Rwanda - Second National Agricultural Research Project

Document of The World Bank FOR OFFICIAL USE ONLY Report No. 11326-RW STAFF APPRAISAL REPORT RWANDESE REPUBLIC SECOND NATIONAL AGRICULTURAL RESEARCH PROJECT SEPTEMBER 16, 1993 Agriculture 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 AND EQUIVALENTS Currency Unit = Rwandese Franc (FRw) US$I = FRw 140 FRw I = US$0.0071 (as of November 5, 1992) MEASURES AND EQUIVALENTS Metric System ABBREVIATIONS AND ACRONYMS CEPGL Economic Community of the Great Lakes Countries (Communaute Economique des Pays des Grands Lacs) CIAT International Center for Tropical Agriculture (Centro Internacional de Agricultura Tropical) CIMMYT International Maize and Wheat Improvement Center (Centro Internacional de Mejoramiento de Maiz Y Trigo) CIP International Potato Center (Centro Internacional de la Papa) DANIDA Danish International Development Agency FACAGRO Faculty of Agriculture of the National University of Rwanda (Faculte d'Agronomie) FAO Food and Agriculture Organization of the United Nations FAO/CP Food and Agriculture Organization/Cooperative Program FRw Rwandese Franc GOR Government of Rwanda GTZ German Aid Agency (Deutsche Gesellschaft fur Technische Zusammenarbeit) IARC(s) International Agricultural Research Center(s) ICRAF International Council for Research in Agro-Forestry IDA International Development Association IFAD International Fund for Agricultural Development IITA International Institute for Tropical Agriculture INEAC Agricultural Research Institute for the Belgian Congo (Institut National pour l'Etude Agronomique du Congo Beige) IRAZ Institute for Agronomic and Zoological Research (Institut de Recherche Agronomique et Zootechnique, of CEPGL) IRST Institute for Scientific and Technological Research (Institut des Recherches Scientifiques et Technologiques) ISABu Institute for Agricultural Research in Burundi (Institut des Sciences Agronomiques du Burundi) ISAR Institute for Agricultural Research in Rwanda (Institut des Sciences Agronomiques du Rwanda) ISNAR International Service for National Agricultural Research MINAGRI Ministry of Agriculture and Livestock NARC National Agricultural Research Council NARMP National Agricultural Research Master Plan (Agricultural Research Strategy) NARS(s) National Agricultural Research System(s) NGO(s) Non-Governmental Organization(s) ODA Official Development Assistance (the donors) PEP Public Expenditure Program PES Permanent Executive Secretariat PFP Policy Framework Paper PNAP National Program for the Improvement of Potatoes (Programme National pour l'Amelioration de la Pomme de Terre) PPF Project Preparation Facility (Advance) PRAPAC Regional Program for the Improvement of Potatoes in Central Africa (Programme Rdgional pour l'Amelioration de la Pomme de Terre en Afrique Centrale) RPC(s) Research Program Coordinator(s) SOE Statement of Expenditure SN National Enterprise (Societe Nationale) UNDP United Nations Development Program UNR National University of Rwanda (Universite Nationale du Rwanda) USAID United States Agency for International Development FISCAL YEAR JANUARY I - DECEMBER 31  FOR OMCIAL USE ONLY RWANDA SECOND NATIONAL AGRICULTURAL RESEARCH PROJECT STAFF APPRAISAL REPORT 1/ TABLE OF CONTENTS Credit and Project Summary ......................................... i Selected Data and Documents Available in the Project File ...................... iv I. BACKGROUND ............................................. 1 A. Project Background ....................................... 1 B. Economic Resources and Their Management ........................ 1 C. The Agricultural Sector ..................................... 3 H. AGRICULTURAL RESEARCH: PAST AND FUTURE ............. 6 A. Agricultural Research in Rwanda ............................... 6 B. The National Agricultural Research Project (Cr. 1546-RW) Objectives, Results and Lessons Learned ........................ 9 C. The National Agricultural Research Strategy ....................... 11 IH. THE PROJECT ............................................. 12 A. Objectives ............................................ 12 B. Summary Description ..................................... 13 C. Detailed Features ........................................ 14 IV. PROJECT COSTS AND FINANCING ............ .. ................ 23 A. Cost Estimates ......................................... 23 B. Project Financing ........................................ 25 C. Government Budget Impact ................................. 26 D. Procurement ........................................... 26 E. Disbursements .......................................... 28 F. Accounting, Finance, Administration and Audit ..................... 29 /I This report is baed on the findings of a Bank appmial minion which visitd Rwanda in September 1992; Measrs./Mmes. J. Weijenberg, Principal Agriculturalist and mision leader, J.C. Fayd'herbe, D. Prevoo, R. Contant (Consultant ISNAR on Research Programming), H. Muhtar (Consultant on Station Management), 0. Leprince (Consultant on Documentation), U. Scheiddeger (designated by Swis Cooperation, asisted in appraising on-farm research and linkages with extension) nd M. Barnaud (Consultant designated by OWZ) were the mission members. The following did not participate in the appris mnission itself but contributed significantly: Mr. A. Yates and Mr. D. Aw aited in appraising the rearch conponent, Mr. J. Weetijen aisted in appraising the resarch/extension linkAges. Mr. A. Pritchard (Agricultual Reoserch Lead Advisor) and Mr. M. Tourd (Permanent Secretary of SPAAR) reviewed the project design. Ms. D. Favreau and Ms. I. Razarumnalina provided ecretl support in the preparation of the repoit. Mr. Nile Tcheyan and Mr. Francisco Aguirre-Saca are the managing Division Chief and Department Director respectively for the operation. r 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.  V. ORGANIZATION AND MANAGEMENT ............................ 30 A. Institutional Reform ...................................... 30 B. Internal Linkages: Research - Farmer - Extension Interaction ......... ... 34 C. External Linkages: Regional/International Collaboration ............ ... 35 D. Research Proposal Presentation and Justification .................... 35 E. Research Program Planning, Monitoring, Evaluation and Review (PMER) Procedures ......................................... 36 F. Research Support Services and Station Management .................. 37 G. Human Resources Management, Training, Recruitment, Staff Evaluation and Terms of Employment ..................... 37 H. Consolidated Funding Mechanism ............................. 38 I. Investment Program (Infrastructure and Equipment) ................... 38 J. Project Implementation and Supervision .......................... 39 VI. BENEFITS AND RISKS ..................... 40 A. Anticipated Project Benefits ................................. 40 B. Employment and Poverty Impact .............................. 40 C. Environmental Impact ..................................... 40 D. Project Risks .......................................... 41 VII. AGREEMENTS REACHED WITH THE BORROWER, CONDITIONS AND RECOMMENDATIONS ................... 41 TABLES AND FIGURES IN MAIN TEXT Table 2.1 ISAR Staff (National Personnel), End 1992 Table 2.2 Evolution of ISAR Research Staff Figure 4.1 Project Costs Summary Pie Chart Table 4.1 Project Costs Summary Table 4.2 Project Financing Plan Table 4.3 Summary of Proposed Procurement Arrangements Table 4.4 Disbursement Categories SUPPORTING ANNEXES, TABLES, CHARTS AND MAPS ANNEX 1 Verifiable Indicators for Research ANNEX 2 NARC Draft Statutes ANNEX 3 National Agricultural Research Master Plan ANNEX 4 ISAR Summary Financial Statements (past and future) ANNEX 5 Project Costs Summary ANNEX 6 IDA Disbursements ANNEX 7 Project Supervision Plan ANNEX 8 Other Donors' Financed Research Programs ANNEX 9 Agricultural Research Policy Letter Chart 1 NARC Organization Chart Chart 2 ISAR Organization Chart Chart 3 Project Procurement and Implementation Schedule Chart 4 NARC: Review and Decision Making Process for Research Funding Mechanism Map: IBRD No. 24298  RWANDA SECOND NATIONAL AGRICULTURAL RESEARCH PROJECT STAFF APPRAISAL REPORT Credit and Project Summary Borrower: Republic of Rwanda Benefidaries: The National Agricultural Research Council (NARC) and the Institute for Agricultural Research (ISAR). Amount: SDR 10.9 million (US$15.0 million equivalent) Terms: Standard IDA terms with 40 years maturity Onlending Terms: Proceeds of the Credit would be passed on as a grant to NARC and ISAR. Project Description: (i) Objectives and Content: The objective of the project is to ensure that adequate technology becomes available to farmers, as a means to increase agricultural growth to at least 4 percent per annum. The project would improve quality, relevance and accountability for results of priority agricultural research programs through institutional reforms. The project would: (a) support the establishment of an autonomous NARC, through which all official funding for agricultural research would be coordinated; (b) support the transformation of ISAR into an autonomous institution, and reorganize and strengthen it to improve research efficiency and accountability for results; (c) strengthen internal linkages (research - farmer - extension) to ensure research relevance and to accelerate the dissemination of research results; (d) strengthen external linkages and support the participation of Rwandese institutions and scientists funded through NARC in regional and international collaborative research programs; (e) strengthen human resources development and management and provide in-country and overseas training; (f) support research contracts with public, private and non-governmental institutions, both inside and outside Rwanda, funded through NARC; and (g) rehabilitate and renew research plant and equipment.  ii (ii) Benefits. Long term benefits are expected from: (a) accelerated agricultural productivity; (b) improved management of Rwanda's natural resources; and (c) increased foreign exchange earnings. (iii) Risks. The principal risks of the project are: (a) implementation delays as a result of continued political instability; (b) inadequacy of public funding, (c) weak research management; and (d) Government reluctance to rely on regional collaboration. The design of the project includes measures to reduce the non-political risks. Estimated Costs: FRw (millions USS (thousands X X TotaL Foreign Sase Local Foreign Total Local Foreign Total Exchangs Costs A. Stations 1,198 470 1,668 8.556 3.354 11.910 28 37 B. Research Program 1,117 554 1,671 7,977 3,959 11,936 33 37 C. Central Services 249 138 387 1,7 989 2.767 36 9 D. MARC Permanent 42 11 53 301 77 378 20 1 Executive Secretariat E. Training 226 135 361 1,611 968 2 579 38 8 F. Consultant Services 20 354 374 140 2,531 2,671 95 8 Total BaseLine Costs 2,851 1,663 4,514 20,364 11,877 32,241 37 100 Physical Contingencies 148 38 186 1,054 272 1,326 21 4 Price Contingencies 310 96 406 2,211 686 2.897 24 9 Total Projects Costs 3,308 1,797 5,105 23,629 12,835 36,464 35 113 Financing Plan: (USS million) Local Foreign Total X Total Goverrnmnt 8.4 0.0 8.4 23 Beneficiary Institutions 1.4 0.0 1.4 4 Other Donors 7.6 3.9 11.5 32 Private Sector 0.2 0.0 0.2 1 IDA 6.0 9.0 15.0 41 TotaL 100  Estimated IDA Disbursements: (US$ million) IDA Fiscal Year (USS miLlion equivalent) 1994 1995 1996 19972 ' 1122 Annual 2.1 2.4 3.0 2.4 2.7 2.4 Cumulative 2.1 4.5 7.5 9.9 12.6 15.0 Rate of Return: Not Applicable Poverty Category: Not Applicable Map: IBRD No. 24298   iv RWANDA SECOND NATIONAL AGRICULTURAL RESEARCH PROJECT STAF F APPRAISAL REPORT Selected Data and Documents Available in the Project File 1. Mission Reports: - ISNAR Reports (Huhn, Corbett, Contant) - Consultant Reports (Winterbottom, Muhtar, LePrince) - Aide Memoires (January 1990, November/December 1991, October 1992 (Appraisal) 2. FAO/CP Preparation Report 3. Plan Directeur de la Recherche Agricole (Draft of January 1993) 4. 1993 ISAR Experiments and Studies: Titles, Objectives, Period and 1993 Indicators 5. 'Recherche et Valorisation de la Recherche", Seminaire d'Echanges et de Reflexion, Rubona, 28-29 Septembre 1992. 6. Compte-Rendu de la Table-Ronde des Bailleurs de Fonds Tenue le 5 Octobre 1992. 7. Proces-Verbal du Seminaire-Atelier sur le Plan Directeur de la Recherche Agricole Tenu I l'Hotel Amahoro du 19 au 21 Novembre 1992.   1 RWANDA SECOND AGRICULTURAL RESEARCH PROJECT STAF'F APPRAISAL REPORT I. BACKGROUND A. Project Background 1.1 Since 1973, the International Development Association (IDA) has been assisting the Government of Rwanda in the development of its agricultural sector. Support to agricultural research was initially made through components within area development projects. A credit for the first self- standing national agricultural research project was approved in December 1985. It was recognized that success in this field required a long-term commitment. For this reason the project was conceived as a first time-slice of assistance in building country capability for this vital service. One of the key objectives of the project was the preparation of a Master Plan for agricultural research. This was achieved in 1990, with assistance from the International Service for National Agricultural Research (ISNAR). It is the foundation for the present project. A first preparation report was completed by the Food and Agriculture Organization of the United Nations/Cooperative Program (FAO/CP) at the end of 1989. It was subsequently modified and amended in the course of further project preparation, to take account of recent economic and political developments and to incorporate ideas generated by the Special Program for African Agricultural Research (SPAAR) initiatives. B. Economic Resources and Their Managernent Economic Performance 1.2 Rwanda is a landlocked country covering an area of approximately 26,000 square kilometers. With a total population estimated at 7.8 million in 1992, increasing at an annual rate of 3.0 percent, it is one of the most densely populated countries in Africa (295 persons per square kilometer). In 1992, per capita GDP was estimated at US$250 and life expectancy at 49 years. 1.3 Over 90 percent of Rwanda's population lives in rural areas and derives its livelihood from subsistence agriculture and the cultivation of coffee and tea. Agriculture accounts for about 40 percent of total output, and for the employment of some 90 percent of the active labor force. Coffee and tea account for about 80 percent of the country's exports. The industrial sector remains small and inefficient. After years of contraction in mining output, the mining of tin, the principal mineral, virtually stopped in 1987, in part because of an uncompetitive exchange rate. 1.4 From its independence in 1962 to the early 1980s, Rwanda followed generally prudent fiscal policies that resulted in financial stability, low inflation, and sustained growth (averaging 4 percent per annum). By contrast, in the second half of the 1980s, Government responded to the precipitous decline in world coffee prices and adverse weather conditions by introducing economic controls rather than correcting macro-economic imbalances and, in particular, adjusting the exchange rate. This  2 policy response proved counterproductive and contributed to declining per capita income and serious budgetary and balance of payments problems. In 1990, after protracted policy discussions, both internally and with the Bank and the International Monetary Fund (IMF), the Government requested the help of these two institutions in preparing an adjustment program, supported by an IDA Structural Adjustment Credit (Cr. 2271-RW) approved on June 19, 1991. 1.5 The adjustment program was initiated following the invasion by Tutsi refugees from neighboring Uganda. A low level of hostilities prevailed in 1991 and 1992, which nevertheless led to unsustainable security related expenditures and disrupted economic activities. In response, the Government initiated a political reform process which has continued and intensified despite the clashes with the invaders. A new constitution was adopted in June 1991 replacing the single party model with a multi-party political system. Nine new political parties have been formed and a free press has emerged. In April 1992, a transition cabinet headed by the representative of one of the opposition parties was set up. It engaged in direct negotiations with the invaders, which led to the peace accord signed on August 4, 1993. This peace accord provides for: (1) a reduced national army (one third of the current 40,000-strong contingent), shared between the Government (60 percent) and rebel forces (40 percent); (2) a transitional period of 22 months leading to parliamentary and presidential elections; and (3) an interim political power-sharing system including all factions, with executive power held in the cabinet rather than the President. The transition government to be appointed shortly will have to confront the daunting task of consolidating the peace process, demobilizing troops, resettling refugees and bringing back the economy onto the path of adjustment. Structural Adjustment, Policies and Results 1.6 Starting in late 1990 (in the midst of the invasion), the authorities have implemented some of the most important--and politically difficult--upfront measures of the adjustment program: (i) devaluing the Rwandese franc by 61 percent in foreign currency terms; (ii) increasing the minimum interest rate on one-year deposits from 6 to 10 percent and the maximum lending rate from 12 to 16 percent, freeing all other deposit interest rates, and introducing a unified rediscount rate of 14 percent; (iii) increasing petroleum prices by 79 percent on average, to pass on to the consumer the full impact of the devaluation (bringing the price of gasoline to the equivalent of US$4.20 per gallon); (iv) setting a minimum import duty of 10 percent; and (v) almost doubling the sales tax to an average of 10 percent. 1.7 The above measures were taken in the context of a new economic strategy, adopted in late 1990 and reflected in the 1991 Policy Framework Paper (PFP). The new strategy aimed at: (i) fostering an environment conducive to private activity and international competitiveness; (ii) improving public resource management; (iii) developing the human resource base; and (iv) improving the management of natural resources. The key targets under the 1991-93 PFP were economic growth of about 4 percent a year and an inflation rate of 5 percent by 1993, while reaching a sustainable budgetary and balance of payments situation. Agriculture was to continue to be the mainstay of the economy and was expected to grow at 3.7 percent per year; growth was expected to average 4.2 percent in industry (including mining) and 4.0 percent in trade and tourism. A key assumption was that the armed conflict would be settled by mid-1991. However, this did not happen. 1.8 At the end of 1992, the economy of Rwanda was in severe distress, mainly from the widespread economic disruptions caused by the war situation, and to a lesser extent from depressed domestic demand--associated largely with low coffee prices. GDP is estimated to have declined by 1.2 percent in 1990 and 2.9 percent in 1991. In 1992, real GDP is estimated to have grown by about 2.5 percent, due to a good coffee crop (35,000 tons) and a substantial increase in beer production.  3 The sectors most severely affected have been manufacturing, transport, trade and tourism. Internal commerce has been thoroughly disrupted by security measures and a large proportion of external trade has had to be re-routed, with a resulting increase in costs. Also, the country has to cope with an increasing refugee problem as people have been fleeing insecure areas along the northern border. 1.9 Inflation, which had increased to 9 percent in 1990, rose further to 15 percent in early 1991. However, these numbers are heavily influenced by the timing of a large devaluation (40 percent) in November 1990 and a smaller one (15 percent) in June 1992. Inflation in 1991 was half the PFP target of 28 percent, it decelerated rapidly through the year to around 10 percent by December 1991 and 8 percent in 1992. With price, credit and monetary targets broadly met, the major departure from the macroeconomic framework was in public finances, which experienced a substantial revenue shortfall (owing to the recession), combined with expansion in military expenditure (due to the persistent hostilities). Rwanda's military spending had traditionally been around 2 percent of GDP, well below the sub-Saharan African average; it jumped to 8 percent in 1992. This resulted in a compression of current civilian expenditure below program targets. Among other things, this compression has badly affected social expenditures and the "safety net" that had been incorporated in the adjustment program. 1.10 The serious deterioration in the composition of public expenditures is a paramount and urgent problem. Unless reversed quickly, the compression of essential civilian expenditure and deterioration in the quality of basic services are likely to have negative effects on long-term development and jeopardize the objectives of the adjustment program. Accordingly, further adjustment lending to Rwanda will be contingent on satisfactory annual agreements on public expenditure programs. C. The Agricultural Sector Sector Profile 1.11 Rwanda represents one of the most extreme cases in sub-Saharan Africa where high population growth, stagnating agricultural production and a degrading natural resource base are mutually reinforcing factors, together severely limiting prospects for economic growth and poverty alleviation. The most important challenge the country has to address is the pressure of population on scarce land resources (400 persons per square kilometers of arable land). Therefore, the Rwanda Population Project (Cr. 2272-RW), is aiming at: (i) reducing the total fertility rate; (ii) decreasing maternal child mortality and morbidity; and (iii) integrating a demographic dimension in socio- economic planning. 