Rwanda - Second Phase Bugesera Gisaka Mogongo Rural Services (Phase II) Project : Rwanda - Bugesera Gisaka Mogongo Rural Services (Phase II) Project

Dowumnt of The World Bank Ion omCIAL USE OMlY RepWt N 12592 PROJECT COMPLETION RPORT RUANDA BUGESERA AND GISAKA MIGONCO - PHASE II PROJECT (CREDIT 1283-RW) DECEMBER 10, 1993 MICROGRAPHICS Report No: 12592 Type: PCR Agricutlure Operations 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 EQUIVALENTS Currency Unit = Franc Rwandais (FRw) FRw = US$1.00 1982 - 92.8 1983 94.8 1984 - 100.2 1985 - 101.3 1986 - 87.6 1987 79.7 1988 - 76.5 1989 - 80.0 1990 - 82.6 1991 - 125.1 10/1992 - 137.0 WEIGHT AND MEASURES Metric System ABBREVIATIONS AVSC Agro-Veterinary Support Center (Poste agro-v6t6rinaire) BGM Bugesera and Gisaka Migongo Project FAC Fonds d'aide et de coop6ration ISAR Agricultural Research Institute of Rwanda (Institut des sciences agronomiques du Rwanda) IITA International Institute for Tropical Agriculture MinAgri Ministry of Agriculture and Livestock Development MINITRAP Ministry of Transport OCIR-Caf6 Agency for Industrial Crops of Rwanda - Coffee (Office des cultures industrielles du Rwanda - Caf6) ONAPO National Population Office (Office national de la population) PSA Agricultural Services Project (Projet de services agricoles) RMEA World Bank Regional Mission in Eastern Africa SAR Staff Appraisal Report SSS Seed Service (Service des semences s6lectionn6es) T&V Training and Visit System of Extension FISCAL YEAR OF BORROWER January 1 - December 31 FOR OFFICIAL USE ONLY THE WORLD BANK Washington, D.C. 20433 U.S.A. Oflo of Dirotorenwal Operatione Evaluation December 10, 1993 MEMORANDUM TO THE EXECUTIVE DIRECTORS AND THE PRESIDENT SUBJECT: Project Completion Report on Rwanda Bugesera and Gisaka Migongo - Phase II Project (Credit 1283-R Attached is a copy of the report entitled Project Completion Report - Rwanda - Bugesera and Gisaka Migongo - Phase II Project (Credit 1283-RW)" prepared by the Agriculture Operations Division, Country Department I, Africa Regional Office. Part II was prepared by the Borrower. The project outcome is considered unsatisfactory. It failed to encourage widespread adoption of new farming technologies and it did not develop resilient institutional services. - The PCR concludes that area-based agricultural extension projects are inherently unsustainable. However, the evidence presented suggests that poor results in crop extension may have been related to inappropriateness of many of the technical recommendations. In general, the evidence is too weak to draw any confident lessons about alternative strategies for improving semi- arid smallholder farming systems. With the above reservation, the PCR is satisfactory. No audit is planned. Attachment This document has a restuicted diaisubutlon and may be used by recipients only in the perormance of their official dutles. Its conitets may not otherwise be disclosed without World Bank authoridon. FOR OFFICIAL USE ONLY RWANDA BUGESERA AND GISAKA MIGONGO - PHASE II PROJECT (CR. 1283-RW) PROJECT COMPLETION REPORT TABLE OF CONTENTS PREFACE ...... i EVALUATION SUMMARY ....... im PART I: PROJECT REVIEW FROM THE BANK'S PERSPECTIVE ..... . 1 Project Identity ................................................1 Background .................................................. Project Objectives and Description ................................... 2 Project Design and Organization ................................... 3 Project Implementation ........................................... 4 Project Results ................................................. 6 Project Sustainability ............................................7 Bank Performance .............................................7 Consultant Performance .........................................8 Borrower Performance ..........................................8 PART II PROJECT REVIEW FROM THE BORROWER'S PERSPECTIVE .......... 9 PART III: STATISTICAL INFORMATION..................................... 18 Table 1: Related Bank Loans and/or Credits ........................... 18 Table 2: Project Timetable ........................................ 19 Table 3: Credit Disbursements...................................... 20 Table 4: Project Implementation .................................... 21 Table 5: Project Costs and Financing ................................. 22 Table 6: Project Results .......................................... 23 Table 7: Compliancw with Credit Conditions ........................... 30 Table 8: Use of Bank Resources .. ................................. 31 Maps IBRD 15961, 15966, 15967 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. i RWANDA BUGESERA AND GISAKA MIGONGO - PHASE II PROJECT (CR. 1283-RW) PROJECT COMPLETION REPORT PREFACE This is the Project Completion Report (PCR) for the Bugesera and Gisaka Migongo- Phase II Project, for which Credit No. 1283-RW in the amount of SDR 14.5 million (US$16.3 million equivalent) was approved on July 13, 1982. The credit was closed on December 31, 1991, after a four year extension. The last disbursement took place on February 29, 1992, at which time the undisbursed balance of SDR 183,000 (US$206,000 equivalent) was cancelled. The French Fond d'Aide et de Coop6ration provided parallel financing in the amount of US$2.5 million equivalent. This PCR was finalized by the Agriculture Operations Division, South-Central and Indian Ocean Department of the Africa Region. The draft report was written by staff of the Food and Agriculture Organization/World Bank Collaborative Program, following a field visit in October 1992. It is based on project documentation such as the Staff Appraisal Report, the Development Credit Agreement, project correspondence and supervision reports. A copy of the report was sent to the Borrower on April 14, 1993, with a request for comments. These comments are presented as Part II of this report. m RWANDA BUGESERA AND GISAKA MIGONGO - PHASE II PROJECT (CR. 1283-RW) PROJECT COMPLETION REPORT EVALUATION SUMMARY Background 1. IDA has been involved in the Bugesera and Gisaka-Migongo (BOM) regions since 1975, when the first BOM (Cr. 668-RW) project was initiated. The project was designed to support a boad range of development activities (rural services and infrastructure) in tne area, building on the foundation of prior development schemes. This was to be achieved through existing Government institutions, including local Government entities, and training of project staff and farmers. Given the prevailing institutional weaknesses in the agricultural sector at the time, emphasis was placed on flexibility in implementation through annual work programs. The implementation record was mixed. Project management was good in many respects (timely submission of work progr.ms and audit reports) and the project was implemented on schedule. However, some fundamental problems had arisen. Instead of a small management apparatus, relying essentially on existing services, closely linked to local authorities (as desired and planned by the Government of Rwanda and IDA), a large, autonomous project unit emerged and little effort was made to involve local authorities and cooperatives in project management. The flexibility built into the project through the annual work program mechAism was not used to address sustainability of project investments. Results fell short of appraisal estimates in terms of agricultural and livestock production, and institutional development. Infrastructure development was substantial and generally impressive,.although less than planned, notably for roads, and at higher cost. Problems had arisen with inappropriate design and poor quality of construction. Extensive beef production proved to be unfeasible in the Rwandese context. Successful project activities included erosion control, distribution of forestry and fruit seedlings, rapid development of vegetable production, and the introduction of the stall feeding (fermette) system. Agricultural credit, marketing and input distribution efforts were disappointing because of management and technical shortcomings and insufficient attention to cost recovery. The first phase was completed in 1982, at which time the economic rate of return was estimated at 5 percent. Objectives and Components 2. The second phase project had as its primary objective to work with the rural population to: (i) carry out research trials and introduce cropping techniques and varieties for the semi-arid areas; (ii) promote soil conservation to prevent loss of soil fertility; (iii) integrate on-farm crop and livestock activities and promote small stock where appropriate; (iv) intensify cultivation of cash and foodcrops in areas of average and above average potential; (v) improve rural roads; and (vi) improve nutrition standards and expanded awareness of family planning alternatives. A further objective was to develop the project planning and implementation capacity of the prefectures, sub-prefectures and the communes. iv Implementatioa 4. The implementation of agricultural and livestock activities was mixed. Adoption of extension messages was low and the demand for veterinary services was initially high, but dropped substantially after the introduction of cost recovery measures. The distribution of improved seeds suffered from poor quality. 5. The applied research component suffered from a lack of on-farm research and did not develop proper interaction with the extension service. Farmers were reluctant to adopt the messages because of poor prospects for marketing incremental production and low cost-benefit ratios of the proposed production techniques. The Rubilizi training center performed satisfactorily. The livestock ranches at Rusumo and Nasho proved to be financially unsustainable and activities at the former were drastically reduced, while they were halted at the latter. Implementation of road rehabilitation and construction experienced considerable delays; in the end, however, appraisal estimates were exceeded by 80 percent. 