Rwanda - Food Security and Social Action Project

C_. - /'i ~A- , (Vk d - Document of The World Bank FOR OFFICIAL USE ONLY MICROFICHE COPY Report No. 10373-.RW Type: (SAR) Report No.10373-RW BAUDOUI,J./ X34030 / J-7223/ AF3PH STAFF APPRAISAL REPORT RWANDA FOOD SECURITY AND SOCIAL ACTION PROJECT MAY 26, 1992 Population and Human Resources Operation Division South Central and Indian Ocean Department Africa 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.  CURRENCY EOUIVALENT Currency Unit Rwandese Franc (FRW) US$1.00 = FRW 120 (January 1992) WEIGHTS AND MEASURES Metric System GOVERNMENT OF RWANDA - FISCAL YEAR January Ist - December 31st ABBREVIATIONS BCP Bureau de Coordination du PNAS (PNAS Coordination Office) CATGA Cellule d'Appui Technique A la Gestion des Aides Alimentaires (Food Aid Management Support Unit) EEC European Economic Community ILO International Labor Office LBWr, Labor-Based Works MINAGRI Ministry of Agriculture MIJEUMA Ministry of Youth and Associative Movement MININTER Ministry of Interior MINIPLAN Ministry of Planning MINEPRISEC Ministry of Primary and Secondary Education MINISANTE Ministry of Health MINITRAPE Ministry of Public Works MINITRASO Ministry of Labor and Social Affairs NGO Non Governmental Organisation PAIB Programme d'Appui aux Initiatives de Base (Micro- Enterprise Promotion and Development) PEP Public Expenditure Program PIC Programme d'Investissement Communal (Communal Investment Program) PIP Public Investment Program PNAS Programme National d'Action Sociale (National Social Action Program) PSAN Programme de Securite Alimentaire et Nutritionnelle (Food Security and Nutrition Program) PSTP/HIMO Projet Special de Travaux Publics A Haute Intensite de Main-d'Oeuvre (Labor-Based Civil Works Pioject) SDA Social Dimension of Adjustment UNDP United Nations Development Program UNICEF United Nations Children's Fund USAID United States Agency for International Development WFP World Food Program This report is based on the fundings of an appraisal mission which visited Kigali in December 1991. Mission members included Jacques Baudouy, mission leader; Eileen Murray, financial analyst; Ghislaine Delaine, stastitician; Qaiser Khan, economist; Marco Ferroni, economist; Kees Tuinenburg, food security specialist; Sven Hertel, civil engineer; and Harley Johnson, community-development specialist. The report was drafted by J. Baudouy with the assistance of E. Murray, G. Delaine and Q. Khan. The peer reviewers are Tonia Marek, Harry Walters, Juan Gaviria and Steen Jorgensen. The report was processed by Ms. V. Fauconnier. Messrs. Francisco Aguirre-Sacasa and Alain Colliou are the Department Director and the ranaging Division Chief, respectively, for the proposed operation.  FOR OFCIAL USE ONLY RWANDA: FOOD SECURITY AND SOCIAL ACTION PROJECT TABLE OF CONTENTS Page No. Basic Data ............................................. i Definitions . ........................................... i Credit and Project Summary ................................. iii I. BACKGROUND ............................ .............. -1- A. Introduction ........................................ . 1- B. The Economic Situation and the Structural Adjustment Program . ........ - 2 - II. POVERTY IN RWANDA AND BANK'S ROLE. . .-.................... 3- A. Poverty in Rwanda . .................................... 33- B. The Government's Social Policies and the Bank's Role . ............. - 7 - II. THE PROJECT . ......................................... -. 9- A. Project Objectives and Design ........ ..................... -. 9- B. Project Description . .................................. . 10 - IV. PROJECT IMPLEMENTATION ..............-................. 25 - A. Status of Project Preparation ............................. -25 - B. Project Management .................................. -25 - C. Project Monitoring and Reporting ......................... -28 - V. PROJECT COSTS AND FINANCINi ......-...................... 30- A. Project Costs . ..................................... - 30 - B. Financing ......................................... - 32 - C. Disbursement . ..................................... - 33 - D. Procurement . ...................................... - -35 - VI. BENEFITS AND RISKS ......... .......................... -. 39 - VI. ASSURANCES AND RECOMENDAON . .................... 41- 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.  (continued) Tables in Main Text Table 2.1 Social Safety Net Table 3.1 Labor Share for Various Types of Works Table 5.1 'Cost Summary by Category of Expenditure Table 5.2 Cost Summary by Component Table 5.3 Financing Plan by Disbursement Category Table 5.4 Disbursement Categories Table 5.5 Procurement Arrangements Table 5.6 Procurement of Works Table S.7 Procurement Implementation Schedule Annexes Annex I Social Safety Net Annex II Food Aid and Nutrition Annex mI : Labor-Based Works Annex IV Micro-Enterprise Promotion and Development Annex V Social Surveys Annex VI Terms of Reference for the BCP Annex VI : Technical Assistance Annex VmII Training Program Annex IX : Project Performance Indicators Annex X Supervision Plan Annex XI Costs and Financing Annex XII Schedule of Disbursement Annex XmII Selected Documents in the Project File Chart I Organizational Chart of LBWs Chart II Organizational Chart of PAIB Chart mI : Organizational Chart of Poverty Monitoring Chart IV Structure of the PNAS M^iz : iBRD No. 23684  BASIC DATA Rwanda Sub-Saharan Africa GNP per capita (1990) USS 300 US$ 340 Population size (1991) 7.3 million 480 million Rural areas 93% 72% Urban areas 7% 28% Estimated Population growth rate in 1991 3.6% 3.2% Population density 280 per km2 20 per km2 Total fertility rate 8.5 6.5 Contraceptive prevalence rate 10% 6 % Active work force (1989) 3.2 million N/A Agriculture 2.7 million Formal sector 0.2 million Informal sector 0.3 million Dependency ratio 102.4% 97.6% Adult illiteracy rate 53% 52% Enrollment primary school 64% 67% Life expectancy 49 years 51 years Infant mortality rate (1990) 120 per thousand 107 per thousand Under-five mortality rate (1990) 201 per thousand 179 per thousand Food production per capita (87/89) 77 95 (100 = 1979/81) Daily calorie supply (% requirements) 85 96 Underweight under-fives 30% 29% Newborns with low birthweight 20% 16%  - ii - DEFINUIN Dependency Ratio: The ratio of the economically dependent part of the population to the productive part, arbitrarily defined as the ratio of the young (those under 15 years of age) plus the elderly (those 65 years of age and over) to the population in the "working 'ges" (those 15 to 64 years of age). Infant Mortality Rate: The number of deaths of infants under one yeat old in a given year per 1,000 live births in that year. Life Expectancy at Birth: The average number of years a newborn would live if current age-specific mortality rate trends prevailing at the time of birth were to continue. Rate of Population Growth: The rate at which a population is increasing (or decreasing) in a given year due to natural increase and net migration expressed as a percentage of the base population. Total Fertility Rate: The average number of children that would be born alive to a woman (or group of women) during her lifetime if during her child-bearing years she were to bear children at each age in accordance with prevailing age-specific fertility rates. Communes: It designates both the smallest territorial division of the country (there are 145 communes) and its local government which has an independant legal status and financial autonomy. The commune is headed by a Bourgmestr. Prefectures: It designates the administrative structure at the regional level. Rwanda is administratively divided into 10 Rrefectures and 22 Msou grffectures (to serve the communes further away from the prefectures).  - iii - CEIT AND PROJECr SUMMARY Boffower : Republic of Rwanda Benefigiary : Not applicable Credit Amou t SDR 14.0 million (US$19.1 million equivalent) Trms : Standard IDA termns, with 40 years maturity Eroect Description: Qbjectives: (a) to improve the food srcurity and social welfare of the poorest popu!ation groups; (b) to hw,rove the government's capability to monitor living standards of the population; and (c) to initiate a long-term poverty alleviation strategy. Target groups include- (a) individuals with insufficient income-earning capabilities, suc.h as landless farmers, poor female-headed households, AIDS-afflicted families, orphans and others (annual target: 75,000 people); (b) farmers faced with acute food insecurity (annual target can vary from 0 to 300,000 people); and (c) farmers with land smaller than 0.2 hectare which will need off-farm employment to survive (annual target: 25,000 workers and 125,000 dependents). In addition, potential micro-entrepreneurs will be the target of a micro-enterprise development component. The Food Aid and Nutrition com,nn1o would aim at reducing food insecurity and malnutrition through: (i) food aid purchased locally or regionally and targeted at vulnerable groups (children and pregnant/lactating women from poor families, AIDS-afflicted households, orphans) in association with a package of nutrition activities (education, monitoring, micronutrient supplementation); (ii) coordinated emergency food relief to respond to local acute food insecurity; and (iii) support to the development of a national food policy. Tbe Labor-Based Works a&BWs) component would increase the purchasing power (and thus contribute to improving food security) of the population of the 43 poorest communes in the country through community-based works including: (i) communal road rehabilitation and maintenance; (ii) anti-erosion and land improvement measures, such as reforestation and agricultural terracing; and (iii) other public works of public interest. The Micro-enterprise Promotion and Develm nt would: (i) promote the creation of profitable micro-enterprises in rural areas as a mean of generating productive off-farm employment; (ii) identify potential entrepreneurs through a community-based network; and (iii) facilitate the creation of these micro-enterprises by appraising their viability and assisting entrepreneurs in obtaining financing from existing financial institutions. Finally, *he Social Surveys component would aim at strengthening the government's capacities to: (i) monitor the living standards of the population through socioeconomic surveys and studies; and (ii) elaborate relevant social policies and strategies.  - iv - Beniefit and Risks: The proposed operation would provide rapid support to poor families in need of food and additional income. In the longer term, the project would contribute to a better monitoring of the living standards of . pulation, and overall social policy development. Indirect benefits will be generat.. tY the LBWs and the micro-enterprise programs which would contribute to demonstratiL, the efficiency of switching ca;al for labor and the potential for off-farm employment, respectively. The main risk is that implementadon delays occur because of inefficient project management. This risk will be addressed by: (a) establishing an adequate project management structure prior to credit effectiveness; (b) decentralizing the decision making process; and (c) involving beneficiaries in the monitoring process. Another significant risk is insufficient budgetary support from the gevermnent. This risk will be addressed by a close monitoring of the agreed-upon public spending on social sectors. Project Cost Estimates ....... USS million ----- Es 2mataed Cot Local Foren Total Food Aid nd Nutrition 8.1 5.9 14.0 Labor Based Works 13.4 4.4 17.8 Support to Micro-entreprises 2.7 0.6 3.3 Surveys 0.8 0.6 1.4 Project Nanagement 0.6 1.5 2.1 PPF 0.3 1.0 1.3 Total Base Costs 25.9 14.0 39.9 Physical Contingencies 1.8 0.7 2.5 Price Contingencies 2.7 1.0 3.6 Totat Project Costs 30.4 15.7 6.1 Financine Plan IDA 12.5 6.6 19.1 UNDP 1.0 2.1 3.1 WFP 8.9 7.0 15.9 Governnent *8.0 --- 8.0 TOTAL 30.4 15.7 **".1 * Part of the government's contrIbution wilL be covered by the Unified Counterpart Fund (pool of counterpart funds adininistered by the Treasury) and by the Uniffed Road Fund (wenaged by NINITRAPE) for conimmtl road mintenance. ** Excluding USS2.3 miLlion equivalent for taxes and USS0.70 million equivalent for import dutfes. Estimated IDA Disbursements IDA's fiscal years f12i FY94 FYff FY9 FY9Z Anmual 1.5 4.4 6.0 4.8 2.4 Cumulative 1.5 5.9 11.9 16.7 19.1  RWANDA FOOD SECURITY AND SOCIAL ACTION PROJECT ETAFF APPRAISAL REPORT I. BACKGROUND A. Introduction 1.1 Wi4 a GNP per capita estimated at US$300 in 1990, Rwanda belongs to the group of the world's poorest countries. Rwanda is also one of the smallest African countries (26,000 km2). The country is confronted with severe development constraints: (a) the population density is the highest in Africa (280 per km2), which is having a negative impact on agricultural land availability and the environment; (b) the population growth rate (estimated at 3.6 percent in 1991) and the population momentuml/ are high and put pressure on the country's economic growth and investment capabilities; (c) foreign exchange earnings are excessively dependent upon coffee; and (d) its landlocked situation makes external trade dependent upon two long and expensive routes from the Indian Ocean. In the 1970s, thanks to politicai stability, high coffee prices and prudent economic and fiscal management, Rwanda enjoyed high rates of economic growth (5 percent per year on average) together with budget surpluses, large external reserves and a low debt service ratio. These were significant achievements in such a difficult environment. In the early 1980s, the economic situation started to deteriorate because of unfavorable terms of trade and excessive and inefficient public spending. This decline was precipitated in 1987 with the fall in world coffee prices. 1.2 In early 1990, the government requested the Bretton Woods institutions' support in preparing an adjustment program. While in the midst of opening its economy, Rwanda was invaded in late 1990 by armed opponents to the regime based in Uganda. Although the invasion was repelled, clashes with these opponents continue to occur. The low intensity fighting - coupled with depressed world coffee prices - have had a severe impact on economic activity; the government has increased secu-ity-related annual outlays to about 7.0 percent of GDP, from an annual average of about 2 percent over the past decade. 1.3 To complement economic liberalization and respond to domestic and external pressure, the government has also initiated a political reform process which has continued and intensified despite the clashes with the rebels. A new constitution was adopted in June 1991 which opened the way for multi-party politics. The political transition process is under way and an outspoken press has emerged. 