1.12 An environmental review took place between 1988 and 1990 and a National Environmental Action Plan was drawn up in 1991. The main findings of the review were: (i) the extent of soil degradation from population pressure; (ii) exhaustion of soil nutrients from over-cultivation; (iii) destruction of forest for fuelwood; and (iv) insufficient knowledge of technologies linked to resources conservation. Because of the pressure on land, farm size is diminishing and holdings are becoming increasingly fragmented. Over half of one million farm holdings are less than one hectare. Farmers have responded to the decline in relative land availability by intensifying production: fallow land has virtually disappeared, much land is double cropped, and inter-cropping practices are widespread. In the absence of any significant use of chemical fertilizers and with the scarcity of mulch and manure, fertility of certain fragile (non-volcanic) soils is rapidly decreasing. Strong and pervasive soil conservation efforts promoted by Government, such as anti-erosion ditches, agro-forestry techniques  4 and terracing, have been able to limit erosion damage, but have not prevented the decline in soil fertility. As a result, average crop yields are declining. 1.13 Rwanda's performance in the agricultural sector in the first two decades after independence in 1962 was a remarkable and rare success story in sub-Saharan Africa. During this period, the increase in agricultural output averaged 7.2 percent per annum and fuelled a 4 percent growth rate in the economy, well above the population growth rate of about 3.5 percent. This was achieved primarily by bringing new land under cultivation, increasing crop intensities and expanding coffee and tea production. Since the early 1980s, however, agricultural performance has deteriorated significantly. From 1980 to 1984, agricultural growth just matched population growth; between 1984 and 1988, agricultural output stagnated. It dropped by 12 percent in 1989 because of crop failure and recovered by only 5 percent in 1990. In 1991, it was virtually stagnant, but again partly recovered in 1992 due to good weather conditions. 1.14 In conclusion, little increase can be expected from area expansion and increased cropping intensity. Future increases must come from productivity gains. This would require an expansion of cash crop production and/or improvement of their quality and the introduction of new high-value crops to provide farmers with the means of purchasing inputs to increase yields and arrest soil fertility decline. To achieve this goal, the development of relevant, environmentally sensitive, agricultural technology and its dissemination should be forcefully pursued by strengthening research and extension services. Sector Institutions 1.15 The Ministry of Agriculture and Livestock (MINAGRI) has a broad mandate for the management of the following sub-sectors: crops, animals, forests and other natural resources (land and water). The Minister is assisted by a Director General and commands five directorates: Plant Production, Animal Production, Forestry, Rural Engineering and Soil Conservation, and Planning (also including the performance monitoring and evaluation function). MINAGRI employs about 2,500 staff. 1.16 At the field level, the organization of MINAGRI follows that of the territorial administration: prefectures (10), sub-prefectures (41), communes (143) and sectors (approx. 10 per commune). In principle, each sector is assigned one extension agent. Recently, a unified extension service has been established under the auspices of the Plant Production Directorate, which now has the mandate to supervise all agricultural extension activities. 1.17 Agricultural extension, research and related services are implemented through about 80 donor assisted projects. The most important donors that support agricultural extension services are: Belgium, Canada, France, Germany, Switzerland, the European Economic Community, International Fund for Agricultural Development (IFAD), United Nations Development Program/Food and Agriculture Organization (UNDP/FAO), and IDA. The different approaches are very much donor- determined. Efforts are under way to unify these approaches through improved donor coordination. 1.18 A number of parastatals are also under MINAGRI tutelage. The most important amongst these are: Institute for Agricultural Research in Rwanda (ISAR), the Rwandese Office for Industrial Crops - Coffee, and the National Office for the Development and Marketing of Food Crops and Livestock. Tea, pyrethrum and sugar are entrusted to parastatals under the tutelage of the Ministry of Industry.  5 Government Sector Policy Objectives and Strategy 1.19 The main objective of the new strategy by the Government for the sector is to reverse the stagnation and decline of agricultural output. To reach this objective, the Government intends to pursue the following three-pronged strategy: (a) shift policy emphasis from production for subsistence to production for market. As the absorptive capacity of the domestic market is still very small, priority has to be given to export growth and diversification; (b) increase agricultural output through the development and dissemination of new technologies, emphasizing those that are likely to bring quick and significant returns (in the first instance, plant and animal husbandry techniques); the ongoing Agricultural Services Project (Cr. 2026-RW) and the presently proposed project would together pursue this goal; (c) arrest environmental degradation resulting from land scarcity in the face of unabated population growth. A National Environmental Action Plan (NEAP) has been prepared (para. 1.12) and was presented to a donor Round-Table in June 1992; Government and donors expressed high commitment to its implementation. Bank Group Involvement in Agriculture 1.20 IDA Financing. Since 1973, IDA lending has amounted to US$135.9 million for eleven operations, of which two are ongoing. The operations are as follows: five in area development, two in forestry, two in extension/agricultural services, one in perennial crops (cinchona) and one in agricultural research. 1.21 Performance Evaluation. Project Completion Reports have been prepared for seven projects and a Performance Audit Report for one project. The implementation experience has been rather poor with economic rates of return ranging from negative to barely satisfactory. Common problems encountered have been: (a) paucity of technical messages affordable to farmers and able to yield quick and significant returns; (b) insufficient knowledge of the complexity of Rwanda's farming systems; (c) weak links between research, extension and farmers; (d) lack of experienced local managers and weak sector policy analysis and implementation capacity; (e) inadequate performance monitoring and evaluation; (f) insufficient counterpart funding, mainly from 1988 onwards, and (g) unclear sector policies and investment priorities, combined with poorly coordinated donor support. The present project's design incorporates lessons learned from this experience. IDA Lending Strategy for Agriculture In Rwanda 1.22 In view of the constraints facing the sector and the implementation experience, IDA strategy for lending to agriculture focusses on: (a) improving the testing and adaptation of new technologies and the dissemination of proven technologies to farmers, through strengthening the delivery of basic services, namely agricultural research and extension, under long-term projects that are well coordinated with other donors; (b) improving the conservation and husbandry of natural resources through financing of forestry sector investments and funding of activities identified in the Enviromnental Action Plan (para. 1.12); and (c) providing an enabling environment for export promotion and diversification, the promotion of private sector and non-governmental organization (NGO) initiatives and involvement through sectoral adjustment operations.  6 II. AGRICULTURAL RESEARCH: PAST AND FUTURE A. Agricultural Research In Rwanda 2.1 Historical Background. ISAR was created just before independence in 1962. It combined on a national scale all of the facilities which had been installed by the Belgian colonial research organization, the Agricultural Research Institute for the Belgian Congo (INEAC). Research in Rwanda has traditionally emphasized food and forestry rather than industrial crops. This tendency was strongly reinforced in the 1970's and 1980's with the establishment of numerous donor supported projects carrying out farming systems research and natural resources management activities. Institutional Framework 2.2 ISAR was given the mandate for all aspects of publicly-funded research dealing with agriculture: soils and natural resource conservation, crops, livestock, forestry, fisheries and apiculture. This mandate is however not exclusive. In the public sector, the Faculty of Agriculture (FACAGRO) and the Department of Geography of the National University of Rwanda (UNR) undertake some agricultural research, as does the Institute for Scientific and Technological Research (IRST). Other significant research initiatives are funded separately by donors. The tendency of donors to create "enclave" projects has limited the degree of coordination exercised by ISAR, especially in farming systems research, where coordination and cooperation between individual projects has been limited; and in forestry where an excellent program has been funded and managed almost entirely by Swiss aid. Regional cooperation between Rwanda, Burundi and Zaire is reflected in the regional research organization, the Institute for Agronomic and Zoological Research (IRAZ). Close cooperation exists with three international agricultural research centers: the Intemational Center for Tropical Agriculture (CIAT), the International Potato Center (CIP) and the International Council for Research in Agro-Forestry (ICRAF). 2.3 The organization of ISAR is governed by Legislative Ordinance No. R/I 18/52 of 1962, and Law No. 21/82 of 1982. ISAR was established as a public enterprise under the tutelage of MINAGRI and the direction of a Board of Directors nominated by the President of the Republic. The Board is composed of representatives of ministries and institutions which have relevant responsibilities. The Director General of ISAR, who is also nominated by the President of the Republic, but acts as its secretariat. He is responsible for the day-to-day management of ISAR, both in the scientific and administrative sense. In this, he is fully responsible for the operating budget which is allocated to him annually by Government. He has little control, however, over the employment of scientific staff, who are proposed by the Ministry of Civil Service and nominated by the Board of Directors or the President of Rwanda. He is assisted in scientific management by a Research Committee. Up until August 1992, there were four research Departments: Crop Production, Animal Production, Forestry, and Farming Systems Research. Fundamental changes instituted in August, 1992, which conform to the concept of the new project, are described in para. 3.9. 2.4 About 90 percent of all agricultural researchers in Rwanda are employed by ISAR; personnel status and staff evolution are shown in Tables 2.1 and 2.2 respectively.  7 Table 2.1: ISAR Staff (National Personnel). End 1992 Grad Oualifications Number Researchers AO Ing.Ag., MSc., or Ph.D. 49 2a Technicians Al BSc 20 Technicians A2/3 High School Diploma 75 Others (as at end 1991) Contractual Staff S/C 379 Permanent labor J Unskilled 465 Seasonal labor S Unskilled 1,834 Table 2.2: Evolution of ISAR Research Staff Year Expatriates Nationals iToal 1962 (at independence) 6 0 6 1975 12 2 14 1980 (departure of Belgians) 4 17 21 1992 (end) 16 49 65 2.5 UNR. The university is severely constrained by limited human resources, and has therefore been unable to provide sufficient personnel to participate in the research programs of ISAR (for which funds were available under the first project). In addition, cooperation by the university was hampered by the cumbersome procedures involved in the cooperation between two institutions falling under different Ministries. Some limited research has been undertaken under more flexible arrangements with other donors, and both UNR and IRST have expressed strong interest in a more active participation in a national research program. 2.6 Donor-funded projects. There have been numerous donor-funded projects with major research components which have not been fully integrated into the national research program. Many have produced, or are still producing valuable results. Of the commodity specific projects, the Swiss funded forestry project continues to be successful and the region specific Birunga Maize Project (IFAD) succeeded in selecting varieties suitable for both high and medium altitude, and fertilizer regimes suitable for the volcanic soils of the north. There have been numerous regional projects with major farming systems components, such as the now completed Bugesera-Gisaka-Migongo Project (Cr. 1238-RW); and the ongoing GTZ project at Rubona. One project which emphasizes on-farm trials, is the Gikongoro Project (UNDP/FAO); this project is on-going and produces annual reports which summarize the results of about 600 trials per season. 2.7 The first phase of the joint FAO and Danish International Development Agency (DANIDA) financed Fertilizer Project was initiated in 1980; a fourth phase has recently been agreed which will continue until mid-1996. It is national in concept and uses three regional coordinators (at Gisenyi, Kigali and Butare) who cooperate with staff of regional projects and extension services for the 2l Including the Director General, but excluding eight researchers pursuing post-graduate training outside Rwanda and a total of 16 expatriate researchers who work in close association with ISAR.  8 installation of trials. The fourth phase aims at consolidation of project results, diffusion of the accumulated data, and the institutionalization of the work within MINAGRI. 2.8 Extension Services provided by various rural development projects are progressively being phased out, and are being replaced by a national extension service. The cooperative research described in the previous two paragraphs is also, therefore, progressively becoming the responsibility of the national extension service. Cooperation exists between ISAR researchers and the projects/extension services; for example, both the grain-legume and root-crop commodity research teams rely on these services to organize and maintain the large network of multi-locational trial sites and the on-farm trials, and to carry out seed multiplication. External (Regional and International) Linkages 2.9 Rwanda is part of the agro-ecological area known as the Highland Region of the Great Lakes. This region is demarcated by the 1400 meter contour, and embraces virtually all of Rwanda and Burundi, a large area of the Kivu Province of Zaire, and adjacent parts of Uganda and Tanzania. Prior to independence, research in this region was conducted in an integrated manner. In 1979, the Economic Community of the Great Lakes Countries (CEPGL: Burundi, Rwanda and Zaire) created a regional research institute, IRAZ, to coordinate and assist research activities of the member states, and to undertake research not done by the member countries. IRAZ has recently moved into newly constructed laboratories and offices at their large experiment station close to Gitega, Burundi. Operating budgets are provided by the three participating Governments I/ plus grants from donors. Technical assistance and/or cooperation is given by several international organizations including CIAT, the International Maize and Wheat Improvement Center (CIMMYT), CIP, the International Livestock Center for Africa (ILCA), the International Board for Plant Genetic Resources, FAO and the International Network for the Improvement of Bananas and Plantains. The center maintains a large gene-bank (mainly as seeds) of species indigenous to the area; it is reputed to have the largest collection of bananas in Africa (in this case, highland varieties), and has an effective tissue culture facility for the multiplication of disease free stock; it has developed an impressive documentation center, and has organized numerous meetings. However, according to a recent evaluation of the institute (March 1990), overall performance has been disappointing. The evaluation mission identified numerous factors as being the cause for poor performance, and made detailed recommendations. Implementation of these recommendations should significantly improve IRAZ's effectiveness in regional research. 2.10 Assuming that political problems can be resolved, there is every reason to integrate the research activities of the region, especially between Rwanda and Burundi. Both countries are following similar lines of research on soil conservation and fertility improvement, and in cattle improvement. As neither country has the means to undertake all of the needed research, cooperation between the national research systems (currently mainly ISAR and its Burundi counterpart, the Institute for Agricultural Research, ISABu) is essential. This is accepted by the Directors of the two institutes who meet regularly for this purpose, and by some donors. Currently, French aid has agreed to provide a single team to assist coffee research in both countries; implementation will depend on agreement for the free exchange of information, personnel and genetic material between the two countries. 2/ In 1992, Rwanda supported IRAZ with FRw 100 million.  9 2.11 Other regional research bodies include the regional bean program, guided by CIAT, and the Regional Program for the Improvement of Potatoes in Central Africa (PRAPAC), supported by CIP; there is also close cooperation between ISAR and ICRAF. CIAT has maintained a small team (funded mainly by Swiss aid) at Rubona since 1983; this team has worked closely with ISAR and ISABu, and the National Agricultural Research Institute of Zaire (INERA), supplying genetic materials and training national researchers. A new 3-year phase, which started in October 1992, reduced the number of CIAT core staff from 4 to 2 as most of the training work had been completed. The CIP supported PRAPAC program has grown into one of the more successful regional collaborative programs in Eastern Africa, now also comprising Ethiopia, Uganda and Tanzania. The Inter-Governmental Authority on Drought and Development, in collaboration with SPAAR, intends to organize a series of workshops between research leaders of participating countries to develop a plan of action to intensify regional collaboration. B. The National Agricultural Research Project (Cr. 1546-RW) Objectives, Results and Lessons Learned 2.12 The National Agricultural Research Project was identified in 1981, prepared in 1982/83 with the assistance of ISNAR and appraised in 1984; the IDA Credit (Cr. 1546-RW) became effective in November 1985. It was conceived as a short-term (3 years) first phase of a long-term support program to strengthen the national agricultural research system of Rwanda, and to make it more responsive to the needs of farmers. Slow implementation caused the postponement of the closing date to December 1991. The specific objectives of the first phase were: to prepare a National Agricultural Research Master Plan (NARMP) which would become the basis for a second phase research project; reorganize ISAR into four research departments; strengthen research programming, budgeting and financial management; provide training; establish an annual research program review system; provide for a donor coordinating mechanism; and rehabilitate four experiment stations. 2.13 Achievements. Progress has been made in many areas during the first phase. Specific achievements include: (a) The NARMP was completed in 1989 and published in 1990. Many of the conclusions remain relevant but, as was noted in the original plan, revisions are required from time to time to adapt it to changing economic and technical circumstances. (b) Standards of technical reporting and, in particular, accounting have greatly improved. High quality annual reports have been published on time since 1989; and the introduction of a computer program now allows reliable and transparent accounts to be kept by research program, research location, nature of expenditure, and source of funds. (c) The number of national research staff at AO level has significantly increased (Table 2.3), although the increase was about 17 percent short of the appraisal forecast. (d) Rehabilitation of most of ISAR's research infrastructure has been completed. (e) A series of approximately 20 external technical evaluations of the various research programs has been completed.  10 (f) Notable achievements in crop research include the release of new disease resistant varieties of beans and root crops which have been widely adopted by farmers. Of special interest are the advances made in farmer participatory research, developed within the grain legume program with support from CIAT. In the forestry program, ISAR now provides on a regular basis seeds of local tree species for farmers' use as well as for re-afforestation. Species have been selected and techniques developed for agro-forestry. Good progress has also been made in the control of erosion. 