6. The monitoring and ev luation unit, established at the beginning of the project, was closed with the departure of the Technical Assistant in 1986. No baseline and follow-up surveys had been undertaken in order to measure project impact on agricultural production. The studies that had been carried out lacked in-depth analysis. 7. Actual project costs are estimated at US$21.3 million (US$21.6 million planned), of which IDA provided US$16.5 million (US$16.3 million planned), the French Fonds d'aide et de coop6ration provided US$2.5 million (US$3.0 million planned) and Government provided, as planned, US$2.3 million. Although total costs were as planned, there were significant variations within certain cost categories. The civil works category of the research component and the National Population Office was 338% and 376% higher respectively than appraisal estimates, and the technical assistance for the rural roads component and the Ministry of Agriculture and Livestock Development were 665% and 139% higher respectively than appraisal estimates; on the other hand, the operating costs for the research and the roads components were only 70% of appraisal estimates. Results & Since no systematic e.- rt was made during project implementation to monitor farm developments and collect information on the participants' pre-project and current activities, project related production increases are difficult to quantify. It can nevertheless be argued that the Project contributed to maintaining yields and cropping intensity. In the without-project situation, yields would have decreased and soil fertility would have continued to decline, as a result of increasing population pressures. Apart from the release of some improved crop varieties and the distribution of trees for fruit and firewood production, other activities (composting pits, etc.) have had only a marginal impact on agricultural development. Livestock production has increased as a result of the cattle and small ruminant stalling practices and associated forage production which is attributable to the project. However, animal health care has not been continued. In general, the majority of smallholders in the project area appear to have continued to practice traditional farming techniques. Project staff and farmers visited confirmed that participating farmers obtained moderate increases in yields of cassava, sweet potatoes, sorghum, maize, and groundnuts which has been achieved largely as a result of improved varieties and improved manuring. Soy beans production was introduced as part of the Project. The project did not succeed in improving the interaction between researchers and extension agents. The social infrastructure (nutrition centers, roads) have no doubt benefitted many thousands V of families. By developing the needed infrastructure, the groundwork has been laid for future agricultural development. 9. At appraisal, the economic rate of return of the Project's directly productive components was estimated at 16%. This was, however, not considered the primary justification for the long-term investments proposed under the project. The primary justification was farmer training and the development of services and infrastructure in support of smallholder agriculture. Available data does not allow for a conventional economic analysis to be carried out; based on the limited available information, the economic rate of return is estimated at less than one percent (Table 6, Part III). Sustanabilty 10. As demonstrated by the rapid decline of activities after the termination of project financing in 1989, project activities have proven to be not sustainable. Support services for improved crop and livestock production have virtually stopped with the closure of 16 out of 24 Agro-Veterinary Support Centers. Much of the experience gained in the implementation of the T&V system has been taken into account during the preparation of the Agricultural Services Project (Cr. 2026-RW). This project, which became operational in 1991, is providing support to the reorganization of the extension services on a national scale. It has introduced specific measures to improve extension delivery. More emphasis has been put on the development of adaptive and on-farm research and the strengthening of links with research, to give momentum to the development of new messages and technical recommendations. Also, the linkages between the extension organization and farmers has been better defined. Its activities in the BOM area are, however, still limited. Lessons Learned 12. Four main lessons can be learned from the experience gained in the project: (i) area based agricultural extension projects are not sustainable; (ii) an extension project should not conduct independent research, not even adaptive research; (iii) reliance on technical assistance for project implementation is not sustainable; and (iv) monitoring and evaluation should be systematically carried out. 1 RWANDA BUGESERA AND GISAKA MIGONGO - PHASE I PROJECT (CR. 1283-RW) PROJECT COMPLETION REPORT PART I: PROJECT REVIEW FROM THE BANK'S PERSPECTIVE 1. Project Identity Project Name Bugesera and Gisaku Migongo - Phase II Project Credit No. 1283-RW RVP Unit AF3AG Country Rwanda Sector Agriculture Subsector Area Development 2. Background 2.1 IDA has been involved in the Bugesera and Gisaka-Migongo (BGM) regions since 1975, when the first BOM (Cr. 668-RW) project was initiated. The orientation of this project was based or the development of mixed farming and improvement of rural services in what were then two largely undeveloped regions of Rwanda. The first phase was completed in 1982. 2.2 The first phase project was designed to support a broad range of development activities (rural services and infrastructure) in the area, building on the foundation of prior development schemes. This was to be achieved through existing Government institutions, including local Government entities, and training of project staff and farmers. Given the prevailing institutional weaknes&s in the agricultural sector at the time, emphasis was placed on flexibility in implementation through annual work programs. The implementation record was mixed. Project management was good in many respects (timely submission of work programs and audit reports) and the project was implemented on schedule. However, some fundamental problems had arisen. Instead of a small management apparatus, relying essentially on existing services, closely linked to local authorities (as desired and planned by the Government of Rwanda and IDA), a large, autonomous project unit emerged and little effort was made to involve local authorities and cooperatives in project management. The flexibility built into the project through the annual work program mechanism was not used by IDA or Government to address sustainability of project investments. The result was that, at the time of credit closing, Government and IDA were faced with serious problems, because the institutional setup and related operating costs were neither sustainable nor replicable. Results fell short of appraisal estimates in terms of agricultural and livestock production, and institutional development. Infrastructure development was substantial and generally impressive, although less than planned, notably for roads, and at higher cost. Problems had arisen with inappropriate design and poor quality of construction. Extensive beef production proved to be unfeasible in the Rwandese context. Successful project activities included erosion control, distribution of forestry and fruit seedlings, and rapid development of vegetable production. In addition, the stall feeding (fermette) 2 system was successfully introduced. Agricultural credit, marketing and input distribution efforts were disappointing because of managerial and technical sLortcomings and insufficient attention to cost recovery. The economic rate of return at project completion in 1982 waa estimated at 5 percent. 3. Project Objectives and Description 3.1 The second phase project had as its primary objectives to work with the rural population to: (i) carry out research trials and introduce improved cropping techniques and crop varieties for the semi-arid areas; (ii) promote soil conservation to prevent loss of soil fertility; (iii) integrate on- farm crop and livestock activities and promote small stock where appropriate; (iv) intensify cultivation of cash and foodcrops in areas of average and above average potential; (v) improve rural roads; and (vi) improve nutrition standards and expanded awareness of family planning alternatives. A further objective was to develop the project planning and implementation capacity of the prefectures, sub- prefectures and the communes. 3.2 The Project included support for. (a) semi-arid crop development through applied and adaptive research; (b) strengthening extension support services through: (i) improvement of in-service training, establishment of local demonstrations and organization of a service network for cropping and livestock activities; (ii) promotion of an improved seed multiplication program; a tree nursery and distribution program for coffee, fruit and forest trees; a program to improve soil fertility through stall-feeding and compost production and erosion control measures; and promotion of improved methods of coffee plantation maintenance and organization of pest control campaigns; and (iii) training for cooperative management and technical support for agricultural credit. (c) livestock activities: support for the Rusomo breeding ranch and a program for upgrading local sheep and goat stock; (d) the Rubilizi Training Center for assistant-level (A3) zootechnicians; (e) nutrition center activities and development of a pilot family planning program in the project area 1/; (f) rural roads construction and improvement; and 1/ The PCR does not cover the component as nutrition centers and famly planning activities were taken over by other projects in the course of project Implementation. 