1.4 Rwanda's economic difficulties have a negative impact on the social situation. Poverty has increased in the past years, as revealed by a recurrent state of acute food insecirity in the southern and western parts of the country caused by a combination of insufficient food production and purchasing power of the population. Public funding is insufficient to respond adequately to the demand for social services. In education, the government has given high priority to increasing primary education enrollment rates, but providing the necessary budgetary resources becomes more difficult because of budgetary I/ The current population is estimated at 7.3 million. Would fertility remain constant, the population size would reach 16 million by 2010.  - 2 - providing the necessary budgetary resources becomes more difficult because of budgetary constraints. In health, the situation has deteriorated with the combination of a worsening health environment (AIDS epidemic, expanding malaria endemic and increasing malnutrition rates) and insuffwient budgetary allocations to the public health system. It is only recently (June 1990) that the government has formally issued a national population policy, therefore recognizing the need for addressing one of the nation's most critical development constraints. B. The Economic Situation and the Structural Adjustment Program 1.5 The economic situation of Rwanda has been deteriorating rapidly during the past five years for three main reasons: (a) world coffee prices have fallen, reducing foreign exchange availability; (b) agicultural production has Ieclined (agricultural output in 1990 was 9 percent lower than in 1987), mostly because of an increasing disequilibrium between agricultural land availability and population pressure; and (c) the government has failed to implement relevant economic policies. Real GDP declined sharply in 1989 (5.8 percent) and again in 1990 (1.2 percent). These negative trends had already started in the early 1980s when the government became more interventionist in the economy to try to compensate for unfavorable terms of tra 4e: price controls were imposed, public expenditures (including civil servant salaries) were reduced, and local enterprises were protected by import restrictions. Restrictive import licensing was introduced to respond to foreign exchange scarcity and caused a slow down in many activities. The government, however, did not correct the highly overvalued exchange rate (one important cause of the lack of export competitiveness) and did not attempt to reduce excessive regulation and control over private enterprises. In view of the continuing deterioration of the situation, the government finally decided in early 1990 to prepare an adjustment program with the help of the Bank and the IMF. 1.6 The economic reform program described in the 1991-93 Policy Framework Paper includes: (a) promoting private sector activities and international competitiveness; (b) improving public resource management; (c) developing human resources; and (d) improving the management of the country's natural resources. In November 1990, the government began implementing this program, starting by one of the most politically difficult measures - the devaluation of the Rwandese franc (67 percent in local currency terms, 40 percent vis-a- vis the SDR). Thereafter, the government took several measures to promote the private sector's involvement in the economy by eliminating price controls and certain cumbersome administrative procedures for enterprise creation. In the agricultural sector, which will remain the mainstay of the economy for, at least, the next generation, self-safficiency strategies are being replaced by specialization and production for both external and internal markets. In parallel with the development of these measures, the government has shown a concern about the social situation and the effects of the adjustment measures. Starting in August 1990, the government has worked with Bank staff and UNDP to design a social safety net aimed at responding to basic needs of the population, particularly the poor. 1.7 Economic Developments in 1991. The economic situation continued to deteriorate in 1991 (GDP decline of 2.9 percent), mainly because of the war situation. Trade has been severely affected by the closure of the route from Mombasa through Uganda. Transport is hindered by security-related traffic restrictions within the country. Private investments are minimal because of the lack of confidence vis-&-vis the country' s overall situation, contributing to poor supply response. Tax revenue fell by at least 10 percent because of the slowdown in economic activities and lower imports. In consequence, the fiscal deficit rose to 14.8 percent of GDP (grants excluded) and was largely financed by  -3- accumulating arrears through the domestic banking system. In the external sector, the current account deficit remained at 13 percent of GDP and the external debt rose to about 31.5 percent of GNP. As noted, the current fiscal imbalance was associated largely with war and one of the most serious problems was the crowding out of civiliQi expenditures by military expenditures now accounting for 42 percent of the budget. The Social Safety Net which had been incorporated in the adjustment program was adversely affected (e.g., spending on essential drugs was only 25 percent of the budgeted allocations). If these trends continue (civilian expenditures have decreased in real terms by 33 percent over the past two years), the social sector situation will further deteriorate In the coming years. II. POVERTY IN RWANDA AND BANK'S ROLE A. Povet in Rwanda Social Indigators 2.1 Social indicators are strongly determined by the poverty level of a country and Rwanda's indicators are slightly below the average for Sub-Saharan Africa (see basic data, page (i)). Health indicators are poor: life expectancy has not yet reached 50 years; under- five mortality rate is still around 200 per 1,000 and the pace of infant mortality reduction is one of the slowest in Africa. The malnutrition rate among young children is estimated at about 30 percent using international standards for weight and age. In many cases, malnutrition starts at birth because of the poor nutritional status of many mothers: about 20 percent of newborns are underweight. Education indicators are about average for Sub- Saharan Africa but have alrA.ady shown signs of deterioration because of an increasing discrepancy between the huge increase of school-age population and limited financial resources. If recent trends were to continue, the primary education enrollment rate would decline from the current 62 percent to about 50 percent within the coming 10 years. Povery Profile 2.2 A poverty profile for Rwanda has been prepared/ on the basis of the national 1983-85 household consumption survey. The premise of the study is that, while Rwanda is a poor country by international standards, certain groups are poorer than others and they have a special claim on targeted assistance to raise their income and consumption possibilities. These groups either do not have the necessary means for adequate physical and mental survival, or they are at considerable risk of falling below that level of entitlement, temporarily or for a longer period owing to fluctuations in their income. While the results of the study are still preliminary, certain interesting elements have already emerged as described below. 2.3 Total real expenditure per capita has been used as a proxy for the standard of living of the households. The "poor" have been defined as the bottom 40 percent of the sample, and the "hard-core poor" as the bottom 20 percent. The incidence of poverty is much greater in the rural areas than in urban areas as shown by the large gap between urban 2/ Tbis poverty profile has been prepared by the Technical Department of the Africa Region (Social Policy Division) as an input for a more comprehensive poverty assessment.  - 4 - and rural expenditure levels (urban expenditures are 2.4 times higher). The rural areas represent more than 98 percent of all poverty in Rwanda, with the South Central zone having the highest incidence of poverty. 2.4 Labor and land are the two most important assets of the poor and their variations have a strong correlation with the level of poverty. The dependency ratio is 1.14 for poor households and 0.87 for the non-poor, which means that poverty is more common among large households. In rural areas, the incidence of poverty is greater among female- headed households but the difference is small (the difference, however, tends to increase with the depth of poverty). By contrast, in urban areas female-headed households tend to be much poorer than their male-headed counterparts (but still better off than their rural counterparts). With regard to land, the association between farm size and poverty is negative (as expected) in spite of smaller farms having a larger per-hectare productivity. 2.5 Income Pagms. Agricultural production accounts for over 70 percent of the income 0f poor households and for 52 percent of that of non-poor households. While policies which strengthen agriculture production will contribute to alleviating poverty, a sizeable portion of new agricultural employment is likely to be absorbed by the non-poor. This applies to any employment creation in rural areas. It would be, therefore, necessary to target employment (through a combination of types of works, level of wages and poverty criteria) in order to absorb a higher proportion of poor. 2.6 Other important findings include: (a) the piedominance of food produced for own-consumption; (b) the food crop sales (including banana beer) significantly higher than cash crop sales; (c) the share of cash crops (including coffee) representing only 7 percent of agricultural production; and (d) the minor role of purchased inputs. These patterns do not vary greatly between poor ard non-poor households, but illustrate the difficulties of promoting cash crop production in Rwanda. 2.7 Expenditure Patterns. The poor (arbitrarily estimated at the bottom 40 percent of the population as previously mentioned) account for about 25 percent of total expenditures and, on average, food represents 70 percent of household expenditure. Non-food purchases are significantly lower for the poor (12 percent) than for the non-poor (26 percent). The study provides also certain interesting findings with regard to the potential for food subsidy: (a) there are no 'inferior" foods; (b) there are no foods for which a subsidy would yield larger benefits for the poor than for the non-poor; and (c) there is no obvious self-targeting tood subsidy in Rwanda. 2.8 With regard to non-food expenditures, the poor account for 40 percent of health services (it means that the poor spend as much in bsolute Jms as the non-poor), and for 28 percent of expenditures on used clothing, school uniforms, tobacco leaves, and kerosene. These findings have certain implications in the context of adjustment (as related to devaluation): the main impact for the poor would be higher prices for used clothing, drugs, and kerosene. In addition, changes in the cost of health services would have a significant impact on the poor.  -5 - pFood Security 2.9 It is useful to analyze the poverty situation from a food security perspective, since food security is the most pressing concern of poor people. Until 1985, food production kept pace with population growth. Since then, population growth rates have steadily increased above those of agricultural production. New land has become scarce and per capita food production slowed down. In 1989, food production per capita was estimated at 77 percent of its 1979 ,evel and the situation has continued to deteriorate. Cash crops have been largely insufficient for providing the income necessary to an increasing number of families in order for them to purchase food. Rwanda has entered into a period of structural food deficits, not only at the national level, but increasingly so at the level of rural households. 2.10 Food security is mostly a rural problem, since 93 percent of the country's population is rural. The underlying calise of food insecurity is the fast rising population density: family holdings are becoming too small for making a living solely out of agriculture. The government has estimated that one-third of tilF rural population (some 2 million people) had plots which were too small to cover 100 percent of household needs through agricultural revenues. This percentage will increase over time. While the new agricultural strategy will aim to enhance productivity through a number of policies, intensification, specialization and production for the market, the solution to rural food security will have to be found through the creation of off-farm employment. In the meantime, special transfer programs will be necessary to prevent increasing malnutrition and famine. 2.11 Because of the increased population pressure, the agricultural system has become so vulnerable that even relatively minor shocks such as erratic rainfall, pests, plant diseases and reduced cash income from coffee have a disproportionately large impact. In 1989, these shocks led to a situation where over 300,000 rural people were estimated to be in need of assistance. For 1991, an identical number of people have been estimated to be at risk of acute food insecurity (most of them in the prefectures of Gikongoro, Butare and Kibuye). Two categories of food insecure have been identified: the chronically food insecure, e.g. the old and the handicapped, as well as the landless; and the transitory food insecure, suffering from a lack of food and/or income during the periode de soudure3/. Both categories receive assistance from certain NGOs (such as CARITAS, an international Catholic NGO specialized in social programs and active in Rwanda for many years), the government (through occasional fond aid provided by bilateral donors and the World Food Program (WFP)), and from relatives, neighbors and friends. However, the latter form of assistance (solidarity-based) is becoming increasingly difficult because the majority of the population is suffering from the deteriorating economic situation. In any case, for the foreseeable future, rural Rwanda will be regularly confronted with the effects of structural food insecurity at the household level, which will be particularly felt during the eriode de soudure. The threat of famine (and actual famine if no action is undertaken) is likely to become a regular phenomenon in rural Rwanda that the government has to address in a systematic way. 2.12 Two new factors increase food insecurity in both rural and urban areas. The first factor is the increasng number of AIDS-related deaths. Rwanda's AIDS epidemic is one of the worst in Africa: in Kigali, about 25 percent of adults between the ages of 15 and 45 are I/ The eriode de soudur is the time between two harvests when food reserves can become insufficien. at the family level.  