and US$1.1 million by the People's Bank and the Bank of Ceylon to some 12,200 youth. Recovery on loans which carry a concessional interest rate of 152 is 802. 3.22 In addition, a number of programs are being offered (of varying quality) by different organizations. No less than 25 Ministries are involved in some kind of skill development, vocational or self-employment training. Many of these so-called self-employment schemes are little more than short 'awareness, courses and stop short of preparing the trainee to start a business. 3.23 Vocational and Skill Training. There are reported to be over 3,300 courses available in the public sector, supplemented by on-the-job training in the private sector and a number of other programs run by NGOs. No less than 19 ministries are involved with the four major ministries (Higher Education; Youth Affairs and Sports; Labor; Education) offering some 50,000 places. There is limited coordination among these programs, no clear overall policy for vocational  - 9 - demand. Thus many rural industries may be capable of responding to the changing pattern of demand associated with rising income levels. Subcontracting as a tool of enterprise development, however, is still in its infancy. 3.19 Entrepreneurship Development. In recent years, GOSL has given high priority to the problem of youth unemployment. Policy has repeatedly emphasized the need to create viable alternatives to wage employment through entrepreneurship development programs (EDPs). These have become the main thrust of the Government's micro-enterprise development strategy. Traditional approaches to micro-enterprise development have stressed the "enterprise" and its environment: markets, raw materials, credit, technology, etc. In contrast, the EDP approach begins with the entrepreneur and prepares him or her for the challenge of going into self-employment or small business. 3.20 The best designed EDPs are based on practical training where trs.inees select an idea which is developed into a business plan. After the training, follow-up advisory services are provided by the implementing agency with some support extended to enhance access to credit institutions. The most significant programs are those under the Ministry of Youth Affairs and Sports (MYAS) the Ministry of Textiles and Rural Industrial Development (MTRID), and the Ministry of Lands and Mahaweli Development. Recent ILO/UNDP assistance to the MYAS and the MTRID to launch the Integrated Entrepreneurship Development Program is an important step towards developing a systematic and coordinated approach to EDPs in Sri Lanka. The objective is to build up the government's capacity both to plan and implement integrated EDPs. Over the next three years an estimated 9,000 youth will receive four-day awareness training, out of whom 1,625 will be trained in a 3-5 week EDP. Folow-up extension services will be provided to an estimated 800 who are projected to start their own business. Government capacity for planning and implementing ED?s is however limited. 3.21 The MYAS program is implemented through its Small-Scale Enterprise Division, the National Youth Services Council and its cooperative banking affiliate, the National Youth Services Cooperative (NYSCO). The program aims to provide entrepreneurship development and business start up assistance (including access to bank credit) to unemployed, rural youth. Initially implemented through Youth Clubs, the program is being re-structured through the creation of youth enterprise clubs. Loans advanced to members to date total only US$750,000 by NYSCO and US$1.1 million by the People's Bank and the Bank of Ceylon to some 12,200 youth. Recovery on loans which carry a concessional interest rate of 15Z is 80Z. 3.22 In addition, a number of programs are being offered (of varying quality) by different organizations. No less than 25 Ministries are involved in some kind of skill development, vocational or self-employment training. Many of these so-called self-employment schemes are little more than short "awareness, courses and stop short of preparing the trainee to start a business. 3.23 Vocational and Skill Training. There are reported to be over 3,300 courses available in the public sector, supplemented by on-the-job training in the private sector and a number of other programs run by NGOs. No less than 19 ministries are involved with the four major ministries (Higher Education; Youth Affairs and Sports; Labor; Education) offering some 50,000 places. There is limited coordination among these programs, no clear overall policy for vocational  - 11 - D. Access to Credit 3.29 Sri Lanka has a long history of efforts to get the formal sector banking institutions, especially the publicly owned banks involved in lending in rural areas. Most of these efforts have been in response to Government pressure, resulting in subsidized interest rates, and because of the perceived grant nature of the loans by the recipient, to high rates of loan default and occasional outright loan forgiveness. The predominant role of the banks has been collection of rural savings for use elsewhere in the economy. As a result the volume of lending has been small and the impact on the poor minimal. Meanwhile a very large percentage of rural producers borrow from the informal credit markets often at very high interest rates. It is clear that their financial needs are not well served by the existing network of institutions and that Government intervention to control interest rates, provide refinancing, guarantees and forgive debt, far from being part of the solution, appears to be part of the problem. 3.30 Credit institutions in Sri Lanka can be classified into three broad types: formal, quasi-formal, and informal. Formal Institutions 3.31 Although commercial banks have a network of over 600 branches in Sri Lanka, the )anks have been constrained in their outreach to small farmers and producers by low profitability, crippling defaults and a corporate culture which is geared to larger commerce, industry and -- in agriculture -- to public estates and large plantation cultivation. Formal systems suffer from high admintatrative costs, inflexible standards of credit analysis, and insistence on security. Much of their rural credit has been developed in tandem with Central Bank of Sri Lanka (CBSL) rediscount and credit guarantee schemes, which have been used to subsidize interest rates to borrowers and depress lending margins. 3.32 Despite a proliferation of innovations and special schemes for the poor or unemployed youth (People's Bank and Bank of Ceylon have over 30), the volume of lending is small -- less than $5 million to date -- with recovery rates generally below 80?. Currently formal lending institutions cannot cover administrative costs and loan losses from the inadequate interest margins allowed particularly under CBSL refinance. 3.33 The large state banks have a mandate to participate in the JSP but are nervous about lending through it. They favor methods of intermediation through NGOs to reduce credit delivery costs and by-pass the problem of training and motivating field personnel. However, they have limited experience in dealing with NGOs and have been unable to formulate specific plans. The smaller private sector banks have only recently considered rural credit as a strategic objective. 3.34 Recently, links between the formal and the informal credit systems have been institutionalized in the Praia Naya Niyamaka schemes of the state banks, and the Sampath Seva Sanyojaka scheme of Sampath Bank, whereby nearly US$2 million has been extended by commercial bank branches to approved, private individuals and merchants acting as loan agents for on-lending to the rural poor. The scheme recognizes that a local agent has better knowledge of individual borrowers and their circumstances than formal lenders. The weakness of the  - 12 - agency system in Sri Lanka lies in the selection of the agents. A weak or poorly motivated agent limits his risk by lending to family and close personal contacts. 3.35 Regional Rural Development Banks (RRDBs). In 1985 the CBSL established the RRDBs to bridge the gap between rural branch bank services and informal credit sources. There are 12 RRDBs with 90 branches. Loan recovery is less than 70? and defaults appear to reflect a widespread attitude that RRDBEs are lending government funds and that these funds are only disguised grants. In addition to defaults, RRDBs have high overhead and administrative costs. RRDBs are constrained in cost-coverage by interest rate ceilings imposed by CBSL refinancing. Interest rates which depend on loan-purpose are typically in the range 12-182. 3.36 RRDBs lent US$4 millien in 1989. With assistance from the International Fund for Agricultural Development (IFAD), RRDBs have recently commenced lending to producer groups relying on peer pressure as collateral. In addition to taking banking to the village level, the RRDBes have also simplified lending procedures and made them more flexible. Credit volume under this scheme is small and performance has yet to be evaluated. Quasi Formal Institutions 3.37 The most significant quasi-formal institutions are the Thrift and Credit Cooperative Societies (TCCSs), the NGOs with their group-savings approach and the Cooperative Rural Banks (CRBs). 3.38 Thrift and Credit Cooperative Societies (TCCSs). The three-tier TCCS movement is composed of village societies, district unions and a national Federation and has close to 6,000-member societies with a total membership of over 560,000 persons. TCCSs are singular in purpose and maintain no formalized services outside of savings and credit. Total shares and savings have trebled in the past 10 years to US$14 million and total loans and advances to US$10 million. Studies have shown that TCCSs tend to be established in the larger, more prosperous villages and appeal to the better-off members of the community and that good management practices are not spread uniformly throughout the system. Despite these weaknesses, TCCSs are seen as a promising instrument to channel the flow of institutional credit to the rural sector. 3.39 Most TCCS members and almost all its borrowers are larger landowners with the cash to buy shares in the societies (lending is based on multiples of an individual's share value). Traditionally, the poorest households in a village were excluded from membership. As a result the Federation recently launched a pilot project funded by Swiss aid to develop small groups of the poor (defined as JSP recipients) within existing societies and extend loans for self-employment. To date, TCCSs have lent US$90,000 to 1,750 borrowers with a 93? on-time recovery. Borrowers are motivated and organized by change agent village workers employed by the District Union. Future plans incluie an ambitious expansion of the scheme to 6,000 societies and development of a cadre of 1,400 credit officers posted at AGA -- divisional level to supervise lending and recovery operations. 3.40 NGOs. The largest NGO savings and credit operation is the Sarvodaya- SEEDS started in 1986, which builds on the foundation of Sarvodaya's (an NGO) community development. SEEDS works with over 500 village societies, organizes  - 13 - savings and credit societies in villages and trains its members in financial management. The savings are re-lent to society members and also used to leverage loans from the SEEDS district level. A member is also required to have made a previous deposit of 102 of the requested loan amount to the village society which is refundable with interest at maturity of the loan. The village society guarantees all loans to its members. Loans are extended at 20% p.a. -- with SEEDS head office earning 52, the SEEDS district offices 102 and the societies 5Z. SEEDS has extended 11,600 loans worth US$500,000. 3.41 SEEDS projects the program to expand to 31,800 loans totaling US$7 million in 1994. SEEDS does not dictate conditions of eligibility for group membership; group members are responsible for selection. Members must have been saving for six months prior to receiving a loan, must have a feasible proposal, and the proposal must be approved by the group. SEEDS, like some of the smaller NGO programs, endeavors to provide some related marketing, training, and technical assistance to borrowers. 3.42 Currently, loans default are a major problem. As of December 1989, on-time recovery was only 65%. As a result, SEEDS has launched a concerted loan overdue recovery drive which is expected to recover most of the overdues, thereby increasing recoveries to 90X. SEEDS' management information system was re-vamped with technical assistance from the Bangladesh Rural Advancement Committee to give loan disbursement and recovery data within 30 days of month-ending (in comparison with six months previously). 3.43 Future plans include participation as a banking institution in an IFAD-CIDA financed "Small Farmer and Landless Credit Projects, development of a business extension service (the Rural Enterprise Development Service - REDS) and expansion of management training for both statf and beneficiaries (to be undertaken by the Sarvodaya Management Training Institute). 3.44 There are over 30 other NGOs and cooperatives with some experience in income-generating activities or small enterprise development. The preferred approach is group-based lending to small solidarity groups which also engage in savings programs. The creditworthiness of individuals is backed by solidarity and guaranteed by mutual savings. Although it is difficult to accurately cost credit support activities as distinct from community development activities, this approach appears to lower transactional costs for both the borrower and the lender. NGOs also benefit from grant funding and the use of volunteers which further reduce costs. The volume of lending is in the region of US$2 million a year involving loans to some 30,000 borrowers. Loans are taken for a wide variety of purposes decided upon by the borrower and include agricultural, non- farm and trading activities. Loan recovery varies by program but several promising programs can be identified where recovery is close to 100X5 in these schemes borrowers save regularly and programs charge rates of interest at or in excess of commercial bank rates. 3.45 Cooperative Rural Banks (CRBs). Created by the People's Bank in 1964 as the banking arm of the Multi-Purpose Cooperative Societies (MPCSs), the CRBs are a major collector of rural savings. However, their lending operation is constrained by: (i) staff unskilled in rural lending operationsS (ii) a conservative approach which emphasizes asset-based lending, essentially through pawn-broking; and (iii) insufficient margins over deposit rates of 12-13% p.a. to  - 14 - cover administration costs and lending losses. With the assistance of CARE, the CRBs have launched a pilot lending program to provide credit to up to 3,000 landless women organized into solidarity groups within 30 MPCSs. CARE has underwritten the program with a partial loan guarantee fund and provision of extension and training services to borrowers. Repayment on the initial 200 loans -- which carry an interest rate of 162 -- has been close to 10OZ. Overall the program has introduced the CRBs to the concept of group lending to the poor, linked credit with technical assistance, and given women from poor households experience of a rural banking institution. Future plans will depend on performance as the program scales up and on the CRBIs willingness to on-lend funds based on a net interest income of only 3-4Z in the absence of any increases in lending rates. Informal Sector 3.46 It has been estimated that the formal credit institutions supply less than 102 of the credit requirements of smallholders and micro-enterprises. Informal credit sources vary from professional money-lenders through landlords, shopkeepers and traders, to friends and relatives. The informal credit system is complex -- with a variety of interest rates and repayment terms. IFAD estimates that 302 of loans are extended free of interest -- from relatives and traders and shopkeepers. The latter then often (but not always) tie the borrowers into purchasing inputs/selling outputs at trader-determined rather than market-set rates. Interest rates where charged by moneylenders and traders/shopkeepers typically range from 602 to 1502 p.a. Such high rates absorb all or most of any production surplus. The chief positive feature of informal credit is ready access by the borrower. Most lending is unsecured and the borrower is rarely required to furnish documentation. 3.47 Women and Credit. There are no hard data on the proportion of women getting credit from formal sources or through NGOs (apart from special women-only programs). However, poor women are fearful of approaching banks and other formal credit societies. They often do not understand how to cope with the paperwork and how to meet the various criteria for loans. Furthermore, land is almost exclusively in the name of the husband, particularly in settlement areas. Women are therefore unable to offer any security in order to obtain a loan. Poor village women also do not have the contacts to provide banks with two or even one guarantor having salaried employment. 