3 (g) monitoring and evaluation and studies. 4. Project Design and Oralzatlon 4.1 The second phase project benefitted from some of the experiences acquired in the first phase, including less numerous components which were to be managed by the relevant local institutions, decentralization of responsibilities to the two Regional Directors in the Project area, discontinuation of unsustainable components, introduction of adaptive research activities, concentration of activities on extension and services provision, and better coordination with local authorities. 4.2 Unfortunately, the second phase project also inherited some of the design flaws of the first phase project: (a) an area based extension project with heavy TA, which is not a sustainable undertaking; (b) an infrastructure out of proportion with real needs; (c) a numerous and often inexperienced staff (an average of 70 agents per commune), disproportionate to real needs; (d) extension methods which were not adapted to local circumstances; moreover, there were few qualified and experienced personnel that could be given the responsibility for extension work; and (e) lack of knowledge of farmers' needs, which is indispensable for proper planning of extension activities. 43 There were major changes to the preparation report during the appraisal and post- appraisal missions. These were: (i) a substantial reduction in the scope of Project activities and a greater emphasis on improvement of extension services and agricultural research; (ii) limitation of the Project area to that of the first phase project; and the inclusion ot nutrition and family planning activities, an applied research component, and a rural roads component. The institutional mechanisms needed to ensure delivery of extension services and promote adoption by farmers, however, were given insufficient attention in project design. 4.4 The Project Coordinating Committee had functioned well under the first phase and remained responsible for overall implementation and the two regional directors for day-to-day project implementation. Responsibility for implementation was to be gradually decentralized from the Kigali bureau to the two Regional Centers. Local authorities were to participate in the project through Regional Development Committees. Project activities were to be, where possible, managed by technical ministries. The Bridges and Roads Department of the Ministry of Public Works (MINITRAP) was to implement the rural roads component, the Agricultural Research Institute of Rwanda (ISAR) was to implement the research component and the National Population Office (ONAPO) was to implement the pilot nutrition and family planning activities. 4 5. Project Implementation General 5.1 The credit was approved by the Board in July 1982 and signed in September of the same year. The credit became effective in March 1983, a delay of three months. Disbursements from the IDA credit proceeded swiftly until June 1989 when over 90% of the credit had been disbursed. This was largely due to a well established project administration unit and staff trained under BOM I. However, the credit closing date was extended twice, to allow for the completion of the rural roads component which had been delayed. All other project components were completed in 1989 and project activities were rapidly scaled down thereafter. The credit was closed on December 31, 1991, and the final disbursement was made on February 29, 1992, at which time the undisbursed balance in the amount of SDR 183,000 (US$206,000 equivalent) was cancelled. 5.2 Actual project costs are estimated at US$21.3 million (US$21.6 million planned), of which IDA provided US$16.5 million (US$16.3 million planned), the French Fonds d'aide et de coop6ration (FAC) provided US$2.5 million (US$3.0 million planned) and Government provided, as planned, US$2.3 million. Although overall costs were as planned, there were significant variations between cost categories. The civil works category of the research component and ONAPO was 338% and 376% higher respectively than estimated; the technical assistance for the rural roads component and MINAGRI were 665% and 139% higher, respectively, than estimated; and operating costs were 70% of the appraisal estimates for the research and the roads components. Implementation of Individual Components 5.3 Semi-arid crop development and farm systems research. The Project supported ISAR's research activities at the Karama Research Station in the Bugesera region. Technical assistance (a team leader and a plant breeder) for the farming systems research program was obtained from the International Institute for Tropical Agriculture (IITA). Although some efforts were made to undertake on-farm trials most of the activities remained on-station. The overall response of farmers was disappointing. In general, farmers claimed that mineral fertilizers were only acceptable if subsidized because of the poor cost-benefit ratio of the technical packages and the absence of market opportunities for excess production and that agro-forestry was wasteful of limited land resources and that it required increased labor with no concomitant increase in income for several years. Alley cropping practices were not accepted by farmers, who preferred traditional contour and boundary tree planting (the latter technique appears adequate for the region, where over 60% of the farms range in size from 1 to 2 ha and where the topography is generally plateau like). The final IITA/ISAR report recommended that a socio-econonic analysis of the region be undertaken and that the deployment of more research staff should be envisaged. Implementation of this component was hampered by: (i) a shortage of field transportation and field equipment; (ii) an inadequate number of properly trained staff; and (iii) poor coordination between research and extension staff, 5.4 Strengthening of extension support services. The Training and Visit (T&V) system was introduced in the project area in late 1984 and was fully operational by 1985. The project focussed on the in-service training of prefectorial and regional agronomists, the communal agronomists and veterinary officers, the staff of Agro-Veterinary Support Centers (AVSCs) and sector level extension agents. These training sessions (generally held in twice yearly sessions of one week each) were directed by the technical assistants. A local consulting company collaborated with project staff in preparing and conducting these courses. 5 5.5 Contact smallholders were often absent during formal training sessions because the time conflicted with agricultural work practices (agricultural work is carried out in the morning and peasants have to participate in communal work twice a week). This resulted in poor farmer participation (on average only 4 to 6 farmers participated in regular group meetings as compared to 12 to 14 estimated at appraisal). Farmers lost interest because only few contact people were visited and because meetings dealt with a small number of extension messages well known to them and often reflecting more the administration's priorities (e.g. coffee tree pruning and spraying, erosion control, etc.) than farmers' interest. The low level of literacy of the population in the zone did not allow for the use of technical leaflets and posters during contact group meetings. 5.6 A total of twenty-four Agro-Veterinary Support Centers (AVSC) were created during Phase 1 (19) and H (5), to provide agricultural and livestock support to smallholders. The role of the AVSC was increased in importance because it was given responsibility for: (i) the sale of agricultural inputs (small tools and equipment, pesticides, and improved seeds, plants and cuttings); (ii) the multiplication of improved seeds; (iii) the monitoring and evaluation of mass campaigns; and (iv) the sale of veterinary drugs and chemicals to be used for tick-control, vaccination campaigns, monitoring of tick-control and deworming programs, operating stud centers, supervising cow credit schemes, and monitoring of the health status of livestock in the area. 5.7 The extension system also listributed, through the AVSCs, fruit, forestry and coffee seedlings (produced in nurseries installed during Phase I and II). The seeds provided to farmers and produced on AVSC land or through seed producing farmers did not prove to be satisfactory. The varieties available for multiplication were provided by ISAR and had often not been properly tested at the Karama Research Station. Moreover, they frequently were delivered too late to be of use and few varieties were superior in performance to farmers seeds. Demand was thus limited. Cost of production was high and seed multiplication and distribution through the AVSC was therefore not sustainable without heavy subsidies. Smallholders were organized to produce and sell their own seeds. Cost of seed production remains high, quality is often mediocre due to insufficient quality control and the cost recovery potential is limited. Avocadoes and trees for firewood were successfully introduced in the project area. The quantities of selected seeds and details of forestry seedlings distributed by the Project are presented in Table 6 F (i) and (iii) of Part IIL 5.8 Uvestock extension and animal health services benefitted a very large number of cattle in the project area while services and supplies were subsidized. Full cost recovery was introduced, which significantly reduced the demand for vaccines and other veterinary chemicals, as farm cash income is low. 5.9 Erosion control measures were promoted actively and contour bunds were established in almost all agricultural areas of the project. Many compost pits were dug but they are mostly empty for lack of materials. An important aspect of maintaining soil fertility was the establishment of small livestock units with stall feeding practices. Between 1984 and 1987 a total of 6,050 and 6,900 such units were established for cattle and small ruminants, respectively. Mortality, however, is reported to be high, particularly since the use of the veterinary services has decreased after the introduction of full cost recovery. 