Somalia - Bay Region Agricultural Development Project

Document of The World Bank FOR OFFICLL USE ONLY Report No. 11874 PROJECT COMPLETION REPORT SOMALIA BAY REGION AGRICULTURAL DEVELOPMENT PROJECT (CREDIT 972-SO) MAY 14, 1993 Agriculture and Environment Operations Division Country Department II Africa Regional Office 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 1981 US$ 1 = So.Sh. 6.295 1982 US$ I = So.Sh. 1.75 1983 US$ 1 = So.Sh. 15.788 1984 US$ 1 = So.Sh. 20.018 1985 US$ I = So.Sh. 39.487 1986 US$ 1 = So.Sh. 72 1987 US$ 1 = So.Sh. 105.177 1988 US$ 1 = So.Sh. 170.453 1989 US$ I = So.Sh. 490.675 WEIGHTS AND MEASURES I hectare (ha) = 10,000 m2 I square kilometer (km2) = 100 ha I metric ton (mt) = 1,000 kg ABBREVIATIONS ADF African Development Fund AFMET Agricultural, Farm Management, Extension and Training Project ERR Economic Rate of Return GOS Government of Somalia IDA International Development Association IFAD International Fund for Agricultural Development M & E Monitoring and Evaluation MTR Mid-Term Review NMEF National Monitoring and Evaluation Facility PADU Pilot Agricultural Development Units PMU Project Management Unit SAR Staff Appraisal Report TSP Triple Superphosphate USAID United States Agency for International Development WDA Water Development Agency GOVERNMENT FISCAL YEAR January 1 - December 31  FOR OFFTCUL USE ONLY THE WORLD BANK Washington, D.C. 20433 U.S.A. Office of Director-General Operations Evaluation May 14, 1993 MEMORANDUM TO THE EXECUTIVE DIRECCTORS AND THE PRESIDENT SUBJECT: Project Completion Report on Somalia - Bay Region Aericultural Development Project (Credit 972-SO) Attached is the Project Completion Report on Somalia - Bay Region Agricultural Development Project (Credit 972-SO) prepared by the Africa Regional Office. The project aimed to increase smallholder production. It was an innovative design, with a first phase of capacity-building followed by a production phase subject to a mid- term review. At the mid-term review a better understanding of the area's natural resource and sociological constraints, together with limited success developing agricultural technology, led to reduction of the project's scope (principally dropping the main incremental farm production objectives in favor of continued capacity-building, technology development and preparation of a follow-up phase). Implementation of the revised project met targets for construction, staffing, training, the organization of agricultural research and preparation of a second phase. Civil strife then interrupted follow-up plans and, according to the PCR, may have caused the destruction of much of the project's investments. The PCR examines performance against only the modified objectives that emerged from the mid-term review. In retrospect, this was a risky investment and was designed accordingly. The risk was worth taking, but events exceeded even the level of risk foreseen. The project may be audited to explore natural resource and agricultural technology issues in semi-arid areas. Attachment This document has a restricted distribution and may be used by recipients only in the porformance of their official duties. its contents may not otherwlse be disclosed without World Bank authorization.   FOR OMlCLAL USE ONLY PROJECT COMPLETION REPORT SOMALIA BAY REGION AGRICULTURAL DEVELOPMENT PROJECT (CREDIT 972-SO) Table of Contents Preface .............................. i Evaluation Summary ................. .................. iii PART I. PROJECT REVIEW FROM THE BANK'S PERSPECTIVE .1 Project Identity .1 Background .1 Project Objectives and Description ......................... 2 Project Design and Organization ........................ 2 Project Implementation ....................... 3 Project Results ....................... 7 Project Sustainability ....................... 8 Bank Performance ....................... 8 Borrower Performance ....................... 8 Consulting Services ....................... 8 Lessons Learned ....................... 9 PART II. PROJECT REVIEW FROM BORROWER'S PERSPECTIVE .10 PART III. STATISTICAL TABLES .11 Table I Related Word Bank Loans and Credit .1 Table 2 Project Timetable .12 Table 3 Credit Disbursement .13 Table 4 Project Implementation .... ... .................. 14 Table 5 Project Costs and Financing ....................... 16 Table 6 Project Results ......... ............... 18 Table 7 Status of Covenants ......................... 20 Table 8 Use of Bank Resources ........................ 21 Map Nos. 13870,14245 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.   PROJECT COMPLETION REPORT SOMALIA BAY REGION AGRICULTURAL DEVELOPMENT PROJECT (CR.972-SO) Preface 1. This is the Project Completion Report for the Bay Region Agricultural Development Project in Somalia for which Credit 972-SO in the amount of USS 12.0 million was approved on December 27, 1979. The credit was closed on March 31, 1989, two years behind schedule. US$ 11.91 million was disbursed. The last disbursement was on December 29, 1989. US$ 0.09 million of the credit was canceled. 