seropositive; in rural areas it is estimated that 3 percent of adults in the same age group are infected. Many families in urban as well as in rural areas are losing income earners killed by this disease and have difficulties purchasing food. Orphans are becoming numerous and are at high risk of food insecurity if they are not adopted. It is difficult to assess the overall economic impact of AIDS on the national economy because of the lack of data. It is likely, however, that this impact will become increasingly significant because AIDS kills people in their most productive years and tends to strike better-off families (even in rural areas). The second factor with a negative effect on food security is the loss of employment caused by the contracting economy particularly in urban areas where a significant number of small enterprises are closing down and laying off their employees. EmPloyment and Labor Market 2.13 With the current population growth rate, the number of new entrants in the labor market will increase from 130,000 per year in 1990 to 360,000 per year in 2015. It will be difficult for the Rwandese economy to absorb this workforce because the agricultural sector is already saturated and the development of off-farm employment will take time. The industrial sector consists mainly of small-scale agro-industries; its expansion will necessitate adjustment measures and one cannot expect major job creations in the near future. The development of small and micro-enterprises is hampered by: (a) excessive government regulations and insufficient promotion; (b) lack of easy access to credit; and (c) poor domestic markets. Migrations from rural areas to cities where migrant workers could find opportunities in the informal sector are made difficult by a law restricting individuals' mobility. Migrations towards neighboring countries are limited either by political reasons (Uganda, Tanzania, Burundi) or by the absorptive capability of the neighboring regions (for example, the southern Kivu region of Zaire is already overpopulated). In 1991, Tanzania decided to expel about 25,000 Rwandese migrant workers whc had to be relocated hastily by the government. Similar events could occur repeatedly in the coming years, increasing the pressure on an already depressed labor market. 2.14 Rwanda has reached a point where a large, and increasing, share of the abundant labor force is not employed productively. The most recent large scale survey on employment (1988-89) shows that 72 percent of the labor force work less than 40 hours per week (the threshold usually used to differentiate part-time from full time employment). In the traditional agricultural sector (which employs 90 percent of the work force), the percentage of part-time work reaches 77 percent, a high figure by any standard. The on-going structural adjustment program should improve the situation by stimulating the creation of new employment opportunities and increasing average revenues, thus expanding local markets. In particular, the elimination of restrictive labor regulations is included in the conditions for second tranche release. But these changes will take time and meanwhile transfer programs will have to be established to alleviate the effects of this pervasive underemployment. In addition, the restructuing or closing of inefficient public enterprises will lead to laying off a limited number of workers who should be taken care of in the context of a social safety net. Social Program 2.15 The healthsystem provides basic medical coverage to about 60 percent of the population. The remaining 40 percent are either too far from any health facility for an easy access or are too poor to afford the cost of drugs and other treatments. Approximately 40 percent of basic services are provided by church-affiliated facilities where the quality of care  -7- is usually better than in public facilities (higher user charges, however, have a disincentive effect on the poor). There is no potential for a rapid increase in health coverage because public funding is insufficient to cover the recurrent cost implications of an expanding health system. Major public health programs (such as immunizations, AIDS control, and malaria control) are heavily subsidized by external agencies. The Prima Education System is effective (including good female enrollment), but the primary enrollment rate is likely to decline because of increasing population pressure and insufficient public funding. The poor are affected by the cost of school uniforms which represent a significant expenditure in small household budgets. There is neither a social security system nor a national health insurance system in Rwanda. Civil servants are partially covered by the government for health costs, disability and pensions. Certain private employers cover employees' health expenses but there is no specific legislation to this effect. Unemployment is not covered by any program. 2.16 Most welfare services are provided by NGOs (more than one hundred have been listed), including religious organizations. Their activities are usually small scale and localized as part of broader development programs (such as intersectoral rural development projects). CARITAS is an exception, with food aid carried out in numerous communes and support programs for arphans and AIDS-affected families. A few NGOs (such as IWACU and ARDI) provide training and technical assistance to other NGOs. There has been some attempt to coordinate the activities of NGOs (such as the creation of the Conseil des Qrganismes Dour les Apis aux Initiatives de Base which regroups a dozen NGOs) but most NGOs prefer to work independently and tend to compete for foreign aid rather than collaborate. B. The Government's Social Policies and the Bank's Role 2.17 Early in the process of preparing the structural adjustment program, the government requested the Bank's assistance to define and prepare a social actior program. Initially conceived by the government as a broad set of activities aimed at alleviating the negative social effects of certain adjustment measures, the concept of the program has evolved towards a comprehensive social policy package with a three-pronged strategy: (a) establishment of a social safety net including: (i) the strengthening of core programs in education, health and population of piblic spending on education, health population and a national social action program (Proframme National d'Action Sociale - PNAS) targeted towards the poorer segments of the population (this program is supported by the proposed operation); (b) resource mobilization to ensure adequate funding of this core social expenditure program from both external aid and domestic budgetary resources (including counterpart fuinds); and (c) close monitoring of the financial and physical implementation of the social expenditure program. The government has summarized this approach in a Policy Statement on human and social development policy which was finalized with IDA during the negotiations of the proposed operation (para. 7.1) and which wiUl serve as a reference during project implementation. The social sector strategy implies a close coordination with the structural adjustment process, particularly with regard to the preparation and monitoring of the core social expenditure program. The programming process for public expenditures on the social sectors has been integrated into the Structural Adjustment Program as a part of the component aimed at improving public sector resource management. This process has two main elements: (i) a review of the public investment program to eliminate all investwmnts inconsistent with sectoral policies or economically and financially inefficient; and (ii) adequate budgetary allocations to cover operating costs of essential development programs, such as the social safety net. The major constraint to adequate budgetary allocations for social sectors are  - 8 - the competing requests from other sectors, including the army which has become particularly expensive due to the recent military conflict. In order to avoid under-funding of social programs, the annual review of social sector public expenditure programs will be a covenant of the proposed operation. During negotiations, agreement was reached with the Government that: (a) the social sector Public Expenditure Program (PEP) for the next three years will be submitted to IDA for review and approval no later than October 31 of each year of project implementation (para. 7.1 (a)); quarterly audit reports on actual non-salary spending of the ministries of health (including spending on population activities), primary and secondary education and planning (for social expenditures only) will be prepared by independent auditors and submitted to IDA within four months following the end of each quarter (para. 7.1 (b)); and (c) the implementation review of the Social Sector PEP for the past year will be submitted to IDA for review and discussion no later than June 30 of each year of project implementation (para. 7.1 (c)). Preliminary estimates of the 1992-94 public expenditures for the Social Safety Net are summarized in Table 2.1 below. A more detailed analysis is provided in Annex I. During negotiations, agreement was reached with the government that the approval by IDA of the 1993-1995 social sector public expenditure program will be a condition of credit effectiveness (para. 7.2 (a)). The monitoring of physical implementation of the social safety net will be done at the sectoral level, with the support of the relevant IDA-financed operations. Table 2.1: 1990-94 Social Sector Public Expenditures US$ million (1991 price) ACTtAL PROJECED 1990 1991 1992 1993 1994 PRIMARY & SEC. EDtATION 74. 8 73.9 773 79.0 - hwivt 21.0 18.3 18.2 18.3 18.3 Reauvrat 53.8 54.6 55.7 59.1 60.7 HEALTH 27. 22L 263 3L7 31.6 h-vedm 17.8 12.7 8.4 6.8 5.6 - Recum 9.7 10.0 18.0 24.9 26.0 POPULAlON S.3 7.7 6.7 3.3 3X4 - iv_" 4.0 6.2 5.4 2.3 2.4 1.3 1.5 1.4 1.0 1.0 PROPOSRD PROJECF 0.0 0.0 .S 10.S 10.7 .wadment 0.0 0.0 0.5 9.5 9.2 -ReowTt 0.0 0.0 0.0 1.0 1.5 TOTAL SOCL4L SAFETY NT 107.6 1X3. 107.6 122 124 . IsvedmmS 42.8 37.1 32.5 36.8 35.5 - Roent 64.7 660 75.1 860 89.3 TOTAL GOVERNME P 359.4 42.1 392 40e.8 414A. e vhA 100.8 141.8 165.0 174.4 175.0 - Pgauwut 258.6 290.3 227.1 231.4 239.4 SAFETY NET AS% ABOVE 29.% 239% 27.4% 30.3% 30.1% -v 42.5% 26.2% 19.7% 21.1% 20.3% R _aurr 25.0% 22.76 3.1% 37.2% 37.3% PRO3ECf AS % GOVT. EXPEND 0.1% 2.6% 26%S kbi_mM 0.3% 5.4% 5.3% -Rg b_ 0.0% 0.4% 0.6% PROJECr AS SAFlTY NET 5% L.S% 8.6% k g br._1.5% 25.7% 25.9% - Rseuit 0.0% 1.2f [ 1.7% Sg Am 1 Mhdm (mbw M1991)  -9 - 2.18 The proposed operation is an integral part of IDA's strategy in ltwanda and complements not only the Structural Adjustment Credit but also several investment operations, including: a Population Project (FY91); an Education Sector Project (FY91); and an ongoing Family Health Project (FY87). In addition to these social sector projects, several other IDA- supported operations are likely to have a positive impact on the social situation: an Agro- Industrial Sector adjustment operation tappraised in FY92) which aims at increasing agricultural revenues and a Financial Sector adjustment eperation (negotiated in FY92). which will help the creation of private enterprises and increase employment. The proposed operation has been designed in such a way that it does not duplicate other projects but complement them and fill the gaps in reaching the poor. IDA will also assist the government in promoting aid coordination and encouraging co-financing. m. THE PROJECT A. Project Objectives and Design 3.1 The proposed operation would support the national social action program iProgramme National d'Action Sociale - PNAS) which is an integral part of the social safety net established as part of the structural adjustment process. As conceived, the PNAS combines three essential elements: (a) a comprehensive transfer program aimed at providing both rapid consumption transfers (mostly food) to the extremely poor and employment opportunities to those who are able to work; (b) a community-based investment program aimed at improving the basic infrastructure and the environment in poor rural areas; and (c) a policy element aimed at improving the development and implementation of poverty alleviation strategies. The objectives of the PNAS are to: (a) reduce food insecurity, malnutrition, and other effects of poverty among the poorest population groups of Rwanda; (b) induce policy changes with regard to the use of the most important asset of the poor, their labor force; and (c) improve poverty assessments as well as the monitoring of the implementation of the national social policy. An indirect objective of the PNAS is to improve the developmental role of the communes and their interactions with the communities. 3.2 The PNAS has three complementary operational components - Food Aid and Nutrition, Labor-Based Works (LBWs) and Micro-enterprise Promotion and Development - and a Living Standards and Poverty Monitoring component. Each operational component has specific target groups. The Food Aid and Nutrition component is aimed at: (a) individuals and families with no or insufficient income-earning capabilities, such as landless farmers, poor female-headed households, AIDS-afflicted families, orphans and others (annual target: 75,000 people belonging to this cate!ory); and (b) farmers faced with acute transitory food insecurity (annual target can vary from 0 to 300,000 people). The LBWs component is aimed at under-employed or unemployed people in rural areas (such as farmers with land less than 0.2 hectare) who will need off-farm employment to survive (annual target: 25,000 workers and 125,000 dependents). The Micro-enterprise component is aimed at potential entrepreneurs who could benefit from small credits and training to start a profitable micro- enterprise (annual target: 300 entrepreneurs, 1000 workers and 4000 dependents). The quantitative estimates, while based on cross-checking of various sources (including primary data collected during field trips), are subject to revision as more data become available. Each component has specific targeting mechanisms to reach the intended beneficiaries, such as the  - 10- establishment of a poverty index to identify the priority communes which will benefit from LBWs. These mechanisms are described in more details for each component. 