3.48 One problem many women face is that a husband may have taken a loan on which he defaulted. As a result, many women are not considered eligible because they belong to the defaulting household. There is a need to differentiatu the different members of a household and to pay particular attention to the special needs of women, an area where most credit programs are currently very weak. 3.49 Many women have sideline businesses and are quite active in organizing and forming credit groups in order to save for income generating activities. The most comm n form is the "Cheettu" or Chit fund. Most of the funds so generated are used for animal husbandry, petty trading and home gardening. The Vinivida grassroot organizations coalition in Puttalam runs eight different credit programs for women based on the cheettu concept ranging from a fund with a yearly cycle using monthly savings to give short-term consumption loans, through funds for latrines, to funds for income generating activities related to agriculture  - 15 - and handicrafts. The payback rate has been over 952 with loans given to individuals and to groups, and with the group taking collective responsibility for loan repayment. Significantly, these schemes state the interest rates in terms of monthly payments, not an annual rate of interest, which helps facilitate comparison with informal sector loans. 3.50 Lessons learned from the review of small producer credit in Sri Lanka indicate that the sustainability of programs designed to provide credit to the poor is weak as the pricing of credit is insufficient to cover cost of funds, administration and loan losses. However, group-based lending offers the prospects of using solidarity groups and peer pressure as a substitute for collateral. It also promises economies of scale in reaching credit-worthy groupe through intermediaries, which would be able to control delivery costs through community participation in borrower selection and project vetting. Given use of peer group pressure and group-based lending, recovery rates can be improved and interest rates charged to cover the costs of funding operations. 3.51 Experience also shows that credit programs are most effective where borrowers save (which the poor appear to be able and willing to do), giving borrowers a financial stake in the lending institution. 3.52 Capacity to organize the poor and facilitate access to resources is currently limited and needs to be built up in programs and institutions within and outside the government. Most promising are TCCSs, RRDBs, Sarvodaya-SEEDS, NYSCO and selected NGOs. As conventional bank training programs (under the CBSL and the Bank of Ceylon) are designed for traditional commercial banking, they do not appear appropriate to build credit management for group-based savings and lending programs; innovative approaches will have to be tried and developed. Promising training methods and materials are being developed and refined (such as under the Redd Barna savings and loans program) but considerable expansion of this capacity needs urgent priority. IV. IDA'S POVERTY ALLEVIATION STRATEGY IN SRI LANKA A. General 4.1 Over the past several decades, the level of IDA's assistance to Sri Lanka fluctuated largely depending on the rate of progress in the policy dialogue with the country. While IDA has always maintained a fairly large presence in Sri Lanka, its support was particularly significant during the major liberalization episode of 1977-83, when IDA's assistance to Sri Lanka incr-ased sharply. IDA projects at that time focused upon rehabilitation and expansion of the power ar.d transport sectors, as well as the agricultural sector, particularly to increase rice production, promote irrigated agricultural development (Hahaweli) and rehabilitate the tree crop subsector. Poverty alleviation was not a specific objective. 4.2 In the mid-1980s, when the country was facing ethnic problems, the Government's inability to pursue a vigorous economic reform program constrained IDA's country assistance in terms of lending support and in particular adjustment lending. More recently, the Government has undertaken a substantial structural  - 16 - reform program, which is being supported by IDA. Along with structural reforms, the Government is keen to address development issues which have aggravated the social and political tensions in the country. In particular it intends to reduce high unemployment levels and at the same time ensure that there is an improvement in the poverty situation. A combination of growth and general development objectives translates into a development agenda with three interrelated core policy objectives: (a) restoring macroeconomic stability; (b) reducing the size of the public sector and improving its efficiency while concurrently increasing the role of the private sector; and tc) reducing poverty. 4.3 In support of these broad objectives, IDA's current country assistance strategy focuses on a stabilization strategy to reduce both internal and external imbalances and to lay the basis for sustained economic geowth with due regard to poverty issues. The proposed lending program for Sri Lanka will continue to support the Government's macroeconomic reform program. In contrast to the past emphasis on lending to the power and major irrigation sectors, the future lending program would underpin the Government's priorities in agriculture, industry and infrastructure, with the objective of stimulating higher rates of growth, employment &ad incomes, as well as in the social sectors, including targeted poverty alleviation. 4.4 IDA's strategy in poverty alleviation has two main aspects. First, it would help to lay the foundation for revival of a dynamic, more broad-based growth which will generate increased employment opportunities and incomes. However, by itself an increased rate of growth is not sufficient to ensure that the lowest income groups share equitably in the benefits. This leads to the second aspect of IDA's strategy which recognizes the socio-political dimension of the poverty problem. 4.5 Sri Lanka has always had sizeable social programs and political pressures over the past few years led to the creation of several poorly conceived and costly new initiatives. These programs need to be cut back and retargeted towards the truly needy. Through its policy dialogue, IDA has been able to persuade the Government to impose fiscal limitations on total welfare expenditures and to start the process of retargeting individual programs. For instance, under the Economic Restructuring Credit (Cr. 2128-CE) the number of food stamps recipients is to be cut back by half and the JSP is to be restructured and made into an effective and sustainable anti-poverty initiative by making it production-oriented. The proposed Poverty Alleviation Project supports this latter objective. 4.6 It is also important to help the longer-term development of Sri Lanka's social infrastructure. In this regard the lending program incorporates a series of social sector projects. A future project in health care will support the Government's efforts to refocus Sri Lanka's medium-term expenditure program towards high priority health interventions, particularly in primary health care. In education, the strategy is to improve the general education system by upgrading schools and strengthening their supervision and management. At the same time IDA would aesist in laying the basis for more fundamental reform of curricula, examinations and post-school training to make secondary and higher education more cost effective and appropriate to the economy's needs.  - 17 - B. Rationale for IDA Involvement 4.7 This project is an important component of IDA's four-pronged country assistance strategy which includess (i) promotion of long-term efficient growth through adjustment lending by addressing internal and external imbalances, reditcing size of the public sector while improving its efficiency and increasing the role of the private sector; (ii) support to agriculture, industry, energy and infrastructure; (iii) rehabilitation and improvement of social services; and (iv) targeted poverty alleviation initiatives to facilitate the poor's participation in the productive process. Under the Economic Restructuring Credit, GOSL has agre d to reduce the number of families in the food stamp program by better targeting, and to make JSP more production-oriented, thereby containing total welfare expenditures within 32 of the GDP. In connection with restructuring of JSP, the Government requested IDA's assistance in designing and financing a production-oriented poverty alleviation project. IDA'o continuing dialogue with GOSL has led to a realization that welfsre expenditures must be contained and streamlined. The project will further help demonstrate the feasibility and sustainability of an alternative approach to poverty alleviation which is a high priority in Sri Lanka's development plans. The proposed project is consistent with IDA's areas of special emphasis and complementary to other IDA-supported projects which do not reach the very poor directly. This project would also support the Government's long-term growth policies by creating additional productive employment. C. Proiect Strategy 4.8 Opportunities for employment in the public and the organized private sectors cannot be expected to solve the problem of poverty. On the other hand, traditional welfarc approaches to poverty have not proven sustainable. There is. therefore, an urgent need to reorient programs of poverty alleviation towards more productive activities. Such a reorientation will require substantial efforts to build the capacity of institutions which organize and serve the poor. 4.9 The project seeks to create short-term employment through infrastructure development, long-term sustainable increases in income earning opportunities through maintenance of infrastructure, self-employment and micro- enterprise development, and improve nutritional status of poor children and pregnant and lactating mothers. The most effective strategy appears to be a more participatory approach based on group formatioi at the village level. This would enable the poor to organize themselves to gain access to necessary resources to undertake a range of income earning opportunities. Experience in Sri Lanka and other parts of South Asia has shown that an organizational structure with high level of staff commitment can effectively assist the poor. The poor and the unemployed generally cannot articulate their needs and have little power to command resources beyond their own labor, and the financial and service institutions are generally reluctant to cater to their needs. There is thus a potential role for intermediaries that can organize the poor and the unemployed and facilitate their use of resources and services. The role of such intermediation is now increasingly being recognized in Sri Lanka, and many NGOs, some financial institutions, and government programs have started to organize the poor for credit supported productive activities; but their capacity and scale of operations is still very limited, and they lack resources to expand, and need  - 18 - reorientation of their approaches. The project strategy emphasizes the reorientation and expansion of existing institutional capacity to serve the poor. 4.10 As there is currently no capacity within Sri Lanka to analyze the employment and poverty implications of macro-economic policy, sectoral programs and specific projects, it is also necessary to develop capacity within the government to undertake policy and operational research, and develop poverty and employment monitoring indicators. V. THE PROJECT A. Proiect Formulation 5.1 The proposed project's design is derived from sector work conducted in Sri Lanka by UNDP/IDA in collaboration with GOSL, and on IDA's policy discussions with GOSL. The sector work analyzed the employment and poverty issues, socio- economic relationships, approaches and institutions involved in self-employment, micro-enterprise development, credit, rural works and social mobilization. These reports recommend innovative approaches which would organize the poor, provide them access to resources, including appropriate training, and thus facilitate their integration into the productive process. B. Project Beneficiaries and Relationship with the Janasaviya PLogram (JSP) 5.2 The ultimate beneficiaries of this project would be families who fall below the poverty line,and malnourished children under the age of three, as well as pregnant and lactating mothers. Although in 1986/87 only 28Z ox iri Lanka's households (900,000) were poor (para 2.7), about one-half of all households currently receive food stamps as supplemental income. 'he food stamp program is now being rationalized to reduce the numbers from 1.9 m11.11-n to 1 million with improved benefits. The food stamp recipients will be further screened when joining the JSP in 11 Rounds. In Round I which started i.n ')ctober 1989 in 28 AGA divisions and is to last till September 1991, some 164,006 of the 224,000 food stamp recipient families were selected for JSP entitlement In Round II which uses more rigorous criteria, benefic:.ary selection will be n three phases. In the first phase 20 poorest families were selected in each haiwlet of the 22 AGA divisions following a community-based screening process in wtich all food stamp recipients participated with assistance from a support team. This phase selected 53,000 families from among some 164,000 food stamp recipiens. Subsequently, following the same procedure, another 10 families from each hiamlet were selected, making a total of 80,000 JSP beneficiaries. In some hamlets where there is a large concentration of the poor, more families may be identified in a third phase. The expectation is that at any given time during Round II, the total number of beneficiaries would not exceed 80-90,000. The third phase will be undertaken only after the beneficiaries selected in the first two phases have been organized and started participating in productive activities. 5.3 In contrast to Round I, Round II ia being made production-oriented; all Round II beneficiaries are to be enrolled in a production-oriented work or training program, which includes civil works, land development, ski-ls and entrepreneurship training for self-employment and micro-enterprise development,  - 19 - to be eligible to receive the Rs 1,458 per month JSP entitlement, consisting of Rs 1,000 in JSP stamps to buy specified consumer goods and Rs 458 to be deposited in an individuals savings account. For those beneficiaries who choose to work on civil works, the JSP entitlement would be in lieu of wages that they would otherwise earn. Others would enroll themselves in a training program and may be given credit when ready to undertake self-employment activities. This nexus between payment of Ru 1,458 and participation in a productive activity would be achieved with assistance from the newly created Janasaviya Trust Fund-the Trust- through the partner organizations (para. 5.11) in accordance with the conceptual framework, methodology and discipline of this project. 5.4 In order to facilitate the production orientation of this program, a team from the Department of Statistics will develop the socio-economic profile of the JSP beneficiaries. In the process those found with incomes above the JSP eligibility level (about Ru 700 per month) will be dropped from the program, streamlining the selection process further. Based on the experience of Round II, the beneficiary selection process for subsequent Rounds will be further refined. 5.5 GOSL would also put in place an exit mechanism, so thats (a) those refusing to participate in a productive activity would be dropped from the program immediately and become eligible for food stamps only; (b) beneficiaries would graduate from the program within the two-year JSP entitlement period as soon as the family's income level reaches Rs 1,500 per month on a sustained basis. 5.6 GOSL would further require that, until it reaches a level of Rs 1,500 per month, the income generated from self-employment and micro-enterprise development be placed in the beneficiary's savings account which would in the first instance be used to repay the loan taken for engaging in the self- employment activity, and the balance would be used for further investment. Alternatively, the additional income generated from self-employment and micro- enterprise development should enable the family to invest an equivalent amount out of the JSP entitlement in a savings account, thus accumulating savings for further investment. 5.7 During negotiations GOSL reconfirmed the above understandings reached earlier and indicated that operational guidelines for making the Round II production-oriented, including the exit mechanism and the provisions of para. 5.6 above, will be published by March 31, 1991 for use of operating staff and beneficiaries. 5.8 With the significant involvement of the Trust in training and credit for self-employment and micro-enterprise development it is expected that the investment component of the JSP (Rs 25,000 at thet end of the two years) would become redundant (para. 3.5). 