Rwanda - Gitarama Agricultural Production Project and MINAGRI Institutional Development Project

Document of The World Bank FOR OFmFCIAL USE ONLY Report No. 5895-RW STAFF APPRAISAL REPORT RWANDA GITARAMA AGRICULTURAL PRODUCTION AND MINAGRI INSTITUTIONAL DEVELOPMENT PROJECT February 19, 1986 Regional Projects Department Eastern and Southern Africa Regional Office This docment has a restricted distribution and may be used by recipients only in the performance of their official duties. Its contents rmay not otherwise be disclosed without World Bank authorization.  CURREE QUVALENTS Currency Unit - Franc Rwandese (RF) US $1.00 = RF 103 RF 100 ' $.97 WEIGHT AND NEASURES Metric British/US Equivalents I meter = 3.3 feet 1 hectare = 2.47 acres I are (100 m2) = 0.02 acre 1 kilometer (km) = 0.62 mile 1 square kilometer (km2) = 0.39 square mile (sq. m.) 1 kilogram (kg) = 2.2 pounds (lb.) I liter (1) = 0.26 US gallon (gal.) 1 metric ton (m ton) = 2,204 (2.b.) GcLOSSAY OF A'3BREVIhIIONS BACAR - Banque Continentale Africaine - Rwanda BCA - Correspondence and Archives/Bureau de Courrier et Archives BDI - Bureau of Documentation and Information/Bureau de Documentation et d'Information BGC - Management of the Budget/Bureau de Gestion des Credits BGP - Personnel Management/Bureau de Getion du Personnel BNR - National Bank of Rwanda/Banque Nationale du Rwanda BRD - Industrial Development Bank/Banque Rwandaise de Developpement CFRC/IWACU - Cooperative Research and Training Center/Centre de Formation et de Recherches Cooperatifs COOPIBO - Belgian Non-Governmental Association for Support to Cooperatives/Compagnons batisseurs CSOG - Cooperative Service Center of Gitarama DCE - Department of Control and Evaluation/Direction Contro^le et Evaluation DEP - Department of Studies and Planning/Direction Etudes et Planification INADES - Institut National pour le Developpement Economique et Social ISAR - National Research Institute/Institut des Sciences Agronomiques du Rwanda MINAGRI - Ministry of Agriculture, Livestock and Forestry/Ministere de l'Agriculture, de l'Elevage et des Forets MINIMARWT - Ministry of Industry and Artisanal Development/Ministere de l'Industrie, des Mines et de l'Artisanat OCIR - Office des Cultures Industrielles du Rwanda OPROVIA - Office National pour le Developpement et la Commercialisation des Produits Vivrieres et des Productions Animales SNEA - National Agricultural Survey/Service National des Enquetes Agricoles SSS - Selected Seed Service/Services des Semences Selectionnees TRAFIPRO - Travail, Fidelite, Progras (Cooperative) FISCAL YEAR Government - January 1 to December 31  FOR OFIL USE ONLY (i) RWANDA GITARAMA AGRICULTURAL PRODUCTION AND MINAGRI INSTITUTIONAL DEVELOPMENT PROJECT STAFF APPRAISAL REPORT Table of Contents Page No. Table of Contents *.. ......................................... i) to (v) Credit and Project Summary ................................ (vi) to (ix) I. BACKGROUND ..... .............................. 1 A. Project Background ....... ......................... 1 B. Economic and Agricult.iral Setting ................. 2 C. Government's Agricultural Sector Strategy and Major Issues of Agriculture ............ .... .... ... .. 3 D. Government Financing of Agriculture ............... 4 E. IDA Assistance to Rwanda's Agriculture Sector ..... 5 II. RWANDA'S AGRICULTURE SERVICES ......................... 7 A. The Ministry of Agriculture, Livestock and Forests (MINAGRI) ............. ...... ............ O................... 7 B. Government's Agricultural Extension and Related Services ......................... ... ..... 8 C. Agriculture Research ..... ........ ................ . 9 D. Farm Inputs .............. .0...................................... 10 E. Agricultural Produce Marketing and Pricing .eeeoe 11 F. Agricultural Credit and Finance ........ ............ 11 G. Cooperatives and Farmers' Groups ....... . ........... 13 H. The Gitarama Region ................ ...... .... . 13 This document has a restricted disribution and may be used b: recipients only in the performance of thir official duties Its contents may not otherwise be disclosed without World Bank authorization.  (ii) Page No. III. THE PROJECT ...... . .............................. 14 A. Sector Strategy and Rationale for the Project ..... 14 B. Summary Description of Project ..................... 16 C. Detailed Features, Extension and Agricultural Services .................................................... O 17 Development and Introduction of Specific Extension Packages ....0 .................... 17 The Extension System .... . . ....... .. ......... .. 20 Support Services .......................... ... . 21 Training .............. O-T..........a 22 Equipment and Transportation ................... 23 Civil Works ...... ...**.. ....... .......* ......... 23 D. Detailed Features: Promotion and Financing of Agricultural Processing, Marketing and Farm I.nput Supply .............................................. 24 The Investment Promotion Office ................ 24 Support to the Banques Populaires .............. 25 Support to Cooperatives and Farmers' Groups .... 25 Support to the Artisanal Sector .............. .. 26 Innovation Fund .. .......... . .............. .0 26 Documentation, Information and Surveys ......... 27 E. ilyabarongo Valley Master Plan Study ............... 27 F. MINAGRI Institutional Development Comporent ....... 28 Component Objectives .......................... 28 Detailed Features ... ............. .... ...... ..... 30 IV. PROJECT COSTS AND FINANCING ............................ 33 A. Cost Estimates ................... ............................. 33 Gitarama Agricultural Production Component ..... 33 MINAGRI Institutional Development Component .... 33 B. Financing ......................................... 0............ 34 Gitarama Agricultural Production Component ..... 35 Cost Recovery and Fiscal Analysis .............. 36 MINAGRI Institutional Development Component .... 37 C. Procurement ................................*...... 37 D. Disbursements ..................................... 40 Statements of Expenditure ...................... 40 Special Account ......... ....................... 40 Government Counterpart Funds .................. 41  (iii) Page No. E. Financial Management ..... ........ ....... ............ 42 Accounting for Gitarama Agricultural Production Component .........o..... ..43 A'counting for MINAGRI Institutional Development Component ............................*..o..e 43 F. Environmental Iumpact ....*. --.. ..- ........ ..... 43 Vo ORGANIZATION AND MANAGEMENT .. .............. .. .... . 44 A. Gitarama Agricultural Production Component ........ 44 B. MINAGRI Institutional Development Component ......o. 47 VI. BENEVITS, JUSTIFICATION AND RISKS ....... oo.. ............ 50 A. Justification and Benefits of the Gitarama Component ..... o....... o.....-o ......... 50 B. MINAGRI Institutional Development Component ..... 51 C. Economic Analysis o ................. 52 D. Risks ............ .. .....-......... 53 VII. AGREEMENTS REACHED ..... ... .. .... .. ...... . 54 TABLES IN MAIN TEXT: Table 3.1: Yields o.o....o ........................ o ..o........ ... ... 19 Table 4.1: Total Project Costs ..... ...................... 34 Table 4.2: Financing Plan ...o. ............. ... . ... ... . . 