5.10 Ivestock ranches and smaH stock promotion. The Rusumo breeding station never became financially viable. Recently, most of the existing stock was sold and the ranch's capacity has been reduced to 1,500 animals compared to its potential of about 3,500. Despite eradication efforts, 6 tsetse flies never completely disappeared. The Government is currently studying the possibility of privatization of the Rusumo ranch. The Nasho ranch has been settled by refugees. 5.11 Rubilliz training center. The project contributed to the financing of the operating costs of the Rubilizi training center. Actual costs exceeded appraisal estimates and Government could not cover its share of operating costs. Training of livestock personnel was e"Qctive and contributed to reduce the shortage of livestock staff at level A3. The center continues to operate satisfactorily. 5.12 Rural road construction. Rural road construction was a major component, accounting for about 30% of estimated project costs at appraisal. The initial target set for this component was the construction/improvement of 218 km of rural roads, of which about 138 km in the Gisaka- Migongo region and 80 km in the Bugesera region. During implementation this component was expanded and by credit closing, 402 km of rural roads had been constructed/improved (80 percent more), which was accommodated within the initial allocation (98% of planned estimates). This translates to either high efficiency or over-budgeting during project appraisal. Supervision missions noted that the roads brigade was well organized and produced a high standard of work. Rural roads construction was made on the basis of force account, a common practice in Rwanda at the time. Roads maintenance has generally been inadequate and this has put the investment in jeopardy. It had been agreed that the responsibility for maintenance was to be transferred to MINITRAP and the Communes. This did not happen and often the roads brigade had to carry out maintenance activities. 5.13 Monitoring and evaluation. A Monitoring and Evaluation (M&E) Unit, established at the beginning of the project was closed down with the departure of the FAC financed M&E Advisor. The latter was employed for four years instead of the five foreseen at appraisal. A number of studies (see Table 6-D, Part III) on livestock, credit, extension, women in agriculture, T&V extension were undertaken by the Monitoring and Evaluation Unit. While being informative, they lacked in-depth analysis. No baseline and follow-up surveys were undertaken to measure project impact on agricultural production and little is known on the evolution of yields, costs and adoption rates of new technology. A proper evaluation of project impact is therefore not possible. 6. Project Results 6.1 Since no systematic effort was made during project implementation to monitor farm developments and collect information on the participants' pre-project and current activities, project related production increases are difficult to quantify. It can nevertheless be argued that the project contributed to maintaining yields and cropping intensities. In the without-project situation, yields would have decreased and soil fertility would have continued to decline, as a result of increasing population pressures. Apart from the release of some improved crop varieties and the distribution of trees for fruit and firewood production, other activities (composting pits, etc.) have had only a marginal impact on agricultural development. Livestock production has increased as a result of the cattle and small ruminant stalling practices and associated forage production. However, animal health care improvements have not been maintained.. In general, the majority of smallholders in the project area appear to have continued to practice traditional farming techniques. Field visits have confirmed that participating farmers obtained moderate increases in cassava, sweet potatoes, sorghum, maize, and groundnuts, largely as a result of improved varieties and improved manuring. Soy beans production was introduced as part of the project. The project did not succeed in improving the interaction between researchers and extension agents. The social infrastructure (nutrition centers, roads) have no doubt benefitted many thousands of families. By developing the needed 7 infrastructure, the groundwork has been laid for future agricultural development. However, the project has, on balance, proven costly in manpower and in financial terms leading to difficulties for the Government to sustain the investments. 62 Econode benefits. At appraisal, the economic rate of return of the project's directly productive components was estimated at 16%. Project benefits were to be derived from the incremental fooderop and coffee production resulting from improved farming practices. The economic rate of return was, however, not considered the primary justification for the long-term investments proposed under the project. The primary justification was farmer training and the development of services and infrastructure in support of smallholder agriculture. 6.3 Available data does not allow for a conventional economic analysis to be carried out since neither the benefits that have accrued from the agricultural services nor the speed of accrual can be quantified. However, for illustrative purposes rough estimations (which have limited significance) were made. The assumptions are detailed in Table 6, Part IIL Based on these assumptions, the net incremental value of production per participating farm household was estimated at FRw 10,273 (US$75) and FRw 7,230 (US$53) for the Gisaka-Migongo and the Bugesera areas respectively and the economic rate of return at less than 1%. 7. Project Sustainability 7.1 As demonstrated by the rapid decline of activities after the termination of project financing in 1989, project activities have proven to be not sustainable. Support services for improved crop and livestock production have virtually stopped with the closure of 16 out of 24 AVSCs. Much of the experience gained in the implementation of the T&V system has been taken into account during the preparation of the Agricultural Services Project (Cr. 2026-RW). This project, which became operational in 1991, is providing support to the reorganization of the extension services on a national scale. It has introduced specific measures to improve extension delivery. More emphasis has been put on the development of adaptive and on-farm research and the strengthening of links with research, to give momentum to the development of new messages and technical recommendations. Also, the linkages between the extension organization and farmers has been better defined. Its activities in the BGM area are, however, still limited. & Bank Performance 8.1 The Project was regularly monitored and supervision missions were clearly helpful in speeding up procurement procedures, and ensuring that major covenants were met. Supervision missions were adequately staffed and were timely. In retrospect, it could be argued that the supervision of the roads component did not propose firm enough measures to ensure that road construction went according to plans and thus avoid the major delays during implementation. Also, not enough attention was given to ensuring that the improved infrastructure (AVSCs and Karama) and equipment financed by the Project would be put to suitable use after it was dc4ded that they no longer served the purpose for which they were built. As such, many of these buildings are no longer used. Details of supervision missions are provided in Part III, Table & 8 9. Consultant Performance 9.1 Consultants (especially the roads component technical assistance) did not sufficiently train local staff to take over their functions. As such their contracts needed to be extended and an extra seven man-years had to be provided. 10. Borrower Performance 10.1 The main project implementing agency - MinAgri - has, given the social and economic context in Rwanda, performed its technicm! responsibilities well. Supplies were adequate and timely. 10.2 The Borrower has been supportive of the project, particularly during the initial phases. However, the Borrower had problems in implementing the extension service in the field. Similarly, the M&E Unit did not function as foreseen even though considerable support had been provided. Also the operating costs for the Rubilizi Training Center had to be provided by IDA as the Government could not cover its share of these costs. The data relevant to the preparation of the PCR were not readily available, or were provided in raw format. Available data was partly conflicting and had apparently not been analyzed. 