5.9 Round II is considered a transition phase which establishes a clear link between JSP entitlement, productive activities and the Trust. It is to be closely monitored by MPPI and the Trust with a view to improving implementation methodology. Implementation of Round II is assisted by four UNDP experts  - 20 - attached to the Trust. As new food stamp beneficiaries join future JSP Rounds III-XI, they will be dropped from the Food Stamp Program and submitted to the production-orientation discipline of this project. In this way the proposed project would complement the restructured Food Stamp/Janasaviya Programs, facilitating the poors' participation in the productive process, and become the poverty alleviation program of the Government. C. Project Objectives and Scope 5.10 The major objectives of the project are to increase income earning opportunities among the poor, and to improve the nutrition status of children under three years of age and pregnant and lactating mothers. To achieve these goals, the project has the following specific objectivess Ca) reorient and expand the existing institutional capacity to serve the poor, and to create additional capacity; (b) develop credit and other services for promoting self-employment and micro-enterprise development through group-based lending and entrepreneurial development; (c) expand productive wage employment for the poor through technically, economically and socially viable rural works projects; (d) develop programs for nutritional interventions for malnourished children and pregnant and nursing mothers; and 'e) create policy research and program formulation capacity to take greater account of issues of poverty and unemployment in overall growth policies and public investment projects. D. Proiect Description 5.11 The project will assist the Trust in financing credit operations, human resource and infrastructure development, and nutrition intervention activities of NGOs and government agencies (called Partner Organizations). The Trust will manage four components of the project: Ca) a Credit Fund (CF) to lend to partner organizations (POs), which will on-lend to the poor in a manner prescribed by the Trust at interest rates which will make the credit fund operations self-supporting; (b) a Human Resources Development Fund (HRDF) for promoting the productive use of credit and for developing the lending capacity of P0; (c) a Rural Works Fund (RWF) for building economically viable infrastructure and creating wage employment; and (d) a Nutrition Fund (NF) for reducing wasting and stunting in children and for reducing incidence of low birth weight and the prevalence of maternal malnutrition. In addition, the project will finance technical assistance for an Employment and Poverty Policy Unit in the MPPI. E. Credit Fund Component Credit Demand 5.12 Credit would be available to the partner organizations for on-lending to the poor for productive purposes. The ultimate beneficiaries would need to  - 21 - complete a preparatory phase of human resource development to achieve minimum standards of creditworthiness. The size of the fund (US$30 million) is based on an analysis of the potential demand, over a five-year period, made by the prospective POs and on an assessment of their capacity to deliver such credit. The proposed IDA project would provide US$20 million, and an additional DM 15 million (US$10 million equivalent) will be provided by the Federal Republic of Germany through KFW as a grant. During negotiations agreement was reached that GOSL would contribute US$5 million to the Trust's Credit Risk Fund -- US$3 million before the IDA Credit becomes effective and US$2 million before December 1, 1992. (para 9.1(a)). Credit Fund disbursement estimates by POs are summarized below: Table 5.1s The Credit Funds Lending Estimates (Project Year and US$ Million) 1 2 3 4 5 Total RRDBs 0.8 1.2 1.7 2.4 2.5 8.7 NYSCO 0.3 0.9 1.0 1.7 2.1 6.0 Sarvodaya-SEEDS 0.3 0.6 0.9 1.1 1.0 3.9 TCCSs 0.8 1.5 1.5 1.5 1.8 7.1 Commercial Banks 0.7 0.8 0.8 0.8 0.8 3.9 Other NGOs 0.0 0.0 0.1 0.1 0.2 0.4 Credit Fund 2.9 5.1 6.0 7.6 8.4 30.0 Credit Risk Fund 3.0 2.0 0.0 0.0 0.0 5.0 5.13 Onlending Terms to the Trust. IDA allocations to the Credit Fund will be passed on to the Trust under a subsidiary loan agreement at 32 repayable in 25 years with a five-year grace period for principal and interest on a fixed amortization schedule, and GOSL will bear the foreign exchange risk (para. 9.1(b)). A draft of the Subsidiary Loan Agreement was agreed in substav.e during negotiations; and the signing of this agreement, satisfactory to IDA, between GOSL and the Trust is a condition of effectiveness of Credit (para. 9.4(b)). The amount retained in the Trust's Credit Fund arising from the faster maturity of loans given by the CF to POs compared with the Trust's repayment obligations to GOSL can be used only for lending to P0. Credit Terms and Conditions 5.14 PO Participation Criteria. Participation by P0 in the Project would be carefully monitored by the Trust, and POs subjected to uniform eligibility criteria accordingly. During negotiations agreement was reached with the Trust that it would use the following eligibility criteria (para. 9.2(a)). (a) POs should have at least two years' experience in lending to the target population;  - 22 - (b) the loan recovery rate of the PO during the two years proceeding loan application should be at least 70Z of the amount (principal and interest) falling due; (c) P0 would have an adequate spread on existing and proposed operations to cover operating costs and risks; and (d) POs must have or establish an accounting and financial reporting system acceptable to the Trust, and agree to have the accounts audited annually by an independent auditor. 5.15 In evaluating the eligibility of POs, the Trust would also take into account such other non-financial criteria as managerial capacity, organization and staffing to implement the project, and a demonstrated ability in eocial mobilization at grassroots level. 5.16 Additionally it was agreed with GOSL and the Trust that eligibility for continued access to credit would be dependent on the ongoing financial performance of the PO including, inter alia, (i) an adequate coverage of operating cost, (ii) net margin sufficient to meet debt servicing requirements and to build up reserves against risk, (iii) subloan recovery of principal and interest of not less than 902 from the subborrowers, and (iv) on-time repayment to the Fund (para. 9.2(b)). 5.17 Onlending Terms to POs. The Trust will onlend IDA funds to P0 at not less than 71 p.a. This rate is designed to provide an adequate margin to P0, given the agreed rate to be charged to the ultimate beneficiaries (para. 5.22), and the high operating cost of extending credit to the poor in small amounts. Although the onlending rate to the POs is below the average cost of non-demand deposits to the commercial banks (presently about 12.52 p.a.), the special nature of the project's credit operations justifies this rate. During negotiations agreement was reached that IDA and GOSL would review the onlending rates with the Trust semi-annually in the light of the P0 operating and risk costs and make appropriate adjustment. The maturities of loans to POs would be based on the composite maturities of subloans from POs to the ultimate beneficiaries but would not exceed three years with six months grace period although the majority of subloans during the initial period are not expected to exceed 12 months (para. 9.2(c)). 5.18 Lending Terms to Ultimate Beneficiaries. The onlending terms and conditions to the ultimate beneficiaries would be determined by individual POs using prudent lending policies and practices (Annex 2). However, POs must enroll all sub-borrowers in a savings program at the time they receive credit and through the period for which they are indebted. The Trust will review these policies and practices periodically as to their adequacy to maintain the financial integrity of the PO. If they are found to be lacking, the Trust would seek their revision to comply with the Trust's eligibility criteria or discontinue lending to the PO concerned. 5.19 Subloans by PO to ultimate beneficiaries would finance both investment and working capital expenditure. During negotiations agreement was reached with GOSL/Trust that POs would limit subloans to a maximum of Rs 40,000 (US$1,000 equivalent). Total loans of a PO to a subborrower would not exceed the subloan  - 23 - limit of Rs 40,000 (US$1,000 equivalent) at any given time. Rescheduling would be considered only once for a maximum period of two years in extreme circumstances such as natural disasters or loss of income earning assets. Subloan maturity would be determined by (i) the projected cash flow from the productive activities financed, (ii) the availability of cash surplus to pay interest and principal, and (iii) the subborrowers subsistence needs, but with an overall ceiling of three years (para. 9.2(d)). 5.20 For the Trust to monitor the performance of the project financed subloans and ensure that they are directed to the target population, POs would be required to maintain separate project ac^1unts, and prepare an annual income statement, source and utilization of funds statement and a balance sheet for project financed activities. These statements would be certified by an independent auditor acceptable to the Trust (para. 9.2(e)). 5.21 Most of the prospective POs are currently onlending at rates varying from 122 to 18Z. Most agricultural credit is subject to the CBSL refinancing schemes and is onlent at 12Z to the ultimate subborrowers. This involves a considerable subsidy, given the high cost and risk of rural credit and commercial banks' weighted average lending rate of about 18.52. Subsidized credit undermines the viability of rural credit, and mostly reaches the better-off members of the society. Lending rates prevalent in Sri Lanka are given in Annex 3. 5.22 POs would fix onlending rates to the ultimate beneficiaries at levels which would be sufficient to cover their operating costs and possible loan lcsses. Given the high operating costs and risks associated with this type of lending, the POs would be expected to have a minimum spread of at least 122 per annum. The Trust would ensure that the initial onlending rate on subloans are sufficient to recover cost of funds, operating costs, risk costs and provide a margin to build up a prudent cash reserve against future loan losses. During negotiations agreement was reached with GOSL/Trust that POs onlending rate would not be less than the average weighted interest rate (for all lending operations) charged by the Commercial Banks, Development Finance Company of Ceylon, National Development Bank of Sri Lanka, Regional Rural Development Banks, Credit Societies and Cooperatives, as published semi-annually by the Central Bank of Sri Lanka. The Trust would review rates determined by the POs semi-annually, and agree on a rate to be used for subloans for the ensuing six months (para. 9.2(f)). However, the Trust should allow flexibility to POs to rebate those subborrowers who pay their subloans on time provided the full cost recovery criteria described above is met. Credit Fund Sustainability 5.23 The credit operations of the CF would be structured to achieve at minimum a break-even position after allowing for cost of funds, operating costs and risk. In order to test the viability of the Credit Fund and its sustainability, financial projections, covering the five-year disbursement period, have been prepared. The key assumptions are that the Trust will receive loan funds at an interest rate of 32, and onlend them at 72, yielding a spread of 42. GOSL will contribute US$5 million to a Credit Risk Fund which would be invested in Government Securities at a yield of no less than 102, the loan recovery ratio after rescheduling would average around 95Z of disbursements, and  - 24 - 25Z of loan amounts falling due would be rescheduled. In addition, the model assumes that 52 of the recoveries would be accumulated in a cash fund to permit the Trust to meet its liquidity requirements. For the sake of simplicity the model assumes that all loans are outstanding for a full year. 5.24 These assumptions and the result they produce for the Trust are detailed on Annex 4 and summarized belows Table 5.2: The Credit Funds Financial Projections (Project year and US$ Million) Year 1 2 3 4 5 Credit Fund Loans 2.9 5.1 6.0 7.6 8.4 Total Loans 2.9 7.2 11.4 16.7 22.0 Rescheduled Loans Outstanding 0.7 2.2 3.7 5.6 7.6 Bad Debts Written-Off 0.0 0.1 0.2 0.5 0.9 Net Income/(loss) from Credit Fund Operations 0.2 0.3 0.2 0.1 (0.1) Net Income/(Loss) 0.4 0.7 0.7 0.7 0.5 Credit Risk Fund 3.0 5.3 5.8 6.4 7.1 Liquidity/Security Fund 0.1 0.4 0.9 1.6 2.5 Total Assets 6.4 14.3 21.4 30.3 40.0 Net Worth 3.4 6.1 6.8 7.5 8.1 5.25 The above data show that provided the loan default rate is less than 5Z and the Credit Risk Fund and the Liquidity/Security Fund continue to earn an income of no less than 102, the Credit Fund will continue to be financially viable and could sustain its operations after the credit period. A sensitivity analysis shows that at a default rate of 1OX, an investment income of at least 152 of the above funds would be required to maintain a break-even position. F. The Human Resource Development Fund (HRDF) Component 5.26 The HRDF will provide the initial costs of human resource and related institutional development. The first includes the process of group formation and awareness building at the beneficiary level, and beneficiary training in technical, micro-enterprise development and leadership skills. As the beneficiaries would be JSP recipients, they would not be paid any stipend during the training period. Institutional development includes the strengthening of the conceptual and management skills of partner organizations which operate credit, rural works and nutrition programs for the target population. These costs are treated as social costs and would be eligible for grant support during the initial period of program operation. As the credit programs develop and expand, the recurrent costs of the credit program would be met from net interest income on lending. 5.27 Estimates of HRDF contained in Table 5.3 are divided into two main components. A basic fund (to support social mobilization, training and selected administrative costs); and capital cost support. The size of the basic fund is estimated by analyzing training costs estimated by the prospective P0 and on the  - 25 - assessment of their capacity to undertake effective training. Support for the incremental capital expenditure (such as motorcycles and computer equipment) is estimated at 25Z of the basic fund. Table 5.3: The Human Resource Development Funds Grant Estimates (Project Year and USS Million) Year 1 2 3 4 5 Total Basic Fund 2.3 2.6 2.6 2.1 2.0 11.6 Capital Cost Support 0.6 0.6 0.7 0.5 0.5 2.9 HRDF 2.9 3.2 3.3 2.6 2.5 14.5 5.28 The HRDF will finance programs of awareness creation and organization of the poor, skills and entrepreneurship development, and assistance in the identification of productive activities, and rural works projects. The training to the partner organizations staff would involve group formation, credit management, MIS, accounting and auditing. 5.29 Eligibility Criteria. During negotiations, agreement was reached with GOSL/Trust that POs receiving HRDF grant support would have at least one year experience in organizing the poor. Additionally, POs would agree to a medium- term development plan linking human resource development with the use of credit and other Trust activities. Continued access to the HRDF would be compliance with previously agreed activities (para. 9.2(g)). G. The Rural Works Fund Component 5.30 Objective. The main objective of the Rural Works Fund is to provide wage employment opportunities to the poor and to create productive assets, including rehabilitation and maintenance of rural infrastructure. The proposed program is expected to maximize the long-term sustainable income generation capacity of the poor through selection of projects that generate long-term income, and/or on-the-job training in construction skills. 5.31 The rural works program would operate mainly in the South and Central parts of the country outside the major metropolitan and estate areas. In the wet zone, the watershed protection and soil conservation/social forestry component is especially relevant in order to protect fragmented and fragile land resources and to reduce the negative downstream effects from upland erosion. As the dense road network in this zone is affected by heavy rainfall and lack of regular maintenance, road rehabilitation not already covered by ongoing rehabilitation projects would be given high priority. Irrigation maintenance and rehabilitation is especially relevant to the dry zone area with its seasonal rainfall pattern. Because of the scarcity of fuelwood and timber in these areas, and also because of the need to protect the tank catchments, community forestry would be an important project activity. Village sanitation, water supply small town infrastructure and fishpond construction would cut across all zones.  - 26 - 5.32 The Rural Works Fund will finance projects that will be generated and submitted by potential POs. It would be implemented to a large extent through government agencies particularly at the AGA division level. The AGA's will be encouraged to work closely with community groups in identifying eligible projects. Interested NGOs would also be encouraged to develop programs and as appropriate they will be assisted through HRDF in developing planning and implementation capacities. The Trust will establish a mechanism to aggregate small proposals that the Trust will not be able to appraise and handle individually, through intermediary organizations. 