2.14 The areas in which the first project was less successful, and that would need remedial action under the second phase, can be summarized as follows: (a) The NARMP has not yet been used by Government for its three main purposes: (i) to guide individual research proposals on priority topics; (ii) to present a vision with respect to future institutional development of the national agricultural research system; and (iii) to coordinate the participation of donors around a common set of objectives to be jointly implemented. (b) The donor coordination mechanism set up under the first project has failed to produce coherence in research planning, execution and funding. Funding by donors has adopted an "add-on", "enclave project" approach. As a result, recurrent overhead and research operating costs are too high, seriously under-funded and far beyond the capacity of Government to finance. (c) Linkages with the extension services, development projects and NGOs suffered from the "enclave" approach, and on-farm participatory research is of variable quality and intensity; efforts must now be made to standardize the quality of these efforts. (d) Regional and international research collaboration among individual NARCs, along the lines of the successful CIP initiated PRAPAC program, are still embryonic. (e) Research planning is seriously flawed, with too much emphasis on traditional single commodity subjects. Too little attention has been given to work done previously or elsewhere, and to external technical evaluations. A formal and regular independent external review system had been planned; this has not been put into operation, and has therefore not provided the desired controls on research quality and relevance, acceptable to both Government and donors. Research management has failed to coordinate individual research sub-programs into coherent and cost-effective national programs. (f) The structure of the institution is still unsatisfactory. ISAR is a public enterprise with full autonomy in financial management, but with no autonomy in personnel management. The absence of a well defined human resources incentive, development and management policy is a serious handicap in making staff accountable for research results. ISAR's Board of Directors consists of civil service appointees representing the Ministries of Higher Education, Agriculture and Livestock, and Plan; there is no representation of farmers, nor of the private sector. (g) Since 1985, ISAR's research funding has encountered serious problems. Government has failed to maintain a stable budgetary support beyond the original three-year project period, and bilateral donors have been unwilling to finance ISAR's overhead  11 costs, limiting their contributions to enclave programs. In addition, Government funding has targeted civil servant salaries, and sacrificed the funding of other key expenditures. The resulting lesson is that funding from all sources of finance needs to be consolidated with the objective of avoiding major annual fluctuations and guaranteeing the availability of funds promised for any given year. Moreover, research programs have to be tailored to the available funds and research managers need to be made accountable for the efficient use of these funds. C. The National Agricultural Research Strategy Original objectives 2.15 The 1989/90 version of the NARMP was designed to support the basic Government objective of food self-sufficiency. Secondary objectives were to encourage the production of import substitution products and to improve the quality of export crops (specifically coffee and tea). Research was to address itself to the problems of the majority of the rural population; it had to deal with all major agro-ecological zones; and it had to aim at intensifying agricultural production whilst reducing environmental deterioration. These objectives were to be viewed against the background of a rapid rate of population increase, the virtual lack of unused land suitable for agriculture, and limited resources availability (both human and financial). Thus the NARMP aimed at improving overall planning, including improved coordination between the various participants in research, and between donors; it aimed at improving the management of programs through improving the calibre of researchers; and it directed research to deal only with priority programs. Revised Objectives 2.16 The NARMP has been adjusted to better support the new strategy of export growth and diversification. A seminar held in November 1992 initiated the first comprehensive review, which was completed in January 1993. This review undertook a re-evaluation of research priorities, and dealt with the improvement of research-farmer linkages, the status of researchers, and the consolidation of research funding. The review was intended to better satisfy the following principles: (a) improving research relevance (responsiveness) and quality; (b) redefining research priorities in response to: (i) farmer constraints (strengthening of interal linkages); (ii) economic policy considerations emphasizing export crop promotion and diversification; and (iii) environmental and natural resource management considerations, de-emphasizing (not abandoning) the traditional varietal improvement approach; (c) focussing research on: (i) the testing of technologies developed elsewhere; (ii) seeking regional collaboration in adapting promising technologies to local circumstances (strengthening external linkages), avoiding expensive duplication of effort; and (iii) expanding and moving faster towards on-farm testing; and (d) implementing institutional reforms to: (i) introduce the principle of accountability for results and improve staff motivation; (ii) improve cost-effectiveness and efficiency based on the principle of competitiveness (comparative advantage for  12 implementation); and (iii) ensure financial sustainability through improved coordination between sources of funds. 2.17 To meet the objectives of the new agricultural research strategy, Government and the research community have been preparing for the following reforms: (a) Institutions: (i) creating a National Agricultural Research Council (NARC) through which all public funding (including all official development aid funds) for agricultural research would be coordinated and consolidated; and (ii) transforming ISAR into a privately managed enterprise, accountable for results and with full authority for scientific, financial, administrative and personnel management. (b) Prioritization: The necessity for a continuous revision of the Master Plan is accepted, and a review has been undertaken which will give greater emphasis to export crop diversification and social sciences research. (c) Pluralism: An increasing role for IRST, and the Agronomic and Economic Faculties and the Geographic Department of the University of Rwanda is planned, and the granting of research contracts and the promotion of partnerships with other research organizations in and outside Rwanda is accepted, including partnerships with private sector operators and agro-industries; also a more formal cooperation with the FAOIDANIDA fertilizer program is being established. (d) Decentralization: Greater autonomy in budget execution for regional stations, and closer coordination between ISAR and the extension services/Projects at the regional level, is being proposed. (e) Human Resources Development: The deficient levels of training (quality and skills mix) need to be corrected by intensified recruitment and directed training. Scientist motivation needs to be strengthened through improved career structures and salary scales. (f) Organization and Management: Extending the membership of ISAR's Board of Directors to include representation of farmers' interests, agro-industries, and possibly private sector operators. The Director of ISAR would become the Director General, and would be supported by two new Directors, one for Scientific Research and one for Support Services, Administration and Finance. The following research support functions would be strengthened: (i) scientific management (programming, justification, design, analysis, communication and information), and (ii) research station management and operation (to improve research efficiency this would include the divestiture of surplus research capacity). III. THE PROJECT A. Objectives 3.1 The basic objective of agricultural research is to help farmers increase agricultural production and income per unit of land in a sustainable manner. Research must identify methods, appropriate  13 for use by peasants in Rwanda, to combat and reverse the decline in soil fertility; it must maximize yields through improved varieties and better agronomic practices; and it must reduce losses due to pests and diseases. It must protect earlier gains through "maintenance' research: for example, it must maintain disease resistance in beans in spite of mutating pathogens. Even "proven' technology needs to be verified by research; for example, experience has shown that about half of the erosion- control ditches installed throughout the country in the 1980s have not worked, as the technology is not appropriate to the steeper slopes; the technology has also proved too expensive for voluntary adoption by the typical farmer. 3.2 The project has long, medium and short-term objectives. The long-term objective is to ensure that adequate technology will become available to farmers as a means to increase agricultural growth to at least 4 percent per annum. The medium-term objective is to ensure that appropriate technology is generated to counter the most acute problems confronting agriculture in Rwanda. The short-term objective is to release for final testing and/or immediate application those technologies which are now ready for dissemination. All of these objectives must be attained within the limits of presently available resources, both in terms of funding and of human resources, it would be unrealistic to expect that they will be greatly increased. The project will further improve agricultural research quality, relevance and accountability for results to stimulate agricultural growth. To achieve this, emphasis will be on: (a) the testing of technologies developed elsewhere; (b) expanding and moving faster towards on-farm testing (strengthening internal linkages); and (c) seeking regional collaboration in adapting promising technologies to local circumstances (strengthening external linkages!, thereby avoiding expensive duplication of effort. The Government has issued an Agricultural Research Policy Letter in support of the project's objectives (see Annex 9). B. Summary Description 3.3 To meet these objectives the Project, in line with the NARMP, includes the following components: (a) support the establishment of an autonomous National Agricultural Research Council (NARC), through which all official funding for agricultural research would be programmed and coordinated; (b) support research contracts with public, private and non-governmental institutions, both inside and outside Rwanda, funded or coordinated through NARC; (c) support the transformation of ISAR into an autonomous institution, and reorganize and strengthen it to improve research efficiency and accountability for results; (d) strengthen internal linkages (research - farmer - extension) to ensure research relevance and to accelerate the dissemination of research results; (e) strengthen external linkages and support the participation of Rwandese institutions and scientists funded through NARC in regional and international collaborative research programs; (f) strengthen human resources development and management and provide in-country and overseas training; and  14 (g) rehabilitate and renew research plant and equipment. C. Detailed Features The National Agricultural Research Council (NARC) 3.4 The Government of Rwanda (GOR) will, together with the major sources of Official Development Assistance (ODA) for agricultural research, establish NARC as an autonomous institution under Rwandese law. NARC will have ultimate responsibility for agricultural research policy and the consolidation of all ODA and Government funding for agricultural research. Ultimately, NARC would only operate through the contracting out of research in line with its established policies, as laid down in the NARMP. Until December 31, 1995, NARC would also coordinate funding for ISAR in accordance with annually agreed research programs, budgets and work plans. Thereafter, all research activities funded under the project, including research activities by ISAR, will be undertaken exclusively through contractual arrangements between NARC and the eligible research institutions. Assurances to this effect were obtained at negotiations (para. 7.1 (d)). For NARC to play its role in overseeing the national agricultural research system and to administer its funding, it will constitute a small Executive Secretariat (ES), consisting of a highly qualified Research Manager and a Finance Officer, supported by secretarial staff. The establishment of the ES and the nomination of the Research Manager and Finance Officer. with terms of reference, qualifications and experience acceptable to IDA. are Conditions of Effectiveness (naras 7.2 (a) (iii) and (b) (i??. 3.5 For NARC to execute its functions, the Project will finance the following: (a) the establishment and operation of the ES; (b) the regular updating of the NARMP; the first such update has started, following the November 1992 seminar (para. 2.16), has provided the underpinning for Government's Agricultural Research Policy Letter (para. 3.2); (e) the establishment of an independent external research program review system through regular scientific and accounting audits of research programs/contracts/institutions that are funded/coordinated through NARC; and (d) priority research contracts established according to criteria and with institutions acceptable to IDA. 3.6 Before January 1, 1996, the date after which ISAR will be funded exclusively through contractual arrangements, NARC will let - alone or in partnership with other national or intemational institutions - contracts with UNR, IRST, agro-industries, private sector organizations and NGOs, collaborating universities and regional or intemational research institutions that meet the eligibility criteria. These criteria concern mainly: (a) the relevance of the research proposal with respect to the NARMP and its justification in terms of client-responsiveness;  15 (b) the scientific capacity and comparative advantage of the institution _/ presenting the proposal for funding; (c) the quality of the proposal (NARC will publish criteria for research proposal presentation (para. 5.10); and (d) the administrative and accounting capacity of the institution (para. 4.17) 3.7 NARC can request institutions to submit proposals. Institutions can also present non-solicited proposals to NARC. The preparation of a research proposal is time-consuming and costly (may take up to 30 percent of the time of a scientist). Therefore, NARC may provide funds to prepare a research proposal without the guarantee of a follow-up contract. Such proposals and possible publications as a result of state-of-the-art analyses will become the property of NARC. Certain agro- industries may be better placed to undertake and fund their own research, though start-up financing might be needed; this participation is of such potential importance that the project will encourage it through the provision of such financing (e.g. for a tea research program). In partnership with private investors, NARC will, through the provision of "seed' money, promote research on the adaptation of technologies available elsewhere and necessary to diversify exports. Similarly, cooperation with NGOs will be encouraged as much as possible. Cooperation within the Great Lakes Region will be encouraged by (i) integrating regional programs (along the lines of the successful collaboration on potato research through PRAPAC, to minimize duplication and maximize economies of scale; and (ii) improving complementarity with IRAZ implemented research (see also para. 5.8). Improvement of the cooperation with international research centers should require no more than limited funding to supplement existing resources. In exceptional cases, when the research topic is beyond the scope of the Rwandese institutions or scientists, and the research topic does not lend itself to a collaborative approach, NARC may decide in Rwanda's interest to contract-out such research on the basis of competitive bidding. The National Agricultural Research Institute (ISAR) 3.8 Until December 31, 1995, ISAR will be funded through NARC on the basis of agreed priority research programs, annual budgets and work plans. Thereafter, ISAR will be funded solely through (pluri-annual) contractual arrangements. Such contracts would include a reasonable share of the indirect and fixed costs of experiment station operation (including staff costs) and maintenance and that of the central support services. 3.9 On August 5, 1992, ISAR's Board of Directors gave the go-ahead for a reorganization of the institute, outlining areas for improvement: appointing a Scientific Director with direct responsibility over all researchers, abolishing the existing central Research Departments (Crop Production, Animal Production, Forestry and FSR), decentralizing budget and accounting management to Research Program Coordinators (RPC) and Station Managers and creating Regional Research Centers representative for each of the agro-ecological zones (see Map no. 24298), under Regional Directors. To avoid adding new layers of management in the decentralization process, Regional Directors would be the RPCs for the mandated crops of the Regional Center. They would thus cumulate their technical functions as RPC with that of being the representative of the Director General in the region _4/ For example, UNR is best able to help with projects that have flexibility in timing (say, socio-economic studies), or which require only low-cost investments (say, small non- ruminant livestock).  16 and assure the formal linkage with development projects, extension services, NGOs, export crop estates and the private sector. They should n have direct administrative or technical responsibility for experiment station management. The changes decided by the Board are an important first step in providing the required framework for improving research quality, relevance and accountability for results. Under the project, these reforms will be further pursued to improve: (a) accountability through delegation of authority; administrative management will be decentralized with more responsibility given to experiment station managers; at the same time, a number of superfluous research installations (para. 4.16) will be closed, or drastically diminished in size; whereas some installations, presently operated by "rural development projects", will be added to provide for a better agro-ecological coverage; (b) research management; under the proposed project, the scientific management function will be further reinforced through the establishment of a support unit responsible for biometry; research programming, monitoring and evaluation; and the establishment of scientific data bases; (c) operation and maintenance of research installations; ISAR will nominate and train professional experiment station managers (and provide for their career opportunides); as a first step, the Director General of ISAR recently appointed an experienced manager for ISAR's largest station at Rubona; the next step will be the appointment of an Administrative and Financial Director (a "technical" function), who will be responsible for all station operations, procurement, accounting and administration, and the payroll; under the project the support services management function will be strengthened through: (i) further reinforcement of the accounting department (in particular the budgeting and internal budget control functions); (ii) maintaining ISAR's central laboratories fully operational (amongst others through the introduction of a charge-back system); and (iii) building-up ISAR's publication, documentation and library capabilities; (d) research relevance; linkages with farmers, export crop organizations, agro-industries, extension services and agricultural development projects would be improved through: (i) emphasizing on-farm testing; (ii) decentralizing and clarifying institutional responsibility (accountability) and arrangements for research - farmer - extension linkages; and (iii) including extension and clients (users of research results) in the review of research proposals and the monitoring of research progress and results; 3.10 To achieve these objectives, ISAR will be given full autonomy for scientific, financial, administrative and personnel management through a change in its legal status. Under the proposed draft law on public enterprise reform currently under review by Government, public enterprises will have to opt between the 'tablissement public' or 'soci6te nationale' (SN) status. The latter is deemed more appropriate in providing greater protection from Government intervention in the critical areas listed above, as SNs are governed by common legislation applicable to the private sector. Engaqnnt of ISAR's revised statutes and by-laws. acceptable to IDA, is a Condition of Credit Effectiveness (para, 7.2 (a) (v). Assurances were obtained at negotiations that the Director General, the Scientific Director and the Administrative and Financial Director will have qualifications and experience satisfactory to IDA (para. 7.1 (p)).  17 3.11 Priority Research Programs and On-Farm Testing. The project will support ISAR's research programs that are consistent with an up-to-date national strategy, are properly justified, and are planned to maximize the efficient use of resources. During appraisal, each of the research programs to be implemented in 1993 in the context of the NARMP, were reviewed and agreed upon, including preliminary individual budgets. Agreement was reached during negotiations on an appropriate 1993 budget for ISAR, which was passed by the Board of Directors of ISAR in April 1993. 