9 RWANDA BUGESERA AND GISAKA MIGONGO - PHASE II PROJECT (CR. 1283-RW) PROJECT COMPLETION REPORT PART II: PROJECT REVIEW FROM THE BORROWER'S PERSPECTIVE VI 1. Project Identity Title: Bugesera-Gisaka-Migongo Rural Services Project - Phase II (BGM II) Credit No.: 1283-RW Region: Africa Country: Rwanda Sector : Agriculture Subsector: Regional development 2. Project Objectives and Description 2.1 Objectives. The BGM II project was financed by a Development Credit (Cr. 1238-RW) in the amount of SDR 14.5 million (US$163 million). This credit was negotiated by the Rwandese Government with the World Bank and signed in September 1982. The project became effective in March 1983 and was closed on December 31, 1991; the last disbursement took place on February 29, 1992, on which date the undisbursed balance of SDR 183,000 was canceled. 2.2 The main objectives of the BGM II project were to work with the rural population in order to: (a) conduct trials and introduce improved farming techniques and varieties in the semi-arid zones; (b) promote soil conservation to prevent loss of fertility; (c) integrate on-farm crop and livestock activities and promote small stock where appropriate; (d) intensify cultivation of cash and foodcrops in areas with average and better-than-average potential; (e) improve the rural roads; and 2 This part of the PCR is a translation of the comments provided by the Government. 10 (f) improve nutrition status and expanded awareness of family-planning alternatives. A further objective was to develop the planning and execution capability of the prefectures, subprefectures and communes. 2.3 Description. The area covered by the BGM project comprises two quite different zones: the eastern part of the Bugesera region, made up of about 65,000 ha south of Kigali, and the Gisaka/Migongo region of Kibungo prefecture in south-eastern Rwanda, made up of about 170,000 ha. These two zones form part of Rwanda's eastern plain; they lie at relatively low altitudes (below 1,500 m on average) and have relatively low rainfall (less than 1,000 mm) compared with other parts of the country. According to ISAR, the zones' soils are made up of eroded hillside soils unsuitable for agriculture (30% of the area), and insufficiently drained bottomland and valley soils forming 60% of the area. The presence of soils requiring good humic maintenance, in conjunction with at times protracted dry periods, raises the problem of keeping the soil moist to allow good humification. In 1989 the project zone had a population of at least 350,000. 2.4 The project included support for. (a) development of crops suited to semi-arid conditions by means of applied and adaptive research; (b) strengthening of extension support services through: (i) improvement of on-the-job training, organization of local demonstrations and setting up of a network of services for crops and livestock; (ii) promotion of an improved-seed multiplication program; establishment of a tree nursery and of a program for distribution of coffee, fruit and lumber tree seedlings; preparation of a program to improve soil fertility by means of stall feeding, compost production and measures to combat erosion; promotion of improved methods of coffee plantation maintenance and organization of disease-control measures; and (iii) training for cooperative management and technical support for agricultural credit; (c) livestocla support for the Rusumo ranch and establishment of a program to improve local sheep and goats; (d) the Rubilizi training center for producing livestock assistants (level A3); (e) nutrition activities and preparation of a pilot family-planning program in the project zone; (f) construction and improvement of rural roads; and (g) monitoring and evaluation studies. 11 2.5 The project was scheduled to be implemented over five years (1983-87) at an estimated cost of US$21.6 million, US$16.3 million of which was to be provided by IDA, US$3 million by FAC and US$2.3 million by the Government. 3. Project Design and Organization 3.1 The design of Phase II of the project was determined on the basis of the results of Phase I, although no evaluation had yet been made. Its main aims were defined as follows: - clear the project of nonagricultural activities by entrusting them to the specialized institutions concerned; * decentralize the management of the project by transferring its responsibilities to the two regional directorates; - concentrate management of all agricultural activities in the regional directorates, particularly as regards training and agricultural and livestock extension; - set up a of monitoring unit. 3.2 Unfortunately, Phase II also inherited certain of the design shortcomings of the Phase I project: (a) infrastructure out of proportion with real needs; (b) large number of inexperienced personnel; (c) extension methods not adapted to local conditions; (d) insufficient knowledge of farmers' needs to be able to carry out planning and extension activities properly. 4. Project Implementation 4.1 Research on farming system and development of crops suited to semi-arid conditions. This component encountered some problems: lack of trained and experienced personnel for research, lack of specific messages for the farmers, lack of research/development liaison among all parties as to the tasks to be undertaken. The results obtained by this component are certainly commendable, but they are only the results of trials conducted on station under more or less ideal circumstances. 4.2 The research program having a long-term focus, lasting results could not be expected within the space of a few years. Especially, as the component in question was focused on different priority topics entailing a whole series of stages that needed to be completed before reliable results could be achieved. For instance, in their work program the researchers started a number of multi- year trials, some running for longer than five years, Le. longer than the duration of the project as funded by IDA. These multi-year research activities should not nonally be included in an extension project, where research should be aimed at adapting technologies already tested by an agricultural 12 research institute. Thus, by the end of the project certain new varieties of maize, groundnuts and peas were under study in a controlled environment on the stations. The on-farm research, however, could not convince the farmers, because results were disappointing. Farmer adoption rates were therefore low. Strengthening of the extension support services 4.3 The Training and Visit System. This system was introduced at the end of September 1984 and the new approach made it possible to correct certain shortcomings noted earlier such as the lack of training, lack of precise programming of activities and irregular monitoring of extension officers, while also serving to reduce the number of people farmers had to deal with. 4.4 Application of this system obliged the project to take certain steps, the most important of which were the redrafting of the terms of reference of each officer to distinguish extension activities from those of a different nature, intensive training of agronomists to make them all-round basic extension officers, the programming of various activities, the setting up of a precise schedule of activities, and the selection of priority extension messages. 4.5 The system was operated in a rigid fashion which made it hard to adapt to the local context and its special features, such as the fact that agricultural work is done sol,* in the morning, the obligation for all farmers to participate twice a week in umuganda 2/ work, the existence of bi- weekly markets to which most of the farmers go, etc. 4.6 The results of the first two years (1985-86) following the introduction of the T&V system showed that the priority messages selected (LAE 41, improved seeds, combination of crop farming and livestock, etc.) were readily accepted by farmers who were members of the contact groups and that the number of farmers from these groups participating in the training and demonstration sessions organized by the basic extension officers was increasing. On the other hand, as of 1987 the records show a marked drop in farmer participation in these training and demonstration sessions. 4.7 Among the main deficiencies noted and which are the basic reason for the low success rate of the T&V system, are the following: - rigidity of the system; - the high cost of the system in terms of personnel; - insufficient use of inputs owing to their high cost; - non-involvement of farmers in choice of technical messages, with the result that messages were selected that did not meet the farmers' real needs; - lack of close collaboration between technical staff and local administrative authorities and also in the training, sensitization and follow-up of farmers; Editor's note: Wu_anda Is communal work. Editor's note: It Is not clear from the text what LAE means. 13 - a very low level of literacy among the population of the area which rendered the traditional approach based on information sheets unusable; - finally, the inadequacy of the sparse technical data on agriculture and research results made it very hard to develop technical packages. 4.8 To provide support for farmers regarding crops and livestock, the project used the services of Agro-Veterinary Support Centers (AVSCs), particularly for making available some inputs and means of production and for improving local breeds by keeping breeding stock (cattle, sheep and goats) in the AVSC breeding centers. A seed program, covering production of improved seeds and plants and their dissemination in rural areas was started by the AVSC stores. The AVSCs were also involved in the sale of veterinary medicines, improvement of prophylaxis for area livestock by means of mass preventive or curative measures and the monitoring of livestock credit The role played by the AVSCs was definitely not insignificant, but unfortunately they were not integrated into the organizational structures that took over from the BGM project. There were altogether 13 AVSCs created in Bugesera and 11 in Gisaka-Migongo. 