5.33 Eligibility Criteria. While the Trust would establish sector-specific criteria, the following general eligibility criteria for RWF projects were agreed during negotiations (para. 9.2(g)): (a) projects should generate substantial employment and use local labor to the extent possible. The labor component should be in excess of 50Z of the total contract cost (compared to 30Z in current works); wage rates would be consistent with those prevailing in the project area, and JSP entitlement in JSP areas would be paid in lieu of wages that they would otherwise earn; (b) projects should generate long-term socialleconomic benefits accruing directly to the poor; (c) projects should build sustainable income earning capacity of the poor; (d) upgraded infrastructure should be maintained on a regular basis either by the responsible agencies or by the users themselves. Beneficiary participation in resource management and community contracting should be favored; and (e) projects must meet technical requirements in each sub-sector. 5.34 Estimates of Rural Works Fund activities are summarized below: Table 5.4: The Rural Works Fund: Grant Estimates (Project Year and US$ Million) 1 2 3 4 5 Total Irrigation 0.4 0.6 0.9 0.9 0.9 3.7 Roads 0.4 0.6 1.2 1.6 1.8 5.6 Forestry & Watershed Management 0.1 0.2 0.4 0.6 0.6 1.9 Water Supply & Sanitation 0.2 0.3 0.4 0.3 0.4 1.6 Fisheries 0.1 0.1 0.1 0.1 0.1 0.5 Small Town Infrastructure 0.2 0.3 0.5 0.5 0.5 2.0 Feasibility Studies 0.2 0.2 0.3 0.3 - 1.0 Rural Works Fund 1.6 2.3 3.8 4.3 4.3 16.3  - 27 - H. Nutrition Fund Component 5.35 Obiective. The Nutrition Fund will support the innovative intervention approach being tested in Galewala AGA Division, and extend it to 180 AGA Divisions through Sri Lanka's 300-odd NGOs. The Trust would set up a four- person Nutrition Division to organize the component, screen NGO applications, develop NGO worker and supervisor training programs and monitor and evaluate the results. 5.36 The objectives of the component would be, after five years of operation in an AGA division, to: (a) reduce moderate and severe malnutrition in children under three years of age, thereby eliminating severe clinical malnutrition as a public health problem, and reducing (i) by 75Z the frequency of wasting in children aged 12-23 months, and (ii) by 30? the prevalence of stunting in children aged 24-36 months; and (b) reduce by 252 the incidence of low birth weight and the prevalence of maternal malnutrition and iron deficiency anemia. 5.37 The Fund would finance training for NGO personnel atd the costs of monitoring, evaluation and special studies. It also would finance grants to NGOs for: scales, utensils and equipment for volunteers; and other program operating costs. Honoraria and salaries, respectively, for volunteers and supervisors (if they are JSP recipients) would be adjusted from their individual JSP entitlement. 5.38 Because the availability of adequate quantities of thriposha is central to the program's success, agreement was reached at negotiations that the Government would maintain thriposha production levels of 870 metric tons monthly during project implementation and would allocate to the Trust as much of that offtake as the nutrition component would require. To promote appropriate linkages between NGO nutrition and Goveranment health personnel, agreement was reached with GOSL at negotiations that, in each division where the component operates, the AGA would chair quarterly meetings of a division nutrition-health coordinating committee to be set up under terms of reference and with representation satisfactory to IDA (para. 9.1(c)). 5.39 Phasing of coverage and grants to NGOs from the nutrition fund are estimated in Table below: Table 5.5: The Nutrition Fund: Grant Estimates (Project Year and US$ millions) Year 1 2 3 4 5 Total No. of AGA Divisions 15 40 70 120 180 180 Equipment Cost 0.2 0.4 0.6 0.9 1.3 3.4 Training 0.3 0.5 0.6 1.0 1.3 3.7 Operating Cost, including consumables 0.2 0.6 1.0 1.9 3.4 7.1 0.7 1.5 2.2 3.8 6.0 14.2  - 28 - I. Technical Assistance Component 5.40 The Janasaviya Trust Fund. The establishment of the Janasaviya Trust and its effective operation require considerable technical assistance (TA). Assistance for start-up consisting of developing operating level policies and establishing systems and procedures has been provided under a UNDP facility (Annex 5). Further assistance to the Trust will be required at two levels: (i) for the Trust's management systems, and (ii) for the Trust's partner organizations. 5.41 The Trust would need assistance in establishing data processing systems necessary for modern enterprises: these include financial modeling expertise, funds forecasting, credit approval, monitorLng procedures and management information systems. The design and early implementation of a computerized management information system would be of considerable benefit to the Trust. 5.42 The Trust would also need general assistance during the project period depending on the iscues that arise and the steps that need to be taken to strengthen operating pergormance. 5.43 Technical assistance by the Trust to the partner organizations in the first instance would focus on social mobilization, credit management, accounting procedures and monitoring performance. Many NGOs and government agencies are particularly weak in those areas. Some of the larger organizations will need help in establishing computerized accounting and control. 5.44 In addition, the Trust would assist P0 in identifying productive income generation opportunities, linking of small rural producers with wider local and international markets, and in other ways as identified by the Trust's evaluation of experience of its partner organizations. 5.45 UNDP would provide this assistance in continuation of its previous involvement with the Trust. Signing of UNDP project document, providing technical assistance to the Trust is a condition of effectiveness of IDA credit (para. 9.4(c)). 5.46 Employment and Poverty Policy Unit. There is currently little capacity within Sri Lanka to analyze the employment and poverty implications of macro-economic policy, sectoral programs or specific projects. Nor are the employment and monitoring indicators readily available in a timely fashion. The project would establish an "Employment and Poverty Policy Unit' (EPPU) within MPPI to develop monitoring indicators and to undertake policy and operational research and analysis itself in addition to commissioning special studies by outside agencies. This unit would act as a focal point for policy planning, coordination and evaluation of employment activities. The main objectives of the unit would be to: (a) develop methodologies for analyzing and monitoring, on a continuing basis, the employment and poverty effect of programs in different sectors of the economy;  - 29 - (b) evaluate the effectiveness of programs and projects aimed at income generation for the poor; (c) develop operational guidelines for employment and poverty strategies and programs at macro and sectoral levels; and (d) recommend changes necessary to make existing or proposed policies more effective in employment generation and poverty alleviation. 5.47 The unit would span two existing divisions -- the Janasaviya and Nutrition Division and the Human Resources and Employment Division in the MPPI and would be coordinated by the Secretary, planning. The unit would be assisted by two advisors skilled in analyzing employment and poverty issues. 5.48 The unit will also develop capacity for small enterprise policy planning. The first phase would involve the development of a data base for future work on small enterprises. The second phase would involve the analysis of national policies and their impact on small enterprises. 5.49 In addition to the need to research specific aspects of both employment and poverty policy and program performance there is an urgent need to strengthen the capacity to undertake important household budget surveys -- in order to identify the characteristics of households (particularly the poor) -- and to monitor trends over time. Currently, survey capacity is divided between two institutions -- the CBSL and the Department of Census and Statistics. Both institutions are under-resourced with the result that timely publication of survey results is delayed. The project will provide financial and technical support to the Department of Census and Statistics to upgrade its household survey capacity. 5.50 The location of the unit in MPPI will ensure that its recommendations receive appropriate attention. The technical assistance staff will work directly on analytical and research studies with counterpart staff in MPPI. The unit will have also the capacity to hire short-term consultants to undertake specific studies. The cost of the unit at US$1.4 is detailed in Annex 6. Agreement was reached with GOSL during negotiations that the EPPU will develop a three-year action plan which will be updated annually (by October 31 of each year) and approved by IDA (para. 9.1(d)). It is expected that this will be provided to IDA during the second quarter of GOSLs fiscal year. VI. PROJECT COSTS AND FINANCING 6.1 Total project cost is estimated at US$85.0 million equivalent of which the proposed IDA credit for SDR 40.6 million (US$57.5 million) would contribute 68X. Project costs are specified in Table 6.1, and the financing plan is given in Table 6.2. Of the remaining project costs, the Federal Republic of Germany through KFW would contribute DM 15 million (US$10 million equivalent), and US$17.5 million would be contributed by GOSL. The proportion of project costs contributed by partner organizations and the ultimate beneficiaries are difficult to estimate and are not quantified here because different GOSL agencies and NGOs have different cost structures and the lending agencies would have different  - 30 - requirements. For the Credit Fund operations, some partner organizations would probably provide 1002 loan financing with the condition that sub-borrowers save regularly. but these savings would not necessarily be available to fund project activities. Others would be more conservative, such as matching with credit on a one-for-one basis sub-borrower's cash contribution to project activities. The wage costs of the RWF, stipend during training and honoraria and salaries, respectively, of volunteers and supervisors (if they are JSP recipients) for the NF are also not quantified as these would be adjusted from the individual's JSP entitlement. The monthly Rs 1,458 JSP entitlement will continue to be paid out of the Government budget. 6.2 During negotiations agreement was reached with GOSL that IDA's contribution of US$20 million would be onlent to the Trust, while US$36.2 million is passed on as grant to the Trust for implementation of the HRDP, RWF and NF activities. Table 6.1: Proect Caot Sumary /a (In Mfllos) Rupee USS % % of Components Locd Foreign Total Local Foreign Total Foreign TAsI A. Credit Fund 1,410.5 - 1,410.5 35.0 - 35.0 0.0% 41.2% B. Hu-mn Resource Development Fund 528.4 56.4 584.8 13.1 1.4 14.5 9.6% 17.1% C. Rural Works Fund 589.3 64.5 653.8 14.7 1.6 16.3 9.8% 19.2% D. Nutrition Fund 523.3 48.4 571.7 13.0 1.2 14.2 8.5% 16.7% E. Employment& Poverty Policy Unit 10.1 48.4 58.5 0.2 1.2 IA 85.7% 1.6% F. Projdct Management 145.9 - 145.9 3.6 - 3.6 0.0% 4.2% Total Project Cost 3,207.5 217.7 3,425.2 79.6 5.4 85.0 6.3% 100.0% a/ Calculated at a buying rate of Rs.40.3tUSS as of November 1990. Table 6.2: Proect Financins Plan (In MMilo) FEDERAL REPUBLIC UOSL IDA OF GERMANY TOTAL Components Rs. USS Rs. USS Rs. USS Rs. US$ A. Credit Fund 200.7 5.0 806.0 20.0 403.0 10.0 1,409.7 3S.0 B. Human Resource Development Fund 100.8 2.5 483.6 12.0 - - 584.4 14.5 C. Rural Works Fund 64.5 1.6 592.4 14.7 - - 656.9 16.3 D. Nutrtion Fund 189.9 4.7 382.9 9.5 - - 572.8 14.2 E. Employment & Poverty Policy Unit 4.0 0.1 52A 1.3 - - 56.4 1.4 F. Projet Management 145.1 3.6 - - - - 145.1 3.6 Total 705.0 17.5 2,317.3 57.5 403.0 10.0 3,425.2 8S.0  - 31 - A. Cofinancing 6.3 The Federal Republic of Germany (FRG) has proposed a grant of DM 15 million (US$10 million equivalent) for financing the credit component of the project on a parallel basis after reviewing IDA's Staff Appraisal Report. FRG has indicated full support for the conditionality of the IDA project and indeed in the broader areas of re-orientation of the JSP. The FRG grant funds for the credit component would be administered and disbursed thiough KFW on a parallel basis to finance subloans of POs which would be identified separately from those to be financed by the IDA credit. IDA, the Trust and KFW would identify the eligible POs for FRG grant proceeds annually based on credit demand projections of POs/eligible beneficiaries. 6.4 Although not an integrated part of this project, UNDP will provide US$2.5 million for technical assistance to the Trust during project implementation. VII. PROJECT MANAGEMENT AND IMPLEMENTATION A. Institutional Responsibilities 7.1 Administration of the proposed Funds, i.e., the Credit Fund, the Human Resource Development Fund, the Rural Works Fund and the Nutrition Fund would be the responsibility of the Trust, which was established as an apex agency as a result of collaboration with GOSL and IDA and with technical assistance from UNDP. It would receive and disburse designated project funds to partner organizations according to criteria agreed with IDA. 7.2 The establishment of the Employment and Poverty Policy Unit will be the responsibility of the MPPI. B. The Janasaviya Trust Fund 7.3 Organizational and Management System. The Trust was chartered in January 1991. Its charter specifies its nor-profit character and its authority to enter into relationships with foreign sources of funding that have received GOSL approval. Its objectives are employment generation and pove..ty alleviation and include providing credit and otherwise assisting government programs, financial institutions, NGOs, and other groups working to alleviate poverty, which are called partner organizations. Agreement was reached with GOSL/Trust at negotiations that any amendments to the i lest's Charter during project implementation will require IDA's prior agreement (para. 9.1(f)). 7.4 The Trust comprises three elements (Annex 7)t a Board of Trustees, an NGO Advisory Board and an executive arm. It will liaise with GOSL through the Ministry of Policy Planning and Implementation. A technical assistance team assists the Board and a Managing Director, in establishing and operating the programs of the Trust. The executive arm of the Trust has established its own personnel and administrative procedures outside civil service rules and regulations.  - 32 - 7.5 The Board of Trustees. As patron of the Trust, H.E. the President of Sri Lanka nominated a Board of Trustees, consisting of people from the public sector, NGOs, academic institutions and the private sector. The trustees act in their individual capacity and are chosen for their professional competence and commitment to poverty alleviation. The Board consists of 16 members, 50Z of whom are drawn from the public sector. Three of the Trustees are women. The President will be kept informed of the activities of the Trust through the Secretary of Finance who is the Chairman of the Trust. Given the demand on the time of the Secretary, a Vice Chairman acts in his absence and on his behalf. The Vice Chairman is nominated by the President from the Board members outside the civil service. 7.6 The Board of Trustees is to meet at least 10 times a year. Operating concerns are delegated to the Managing Director who is a competent Sri Lankan, selected by the Board for his managerial competence and familiarity with poverty alleviation and self-employment schemes. The Managing Director is responsible to the Board for the running of the Trust. The Chairman, the Board of Trustees and the Managing Director were nominated before appraisal and the heads of the five contemplated departments and necessary support staff were in position before negotiations. 7.7 The NGO Advisory Board. The Trust has an advisory committee consisting of NGOs and peoples' organizations as representatives of the poor, with the objectives of advising the trustees and the Managing Director on the problems and needs of the poor, as well as on the suitability and effectiveness of the programs that are being implemented at the grass root level. 7.8 MPPI Liaison. For many of its activities, particularly for the Rural Works Program, the Trust will need to gain the cooperation of officials in the central line agencies and the regional administrations at district and divisional levels. This cooperation will be assisted by MPPI, the Secretary for which is the Chairman of the Board of Trustees, and senior officials in which have the responsibility for the development of the Trust, and who will be responsible for liaison effort with other agencies during implementation. 7.9 The organization structure of the Trust shall be flexible, and is the responsibility of the Managing Director in consultation with and approval of the Board. The organizational structure combines managerial, promotional, appraisal, supervision, accounting, monitoring and credit skills. The initial grouping of the functions of the Trust under six departments together with an outline of the tasks and responsibilities of each of the functions is given in Annex 1. Operating Strategy 7.10 The Credit and HRD Funds. In order to ensure that the Trust's resources reach the intended beneficiaries for financially viable and sustainable income generating activities, the Trust's management is expected to operate independently of any political or bureaucratic interference. The staff of the Credit Fund would be high caliber professionals with extensive experience in rural credit. The Trust would identify innovative methods of credit delivery t!rough grassroots level government and non-government organizations (Annex 1). The Trust management would support POs which have the ability to reach the target population, are able to lend and recover loans and have sufficient backing to  - 33 - bear the costs and risks of lending. The Trust would also assist and support the POs in steering clear of possible political interference by adhering to the agreed definition of ultimate beneficiaries. 