3.3 They are significant differences among the components in terms of urgency, scope and degree of public sector involvement. The Food Aid component is designed to respond to immediate needs of the most vulnerable population groups and this component will benefit from large amounts of extemal aid. Tbe LBWs component is of a longer-term nature and addresses the needs of people able and eager to work, with a mix of public sector involvement (through financing and some technical assistance) and private sector participation (through the contractors). In addition, this component, while providing cash transfers to the poor, contributes to basic infrastructure development at the community level. Finally, the Micro-enterprise component has a long term purpose: its goal is to explore the potential for profitable and sustainable off-farm employment. The quantitative impact of this component will be rather limited but it could contribute to the development of an approach to job creation which would prevent a certain proportion of the entrants in the labor force in rural areas from falling into poverty. 3.4 The PNAS does not intend (nor has the capacity to) to respond to all social problems in Rwanda. The PNAS is an important element of the overall social safety net that the government wants to implement in parallel to the adjustment program. It would complement in a rational way many other actions that are carried out by line ministries (Health, Population and Education), ionor-supported rural development projects and certain NGOs. The PNAS would not replace these actions but would increase the amount of resources allocated to poverty alleviation activities. The proposed project, in addition to supporting the PNAS has an important policy dimension: the comprehensive conditionality linked to the preparation and monitoring of the social sector public expenditure program (the social safety net) would allow the government to assess on a regular basis the implementation of the national social policy and make the necessary adjustments. B. Proiect Description Food Aid and Nutrition 3.5 Background and Justification. As already indicated (paras. 2.9-12), Rwanda has entered a period of permanent risk of food insecurity. This has been demonstrated in 1989-90 by the famine situation in the southern part of the country, initially caused by a crop failure (itself caused by the combination of adverse climatic conditions, pests, plants diseases and soil acidity). The government estimated the number of people affected by this famine at 300,000. The main cause of this situation was not the absence of food (certain northern regions have excess production of food crops), but the lack of purchasing power among poor rural households. One-third of these families have plots that are too small to provide any cash income. This situation will worsen with the high population growth rate which is making land ever more scarce. In addition, the prevailing conflict in the northern parts of the country has greatdy reduced border trade, thereby further increasing the number of people requiring food aid. As already mentioned, the AIDS epidemic will also have an impact on food insecurity and malnutrition among urban and rural families as an increasing number of income earners will die of the disease. The need to reduce food insecurity in Rwanda will continue until the purchasing power of the population has increased through economic adjustment and sectoral strategies which will: (a) reduce population pressure; and (b) create more lucrative off-farm employment. IDA's contribution to this component (which will be  - 11 - mostly funded by WFP) will consist of technical assistance and operational support aimed at: (a) improving the targeting and use of food aid; (b) monitoring the social and nutritional impact of this component; and (c) developing a more effective national food aid policy, including program food aid. 3.6 Objectives and Scope. The main objectives of this component will be to: (a) reduce chronic food insecurity among vulnerable groups which include malnourished mothers and children, orphans and families of AIDS victims; (b) improve the government's ability to provide assistance to target groups during transitory periods of acute food insecurity; and (c) contribute to the elaboration and implementation of a food aid policy with a special focus on program food aid. Several guiding principles will be applieed during the implementation of this component. First, a large share of food aid will be purchased locally so as to stimulate agricultural production and avoid gluts which would have a negative impact on the domestic markets. Second, food aid will be linked to nutrition programs, including nutrition education and efficient food allocation within families. Third, an exception to the second principle will be emergency responses to famines or other disasters. Finally, food aid will be part of a comprehensive food policy in order to allow the government and donors to keep a long term perspective on key factors influencing the national food security situation. 3.7 Activities. A three-pronged strategy has been identified to reach the above objectives: (a) targeted food aid and nutrition activities for the reduction of chronic food insecurity; (b) rapid response to acute food insecurity; and (c) program food aid policy development. Most activities involved in this strategy already exist in Rwanda. They require, however, strengthening, rationalization, restructuring, targeting and intensification. lTe PNAS would contribute to all these operational aspects, as described below. (a) Alleviation of chronic food insecurity. 3.8 This sub-component is targeted to the most vulnerable groups (malnourished mothers and children, AIDS victims and orphans). It has been estimated during project preparation that a total of approximately 150,000 people were currently in need of permanent food aid under prevailing conditions in Rwanda. The proposed operation would aim at reaching about 50 percent of this target group, assuming that the remaining half is and will be covered by existing programs supported by various external agencies, including NGOs, such as CARITAS, OXFAM and CARE. This assumption will be reviewed during the mid-term implementation review of the project and the necessary adjustments will be made in collaboration with the World Food Program (WFP) which will finance most of this sub- component. The key elements of the proposed activities are summarized below. A more detailed description is provided in Annex II. (i) Targeted food aid for vulnerable groups. This program will be supported by WFP which has developed a specific project (accepted by its Governing Body in December 1991) for this purpose. The initial duration of the program will be four years and might be extended for two additional years if necessary. As previously stated, the target groups eligible for food aid include: malnourished children in the 0-30 months age group and their families; pregnant and lactating women; AIDS-afflicted families; and orphans. The first and second target groups will be reached through the existing network of nutrition centers; the third and fourth target groups would be reached  - 12 - through programs and institutions run by CAR1TAS. This project of US$15.9 million will be partly financed by food aid - wheat (8,000 MI), milk powder (1,400 MT), oil (1,000 MT) and sugar (700 MT) and partly by a cash contribution of US$7.0 million for local purchases of food (or purchasing in the region) of maize and sorghum. The wheat will not be distributed through rations but sold on the market (private mills) and the funds generated (US$1 million) will also be used for the purchasing of maize, sorghum and beans on the domestic or the regional markets. It will be distributed through the nutrition center network in a phased approach starting with approximately 60 centers selected on the basis of operational criteria established by the Ministry of Health (MINISANTE) and poverty levels of the communities they serve. The program will be progressively expanded to 160 nutrition centers over a four-year period. CARITAS will be responsible for executing the AIDS and orphans program. Part of the cash generated by the sales of imported food aid would be used to finance school tuition for children of AIDS-afflicted families. The PNAS coordinating office (Bureau de coordination du PNAS - BCP) will ensure the financing and overall monitoring of the logistics for food aid transportation and distribution. It is expected that other poverty alleviation activities of the PNAS and economic improvements induced by the adjustment program will allow permanent food aid to be strictly limited to individuals with no other means of survival. A thorough assessment of the food aid impact would be part of the mid-term review of project implementation. (ii) Nutrition activities. The Nutrition Division of the MINISANTE would be responsible for undertaking a major strengthening of the national nutrition program, including: (a) definition of a package of activities to be carried out by each nutrition center, such as: (i) growth monitoring and diagnostic of malnutrition among children and women; (ii) nutrition education and promotion of family food production; and (iii) provision of complementary measures such as dewormization and supplementation in Vitamin A, iron and iodine (as necessary); (b) the programming of training and supervision activities that would be needed to carry out such activities at the periphery; and (c) the establishment of a monitoring system. Several donors (including IDA through the on-going Family Health Project) would provide support to these activities; UNICEF and WHO have agreed to provide the necessary leadership and guidance at the operational level. Nutrition centers benefitting from food aid will be the first to participate in this nutrition program because of the expected increase in attendance. (b) Rapid Response to Transitory food insecurity. 3.9 This sub-component will address those households who face acute transitory food insecurity. Such a situation may occur because of unexpected circumstances (people displaced by war), natural disasters (hailstorms! landslides, drought) or increasingly because,  - 13 - in certain areas of the country, the balance between population and land has been jeopardized. Many farmers are now left with so little land that their precari^lus food security situation can be rapidly upset by small changes in food production and income. The famine of 1989 in the southern part of the country is an example of this. The PNAS team has prepared a rapid response system to the threat of acute food insecurity, as described below. 3.10 Early Warning System. Famine situations can be prevented more easily in Rwanda than in Sahelian countries because: (a) it is a smal! country with a good communication system; (b) there are a wide diversity of food crops and at least two harvest seasons; and (c) regional exchanges can contribute to food security. On this basis, no sophisticated early warning system is proposed but simply to strengthen data collection carried out at the commune level by MINAGRI. Data to be collected will include harvest estimates for all key products, agricultural prices and wage prices (to assess the purchasing power, a key element for food security). As soon as certain indicators show a potential risk of food insecurity, MINAGRI will convey the information to a food security task force whose constitution and role are summarized below. 3.11 The Food Security Task Force. This task force is constituted by two commissions and the Cellule d'Appui technique a la Gestion des Aides Alimentaires (CATGA). The first commission (Inter-ministerial Committee on Food Aid) is chaired by 'INIPLAN and is responsible for monitoring program food aid to the country and making projections. The second commission (Technical Committee of Emergency Food Aid) is chaired by the Ministry of Labor and Social Affairs (MINITRASO) and deals with emergency requests from communes ("project" food aid). It includes all donors involved in food aid, under the informal leadership of Canada and WFP. The CATGA has been created with Canadian support to rationalize food aid. It is responsible for collecting, analyzing and disseminating all information pertaining to food aid in Rwanda. Its mandate includes food aid related to balance of payments support (program food aid), specific projects (such as the WFP project previously described) and emergency food aid. In practice, the CATGA acts as a technical secretariat to both commissions which meet on a periodic basis or when there is an emergency situation. Management of food aid logistics and monitoring will be under the responsibility of the executing unit for the component (see para. 3.17). 3.12 The Reserve Stock. This stock would be constituted of food purchased locally with the unified counterpart fund (see para. 5.7) and would allow the first rapid response to the threat of acute food insecurity in a given area, on the basis of reports from the Ministry of Agriculture (MINAGRI)'s early warning system. The decision to intervene is taken by the MINIPLAN after a plenary session of the Food Security Task Force. The stock of approximately 4,000 tons would cover the needs of approximately 60,000 people over a one- month period. The stocks will be maintained in a few selected storage facilities run by the national food storage authority (OPROVIA). Food aid transport will be financed by the BCP (PNAS coordinating office) which will contract out to private companies for the transport of food to the prefectures. At the prefectures' level, the local authorities will be responsible for transporting the food to the communes at need (in most cases small private truckers will be used under BCP financing). At the commune level the distribution of food will be carried out by the local administration on the basis of well publicized lists of priority needy families. Communal councils, which include representatives from the churches and other NGOs will ensure the transparency of this process. Simultaneously, with the use of this food reserve, the CATGA will request external contributions on the basis of the anticipated gravity and duration of the food insecurity situation: at the minimum, a donor will be asked to replace the part of  - 14- the reserve that has been used; if the situation is more severe, a documented request stating the magnitude of the problem and the specific needs will be prepared by CATGA and transmitted to the donor community. 