3.12 Crop Production. The medium-term program of ISAR envisages national research on 13 commodities: bananas, beans, sweet potatoes, potatoes, maize, rice, coffee, cassava, sorghum, wheat, tea, soybeans, and sugarcane. Most work on fertilization and sustainability involves collaboration with the soil and water conservation and improvement program. Experiments on crop associations usually involve several programs. (a) The socio-economics program will support all other programs, by aiming at a better typology of farming systems, economic evaluation of proposed technologies, and post-diffusion monitoring of adoption and impact; this program will work particularly closely with MINAGRI's extension services. The commodity programs are clearly moving towards greater focus on the priority themes, with 1993 being a transition year in which some ongoing work of lower priority will be concluded and phased out; (b) The banana program will focus on organic/chemical fertilizer dose/response curves, the effects of widespread banana/food crop associations on productivity, soil fertility and erosion, optimal spraying against Cercospora, and screening for resistance to the most harmful diseases in Bugarama and Kibungo regions; (c) For beans, greater emphasis will be placed on fertilization and variety/manure/fertilizer interactions, followed by genetic improvement of yield, disease/pest resistance, tolerance to poor soils, and promotion of climbing beans; research is also underway on seed production and varietal differences in nodulation capacity; (d) The sweet potato program will concentrate on identification of and selection for multiple resistance to viruses and Alternaria, and for tolerance to poor soils, as well as purification of planting material, and monitoring of variety adoption/impact in the rural areas; priority research covers fertilization on poor soils and the study of associations/rotations; the program also foresees a marketing study and work on conservation of planting material; (e) For potato research, the accent will lay on selection for horizontal resistance to mildew and tolerance to bacterial wilt, chemical crop protection, seed production including purification from disease and storage, determination of optimal variety/fertilizer/manure combinations; research also covers integrated bacterial wilt control, and potato drying and processing; (f) The maize research proposals and their justification in light of needs and previous results, will need to be improved before being finalized;  18 (g) The rce program will focus on selection for disease resistance, cold tolerance, early maturity, grain characteristics, and pest tolerance; (h) The coffee program will reflect more rigorously the NARMP's main thrusts: maintenance/improvement of coffee quality, reduction of diseases (mainly through selection) and cultural practices (soil cover, density, economics of fertilization, pruning); (i) Priority for cassava research goes to protection from pests and diseases through selection and integrated pest management; to cultural practices, fertilization, associations and rotations, aimed at system sustainability; and to technological and economic studies of drying, flour conservation, and flour utilization; (j) Sorghum research will focus on control of Siriga (varietal screening, integrated control, species inventory), fertilization, selection for drought tolerance and disease resistance as appropriate to the various regions, and sorghum utilization technologies; (k) For wheat, the research goal will be to produce two new disease resistant early varieties every second year; however, the economic justification of the wheat program needs to be clarified first in the larger context of regional collaboration; (1) The tea research program will be finalized in collaboration with the tea production and processing units; for the moment the most urgent priority concerns fertilization; (m) Research on soybean will emphasize selection of short-season varieties and corresponding Rhizobium strains, and development of food recipes; and (n) ISAR will do sugarcane research in collaboration with the Regie Sucriere de Kabuye, under a program financed by France, focusing on selection for disease resistance, high sugar yield/ha, and earliness; and on finding the economic optima for fertilizer/lime combinations. All programs include participation in the training of extension staff. Funding for the maize, wheat and coffee programs will only be made available upon presentation of satisfactory research proposals in line with the revised NARMP. Similarly, research on sugar cane and tea will only be undertaken upon the satisfactory conclusion of research contracts, whose funding will include a reasonable share of ISAR general overhead, indirect station operating and maintenance expenditures, and personnel costs. Assurances to this effect were obtained at negotiations (paras. 7.1 (a) and (b)). 3.13 Animal Production. Animal feeding and agrostology and animal health are no longer separate programs, but are dealt with under small ruminants (goats/sheep) and cattle; for pigs and poultry, the FACAGRO is preparing national programs. Research on sheep will be limited to the Zaire-Nile watershed area and will have more focused objectives: utilization of the undergrowth (inventory, feed value, management techniques), determination of feed rations, and genetic improvement (crossing) for wool production have highest priority; improvement of the local breed for meat has a lower priority. Research on goats will emphasize utilization of available feed resources (inventory, feed value and its enhancement by additives, production and conservation techniques); identification of outstanding local animals for meat and milk, and determination in on- farm research of the optimum level of crossing with the Anglo-Nubian breed. Work on alternative feeds for both goats and sheep are envisaged at a later stage, including testing/selection of fodder  19 species, use of additives, introducing rapidly growing fodder species into the rotation as dry-season crops, and agroforestry trees and shrubs (para. 3.15). Cattle research is based at Songa; it will focus on determining through on-farm research the optimum Ankole x Jersey crossing level for milk production; while no further research will be done on the national Ankole herd. A herd of approximately 1,000 heads will be maintained on about 1,000 hectares at Karama, by controlled mating based on standard records. The Government will adopt a private sector style management with separate accounts for this operation. Assurances on this point were obtained at negotiations (para, 7.1 (c). As of March 31, 1994, ISAR's station operations at Karama will be limited to approximately 50 hectares, including fallow, on both upland and lake-shore soils. 3.14 Soil Conservation and Improvement. In the new, integrated soil conservation program (which includes water harvesting and moisture conservation), the highest priority goes to understanding the erosion processes and development of anti-erosion methods; this work, conducted with the extension services, will include a survey of farmer perceptions and strategies. Second comes the physico-chemical characterization of soils. This is a prerequisite for the third program, on fertilization, carried out in collaboration with the FAO/DANIDA financed Fertilizer Project (fertilizer formulas and fertilizer/manure interactions, for different crops, rotations and regions, including valley bottoms). The fourth priority area is promotion of organic matter: biomass from different production systems, green manure, agroforestry (para. 3.15), improved fallow, use of human excreta and Azolla. 3.15 Forestry and Agrotorestry. This area comprises three programs: natural forests, sylviculture, and agroforestry. Research on naturalfforests focuses on system dynamics (Kibungo Est, Nyungwe, Southern part of Plateau), development of sylvicultural methods to restore degraded forests, and establishment of management standards. The sylviculture program aims at improved knowledge of species, natural regeneration in buffer-zones (Nyungwe Natural Forest), fertility maintenance, sylvicultural techniques, production standards, and woodlot conversion methods; lower priority work includes genetic improvement and forest seed technology, and phytosanitary protection. The agroforestry program includes selection of multi-purpose species for different regions, efficiency of agroforestry systems for improvement of soil fertility and regeneration of degraded soils, development of agroforestry and sylvopastoral management techniques acceptable to farmers, and simplification of seedling production. 3.16 Experiment Station Management. Under the project, the role of the experiment station manager would alter substantially, in that he/she would have two essential responsibilities: (a) organize and implement experiment station operations, so as to provide maximum and efficient support to the scientists in the execution of their experiments, including research farm management, administration, budget control, and the maintenance of research installations; the largest stations may have an additional farm manager, responsible for field operations on the station, whereas the smallest stations without resident scientists may only have a farm manager reporting to the manager of a regional station; for all matters related to station operations stricto senso the managers will report to the Administrative and Financial Director; and (b) provide logistical support to scientists for the implementation of on-farm testing programs in the eco-region. 3.17 From January 1, 1994 onwards, experiment stations will have for the first time budgets that are separate from research budgets. The former include the indirect and fixed costs to operate and  20 maintain the experiment station and its equipment. Preliminary 1993 station budgets were reviewed during appraisal in preparation for the 1994 budget exercise. The preliminary budgets have been used to estimate experiment station operating expenditures for the duration of the project. They amount to US$24,600 per scientist per year and represent 37 percent of total cost including overhead. Strengthening Internal Linkages (Research - Farrner - Extension) 3.18 The responsibility within ISAR for the improvement of these linkages will rest with the Director General, together with the Regional Directors (para. 3.9), and the Scientific Director. Through their contacts with farmers and extension staff, the scientists have an important role to play in making sure that research objectives (in particular on-farm testing objectives) are responsive to the constraints and concerns of the farmers. Scientists are responsible for the quality of on-farm testing, both in terms of experimental design (correspondence to the research objective, replicability and applicability of expected results) and in terms of implementation of the experiments. One of the paramount objectives of the project is to accelerate the availability of new technologies by speeding up on-farm testing and greatly expanding it. All scientists will in the future work both on-farm and on-station. All on-farm testing proposals will in the future include precise arrangements for the involvement of farmers and collaboration/participation of extension staff. The project will specifically include: (a) institutional arrangements to establish research - farmer - extension links (para. 5.7); (b) an incentive program for researchers to expand on-farm testing, to be included as an integral part of a scientist's scheme of service (para. 3.20); (c) training to improve the quality of on-farm testing (experimental design and data analysis); (d) funds to cover additional costs of on-farm research (renting of land, payment of farmers, etc.), in particular the expense of increasing scientist mobility to supervise field experiments; and (e) restructuring the research information system, to facilitate researchers' access to documentary information (national and external) and make research results widely available. The Agricultural Research Policy Letter reflects the specific recommendations of the Research/Extension linkage workshop, organized during project appraisal, the results of which are recorded in a working document, available in the project file. Strengthening External Linkages (Regional and International Collaborative Research) 3.19 The main difference with past approaches to collaborative research and networking will be that NARC and ISAR, rather than waiting for what comes their way, will actively seek out scientific partners to collaborate on adaptive research programs that would otherwise be beyond the scope of the national system. Under the project, funding has been included for Rwanda's participation in collaborative research. The only condition would be that the entire program is acceptable to NARC (i.e. is in accordance with the NARMP) and that the components of the program that concern the participation of other partners are fully subscribed. Depending on where the lead institution of a collaborative program is located, Rwandese scientists could be seconded to a neighboring country or  21 a regional/international institution, to carry out the component that NARC has contracted to subscribe to. Alternatively, the component may be suitable for execution in Rwanda, in which case there is no need for such a secondment. Human Resources Development and Management 3.20 Under the project, ISAR will develop by March 31, 1994 with assistance from ISNAR, a Human Resources Development and Management Plan for all of its staff (para. 5.17). The plan will be developed according to terms of reference acceptable to IDA. In terms of future staffing and skills mix requirements, the plan will be consistent with the revised NARMP. The plan will contain detailed function descriptions according to the new ISAR organization, including a description of responsibilities and accountability. Profiles will be prepared, describing education, training, experience and other specific skills required to fulfill these functions. Staff will be evaluated, followed by a skills gap analysis through the matching of profiles and evaluations. On the basis of this exercise, function descriptions and in some cases the organization itself or operational procedures will be changed to adapt these to available staff skills. The end of the exercise will be a training program for existing staff; some dismissals; clear career paths; procedures for growth promotion, including annual evaluation procedures and criteria E/ for all categories of staff; and clear recruitment criteria for new staff (para. 5.17). This exercise will also provide the basic elements for the establishment of new staff rules, including: staff contracting arrangements, 'tenuring' criteria, incentives for scientists to emphasize on-farm testing (para. 3.18), and 'bonding' criteria for staff benefitting from external training (para. 3.21). 3.21 Training. Laboratory technicians, tractor drivers, equipment operators and workshop staff will be tested to determine on-the-job training needs. Funds have been included under the project to provide professional assistance to ISAR for in-house training courses. Special in-country training courses for station managers were started during project preparation; these will be continued during the project. The project will develop this cadre as a separate professional group, with their own career development stream. Formal academic training (MSc and PhD) will be supported only insofar as they serve to improve the scientific skills of existing staff and are in accordance with ISAR's long term skills mix requirements. Staff benefiting from such training will be required to stay with the institution for at least five years after successfully completing their studies. Funds available under the project for training are not limited to ISAR staff only, but will also be available for contractual staff. Training funds will be administered by NARC. For the purposes of costing the training component, distinction has been made between: (i) long-term MSc/PhD training abroad; and (ii) short-term training. The IDA credit will provide financing for long-term MSc/PhD training abroad for up to six researchers, provided these researchers have signed up with a university prior to October 1, 1994. All other training is expected to be provided through bilateral funding arrangements. Short- term training, for which funding is also provided under the proposed Credit, will include: scientist participation in international congresses and seminars; PhD training, based on research undertaken in Rwanda, with short-term course requirements abroad; workshops, seminars and congresses organized by Rwandese institutions; in-house training in research and station management, and accounting; and on-the-job training of scientific, technical and administrative staff. 3.22 Technical Assistance. The project will support improved scientific management at three levels; the first two of these interventions will be a continuation of the work done during the 5/ Including criteria to value a scientist's client orientation, knowledge of farming conditions and responsiveness to farmer constraints.  22 preparation of the project. First, for as long as NARC will deem that such assistance is necessary, the project will continue to provide short-term advisory visits to assist in the planning of individual research proposals and the integration of these into a cost-effective national program. Second, it will continue to provide, short-term advisory visits to assist in the setting-up and operation of the Programming, Monitoring and Evaluation Unit of ISAR (paras. 5.11-5.13), including support to the biometry function (experimental design, data analysis and data storage and dissemination). Third, it will provide financing for the regular external reviews organized by NARC (para. 4.13). 3.23 Within the central administration of ISAR, the project will continue to provide technical assistance for two years, to ensure that the newly installed cost accounting system is used in all sections, that it is properly understood, and that an operations manual is prepared. Accounting and management audits by independent auditors will be organized by NARC for all participating institutions, including ISAR. The Human Resources Development and Management Unit of ISAR will receive assistance from ISNAR in the form of regular visits to prepare and implement the human resources development plan and the staff rules (paras. 3.20 and 5.17). The project will continue to support technical assistance to the central analytical laboratory, until such time as a qualified Rwandese can take over the direction of the laboratory. Finally, ISAR will seek professional advice for the rehabilitation of its stations. Rehabilitation and Renewal of Physical Plant and Equipment 3.24 Capital investments to be funded under the project will include: the rehabilitation of experiment stations (buildings and grounds) to be retained in the system; and laboratory, agricultural and computing equipment; vehicles; and publications, library and communication equipment. Stations included for rehabilitation/delayed maintenance are shown in Table 3.1. Proposals for new buildings at ISAR's Head Office in Rubona include: an in-vitro laboratory for which donor funding has already been identified (European Economic Community) and a conference center for which donor commitment is being sought. IDA financing would be directed towards support to research activities through the construction of a cold storage unit for the conservation of germplasm and an extension to the existing documentation center. Other investments include improvement to the existing research information system, which would involve bringing under a single improved management system (System for Documentation Information and the Publication of Research Results) 6/ the operating structures of the existing Central Unit for Information and Dissemination at Rubona, the three documentation centers at Rwengeri, Ruhande and Karama and other libraries of ISAR in the field. Investments would include the purchase of scientific publications, software, improved communications and computer network systems and better coordination with existing documentation centers at UNR, IRST and MINAGRI. Better known by its French acronym SIVRA or Systeme d'Information Documentaire et de Valorisation des Publications de la Recherche Agricole.  23 Table 3.1: ISAR Network of Research Stations Main Experiment Station Multi-Locational Agro-Ecological Region Principal Mandate ** (Regional Center) Site 1. Rubona Ruhande (*) Plateau Central Beans, coffee, Songa (*) horticulture, sweet potatoes, cattle 2. Gakuta Rukoko Hautes Terres tea, sheep 3. Ntendezi Bugarama Terres du Bugarama/ rice, coffee, bananas, Mwito Bordures du Lac Kivu soya 4. Ruhengeri/Kinigi Tamira Terres de Laves potatoes, maize, climbing beans, bananas 5. Rwerere Byumba Hautes Terres du horticulture, potatoes, Buberuka climbing beans, peas 6. Nyagatare/Cyabayaga N/A Terres du Mutara/ rice, sorghum, cassava, Kibungo Est bush beans 7. Kibungo Gahoro Plateau de l'Est bananas, sorghum, Mubago groundnuts 8. Karama Rutongo Bugesera-Mayaga sorghum, bush beans, Icynya cassava, sugar cane, goats Specialized sites: Ruhande (Forestry); Songa (cattle) $* Soil conservation, bananas and agro-forestry are programs that are executed at almost all stations. 