35 Table 4.3: Procurement Procedures - Gitarama Agricultural Production Component ...... .-...... ........ 39 Table 4.4: Procurement Procedures - MINAGRI Institutional Development Component ..... -.-.-....... 40 Table 4.5: Gitarama Agricultural Production Component, Proceeds of the IDA Credit ............... 41 Table 4.6: MINAGRI Institutional Development Component, Proceeds of the IDA Credit ...... o .......... 42 Table 6.1: Gitarama Prefecture. ..... o .............. 51 Table 6.2: Farm Models o ................... ....... 51 Table 6.3: Gitarama Prefecture, Project Year 10 * ......... 52  (iv) Page No. SUPPORTING ANNEXES, CHARTS AND MAPS: Annexes: Annex 1 - Implementation Schedule ........................ 57 Annex 2 - Schedule of Recruitment of MINAGRI Institutional Development Project Staff .................... 58 Table 1 - Local Staff ........ 0-00 ......... 58 Table 2 - Expatriate Staff ......... ............ 59 Table 3 - MINAGRI Secretariat-General, Project Staffing Requirements . 60 Annex 3, Table 1 - Gitarama, Summary Accounts Cost Summary 62 Table 2 - Summary Accounts by Year ... 63 Table 3 - Breakdown of Summary Accounts 64 Table 4 - Project Components by Year . 65 Table 5 - Project Components by Year . 66 Table 6 - MINAGRI, Suimmary Accounts Cost Summary 67 Table 7 - Summary Accounts by Year ... 68 Table 8 - Breakdown of Summary Accounts 69 Table 9 - Project Components by Year . 70 Table 10 - Project components by Year . 71 Annex 4, Table 1 - Schedule of Disbursements - Gitarama . 72 Table 2 - Schedule of Disbursements - MINAGRI - IDA ...9 ..... .............. 73 Annex 5 - Projected Government Cash Flow ................. 74 Annex 6 - Expected Adoption Rates/Taux d'Adaption ........ 75 Annex 7 - Yields of Selected Commodities/Rendement des cultures .... 76 Annex 8, Table 1 - Economic Analysis ..................... 77 Table 2 - Economic Analysis, Sensitivity at Expected Adoption Rate .............. 79 Annex 9, Table I - Farm Budgets No. 1.o ................ 80 Table 2 - Farm Budgets No. 2 ..................... 82 Table 3 - Farm Budgets No. 3 ......... o .......... 84 Table 4 - Farm Budgets No. 4 .................... 86 Table 5 - Farm Budgets No. 5 .................. . 88 Table 6 - Farm Budgets No. 6 .................... 90 Annex 10, Table 1 - Farm Model No. 3 ................. .. 92 Table 2 - Farm Model No. 4 ...o ................ 96 Annex 11 - Selected Data and Documents Available in the Project File .....*... ......*................ 100  (v) Page No. Charts: Chart I - Organizational Chart - Gitarama ................. 104 Chart 2 - Organizational Chart - MINAGRI .................. 105 Chart 3 - Organizational Chart - Ministe-re de l'Agriculture, de l'Elevage et des Forets .... 106 Maps: IBRD 19380 - Rainfall and Agricultural Pattern and Location of Bank-financed Projects IBRD 19381 - Gitarama Prefecture IBRD 19382 - Main Existing Agricultural Projects Financed by External Donors  (vi) GITARAMA AGRICULTURAL PRODUCTION AND HINAGRI INSTITUTIONAL DEVELOPMENT PROJECT Credit and Project Summary Borrower: Republic of Rwanda Beneficiary: Ministry of Agriculture, Livestock and Forestry (MINAGRI) Amount: SDR 11.617 million (US $12.7 million equivalent) Terms: Standard IDA terms Project The proposed Credit would finance a set of activities Objectives designed to support Government's agricultural sector strategy and with the twin objectives of stimlating agricultural Description: production, employment and income, and improving management of the sector. It comprises two distinct components which would be managed separately. The Gitarama Agricultural Production component would be the pilot phase of a national agricultural extension and agriculture service organization, and of a national program for promoting private and cooperative investment in agriculture, agro-processing and marketing. The service systems developed on a pilot basis under this component are intended as a model which would be replicable at the national level and institutionally and financially sustainable. The component would include: (a) identification and introduction to farmers of specific recommendations to increase productivity and income; (b) introduction of a new extension management system; and (c) support to the development of other selected services (e.g. provision of plant material, forestry and veterinary services). In addition, it would provide support and assistance to credit institutions, cooperatives, farmers' groups, artisans and private enterprises to develop agro-processing, marketing and input supply activities. The Credit would also finance a study for the development of the Nyabarongo Valley, an area of considerable potential for agricultural intensification. The MINAGRI Institutional Development component would strengthen the institution building process to improve agricultural sector planning and management. The component would provide support mainly in the form of technical assistance, equipment and training for project preparation, monitoring and evaluation of agricultural services and projects, accounting and financial management, documentation and general administration.  (vii) Benefits: The Gitarama Agricultural Production component would benefit directly a large number of farm families (about 50,000 suallholders, all in the poverty target group) in the Gitarama prefecture. In time, the entire population of Rwanda is expected to benefit from project interventions when these are replicated throughout the country. Farmers would benefit from sustained and systematic support services designed to address their specific needs. Production, especially that of food crops and farm income, would be increased. The component would also benefit a large number of artisans and small private entrepreneurs in the prefecture, especially through professional training. The country would generally benefit from a more efficient 'uad cost-effective administration of key services to the rural population. The MINAGRI Institutional Development component would have sector-wide benefits through better planning, organization and management of the ministry. The project would sharpen the technical skills of ministry staff and improve the design, management and cost-effectiveness of services to the sector. Risks: The most important project risk relates to budget allocations from MINAGRI. Measures have been foreseen to assure adequate funding to minimize project costs and to limit the financial burden on the Government to acceptable levels. Another important risk relates to the innovativeness of the project proposals and the resulting changes required in the institutional behavior of the Ministry of Agriculture, especially in terms of management. This risk is greatly reduced by the strong commitmert of the Government to the project and it would be further minimized with sustained IDA support to project implementation. A specific risk of the Gitarama Agricultural Production component is that of running out of improved technical recommendations, but this will be greatly reduced as the Agricultural Research Project (Credit 1546 RW) begins to produce results. A specific risk of the MINAGRI Institutional Development component relates to the need for qualified personnel. This risk is greatly reduced by the important training component.  (viii) Estimated Local Foreign Total Project Costs: - US $ million -- I. Gitarama Agricultural Production Component Extension 3.24 1.87 5.11 Adaptive research .46 .51 .97 Other agricultural services 2.06 .95 3.01 Investment Promotion Office .33 .43 .76 Support to Banquee Populaires .04 .01 .05 Suppurt to cooperativ.-s and farmers' groups .18 .08 .26 Support to the artisanil sector .12 .17 .29 Technical assistance - .58 .58 Nyabarongo Valley Study .24 .97 1.21 Base cost 6.67 5.57 12.24 Physical contingencies .36 .46 .82 Price contingencies 1.55 .87 2.42 Total component cost 8.58 6.90 15.48 II. MINAGRI Institutional Development Component Studies and planning .86 1.69 2.55 Monitoring and evaluation .22 .14 .36 Management and accounting .37 .66 1.03 Computerization .28 .71 .99 Documentation .27 .25 .52 Personnel and financial management .18 .14 .32 Files and archives .06 .04 .10 Base cost 2.24 3.63 5.87 Physical contingencies .10 .22 .32 Price contingencies .46 .56 1.02 Total conponent cost 2.80 4.41 7.21 III. Total project costs 11.38 11.31 22.69 Financing Plan: IDA Gitarama Ag. Production component 3.05 7.15 10.20 MINAGRI Inst. Dev. component .75 1.75 2.50 Sub-total 3.80 8.90 12.70 UNDP Technical assistance and studies .60 2.41 3.01 Government Gitarama Ag. Production component 4.45 - 4.45 MINAGRI Inst. Dev. component 1.43 - 1.43 Taxes 1.10 - 1.10 Sub-total 6.98 - 6.98 Total (including taxes) 11.38 11.31 22.69  (ix) Estimated Disbursements: IDA fiscal year: 1986 1987 1988 1989 1990 1991 1992 US $ million - A;nnual .47 1.33 2.35 3.25 3.20 1.87 .23 Cumulative .47 1.80 4.15 7.40 10.60 12.47 12.70 Economic Rate of Return: 272 on the directly productive components of the Gitaramua Agricultural Production component Maps: IBRD 19380, 19381 and 19382  RWANDA GITARAMA AGRICULTURAL PRODUCTION AND MINAGRI INSTITUTIONAL DEVELOPMENT PROJECT I. BACKGROUND A. Project Background 1.01 The proposed project is the first of several possible operations aimed at addressing, on a small/pilot basis, a priority list of development areas established during agricultural sector discussions between Bank/IDA staff and the Government in November 1983. This approach would enable the Government to address simultaneously several hirh priority secto-ral issues and to test concepts thoroughly before applying them on a larger or national scale. The proposed project would include two sub-projects: (a) the Gitarama Agricultural Production component as a pilot phase of a national agricultural extension, promotion and service organization; and (b) the MINAGRI Institutional Development component to strengthen the capacity of the Ministry of Agriculture to plan, monitor and coordinate agricultural activities. 1.02 This project takes into account Government's main objectives for agricultural development in the Gitarama Prefecture as expressed in a proposed Integrated Rural Development Project for the prefecture presented to the Bank for consideration in early 1982. It differs, however, from the original proposal because the i983 sector discussions concluded that it was essential to address as a priority matter various sectoral issues highlighted by the sector analysis. Further, the Bank's experience from past and ongoing projects suggested that the integrated rural development model had proven excessively costly in terms of overhead and manpower requirements, redundancy in parallel regional development structures and inefficiency of central coordination and control. The present operation was identified over a series of IDA missions carried out in early 1984. During that period, the Agricultural Research Project was appraised as the first of a series of institution strengthening projects to develop coherent national agricultural services and to reinforce the central ministry's planning and management functions. 1.03 Preparation of the project was carried out by MINAGRI staff, supported by external consultants financed under an IDA-financed agricultural project (BGM II) and followed up by IDA staff missions in the field. Project proposals were completed in March 1985 and the Government invited IDA to appraise the proposed project promptly. This report is based on the findings of a pre-appraisal mission in April 1985, followed by an appraisal mission in April/May and June/July 1985 composed of Philippe Blanc, Susanna Badgley, Xavier Legrain and Armand de Largentaye (EAPCA) and Jean-Fransois Barres (Consultant).  - 2 - B. Economic and Agricultural Sr. xing 1.04 Rwanda is a landlocked country covering an area of approximately 26,000 km2, The topography is hilly to mountainous with elevation and rainfall increasing from east to west. The population of Rwanda is estimated at about 5.5 million with an annual growth rate of approximately 3.7Z. Population density is one of the highest in Africa, about 400 inhabitants per square kilometer of usable land. Only 5Z of the population lives in urban areas. Smallholder and rainfed agriculture is the predominant feature of the economy. There are about one million smallholdings. Population pressure has resulted in diminishing farm size wit't more than half of the farm families occupying less than 1 ha and another 30% between 1 and 2 ha. Average annual rural family income is about US $550 (US $70 per capita). 1.05 In recent years, Rwanda has been a good economic performer; real GDP grew at an average rate of 5.3Z per annum between 1970 and 1982. The agricultural sector is the most important contributor to the national economy. Its share of GDP at current prices has averaged about 40Z in recent years with food crops contributing about 80Z of this. In 1983, agricultural products accounted for 82Z of the total value of exports, with coffee alone representing 66% and tea 12%. 1.06 About half of Rwanda's land area can be used for agriculture. Of this, 5C% is planted in food crops: beans, sweet potatoes, cassava, beer banana and sorghum (and more recently, potatoes, maize and peas). About 5% is under export or industrial crop cultivation, mainly coffee, but also tea and approximately 40% is used as pasture land, wood lots and fallow. Rapidly growing population pressure has resulted in an expansion of the area under cultivation at the expense of forested areas, grazing lands and the practice of fallow. Declining land availability has led to an intensification of agriculture through double cropping. Despite the high fertility of soils in much of the country, the adaptability of farmers and relatively well-developed traditional agricultural systems, there is mounting evidence that productivity is falling as a result of overcropping, increased cultivation of marginal lands and erosion in some fragile areas (see W. Jones and R. Egli -Farming Systems in Africa - the Great Lakes Highlands of Zaire, Rwanda and Burundi", World Bank Technical Paper No. 27, 1984). 