4.9 In the extension context, the BOM project has also provided the farmers with coffee, fruit-tree and lumber-tree seed and seedlings. Demand for the improved seed and seedlings was so heavy that multiplication had to be organized chiefly in certain AVSCs. Unfortunately, caution had to be exercised because the stock multiplied by the project came from either ISAR or SSS and had not been previously tested at the local ISAR station. Moreover, production costs were high, yet the stock was sold at below-market prices. 4.10 Between 1983 and 1987, the following inputs were made available for the Bugesera region: - Seed (kg) - : 107,366 - Cassava cuttings (No.) : 931,250 - Sweet potato cuttings N/ : 8,567 4.11 It should be noted that as the years progressed the quantities of seed provided declined for the simple reason that the price was too high, for one thing, while the farmers were also using seed obtained from their own crops grown from the improved seed. Then there was also the fact that certain local varieties in some cases seemed to do better than the improved ones, which was an important reason for farmers not to buy the improved seed. 4.12 Some activities were started in the context of the soil fertility improvement program. Thus in the Bugesera region a marshlands program was started after 1984. Between then and 1987, 5,904 ha in the Nyabarongo and Akanyaru marshes was sown with various improved seeds (beans, maize, potatoes, soybeans, sweet potatoes, etc.). 4.13 Anti-erosion measures were also initiated by the project in two phases, by establishing multiplication fields of soil-anchoring grasses and providing technical advisory services to the population. By the end of 1987, 85% of the Bugesera region's area was covered by anti-erosion measures. The creation of collective pastures and forage-crop fields also contributed to soil 5/ Editoes note: the unit was not specified. 14 conservation. In addition to the anti-erosion measures, the project undertook a vast program for producing organic manure and for introducing the use of mineral fertilizer to improve fertility. As regards organic fertilizer, the objective was to have two composting units per farm. By the end of 1987 this objective had only been 70% accomplished in Bugesera while in Gisaka as of the first half of 1991 the figure was 82%. 4.14 Promotion of small farms was a major project aim. Thus, in Bugesera as of December 31, 1987, 15.358 small farms had been created while by the first half of 1991 323 had been set up in Gisaka. J 4.15 A survey was undertaken to determine what needed to be done so that each farmer could have a livestock operation in the area. 4.16 As to the use of mineral fertilizer, trials were started in 1987 in collaboration with the FAO Fertilizer Project and the demonstrations set up in the project zone were attracting farmers' interest. 4.17 The forestry and agro-forestry programs were a great success in the project zone. Between 1983 and 1987, 8 million seedlings were produced in the nurseries covered by the project. These seedlings were used to set up plantations; by the end of 1988, about 2,600 ha had been planted in this way by the BGM project. The fruit-tree campaign, with distribution of avocado and pineapple seedlings, proceeded at the same pace as the lumber-tree and agro-forestry programs. During the second phase the project distributed 620,277 fruit-tree seedlings in the Bugesera region. The experiment was not started in the Gisaka region until 1987. 4.18 With regard to industrial crops, and specifically coffee, the project placed special emphasis on technical operations such as pruning and pest control, the equipment for which was provided free of charge at the start of the project in order to encourage the farmers. Subsequently, the project was to ensure that equipment and materials were available in the AVSCs for cash sale. The insecticide (sumithion 3%) was provided by OCIR/CAFE. In Gisaka, the borer that is a major pest in certain zones was a source of considerable discouragement for the coffee growers. Apart from pest control for coffee, the BGM project has assured crop protection by providing instruction in cultivation techniques that help to reduce disease and by making disease control products available to farmers through the AVSCs. The absence of records of products provided means that no data are available to determine the farmers' interest in using these products. 4.19 In the livestock sphere several actions were undertaken to assist farmers in the BGM pro:ct area. Livestock feeding has been improved by the establishment of collective pastures and forage-crop fields. In Bugesera the project invested considerably in three pastures of 2,596 ha. According to a count made at the end of 1990 there were more or less 30,000 head of cattle in the area. 4.20 Between 1983 and 1987, 460 ha of forage-crop fields were established in the region. The Gisaka region alone had 4,997 ha of collective pastures and forage-crop fields and 4,826 head of cattle according to a 1990 count. Editor's note: It Is not clear whether this refers to resettlement of farmers or to other activities. 15 4.21 Genetic improvement of livestock was also one of the project's main thrusts. In this context, breeding centers were set up in each AVSC with bulls, rams and goats for breeding purposes. Only limited use was made of this breeding stock, despite the fact that servicing was free of charge. In view of the very high cost of keeping these breeding animals in the AVSCs, these activities were halted and AVSC staff have started looking for farmers or groups of farmers interested in purchasing the animals. In Bugesera, nine bulls have thus been placed with stockfarmers. In Gisaka the centers have been abandoned and the animals have been placed with groups of stockfarmers with improved pastures. Artificial insemination has also been started in the area, but the success rate is still low. 4.22 Much effort was also invested in the animal health field, especially prophylaxis, namely tick control, worming and vaccination by means of systematic treatments. It is gratifying that a large number of animals have been treated in this way, although this has admittedly been due to the fact that the services and supplies have been subsidized, because it is the cost of treatment that ik the main obstacle to expansion of the health programs. 4.23 The BOM project further initiated a program entitled "Action Fermette," which was to encourage the integration of cropfarming and livestock through stall feeding. Credit for cattle purchases was provided to promote this program. Credit was provided for sheep purchase in some cases, but the farmers displayed a very definite preference for cattle. By the end of 1987 there were 15,358 "fermettes" in the Bugesera region, for about 44,609 families, thus making 0.3 "fermettes" per family. This figure is lower than expected, since the aim was to have two "fermettes" per farm. 2 4.24 The Rusumo ranch was a full component of the BGM project, with its own management, while its accounts were handled by the Kigali coordination office. In the second phase of the project the ranch's initial role was to provide farmers with improved heifers within the framework of the "fermettes" under the livestock credit scheme. Despite the considerable investments put into it, the ranch's financial position remains catastrophic. This is also aggravated by cattle mortality, since the tse-tse fly remains a major problem in the region. In 1993 the ranch was estimated to have 1,500 head of cattle. 4.25 RubilizI Training Center. The project contributed to the financing of this center until December 1986. As of March 1988, EDF has taken over the funding of fellowships. Up to the present, the center has trained nearly 450 livestock assistants level A3, who are now working throughout the country. The Ministry of Agriculture and Livestock is presently negotiating with the Ministry of Primary and Secondary Education to have the latter assume responsibility for the center, since long-term financing for it by EDF is not assured. The center is continuing to operate satisfactorily. 4.26 Construction of rural roads. This component was entrusted to the Ministry of Public Works, which was to handle execution and maintenance thereafter. By and large the work has been properly carried out, although maintenance is inadequate. 7 Editor's note: It is not clear where this goal came from. There may have been some confusion between the goal for compost heaps and and the number of families participating in the stall feeding program. 16 5. Project Results 5.1 Because project activities have not been monitored regularly, it is hard to evaluate its impact in the area owing to the lack of reliable data. However, some noteworthy accomplishments have been achieved in certain areas. Yields of certain crops have been maintained thanks to the introduction of improved seed and fertilizer use. Treefruit crops and also forestry are similarly becoming more profitable undertakings. Livestock has become a key focus for the farmers. Each production unit keeps poultry or small livestock. In this way, the "fermette" campaign has more or less developed to the point that there are some "free" fermettes which have been set up without benefit of livestock credit. 5.2 Development of livestock health measures, both preventive and curative, has slowed down after the introduction of full cost recovery for most services, which are furthermore only available in a limited number of centers, which means that they are at a considerable distance from the farms. The.same applies to the genetic improvement programs. The crossbred or purebred breeding stock placed in AVSCs were not considered robust enough by the farmers, who had misgivings as to how their progeny would turn out. 5.3 The research component, which was one of the main components of the project, has not produced the expected results. Research has to be a long-term undertaking, and whatever structure is made responsible for extension in this area will always have to find solutions to problems put to it by the farmers. 