3.25 Most of the equipment to be bought under the project will be for renewal or modernization. In the case of agricultural, library, communication and computing equipment, additional equipment is included to improve the quality (e.g. soil preparation equipment) or efficiency (e.g. CD-ROM equipment) of the operations. Institutions other than ISAR can present proposals for equipment purchases in the context of a research contract (for procurement rules in such cases see para. 4.13). IV. PROJECT COSTS AND FINANCING A. Cost Estimates 4.1 The overall agricultural research activities have been considered as a capital good and consequently project costs have been calculated to include both ongoing and new research activities, irrespective of the financing source. Costs also include preparation and start-up activities financed under the Project Preparation Facility (PPF). Only grant financed technical assistance has been  24 excluded (about 7 percent of total costs), as these funds are generally not fungible or the assistance is provided in kind. This procedure has been adopted to emphasize the integration of the IDA crodit into a coherent national plan as expressed in the NARMP (paras. 2.16 and 3.2). 4.2 Project costs have been estimated at FRw 5,104 million or US$36.5 million equivalent, of which FRw 1,786 million or US$12.8 million equivalent (35 percent) will be in foreign exchange. Base costs reflect prices and exchange rates in effect at appraisal in October 1992. A summary of project costs is given in Table 4.1 and its percentage breakdown by component is also graphically shown (Figure 4.1). A detailed breakdown by year and component is given in Annex 5. Physical contingencies varying from 5 to 15 percent have been applied to selected categories to reflect uncertainties regarding detailed quantities and possible design modifications. Price contingencies have been calculated for local inflation according to the following annual forecasts: 1993, 5 percent; 1994 through 1997, 4 percent. Inflation on foreign goods and services has been calculated at the fbilowing rates: 1993, 3.8 percent; 1994, 1.9 percent; 1995, 2.7 percent; 1996, 3.4 percent; 1997, 3.6 percent. Taxes and duties have been calculated on all costs at prevailing rates. Table 4.1: Project Cost Summary 7/ _______ ________ S~~~~~~ ~~~~~~ Total, Foreign gms Local Foreign Total LocaL Foreign Total | ch1 a e A. stations ..J.J..11198 . 470 1 1.668 . 8.556 . 3.354 j 11.910 1 28 ...37 1 B. Research Progrm 1.117 554 1.671 7.977 3.f59 11.936 33 37 C. Central Services 249 I 138 ! .387 !.1.778- 989 2.767 36 9 D. NARC Pcrmanent 42 11 53 301 77 378 20 _ .1 1lE. TroSinfn 226 _135 361 1.611 968 2.579 38 a F. Consultant Servic k 20 354 374 140 2.531 2.671 95 8 Total Baseline Costs 2,851 1,663 4,514 20,364 11,877 32,241 37 100 Physical Contingencies 148 38 186 1,054 272 1,326 21 4 Price Contingencies 310 96 406 2,211 686 2.897 24 9 Total Projects Costs 3,308 1,797 5,105 23,629 12,835 36,464 35 113 1/ Totals may not add up due to rounding.  25 Figure 4.1: Project Cost Summary (Pie Chart) NARC TrainingConsultant Centra Servces NRC TriningServices Cenra Sevc 1% 8% 8% ~~~~~~~~~~~~~. . ............................................ ... . ~~~~~~~4 _ ' " ; ' t " ~ ,._,. Research Station Programs Management 37% 37% B. Project Financing 4.3 The proposed IDA Credit of US$15.0 million would finance 47 percent of project costs net of taxes and duties. Financing will cover: civil works (new construction, rehabilitation of existing facilities and construction and rehabilitation of terraces and access roads), vehicles and equipment, training, technical assistance and consultant services for ISAR and NARC, and operating costs (both existing and incremental). Incremental and non-incremental operating expenditures have been considered a capital good and therefore eligible for IDA financing. The IDA Credit will be passed on to NARC and ISAR on a grant basis. Agreement to this effect was reached at negotiations (para. 7. 1 (e)). 4.4 Government will finance 13 percent of net project costs (23 percent when including taxes and duties). Starting with the adoption of the new statutes for ISAR (para 3.10), Government funding will be directed towards financing a share of research costs and will cease to be targeted to salaries. Budget allocations for all agricultural research in Rwanda will be channeled (or at least programmed) through NARC (paras. 3.4 and 5.1 (d)). However, because of the financial implications for ISAR, Govermment funding for ISAR's research programs will continue to be channeled directly to that institution until December 31, 1995. 4.5 Expected donor financing, based on present commitments is shown in Table 4.2. A breakdown of donor financing by research program is given in Annex 7.  26 Table 4.2: Project Financing Plan (US$ '000) Total GOR % ISAR % Private % Donors % Donors % IDA % committed uncommined Invenrnent Cogs Civil Works, Vehicles 10,052 354 4 1,019 243 2 8,436 84 and Equipment, and 10 Consultant Services Training 2,425 - 0 - - 1,925 79 500 21 Overatin2 Costa Research Progrms and 17,593 2,550 14 1,400 8 200 1 5,413 31 2.895 16 5,130 29 Stations Central Services 1,767 1,080 61 - - - - - - - - 687 39 NARC 247 - - - - - - - - - - 247 100 Sub-totl 32,085 4,060 13 1,400 4 200 1 6,437 20 5,063 16 15,00 47 0 Taxes 4,379 4,379 100 - ToUl 36,464 8,439 23 1 ,400 4 200 1 6,437 18 5,063 14 15,00 41 0 C. Government Budget Impact 4.6 Government's budgetary allocation to ISAR for 1992 was FRw 105 million (US$750,000) or 28 percent of ISAR's budget. This represents a 40 percent increase over 1991 (47 percent over 1990), equivalent to 10 percent of the budgetary outlays for agriculture under the 1992 Public Expenditure Program (PEP), which is considered reasonable. During the 5-year project period, Government's contribution, net of taxes and duties, will represent FRw 568.7 million (US$4.06 million) with each annual contribution required to increase by 4 percent (the average estimated local inflation rate). Government's average annual budgetary contribution will remain in real terms at the 1993-95 PEP percentage level. Demands of this magnitude on fiscal resources, are in line with the country's current macro-economic framework. D. Procurement 4.7 Procurement procedures are summarized in Table 4.3 below. Figures are, however indicative, as the objects of IDA financing will be set each year in the light of the annual work program and other available funding. ISAR's staff has benefitted under the predecessor project from training in procurement by Technical Assistants and Bank disbursement staff. To simplify procurement, the Government has provided the Bank with standard procurement documents (standard bidding documents and letter of invitation for consultant services) for civil works, goods and services (including research contracts), prepared on the basis of the Bank's sample documents, that will be used during project execution.  27 4.8 Civil works contracts to be financed by IDA (US$3.5 million) will be procured under LCB. This procurement procedure is warranted by the nature of the work over 22 geographically widely dispersed sites, which makes packaging of the bids in large lots impossible. Table 4.3: Summary of Proposed Procurement Arrangements (US$ '000 equivalent) Project Element ICB LCB Other N.B.F. Total 1. CiviL Works 3,545 760 4,305 (3,191) (3,191) 2. Vehicles, Equipment 2,983 440 2,109 5,532 (2,230) (344) (2,574) 3. Consultants & T.A. / 2,671 2,671 (2,671) (2,671) 4. Training Long-term 174 1,067 1,241 (150) (150) Short-term 406 1,179 1,585 (350) (350) 5. Operating Costs: ISAR 3,734 8,355 12,089 (3,490) (3,490) NARC 275 275 (247) (247) 6. Research Contracts (NARC) 2,490 6,276 8,766 (2,327) (2,327) Total 2,983 3,545 10,190 19,745 36,464 (2,230) (3,191) (9,579) 0 (15,000) Notes : Figures in parenthesis are the respective amounts financed by the IDA credit. N.B.F.: Non-Bank Financed. 4.9 Vehicles, Equipment and Spare Parts will be packaged in lots of US$200,00 or more and procured using ICB (US$3.0 million). In the case of specialized scientific and agricultural equipment, where only a limited number of suppliers that can meet product specifications exist, procurement would follow international shopping procedures (US$340,000); prior agreement from IDA will be needed on the justification for international shopping and on the list of potential suppliers. The purchase of minor spare parts and replacement of hand tools, estimated to cost less Services will be procured in accordance with World Bank guidelines: Use of Consultants by World Bank Borrowers and by the World Bank as Executing Agency (Washington, D.C. 1981).  28 than US$20,000 per package, with an aggregate limit of US$100,000 will be procured using local and international shopping procedures (minimum three quotations), except in the case of sole suppliers, where direct procurement will be permitted. 4.10 Technical Assistance and Consultants contracts will be procured on a competitive basis in conformity with Bank guidelines. Terms of reference, evaluation criteria, conditions of contract for these services will be subject to prior approval by the Bank. 4.11 Long-Term Training abroad for MSc/PhD candidates, for an estimated amount of US$174,000, will be procured following Bank guidelines. 4.12 Short-Term Training on specific pre-identified subjects, such as scientific programming, station management or valorization of research results, for an estimated amount of US$406,000 will be conducted by specialized training agencies recruited following Bank guidelines. 4.13 Research Contracts would be progressively introduced after December 31, 1995, for all IDA financed research activities. Pre-qualified research institutions represented within and outside Rwanda would be eligible (para. 3.6). It is estimated that the number of such contracts will be around 40, averaging about US$220,000 each. The pre-qualification process will be carried out by the Scientific Committee (SC) of the NARC and results submitted to the Bank for prior approval. Selection procedures will follow Bank guidelines for consulting services. All contracts of a duration exceeding 3 years will be subject to a mid-term review and all research activities will be subject to independent ex-post reviews mandated by NARC. All research contracts with pre-qualified research institutions would be on standard terms (para. 4.7), only eligible research contracts to these institutions above a threshold of US$100,000 would be subject to the Bank's prior review procedures. Agreements were obtained at negotiations on the content of the pre-aualification mechanism and on the elements of the review process. In addition, assurances were obtained that research proposals under contract funding will be handled by NARC and not be subiect to Central Tender Board scrutiny (7.1 (f)). 4.14 Bank prior review is expected to cover about 72 percent of the value of works and goods contracts, all contracts for consultant services and at least 80 percent of the research contracts. The Project Implementation and Procurement Schedule is outlined in Chart No. 3. Procurement information will be collected and recorded as follows: (a) prompt reporting of contract award by borrower; and (b) comprehensive 6-monthly reports by NARC and ISAR to IDA indicating: (i) revised timing of procurement actions, including advertising, bidding, contract award, and completion time for individual contracts; and (ii) compliance with aggregate limits on specified methods of procurement. Assurances on these points were obtained at negotiations (para. 7.1 (g)). Procurement relating to co-donors' financed components will be carried out according to their own procedures. E. Disbursements 4.15 A schedule of disbursements is in Annex 6. The actual distribution of the proceeds of the IDA Credit will be done on the basis of annual work programs. Hence, the allocation by disbursement categories (Table 4.4) is only indicative. All categories will finance 100 percent of eligible expenditures, except for civil works, and vehicles and equipment, where 100 percent of eligible foreign expenditures and 90 percent of eligible local expenditures will be financed. All disbursements will be net of directly identifiable taxes and duties.  29 Table 4.4: Disbursement Categories Category Amount of the X of Expenditures to be Credit Atlocated Financed _______________________________ (USS '000) (1) Civil Works 2,600 100X of foreign expenditures ___ _ _ _ 90X of locaL expenditures (2) Vehicles & Equipment 100X of foreign expenditures (a) ISAR 1,900 90X of local expenditures (b) NARC 70 (3) Consultants Services 100X of expenditures (a) ISAR 1,000 (b) NARC 1.000 (4) Training 100X of expenditures (a) Long-Term 150 (b) Short-Term 350 (5) Research Contracts 2,000 100X of expenditures (6) Operating Costs 100X of expenditures (a) ISAR 3,000 (b) NARC 200 (7) Refinancing of the PPF 1.500 _ Amount due (8) Unaltocated 1,230 TOTAL 15,000 4.16 To reduce the number of disbursement claims and the need for working capital, two Special Accounts in Foreign Exchange will be established (one each for ISAR and NARC). Upon effectiveness, IDA will advance US$500,000 into the ISAR Special Account and US$150,000 into the NARC Special Account, based on three months average estimated expenditures on eligible items. All disbursements will be fully documented and expenditures on single items under US$20,000 and all operating costs will be made on the basis of Statement of Expenditures. Documentation will be retained by ISAR and NARC and will be made available for review by IDA supervision missions and independent auditors. With the introduction of research contracts (involving operating costs only) disbursements will be in tranches according to contract specification. The implementation of agreed standard contract specifications for research contracts by NARC (para. 4.7) and agreement on pre- qualification procedures acceptable to IDA for contracting institutions (para. 3.6). are Conditions of Disbursement for research contracts (para. 7.3 (a)). The application of agreed research proposal presentation guidelines (para. 5.10) will be a Condition of Disbursement for research operating expenditures and contracts (para. 7.3 (b)). No disbursements under the Development Credit Agreement will be made for civil works and land improvement at any given location, until development plans have been prepared for the experiment stations and an action program for the implementation of civil works. acceptable to IDA. has been Drepared (para. 7.3 (,c). and ISAR has obtained title to the land at that location, and excess lands and buildings have been transferred to MINAGRI. leased or sold (para. 7.3 (d)). F. Accounting, Finance, Administration and Audit 4.17 Separate accounts will be kept of all project expenditures. ISAR will keep accounts relating to the components under its management and NARC for its own expenditures and those relating to  30 research contracts. ISAR has demonstrated during the earlier project (para. 2.13 (b)) and at appraisal, its ability to run a performing accounting and budgetary control system. The existing system will form the basis for the introduction of a decentralized budget under the project. Technical assistance in this area, which started under the first project, will continue for two more years, or until local staff are properly familiarized with the newly introduced (1991) computer software. Equally, all contractors to the research programs will have to demonstrate the existence of an accounting system satisfactory to IDA (para. 3.6 (d)). Accounts will be kept according to internationally accepted accounting standards to cover as a minimum, all expenditures, reimbursements received, commitments and IDA funds position. The set up of an accounting system at NARC. satisfactory to IDA. is a Condition of Effectiveness (para. 7.2 (a)(iv)). 4.18 To ensure timely and sufficient Government counterpart funding, ISAR and NARC will each establish, not later than August 31 of each year, their budget funding requirements on a quarterly basis. Government will not later than December 31 of the same year confirm its budgetary commitment and will deposit, not later than January 15 of each year, the first quarterly advance. Prior to the start of each subsequent quarter, the agreed advance will be deposited. Agreement to this effect was obtained at negotiations (para. 7.1 (h)). The quarterly advance for NARC will be deposited in a project account at a commercial bank to be opened by Government on terms and conditions acceptable by IDA. Agreement to this effect was obtained at negotiations (para. 7.1 (i)). The deposit of an amount of FRw 26 million into the NARC proiect account representing a Quarterly fundinLi estimate for the first Droiect year is a Condition of Effectiveness (para, 7.2 (c)). 4.19 The NARC and ISAR accounts will be audited by independent auditors acceptable to IDA. The auditors' reports will be submitted to IDA within six months of the end of the NARC and ISAR fiscal years; and these reports will be of such scope and in such detail as IDA may reasonably request, including a statement as to whether or not IDA funds have been used for their intended purpose, and a certification that all expenditures claimed under the Statements of Expenditure (SOEs) during the year have been properly incurred for the Project as designed. Agreements on accounting and auditing arrangements were obtained at negotiations (para. 7.1 (j)). 4.20 Government will prepare a project completion report within six months of the Credit closing date. Assurances to this effect were obtained at negotiations (para. 7.1 (k)). V. ORGANIZATION AND MANAGEMENT A. Institutional Reform NARC 5.1 The objective of the institutional reform of the Rwandese agricultural research system is to rationalize public and ODA support by unifvJng crop, livestock, forestry and natural resources management research under the umbrella of NARC (para. 3.4). NARC's draft statutes contain the following critical elements: (a) A clear statement as to its responsibility for formulating and implementing the national policy expressed through the revised NARMP (para. 3.5 (b)) with respect to agricultural and natural resources management research.  31 (b) Authority to implement the national policy, through letting research contracts 2/ with national, regional and international institutes, NGOs or the private sector; and/or enter into partnership arrangements with such institutions. (c) By-laws providing a clear definition of the terms of reference and responsibilities of the ES. (d) Control over all Government (including salaries) and ODA funds will be channeled (or at least programmed 10/) through NARC (para. 3.4). Annual decisions on the application of these funds will be vested in the Finance Committee, which will be chaired by the Cabinet Director of the Ministry of Finance, and which will include duly authorized representatives of all sources of funds, Finance Committee decisions would be binding. (e) Review of individual research proposals or entire programs; responsibility for this function will be vested in an independent Research Committee, which will include scientists of international repute; to ensure transparency and accountability, committee recommendations will be available to the Finance Committee prior to its deliberations and to the scientists concerned. (f) Composition of NARC and its statutory organs the Research Committee and the Finance Committee, and eligibility criteria and selection procedures for the nomination of its members (ex-officio and non ex-officio). 5.2 To exercise its function as an independent research policy-maker and representative of both the users of research results and Government/donor interests, will demand both intellectual leadership and private entrepreneurship. NARC's membership will reach beyond the sphere of public administration (ex-officio) into a broad range of interested parties from agro-industries, farmers associations, NGOs, academic and technical institutes (in non ex-officio capacity). A membership 5:4 ratio in favor of the non ex-officio members will lend the desired weight (with the necessary derogation obtained). The Council will be assisted by a Research Committee and a Finance Committee. The ES will be responsible for day-to-day implementation of policy decisions. The ES will have the following tasks: (a) to organize the preparation, validation and regular updating of the NARMP, as a yardstick for the decision-making process on priority research objectives and research funding; (b) to oversee, implement and administer research contracts funded through NARC, including the organization of independent scientific reviews and financial audits as provided for under the terms of individual contracts and mandated by the Research and Finance Committees, respectively; 2/ Including contracts for collaborative programs between institutions, or as a partner with other sources of funds. 10/ To allow for variations in procurement and disbursement procedures between donors.  32 (c) to prepare guidelines for research proposal presentation, progress reporting and financial accountability; (d) to establish a data base to monitor research programming, program implementation, research results and program funding, basic elements to be presented annually to the Research Committee and Financial Committee; (e) to act as a secretariat to the meetings of the Research Committee and the Financial Committee and disseminate their reports; and (f) to ensure the appropriate and timely mobilization of funding from national and external sources. 