1.07 Up to tne present, food production seems to have kept pace with population growth, roughly maintaining national food self-sufficiency. However, national statistics indicate that the growth of agricultural output may be slowing down after a period of relatively strong growth, 5.8% per annum in the period 1974-81. During the period 1981-84, the average annual growth rate was about 3.5%, less than the population growth rate. Commercial food imports (mainly wheat, rice, oils, salt and sugar) accounted for about 8% of total value of merchandise imports in 1984 (US $11 million). Normal food aid flows in recent years have been in the order of 10,000 to 15,000 tons of cereals, beans, powdered milk and vegetable oils.  1.08 Rwanda experienced substantial crop failures in 1984 due to poor rain early in the year. The Goverrment responded quickly to a potential food crisis by assisting farmers to plant dry swamp lands and river valleys. Donors provided an additional 27,000 tons of food aid, the arrival of which was delayed due to transportation bottlenecks in Kenya. The rainy season in late 1984 and 1985 has been good and general recovery is proceeding well. C. Government's Agricultural Sector Strategy and Major Issues of Agriculture 1.09 The Third Development Plan (1982-86) reaffirms the predominant importance of the agricultural sector to the Rwandese economy and stresses the need to focus greater resources on the critical constraints to increasing production at the farm level. The means of achieving these objectives include the following: (a) Reorientation of agricultural research to the needs and constraints of the traditional farm system and strengthening of links between research and extension. (b) Improved organization of field extension services. (c) Expansion of agricultural credit. (d) Development of food crop marketing, expansion of input services and improving food crop storage. (e) Development of the cooperative movement. (f) Strengthening of technical manpower capacity through expansion of technical schools. 1.10 The major constraints to implementing this strategy include the Government's difficult budgetary situation, limited sector planning and management capacity, and the shortage of trained staff in technical and management fields. A highly decentralized approach to development featuring numerous integrated rural development (IRD) projects has added complexity in coordinating interventions in the agricultural sector and increased the Government's difficulty in formulating and implementing new policies and national actions. Each IRD project has introduced a separate extension service, input supply mechanism, credit system, etc., contributing to the fragmentation of such services throughout Rwanda. External agencies have often contributed to this situation because of insufficiently coordinated actions. 1.11 There is an increasing concern in the Government that past development efforts have generally not led to significant increases in agricultural productivity or to viable and sustainable enterprises and  - 4 - field services. This has prompted a closer scrutiny of proposed investments, a review of the organization of agricultural field services and of policies toward agricultural subsidies. In addition, there is an increasing awareness of the need for better policy analysis, project prepara:ion and for more effective monitoring, evaluation and control of projects and services. D. Government Financing of Agriculture 1.12 Agricultural investments and services are financed from five sources: the Government, external aid agencies, non-governmental organizations, the local communes and the general population's collective and compulsory contribution through public works projects (umuganda). Government allocations to the sector in recent years have averaged about RF 1,400 million (US $14 million). About 60% of this has gone to operating costs of ongoing activities and 40% to the Development Budget. Agriculture accounts for about 5% of the total Operating Budget and about 15% of the Development Budget. Allocations to the sector have remained largely unchanged in nominal prices over the last three years and its share of the total has also remained constant. The Government has been very successful in attractin- foreign assistance for the sector. The level of foreign aid (mainly on concessional terms) has grown steadily in recent years; expenditures in 1983 from multilateral and bilateral aid sources amounted to approximately US $32 million. Non-governmental organizations were estimated to contribute about US $700,000 in 1983. Communal financing and the contribution of umuganda vary significantly from one commune to the next, depending on such factors as the level of economic development of the area, the resource base of the commune and the dynamism of local leadership. Communal participation in agricultural activities typically includes the financing of some field level extension agents, establishment of tree nurseries, erosion prevention measures, forestry plantations on denuded areas and other specialized activities. 1.13 Cost Recovery. Direct cost recovery on agricultural services is low, although it should be noted that the umuganda system can be considered as a payment in kind, a sort of cost recovery. Under the current Government policy, most inputs are subsidized to some extent: all tree seedlings, shoots and cuttings are free as are natural livestock insemination services. Improved livestock (heifers and smallstock) and selected seeds are generally sold at prices slightly higher than those for traditional stock, but much lower than production costs. Most agricultural tools are sold through private sector channels at market prices although some newly introduced tools are sold by IRD projects at cost. Veterinary medicine and services (castrations, wound treatments and small operations) are charged to stock owners at cost. In principle, farmers pay indirectly for extension services through local taxes although the communal taxation system remains poorly organized and most communes cannot afford to pay more than a few agricultural employees. The impact of subsidies on the Government's budget has not been large to date because most subsidies are  financed by externally financed projects (seed multiplication, insemination services, production of tree seedlings and fertilizers imported on a grant basis, etc.). The problem will increase as such donor financing phases out. 1.14 There is an increasing awareness in Government of the potential future impact of subsidies on the Government's budget. Improved cost recovery for Government input distribution and services is now a Government objective. Recent initiatives have been taken in some projects to provide vehicles (motorcycle and bicycles) to staff on a credit basis, to rent underutilized transport or storage capacity to the private or cooperative sectors at cost and to provide equipment services at a fee to the local community. Some projects have introduced greater cost recovery on project staff housing. Any future project should incorporate cost recovery measures on inputs and services provided to farmers, should minimize the need for operating cost outlays by the Government and should be carefully designed to ensure that such Government outlays, which are required, can be supported. E. IDA Assistance to Rwanda's Agriculture Sector 1,15 Previous and Ongoing Bank Group Involvement. IDA lending for the rural sector has amounted to US $92.4 million for eight operations as follows: (a) four area development projects in the eastern savannah, lower altitude zones, Mutara I and II Projects (Credit 439 RW - US $3.8 million in 1974 and Credit 937 RW - US $8.8 million in 1980), and the Bugesera-Gisaka/Migongo I and II Projects (Credit 668 RW - US $14.0 million in 1977 and Credit 1283 RW - US $16.5 million in 1983); (b) two area development projects in the Lake Kivu and Zalre-Nile divide zones, the Lake Kivu Coffee and Food Crop Improvement Project (Credit 1126 RW - US $15.0 million in 1981) and the GBK Forestry and Livestock Project (Credit 1039 RW - US $21.0 million in 1981); (c) the Cinchona Development Project financed by Credit 656 RW - US $1.8 million in 1977; and (d) the Agricultural Research Project (Credit 1546 RW - US $11.5 million). Through these projects, IDA has been involved in development efforts in several ecological zones covering a broad spectrum of regional development activities in all of the major sub-sectors, food crops, export crops (coffee and cinchona), forestry and livestock. With the exception of the Cinchona Development Project, the other projects have been of an integrated development nature, involving substantial investments in agricultural extension services, construction of roads, and various economic and social infrastructure and project related research. The Agricultural Research Project represents the first of a series of new types of projects aimed at addressing critical sectoral constraints (para. 1.18). 