5.4 The planned facilities were built, but the recurrent costs induced by certain of them meant that they could not be maintained. The roads and tracks built have improved communications and trade with other regions. 6. Project Sustainability 6.1 The lack of adaptive research and on-farm research has been the main obstacle that explains the absence of new extension messages and the monotonous repetition of the old messages. 6.2 Mention must also be made of the population's limited participation in the identification, solving and follow-up of problems. The links between ISAR, responsible for agricultural research, and extension have displayed weaknesses. 7. Bank Performance 7.1 The World Bank fulfilled all its commitments. The programmed supervision missions were performed as scheduled. However, when the project was designed, no consideration was given to the assumption of project activities by the beneficiaries or the local government. Thus, certain infrastructure facilities put in place by the project have been neither used nor maintained. The World Bank rightly assigned certain activities to specialized institutions for execution, which certainly enabled the project to concentrate on the agro-livestock technical and extension activities for which the Ministry of Agriculture was responsible. 17 8. Consultant Performance 8.1 A deficiency was observed regarding execution of the road component, both the duration and the cost of which were over estimate. 9. Borrower Performance 9.1 The Borrower also met its commitments in the sense that it assumed its technical responsibilities. It made the required technical personnel available to the project in sufficient numbers. It also fulfilled its commitments in terms of provision of counterpart financing. However, the project ran into several problems in the organization of the extension system. The project did not have specific technical agricultural messages to put before the farmers. 9.2 The monitoring and evaluation unit set up in the BGM project was attached to the Ministry of Agriculture which handicapped its activities at the project level. 9.3 The Government has not been able to fund the operating costs of the Rubirizi training center. The center is currently financed on a short-term basis by EDF, which means that there is a risk that its activities may have to be discontinued in the future, if the government is unable to fund its continued operation. 10. Lessons Learned 10.1 Several lessons can be learned from this project. (a) Emphasis should be placed on adaptive and on-farm research and on strengthening the links with the research institutions so as to speed the formulation of new messages and new technical recommendations. (b) The extension system should be adapted to the socio-cultural and economic circumstances of the beneficiaries. (c) The monitoring and evaluation of extension needs to be set out more precisely. (d) At the time projects are prepared, attention should be given to the transfer of responsibility for project activities once the credit is closed. 18 RWANDA BUGESERA AND GISAKA MIGONGO - PHASE II PROJECT (CR. 1283-RW) PROJECT COMPLETION REPORT PART II STATISTICAL INFORMATION Table 1: Related Bank Loans and/or Credits Loan/Credit Year of Project US$ Status Title Approval Amount Cr. 668-RW 1977 Bugesera and Gisaka 14.0 Closed 1982 Migongo Rural Development Project Cr. 1126-RW 1981 Lake Kivu Coffee and 11.8 Closed 1988 Foodorop Production (PCCV) Project Cr. 1546-RW 1985 Agricultural Research 11.5 Closed 1991 Project Cr. 1669-RW 1986 Gitarama Agricultural 11.7 Closed 1991 Production Project Cr. 2026-RW 1989 Agricultural Services Project 19.9 Ongoing 19 Table 2: Project Timetable Activity Planned Date Revised Date Actual Date Identification 10/79 Preparation 04/81 Appraisal - 05-06/81 Post-appraisal 11/81 Negotiations 05/82 Board Approval - 07/13/82 Credit Signature - 09/28/82 Credit Effectiveness 12/29182 - 03/29/83 Project Completion 12/31/88 12/31/89 12/31/91 Credit Closing 04/30/89 12/31/89 12/31/91 Final Disbursement 10/31/89 04/30)92 02/29/92 Source: Project Files. A two-year extension was needed to complete the roads component. All other project activities were completed by December 31, 1989. 20 Table 3: Credit Disbursements Culutative Disbursement Estimates Actual Cumulative Disbursements IDA Fiscal month/Year SOR *000 us$ *000 Revised SOR 4000 us$ o000 Actuali as a Year Estimates of Estimates IerUS 000US$) 1983 12/82 900 06/83 1,200 1,083 1,171 97 1984 12/83 2,500 2,526 2,690 10? 06/84 4,300 4,301 4,562 106 1985 12/84 6,000 5,279 5,547 92 06/85 7,500 6,057 6,313 84 1986 12/85 8,800 7,328 7,623 86 06/86 10,000 8,978 9,535 95 1987 12/86 11,300 9,533 10,203 90 06/87 12,500 10,856 11,904 95 1988 12/87 13,800 11,008 12,102 87 06/88 15,000 12,243 13,766 91 1989 12/88 16,300 12,672 14,334 88 06/89 16,300 15,000 13,165 14,973 92 1990 12/89 15,500 13.475 15,366 99 06/90 16,000 13,765 15,750 98 1991 12/90 16,300 13,989 16,067 98 06/91 16.500 14,174 16,325 99 1992 12/91 14,500 16,700 14,299 16,495 99 02/92 14,500 16,700 14,316 16,518 99 Cancellation 02/92 - - 183 206 Source: SAR and Project Files. 21 Table 4: Project Implementation Indicators Unit Appraisal Estimate Actual 1. Project management Vehicles: Cars unit 7 5 Motorcycles unit 3 1 Technical Assistance FAC: Extension man-years 10 6 Financial Advisor man-years 1.25 0 Monitoring and evaluation man-years 5 4 IDA: Research (IITA) man-years 9.8 9.8 Family planning man-months 9 9 Roads man-years 8 15 II. Extension AVSC constructions unit 5 5 Vehicles: Cars unit 10 3 Trucks unit 2 20 motorcycLes unit 0 2 11I. RancS Rusuo/Nasho Purunase of heifers unit 200 200 Import of sheep/goats unit 100 0 IV. Research (ISAR) Construction: Laboratory unit 1 1 Office unit 1 1 Housing unit 1 6 Vehicles: Cars unit 4 1 Motorcycles unit 4 2 Trucks unit 2 4 V. Family planning/nutrition Construction new unit 6 6 Vehicles: Cars unit 3 2 Motorcycles unit 6 15 Bicycles unit 0 39 VI. Roads Heavy equipment: Dozer unit 1 2 Tractor unit 2 3 Vibrant rotter unit 3 3 Pick-up unit 1 5 Source: Project Files. 23 Table 6: Project Results A. Crop Yields of Participating Farmers Pre-project SAR Estimate at Full Development PCR Estimaye in Without With Project With Project Project Increase Yield kg/ha Zone 1 - Coffee groving Coffee (g/tree) 400 450 50 450 Banana (beer) 8,000 8,000 8,000 Banana (fruit) 8,000 - 8,000 8,000 Cassava (fresh) 6,000 5,700 7,500 1,800 7,500 Sweet potatoes 4,500 4,300 5,500 1,200 5,500 Maize 800 750 900 150 900 Sorghum 1,000 950 1,200 250 1,200 Groundnuts 750 700 800 100 800 Soya beans * 550 550 550 Beans 650 650 650 Zone 2 - Semi-arid area Coffee (g/tree) 350 350 350 Banana (beer) 8,000 8,000 8,000 Banana (fruit) 8,000 8,000 8,000 Cassava (fresh) .5,000 4,750 7,000 2,250 7,000 Sweet potatoes 4,000 3,800 5,000 1,200 5,000 Beans 600 - 650 50 650 Maize 800 570 800 230 800 Sorghum 900 850 1,000 150 1,000 Groundnuts 650 620 700 80 700 Soya beans - - 500 500 V No hard data on current yields are available and for comparative purposes the PCR mission retained yield increases as described in SAR. Project staff and farmers confirmed that participating farmers have obtained a moderate increase in yields of cassava, sweet potatoes, sorghum, maize, and groundnuts largely due to improved varieties and improved manuring. Soya beans was introduced as part of the Project. Livestock production has increased as a result of the cattle statt feeding and fordge production practices. Table 6: Project Results B. Farm Model Development Pre-pro ect SAR Estimate at Full Devetcoment PCR Estimate in 1992 Area Production Vatue Cash Area Production Vatue I Cash Area Production Value Cash Income IncomeIncome (ares) (kg) (FRw) (ares) 9(k) (FRW) (ares) (kg) (FR) Zone 1 - Coffee growing Crops Coffee 12.5 60.0 7,200 7,200 15.5 84.0 10,080 10,080 15.5 84.0 10,500 10,500 Banana (beer) 30.0 2,400.0 16,800 6,720 30.0 2,400.0 16,800 6,720 30.0 2,400.0 24,000 9,600 Banana (fruit) 5.0 400.0 4,000 1,000 5.0 400.0 4,000 1,000 5.0 400.0 5,200 1,300 Cassava (fresh) 10.0 600.0 6,000 600 10.0 750.0 7,500 1,125 10.0 750.0 7,500 1,125 Sweet potatoes 5.0 225.0 2,250 340 5.0 275.0 2,750 550 5.0 275.0 4,100 820 Beans 25.0 162.5 2,925 150 33.5 218.0 3,920 785 33.5 218.0 8,700 1,740 Maize 5.0 40.0 400 40 5.0 45.0 450 45 5.0 45.0 1,800 180 Sorghum 20.0 200.0 2,800 700 20.0 240.0 3,360 840 20.0 240.0 7,200 1,800 Groundnuts 5.0 37.5 2,625 400 7.0 56.0 3,920 980 7.0 56.0 7,800 1,950 Soya beans - - * 3.0 16.5 412 - 3.0 16.5 1,320 - Livestock Mitk (titers) 30 750 250 60 1,500 750 30 3,000 1,500 Meat - miscettan. 2,500 - - - 6,375 3,187 Fruit trees 2.0 65.0 650 - 8.0 260.0 2,600 1,300 8.0 260.0 2,600 1,300 49,000 17,400 62,000 24,175 90,095 35,000 Zone 2 - semi-arid Area Crops Coffee 6.0 25.6 3,000 3,000 6.0 25.2 3,000 1,000 6.0 25.2 3,150 3,150 Banana (beer) 15.0 1,200.0 12,000 - 15.0 1,200.0 12,000 * 15.0 1,200.0 12,000 Banana (fruit) 2.5 200.0 2,000 400 2.5 200.0 2,000 400 2.5 200.0 2,600 520 Cassava (fresh) 10.0 500.0 5,000 1,650 10.0 700.0 7,000 2,300 10.0 700.0 7,000 2,310 Sweet potatoes 8.0 320.0 3,200 480 9.0 450.0 4,000 800 9.0 450.0 6,750 1,350 Beans 25.0 150.0 2,700 270 30.0 195.0 3,510 470 30.0 195.0 7,800 1,170 Maize 5.0 40.0 400 * 5.0 40.0 400 - 5.0 40.0 1,600 - Sorghum 15.0 135.0 1.900 500 15.0 150.0 2,100 525 15.0 150.0 4,500 1,125 Groundnuts 5.0 32.5 2,275 1,820 8.0 56.0 3,920 3,136 8.0 56.0 7,840 6,272 Soya beans . * 4.0 20.0 375 - 4.0 20.0 1,600 - Livestock Milk - - - -- Meet 1,500 400 3,000 750 4.250 2,125 Fruit trees * * * 4.0 130.0 1,300 - 4.0 130.0 1,300 - 32,500 8,520 42,605 11,381 60,390 18,012 a are 0.01 hectare Table 6: Project Results C. Producer Benefits SAR Estimate PCR Estimate in 1992 - Pre-project Full Development With Project Increase Actual Senefits Constant Terms With Project Increase FRw us$ FR; US$ FRw US$ FRw = US$ FRw FRw Zone I Value of production 49,000 533 62,000 675 13,000 142 90,095 658 59,273 10,273 Cash income 17,400 200 24,175 265 6,775 65 35,000 255 23,026 5,626) Zeo 2 Value of production 32,500 353 42,600 463 10,100 110 60,390 441 39,730 7,230 Cash income 8,F,20 93 11,400 124 2,880 31 18,012 131 11,850 3,330 I Actual benefits are 1992 values. Benefits expressed in constant terms have been recalculated in 1982 FRw equivalent. 