5.3 This process replaces a complicated and cumbersome system: (a) The two NARC committees replace a multitude of individual donor interventions/systems/reviews, separate accounting systems, etc.; all donors are members of the Finance Committee and formally represented in NARC; and (b) Including 'outsiders' (extension services, farmer organizations, the private sector, etc.) in the internal priority setting and decision making processes of ISAR will ensure a much closer 'bondage' between scientists and their clients, and thus ensure research responsiveness to farmer constraints, one of the principal objectives of the project. It is therefore of the utmost importance that Government seeks adherence of all donors to the proposed decision-making processes. Agreements are established with all donors, in particular, on joint, independent, financial and scientific audits, which will replace the multitude of time consuming reviews organized by individual donors. To achieve this, donors were invited to send formal observers to participate in Credit negotiations, and Government will organize a donor Round Table to discuss options prior to the adoption of the statutes of NARC and ISAR. Proposed draft statutes for the NARC are in Annex 2 and its Organization Chart is in Chart 1. NARC's and ISAR's statutes have been transmitted to the Council of Ministers. Enactment of the statutes and by-laws of NARC. acceptable to IDA, is a Condition of Effectiveness (para. 7.2 (a) (ii)). Detailed arrangements for the channeling of Government funds for agricultural research, including those destined for IRAZ, will be the subject of an Implementation Agreement between Government and NARC. Assurances to this effect were obtained at negotiations (para. 7.1 (1). Thne execution of this agreement will be a Condition of Effectiveness (para 7.2 (a) (i!). 5.4 ISAR: Research Management and Resource Allocation. Improvements in the research management, programming and review system of ISAR, tested under the first project, will need to be reinforced. Essential elements of this system will be formalized under the second project, i.e. be included in ISAR's by-laws under its new statutes. This will provide the necessary guarantees for the participation of the users of research results (farmers, NGOs, extension services and the private sector) in the setting of the research agenda. Also, the system will provide for timely interaction with NARC and its statutory committees. 5.5 The three principal elements of the system are:  33 (a) Research Program Task Forces (RPTFs); each of the previous 4 research departments will constitute a RPTF, including all scientists and selected representatives from farmer organizations, NGOs, agro-industries, the private sector and extension services; RPTFs review individual research program objectives and advise the Scientific Director and RPCs for decisions on priorities and resource allocations at the program level; all on-farm testing program proposals are reviewed by the RPTFs; (b) Annual Research Review; ISAR's existing Scientific Committee consisting of the Scientific Director and the 8 Regional Directors, will be expanded to include the Administrative and Financial Director and decision-making level representatives (technical staff, not administrators) of the extension services, the agro-industries, and selected collaborating scientific partner institutions; the Scientific Committee meets annually to review progress on the basis of ISAR's Annual Reports, to decide on priorities between programs, review ISAR's consolidated program and budget proposals, and on future directions; previous NARC Research Committee (para. 5.2 (e)) reports and findings concerning ISAR programs must be available prior to Scientific Committee meetings; and (c) ISAR's Management Committee consisting of the Director General and the Scientific, and Administrative and Financial Directors, has joint responsibility for consolidating ISAR's research programs and budgets, obtaining Scientific and Finance Committee approval for all research funded through NARC, and presenting it with its own recommendations to ISAR's Board of Directors for final approval. 5.6 Many agricultural research activities are seasonal. Most field trials have to start in October, at the onset of the most important one of the two growing seasons. This means that preparation (preparing research-protocols and field books, ordering materials, etc.) has to start at least six months earlier and that decisions on available resources have to be made before December 31 of each year. Therefore, the by-laws of ISAR would provide for the following annual programming, budgeting and review cycle: (a) by July 31: publication of the annual report of last year's research results, thus allowing for last minute changes in research protocols before the start of the agricultural season; (b) by September 30: preliminary conclusions of RPTF meetings with respect to next year's research programs and budgets to be presented to NARC's Research Committee; (c) by October 31: RPTF conclusions with NARC's Research Committee comments to be discussed by the Scientific Committee; (d) by November 30: ISAR's Management Committee (para. 5.5 (c)) presents and discusses Scientific Committee conclusions with NARC's Finance Committee; and (e) by December 31: ISAR's Director General presents research program proposals, detailed budgets and financing plans with NARC's Research and Finance Committees' recommendations to its Board of Directors for approval.  34 B. Internal Linkages: Research - Farmer - Extension Interaction 5.7 The success of a small research system depends on the strength of its linkages: internal with the users of technology and external with the sources of technology. The most direct internal linkages between research and farmers are through on-farm experimentation and testing programs, conceived in partnership with the extension services, NGOs, farmer organizations, agro-industries, etc. These linkages try to give the farmer a more direct voice in setting the research agenda and immediate program goals. In addition to formalizing such internal linkages with research clients and making them inescapable, the project will improve certain of the processes previously adopted, and add others to provide liaison at five levels: (a) ISAR will organize a limited number (two to four) of field days for farmers, scientists and extension staff. These field days will focus on specific themes, identified in advance, and each field day will be held at the most appropriate site (either at an experiment station or at a site of on-farm trials). The objective would be a two-way dialogue. (b) Each Regional Director will attend monthly meetings with extension staff and representatives of farmer associations, NGOs, agro-industries and projects in each Prefecture served by the station (para. 3.9). Depending on the subjects to be discussed, the Regional Director may designate a scientist to attend in his place. These meetings will be to discuss themes for extension, and for the ISAR representative to receive feed-back from extension staff. The Regional Director will also use these meetings to arrange for contacts to plan on-farm experiments. (c) Scientists will inform extension staff and representatives of projects and NGOs of their plans for on-farm experimentation; they will discuss the design of these experiments, and arrange for the full cooperation of the field staff of these other organizations, including their assistance in installing and operating the trials. Senior field staff will be invited to participate in inspection visits and in the evaluation of the trials. (d) ISAR will invite extension staff to participate in the RPTF and Scientific Committee meetings (para. 5.5 (a) and (b)) which will discuss the results of the research of the previous year, and plans for the coming year. (e) NARC will ensure that a senior representative of the extension services who has special experience in the subject(s) being studied will participate in the external technical reviews of research (para. 5.1 (e)). (f) ISAR will negotiate with an agency specialized in the management and dissemination of agricultural information, in partnership with a local NGO ('INADES-Formation' .11), a turn-key contract for the dissemination of research results. I1/ INADES-Formation has a wide ranging experience in the dissemination of agricultural information to extension agents, farmers and in the publication of documents in Rwanda.  35 These points were discussed in detail at a two-day workshop of research and extension staff, organized during the appraisal of the project. The results of this workshop are recorded in a working document, available in the project file. C. External Linkages: Regional/International Collaboration 5.8 The considerable degree of dependence of the Rwandese agricultural research system on technology generated elsewhere (para. 3.2), as a reality and as an explicit policy, make it vulnerable to weaknesses in its links with outside sources of technology and information. The project therefore proposes a much more pro-active approach to Rwanda's participation in regional and international collaborative research programs (para. 3.19). Improved linkages with research databases and other outside sources of information (documentation, publications, CDROM, etc.) supported under the project will complement such collaborative efforts. Well-directed overseas training (para. 3.21) and the participation of Rwandese scientists in relevant international and regional scientific conferences and workshops will provide opportunities to strengthen and increase external contacts. In the context of regional collaborative research, the project strongly promotes the exchange of scientists between national agricultural research systems in the region. As a first step to promote increased collaboration between national agricultural research systems in the region, maximizing economies of scale and minimizing duplication of effort (para. 3.7), Government will channel its funds for IRAZ through NARC. Through this, Rwanda's participation in the IRAZ programs will become subject to the same scientific and financial scrutiny (audits) as its national research effort (para. 5.1 (d) and (e)). D. Research Proposal Presentation and Justification 5.9 The Master Plan identifies research priorities in broad terms only. Each general subject must be converted into its component themes; each theme will have different potentials for success, so that each must be separately justified; and each proposal must take account of all the relevant research including the most recent results and those obtained by other research organizations (both within Rwanda and elsewhere). One of the first orders of business of the ES will be to prepare detailed guidelines for research proposal presentation and to provide regular training in its use. 5.10 In principle a research proposal should always include the following elements: - a constraints analysis; - problem definition; - a 'state-of-the-art' analysis (including a review of on-going research in neighboring countries); - definition of research objectives; - definition of individual research themes, corresponding to the objectives; - selection and justification of research methodology; - proposed experimental design and statistical analysis methodology; - research protocols and 'field books'; - key indicators (bench marks) for the monitoring of progress and evaluating impact; and - a detailed budget Preparation of research proposals is costly, that is why NARC will provide funding for proposal preparation (para. 3.7). This is justified by the fact that a well done 'state-of-the-art' analysis will  36 lead to applicable research results (publications), and costly duplication of effort is avoided. By March 31. 1994. NARC will adopt proposal presentation guidelines, acceptable to the Association. which have been reviewed by its Research and Finance Committees. During negotiations. assurances were obtained that these guidelines will include mandatory provisions for economic justification (para. 7.1 (). They should also provide for impact evaluations. In all cases, these guidelines should call for integration of the proposed research in a regional context (coherence and complementarity). E. Research Program Planning, Monitoring, Evaluation and Review (PMER) Procedures 5.11 The Rwanda agricultural research system recognizes three levels of PMER: (a) strategic planning under the responsibility of NARC through the preparation, regular review and updating of the NARMP; (b) regular and independent technical reviews of individual research programs (mostly to advise on the quality and relevance of research proposals, mid-term or every two to three years for long duration programs, and at program completion) and research contracts (as per the progress reporting and review requirements spelled out in the terms of the contract) to be organized by ES (para. 5.2 (b)); and (c) the annual programming, progress reporting, review and budgeting cycle described in paras. 5.4 through 5.6. 5.12 Individual institutions such as ISAR will have their own research programming and program (or contract) implementation monitoring system, as a day-to-day management information system. The ES will be responsible for setting up a system for the entire Rwanda agricultural research system, based on information furnished by ISAR and reporting and review requirements included in individual research contracts funded under NARC. The ES will commission the design of the system and subsequently present detailed proposals to the Research Review and Finance Committees prior to its adoption by NARC. Prior agreement on the terms of reference for the design of the national agricultural research planning, programming and budgeting system, also including the establishment of research databases and linkages with outside sources of scientific information, for implementation by July 1, 1994. Agreements to this effect were obtained at negotiations (para. 7.1 (n)). 5.13 At the level of ISAR, the conceptualization of research programs, its detailed planning, the preparation of budgets (for all research, both on-station and off-station), and the control of expenditure will be the responsibility of the Scientific Director and the researchers. The researchers will be responsible at the level of individual sub-programs. The Scientific Director will be responsible for the overall planning and coordination of research, and for the presentation of research results to all interested parties, ranging from the reports prepared for NARC to releases prepared for the extension services. His decisions on the content of the research program, and his interpretation of results will be subject to review by NARC. In all of this work, the Scientific Director will be assisted by the Scientific Committee (para. 5.5 (b)) and by a small Data Analysis, Monitoring, Evaluation Unit (para. 3.9 (b)), which will record (if needed analyze), summarize and evaluate all experimental results. The unit will also establish the scientific data bases for ISAR and be linked with similar bases elsewhere. In this respect, the unit will collaborate closely with the Documentation and Publication Unit.  37 F. Research Support Services and Station Management 5.14 The fixed costs of operating and maintaining research facilities have in the past been under- budgeted. Therefore, the program/budgeting system will need to be further developed into a management tool for the efficient operation of experiment stations and laboratories. This is of particular importance when a few years hence the system will be managed to a large extent through contractual arrangements (para. 3.4). This implies that a reasonable share of the indirect and fixed costs of station operation and maintenance, and overheads, must be funded through such contracts. For this to happen, ISAR will have to be able to estimate such costs, and NARC will have to audit them to ascertain that they are reasonable. This means that the budgeting and accounting system will need to separate these costs from the direct research operating expenditures. Before March 31, 1994, ISAR will seek consultant assistance to train its staff in efficient station operation procedures and to further develop its cost-accounting and budgeting system to operate and maintain its research facilities. Assurances on these points were obtained at negotiations (para. 7.1 (o)). 5.15 ISAR's management systems will be improved and responsibility for station operations will be delegated to professional station and farm managers (para. 3.16). At the level of senior management, the Director General of ISAR will be supported by an Administrative and Financial Director. The Administrative and Financial Director will be responsible for all central support services. 12/ The newly installed cost-accounting system will be used to record and control all expenditures, for research activities and support services. The Administrative and Financial Director will be responsible for the integration of budgets and the control of expenditure for all the individual stations. ADpointment of an Administrative and Financial Director with qualifications and experience satisfactoly to IDA is a Condition of Credit Effectiveness (para. 7.2 (b) (iffl. 5.16 Each experiment station will have a station manager who will report to the Administrative and Financial Director, and who will be responsible for the administration and all routine operations of that station; for these purposes, he will be responsible for his own budget; expenditures incurred in the conduct of field trials will be the responsibility of the scientists. The station manager will participate in the planning of ISAR's annual work programs, and will be responsible for ensuring that the needs of his station (maintenance, crop rotation, etc.) are addressed. He will also be responsible for the provision of the facilities required and requested in advance by the researchers Oand, labor, etc.), but he will not be responsible for the scientific content of their trials. G. Human Resources Management, Training, Recruitment, Staff Evaluation and Terms of Employment 5.17 One of the main problems for scientists in Rwanda has been their civil servant status. They are managed by the Ministry of Civil Service and can be transferred at short notice. This has prevented ISAR from developing its own personnel policy with respect to staff evaluation, career development, growth promotion and an incentive-based scheme of service. The result has been that some staff have been appointed scientist, without having the necessary training or even aspiration. As long as ISAR remains a para-public institution and its scientists civil servants, little will change. Success in research depends on the motivation and perseverance of scientists. This is one of the important reasons for transforming ISAR into an institution with its own human resources 12/ Station operations, administration (budgeting, budget control, general accounting and the payroll), purchasing and stores, and financial control.  38 development and management policies, and revising its statutes accordingly. On the basis of the Human Resources Development and Management Plan (para. 3.20), to be prepared with assistance from ISNAR, ISAR will prepare by March 31, 1994 for comments by the Finance Committee and NARC: (i) proposals for criteria for the recruitment, evaluation/promotion, tenuring and termination of staff; and (ii) draft employment conditions including salary scales and employment under fixed term contract. ISAR will seek IDA approval for the proposed scheme of service prior to its adoption by ISAR's Board of Directors. Agreement on these points were obtained at negotiations (para. 7.1 5.18 ISAR's Personnel Section will be separated from the pay-roll (an accounting function) and will become responsible for the development and implementation of the Human Resources Development and Management Plan, designing a new scheme of service, and organizing and monitoring staff training. H. Consolidated Funding Mechanism 5.19 All ODA and Government funds for agricultural research will be administered by and/or coordinated through NARC (para. 3.4), either until December 31, 1995 in direct support for ISAR on the basis of annually agreed research programs and budgets, or in the form of research contracts. In the short term, NARC would operate more as a 'clearing house' for research contracts managed according to commonly agreed procedures (para. 5.2 (b)), rather than a trust fund. This should change in the future as donors gain confidence in the system, i.e. would be willing to commingle their funds and trust the accounting systems of participating institutions to trace the usage of funds by source of finance. 5.20 Research by NGOs, farmer associations, seed and other agro-industries, IARCs and other foreign institutions, that are eligible for NARC support according to the terms of the NARMP (paras. 5.1 (a) and 5.2 (a)), could be funded directly by NARC under contract, provided: (a) the research proposals presented according to NARC guidelines (para. 5.10) are accepted by NARC's Research Committee; and (b) the implementing institutions meet the accountability criteria set by NARC's Finance Committee. Alternatively, these organizations could present proposals jointly with ISAR. 5.21 ISAR will from the onset also be eligible for funds from NARC for individual research contracts. Ideally, by January 1, 1996 (paras. 