1.16 Performance Evaluation. To date, there is only one project (Mutara Agricultural Development Project - Phase I, Credit 439 RW) for which both a completion report and a PPAR (Sec. M 81-570, 1981) have been prepared. A Project Completion Report was recently issued for the  - 6 - Bugesera-Gisaka/Migongo - Phase I Project (Credit 668 RW) and a further Project Completion Report for the Cinchona Project is under preparation. These projects resulted in a lower than expected increment in production and their economic impact was marginal. Being too costly, they were not sufficiently sustainable. However, there were substantial achievements in the areas of infrastructure (BGM and Mutara), population settlement (particLlarly in Hutara), and improved efficiency of administrative and financial control systems. Among the successful activities developed in these projects are erosion control, production and distribution of forestry and fruit tree seedlings, development of vegetable production and improvement of animal health care. 1.17 Experience from these and other ongoing IDA-supported projects highlights a number of common issues affecting project performance: (a) poor database, insufficient knowledge of Rwanda's complex farming systems leading to poor identification and adaptation of appropriate technicdl packages and inadequate monitoring of project progress; (b) project complexity and lack of training and experience of Rwandese management and staff; (c) excessive diversity of project institutional and organizational systems made unmanageable by weak agricultural planning, project preparation, financial control and technical central ministry departments; (d) insufficient attention to recurrent costs financing; and (e) scarcity of technical messages for agricultural improvement at the farm level. Other important constraints to agricultural development are inadequate Government operating budget allocations to agriculture, inexistent or unclear sub-sector policies and investment priorities and poor coordination of donor-suppGrted activities. 1.18 IDA's sectoral dialogue and lending strategy is based on these lessons of experience and on a more general analysis of sector performance, In-depth discussions with the Government in November 1983 led to the main recommendation to move away from an exclusive emphasis on IRD-type projects (still the most common model of bilateral and multilateral donor financed projects in Rwanda) which have not resulted in the development of economically viable and replicable technical support services to the agricultural sector. As a result of sector discussions, priority areas retained for medium-term lending include: (a) development of Government agricultural services essential for agricultural intensification, agricultural and veterinary extension, credit services and other agricultural services; (b) institution building to strengthen planning, project preparation, monitoring and evaluation, financial budgeting and control and training; (c) development of private sector services for agricultural marketing, input supply and agricultural produce processing; (d) further export crop development; and (e) identification and implementation of technical training priorities. The proposed operation would be the second IDA-financed project to fully support the above sector strategy, as it would address key issues of agricultural services (extension in particular) and sectoral management. The first was the Agricultural Research Project which addressed agriculture research sub-sector development issues at the national level. Other projects currently under preparation are designed also to contribute to the  - 7 - implementation of this strategy. A Second Forestry Project is intended to strengthen Government's efforts to Lmplement its national forestry policy, with strong emphasis on the developm!ent of the institutional capacity needed for the management of the sub-sector. Other projects are intended to strengthen the national services related to livestock and coffee. II. RWANDA'S AGRICULTURE SERVICES A. The Ministry of Agriculture, Livestock and Forests (MINAGRI) 2.01 Constraints to Sector Planning and Management. MINAGRI has a broad mandate for the management of agriculture, livestock, forestry and other land and water resources. There are four principal technical departments, including: (a) the Direction Generale for Livestock which covers animal production and health, fisheries and poultty production; (b) the Direction Generale for Agricultural Production which covers agricultural extension, selected seeds, plant protection, industrial crops and food crops; (c) the Direction Generale for Rural Engineering and Soil Conservation covering irrigation, farm tools and mechanization, topography and soil analysis, anti-erosion and soil conservation activities; and (d) the Direction Generale for Forestry covering the maintenance and exploitation of natural forests, reforestation and agro-forestry. The Secretariat-GEn6ral of the ministry is responsible for coordination and control of the technical service departments, supervision of agricultural projects (approximately 50) and general sector planning, management and administration. MINAGRI has about 2,500 staff in the four technical departments and the Secretariat-General. Most of this personnel is posted in the field as prefectoral and communal technicians and extension staff. Two organizational charts are provided (Charts 1 and 2). IBRD Map 19382 shows the main agricultural projects under implementation. 2.02 A number of separate services and agricultural parastatals also come under the jurisdiction of the ministry, including the Selected Seed Service (SSS), the National Agricultural Survey (SNEA) (Service National des Enquetes Agricoles), several agricultural training centers and four parastatals in charge of coffee (OCIR-Cafe), food crop and livestock marketing (OPROVIA), development of the Mutara region (OVAPAM) and national agricultural research (ISAR). The management of several other agro-industrial commercial services and parastatals for tea (OCIR-ThE), pyrethrum (OPYRWA), sugar (La Sucrerie) and rice (La Riziculture), etc., was recently transferred to the Ministry of Industry and Artisanal Development (MINIMART). 