7.11 The success of the credit operations would also depend on the ability of the POs in assisting the beneficiaries in selecting sound income generating activities and providing appropriate services to ensure successful project implementation. The Trust through its HRD Promotion and Field Support Department would assist POs to develop appropriate capacity, and in collaboration with its Monitoring and Evaluation Unit disseminate to POs those models which are found successful. 7.12 It is important, however, that the Trust does not get involved in detailed field work or in propping up wfeaker organizations. It should play the role of promoter, stimulator of innovation and mobilizer of the institutional support to the poor. In order to carry out these functions effectively, the Trust should be well informed about the activities of its POs, knowledgeable about those activities that work and that do not, and willing to facilitate interchange of lessons and ideas among its borrowers. 7.13 Operationally, the Trust would develop a collaborative partnership role vis-a-vis its partner organizations. Rather than responding passively to institutional requests for project finance, it will develop a promotional outreach, enabling it to actively solicit not only credit or rural works project proposals but also related capacity building proposals. Specifically, the Promotions and Field Support Department of the Trust will be charged with the task of identifying the grassroots organizations and assisting them in design and development of sustainable and effective programs. It will also follow an experimental approach documenting and sharing experiences through monitoring and evaluation. 7.14 Rural Works Fund. The RWF would seek to collaborate with institutions at regional levels taking advantage of the decentralization process in government agencies and of the emergence of local-level people's organizations and their networks. Working with community organizations and smaller NGOs on small projects and government agencies and established NGOs on larger projects, a key element of the strategy would be to encourage communities to participate fully in the identification and planning of all projects financed by the RWF. The Fund would operate through a process of project formulation, application, negotiation, approval and release of funds accompanied by an appropriate support package. An organization will become a partner organization when an agreement is reached that commits that organization to implement a project with Trust funds. 7.15 The implementation of the rural works program relies on the decentralized levels of the line agencies and on the planning and coordinating functions of the MPPI, now being developed at the AGA divisional level. The Trust will encourage bottom-up identification and local participation and a project-by-project cycle rather than one based on annual plans and budgets. 7.16 With respect to phasing, the strategy is to start with sectoral activities and collaborating institutions which have experience and capacity. At the same time, Trust staff would need to engage in a massive effort to identify promising implementing agencies and design procedures, as well as set up a  - 34 - technical support and monitoring service. This implies a gradual build-up of the volume of grants from a relatively modest start. 7.17 The initial attention would focus on road and irrigation maintenance where the infrastructure is currently operated at a fraction of its capacity but could be upgraded quickly by labor-intensive methods. Rural sanitation offers another quick starting activity as there are already well functioning programs in place that could be supported. Village irrigation, forestry and soil conservation, on the other hand, would need to be approached more cautiously and systematically. In establishing the sector-specific criteria for irrigation schemes, the Trust would ensure that it is consistent with the criteria applied under the proposed National Irrigation Rehabilitation Project. 7.18 Grass root participation will require a gradual build-up of the capacity of local institutions to undertake contracting and to assume responsibility for the regular maintenance of the upgraded facilities. In the irrigation sector such an approach has recently become national policy. By working with village organizations and NGOs in particular, opportunities would be created for direct community participation in the identification and formulation of proposed projects. Community control of assets and of income from use of these assets would also be an important program feature, underwritten where necessary by credit and technical assistance to develop business opportunities for the poor in related activities like aquaculture in rehabilitated tanks, tree nurseries, woodlots in forestry, for example. 7.19 The Nutrition Fund. Part-time volunteers would be deployed at the average rate of one per 500 population, providing typical rural coverage of 35 under-three children and 12 pregnant or lactating women. Supervision would be at the rate of one per ten volunteers. An implementing NGO in an AGA division typically would recruit, train and manage 100-120 volunteers, 10-12 full-time supervisors and a project manager. The Nutrition Division would contract with institutions such as the Agrarian Research and Training Institute for development of curricula, materials and the conduct of pre- and in-service training for NGO nutrition volunteers and staff. The scale of incremental AGA implementation would bes Year 1-15; Year 2-25; Year 3-30; Year 4-50, and Year 5-60. C. Partner Organizations 7.20 Borrowers from the Credit Fund and recipients of the grants funds will be government agencies, NGOs, GROs, cooperatives and financial intermediaries which lend to the poor. These organizations would become eligible for project support provided they have (a) socio-economic objectives consistent with the overall objectives of the project; (b) proven track record in organizing and assisting community groups; and (c) manaperial capacity to identify and supervise small borrowers and community rural works teams. Eligible POs would enter into a participating agreement with the Trust. A financial simulation of a typical PO is given in Annex 9. During negotiations a draft of a participating agreement was agreed in substance (para. 9.3(b)). 7.21 The likely partner organizations for the Credit Fund and the Human Resource Development Fund are the Regional Rural Development Banks (RRDBs), Thrift and Credit Cooperative Societies (TCCSs), Commercial Banks (for selected schemes), National Youth Services Cooperative Organization (NYSCO), Sarvodaya-  - 35 - SEEDS and other selected NGOs. Likely partner organizations for use of the Rural Works Fund will be the government agencies involved in rural infrastructure development, selected NGOs and Rural Development Societies (RDSs). Nutrition interventions will be primarily implemented through NGOs and GROs. D. Administration Procedures 7.22 Procurement. Under the Credit Fund, individual subloans to ultimate beneficiaries are expected to be below US$1,000 per subloan and the use of subloans would be governed by regulations and procedu.es specified by the POs. Subloans would be used for productive purposes related to agriculture, artisanal production, transport, petty-trading and other services. The funds would be used by individual beneficiaries for purchase of goods and services as agreed by the POs through established commercial practices. The Rural Works Fund would finance purchase of small tools and equipment and operating costs of implementing agencies. The Trust would ensure that small tools and equipment and other materials costing more than US$15,000 per contract would be purchased through local competitive bidding while other items of lower value would be purchased through prudent shopping. Small works would be carried out by force account because their nature and size do not justify competitive contracting. 7.23 Vehicles and motorcycles, computer equipment, office equipment, furniture, and other goods to be procured under the project would be purchased thrcugh prudent shopping when the value is less than US$15,000 per contract, up to an aggregate of US$4 million. Where the value of items or groups of items exceed US$15,000 but is below US$200,000 local competitive bidding procedures acceptable to IDA would apply. Items or group of items exceeding US$200,000 would be purchased under the international competitive bidding (ICB) procedure. However, as the above items would be bought in batches over the five-year project implementation period, it is unlikely that they would be large enough for ICB. The Trust should ensure that goods financed by IDA, would, to the extent possible, be grouped to encourage competitive bidding. 7.24 The consultants required under the Technical Assistance Component would be hired in accordance with the Bank's Guidelines for the use of Consultants.  - 36 - Table 7.1: Procurement Arrangements ICB LCB Other N/A Total Cost Project Elements - (US$ million) ------------- Credit 35.0 35.0 (20.0) (20.0) Tools, Equipment, Materials & Works* - 8.0 8.3 - 16.3 - (7.2) (7.5) (14.7) Computers, Equipment, Furniture and Supplies* - 3.2 2.6 - 5.8 (3.2) (2.6) (5.8) Vehicles & Motorcycles* - 0.6 0.5 - 1.1 (0.6) (0.5) (1.1) Technical Assistance Policy Formulation - - 1.4 - 1.4 (1.3) (1.3) Training - - - 14.6 14.6 (14.6) (14.6) Recurrent Costs - - - 10.8 10.8 TOTAL - 11.8 12.8 60.4 85.0 (11.0) (11.9) (34.6) (57.5) Note: Figures in parenthesis are the respective amounts financed by IDA. * As most items would be purchased in batches over a five-year period, it is unlikely that they would be large enough for ICB. 7.25 Bidding documents for all contracts exceeding $50,000 to be procured under LCB procedures, the first four contracts under $50,000, and all contracts to be procured under ICB, if any, would be sent to IDA for prior review. Other contracts and bid evaluations would be subject to random ex-post review by IDA staff. 7.26 Disbursement Procedures. The proposed IDA credit would be disbursed over a period of five years (Annex 10) against the categories of expenditure shown in Table 7.2 below:  - 37 - Table 7.2: Allocation of Disbursement of the IDA Credit Proposed IDA Category of Expenditures Allocation 2 of Expenditures (US$ million) Financed by IDA 1. Line of Credit 20.0 100 2. Civil Works 13.2 90 3. Computers, Equipment, Furniture 100X of foreign costs; and Supplies 5.2 100% of ex-factory cost 4. Vehicles and Motorcycle 1.0 of local expenditures 907 of local expenditure 5. Technical Asst. & Studies 1.3 6. Training 13.1 100 7. Unallocated 3.7 7.27 All disbursements against contracts for goods and services exceeding US$20,000 would be made against full documentation. Disbursements below that level would be made against Statement of Expenditures (SOEs) for which supporting documents would be maintained by the Trust and would be made available for review by IDA staff. A Special Account in US dollars would be opened in the CBSL, with an initial deposit of US$3.0 million corresponding to about four months estimated disbursements. The application for replenishment will be submitted monthly or whenever the special account is drawn down to 40? of its initial deposit, whichever occurs first. 7.28 Accounts and Audit. The Trust would maintain its accounts in accordance with acceptable commercial accounting practices. During negotiations, agreement was reached with GOSL and the Trust that an independent external auditor satisfactory to IDA would be appointed and that certified copies of Trust's financial statements including corresponding audit reports will be submitted to IDA within four months of the end of Trust's financial year. The audit report would include an opinion on controls and procedures for the use of the Special Account and Statement of Expenditures (SOEs) under the Project (para. 9.2(i)). 7.29 The Trust would compile a list of auditors which the recipients of loan and grant funds would be required to use for providing an annual audit and specify the terms of engagement and scope of audit. 7.30 Reporting Requirements. The Trust would provide operational data on the four funds it operates reasonably requested by IDA. These would include an annual budget and three years projections by the start of each accounting year, and quarterly operational summaries. These reports would include breakdowns by partner organization and the global totals of loans disbursed from the Credit Fund, recoveries, and amounts rescheduled and amounts fallen due but not recovered, and the capital/reserve position of the borrowers. Subloan purposes would also be reported, as would the number of loans made. The Trust's  - 38 - operational summaries for the Credit Fund would also include a brief description of each borrower and each borrowers operations with a focus on gender and location. 7.31 Similarly, .-he operational suummaries of the HRDF, RWF and NF would include nature and extent of expenditure, i.e., on social mobilization, training, capital and recurrent costs with a focus on location and gender. The RWF summaries would also include the wage bill and its adjustment with JSP entitlement, employment and physical infrastructure created. The EPPU would provide an updated annual and a three-year plan of studies every year for IDA's approval. 7.32 Monitoring and Evaluation. Regular program reviews are necessary to check overall program direction, performance of specific components and periodic assessment of the ability of the program to reach the intended beneficiaries. These reviews would focus on programmatic and financial progress as well on their impact on the poor and the unemployed. Such monitoring and evaluation would be conducted primarily through the review of the standardized and other records of the operations of the partner organizations, collected and analyzed by the Trust. 7.33 The objective is to warn managers of the problems in program implementation and alert them of the need for investigation of causes of performance deviation and recourse to necessary corrective action by appropriate modifications in the program. The system would avoid the relentless pursuit of targets and over-rapid program expansion, and provide more attention to quality of implementation. 7.34 In addition, a mid-term impact assessments will be made by an independent research organization. This will cover the impact of the Trust's act"vities on women, adoption of new technologies, environmental impact and other dimensions of poverty. Such external evaluations would also review the effectiveness of the Trust in achieving its objectives. During negotiations agreement was reached that the Trust would appoint an independent research organization to conduct such an assessment in accordance with terms of reference acceptable to IDA by June 30, 1993 for the report to be available by December 31, 1993 (para. 9.2(j)). The proposed monitoring and evaluation system is described in Annex 11. 7.35 An IDA supervision plan for the project is in Annex 12. VIII. BENEFITS AND RISKS A. Benefits 8.1 The major direct benefit of the project is institution-building in the area of human resource development and credit for the poor. Organizations with considerable geographic outreach, but limited poverty focus, would be assisted to reorient their capacity, and those with a stronger poverty focus, but limited outreach and credit management procedures would be strengthened and assisted to expand. NGO networks would also be strengthened and improved linkages between banks and NGOs developed.  - 39 - 8.2 About one million poor are expected to benefit from the resources made available under the project over the next five years. Nearly 175,000 poor household would have the opportunity to increase their earning through the Credit Fund; about half of these would be women. Over 10 million work days of employment, representing Rs 600 million in wage income would be generated from rural works. The RWF would also create long-term employment for about 45,000 men and women, in addition to building economically viable infrastructure. 8.3 The nutrition component will benefit up to one million children under three years of age and 300,Al'D pregnant and nursing mothers. 8.4 Important intangible social benefits include improvement in the social status of women, creation of the organizational base of the poor giving them independence from traditional sources of village-level patronage. 8.5 Through the project's activities indirect benefits will arise, including more active, informed and effective government programs for the poor, an increase in NGO activities among the rural poor, and a larger prospective on how poverty alleviation occurs and how the barriers to progress by the poor can be addressed. B. Risks 8.6 A major risk is that over time the Government would subject the project to pressures of patronage. Furthermore, political interference and bureaucratic controls could stifle the creativity and impartiality of the Trust. 