3.13 The Emergency Fund. In the event of a severe and durable famine situation, there might be a gap between the use of the food reserves and the arrival of additional food aid. The Food Security Task Force could then use an emergency fund of approximately US$1 million drawn from the unified counterpart fund and centrally managed by BCP. This fund would be used by WFP to buy food either in food surplus regions of Rwanda. Logistics are the same as the ones described for the food reserve. When additional food aid becomes available, the fund is reconstituted through the sale of food on the local markets. (c) Food Aid Poliey Development. 3.14 Rationale and Objectives. The need for developing a coherent food aid policy arises from: (a) the likelihood that food insecurity will increase in the coming years because of the demographic pressure and the expected slowness of economic recovery; and (b) the significant risk of negative effects of an unplanned food aid in a country with small markets. Food aid, if poorly managed, can easily compete with local and/or regional productions and thus depress further the supply response of an essentially agricultural economy. In addition, food aid can introduce new and more expensive eating habits in the population that are not sustainable on the long term. In this perspective, a food aid charter between the government and key donors would provide some guarantee that food aid would be used in a prudent fashion. 3.15 The charter would deal with quantities and prices and would stipulate that no government request would be considered unless a technical annex is appended to the request with the following elements: effective demand (on the basis of the purchasing power); national production of the food solicited; national reserves of this food; commercial imports; cross- border trade; program food aid needs; and specific amounts requested from each donor. The charter would not be considered as a binding legal document but rather as a planning tool that all parties involved agree to use. 3.16 Preparation of the Charter and Monitoring. A food aid expert will be recruited to assist CATGA to prepare the Charter. The draft will be submitted first to the technical food aid committee for inputs from the donors involved in food aid and then to the Inter-ministerial Committee for final approval. The monitoring of the use of the Charter will be done by the CATGA which will periodically report to the Food Security Task Force. The task force will use these reports to assess overall food aid needs, taking into account other interventions which contribute to food security (such as LBWs and other investment projects). Component Coordination and Monitoring 3.17 An executing unit (Bureau Executif du Programme de Securite Alimentaire et Nutritionnelle. Bureau Executif PSAN) has been created within MINITRASO as part of the institutional framework of the PNAS (see para. 4.2). This unit will be responsible for coordinating all actors involved in the component and monitoring progress at the commune level. It will work in close collaboration with the WFP office for all logistical aspects of food aid and with the BCP for all financial matters. The executing unit will include a technical assistant whose specific tasks will be to: (a) assist the WFP office in food distribution  - 15 - management; (b) establish a monitoring system to assess the impact of food aid associated with nutrition activities; and (c) follow up on the progress on food aid policy development. This technical assistant will be part of the overall TA project described in para 4.7. Labor-Based Works (LBWs) Component 3.18 BackgroWupd and Justification. The long-term goal of this component is to contribute to the promotion of labor-based methods in public works to take advantage of Rwanda's large and under-employed work force. The short-term purpose is to provide income earning opportunities for the poorest segments of the population. Through this four- year project, over 100,000 families (25,000 each year) will have access to income earning opporunities. This will provide some relief to the poorest Rwandese until the benefits of the various economic and sectoral adjustment programs (including the nascent population control program) reach them. The project will also train and contract work to Rwandese private sector enterprises and NGOs in labor-based construction. Thus, it expects to be a catalyst in switching Rwanda's construction sector to labor-based methods more suitable to the country's situation. 3.19 Labor-based public works (LBWs) in Rwanda have numerous advantages over equipment intensive works. For example, in the case of communal roads, LBWs: (a) cost less; (b) create much more employment; (c) reduce import needs significantly; (d) increase exponditures on local goods and services; and (e) benefit mostly workers from the poorest households. The above are findings from a study of the pilot special labor-based public works program (Proiet Special de Travaux Publics a Haute Intensite de Main-d'Oeuvre - PSTP/HIMO). It was carried out before the 1990 devaluation. The likelihood is that, following the devaluation, the cost advantage of LBWs over equipment intensive works has increased significantly. The only significant negative consequence of the LBWs when compared to equipment intensive works was the loss in tax revenues from imported goods such as fuel and heavy equipment. The tax system reforms which are part of the structural adjustment program are expected to reduce the heavy reliance on trade taxes on a few goods and consequently the revenue loss in switching to LBWs will become less significant. 3.20 The reasons why LBWs have failed to be generalized despite the above findings include: (a) preference among technical specialists for equipment intensive works on the grounds of quality despite evidence to the contrary from other densely populated countries such as China or India; (b) lack of private sector capacity (including NGOs) to implement LBWs; (c) shortage of the specific management and technical skills required for LBWs; and (d) lack of involvement of private sector firms in the PSTP/HIMO. Closer examination revealed that, while stating a verbal preference for LBWs, the Ministry of Public Works (MINITRAPE) uses bidding documents and procedures that have a built-in bias in favor of equipment intensive works. For example, by using fairly complex bidding procedures, it provides an advantage to the larger equipment intensive contractors rather than smaller labor- based contractors. Furthermore, design and bidding documents were often prepared by external consultants with training and experience in labor scarce economies who did not adjust their technical standards to Rwanda's particular situation. In addition, the PSTP/HIMO which  - 16 - was responsible for implementing the pilot LBWs program, did not make enough effort to train private enterprises in LBWs and work through them rather than through force account4/. 3.21 Objectives and Scope of the Proposed LBWs Program. The key objectives of this program are: (a) to finance LBWs in the poorest communes of the country (43 communes out of a total of 145); (b) to develop local private sector capacity in LBWs; (c) and to strengthen local government capacity in managing and implementing small-scale infrastructure investments. Small and medium private enterprises (SMEs) and NGOs expressing an interest in LBWs will be trained and given the opportunity to execute LBWs contracts. Local government capacity will be strengthened by supporting communes in preparing three-year rolling LBWs investment programs (a mini-PIP for communes to be called Progoamm d'Investissements Communal or PIC) and implementing them mostly through SMEs and NGOs with financial and technical support from the national LBWs program. To avoid duplication of efforts, communal roads will not be funded in those communes where the Communal Roads component of the IDA-financed Transport Sector Project is active and will use labor-based methods. A detailed description of the program is provided in Annex m1. 3.22 The LBWs component would finance US$18 million ($21 million including contingencies) of works, 90% to be spent in the 43 poorest communes distributed among the five poorest prefectures of the southern regions of Rwanda. These communes have been selected using an index of vulnerability which ranks the communes on the basis of indicators including per capita caloric availability, number of potentially food insecure people (from surveys carried out by the Ministry of Interior (MININTER)), land pressure, estimated surplus labor and number of people receiving food aid. The index does not represent a ranking of the development level of the commune but rather its needs in terms of poverty alleviation. Since new data on poverty might become available in the coming years, the list of communes will be revised during the proposed mid-term project implementation review. 3.23 Contractors (including for profit enterprises and NGOs) interested in executing LBWs would have to prove their implementation capacity, either by demonstrating reliable past experience, or by joining a training program which will be initiated before launching project activities through PSTP/HIMO with technical assistance from the International Labor Office (ILO). This training program would cover labor management (which is far more important for LBWs than for equipment intensive works), financial management and technical skills. Newly trained firms with no previous experience would initially be limited to simpler works (such as routine road maintenance) and would become eligible for all types of works as they become competent. Following the training period, newly trained firms will be given sole source contracts to be executed under close supervision of the trainer. If they successfully execute the contract, they will be given another contract which would not be supervised as closely. At the successful conclusion of the second contract, they would be included in the eligible list of contractors. Details on how works will be supervised are provided in Annex III. During negotiations, agreement was reached with the government that the training programs for the LBWs component would have been submitted to IDA for approval prior to credit effectiveness (para. 7.2 (b)). _4/ A similar ELO supported project in Ghana trained 32 contractors each capable of doing 24 km. of rural roads per year.  - 17 - 3.24 Types of Works and Expected Impact. The proposed works will consist of several sub-components such as communal roads, forestry, soil conservation, rural water supply, markets and other infrastructure important to rural communes. In addition urban services are expected to be financed. The roads program will support road rehabilitation, maintenance and construction of footpaths. The forestry program will include some reforestation and some maintenance of forests but the major share ,..: go to the rehabilitation of existing forests which have deteriorated. The soil conservation projects will support a limited number of agriculture terraces (on a demaonstration basis, jointly with pilot agricultural projects) as well as other soil conservation activities while the water supply component will finance gravity based water systems and tapping of springs. The exact nature of urban works is yet to be defined. The table below summarizes the share of labor in the total cost for each type of works. Table 3.1: Labor Share in the Total Cost for Various Types of Works Labor ___ __ __ __ __ ___ __ __ __ __ Share _CX Communat Roads 45 Forestry 80 AgricuLtural Terraces 75 Wiater Suvty 30 Ravine Control Program 50 Communal Infrastructure 45 Urban Services 50 Souroe: Apprailsa Minion (Deoenber 19911 3.25 On the basis of the allocation of funds among the different types of LBWs used in the preliminary costing, this component would assist approximately 125,000 persons per year. The proposed flexible system of project selection by communes (see following paras.) may result in a different allocation of funds among the various types of LBWs even though the total cost of LBWs is unlikely to change. The impact on employment will be higher if the proportion of funds going to forestry, terraces, footpaths, markets and urban services increase and lower if other works increase. The estimated impact on employment and share of costs going to labor is based on the assumption that the unskilled workers are paid the minimum wage (120 FRW per day) and that their productivity is similar to that of unskilled workers in the pilot labor-based works program (PSTP/HIMO). It is possible that private entrepreneurs, who are expected to execute a large share of the works, may pay their workers more and that the productivity of their unskilled labor may be higher. If that happens to be the case, the share of the total cost going to labor (and therefore the level of revenue transfer) may be higher though the expected impact in employment days would be lower.  - 18 - 3.26 Program Management and Implementation Plan. The principal objectives for the organization and implementation of this component are to: (a) strengthen the communal administrative capacity to plan and monitor the execution of LBWs; (b) use effectively technical capacities of the prefectures and the Ministries of Interior and Public Works without slowing down the implementation process; (c) involve as much and as rapidly as possible private contractors (or N30s) in the execution of the works; and (d) ensure prompt financial transfers to the workers without interference from the central administration. In order to be able to implement this component, the existing PSTP/HIMO unit (renamed Cellule HIMO) will be streizgthened both in terms of institutional capabilities and operational autonomy 5I. The Cellule HIMO will be part of the overall implementation structure of the PNAS. Through this structure, the Cellule HIMO will have full power to procure goods and services with limited control from the central administration. The Cellule HIMO will be constituted of three technical units (roads, forests and other works) and a financial unit. During negotiations, agreement has been reached with the govermnent that evidence acceptable to IDA that the Director and the four unit chiefs of the Cellule HIMO have been appointed will be a condition of credit effectiveness (paras. 4.6 and 7.2 (c)). 