2.03 Government has made important efforts to improve its general planning capacity, especially with the UNDP/IDA-financed Technical Assistance Project chiefly designed to strengthen the Ministry of Planning. Parallel efforts are also underway to improve the public finance management capability of the Ministry of Finance. In addition, the  Government has recognized that reinforcement of the programming, monitoring and evaluation, and budgeting and financial control functions of MINAGRI is a prerequisite to improving the effectiveness of Government policies and interventions in the sector. Currently, these planning and management functions are the responsibility of the Secretariat-General. The major constraints include: (a) insufficiency of qualified and adequately trained staff in both technical and management field; (b) the absence of well-defined objectives, systems and procedures for the efficient implementation of priority activities; (c) the lack of well-developed information systems and technical documentation at the field and central level; and (d) ina4equate physical facilities, office equipment and logistic support for ministry staff and operations. B. Government's Agricultural Extension and Related Services 2.04 The Government's agricultural field services manage extension and other agricultural services. Each prefecture in Rwanda (ten in total) has a resident agronomist, a veterinarian and sometimes a forester supervising comimnal level agronomists, veterinarians and foresters (in principle, one of each per commune). There are an average of 14 communes per prefecture. The agronomist at the commune level supervises the extension workers, of whicb there is in principle, one per "secteur (average of 11 secteurs per commnue). Total staff available in the field to carry out extension and service work amounts to 68 agronomists having some post-secondary education and 2,075 field level extension agents (Annex 2). These staff are responsible for applying the Ministry of Agriculture's directives on extension and agriculture services, but are paid by a number of different sources (MINAGRI, parastatals, projects, other ministries and private sector). 2.05 MINAGRI issues yearly quantitative directives focussing on a yearly priority (year of the tree, year of agriculture-livestock association, of erosion control, etc.). These directives also include a number of general themes like eroaion control, reforestation, improved agricultural production techniques, use of improved seeds, stall feeding of cattle, etc. MINAGRI gives directives to the prefecture agriculturalists, veterinarians or foresters who are supposed to pass them on to field level extension agents. Agricultural staff and farmers are legally obliged to adhere to the directives. 2.06 The existing extension system as presently conceived, organized and implemented has received little endorsement and support, particularly  -9- from donor-assisted projects, which at present cover more than 100 of the 143 coumunes of Rwanda (IBRD Map 19382) and has proved ineffective. The core problems of the system are the following: (a) there are a multiplicity of activities undertaken by agricultural staff at all levels who are paid and managed by a number of different institutions, and therefore, not coordinated within a single national extension system; there are as many extension systems in Rwanda as there are donor projects and agricultural parastatals; (b) the centrally designed national extension themes are usually ill-adapted to the different farm conditions and circumstances characterizing various agro-ecological regions; (c) related to the above, adaptive research is insufficient in quality and quantity (para. 2.08); (d) the motivation of extension staff is poor due to insufficient equipment, training, supervision and salaries; and (e) the availability of farm inputs is limited and the improved seed production distribution system is ineffective (para. 2.10). The extension and other agriculture service systems are not contributing to needed agricultural intensification. They are costly and have not proven to be replicable although several regional projects present features which could be introduced into a national program. 2.07 Livestock Activities. Government efforts to develop the livestock sector have been directed both at ranching and integration of livestock and smallstock within the farm system. There are numerous ongoing efforts throughout the country to encourage permanent and semi-stall feeding of livestock (the fermette" concept), and to upgrade the quality of smallstock. MINAGRI has also tried to develop a network of livestock services including veterinary services. These efforts have been developed in the framework of externally-financed IRD projects and have not led to institutionalization of these services. C. Agriculture Research 2.08 L'Institut des Sciences Agronomiques du Rwanda (ISAR) created after independence in 1962 is responsible for agricultural and livestock research. Up to 1981, ISAR was supported mainly by Belgian aid and received limited funds from IDRC-Canada (sorghum, triticale and oil seed crop breeding), the Swiss Government (forestry, coffee and some food crop research) and from other donors through rural development projects. ISAR manages seven research stations throughout the country (IBRD Map 19380). Traditionally, ISAR's efforts have been mainly directed to improving husbandry standards for industrial crops (coffee and tea) and introducing and testing new varieties of selected food crops (e.g. maize, sweet potato and beans). In general, research efforts have been conducted in isolation from small farmer problems and farming systems, and without effective coordination. ISAR's activities were reduced considerably after 1981 when Belgian support stopped. To remedy the situation, the recently approved IDA-financed Agricultural Research Project (Credit 1546 RW) was designed to: (a) provide for technical assistance to develop a long term national research strategy; (b) introduce a national multidisciplinary research  - 10 - approach organized by commodity; (c) strengthen farming systems and adaptive research; (d) strengthen research management, programming, evaluation and review; (e) establish a documentation and information system; (f) strengthen capabilities for research monitoring and evaluation; (g) strengthen linkages with the international research centers; and (h) develop a seed certification, multiplication and distribution system. D. Farm Inputs 2.09 Modern farm inputs are not widely used in Rwanda despite a relatively dense network of small rural shops and markets, cooperatives and informal farmers' groups which have the facilities to distribute them. Farm inputs are imported by several private importers located in Kigali by a large cooperative enterprise (TRAFIPRO) and by several bilateral aid supported projects. The limited use of modern farm inputs is caused by their limited availability, lack of farmer knowledge and demand, and often because of farmer aversion to taking the risk of spending limited cash income on inputs with unproven benefits. There are, however, cases in Rwanda where the value of certain kinds of inputs have been proven (potato seed in the west and insecticide for coffee plants), creating a demand met swiftly by the network of cooperatives and private suppliers. Government involvement in input procurement and distribution is mainly related to inputs for coffee, and to a lesser extent, to animal health drugs and vaccines. The Government's strategy for the intensification of agricultural and livestock production relies largely on increased and more widespread use of agricultural inputs. The pace at which this can occur depends on a number of factors such as: (a) the time necessary to adapt new inputs to farming systems (crop treatments, farm tools, improved plant material, tree varieties and breeds of livestock); (b) the need to redefine Government policy on subsidization and cost recovery both of which affect adoption by farmers; (c) the necessary development of efficient channels for distributing and servicing new inputs; and (d) the development of marketing opportunities for the increased marketable production. 