3.4 and 3.8), ISAR will be funded entirely through research contracts, not all of which are necessarily funded through NARC. The future research system should allow for maximum flexibility (i.e. Dg monopolies) and introduce an element of competition for research contracts. The review and decision making process is pictured in Chart No 4 in annex. I. Investment Program (Infrastructure and Equipment) 5.22 Before any investments are made, ISAR will seek professional advice to prepare development plans for experiment station rehabilitation (para. 3.23) and an action program for its implementation. Such plans will indicate the future use of buildings and grounds, including: proposed building occupation and working relationships, internal communications, perimeter fencing, experimental field layout (based on soils and topographical data, crop rotations, regeneration requirements, etc.), and staff housing. Before the action program is implemented, ISAR will obtain title to all land on which its experiment stations are located and lease or sell excess land and buildings or transfer these to  39 MINAGRI. The execution of the development plan at fairly scattered work sites will require the hiring of two or three separate architects responsible for drawing up the bidding documents, preparation, analysis and evaluation and surveillance of works. The introduction of mechanical equipment will also require that the role (cost/benefit) of the central workshop at Rubona be examined by an independent consultant. Until now, ISAR has acquired (or been granted) a multiplicity of personal computers on an ad hoc basis. The increasing use of PCs under the project, both at the central units and at experimental stations will require that a proper implementation strategy be defined both in terms of hardware and software and for which expertise will be sought. Strengthening of the existing cost/budget accounting software through networking at Head Office and software updating will be implemented under a separate plan. J. Project Implementation and Supervision 5.23 The project Procurement and Implementation Schedule in Chart No. 3 provides for the phasing of civil works over a period of four years owing to the need to proceed to an initial lay-out study in the first year, and to the limited capacity of the local industry. Research contracts extending beyond the expected credit closing date of March 31, 1999, will need to be tailored to the availability of funds. Consultancies for training on research programming, station management and research valorisation will extend over the initial three years after project start-up. Technical Assistance at ISAR will end after the first two years, but will remain for 5 years at the ES. The purchase of vehicles will take place initially in the second year of the project, investments thereafter will serve fleet renewal. Agricultural and laboratory equipment will be purchased in the second and third year after the initial station lay out review has been done. Acquisition of documentation for the information center will be done over the project period. 5.24 Supervision Plan. A project launch workshop will be organized in June/July 1993, for which full participation of all donors will be sought. Thereafter, the project will be supervised formally twice a year but given timely and close attention from the Agricultural Country Officer in Rwanda in the early phases. Donors and IDA participation in NARC's Financial Committee's review of: annual work programs and budgets, research contract execution, annual financial audits and Research Committee scientific evaluation reports, will provide an additional element of supervision and donor coordination. NARC will organize a mid-term review of the project by December 31, 1995, to, inter alia, assess: (i) the progress made in the implementation of research contracting arrangements (percentage in terms of value of research programs implemented under contractual arrangements) and improvement in quality and relevance of research proposals; (ii) the progress in the implementation of the human resources management and development plan (staff evaluation procedures, new conditions of employment, recruitment and dismissal practices, and training program); (iii) the effectiveness of regional and international collaboration in research program implementation; (iv) donor coordination arrangements, including the effectiveness of the Finance Committee in consolidating research activities and their funding; and (v) the quality and relevance of scientific and financial audits and the implementation of their recommendations. Agreement on this was obtained at negotiations (para. 7.1 (r)). The Project Supervision Plan is in Annex 7.  40 VI. BENEFITS AND RISKS A. Anticipated Project Benefits 6.1 Project benefits are expected from: (a) accelerated and sustained agricultural productivity growth; (b) higher returns to farmers; (c) improved management of Rwanda's natural resources; and (d) increased foreign exchange earnings. Most of these benefits are indirect and obtained through improved priority setting, better research quality and relevance, greater research efficiency (earlier availability, higher returns to investment, and lower cost), and improved and more dynamic linkages with extension (complementarity with the ongoing Agricultural Services Project (Cr. 2026-RW)). 6.2 As in most other Bank Group-assisted agricultural research projects, no attempt has been made to quantify the economic rate of return a, due to: (a) inherent uncertainties regarding the timing and value of research findings, and the timing and extent of adoption of research findings by farmers; (b) difficulties in separating potential benefits of research and complementary investments, such as input supply, extension and training; and (c) the problem of assessing the value of negative research, which adds to the pool of knowledge and frequently benefits subsequent research efforts. Considerable evidence of ,j pV analysis shows that investment in agricultural research in both developed and developing countries has generated high economic rates of return, exCeeding 40 percent in most cases. 13/ B. Employment and Poverty Impact 6.3 The project will have only a comparatively small direct employment impact during its implementation period. Far more important will be its possible indirect long-term impC*4 on income generation in rural Rwanda. Through the adoption of improved agricultural produetoa and farm management techniques, the project will help to increase small holder productivity and farm output, thus generating additional income for farmers and their household members that constitute about 85 percent of Rwanda's population. Through its focus on generating new and improved technologies for the production of food and export crops, the project is ultimately aiming at helping to secure the food supply and increase incomes, and will thus be an important component of the country's antipoverty strategy. C. Environmental Impact 6.4 With only limited areas of natural forest and park land remaining, farmers in this densely populated country are the major custodians of its natural resources. Generating technologies fbr the sustainable use of these resources is one of the main objectives of the project. Through its emphasis on soil and moisture conservation, soil fertility management, agro-forestry and natural forest management, integrated pest management, and participatory on-farm and farming systems research, the project aims at stabilizing and improving farming in a degrading environment. Although most of the environmental degradation in Rwanda is man-made, the situation has evolved to a point where the farmer's interest and that of protecting the environment are now the same. The environmental Ia/ Agricultural Research (Sector Policy Paper, World Bank, June, 1981), Annex 3, pp. 64- 65.  41 impact of the project will therefore be largely positive. The focus of agricultural research is necessarily limited to: (i) what farmers can do on their own fields; and (ii) what communities can do for their common interest (e.g. watershed protection, etc.). The bigger questions of the protection of Rwanda's remaining bio-diversity, finding solutions for land tenure problems affecting the environment, pesticide usage (legislation and control) are beyond the scope of this project, but they are the main focus of the recently adopted National Environmental Action Plan (para. 1.20 (c)). The proposed project with its focus on sustainable management of Rwanda's natural resources is fully consistent with this Plan and constitutes an important component of it. D. Project Risks 6.5 A risk that may delay the realization of project benefits and affect the credibility of agricultural research in Rwanda is the desired change in attitudes of research staff needed to make research more demand-driven. The project emphasizes on-farm participatory research and close collaboration with extension staff and a number of measures have been included to help in bringing about such a change in attitude: training in participatory research methodologies (already successfully practiced in ISAR's bean program); and mandatory coordination with field extension staff for all on- farm trials and training of extension staff by researchers. However, changes in attitude take time, unless accelerated by motivation. A scientist is motivated by research results. A scientist working in applied agricultural research should find his motivation in seeing the results of his/her research being applied by farmers. 6.6 The major risks of the project are: (i) implementation delays as a result of continued political instability; (ii) inadequacy of public funding; (iii) weak research management; and (iv) Government reluctance to rely on regional collaboration. As to the first point, a peace accord between the Government and the invading refugees has been signed and a new Government is expected to be in place by mid-September (para. 1.5). The design of the project includes measures to reduce the last three types of risk, i.e: the introduction of contractual arrangements for research funding and Government funding through "block grants"; improving staff training and introducing independent staff selection procedures; and the earmarking of funds for Rwanda's participation in regional collaborative research. VII. AGREEMENTS REACHED WITH THE BORROWER, CONDITIONS AND RECOMMENDATIONS 7.1 Assurances obtained at Negotiations (a) Funding for the maize, wheat, triticale and coffee programs will only be made available upon presentation of satisfactory research proposals in line with the revised Master Plan; and which for wheat and triticale would include an economic justification (para. 3.12); (b) ISAR will only undertake research for third parties under contract (e.g. tea and sugar) when such contracts cover a reasonable share of indirect station operating and maintenance costs, personnel expenditures and ISAR general overhead (para. 3.12);  42 (c) By January 1, 1994, ISAR will discontinue research on the national Ankole herd, but will maintain a herd of approximately 1000 heads on approximately 1000 hectares at the Karama station, for which it will maintain a private sector style management, with separate accounts (para. 3.13); (d) After December 31, 1995, all research activities funded under the project will be undertaken exclusively through contractual arrangements between NARC and the eligible research institutions (para. 3.4); (e) The IDA Credit will be passed on to NARC and ISAR as a grant (para. 4.3); (f) NARC statutes will provide for exemption from Central Tender Board scrutiny for the procurement of research contracts, provided prior agreement has been reached on pre-qualification criteria and contractual standards. The Bank's guidelines will be applied with respect to the pre-qualification of research institutions for research contracts and on the elements of their review (para. 4.13); (g) Procurement monitoring will include the following elements: (i) prompt reporting contract award by borrower (para. 4.14 (a)); and (ii) comprehensive 6-monthly reports by NARC and ISAR to IDA indicating: revised timing of procurement actions, including advertising, bidding, contract award, and completion time for individual contracts; compliance with aggregate limits on specified methods of procurement (para. 4.14 (b)); (h) Each year, ISAR and NARC will each establish not later than August 31, their quarterly budget funding requirements for the next fiscal year. Government will not later than December 31 of the same year confirm its budgetary commitment and will deposit not later than January 15 of each year, the first quarterly advance. Prior to the start of each subsequent quarter, the agreed advance will be deposited (para. 4.18); (i) A special project account will be opened by Government for NARC at a commercial bank on terms and conditions acceptable to IDA, in which Government's quarterly contribution will be deposited (para. 4.18); (j) NARC and ISAR accounts and the Special Accounts will be audited by independent auditors acceptable to IDA; the auditors' report will be submitted to IDA within six months of the end of the NARC and ISAR fiscal years; and the report of the auditors will be of such scope and in such detail as IDA may reasonably request, including a statement as to whether or not IDA funds have been used for their intended purpose, and a certification that all expenditures claimed under SOEs during the year have been properly incurred for the project as designed (para. 4.19); (k) Government will ensure that a project completion report is prepared within six months of the Credit closing date (para. 4.20);  43 (1) Government will channel all its funding for agricultural research, including funds destined for IRAZ, through NARC (para. 5.3); (m) By December 31, 1993, NARC will adopt research proposal presentation guidelines, acceptable to the Association, which shall have been reviewed by its Research and Finance Committees and which include, inter alia, a review of ongoing research activities and research results in neighboring countries (and elsewhere) as part of the 'state-of-the-art' analysis, a mandatory economic justification (in particular when research is only economically justified when undertaken as a component of a regional program), and proposed provisions for impact evaluations, for review and approval (para. 5.10); (n) Terms of reference for the design of a national agricultural research planning, programming, budgeting and monitoring system, including the establishment of research databases and linkages with outside sources of scientific information, acceptable to IDA and for implementation by July 1, 1994 (para. 5.12); (o) Before December 31, 1993, ISAR will seek consultant assistance to train its staff in efficient station operation procedures and to further develop its cost-accounting and budgeting system to operate and maintain its research facilities (para. 5.14); (p) The Director General of ISAR, the Scientific Director and the Administrative and Financial Director will have terms of reference, qualifications and experience satisfactory to IDA (para. 3.10); (q) On the basis of the Human Resources Development Plan, to be prepared with assistance from ISNAR, ISAR will prepare by December 31, 1993 for approval by the Finance Committee of NARC: (i) proposals for criteria for the recruitment, evaluation/promotion, tenuring and termination of staff; and (ii) draft employment conditions including salary scales and employment under fixed term contract. ISAR will seek IDA approval for the proposed scheme of service prior to its implementation (paras. 3.20 and 5.17); and (r) NARC will organize a mid-term review of the project by December 31, 1995, to, inter alia, assess progress made with the institutional reforms and next steps to be taken in coordinating donor funding under a consolidated mechanism, including the generalization of contractual arrangements for the implementation of research programs administered through NARC (para. 5.24). 7.2 Actions undertaken prior to Board Presentation (a) Issuance of a Letter of Agricultural Research Policy, supporting project objectives and signed by the Prime Minister; (b) Submission of the revised draft statutes for ISAR and NARC to the Council of Ministers; (c) Agreement on ISAR's 1993 consolidated budget; and  44 (d) Submission of an action program to prepare a Human Resources and Management Development Plan. 7.3 Conditions of Effectiveness (a) Institutional - (i) execution of an Implementation Agreement between Government and NARC (para. 5.3); (ii) enactment of the statutes of NARC, acceptable to IDA (para. 5.3); (iii) establishment of the Executive Secretariat (ES) (para. 3.4); (iv) the setting up of an accounting system at NARC, satisfactory to IDA (para. 4.17); and (v) enactment of ISAR's revised statutes acceptable to IDA (para. 3.10); (b) Staffing - appointment of (i) a Research Manager and a Finance Officer in the Executive Secretariat of NARC (para. 3.4); and (ii) the Administrative and Financial Director of ISAR, with qualifications, terms of reference and experience acceptable to IDA (para. 5.15); and (c) Finance - the deposit of an amount of FRw 26 million into the NARC Project Account representing a quarterly funding estimate for the first project year (para. 4.18). 7.4 Conditions of Disbursement (a) (i) implementation of agreed standard contract specifications for research contracts by NARC; and (ii) agreement on pre-qualification procedures acceptable to IDA for contracting institutions will be conditions of disbursement for research contracts (para. 4.16); (b) The application of agreed research proposal presentation guidelines (para 5.10) will be a condition of disbursement for research operating expenditures and contracts (para. 4.16); and (c) Development plans have been prepared for the experiment stations and an action program, satisfactory to IDA, has been formulated for the implementation of civil works (para. 4.16); (d) No investments for civil works and land improvement will be made under the Credit at any given location (para. 5.22), until ISAR has obtained title to the land at that location, and excess land and buildings have been transferred to MINAGRI, leased or sold (para. 4.16). 7.5 With the indicated assurances and conditions the proposed project would be suitable for a Credit on standard terms of SDR 10.9 million (US$15.0 million equivalent) to the Government of Rwanda.  45 ANNEX 1 page 1 of 2 RWANDA Second National Agricultural Research Project Verifiable Indicators for Research 1. Initial proposal. Each component sub-program will be justified by an initial proposal which will: (a) present a constraints analysis in order to identify the problems; (b) summarize the existing state of knowledge; (c) identify research objectives; (d) estimate the anticipated impact of each distinct sub-group of objectives; (e) define research themes, with approximate estimates of rates of return; (f) define research methodology - both in the technical sense (including method for farmer participation), and in the locational sense (coverage of agro-ecological zones; on-station trials, and off-station trials of varying degrees of complexity); and (g) identify cost-effective work-plan (including cooperation with others such as extension services for down-stream work, and international research institutes for up-stream research). These proposals will be re-examined by each research team every year in the light of any new information obtained during the year. Such new information will include, amongst others: the results obtained from the research during the previous year; new data obtained by researchers elsewhere; advice given by external reviewers; the emergence of new constraints, or change in the importance of previously identified problems; and changed economic conditions. Verifiable Indicators 2. Adoption of new technologies in accordance with predicted schedule. The objective of each research program will be the improvement of sustainable agricultural production by small- holders through the adoption of improved techniques. Special impact assessment studies will be required (using methods similar to those developed for beans). As most research programs have been underway for some time, most should be ready for immediate impact assessment. 3. Release of new technologies. In this context, 'release' implies release for final on-farm testing/ demonstration. New research proposals will be required to predict when new technologies are likely to be available for dissemination - that is when some impact can be expected; those responsible for on-going programs will be required to make, and will be able to make better assessments of the dates of release of new technologies.  46 ANNEX I page 2 of 2 4. Annual evaluation of new information. Each research team will evaluate their own results (for the previous year and cumulative to date) together with any other relevant data which has become newly available from elsewhere (both technical and economic, see para 1 above). These data will be analyzed and recorded in forms suitable for detailed research records, and for reports for external consumption. 5. Adjusted research program. The annual evaluation of new information will be accompanied by appropriate modifications to the proposed research program - both in the short term (the next year's work plan) and in the long term. Such modifications will include: the acceleration or delay in the development of technology and, thus, in the anticipated dates of release of new techniques; related changes in the methodology to be adopted (both technical and locational); changes affecting the cost of the program; and any necessary revision to the economic justification for the program. 6. Revised annual work plan. All changes proposed will be incorporated into a revised work-program for the following year, a program which will redefine the sites proposed for research, and the cooperation anticipated both up-stream and down-stream; and it will develop a cost-effective method for implementing this plan.  47 ANNEX 2 page 1 of 6 RWANDA Second National Agricultural Research Project Conseil National de la Recherche Agricole (CNRA) Projet de Statuts CHAPITRE 1- STATUT DU CNRA Selon la loi No - du - - 199- et le decret No -/- de - 199-: Le CNRA est un etablissement public dote de la personnalite morale et jouissant de l'autonomie administrative et financiere. II a pour objet d'orienter la politique de formation scientifique et professionnelle de la recherche en coherence avec les besoins de l'economie et la politique agricole nationale. Afin de degager les moyens necessaires I l'accomplissement de sa mission, le CNRA dispose d'un Fonds d'Intervention, specialement cree I cat effet, et I partir duquel il peut apporter un concours financier I la realisation de toute action scientifique et professionnelle contribuant I cette politique. Le CNRA est place sous la tutelle administrative du Ministre de l'Agriculture et de l'Elevage et des Forets et sous la tutelle financiere du Ministre des Finances. II dispose d'un budget autonome. Le CNRA est compose de quatre organes (cf. organisation institutionnelle annexe 1) - le Conseil d'Administration, - le Comitd de Recherche, - le Comite Financier - le Secretariat Executif. 