8.7 Another risk is that enthusiasm to do something for the poor might overwhelm the stringent requirements that have to be m.t for the credit portfolio to be self-sustaining, training to be meaningful and infrastructure created to be economically viable. 8.8 To minimize these risks the project includes the following factors that offer prospects for self-correcting behavior during implementation that should promote sustainability in delivering financial and other services to the poor. (a) Performance criteria for partner organizations are clearly spelled out and enforceable; POs which do not meet performance standards will be denied access to project resources. (b) Selection of beneficiaries will be agreed upon before project implementation begins and they will be organized into groups which will discourage those outside the target groups to attempt to gain control of project resources. (c) The proposed project engages NGOs, which are expected to provide a grass root level commitment and flexibility in the development of delivery systems. The proposed advisory committee consisting of only NGOs would be able to collect information on the use of project funds at the grassroots level.  - 40 - (d) Finally, the Government's commitment to poverty alleviation and the selection of a Board of Trustees composed of individuals committed to poverty alleviation can be expected to ensure the success of the Trust' operations by respecting its professionalism and impartiality. C. Programs of Special Emphasis 8.9 Women. Nearly one-half of all beneficiaries of the project a-' expected to be poor women (Annex 13)9 The Trust will encourage partner organizations to organize poor women in separate groups, provide them training in skills and micro-entrepreneurial development and credit. In addition, nearly 300,000 women will benefit from the nvtrition component. 8.10 Environment. No specific environmental dimensions will be included in project activities, which will be small-scale and primarily rural. However, GOSL programs and NGOs that would in principle be eligible for project support, attempt to assist the poor in several ways simultaneously, such as group formation and skills acquisition. Experience shows that when poor become empowered through organization they are quick to address environmental issues, especially as they relate to public health and nutrition. IX. AGREEMENTS REACHED AND RECOMMENDATION 9.1 During negotiations the following agreements were reached with GOSL: (a) GOSL will contribute US$5 million to the Trust's Credit Risk Fund - US$3 million before IDA Credit becomes effective and US$2 million before December 1, 1992 (para. 5.12). (b) GOSL would onlend the rupee equivalent of IDA's allocations to the Credit Fund (US$20 million) at 3Z repayable in 25 years with a five- year grace period for principal and interest, and would bear the foreign exchange risk (para. 5.13). (c) GOSL would maintain thriposha production levels of 870 metric tons monthly during project implementation and would allocate to the Trust as much of that offtake as the nutrition component would require. And, in each division where the nutrition component operates, the AGA would chair quarterly meetings of the division nutrition-health coordinating committee to be set up under terms of reference and with representation satisfactory to IDA (para. 5.38). (d) The Employment and Poverty Policy Unit will develop a three-year action plan which will be updated annually by October 31 of each year and approved by IDA (para. 5.50). (e) IDA's contribution of US$36.2 million is passed on to the Trust as grant for implementation of HRDF, RWF and NF activities (para. 6.2). (f) Amendments to the Trust's Charter during project implementation will require IDA's prior agreement (para. 7.3).  - 41 - 9.2 During negotiations agreement was reached between GOSL, the Trust and IDA that the Trust would: (a) use the following eligibility criteria for P09 using the Credit Funds (i) partner organizations should have at least two years' experience in lending to the target populations (ii) the loan recovery rate of the PO during the two years proceeding loan application should be at least 70Z of the loans falling duet (iii) all subborrowers have been enrolled in a savings program at the time they receive credit and through the period for which they are indebted; (iv) PO must establish an accounting and financial reporting system acceptable to the Trust, and agree to have accounts audited annually by an independent auditor (para. 5.13). (b) require PO to achieve subloan recovery performance of no less than 9o0 from the subborrowers and on time repayment to the Fund for continued eligibility to the Trust's resources (para. 5.16). (c) review on lending rates to the POs from the Trust with IDA and GOSL semi-annually in the light of the POs operating and risk costs and make appropriate adjustment. The maturities of loans to POs would be based on maturities of subloans from POs to the ultimate beneficiaries but would not exceed three years with six months grace period (para. 5.17); (d) require POs to limit all subloans to a maximum of Rs 40,000 (US$1,000 equivalent). Total loans of a PO to a subborrower would not exceed the subloan limit of Rs 40,000 (US$1,000 equivalent) at any given time. Rescheduling would be considered only once for a mximum period of two years in extreme circumstances such as natural disasters or loss of income earning assets. The subloans maturity would be determined by the projected cashflow from the productive activity (para. 5.19). (e) PO would maintain separate project accounts, and prepare an annual income statement, source and utilization of funds status, and a balance sheet certified by an independent auditor (para. 5.20). (f) insist that POs onlending rate would not be less than the average weighted interest rate charged by the Commercial Banks, Development Finance Company of Ceylon, National Development Bank of Sri Lanka, Regional Rural Development Banks, Credit Societies and Cooperatives, as published semi-annually by the Central Bank of Sri Lanka. The Trust would review rates determined by POs semi-annually, and agree on a rate to be used for subloans for the ensuing six months (para. 5.22);  - 42 . (g) use the following eligibility criteria for providing POs with HRDF grant support: (i) POs would have at least one year experience in organizing the poor; (ii) POs would agree to a medium-term development plan linking human resources with use of Credit and other Trust activities; and (iii) Continued access to HRDF will be against compliance with previously agreed activities (para. 5.29). (h) use the following eligibility criteria for RWP project grantss (i) projects should generate substantial employment and use local labor to the extent possible. The labor component should be in excess of 50Z of the total contract cost (compared to 30Z in current works); wage rates would be consistent with those prevailing in the project area, and JSP entitlement in JSP areas would be paid in lieu of wages that they would otherwise earn; and (ii) projects must meet the Trust's technical requirements in each sub- sector (para. 5.33). (i) appoint an independent external auditor satisfactory to IDA; certified copies of Trust's financial statements including corresponding audit reports will be submitted to IDA within four months of the end of the Trust's financial year. The audit reports would also include an opinion on controls and procedures for use of Special Account and Statement of Expenditures under the project (para. 7.28). (j) appoint an independent research organization to conduct a mid-term impact assessment with terms of reference acceptable to IDA by June 30, 1993 for the report to be available by December 31, 1993 (para. 7.34). 9.3 During negotiations, drafts of the following documents were agreed in substances (a) subsidiary loan agreement between GOSL and the Trust (para. 5.13); and (b) participation agreements between the Trust and POs (para. 7.20). 9.4 The following are conditions of credit effectiveness: (a) transfer of US$3 million from GOSL to the Trust's Credit Risk Fund (para. 5.12). (b) signing of a subsidiary loan agreement, satisfactory to IDA, between GOSL and the Trust (para. 5.13); and  - 43 - (c) signing of UNDP project documents, providing technical assistance to the Trust (para. 5.45). 9.5 With the above agreements, the proposed Project constitutes a suitable basis for a Development Credit of SDR 40.6 million (US$57.5 million equivalent) on standard IDA terms with a maturity of 40 years to the Democratic Socialist Republic of Sri Lanka.  - 44 - Annex 1 Page 1 of 8 SRI LANKA POVERTY ALLEVIATION PROJECT Operation of the Credit and Grant Funds 1. This annex describes the operating procedures for each of the three funds operated by the Janasaviya Trust. These funds are: (a) The Credit Fund (b) The Human Resource Development Fund (c) The Rural Works Fund (d) The Nutrition Fund TEE CREDIT FUND - ELIGIBILITY CRITERIA AND OPERATING PROCEDURES obiective 2. The Credit Fund would be managed by the Credit Division of the Trust. Its purpose is to supply credit through P09 for micro-enterprise and self-employment schemes by the poor and unemployed youth. 3. Identification of potential POs would be undertaken with the assistance from the Promotion and Field Support Division. Responsibility for helping POs and beneficiaries to develop capabilities to access credit is with the Human Resource and Institutional Development Division. Eligible Partner Organizations 4. Credit funds are available for each primary outlet (branch) of a partner organization, provided the primary outlet hass at least two years experience in credit operations; an operational savings scheme; a 702 or better on-time collection rate on its lending (90? after one year for continued access); established loan administration procedures and is financially viable; and undertaken mobilization and training for intended beneficiaries by animators or field officers for at least three months before the first credit is disbursed.  - 45 - Annex I Page 2 of 8 Prequalification of Eligible Organizations 5. National P09 would identify beforehand, in agreement with the Trust, those branches or primary outlets that meet the above criteria. Trust Officers (from the Credit Division or via Field Support Staff) would visit the organization and establish that the Trust's requirements in administrative practice and beneficiary support have been met. If the applicant has received a grant from the Human Resources Development Fund, the Trust will verify that the objectives behind the grant have been achieved. 6. New credit outlets established by national or large regional POs with existing credit operations need only meet the requirement of three months' experience in mobilizing and training beneficiaries, and have estab- lished accounting and reporting procedures that meet the Trust's approval. Operation of the Fund 7. Prequalified POs would borrow against subloans made to the benefi- ciaries each quarter (or at other such frequency that is decided in conjunc- tion with the Trust). 8. Information contained in the quarterly application for funding request would be: - Name of organization and location; - Names of applicants for subloans, amounts, terms, conditions and brief purpose. 9. National or regional organizations would summarize applications of the primary outlets and would submit this summarized application to the Trust. Details on applicant, amount and purpose of each subloan would still be provided. 10. For established partners that are meeting criteria, the Trust would forward the requested amount immediately. Data required for monitoring and evaluation purposes would be recorded in the Trust's information system. Operating Costs of Partner Organizations 11. The rate of interest on borrowings from the Trust will be 7Z. The minimum onlending rate to borrowers from the beneficiary target crop will be the weighted average commercial lending rate plus 12 (currently totalling 19.6Z). 12. As early as possible in the project, POs are expected to meet costs of social mobilization, administrative overheads and loan defaults from the interest spread between the cost of funds obtained from the Trust and the onlending rate. Trust staff will review arrangements that the partner organization is making in the achievement of this objective. Partners whose  - 46 - Annex l Page 3 of 8 operating costs and default loans are less than the maximum available under the Trust would use the funds that are made available to finance additional loans. 13. Partner organizations will be expected to hold back from subloans an amount equivalent to 5Z of the sum that is borrowed. This holdback will be invested on the subloan borrower's behalf and earn tnterest at passbook rates. The purpose of the holdback is to provide additional income to that outlet of the partner organization, to provide a cushion against default and to further inculcate a savings discipline in the beneficiaries. Rescheduling of Loans 14. Partner organizations can reschedule each subloan and the equiva- lent amount with the Trust over a two-year period, if conditions of regional disaster or of extreme personal difficulty are encountered by subloan bor- rowers. The Trust would permit only one rescheduling, which would be for a maximum of two years. THE HUMAN RESOURCE DEVELOPMENT FUND - SELECTION CRITERIA AND OPERATING PROCEDURES Obiectives 15. The Human Resource Development Fund will be managed by the Human Resource and Institutional Development Division of the Trust. The fund is intended to help partner organizations (POs) to increase their capacity to mobilize, supply credit to and support unemployed youth and the poor in income generatir..: activities. The fund operates by providing grants to the P0O for the exz .'nditures listed below: (a) training of staff and the provision of consultancies to strengthen knowledge and skills in the development of credit and related bene- ficiary support activities; (b) study visits to similar organizations in the Asia Region; (c) consultancies designed to strengthen the management of poorly performing branches and outlets; (d) costs of networking workshops and seminars designed to share exper- ience and learning among P09 and between P0 and the Trust, and to establish a network of organizations responsible for implementing the project at the grass roots level; (e) training of beneficiaries in income generating capabilities; (f) purchases of equipment and vehicles needed to manage the program;  - 47 - Annex I Page 4 of 8 (g) recurrent costs, such as salaries of additional staff, incentive payments, travelling costs of staff, maintenance of equipment and vehicles associated with human rasource mobilization and skills training, as agreed by the Trust. The maximum period for branches and outlets to receive these grants is three years. It is antici- pated that these costs would be met by the interest rate spread within this period; and (h) grants would also be made available to training institutions for upgrading staff capabilities and for developing curricula that meet the Trust's training needs. Eligible Partner Organizations 16. Grants under the Human Resource Development Fund are made for the above expenditures to organizations which meet the following criteria: (a) the organization has at least 12 months' experience in group mobi- lization and training; and (b) it has an established savings scheme operating with beneficiaries. Selection Criteria 17. Grants from the Trust are made on the condition that the partner organization: (a) commits to undertake mobilization and training activities with unemployed youth and the poor, in order to assist them in identify- ing viable micro-enterprise activities and in obtaining credit; (b) obtains -n "in-principle" agreement from the Credit Division of the Trust to .1-1ply credit if the human resource development activities prove to be effective; (c) provides a medium term development plan for each PO which links the use of the Human Resource Development Fund with the proposed use of credit from the Trust; (d) commits (for national or regional POs with several primary outlets) to recruit and train staff and to establish procedures that meet the Trust's criteria. New primary outlets would not require prior experience to obtain grant funds under these conditions; and (e) for repeat grants, HRDF grants are made against satisfactory achievements under previous grants.  - 48 - Annex I Page 5 of 8 OperatinR Procedures of the Fund 18. Commitments by the Trust to provide grants for capital and recur- rent expenses will be made under a set of conditions on which the partner organization and the Trust will agree. These conditions are that the grant: (a) will be for a maximum period of one year for recurrent expenses, renewed on agreement from the Trust; (b) is reimbursed each quarter against a statement of expensess (c) is appraised against cost guidelir.es established by the Trust for salaries, training expenses, travel, per diem allowances, and consultancy costs; and (d) if required for purchases over Rs 50,000 (or alternate amount approved by the Board of the Trust), will be expended only after obtaining prior competitive quotations. 