26 Table 6: Project Results D. (i) Economic Impact SAR Estimate PCR Estimate Rate of Return measuring net incremental benefits of directly productive project 16% <1% investments The main assumptions are as follows: Staff Appraisal Report. (i) Estimated average life of investments: 20 years. (ii) Value of net incremental production projected to result from the project: 45% of the farmers would be using selected seeds and applying improved cultural practices. (iii) Estimated yield increases are summarized in Table 6. A. (iv) All expenditures for capital investments, replacements, operation and maintenance costs related to directly productive components (i.e. project management, extension activities, Rusumo ranch and rural roads). Project Completion Mission: (i) Estimated average life of investments: 20 years. (ii) Value of net incremental production projected to result from the project: 10% of farmers using selected seeds and applying improved cultural and cattle staff feeding practices. (iii) Estimated yield increases are as summarized in Table 6. A. (iv) Only expenditures related to the MINAGRI (47.5% of total Project costs) components (i.e. project management and extension activities) have been used. As the extension component was significantly reduced as from 1989 pending the start-up of the Agricultural Services Project, which was not yet operational in the field at the time of the FAO PCR mission, future operational costs have not been counted as part of the BGM II project. 27 Table 6: Project Results D. (ii) Rural Road Construction SAR Estimate Actual Construction/improvement 218 km 402 km E. Studies Date Origin/Consultant Title 03/84 Division études et projets. Recensements sur l'élevage bovin en Project BGM. Cellule de suivi- stabulation dans le BGM. Situation évaluation. au 1er janvier 1984. 08/85 Cellule de suivi-évaluation Enquête sur le crédit outillage dans project BGM II. le Bugesera. 05/86 République française. Ministère Le project BGM au Rwanda. de la coopération. 10/86 MINAGRI. Project DERVAM. La vulgarisation agricole du project BGM-Bugesera. Etude de cas. 02/88 Réseau des femmes oeuvrant Etude sur la valorisation et pour le développement rural. l'intégration des activités de la femme dans le domaine agro- pastoral. Zone d'étude: project BGM. 1988 Project BGM. Direction Enquête sur le système de régionale Bugesera. vulgarisation par formation et visites au project BGM Bugesera. Table 6: Project Results F. (i) Selected Seed Distribution 1983 1984 1985 1986 - 1987 Appraisal Actual Apprasat Actual AppraisaL Actual Appraisal Actual Appraisal Actual Estimate Estimate Estimate Estimate Estimate Bugesera Soya (kg) 5,000 n.a 6,000 n.a 7,500 4,382 9,000 4,107 10,000 1,746 Groundnuts (kg) 20,000 22,000 * 26,000 10,657 30,000 7,968 30,000 5,065 Maize (kg) 8,000 U 9,000 U 10,500 2,678 12,000 1,272 12,000 1,828 Beans (kg) 10,000 11,000 U 13,000 5,190 15,000 3,060 15,000 4,574 Vegetable Seeds (20 g packs) - * * 3,013 3,013 * 6,335 Ss*a NMgongo Soya (kg) 13,000 1,786 14,500 4,280 17,000 1,034 19,500 250 19,500 20 Groundouts (kg) 40,000 123 48,000 60,000 580 72,000 3.944 80,000 3,385 Maize (kg) 6,000 515 6,500 1,916 8,000 5,586 9,000 3,093 9,000 343 Bean (kg) 12,000 - 14,500 17,000 18,000 293 21,500 538 24,000 129 Vegetable seeds (20 g packs) * 2,859 - 4,520 3,084 - 7,124 - 7,531 F. (ii) Erosion Control and Composting 1983 1984 1985 1986 Bugesera Gisaka- Bugesera Gisaka- Bugesera Gisaka- Bugesera Gisaka- Migongo Mi19ongo Mi1gongo- ifogongo Vegetative protection (kt) 5,950 1,347 2,865 1,748 2,958 1,548 - - Contour bunds dug (k) 3,095 713 2,337 1,089 77 1,094 3,572 2,036 Planted (km) 881 856 178 822 3,604 635 3,073 1,501 Protected (cumulative ha) 1,475 9,285 4,674 10,930 7,319 13,119 18,162 16.714 Compost pits (nunbers) 1,784 30,376 1,652 - 22,752 39,857 668 29,978 Table 6: Project Results F. (iii) Evolution of Number of Forestry and Agro-forestry Plants Produced in Nurseries 1983 1984 1985 1986 1987 Bugesera Gisaka- Bugesera Gisaka- Bugesera Gisaka- Bugesera Gisaka- Bugesera Gisaka- Migongo Migongo Migong Migongo Migong 0 0 Numbers ('000) 2,769 2,210 1,814 1,450 1,476 1,504 962 1,196 974 1,120 F. (iv) Evolution of Number of Coffee Plants 1983 1984 1985 1986 1987 Bugesera Gisaka- Bugesera Gisaka- Bugesera Gisaka- Bugesera Gisaka- Bugesera Gisaka- Migongo Migongo Migong Migongo Migong o a Numbers ('000) 3,212 3,499 3,947 3,775 3,947 3,976 4,206 4,150 4,206 4,213 30 Table 7: Compliance with Credit Conditions Section Content Status of Compliance 3.01 Borrower conitment to project Compled with. implementation. 3.02 Employment of consultants and Revised. During implementation changes have experts. been agreed as follows: (1) the 2 extension specialists have been employed for 3 years instead of 5; 31) no financia advisor was deemed necessary; (iii) the o & E advisor has been employed for 4 years instead of 5; only short-term consultants have been employed in the Pilot Family Planning component; (iv) Procurement Manager and Work Supervisor for Rural Roads component have been required for 7.5 years each instead of 3 and 5 respectively; Cv% no Chief Auditor in the Studies and Monitoring Unit could be assigned to the project. 3.03 Appointment of key national staff. Complied with. 3.04 Submission of annual work plan. complied with. 3.05 Recovery of full cost of veterinary Compled with. medicines by 1986. 3.06 Recovery of full cost of seeds Efforts were made to achieve cost recovery. provided they remain attractive. This has not been possible due to high costs resulting from diseconomies of scale at project level and subsidization at national level. 3.07b08 Implementation of farming systems Complned with. research components. 3.09 Insurance of goods procured with Compoled with. credit. 3.10 Submission of PCR within 6 months Not compiled with. after closing. 4.01 Audit. CompLied with. Audit of roads component for 1989-91 were completed with delays. 31 Table 8: Use of Bank Resources A. Supervision Mission Supervision Month/Year Number of Period In Speciettzation Performance Mission Persons Field (weeks) Represented Rating status 1 5/83 3 3.0 A, E, P 2 2 7/83 1 0.5 R n.e. 3 1/84 3 2.0 E, E, k 2 3 (b) 5/84 2 1.0 E, L n.s. 4 7/84 2 1.0 E. L 2 S 11/85 3 1.5 L, E 2 6 2/86 3 1.5 L, E 2 7 9/86 2 1.5 A, E 2 8 4/87 3 0.5 A 2 8 (b) 6/87 1 0.5 A n.a. 9 12/87 2 0.5 A 2 10 10/88 2 1.0 E, A 2 11 1/90 1 1.0 A 2 ./ A: Agriculturalist. P: Population Specialist. L: Livestock Specialist. E: Economist/Financfal Analyst. R: Roads Specialist. / 1: Problem free or minor problems. 2: Moderate problems. B. Staff Inputs (staff-weeks) Fiscal Year 81 82 83 84 85 86 87 88 89 90 91 92 93 Total Up to AppraisaL 9/ n.a. 0.1 n.e. Appraisal 79.1 --*- Negotiations 5.3 0.5 5.8 Supervision 15.2 16.4 7.3 9.2 11.2 9.5 3.8 3.6 1.0 0.2 5.0 82.4 Project Copletion Report - - Total n.a. 84.5 15.7 16.4 7.3 9.2 11.2 9.5 3.8 3.6 1.0 0.2 5.0 167.3 .9 Infornnatio on time spent on project preparation prior to FY198 was not avaiable from the MIS. IBRD 15961 30jW OCTOBER 1981 RWANDA KItOMETER5 BUGESERA- Ndom GISAKA-MIGONGO PROJECT ' M E 2 Project Locations ry~øe W~r -m* Byitm N A*Tv10 NAL 4 ø AKAGERA Srhnéi * - ~ ub NATIONAL IPARK N KA. Kayonza Gitaramao KA ZENZE Rs RKIRA . ; Kibun - ~~JCT BOUNDYARIES ZuMII.%S DUPR~Er - MAIN ROAS / ~rr-TS MAIJPAlES 3 ~UGE - RiiS M1PS RGEAVI KENYAB MRECAf5 kik~ ~ LA4ES/t4c5- RWAN ý'MIUITARY RÉURVATION 2 ZONE MII ITAIRE lNATIONA,PARK, ~OUNRIES / U44IrES UPARIC NAN Ol lCTN BNDARIES / IM/TS DMUNE ZAIRE £ k30 PREFMAIt R OUNDARIES / M/PIS 2° PFECTUP30 ATANZEAN IA4 -~RTN 0tDARIEs9/F / ZAMBIA IBRD 1 '3045 OCTOBER 1981 RWANDA BUGESERA GISAKA-MIGONGO PROJECT GISAKA-MIGONGO REGION ECOLOGICAI ZONES ZONES ECOLOGIQUES 0 5 10 15 KILOMETERS 0 S 10 MILES 20 24 22 Gi 18 KIUNG 1 9 U R A -17 191N 34.* 23 ONya~uuy j4 8 4 5 Ruk 25 R U S O 2*15' 6 26 21 2 30 j 29 30 12 1 Rusumo BIR GA 27 k) 32 6 7 28- BURUNDI " REGION BOUNDARIES / /M/rVE O/E~ ECOLOGICAL 8OUNDARY/ / /M/Sf ItOW/4/W i COF FEE GROWMNG AREA / ZONt CAFICOLf UGANDA I1 SEMI-ARID AREA /lN mRIVERS / C:~R D'~A x ® COMMUNAL HEADMUARTERS /I~AU M2MMUNAL BU< ?3# Z A l R E PREECTURE HEADMUARTERS / BWfc.AU AMEFICEf ' vitora . 32 SECTORS (SEE REPORT FOR NAMES) / S$Wrf~M --- SEClOR 80UNDARIES / 115~/7f&PSIMr~u ••COMMUN E 80U~DÅARlES / I/M/T~ ~OMUNE B D ~~NT ERNAT SO 8UNDARIES / AONrIRES INWFANATONAI ES TANA NZANIA .Wa.a!Men . 05o ae e.f IBRD 15967 30 30°15' OCTOBER 1981 RWANDA BUGESERA GISAKA-MIGONGO PROJECT BUGESERA REGION ECOLOGICAL ZONES ZONES ECOLOG1QUES 2°00 2400' KILOMETERS 0 5 10 15 1 1CH ANYA -- 5510 12 15 2 1 T 9 8 ý3- K 7A N 7 \/ 6 )k 6jý S/9H:0H RAI - 1 ~Korea a 8 12 9 Rha 5 2 10 N'N a B U R,U N D l REGION BOUNDARIES /L/M1TESDE REGION AJGANJ>A CCECOLOGICAL BOUNDARY / 1M/TEr ECOLOG/OUE i COFFEE GROWING ARE A /ZONE CAFE/COLE 12 SE MI-AR|D ARE A / ZONE SEMI-AP/D£ RIVERS / COURS D'EAU 2*30'- > MILITARY ZONE / TEfRAIN MIIITAIR I&GM. RANCN / 2ANCH B.G.M. ®COMMUNAL HEADQUARTERS / BUREAU COMMUNAI S U8PREFECTURE HEADQuARTERSI/BuREAuDFSOUSECTE k,ý10 SECTORS (SEE REPOR T FOR NAMES) / SEMTURS --SECTOR BOUNDARIES / L/M/TESOE SECTEUR T N ANi A -•••COMMUNE BOUNDARIES / t/m/TESDE COMMUNE a -----INTERNATIONAL BOUNDARIES /APONr/ERE /NTrE£NA /ONAES 30°15