3.27 The selection of works at the commune level would be done by the local authorities in the form of the three-year rolling communal LBWs investment program with technical support from long-term consultants recruited by the Cellule HIMO. The consultant team will work with groups of ten communes at a time, for a period of 3-4 months. The Cellule will consolidate all the proposals from the communes in annual work programs and budgets to be submitted to the BCP which, after approving the program, will advance funds regularly to cover the costs for the next six months of work. Up to December 1992, the communes, with technical support from the Cellule HIMO, will prepare an exhaustive list of potential LBWs projects including preliminary cost estimates. Following appropriate discussions with local representatives of technical ministries particularly MINITRAPE and MINAGRI, the Communal Councils will establish their priorities among the potential projects by voting their three-year rolling investment program. The total value of the projects in the communal investment programs would be constrained by the indicative budgetary allocation prepared by the Cellule HIMO for the respective communes. These budgetary allocations are simply indicative to give the communes an order of magnitude of a reasonable level of investment based on their implementation capabilities. Since priorities may change, the communes, with technical help from the Cellule HIMO, will annually review their three-year rolling investment programs and get them approved by the Communal Councils. No project proposal will be accepted for funding unless they are included in the investment program but their inclusion in the program does not guarantee funding. The Cellule HIMO will review the projects proposed by the communes and wi:i select projects to be financed according to procedures and criteria described in Annex m. Communes that do not submit a sufficient number of acceptable projects will have their funds transferred to a common pool and used according to procedures described in Annex Im. Projects and contracts below US$40,000 would be approved by the Chief of the concerned technical unit (roads, forests or other LBWs) following consultationt with the Director of the Cellule HIMO. Projects and contracts above US$40,000 would be approved by a committee including the Director, and the chiefs of all technical units (roads, forests, other works). During the first two years, the committee S/ The PSTPIHIMO unit is located within MININTER to benefit from its institutional support at the prefectoral and communal levels. The unit, however, is fairly independent from the Ministry with regards to operational matters.  - 19 - will also include the head of the technical assistance team. Approval (or rejection) of all projects would be based on the selection procedures and criteria described in Annex III. 3.28 The proposed LBWs will be executed using cash-for-work in preference to food-for-work for the following reasons: (a) payment n'i cash would create a demand for food and help the development of markets (two objectives of the agricultural strategy); (b) cash availability would also stimulate cross-border trade, a traditionally important factor for food security; (c) cash-for-work is easier to manage than food-for-work which requires complex logistics; and (d) private contractors for the works would find it difficult to remunerate workers with anything else than cash. Poor workers will be targeted using the communal authorities' lists of the poor. The program will recommend that preference be given to women workers whenever possible and will encourage the establishment of on-site child care arrangements and bonuses to contractors who use women. The Cellule HIMO will be responsible for paying contractors and supervising technical aspects of their work and for submitting regular accounting statements on its use of funds to BCP. BCP will also take responsibility for grouping equipment requirements and purchasing them in large lots. 3.29 Sustainability. The proposed program will help to increase the communes' general revenues by financing markets and supporting sustainable agro-forestry. The development impact of the roads and agricultural terracing is also expected to increase communal revenues. Finally, by requiring specific maintenance programs for roads, markets and other infrastructures, the project will help increase the economically useful life of these investments. 3.30 Environmental Issues. This component will have positive environmental effects through the proposed anti-erosion measures (reforestation, agricultural terracing). Potential negative effects to the environment are minimal and will be prevented through a set of measures which have been identified during project preparation. The risks are related to preserving top-soil, existing natural forests and diversity of tree species. The proposed actions to avoid environmental risks include reliance to the maximum extent possible on labor- based construction using medium-sized equipment only when there is no other alternative. No new roads are planned and only existing roads will be rehabilitated. The agriculture terracing will be done on existing agriculture land and will be done on a pilot demonstration basis. Care will be -aken to remove the top-soil first before terracing and then replacing it on the terraces. In order to ensure a proper technical approach, no terracing will be undertaken unless an agricultural project has taken the lead. Reforestation will be on treeless hillsides with little agriculture potential. Native species (instead of exclusive reliance on pines and eucalyptus) will be included following the recommendations of the National Environment Plan. Micro-Enterpdrise Promotion and Development 3.31 Background and Justification. One of the main constraints that Rwanda is facing is the lack of off-farm employment in rural areas in order to compensate for the saturation of agricultural activities in most regions of the country. The government has been aware of this need, but had not been direcdy involved in the promotion of off-farm employment programs. However, with the structural adjustment program in place, the government seems willing to tackle this issue in an in-depth manner. Most efforts to capitalize on local entrepreneurship have come from NGOs, community organizations, and development projects which are exploring avenues for stimulating off-farm employment. In  - 20 - most cases, these efforts were successful on a small scale as they were undertaken on a regional basis (1-2 prefectures). In the rural areas of Rwanda, the notion of obtaining credit for a productive activity is virtually non-existent. In a 1990 survey undertaken by the Ministry of Industry, only 10 percent of micro-entrepreneurs had benefitted from a formal bank loan. The network of decentralized administrations - the 145 communes - could have played an important role in creating awareness among the population. However, in most cases, the communal administrations have lacked the technical expertise, knowledge, and financial resources to adequately promote micro-entrepreneurship. The most extensive and successfil program has been the ProLjet d'Appui aux Micro-realisations et Micro-entreprises financed by LUNDP with technical assistance from ILO which began in 1988. Its objective has been to promote and create rural off-farm employment on a national scale by providing financing, training, and technical assistance to potential micro-entrepreneurs. 3.32 Objectives and Scope of the Component. This component will provide support to the Programme d'Appui aux Initiatives de Base (PAIB) which is an expansion and enhancement of the above mentioned UNDP/ILO project. The objectives of this component are to: (a) promote the creation of profitable micro-enterprises in rural areas as a mean of generating productive non-agricultural employment; and (b) facilitate the creation of these micro-enterprises by appraising their viability and assisting micro-entrepreneurs in obtaining financing. The proposed activities of the PAIB are operational research-oriented and will have an impact on a longer term than the other project components: this component will provide feedback to the government and the development agencies on the lessons to be learned on the way off-farm employment can be developed in Rwanda. A more detailed description of the component is provided in Annex IV. 3.33 This program targets dynamic individuals who have the potential for creating and operating micro-enterprises and need assistance in obtaining credit and management training. These individuals will be identified through a community-based network and assisted by NGOs with experience in micro-enterprise development. 3.34 Activities and Imnlementation. The activities of this component will be demand-driven by the target populations who will be involved in all stages of this process. This means that the interventions proposed will be tailored to the needs and absorptive capacity of micro-entrepreneurs. An essential element of this approach is that this program will build upon existing local NGOs since most field activities will be sub-contracted to them. The three sets of activities are described below: (i) Creation of a Decentralized Network. This network will mainly consist of dynamic persons from the local communities (traditional leaders) who will act as "front line" promoters of this program. Their main responsibilities will be to create awareness among their community of potentially more lucrative off-farm employment opportunities through micro-enterprise development. Individuals willing to seize this opportunity will then be encouraged to transmit their ideas to the PAIB's agents located at the prefecture level. These agents will be responsible for processing the collected proposals: pre- selection, feasibility studies, and selection for credit funding. This approach will facilitate the interface between creative and ambitious individuals or groups and their technical and financial requirements. These network agents will be volunteers, but will receive the  - 21 - necessary training by PAIB's staff, as well as materials and travel allowances. (ii) project Processing through NGOs. Implementation of the PAIB at the periphery will be mainly the responsibility of NGOs. The selection of NGOs will be based on their expertise and capacity to effectively intervene in the various actions. They will also receive specialized training in order for them to: (a) carry out adequate feasibility studies for the proposed micro-enterprises; and (b) provide training and technical assistance in the fields of management, accounting, marketing, appropriate technologies, and commercialization. These services will be sub-contracted by PAIB management to NGOs. During the pre-implementation stage of this component (April- December 1992), the institutional capabilities of each NGO joining the program will be strengthened by the project team. Institutional strengthening will include: (a) definition of objectives and agreement on the work programs; (b) training of trainers and supervisors; and (c) establishment of adequate accounting systems. NGOs' performance will be monitored by PAIB management. (iii) Creation of Micro-enterprises. A mechanism will be put in place with participating NGOs to assist in the financing of potentially profitable micro- enterprises by: (a) conducting feasibility studies based on the pre-screened proposals; (b) preparing and presenting proposals for financing; and (c) negotiating with financial institutions (such as BaMes Popuaires, BWWe Rwandaise de DdveloDpement and community-based savings and credit associations) which are willing to expand their services to small rural credits. PAEB management will be directly responsible for initiating activity (c). Decentralized PAIB staff (see para. 3.37) will assist NGOs for activities (a) and (b). In addition to the funds, micro-entrepreneurs will benefit from training in management, accounting, production techniques and other relevant skills. Staff from the NGOs involved will also be responsible for regular follow-up visits. 3.35 By building on existing NGO presence in the six prefectures not covered by the existing UNDPIILO project, nationwide coverage is expected to be attained by 1995. At present, the project is intervening in four prefectures; three additional prefectures will be added every year so as to reach this target by 1995. Based on previous results by the UNDP/ILO project, it is anticipated that on the average thirty micro-enterprises wiUl be created and/or assisted in each prefecture annually. The total number of micro-enterprises to be created throughout the project is expected to be around 1,500. 3.36 ram Mana . The proposed management structure of this component wiUl also build upon the experience acquired from the existing UNDP/ILO projectf/ while reflecting the enlarged mandate and broader partnership base of the fi/ The Project Unit is located within the Ministry of Youth and Associative Movements (MIJEUMA) to benefit from its institutional support, particularly at the prefectoral level. The unit, however is independent from this Ministry with regards to operational matters.  - 22 - component (NGOs, community groups, communal administrations). At the central level, the executing unit (Breau Executif PAIB will be composed of a director, two financial analysts and the necessary support staff. This unit will be responsible for: (a) preparation of annual work programs and corresponding budgets; (b) coordination between NGOs, local authorities, and central and regional authorities; (c) contacts with financial institutions; (d) coordination with the other project components; and (e) monitoring of program performance and impact. The executing unit will benefit from the technical guidance and operational support of MIJEUMA, particularly at the prefecture level. In each of the prefectures, two PAIB agents will provide support to NGOs willing and able to participate in project activities. These NGOs will be responsible for reviewing requests for loans, performing feasibility studies, selecting the most viable projects, and requesting the central management unit's assistance in obtaining loans for the selected micro-enterprises through the formal banking system. Hence, the PAIB agents will be the operational links between the executing unit, at the central level and the NGOs, network agents, and entrepreneurs, at the field level 3.37 Potentially profitable micro-enterprises with credit needs exceeding US$1,000 up to a maximum of US$10,000 will be directed towards the formal banking system. Both the Bangue Rwandaise de Developpement (BRD) and the Banques Populaires (BP) have experience in approving loans for the creation of rural micro-enterprises. The PAIB will deposit the equivalent of US$250,000 with these institutions in the form of loan guarantee funds. Based on these two bank's evaluations of loan requests submitted by prefectoral units (approved by the PAIB central management unit and the sub-contracting NGO), these institutions will finance micro-entrepreneurs who lack the necessary collateral for ordinary loans. Forty percent of each of the approved credits will be guaranteed by the PAIB guarantee fund and 60 percent assumed by the bank. Potentially profitable, but very small micro-enterprises (requiring less than US$1,000 investment) will be directed towards community-based savings and credit associations. It is expected that this type of scheme will benefit extremely small potential entrepreneurs who are not in a position of obtaining funds from a bank due to the limited size of their activities and their lack of collateral. A service charge (of at least 10 percent per year) will be added to the loan amount in order to partially defray service costs and to acquaint entrepreneurs with the notion of credit and the payment for financial services. 