19. The Human Resources and Institutional Development Division will start operations by identifying potential POs. The division will assess their programs and agree with them on the nature of assistance required to build their capacity for achieving the Trust objectives. In this regard, the staff will also assess the capabilities of local and regional training organizations which could be contracted to provide specific expertise in training or techni- cal assistance. In collaboration with the Promotion and Field Support Division, the division's staff will also identify smaller NGOs with little track record requiring capacity building and training to become effective partners. The Trust would agree with the partner organization on a devel- opment plan which would link the grant assistance to the Credit Fund. This plan would clearly state the number and type of beneficiaries (lending, rural works or nutrition beneficiaries) that would be ultimately reached by the POs, the expected amount to be drawn from the Credit Fund for the lending target groups, and the specific training and technical assistance interventions required to build the capacity of the POs. Costing guidelines would be developed by the HRDF Division for, among others, training costs, trainers' fees and consulting services. Allowance for a variety of approaches and flexibility would be required in view of the different nature of the POs. An individual agreement with each PO would commit grants for training and institution building for a maximum period of three years, to be reviewed on a yearly basis. RURAL WORKS FUND - SELECTION CRITERIA AND OPERATING PROCEDURES Obiectives 20. The management of this Fund is the responsibility of the Rural Works Division, with the assistance of the Promotion and Field Support Division and Human Resource and Institutional Development Division. Its objectives are:  - 49 - Annex 1 Page 6 of 8 (a) to provide income for the poor through the labor intensive construction of viable rural infrastructure; and (b) to the extent possible, to maximize the long-term sustainable in- come generation capacity of the poor. It would achieve this objec- tive through the choice of the infrastructure projects that incre- ase the incomes for the poor, in the establishment of maintenance contracts or in training and upgrading skills that will lead to further employment. Selection Criteria 21. Grants for rural works project will be made to AGAs, NGOs, and grass roots organizations able to identify and manage rural works projects that meet the Trust's criteria, as follows: (a) must be at least 50Z labor intensive, unless otherwise approved by the Director, Rural Works (the exception would be for small pro- jects that have a high material cost content); (b) employment is aimed at project beneficiaries in the region; (c) the project must has significant social and/or economic benefits which must meet one or more of the following priorities of the Fundt Ci) highest priority to projects which provide employment to the poor, and where the long-term economic or social benefits accrue to the poor (e.g. irrigation schemes for marginal farmers, market or job access facilities for the poor, water sanitation or other social facilities in deprived areas. etc.); (ii) next priority to projects which build a sustainable income generating capability (through on-the-job training and assis- tance in the acquisition of employable skills and/or through the establishment of operations and maintenance contracts); and/or (iii) lowest priority is for projects where the poor receive wage employment, but where the long term benefits go to the non-poor; (d) adequate arrangements for design, management of implementation and maintenance can be established; (e) to the extent possible, the project promotes grass roots particip- ation in the identification and planning of the project; and (f) the project meets the technical requirements in cash sub-sector (irrigation, roads, etc.) prepared by PACT Consultants.  - 50 - Annex I Page 7 of 8 Assessment of Application for Grants 22. The information required for an assessment of an application is designed to provide Trust staff with an indication that the project is viable and will meet Trust criteria. This information provides details ons (a) location, duration and type of projects (b) numbers and wage cost of beneficiaries in work force; (c) evidence of compliance with Trust's selection criteria; (d) organizational and management arrangements for implementation and maintenance; (e) funding required of Trust; (f) arrangements made for design and technical supervision of construc- tion, and where appropriate, agreement of concerned technical agency that project is feasible and conforms to regional develop- ment plans; (g) preliminary technical details; and (h) proposed financing and reimbursement arrangements. 23. A preliminary screen and an in-principle approval would be given on basis of above information. Rural Works Division Staff would visit the site and meet with implementing agencies to review arrangements for engineering design that may be needed. Costing and management details, together with any further technical assistance needed that may be required would also be assessed. 24. Compliance with complete technical and operational criteria would also be assessed. Smaller projects could be reviewed on site by field support staff on behalf of the Rural Works Division. Operations of the Rural Works Projects 25. Rural Works Division staff would visit the project at least once during the construction period, and on completion. Staff of the Field Support section would inspect smaller projects on behalf of the Dlvision. 26. Payments would be made against an agreed schedule of completed work. Projects requiring inspection and certification of completed work would use Trust staff to the maximum extent possible. OPBRATION OF TI= NUTRITION COMPONENT 27. The nutrition program, managed by the Nutrition Division of the Trust, shall operate through NGOs who are willing to organize and supervise volunteers on an AGA divisional basis to provide services, advice and nutri- tional feeding supplement to at-risk pregnant women, nursing mothers and severely malnourished or growth faltering children to three years of age. 28. The task of the NGO volunteers is to spend approximately 2-3 hours daily in their selected villages in weighing children, recording and analyzing  - 50 - Annex I Page 7 of 8 Assessment of Application for Grants 22. The information required for an assessment of an application is designed to provide Trust staff with an indication that the project ls viable and will meet Trust criteria. This information provides details on: (a) location, duration and type of project; (b) numbers and wage cost of beneficiaries in work force; (c) evidence of compliance with Trust's selection criteria; (d) organizational and management arrangements for implementation and maintenance; (e) funding required of Trust; (f) arrangements made for design and technical supervision of construc- tion, and where appropriate, agreement of concerned technical agency that project is feasible and conforms to regional develop- ment plans; (g) preliminary technical details; and (h) proposed financing and reimbursement arrangements. 23. A preliminary screen and an in-principle approval would be given on basis of above information. Rural Works Division Staff would visit the site and meet with implementing agencies to review arrangements for engineering design that may be needed. Costing and management details, together with any further technical assistance needed that may be required would also be assessed. 24. Compliance with complete technical and operational criteria would also be assessed. Smaller projects could be reviewed on site by field support staff on behalf of the Rural Works Division. Operations of the Rural Works Projects 25. Rural Works Division staff would visit the project at least once during the construction period, and on completion. Staff of the Field Support section would inspect smaller projects on behalf of the Division. 26. Payments would be made against an agreed schedule of completed work. Projects requiring inspection and certification of completed work would use Trust staff to the maximum extent possible. OPERATION OF THE NUTRITION COMPONENT 27. The nutrition program, managed by the Nutrition Division of the Trust, shall operate through NGOs who are willing to organize and supervise volunteers on an AGA divisional basis to provide services, advice and nutri- tional feeding supplement to at-risk pregnant women, nursing mothers and severely malnourished or growth faltering children to three years of age. 28. The task of the NGO volunteers is to spend approximately 2-3 hours daily in their selected villages in weighing chlldren, recording and analyzing  - 52 - Annex 2 Page . of 2 SRI LANKA POVERTY ALLEVIATION PROJECT Eligibility Criteria for Sub-Loans to Beneficiaries from Partner Organizations 1. The partner organizations will finance the productive activities of the intended beneficiaries who are among the poorest members of the concerned community. These activities may be ongoing activities (such as in Schedule A), or those suggested by the partner organization and/or the local community. 2. The partner organizations will also ensure that: (a) the activities financed by them are technically feasible and manageable by the intended beneficiaries; (b) such lending would improve the sub-borrower's standard of living/quality of life; (c) the intended beneficiaries are enrolled in a savings program before loans are made to them. Schedule A: SOME ACTIVITIES TO BE FINANCED BY PARTNER ORGANIZATIONS 1. Homestead cultivation 2. Land leasing 3. Shallow tubewells for irrigation and selling water 4. Paddy husking 5. Coir products 6. Coconut oil making 7. Fishing 8. Fish net making 9. Improvement of fishing boats 10. Purchase of fishing nets and boats 11. Dairy cattle 12. Cattle fattening 13. Agricultural produce trading 14. Grocery shop 15. Transport equipment 16. Polleny products 17. Cane and metal works 18. Garment making 19. Weaving 20. Group based poultry farming 21. Group based rice hutters and oil expellers 22. Group based fish ponds 23. Carpentry  - 53 - Annex 2 Page 2 of 2 24. Blacksmithy 25. Soap making 26. Aluminum utensils 27. Fuel briquetting 28. Fish meal and animal feed 29. Shoe making 30. Furniture making 31. Brick making 32. Piggery development  -54 - SRI LANKA POVERTY ALLEVIATION PROJECT Denosi and Ut Rale8 (Percent Per Annum) 1987 1988 1989 1990 TREASURY BILLS Primary market 10.6% - 11.0% 18.5% - 18.9% 15.2% - 17.5% 15.5% - 18.5% Secondary market 11.3% 23% 20.5% 21.5% DEPOSIT RATES Commercial Bank Savings deposits 6.0% - 11.0% 5.0% - 11.0% 5.0% - 14.0% 6.0% - 15.0% 6-month fixed 7.0% - 13.0% 8.0% - 13.5% 8.0% - 16.0% 8.0% - 16.0% 12-month fixed 8.5% - 14.0% 9.0% - 15.5% 11.0% - 20.5% 11.0% - 21.0% LEDING RATES Weighted average prime lending rate 14.2% 16.9% 18.0% 17.3% Development Finance Corporation 14.0% - 19.0% 14.0% - 19.0% 14.0% - 19.0% 14.0% - 19.0% National Development Bank 7.0% - 14.0% 7.0% - 18.0% 7.0% - 18.0% 10.0% - 21.0% OTHER SELECTED LENDING RATES Agriculture loans refinanced through Central Bank n.a. n.a. n.a. 12.0% - 14.0% Bank of Ceylon/Peoples Bank loans to Youth (NYSCO) n.a. n.a. n.a. 18% Sarvodaya SEEDS for Small Enterprise Development n.a. n.a. n.a. 20% Thrift and Credit Cooperative Societies n.a. n.a. n.a. 12.0% - 18.0% Source: Central Bank of Sri Lank - Annual Report. Average Minimum and Maximum Rates during the year.  - 55 - Annex 4 Page 1 of 14 SRI LANKA POVERTY ALLEVIATION PROJECT Financial Prolections for the Credit Fund Operations 1. The following financial projections for the Credit Fund operations are prepared to demonstrate the effect various assumptions of interest spread, operating cost and portfolio performance would have on the financial viability of the Fund operations. The principal assumptions used in the simulation presented are summarized belows (a) The GOSL loan is disbursed 1002 to POs in the year received. (b) The subloans are repayable within 12 months and 95Z of subloan collections are re-lent in the following year. The balance 52 is invested in a liquidity/security fund. Cc) 25Z of subloan principal falling due is rescheduled over a period of two years. (d) Bad loans would amount to 12 of the second year operations, and increase at 12 per annum until it reaches a maximum of 5Z of total disbursements in a year. Se) Interest spread is 4Z. 5f) Operating costs are based on staffing and other cost budget prepared at appraisal. (g) Investment income would be at 102. 2. The above assumptions are variables in the model. Given a set of assumptions the model simulates the assets, liabilities, net income and the cashflow position of the Trust. In addition, the model works out the collection ratio before and after rescheduling. The two constants in the model are the GOSL loan financed out of the IDA Credit and the FRG grant of $30 million and the GOSL contribution to the Risk Fund of $5 million. 3. The following paragraphs describe the outputs resulting from the above assumptions. Subsequently, one or two key assumptions are varied to demonstrate the effect on financial performance. Lending Activity - Table 1 4. Table 1 summarizes the lending activity of the Credit Fund. It projects the volume of funds disbursed onlending to the rural poor and the  - 56 - Annex 4 Page 2 of 14 unemployed. Lending funded from the GOSL loan is termed first round lending, as disbursements would be made against this activity. As the Trust activities continue to grow, an increasing share of lending would be funded out of collections of loans falling due. Collections of Principal - Table 2 5. Table 2 projects the collection performance of the Credit Fund on its loans to partner organizations. Twenty-five percent of amounts falling due are assumed to be rescheduled, and the amounts uncollectible grow from 12 of total disbursements in year 2 to 41 in year 5. The collection ratio before rescheduling averages around 801-85Z and after rescheduling around 96Z-982. Although 252 of new loans falling due are rescheduled, the collection ratio before rescheduling is higher than 75Z due to the effect of collecting previously rescheduled amounts. Credit Fund Income Statement - Table 3 6. Table 3 projects the Trust's income statement assuming a 31 cost of funds and a 71 onlending rate. Important sources of income in addition to this spread are interest earned on balance in the Credit Risk Fund and in the liquidity/security fund, which would be invested in GOSL securities. The current yield on six-month Treasury Bill is 152. However, a modest 10 income on investments is assumed in the projections. The projections are simplified by assuming that loans are disbursed in the first day of the year and recovered on the last day of the year. 7. The amount of Credit Fund loans falling into arrears is expected to be negligible at the outset, but increases to 42 of disbursement in year 5. The amounts falling into arrear are assumed unrecoverable and written -off against income. As the Credit Fund would be wholesaling credit, its operating cost is assumed to remain low. Credit Fund Assets - Table 4 8. The asset side of the balance sheet is presented in Table 4. The Credit Fund's major asset Is cash held. The presentation assumes that all lending during the disbursement period is short term. However, the introduction of term lending should not have a material effect on Credit Fund's income or on the size of the Credit Fund. The other significant assets are the Credit Risk Fund created out of the initial GOSL grant and the liquidity/security fund created by retaining 5 of loan collections. The tws funds together ($9.6 million in year 5) will exceed the value of the rescheduled loans outstanding ($7.6 million in year 5) in any given year. The model assumes that the rescheduled amounts will be collected in full. As long as the risk fund assets exceed the rescheduled loan balance, the Credit Fund would be able to avoid financial collapse due to large scale default on rescheduled loans.  -57 -Annex 4 Page 3 of 14 Credit Fund Liabilities and Net Worth - Table 4 9. The Credit Fund's major liability is the long-tenm loan from GOSL funded by the proceeds of the IDA credit and the PRG grant. The loan is assumed to carry an interest rate of 32, with a maturity over 25 years including five years of grace. The interest is capitalized during the five- year grace period. Therefore, there is no repayment of principal or interest during the credit disbursement period. 10. The Credit Fund's net worth consists of the initial GOSL contribution of US$5 million and retained earnings. Credit Fund Sources and Uses of Funds - Table 5 11. Table 6 summarizes the Credit Fund's flow of funds, which mathematically integrates the balance sheet and the income statement. The sources and uses of funds statement is presented in several parts correspondinq to the major activities of the Fund. 12. Part I shows the financing of the Fund's administration. This is shown separately as funding of the Credit Fund's administration costs out of its income is fundamental to the viability of the Fund. The major sources of funds for this purpose include net income from the Credit Fund operations, income from investments in the Credit Risk Fund and the liquidity/security fund, and the retention of cash due to capitalization of interest on the GOSL loan. 13. Part II shows the mechanics of the Credit Fund's use of outside funding. The GOSL loan received is fully available for credit operations. No loan repayments are due in the disbursement period. GOSL grant received is fully invested in a Credit Risk Fund. The net effect of these flows show the new funds available for first round lending each year. 14. Part III continues the progression by showing how funds are used in credit operations. Here the amounts disbursed constitute a use of funds and the recoveries provide a source of funds. As all loans are assumed to be of short-term, the use of funds in a year should be zero if all borrowers pay on time. However, this rarely happens and 25Z of the annual disbursements are assumed to be rescheduled. Some delayed payments and bad debt losses are inevitable, especially when lending is innovative, as in the case expected to occur under the project. The first instance of failure to repay according to contract is provided for by rescheduling of amounts that are not received when originally due. Rescheduling of subloans are done by the partner organizations, and their action reschedule their loans from the Credit Fund. The rescheduled loans outstanding and loans in arrear and not considered recoverable become a use of Trust funds.  - 58 - Annez 4 Page 4 of 14 15. Par IV summarizes the first three parts and shows the change in the cash position of the Credit Fund, which is a year-end position reflecting the assumption that all loans are repayable on the last day of the year. Sensitivity Analysis. 16. The assumption of loan losses is the most critical fc: the Fund's operations. As the Fund has a spread of only 42, any loan loss above this would result in a financial loss to the extent not covered by income from Fund's investment activities. The simulation can be manipulated to determine the rate of investment income required to offset different rates of loan losses. Of course, one could assume higher rates of spread to cover increasing loan losses. However, this is not considered feasible due to limitations on ultimate onlending interest rate that could be charged to beneficiaries and the need to allow sufficient spread to POs. 17. If the loan loss rate is assumed at 52 in year 2 and assumed to increase to 102 by year 5, with all other assumptions remaining same the following changes to the financial position of the Credit Fund take placet Year 1 Year 2 Year 3 Year 4 Year 5 Collection Ratio Before reschAeduling 75Z 752 79Z 802 812 After rescheduling 1002 94Z 92Z 902 89Z Net Cashflow from Operations 415.5 293.0 -30.5 -329.8 -585.3 18. The net cashflow will turn negative due to the net income from Credit Fund operations, i.e, the net interest income minus loan losses would turn negative and the investment income from the Risk Fund investments would not be sufficient to cover it. However, if the investment income is assumed to increase to 152 compared to 102 assumed in the model, the Credit Fund would be able to maintain a positive cashflow provided a minimum collection ratio after rescheduling of 922 is achieved: Year 1 Year 2 Year 3 Year 4 Year 5 Collection Ratio after Rescheduling 1002 942 922 922 922 Net Cashflow 568.2 439.9 59.6 110.0 192.0 19. The above simulation shows, given the current yield of Treasury Bills of about 15Z, the Credit Fund should aim to contain its loan losses well below 82. This is very critical since from year 6 onwards the Trust will start repaying the principal and interest on the GOSL loan which would require it to generate additional funds.  ANNEX 4 - 59 - Page 5 of 14 ~~~ 59 ~~~~~~ ~Scenario 1 Table I SRI LANKA Page I of 5 POVERTY ALLEVIATION PROJECT CREDrr FUND Projected Financial Statements (US$ '000) Trust Lending Activity Summary Year I Year 2 Year 3 Year 4 Year S Principal Disbursed to Trust clients from IDA credit & FRO Grant (first round lending) to Government Credit Programs 0.0 0.0 0.0 0.0 0.0 to Non-government Programs 2,900.0 5,100.0 6,000.0 7,600.0 8,400.0 to State Banks' Programs 0.0 0.0 0.0 0.0 0.0 Total first round disbursements 2,900.0 5,100.0 6,000.0 7,600.0 8,400.0 Subsequent round disbursements funded by 95% of previous year's principal collections of Credit Programs 2,066.3 5,382.2 9,089.0 13,617.9 Total Disbursements 2A900.0 7,166.3 11,382.2 16,689.0 22,017.9  ANNEX 4 Page 6 of 14 60 - Scenario I Table 2 SRI LANKA Page 2 of 5 POVERTY ALLEVIATION PROJECT CRED1T FUND Colections Worksheet (US$ '000) Year I Year2 Year3 Year4 Year5 Pncipal from first round lending faUing due 2,900.0 5,100.0 6,000.0 7,600.0 8,400.0 Total first round principal faUing due 2,900.0 5,100.0 6,000.0 7,600.0 8,400.0 Total subsequmt rounds' principal falling due 0.0 2,066.3 5,382.2 9,089.0 13,617.9 Total principal faUing due before reschedulig 2,900.0 7,166.3 11,382.2 16,689.0 22,017.9 less: Principal rescheduled (assumed 25% falling due) 725.0 1,791.6 2,845.6 4,172.2 5,504.5 Maturities of rescheduled principal 362.5 362.5 (assumed rescheduled over 2 years) 895.8 895.8 1,422.8 1,422.8 2,086.1 Rescheduled principal falling due 0.0 362.5 1,258.3 2,318.6 3,508.9 Total Principal Falling Due after rescheduling 2,175.0 5,737.2 9,795.0 14,835.3 20,022.3 less: Principal falling into arrears 0.0 71.7 227.6 500.7 880.7 (see Income Statement provision) Total principal collections 2,175.0 5,665.5 9,567.3 14,334.6 19,141.6 Principal collections recycled into subsequent round lending (0 95%) 2,066.3 5,382.2 9,089.0 13,617.9 18,184.5 Noruecycled portion to liquidity/security fimd 108.8 283.3 478.4 716.7 957.1 Principal collection rate before rescheduling 0.75 0.79 0.84 C.86 0.87 Principal coUection rate after rescheduling 1.00 0.99 0.98 0.97 0.96  ANNEX 4 - 61 - Page 7 of 14 Table 3 SRI LANKA Page 3 of 5 POVERTY ALLEVIATION PROJECT CRED1T FUND Projected Income Statement (US$ '000) Year I Year2 Year3 Year4 YearS Interest income from Credit Fund Operations @ 7% 203.0 501.6 796.8 1,168.2 1,541.3 less: Interest expense on project loan @ 3 % (Capitalized during 5 yr grace period) 44.3 165.7 335.4 546.9 800.1 Net interest income from credit operations 158.7 335.9 461.4 621.3 741.1 less: Provision for loans in arrears: % of annual disbursements assumed going into arrea 0.0 1.0 2.0 3.0 4.0 Amount of provision 0.0 71.7 227.6 500.7 880.7 Net Income from Credit Fund Operations 158.7 264.3 233.7 120.7 (139.6) Interest income. at 10% of average balance: Credit Risk Fund 300.0 530.0 583.0 641.3 705.4 Liquidity/security fund 5.4 19.6 43.5 79.4 127.2 Net Financial Income 464.2 813.9 860.3 841.3 693.1 Staff costs 22.4 24.4 33.0 35.0 36.6 Other operating expenses 60.2 73.0 88.4 104.4 109.4 Depreciation expense @ 25% 0.0 0.0 0.0 0.0 0.0 Net Income (or Loss) 381.6 716.5 738.9 701.9 547.1 GOSL grants received 3,000.0 2,000.0 0.0 0.0 0.0 Other grants received 0.0 0.0 0.0 0.0 0.0 Change in Net Worth 3,381.6 2,716.5 738.9 701.9 547.1  ANNMA - 62 - Page 8 of 14 Scenario I Table 4 SRI LANKA Page 4 of 5 POVERTY ALLEVIATION PROJECT CREDIT FUND Projected Balance Sheet (US$ '000) Year I Year2 Year 3 Year4 YearS ASSETS Cash (60 days' staff cost & other oper. experditures) 13.8 16.2 20.2 23.2 24.3 Credit Fund: Cash available for lending 2,481.8 6,449.8 10,925.3 16,618.5 22,770.9 (balancing figure from sources & uses) Accrued interest (assumes all due 31 Dec) 0.0 0.0 0.0 0.0 0.0 Unrescheduled loans outstanding 0.0 0.0 0.0 0.0 0.0 Rescheduled loans 725.0 2,154.1 3,741.3 5,595.0 7,590.6 Gross Credit Fund 3,2u.8 8,u;.8 14,O66./ 2,2z13.5 30,301.5 less: Alowance for loans in arrears 0.0 0.0 0.0 0.0 0.0 (assumes provision = new arrears) Net Credit Fund 3,206.8 8,603.8 14,666.7 22,213.5 30,361.5 Liquidity/socurity fimd 108.8 392.0 870.4 1,587.1 2,544.2 Credit Risk Fund 3,000.0 5,300.0 5,830.0 6,413.0 7,054.3 Total Financial Assets 6,329.3 14,312.1 21,387.3 30,236.9 39,984.3 Misellaneous Assets 0.0 0.0 0.0 0.0 0.0 Fixed Assets 51.9 51.9 51.9 51.9 51.9 less: Accumulated depreciation 0.0 0.0 0.0 0.0 0.0 Net fixed assets 51.9 51.9 51.9 51.9 51.9 TOTAL ASSETS 6,381.2 14,364.0 21,439.2 30,288.8 40,036.2 IUABHL1TES Current Liabilities Accrued interest expense 0.0 0.0 0.0 0.0 0.0 (assumes all due 31 Dec) Otheraccrualsandpayables 3.4 4.1 5.1 5.8 6.1 (15 days staff costs & other op exp) Current portion of project loan 0.0 0.0 0.0 0.0 0.0 (assumed paid on 31 Dec) Total CurretLiabilities 3.4 4.1 5.1 5.8 6.1 Long Term Liabilities Project loan from GOSL 2,951.9 8,096.2 14,261.9 22,197.3 31,144.2 Capitalized interest @3%, 5 yrs grace 44.3 165.7 335.4 546.9 800.1 less: Current portion (paid 31 Dec) 0.0 0.0 0.0 0.0 0.0 Outstanding project ioan balance 2,996.2 8,261.9 14,597.3 22,744.2 31,944.3 TOTAL LIABILITIES 2,999.6 8,266.0 14,602.3 22,750.0 31,950.4 NET WORTH GOSL grants 3,000.0 5,000.0 5,000.0 5,000.0 5,000.0 Other gants 0.0 0.0 0.0 0.0 0.0 Retained earning 381.6 1,098.0 1,836.9 2,538.8 3,085.9 Total Net Worth 3,381.6 6,098.0 6,836.9 7,538.8 8,085.9 TOTAL LIABILITES AND NET WORTH 6,381.2 14,364.0 21,439.2 30,288.8 40,036.2  ANNEX 4 - 63 - Page 9 of 14 Scenario 1 Table 5 SRI LANKA Page 5 of 5 POVERTY ALLEVIATION PROJECT CREDIT FUND Projected Sources and Uses of Funds (US$ '000) Year I Yoar2 Year3 Year4 YearS Pat I Fu-di of Foundatln Adminliatlon Not Income from Credit Fund Operations 158.7 264.3 233.7 120.7 (139.6) Interest capitalized on projet loan during 5 yer grace period 44.3 165.7 335.4 546.9 8C0.1 Credit Risk Ftmd interest income 300.0 530.0 583.0 641.3 705.4 Liquidity/sccudty ftmd intrest income 5.4 19.6 43.5 79.4 127.2 IDer (deer) in other acemals & payables 3A 0.6 1.0 0.8 0.3 less: Increase (decreae) in Misc. Aetb 0.0 0.0 0.0 0.0 0.0 Increasoin60daycashrequirment 13.8 2.5 4.0 3.0 1.1 Staff cots 22.4 24.4 33.0 35.0 36.6 Other operating expcses 60.2 73.0 U8.4 104.4 109.4 Investment in Credit Risk Fund 300.0 530.0 583.0 641.3 Fus available (qued) after aduiln. 415.5 580.3 541.2 663.6 705.1 _-t H Fuig of Fixed Assets & Crdit Operations from Outde Sourcs: Project loan received from COSL .,951.9 5, 100.0 6,000.0 7.600.0 8,400.0 Ic: Inveted in Fixed Aset 51.9 0.0 0.0 0.0 0.0 Project loa proceeds available for Crdit Fund operations 2,900.0 5,100.0 6,000.0 7,600.0 8.400.0 OOSL gratm received 3,000.0 2.00D.0 0.0 0.0 0.0 lss: Repayment of project loan to OOSL 0.0 0.0 0.0 0.0 0.0 Less Investment In Credit Risk Funkd 3,000.0 2.000.0 Net mw finds avaiable for lentn 2,900.0 5, 00.0 6,000.0 7,600.0 8,400.0 Pa tm Fid Used in Credit Operatiwos: Total Disbements 2,900.0 7,166.3 11,382.2 16,689.0 22,017.9 less: Total Principal Pkiling due from borowr befor Rescheduling 2,9C0.0 7,166.3 11,382.2 16,689.0 22,017.9 Funds used before reocebduling 0.0 0.0 0.0 0.0 0.0 Principal rescheduled 725.0 1,791.6 2,S45.6 4,172.2 5,504.5 les: Rescheduled pdncipal falling due 0.0 362.5 1,258.3 2,318.6 3,508.9 Fands used aftr schedug 725.0 1,429.1 1,587.3 1,853.7 1,995.6 Pridcipal falling into arar 0.0 71.7 227.6 500.7 880.7 Actual ds used in et operadIons 725.0 1,500.7 1,814.9 2,354.3 2,876.3 Put IV _3umasy of Sore nd Uses of Pds: Fund available (rqed) after admnin (Pt1) 415.5 580.3 541.2 663.6 705.1 Net ne funds available for lending (Pt II) 2,930.0 5,100.0 6,000.0 7,6C0.0 8,400.0 les: Actual funds ued in credit operations (Pt I) 725.0 1,500.7 1,814.9 2,354.3 2,876.3 Alocaton to Uqudilty/lsouity find 108.8 283.3 478.4 716.7 957.1 Lon In afrrea writtn off 0.0 71.7 227.6 500.7 880.7 Other grsb eceivd 0.0 0.0 0.0 0.0 0.0 IDCr (deer) in scemed Interest expense 0.0 0.0 0.0 0.0 0.0 Change In C"dit Fund Cash Position 2,481.8 3,968.0 4,475.6 5,693.2 6,152.4 (ae Credit Pund asuet n bale iheet) Memo: Credit Risk Fund Conrbution 3,000.0 2,000.0 O.C 0.0 0.0 Interert Income Invesd 0.0 300.0 530.0 S83.0 641.3 Credit Risk Fund Yr end 3,000.0 5,300.0 5,830.0 6,413.0 7,054.3 Credit Risk Fund Yr. BDe. 3,000.0 3,030.0 5,300.0 5,830.0 6,413.0  ANNEX 4 - 64 - Page 10 of 14 Scenario 2 Table I SRI LANKA Page 1 of 5 POVERTY ALLEVIATION PROJECT CREDIT FUND Projected Financial Statements (US$ '000) Trust Lending Activity Summary Year I Year 2 Year 3 Year 4 Year 5 Principal Disbursed to Trust clients from IDA credit & FRG Grant (first round lending) to Government Credit Programs 0.0 0.0 0.0 0.0 0.0 to Non-government Programs 2,900.0 5,100.0 6,000.0 7,600.0 8,400.0 to State Banks' Programs 0.0 0.0 0.0 0.0 0.0 Total first round disbursements 2,900.0 5,100.0 6,000.0 7,600.0 8,400.0 Subsequent round disbursements funded by 95 % of previous year's principal collections of Credit Programs 2,066.3 5,109.9 8,372.4 12,185.0 Total Disbursements 2,900.0 7,166.3 11,109.9 15,972.4 20,585.0  ANNEX 4 - 65 - Page 11 of 14 Scenazio 2 Table 2 SRI LANKA Page 2 of 5 POVERTY ALLEVIATION PROJECT CREDIT FUND Collections Worksheet (USS '000) Year I Year2 Year3 Year4 Year5 Principal from first round lending falling due 2,900.0 5,100.0 6,000.0 7,600.0 8,400.0 Total first round principal falling due 2,900.0 5,100.0 6,000.0 7,600.0 8,400.0 Total subsequent rounds' principal falling due 0.0 2,066.3 5,109.9 8,372.4 12,185.0 Total principal falling due before rescheduling 2,900.0 7,166.3 11,109.9 15,972.4 20,585.0 less: Principal rescheduled (assumed 25% falling due) 725.0 1,791.6 2,777.5 3,993.1 5,146.2 Maturities of rescheduled principal 362.5 362.5 (assumed rescheduled over 2 years) 895.8 895.8 1,388.7 1,388.7 1,996.5 Rescheduled principal falling due 0.0 362.5 1,258.3 2,284.5 3,385.3 Total Principal Falling Due after rescheduling 2,175.0 5,737.2 9,590.7 14,263.8 18,824.0 less: Principal falling into arrears 0.0 358.3 777.7 1,437.5 2,058.5 (see Income Statement provision) Total principal collections 2,175.0 5,378.9 8,813.0 12,826.3 16,765.5 Pnncipal collections recycled into subsequent round lending (0 95%) 2,066.3 5,109.9 8,372.4 12,185.0 15,927.3 Nonrecycled portion to liquidity/security fund 108.8 268.9 440.7 641.3 838.3 Principal collection rate before rescheduling 0.75 0.75 0.79 0.80 0.81 Prncipal collection rate after rescheduling 1.00 0.94 0.92 0.90 0.89  ANN-EX 4 - 66 - Page 12 of 1 Scenario 2 Table 3 SRI LANKA Page 3 of 5 POVERTY ALLEVIATION PROJECT CREDIT FIJND Projected Income Statement (US$ '000) Year I Year2 Year3 Year4 YearS Interest income from Credit Fund Operations @ 7% 203.0 501.6 777.7 1,118.1 1,440.9 less: Interest expense on project loan @ 3 % (Capitalized during 5 yr grace period) 44.3 165.7 335.4 546.9 800.1 Net interest income from credit operations 158.7 335.9 442.3 571.2 640.8 less: Provision for loans in arrears: % of annual disbursements assumed going into arrea 0.0 5.0 7.0 9.0 10.0 Amount of provision 0.0 358.3 777.7 1,437.5 2,058.5 Net Icome from Credit Fund Operations 158.7 (22.4) (335.4) (866.3) (1,417.7) Interest income at 10% of average balance: Credit Risk Fund 300.0 530.0 583.0 641.3 705.4 Liquidity/security fund 5.4 18.9 40.9 73.0 114.9 Net Finncial Income 464.2 526.5 288.5 (152.1) (597.3) Staff costs 22.4 24.4 33.0 35.0 36.6 Other operating expenses 60.2 73.0 88.4 104.4 109.4 Depreciation expense @ 25% 0.0 0.0 0.0 0.0 0.0 Nt Income (or Loss) 381.6 429.1 167.1 (291.5) (743.3) GOSL grants received 3,000.0 2,000.0 0.0 0.0 0.0 Other grants received 0.0 0.0 0.0 0.0 0.0 Change in Net Worth 3,381.6 2,429.1 167.1 (291.5) (743.3)  ANNEX 4 - 67 - Page 13 of 14 Scenario 2 Table 4 SRI LANKA Page 4 of 5 POVERTY ALLEVIATION PROJECT CREDIT FUND Projected Balance Sheet (US$ '000) Year I Year 2 Year 3 Year 4 Year 5 ASSETS Cash (60 days' staff cost & other oper. expenditures) 13.8 16.2 20.2 23.2 24.3 Credit Fund: Cash available for lending 2,481.8 6,76.7 10,186.4 15,106.7 20,322.1 (balancing figure from sources & uses) Accrued interest (assumes all due 31 Dec) 0.0 0.0 0.0 0.0 0.0 Unrescheduled loans outstanding 0.0 0.0 0.0 0.0 0.0 Rescheduled loans 725.0 2,154.1 3,673.3 5,381.8 7,142.8 Gross Credit Fund 3,206.8 8,330.8 13,859.6 20,488.5 27,464.9 less: Allowance for loans in arrears 0.0 0.0 0.0 0.0 0.0 (assumes provision = new arrears) Net Credit Fund 3,206.8 8,330.8 13,859.6 20,488.5 27,464.9 Liquidity/security fund 108.8 377.7 818.3 1,459.7 2,297.9 Credit Risk Fund 3,000.0 5,300.0 5,830.0 6,413.0 7,054.3 Total Financial Assets 6,329.3 14,024.7 20,528.2 28,384.4 36,841.5 Miscellaneous Assets 0.0 0.0 0.0 0.0 0.0 Fixed Assets 51.9 51.9 51.9 51.9 51.9 less: Accumulated depreciation 0.0 0.0 0.0 0.0 0.0 Net fixed assets 51.9 51.9 51.9 51.9 51.9 TOTAL ASSETS 6,381.2 14,076.6 20,580.1 28,436.3 36,893.4 LIABILITIES Current Liabilities Accrued interest expense 0.0 0.0 0.0 0.0 0.0 (assumes all due 31 Dec) Other accruals and payables 3.4 4.1 5.1 5.8 6.1 (15 days staff costs & other op exp) Current portion of project loan 0.0 0.0 0.0 0.0 0.0 (assumed paid on 31 Dec) Total Current Liabilities 3.4 4.1 5.1 5.8 6.1 Long Term Liabilities Project loan from GOSL 2,951.9 8,096.2 14,261.9 22,197.3 31,144.2 Capitalized interest 03%, 5 yrs grace 44.3 165.7 335.4 546.9 800.1 less: Current portion (paid 31 Dec) 0.0 0.0 0.0 0.0 0.0 Outstanding project loan balance 2,996.2 8,261.9 14,597.3 22,744.2 31,944.3 TOTAL LLABILITES 2,999.6 8,266.0 14,602.3 22,750.0 31,950.4 NET WORTH GOSL grants 3,000.0 5,000.0 5,000.0 5,000.0 5,000.0 Other grants 0.0 0.0 0.0 0.0 0.0 Retained enings 381.6 810.6 977.8 686.3 (57.0) Total Net Worth 3,381.6 5,810.6 5,977.8 5,686.3 4,943.0 TOTAL LIABILITIES AND NET WORTH 6,381.2 14,076.6 20,580.1 28,436.3 36,893.4  A-MX 4 - 68 - Page 14 of 14 Scenario 2 Table S SRI LANKA Page S of 5 POVERTY ALLEVIATION PROJECT CREDIT FUND Projected Sources and Uses of Funds (US$ '000) Year I Year 2 Year 3 Year 4 Year S PAt I Funding of Foundation AdminitraUion Net Income from Credit Fund Opertions 158.7 (22.4) (335.4) (866.3) (1,417.7) Interest capitalized on project loan during 5 year grace period 44.3 165.7 335.4 546.9 SOD. I Credit Risk Fund interest income 300.0 530.0 583.0 641.3 705.4 LIquidity/security fund interent Income 5.4 18.9 40.9 73.0 114.9 itor (deor) in other acoruals & payablea 3.4 0.6 1.0 0.8 0.3 less: Increas (decrease) in Misc. Assets 0.0 0.0 0.0 0.0 0