2.10 Improved breeder seeds, shoots and cuttings are produced by the research establishment (ISAR). From these, a national Selected Seed Service (SSS) project supported by Belgian aid produces base seeds and cuttings which are multiplied and distributed by regional agricultural development projects. This system is characterized by: (a) insufficient collaboration between ISAR, SSS and the regional projects; (b) lack of trained staff and infrastructure for seed production, multiplication and distribution; (c) no formalized seed control and certification system; (d) very high production costs at all levels; (e) insufficient multi-location trials of varieties proposed to farmers; and most importantly (f) scarcity of productive varieties and poor quality of breeder seeds. One of the objectives of the IDA-supported Agricultural Research Project (para. 2.08) is to develop a national certified seed multiplication and distribution policy. Seedlings (forest and fruit trees) are produced by MINAGRI and the regional projects through tht establishment and operation of tree nurseries in all communes.  - 11 - E. Agricultural Produce Marketing and Pricing 2.11 Price incentives and local marketing conditions are critical determinant of farmers' production decisions. With mounting land pressures and a growing demand for food crops, the development of more efficient local and inter-regional marketing will become increasingly important in providing incentives for intensification of production. The private sector is dominant in agriculture and in agro-industry and recent surveys (USAID) have found the business climate to be good in Rwanda. However, the marketing system is largely undeveloped for crops other than the industrial crops. The main constraints to development of better marketing systems and of rural businesses include Jack of credit, lack of entrepreneurs and managers, lack of investment ideas and of knowledge about how to develop investment projects. Finally, the marketable surplus of agricultural products is limited. Government involvement in agricultural produce marketing is mainly in coffee and tea, which Government controls for strategic reasons given the overwhelming importance of these two crops for Rwanda's balance of payments. Government interventions in food marketing and in agricultural prices are limited at present to the distribution of food aid, handling of public sector procurement and limited price stabilization measures (through the crop purchases of a parastatal, OPROVIA). Government has pursued a policy of some balance between private and public sector involvement in agricultural marketing and processing. For tea and pyrethrum, the Government is considering divesting existing parastatal enterprises to the private sector. F. Agricultural Credit and Finance 2.12 Rwanda's financial system is simple and consists of the following institutions: the central bank (BNR), three commercial banks (Banque Commerciale du Rwanda, Banque de Kigali and Banque Continentale Africaine - Rwanda (BACAR), a network of cooperative banks (Banques Populaires), a Savings Association (Caisse d'Epargne du Rwanda), a mortgage bank (Caisse Hypoth6caire) and a development bank (Banque Rwandaise de Developpement - (BRD) ). All these institutions have headquarters in Kigali, but some of them have branches in the interior of the country. With the exception of the commercial banks, Government-owned financial institutions are supervised by the National Bank of Rwanda (BNR). BNR regulates Bank equity, equity investments by banks and exposure to any single erterprise. Rwanda's financial system is highly specialized. Because of the short-term nature of bank resources, commercial banks may not, unless authorized by BNR, engage in long-term lending (more than four years). Long-term lending is the domain of institutions such as BRD and the Caisse Hypoth&caire. Commercial banks are mainly active in import and export financing and overdrafts to local companies. 2.13 The rural banking establishment is the Union des Banques Populaires (Union of Village Cooperative Banks). There are now 65 such  - 12 - village cooperative banks in Rwanda, of which 61 are in rural areas. Their number is expanding at a rate of five to seven new banks per year. In 1983, total credit outstanding grew by 13% and savings by 27Z. The objective of the Banques Populaires is to promote savings, provide credit to members in the agricultural, artisanal, small-scale industry and commercial domaine and contribute to the development of cooperatives in Rwanda. These banks hold 23% of the national savings and 4% of the national credit outstanding. In 1983, they had 60,400 individual members, 66,700 saving deposits and 5,050 loans outstanding (most of them for housing and short-term commercial operations). 2.14 The lending policy of the Banques Populaires is conservative. The banks are managed independently from the Government but are subject to Government (BNR) rules and regulations. Most loans are extremely small (65% lower than US $500 equivalent). The Banques Populaires have wisely concentrated on institution building, on gaining the confidence of their largely rural members and on good financial management. They are eager to develop their lending activities for production purposes in the agricultural and agro-industrial sector. To this end, the Banques Populaires have engaged in a strategy to decentralize their apex office to the prefectures, to continue the expansion of their local bank network, to develop training and management assistance, and to identify new lending methods. 2.15 Timely loan recoveries (as a percent of amounts due) have averaged 80%. This results from the good banking management system established by the Banques Populaires including a relatively good loan recovery mechanism. In addition, the Banques Populaires have established their own guarantee fund which currently amounts to RF 24.0 million. The interest rate on loans by the Banques Populaires is 9% to lZ compared to a rate of price inflation in Rwanda estimated at 6.6Z in 1983 and 5.4% in 1984. However, this lending rate is lower than that of BRD which varies from 11% to 13% depending on the type of loan. This lower lending rate is justified by lower costs of their own resources (demand deposits) and low operating costs due to their cooperative structure (unpaid volunteer Board members, free office accommodation provided by collectivities, etc.). The Banques Populaires receive an operating subsidy from the Swiss Government (US $500,000 equivalent in 1983). Since the Banques Populaires' operating results have considerably improved over the previous years (the operating ratio excluding subsidy was 1.96 in 1980, 1.44 in 1981, 1.27 in 1982 and 1.05 in 1983), the level of the subsidy is expected to be substantially lower in the future. 