3.38 Enviromental Issues, To avoid potentially negative environmental effects arising from certain types of micro-enterprises, a positive list of micro-enterprises is being established as part of the completion of the component preparation. For each type of project specific, environmental guidelines will be established to cover both project selection and project execution. The positive list and its corresponding guidelines will be prepared during the final preparation stage with help from experts and submitted for appropriate approval by the Ministry of Environment. The positive list and corresponding guidelines may be revised from time to time as necessary and the revised list submitted for approval by IDA. lving Standards and Poverty Monitoring Component 3.39 Rationale and Objective. From the inception of its structural adjustment program the Government of Rwanda has stated an interest in monitoring the micro-economic and welfare impact of the process of policy reform. The real economic effects of a process of structural adjustment find expression in the behavior of households and individuals regarding labor supply, production, trade, and consumption. The objective of the poverty monitoring system to be established under PNAS is tro trace the evolution of selected determinants of  -23 - welfare, and welfare outcomes, in the context of the country's adjustment process. Under the poverty monitoring system, particular attention would be paid to the following aspects: (a) the evolution of living standards over time (consumption of food and other essentials, nutritional status, satisfaction of basic needs, demand behavior in such areas as health, schooling, and housing); (b) the identification of poor groups in need of targeted assistance to raise their income and consumption possibilities; (c) the participation of different population groups in specified sectors of activity and economic growth; and (d) private reactions to, and the distribution of benefits from, changes in relative incentives between sectors and lines of production as induced by the process of structural adjustment. 3.40 Thus, the proposed poverty monitoring system would generate data on the evolution of key aspects of welfare over time. The data required to analyze the distributional consequences of a process of structural adjustment are currently unavailable in Rwanda. A household survey auete Nationale sur les Budgets et la Consommation des Menages" has been carried out in the eid 1 980s, but the size of the sample (ess than 300 households in the rural areas) was small, and imporiant changes have occurred in the Rwandese economy since that period. The survey needs to be updaLc- However, as already mentioned, a poverty profile of Rwanda has been developed on the basis nf this household survey as part of the preparatory work for PNAS. The proposed poverty moiitnring system calls for repeated data collection over time, because adjustment is a dynamic process1 its benefits and burdens are distributed in time. The changes currently affecting the Rwandese e&c'omy (internal and external shocks, adjustment, high population growth, war, AIDS epidemic, .cc.) have different employment and income consequences for different groups. A policy-oriented analysis of these consequences pre-supposes a permanent and economy-wide monitoring system. 3.41 Proframmatic Approach and Instruments. A first issue in the context of planning a welfare monitoring system over time pertains to the type of data to be collected. Generically, it is appropriate (for reasons of cost and expediency) to distinguish between indicator-based and survey-based poverty monitoring. Indicators include variables such as wage rates (particularly for unskilled labor), rural-urban terms of trade, indices of key food prices, and nutrition surveillance (growth monitoring in health centers and at sentinel sites). Time series of these poverty indicators (with various observation points per year) will provide a fairly accurate picture of broad poverty trends. But indicators do not provide information on causes of change, and they cannot account for population heterogeneity as sample surveys do. Observation of basic trends via indicators is less costly, but also much less revealing than survey-based analysis. 3.42 In the area of household surveys, it is useful to distinguish between integrated and priority surveys (PS). In the former, the full income generation and expenditure processes, and basic needs achievements are observed. Examples of this are the World Bank's Social Dimension of Adjustment (SDA) multi-topic integrated survey, the World Bank's living standards measurement surveys, and Rwanda's household consumption survey of the mid-1980s. An example of the latter is the World Bank's SDA priority survey, a household survey instrument designed for the rapid identification and monitoring of policy target groups. The data content of the priority survey covers a range of welfare aspects (including employment, assets, anthropometry, migration, agricultural and non-farm household enterprises, health consultations, school attendance and educational achievements,  - 24 - housing and amenities, and selected household expenditures) 2/. Compared with the integrated survey, the content of the priority survey is quite restricted, but the data, although collected over a larger sample, permit much faster tabulation and analysis. Ideally, the priority survey is conducted in the first phase of a survey program, hence providing policy makers with an initial empirical basis for targeting essential and urgent interventions. When repeated, the priority survey becomes a monitoring tool for tracing changes in welfare and basic needs fulfillment across socio-economic groups. Repetition of the priority survey make sense, because (as stated earlier) the micro-level effects of macroeconomic policy reform (or external shocks) are distributed over time and across groups and thus call for a combination of cross-sectional and time-series data. 3.43 Given the existence of the priority survey as an innovation in household survey "technology" (this survey has been developed under the Bank's SDA Program and is now being implemented in a number of African countries), the government has decided to refrain from investing in an indicator-based monitoring system in the context of PNASI/ and to deploy the SDA priority survey as a tool for poverty monitoring in Rwanda. The priority survey would be complemented once during the life of PNAS by an integrated survey to permit multi-variate analysis, update the weights for the calculation of regional price indices and, in general, permit comprehensive policy analysis on living standards and their determinants. 3.44 Overall, the welfare monitoring system proposed under PNAS would consist of one Integrated Survey (to be carried out in 1994-95) and two Priority Surveys (1993 and 1996). Technical details regarding the surveys and the program of field implementation, processing and analysis are given in Annex V. 3.45 Institutional ADets. The Director-General of Statistics in MINIPLAN will have overall responsibility for implementing this component. The Director of Surveys (a department within the Directorate of Statistics) will be the technical manager for the surveys. Statisticians, computer specialists and other relevant staff from this department will be mobilized to carry out the surveys and analyze the data. The welfare monitoring system to be established should integrate the two aspects of data collection and processing, on the one hand, and policy-relevant analysis, on the other hand. Thus, the BCP will be responsible for forming a coalition, and foster contact, between data users (analysts in the Ministries of Planning, Finance, Central Bank, line ministries; universities, donors) and data gatherers (Direction de la Statistique of MINIPLAN). To this end, a Comite des utilisateurs (Users' Committee) consisting of the above agencies (see organizational chart No HI) has been created 7.1 A full description of the PS (including a prototype questionnaire) is given in Grootaert, C. and T. Marchant et al., "The Social Dimensions Priority Survey", SDA Working Paper No. 10, The World Bank, 1991. E/ The indicator-based system is being imaproved through sectoral operations (such as Health, Education and Agriculture projects) and will contribute to the overall poverty monitoring system.  - 25 - to ensure a constituency for the survey program among planners and policy makers and maximize the use of the data for analytical purposes. The BCP will also be responsible for the dissemination of the results of the survey analysis, in close coordination with the technical committee in charge of monitoring the implementation of the structural adjustment program. IV. PROJECT IMPLEMENTATION A. Status of Project Preparation 4.1 The proposed operation was identified in October 1990, in parallel with the preparation of the structural adjustment program. A follow-up mission took place in January 1991 to: (a) define further the design of each component and agree on the terms of reference for project preparation; (b) help finalize the 1991-93 social sector public expenditure program; and (c) start building up donor coordination. At this stage, UNDP agreed to finance initial project preparation and to provide technical guidance to the Rwandese preparation team. The preparation mission which took place in May 1991 to review progress in project preparation focused on institutional arrangements for preparing the proposed operation. During this mission, the government set up a multi-sectoral preparation team, organized in five working groups (Food Security, LBWs, PAIB, Education and Social Surveys). MINIPLAN's directorate of studies and projects was put in charge of coordinating the project preparation process. International technical assistants were recruited through UNDP financing for the LBWs and PAIB components and preparation of the project document. The Canadian cooperation provided a consultant for the Food Security component. In addition, several local consultants were recruited by UNDP to contribute to the preparation process. Pre-appraisal took place in September 1991, and the mission was able to review comprehensive working documents for all components, and to assess the feasibility of certain activities during field visits. The mission left detailed recommendations to the preparation team to improve technical and managerial aspects of the proposal. Appraisal took place in December 1991. Aa advance under the PPF of US$1.3 million has been provided to the government to complete project preparation and initiate project-launch activities (such as training of key staff and detailed feasibility studies and preparation of bidding documents). B. Project Management Overall Institutional Framework 4.2 The proposed institutional framework has been designed to respond to specific operational characteristics of the PNAS, including: (a) its multisectoral nature which implies an effective collaboration between all ministries involved; (b) the large number of small-scale activities scattered throughout the country; (c) the involvement of private contractors in the LBWs; (d) the participation of several external agencies and various NGOs; and (e) the expected high disbursement rates. These characteristics have several important implications, including the need for: (a) providing PNAS with as much operational autonomy as possible, in particular to avoid undue interference from the administration; (b) separating the functions of overall programming and financial management from technical management of specific programs; and (c) ensuring that PNAS management has the required managerial skills for running such an operation. On the basis of these considerations, the institutional structure of the PNAS has been organized as follows: (a) the MINIPLAN is institutionally responsible for program execution because the PNAS is part of the social safety net that MINIPLAN  - 26 - monitors; (b) a program management unit (Bureau de Coordination ciu PNAS - BCP) will be responsible for the day-to-day financial management and implementation monitoring of the PNAS; (c) three executing units (one for each operational component LBWS, PAIB and Food Aid and Nutrition); and (d) a national coordination council (the Conseil de Coordination du PNAS will be responsible for overall supervision and coordination of the program. The BCP will be formally created as an self-managed project entity under the umbrella of MINIPLAN. This institutional arrangement would allow the BCP to carry out programming, budgeting and financial management without bureaucratic controls from the administration. The main elements of the proposed institutional framework are described below. An organizational chart is provided at the end of the report. 4.3 Conseil de Coordination. This advisory body would include representatives from the Prime Minister's office, MINIPLAN, MININTER, MINrTRAPE, MINLJEUMA, MINISANTE, MINITRASO, the Ministries of Finance, Trade, Agriculture, NGOs and External agencies. The Conseil's functions are to: (a) monitor the progress of the PNAS against the objectives and strategies stated in the Social Policy Statement issued by the government; (b) ensure coordination and collaboration among the sectors and various external agencies involved; and (c) review the recruitment process of key staff of the BCP to ensure that the best candidates have been selected. The regular schedule of the Conseil's meetings is every 6 months. The president is elected during the first assembly until the next formal session. Extraordinary sessions can be called upon request of one or several members and approval by the president. Meetings will be systematically called during IDA's supervision missions. 