1.1. Le Conseil d'Administration Associant l'Etat, les associations paysannales, les ONG, le secteur prive, le Conseil d'Administration eat compose, outre le president, de 8 membres, preside par le Ministre de l'Agriculture et de l'Elevage (MINAGRI). Sa composition est etablie sur une base exofficio et nQ= ex-officio. Le nombre des membres ex-officio est fixe I quatre et sa composition eat la suivante: le President du Conseil d'Administration, le Directeur du Cabinet du Ministre des Finances, le Directeur de Cabinet du MINAGRI, le Directeur de Cabinet du Ministre de l'Education Superieure et de la Recherche (MINESUPRES). Celui des membres mn eoffco cinq: ses membres sont recrutes I partir des utilisateurs des resultats de la recherche: associations paysannales, ONG, secteur prive. I1 est constitue des comites independants de recrutement pour los postes de ce dermier groupe pour lequel il selectionne au moins trois candidats par poste. Le statut d'observateur est accorde aux personnalites suivantes: le Directeur de L'ISAR, le Directeur de l'ERST, le Doyen de l'Universite Nationale du Rwanda.  48 ANNEX 2 page 2 of 6 Sur la base des donn6es relatives I la recherche agricole, aux besoins de 1'economie, aux exigences de la politique gouvernemnentale etablis dans l Plan Directeur National de la Recherche Agricole, le Conseil d'Administration dlabore pour le Gouvernement des propositions d'orientation en matiere de recherche agricole et de formation profesionnelle, fixe pour le CNRA le programme d'actions prioritaires en termes d'objectifs qualitatift et quantitatfs et alloue les moyens financiers necessaires par l'approbation du budget, prepare par le Secretariat Executif et vote par le Comite Financier. Le Conseil d'Administration est nommd pour une duree de trois ans. 1.2. Le Comitd Financier Compose de - membres, le Comite Financier associe paritairement I'Etat en la personne du Directeur de Cabinet du Ministre des Finances, le Directour de Cabinet du Ministre du Plan et les representants des bailleurs de fonds nationaux et internationaux de la recherche. Sur decision du Comite, le Directeur de l'ISAR peut y Wtre present a titre d'observateur. Le Comite Financier definit lea conditions d'application des directives de l'AssemblEe Generale en particulier en ce qui concerne les criteres d'acces au Fonds d'Intervention auxquels les contrats de recherche agricole soumis I finmncement devront satisfaire. Le Comite Financier a pour objet de traduire en termes opdrationnels les programmes d'intervention du CNRA definis par le Conseil d'Administration notamment en decidant des financements a accorder aux actions de formation professionnelle entrant dans ces programmes. II vote le budget du CNRA, en controle son execution et arrfte les comptes financiers et le bilan de fin d'exercice. 1.3. Le Comitd de Recherche Compose de S membres, le ComitE de Recherche associe partiellement l'Etat, en la personne du Directeur de Cabinet du Ministre de L'Agriculture de l'Elevage et des Forets, le Directeur General de l'Agriculture et trois personnalitEs inddpendantes du milieu scientifique international. Sur decision du Comite, le Directeur de l'ISAR peut y sieger a titre d'observateur. 1.4. Le Secretariat Executif Le Secretariat Executif consdtue l'organe d'execution des decisions prises par le Consul d'Administration. Composo d'un Responsable Scientifique qui an est le chef, ot d'un Responsable Financier, agents sous contrats, appuyds par un personnel administratif (assistant administratif et une secretaire) il est nomme par le Comite Financier. Outre les pouvoirs qui peuvent lui Wre dElEgues par l Comite Financier, le Secrtariat Ex6cutif est plus particulibrement charg6: - d'organiser la prdparation, validation et la mis. I jour rgulibre du Plan Directeur, outil do base pour l'orientation des decisions pour la fixation des priorites de la recherche et de son financement;  49 ANNEX 2 page 3 of 6 - d'haluer leas requets de financement en fonction de criteres arretds par l1 Comite Financier et de leg soumettre I ce dernier pour decision, d'en assurer le suivi, d'organiser lea audits financiers et les evaluation scientifiques tels que prevus dans les contrats; - de prEparer los procedures visant la presentation des tableaux de bord pour le suivi technique et financier des propositions de recherche; - sur la base de donnees de l'ISAR et sur celles des tableaux do bord issus do I 'execution des contrats, d'otablir une base de donn6es pour assurer l1 suivi do la programmation scientifique, de l'execution des programmes, des resultats et de son financement; ces dElments de base devant tre prdsentes annuellement au Comito de Recherche et au ComitE Financier; - d'assister et conseiler les organismes de recherche dans l'elaboration do leurs programmes de formation; - d'assurer la gestion administrative, comptable et financibre du Fonds d'Intervention; - do preparer pour le Comite Financier, le budget at le programme d'action annuel qui sera soumis au Conseil d'Administration; - d'entreprendre des etudes et des activites d'animation et de sonsibilisation. - d'assurer le secretariat des Comites do Recherche et Financier; - d'assurer la mobiisation en temps voulu des sources de financoment des bailleurs de fonds et de l'Etat; A partir des orientations politiques et des critbres d'ligiitEb ddfinis par le Conseil d'Administration et le Comite Financier, le Secretariat Exdcutif procde I I'Evaluation des propositions de recherche et les transmet pour decision au Comite Financier avec un rapport de prEsentation et de synthese des evaluations. La dEcision d'allouer ou non un financement at le cas echdant, d'en fixer le montant, eat subordonnfe aux resultats do I'hvaluation, au programme d'actions prioritaires ddfinis dans lo Plan DIrocteur, aux ressources financitres disponibles (etat des engagements) et aux accords particuliers conclus avoc les bailleurs do fonds. Tout. proposition retenue fait alors l'objet d'un contrat d'exdcution conclus au nom du CNRA par le Secrtariat Executif et le promoteur de la proposition. Ce conut fixe en particulier, les conditions d'ex6cution de la proposition, do son Evaluation a mi-parcours, los resultats attendus et les modalitos de financement. Le Secretariat Executif en assure le suivi, voille au respect des clause contractuolles, assure les reglements financiers et hvalue lea rdsultats dont il soumet I la revue du Comit6 de Recherche.  50 ANNEX 2 page 4 of 6 CHAPrlRE 2 - STRUCTURE ORGANISATIONNELLE DU SECRFEARIT EXECUTIF 2.1 L fncion On peut identifier quatre fonctions dans le Secretariat Exdcutif plac6es sous la responsabilite du Responsable Scientifique: a) une fonction administrative dont les tiches principales sont notamment: d'administrer les affaires gdnErales; la gerer le personnel et de la logistique; de concevoir et proposer des ameliorations du systbme administratif du CNRA; de veiller au respect des procedures administratives et controler leurs applications; b) une fonction financiere charges notamment: do gErer les ressources financibres du CNRA; d'assurer le financement, le suivi et le contrOlo finacier des projets do formation; d'assurer la comptabilitd (gEneral. et analytique); de produire les etats financiers par periode et d'ean faire les rapports pdriodiques; de veiller au respect des procEdures financieros at contrOler lers applications; de concevoir et proposer des amEliorations du systbme financier du CNRA; c) une fonction technique qui consiste a: l'etablir des propositions do recherche; former ot assister los promoteurs de proposition do rechecho; instruire les dossiers de demande de contribution selon les procddures etablies I cet effet; soumettre au Comite financier los dossiers do demande contribution; suivre at controler l'exdcution des projets do recherche; *valuer la realisation des projets; concevoir at proposer des amdliorations des procddures d'1valuation dos dossiers de projets de formation; d) une fonction d'etudes at do ddveloppoment qui a pour tidhes essmtidllement: de rechercher des posuibiitEs de financemont; d'identifler des besoins do formation dos chercheou; d'animer des seminaires d'information et do sensbiliustion des patenairas sociaux sur los rOles devolus au CNRA; * do ddvolopper des syst&mes d'informatdons et dea support meddatiques sur la rochercho; d'assurer uno documentaot dos Informations aclisdo sur la recherche en crEat en mettam & la disposition des wsvica ue bmuq do doznearopri. ClAPITRE 3 - GESTION DU PERSONNEL DU CNRA Lo Personnel participant a la rEalization des objectifs et des activitds du CNRA sont soUs I'autorite du President du Conseil d'Administration par le Code du Travail.  51 ANNEX 2 page 5 of 6 Toute personne rdunissant lea conditions prevues par l'article - du rbglement general du personnel peut participer au concours de slelction organise par le CNRA en vue de doter en personnel lea differents postos dont l'ouverture ost prevue dans Is budget. Lea modalites des concours seront definies par les differonts responsables du CNRA au fur et I mesure des besoins. Les contrats de travail peuvent Wtre, soit I duree determinee, soit I duree indeterminee. Cependant durant la periode de demarrage, tous les contrats seront etablis k duroe determinee. Des systbmes d'evaluation du personnel soront mis on place pour orienter, ddfinir et spEcifier le deroulement des carribres au sein du CNRA. Enfin, la resolution des eventuels differents entre le personnel et entro l CNRA et son personnel se fera par reference au rbglement general du personnel et I la culture de l'etablissement dans leur specificite et I la legislation du travail en viguour au Rwanda dans lour globalitd. En ce qui concerne la formation du personel, uno des t*ches do l'assstance technique prevue pour ie CNRA est de former le personnel en vue de le preparer I la maItrise des principales operations de gestion de l'etablissement. CHAPfTRE 4 - LA GESIION FINANCIERE DU CNRA Trois categories de budget composent le systeme de gestion financiere du CNRA: - le budget de fonctionnement du CNRA; - le budget d'investissement propre au CNRA; - le fonds d'intervention de la Recherche Agricole (FIRA). Le budgetde fonctionnement du CNRA: II est fonde sur: une comptabilitd budgEtaire des charges do fonctionnement du Conseil d'Administration, du Comite Financier et du Comite do Recherche, du SecrEariat ExEcudtif et de ses diffdrents services. Elle eat mise en oeuvre par le Secrttaire Executif apres approbation du Comite Financier et du Consul d'Administration. Le budget d' istsement pore du CNRA: Qui east ogalement fondE sur une comptabiitd budgetaire des investissemonts, mise en oeuvre par le Secretaire Executif aprbs visa et approbation du Conseil d'Administration et du Comite Financier. Ces deux catEgories do budget sont annuelles et ajustdes semostriellement. Les procodures detaillees de mise en oeuvre swont dlaborees et prEcisees par le Secrdtaire Executif des sa nomination et son entroo en fonction.  52 ANNEX 2 page 6 of 6 La Fonds d'Intervention de la Recherche Agricole (FIRM: Les principes de mise en oeuvre de ce fonds sont prdsentes at precises dans la deuxieme partie indtul6e 'Fonds d'Interventon do la Recherche Agricole", dont l'utilisation est autoris6e par le Comite Financier et mise en oeuvre par le Secretaire Executif. D'une maniere generale, ces cat6gories de budget auront chacune des etats financiers distincts lesquels seront ensuite coordonnds (agr6g6s at consolidds) au niveau du CNRA dans son ensemble. Cette organisation n6cessite donc: - I'elaboration et la mise en place d'un plan comptable propre au CNRA, dans le cadre du plan comptable OCAM; - I'elaboration et la mise an oeuvre d'un programme informatique de gestion comptable, financiere et budgdtaire; at - I'6laboration des procedures d'appel d'offres fixees et arr6t6es par le Comite de Gestion, dont lea principes sont d6Ji enonces dana celles de la Banque Mondiale. Ces activites sont effectuees par le Responsable Financier.  53 ANNEX 3 page 1 of 2 National Agricultural Resarch Master Plan Priorities identified in 1990 version Original objectives 1. The 1989/90 version of the National Agricultural Research Master Plan (NARMP) was designed to support the basic requirement of national policy: food self-sufficiency. Secondary objectives were to encourage the production of import substitution products and to improve export crops (specifically coffee and tea). Research was to address itself to the problems of the majority of the rural population; it had to deal with all major agro-ecological zones; and it had to aim at intensifying agricultural production whilst reducing environmental deterioration. These objectives had to be conceived against the background of a rapid rate of population increase and the virtual total lack of unused land suitable for agriculture, and they had to be achieved with resources (both human and financial) which the NARMP realized would remain limited. Thus the NARMP aimed at improving overall planning, including improved coordination between the various participants in research, and between donors; it aimed at improving the management of programs through improving the calibre of researchers; and it directed research to deal only with the priorities outlined in the following paragraphs. To support these programs, it was envisaged that laboratory research capacity would increase from 6 scientific staff in 1990 to 12 in 1995 and 17 by year 2000. Laboratory services would comprise soils, plant protection, biometry, micro- biology, tissue-culture, food technology and veterinary. 2. Crop Production. The NARMP identified 12 crops as being of first priority: banana, beans, coffee, sweet potato, potato, cassava, sorghum, maize, wheat, rice, and soybean. Priority activities for banana are selection for disease resistance and yield, and cultural practices including crop associations and soil conservation. For beans, emphasis is on disease resistance, seed treatment, organic fertility maintenance, promotion of climbing beans, dynamics of varietal mixtures, and seed production. For coffee, the main thrusts are in maintaining/improving coffee quality, reducing the effects of diseases (mainly through selection), and improving cultural practices (soil cover, density, economics of fertilization, pruning). The swet potato program concentrates on the viral complex and purification of planting material. In potato, the accent lies on selection for horizontal resistance to mildew and tolerance to bacterial wilt, chemical crop protection, purification from disease and storage of seed potatoes. Priority activities for cassava relate to pest and disease control, and processing/preservation methods. Sorghum research focuses on agronomic approaches to striga control, on control of diseases and insects as appropriate to the various regions, and on utilization technologies. Maize research concentrates on selection, notably for zonal adaptation, yield, and disease resistance. For wheat, disease (notably rust) resistance is the main preoccupation, while for triticale it is the identification of adapted varieties and methods of utilization. In rice, priority is given to selection (yield and pest resistance) and maintaining soil fertility (rotations, weed control, water control and other cultural practices). Finally, for soybean, research emphasizes selection of adapted short-season varieties, nodulation studies, and food technology (recipes). For the crops' research agenda, the master plan's target was to have 31 scientists by 1995 (and 37 in the year 2000), as against 22 in 1990. The size of the agenda suggests that, even with the planned increase in the number of scientists, early results are entirely dependent on increased regional and international collaboration. Rwanda alone cannot deploy the necessary,critical mass to generate early results.  54 ANNEX 3 page 2 of 2 3. AnImal production. The master plan outlines research on small ruminants (goats and sheep) and catdo; it also lists distinct programs in animal feeding and agrostology, and on animal health; no reference is made to pigs or poultry. For sheep, the main objective is improved genetic stock for meat and wool. For goats, it is selection for rapid growth, carcass weight, and milk production. The priorities in respect of cattle are selection for milk yield, and development of technologies for farm-level milk processing and conservation of dairy products. These activities are linked to those for animal nurlfon (determination of nutritional value of available feeds) and agrostology (evaluation and selection of fodder species, sylvopastoral studies, pasture installation and management). The aninal health program in the master plan ambitiously comprises the pathology of small ruminants, cattle, and pigs, as well as veterinary public health, reproduction problems, traditional veterinary medicine, and mineral/vitamin deficiencies. The number of researchers in the above programs would rise from four in 1990 to nine in 1995, and 14 in the year 2000. 4. FarnIng systen, soil and water ansrvatlon and soil fertiliy. There was much overlap among these programs In the original Master Plan, and duplication with the commodity programs. Systems research includes socioeconomic surveys, analyses and monitoring; the study of cultural practices that contributo tO soil conservation; on-farm verificadon and demonstration trials on crops; studies of associations and rotations, of improved fallow, and on the use of farm residues; also a wide rnge of studies on animal feeds and feeding, manure production, and integration of fodder crops into farm systems. The soU and water conseraton program includes a study of farmer perception of erosion problms and their strategic decisions relating to conservation and soil fertility. It also includes evaluation studies of various soil conservation practices (both cultural and agroforestry systems) in terms of reduced runoff and erosion, water and mineral balances, and fertilizer ffectiveness. Soil fenrUty resoarch starts with a synthesis of all past work aimed at identifying technologies ready for diffusion. It further includes water, mineral and organic matter balances for different crops and soil types, evolution of soils under crops, studies on minor elements, and production of maps. There were seven researchers in 1990, anticipated to rise to 15 by 1995 and 26 in the year 2000.  MDANDA Secan N*mU AonRVAMd Ih.wdP As iDocumbw3 IM 1wmug 1h I COTANT 1112 TES 195 1966 1967 im 1m m 1991 ------------------------- ACTUAL -------------- CUREN4TAES8: 173 In1 211 13S In 21 176 Cb & Snads s0 64 43 62 115 144 11S R.oCebhs 32 S7 101 3S 39 51 35 hw.doFm ((N4 61 as 31 3I 33 24 F-ASEMS 116 204 36 624 m 761 E.hblWn_i Coob 4 5 a 14 16 16 16 BidbW & b*au.s 176 251 454 545 61S 3 a" 72 R_d-.Eqsdp. IS 62 as 162 211 236 246 Vald.. 42 57 5U s0 96 9 UWS1CII 0 0 0 47 55 50 46 06 2 15 1S 5 4 3 4 L*ss:AccuminmAd Doepsaon (210 (216) (219) (250 PM9) (4 (464 TOTALAS8ETUS 2s S92 007 757 11105 11X11 966 ______- a-------a a- a a --a---aa a --aaaa _ _ - -_. _a, UOOAtChlTEWdl EBOfR0WVtOS2k 49 9S 12i 179 147 135 156 Gkmtrnua 20 23 30 43 62 61 73 CGdias & Awa. Lb. 2 67 97 13S a5 74 S kid. lmAOM LaS 0 1 1 1 1 0 0 LONGTERMA BOBWQS: 4 3 2 1 0 0 soul1y: so6 763 1.165 1A91 2125 2.360 2.496 (Qanan FraL G3wnmw0s 07 5S1 640 739 1A1 1.107 1.162 IDA 57 151 354 61 *03 m 9651 FIDA 0 16 46 110 149 175 182 0"w Dons. 27 G6 a6 76 1C9 149 16S Gui Advanos 2B 29 56 36 3 41 40 Acomaw S.upkuA)uick (273 PM69 (1.014) (1.37) (1.459) (1.666 TOTAL UA3ILJIES & EWUlY 2n 3u2 607 757 11106 1.036 966 ' m m-- ----- -mamma mamm. mamma mamma mama 0  IWMADA AW hicm& pnhsAea Fa Yam ig Dc_mbw 3i CONBTUIT 191 TS 1965 19" 1967 1966 19 1og0 199b1 1912 ------------------------- ACUAL-------------- BUDGET OPRTNG EXPEN:S A -&'.- A 104 146 167 99 2n 177 164 164 Idub. a 6supIm 20 44 47 74 61 52 6 75 Tech. Asai A CmuL 2 111 74 as 41 33 3S 53 8umiy - 4 6 17 31 29 23 13 10 PwUnU an Ltab _aOd 0 0 0 a 1 16 20 0 TdOh Ea m 16S 310 30 372 3 302 344 22 LI: Sal _ard PAd m &1m od Pde u 45 56 44 12 66 So 46 41 &1_ csA_ (1) 2 3 5 1 0 5 0 0Iim he_ams 5 6 10 2 15 6 21 S EquIpmatDpedoan Subsh by O6wDonwr 16 3 5 6 6 5 5 S Op.Co.b8.&hAn*by OIwDcw 19 56 56 27 73 71 95 71 84 123 119 52 150 2D1 172 12D Not Cah Ded (79 (" (167) (320 (166) (102) (tOQ (17X Dqwec~n p32 2t 17 53 72 64 55 0 Not DeA (Ill) (213) ( PM (ml) (165) (2 (2 _____ ----- -----....... _____ -----....... __-....... .............. .......... .. 0p.  II 1 Ado i - kLf i -fd-Ro AWs Soto* A App&ml d Fwi sk1mmb Fa Yom dht Deosnb. 31 CONSTANT 192TERMS 195 196 l9o7 im 1#9 log0 tWt 192 SOURDEB ___,....- .u~ Ganb: Gh3 63 144 122 go 91 a5 50 tO5 DA 31 so 225 256 163 a 63 82 FDA 0 1S 33 78 SI 25 7 X Donors Equlpns4 34 32 15 16 37 27 19 45 Dars fOpwdm) 19 56 56 27 71 70 95 62 Olh Doorsm To 54 t0S 104 121 136 123 121 135 Sagasd PRw 49* so s II 72 125 71 44 Sale d A (1) (a 4 a I 0 (5) 0 Doaes in Wortkig S 3 12 as 0 0 71 t TOTAL SOURCES 218 413 517 60 467 3S# 372 367 APPLCTO Op. Cosas SO Fb1d 49 64 57 19 72 125 71 44 Op. Coash Fh-d by w L6a 131 S7 78 92 72 75 105 Op.ComaFine- by Date. 46 116 151 274 175 106 197 173 L_s: D _pr.c 0 0 0 0. 0 0 0 0 Nt OpsV Dat: 163 310 306 372 339 32 344 322 IhIANAIndinft 5S 90 212 220 65 35 25 45 Short Tm Borra ns I 0 0 0 hwbo msWwlu Cqa^l 0 tS 0 83 GI TOTAL APPUCATION 218 413 517 600 467 S 372 367 fe ; L.z C'.w 0  ANNEX 5 58 page 1 of 12 Rwwnde Second A giWd Pe_h Pr*4 C.oenees OeI C"t S TOtW IPRw 'O00I1 "Al '0003 resgu on" Local Foremign Total Local Forein Tota Exchae Cat A. Station 1. Resear