2.16 The Government is encouraging the Banques Populaires to expand their agricultural credit activity in rural areas. The Banques Populaires have also set more ambitious institution building objectives. These include improved financial management and accounting consistent with an increasingly large and complex operation, increased infrastructure development, improved staff training, establishment of more formal contractual agreements with the Government, improved personnel policy, development of regional banking centers (at the prefectoral level),  - 13 - expanded lending in the area of agricultural marketing, agro-industry and artisanat. The Banques Populaires do not require a line of credit since member savings are considered adequate to cover expanded rural lending in the near future. G. Cooperatives and Farmers' Groups 2.17 The number of cooperatives and farmers' groups in Rwanda is estimated at about 3,500, of which 80% have less than 50 members. Support to cooperatives is provided by the Direction Generale du Mouvement Cooperatif in the Ministry of Youth and Cooperatives (MINIJEUCOOP). The Direction Generale has a budget equivalent to US $300,000, of which 80% is for salaries. Its 143 front line agents based at comninal level are under the administrative authority of the bourgmestres. Cooperatives are supported by a number of donors: (a) the local Crop Storage Project (financed by USAID with technical assistance from the Cooperative League of USA-CLUSA) was designed to construct 40 cooperative silos; (b) the Cooperative Research and Training Center (CFRC-IWACU) supported by USAID and Swiss aid has emerged as a central institution of support to cooperatives; IWACU carries out training and research on the economic, social and management aspects of cooperative development, its budget partly covered by training and service fees; and (c) a Belgian NGO, the Compagnons bStisseurs or COOPIBO, appraised a cooperative support project in the Gitarama prefecture to be implemented beginning in July 1985 in association with IWACU; this project will establish a Cooperative Service Center in Gitarama (CSCG) to provide cooperatives with management, marketing, legal and accounting assistance. H. The Gitarama Region 2.18 The region chosen for the implementation of the proposed pilot activities is the Gitarama prefecture (IBRD Map 19381). It covers an area of 2,188 km2 or 8.3% of the country's total area. It is.the only prefecture with no international border and it occupies a central location in the country. The population is estimated at about 723,000 inhabitants in 1983 (with a 3.7% growth rate), or 12.1% of the country's population. The prefecture has 17 com-ines whose area vary from 92 to 121 km2. It has 171 sectors. The population of the commines varies from 26,000 to 57,000 (1978 figures) with densities from 200 to 400 inhabitants per ki2. The population is almost entirely agricultural. Women are providing more on-farm labor than men (slightly more than 50%) and are more frequently present on the farm. Twenty-six percent of all smallholdings are directly managed by women. Work on the farm is divided between sexes with women more occupied with food crops and smallstock on the fields closer to the house. Men generally handle the cash crops and cattle and crops cultivated on fields away from the farm house.  - 14 - 2.19 Food crops are the principal economic activity. The main food crops are beans, cassava, sweet potatoes, sorghum, maize and banana. Cassava cultivation is increasing everywhere and particularly in the poorest areas where it serves as a critical food reserve. Coffee is the only significant industrial crop. There are 11 million coffee trees shared between 98,000 farms, but they occupy less than 1Z of the total cultivated area (against 4X for banana cropping). Despite natural handicaps and the large dispersion of individual plantations, the Gitarama prefecture produces 20% of the country's coffee. Other industrial crops are limited: rice is grown on 140 ha in the Mukunguli valley and sugar cane trials are being carried out in the Klgoma area. Fruit and vegetable crops are relatively well developed in most of the communes. The status of planted forests (9,500 ha) is poor. They tend to be overexploited partly because there are no management arrangements for rational exploitation and because of strong pressures on the demand side as alternative sources of construction wood or domestic energy are limited. Although there is little space available for new forest plantations, the potential for increasing forestry resources remains significant and existing forests (eucalyptus, grevillea and cypress) should in many cases be upgraded or reconverted for better productivity. Livestock traditionally plays an important role in the prefecture with 130,000 cattle (16% of the country's herd) and 143,000 sheep and goat. The number of pigs is also significant at 15,000. 2.20 The Gitarama prefecture displays all of the institutional and farming problems faced by Rwanda as described above. There are several specific agricultural problems in the Gitarama prefecture which improved agricultural services would have to address: (a) mounting evidence of a decrease in soil fertility due to the progressive decline in fallow practices, mounting pressure on land characterized by increasingly small farm size and the intensification of cropping resulting in continuous cropping; (b) shortage of animal feed because of increasing competition between agricultural production, forestry and livestock; (c) plant and animal diseases which cannot now be treated adequately due to lack of services and appropriate products; and (d) chronic shortage of wood for construction and domestic energy. Although it has benefitted less from externally financed agricultural development operations than the rest of the country, the Gitarama prefecture has been the site of a number of interventions by official multilateral (ILO and UNICEF) and bilateral (Federal Republic of Germany, France and Belgium) assistance and NGO's (e.g. Catholic Relief Services). Most of these interventions have been on a limited scale, covering a relatively small part of the prefecture and relating to a wide range of activities. III. THE PROJECT A. Sector Strategy and Rationale for the Project 3.01 The proposed project would be a key instrument for the implementation of the Government's sector stragegy outlined above with the  - 15 - twin objectives of stimulating agricultural production, employment and income and improving management of the sector. In line with these objectives, the project would include two main components: (a) the Gitarama Agricultural Production component; and (b) the MINAGRI Institutional Development component. The major objective of the Gitarama Agricultural Production component would be to test a model of agricultural extension and services and of agricultural promotion which would be more efficient, replicable at the national level and institutionally and f