4.4 Bureau de Coordination (BCP). A detailed description of BCP's composition functions and staffing is provided in Annex IV. Key functions of the BCP include: (a) preparation of general annual work plans for the PNAS; (b) review and consolidation of annual work programs and budgets for each project component; (c) preparation of annual financing plans in coordination with all donors involved in the PNAS; (d) management of disbursement procedures (including the special account) for the IDA credit; (e) provision of adequate budgetary allocations through program accounts for the recurrent expenditures of all project activities; (f) procurement of works, goods and strvices either directly or on the basis of draft bidding documents prepared by the executing units; (g) arrangements for the audit of program accounts and general audit of the PNAS; (h) preparation of annual reviews of project implementation on the basis of indicators collected from the executing units; (i) adjustment of the PNAS overall strategy on the basis of the findings and recommendations of the mid-term implementation review (see para 4.15); and () overall monitoring of PNAS activities and coordination with the administration, external agencies and NGOs involved in the operation. 4.5 Executing Units. There are three executing units: (a) the Cellule HIMO (described in para. 3.26) which will be responsible for managing the implementation of the LBWs component; (b) the PAIB unit (Bureau Executif PAIB, described in para. 3.36) which will be responsible for managing the implementation of the Micro-enterprise component; and (c) the PSAN unit (Bureau Exdcutif PSAN described in para. 3.17) which will be responsible for coordinating the implementation of the Food Aid and Nutrition component. Two of these units are already operational for LBWs and Micro-enterprise Development with UNDP/ILO support but will be strengthened to respond to the heavier workload. These units are located in the relevant technical ministries (MININTER and MIJEUMA, respectively) to benefit from their expertise and operational support at the prefecture level. The units, however, are administratively independent from these ministries for their operations: they will receive their  -27 - financial resources from and will be accountable to the BCP. The same principles will be used with MINnTRASO while creating the PSAN executing unit, which will have a coordinated work program with the local WFP office. Representatives of the line ministries are members of the Conseil de Coordination where they have opportunities for expressing their views and recommendations either during the regular sessions or by calling for a special meeting if an important implementation issue has to be solved. 4.6 Establishment of the Institutional Structure and Staffing. During negotiations, agreement was reached with the government that the establishment of the Cgnseil d Coordination and the BCP and the selection and appointment of staff for all key positions of the BCP and the Cellule HIMO, including the four unit chiefs of the Cellule will be a condition of Credit effectiveness (paras. 3.26 and 7.2 (c)). The qualifications and experience of the personnel of BCP and the three technical units are essential for effective project implementation. This staff will be recruited under contractual arrangements with MINIPLAN after selection through local advertising and review by the Conseil de Coordination. Experience in private sector management will be an important selection criteria, given the type of activities to be carried out through the PNAS. The staff v-!Il be hired as local experts under contract with the govermment. This personnel will receive salary benefits comparable to those in the private sector, in order to recruit the most qualified individuals and ensure adequate performance. During negotiations, agreement was reached with the Government that high level staff of the BCP (national coordinator, accountant and procurement specialist), and the heads of the three executing units (Cellule HIMO, Bureau Executif PAIB and Bureau Executif PSAN) will be satisfactory to IDA at all time during project implementation (para. 7.1 (d)). 4.7 Technical Assistance CTA . Rwanda does not have enough individuals with the experience and skills required for rapidly and efficiently implementing the PNAS. Technical assistance needs have been assessed for central management (BCP) and the four components. A summary of these needs is provided in Annex VII. In order to increase the coherence and coordination of technical assistance, UNDP has agreed to finance a comprehensive TA program to support PNAS implementation. This TA program will be managed by the BCP and will bring in the necessary skills for each aspect of project implementation, including its monitoring. The technical assistants will be distributed between the BCP (programming, monitoring and financial management) and the executing units of each component (technical specialists). The BCP director and the principal technical advisor will have joint responsibility for ensuring the coordination and monitoring of the technical assistance program. The TA program will also contribute to the supervision of the PNAS by providing on a regular basis the status of key program performance indicators. Finally, a technical advisor will assist the MINIPLAN in monitoring the overall implementation of the Social Safety Net. The government, UNDP and IDA will coordinate their efforts to have this TA program operational by the time of credit effectiveness (expected by January 1993). Project Implementation Strateg 4.8 Guiding Principles. First, each component is operationally independent from the others in the sense that implementation delays for one component would not have a negative impact on the others. All components, however, have complementary and synergistic effects on poverty alleviation and their effective coordination would result in efficiency gains. Second, the BCP will be responsible for ensuring a rapid and efficient  -28 - implementation through: (a) effective planning/budgeting of all PNAS activities; (b) adequate procurement, disbursement and accounting; and (c) monitoring of financial and physical performance indicators. Finally, all PNAS activities will: (a) imply a certain participation of the beneficiaries (communes, selected NGOs and the target populations); and (b) make use of the private sector (contractors, consulting firms) to the largest possible extent. 4.9 Integration in the PEP. The PNAS is a new investment in the core social expenditure program of Rwanda and as such it is included in the general three-year rolling public expenditure program (PEP). PNAS appears in MINIPLAN's program because it is under the administrative umbrella of this ministry. The integration in the PEP would permit an annual review of the PNAS as part of the social safety net of public expenditures, including Health, Education and Population. This would facilitate the coordination of external financing and budget preparation for government contribution. 4.10 Donors' Coordination. Several donors have already expressed interest in supporting the PNAS. UNDP has been involved in the project process since identification and has contributed to the financing of project preparation. Canada has also participated in the preparation, specifically for the Food Aid component. Other agencies, such as the European (EEC) and the Belgian and Swiss cooperations have been actively involved in discussions on the concept and design of the PNAS. At the implementation phase, the coordination of these diverse contributions will be under the responsibility of the BCP. 4.11 Sustainability and Evolution. PNAS has been created to respond promptly to a deteriorating social situation, with means and management practices that are not available to the administration. But, beyond this primary goal of poverty alleviation, the PNAS should be used to: (a) demonstrate the effect of better managerial practices on project implementation; (b) better define the role of line ministries in poverty alleviation; (c) improve the performance of the communes in community-based development; and (d) test the potential of the collaboration between the public and the private sectors in social service delivery. During the mid-term implementation review (see para. 4.15), an assessment of all these points will be made to determine what should be the evolution of the PNAS. It is likely that certain activities will be integrated into existing ministries while others might be completely decentralized or contracted out to the private sector. The BCP will be responsible for monitoring the implementation of these recommendations while avoiding a disruption of on- going activities. C. Proiect Monitoring and Reporting 4.12 Monitoring. The overall monitoring strategy is based on a bottom up approach. Information on the various PNAS activities will be collected at the field level by each executing unit and transmitted to the BCP. The BCP would aggregate the data in a standard report which will be transmitted to the Conseil de Coordination and IDA every 6 months. The set of project performance indicators against which the progress of the PNAS will be measured are presented in annex VIII. The BCP will not have to wait for feed back to undertake corrective actions on the basis of the data they have collected if these corrective actions concern managerial or operational issues. On the other hand, policy or strategic decisions would be taken by the Conseil de Coordination on the basis of the analysis of field reports and would follow a top down approach. In all this process, the TA team will have a key role to play for ensuring both timely collection of indicators and accurate data analysis.  - 29 - 4.13 The project performance indicators described above give a good picture of the degree of penetration of PNAS activities in the target population, but they fail to provide information on: (a) the impact of poverty alleviation measures at the family level; and (b) the socio-economic impact on communities of certain PNAS activities, such as the development of micro-enterprises. Point (a) will be addressed through the Living Standards and Poverty Monitoring component, which will establish regular social surveys at the household level. Point (b) will be addressed through specific social studies aimed at assessing: (i) the potential of micro-enterprise as providers of off-farm employment in rural areas; and (ii) the socio- economic impact of LBWs on communities. 4.14 Reporting to IDA. During negotiations, agreement was reached with the government that: (a) the consolidated work prograrn, budget and financing plan for all project components will be submitted to IDA for review and approval no later than October 31 of each year of project implementation (para. 7.1 (a)); and (b) a progress report including: (i) the status of all project indicators and an analysis of the causes for the discrepancies between these indicators and the targets set in the annual work programs; (ii) a general assessment of the progress towards the achievement of the PNAS objectives and lessons learned during the past year; (iii) the status of procurement for all IDA-financed activities; (iv) the status of disbursement for each credit category and the situation of the special account; and (v) the identification of the major implementation problems and proposed remedial actions will be submitted to IDA for discussion no later than June 30 of each year of project implementation (para. 7.1 (c)). The annual audit report will be submitted separately, as specified in para. 5. 11. IDA's supervision plan is summarized in Annex IX. 4.15 Mid-Term Implementation Review. After two years of project implementation, the government and IDA (during a specific supervision mission) would carry out a mid-term project implementation review. About three months before the date set for this review, the government would recruit a team of independent consultants to carry out a preliminary project implementation review under terms of reference acceptable to IDA. The review would focus on certain areas, including: (a) the impact of PNAS activities on target groups at the community level; (b) the relevance of the initial targeting (particularly for the priority communes of the LBWs component) and the necessary changes; (c) the performance of managerial arrangements (particularly of the BCP) and the quality of coordination among all entities involved in PNAS implementation, including line ministries; (d) an assessment of the sustainability of PNAS and recommendations for further development in its scope and design, particularly with regard to the intehration of certain activities into the public or private sectors; (e) an assessment of the adequacy of the PEP process in providing a social safety net; and (t) a set of recommendations for the second phase of project implementation. The results of the review would be discussed by a joint government and IDA team. Agreement would be reached on the necessary changes and adjustment in PNAS scope and design. During negotiations, agreement was reached with the government that a mid-term implementation review of the PNAS will be carried out no later than June 30, 1995 under terms of reference acceptable to IDA (para. 7.1 (e)). 4.16 Project Completion Report. The government will be given the options of either preparing part II of the project completion report or preparing an independent evaluation. In any case, the project completion report will be put together within four months of the credit closing date.  -30- V. PROJECT COSTS AND FINANCING A. Erject Costs 5.1 The project will be Implemented over a four-year period. Total costs for the proposed project are estimated at FRW 553 million (US$46.1 million equivalent). Base costs are estimated at US$40.0 million equivalent contingencies amount to US$6.1 million equivalent (14 percent of total project costs). In addition, taxes amount to US$2.3 million and import duties to US$0.7 million. Foreign exchange costs account for US$15.7 million or 34 percent of total project costs. Project costs by category of expenditure are summarized in table 5.1 below and details are provided in Annex X. Food accounts for 32 percent of base costs, labor cost for the LBWs program accounts for 18 percent, operating costs account for 12 percent, tecinical assistance accounts for 12 percent and equipment for 10 percent of base costs. These costs b-ve been estimated during the December 1991 appraisal mission. Table 5.i1: Cost Summary by Expenditure Category FM.0001 .53.00 % Fwdpe S Tomd L.d FoRi Totd lod Forn TOWle A. Equipmt. Fum., Ma. 1476 322062 470650 1233 2660 3922 6s t0 a. Vdd 24000 102400 126400 200 86S 1063 st 3 C Food 606s61 70698 1512604 6m 562 12004 47 32 0. Tab" & Sm*las 06400 52900 161200 820 440 1260 35 3 E ChUWo*S s66e60 0 6e8m0i 7393 0 7383 0 18 F. Fund 2100 0 21BODO 1820 0 I6o0 0 6 0. Tech. Au_isto. 224160 382620 58s680 l8w 3021 4680 62