2. Parts I and III of the report were prepared by the Agriculture and Environment Operations Division of the Eastern Africa Department, World Bank. A draft was sent to the cofinanciers (ADF, USAID and IFAD) for comment, but none has been received. A copy could not be sent to the Borrower for comment due to the security situation in Somalia. 3. The report is based, inter alia, on the Staff Appraisal Report; the Credit Agreement; supervision reports; correspondence between the Bank and the Borrower; and internal Bank memoranda.   - iii - PROJECT COMPLETION REPORT SOMALIA BAY REGION AGRICULTURAL DEVELOPMENT PROJECT (CREDIT 972-SO) EVALUATION SUMMARY Introduction 1. An economic objective of Somalia was self-sufficiency in major food commodities through an expansion of the area under cultivation and gains in productivity. Development of the Bay Region was one of the programs to attain this objective. The region contained nearly 60% (320,000 ha) of Somalia's existing rainfed agriculture, 370,000 ha of cultivable rangeland, and was considered to have the greatest potential of all regions for a rapid increase in the area of rainfed cultivation and crop yield. The remainder of the Region (3.3 million ha) was rangeland which supported nearly 900,000 heads, or 9% of the country's livestock. The Bay Region had a long tradition of arable farming and was one of Somalia's major producers of sorghum and to a lesser extent of cowpeas and mungbeans. Project Objectives 2. The project was to lay the groundwork for increasing crop and livestock production in the Bay Region. Over a period of six years, it was to improve the potable water supply, upgrade the road network, support the development of a farming system that would preserve the productivity of the land, train Somalis and help build institutions capable of carrying on agricultural development in the Bay Region. 3. The project was conceived as the first phase of a long term development program. It was to provide the infrastructure, institutional strength and information base, funds and facilities required to prepare a second phase project that would focus more directly on increasing agricultural production. Because of the many technical and institutional questions that remained unanswered, the first phase project was to be implemented in two three-year stages. The first was to concentrate on procuring project goods, staffing, training, gathering information, extending assistance to farmers, and initiating an applied research program to find solutions to already identified farming problems. The second stage was to be triggered by a Mid-term Review (MTR) to assess the status and future direction of the project. Project Implementation 4. Initial implementation was slow due to the complicated financing arrangement involving four financiers, delays in the recruitment of technical assistance experts and the reluctance of the government agencies to transfer staff, equipment and facilities to the Project Management Unit (PMU). The unit was to work closely with the regional offices of the Ministries involved and coordinate activities with other development activities undertaken by the regional government. However, this did not work well during implementation. The unit was isolated from the rest of the regional government and was a drain on Government resources.  - iv - 5. The MTR was carried out in late 1983. The Government and financiers agreed to proceed with the second stage but at a reduced scope. While based on the original project objective, the project was redesigned to concentrate resources on trying to find ways to reduce the risks of rainfed farming; consolidating the veterinary services; providing reliable water supply; and rebuilding and maintaining the economically indispensable roads of Bay Region. Two components were dropped: the Pilot Agriculture Development Units (PADUs) and the associated range management. 6. Project cost overruns were foreseen during the first project year due to delays in effectiveness and unexpected domestic inflation. At the MTR, project costs were revised upwards by 14 percent. The increase was mainly for the water supply component, and the financing gap was to be filled by USAID. 7. The Water Supplies program encountered several problems, especially cost overrun. At the MTR, it was decided that drilling should be confined to the limestone plateau, and if further exploration drilling was successful, to the limestone depression. Additional financing was provided by USAID through an extension of the Comprehensive Groundwater Development Project. A total of 116 wells and 38 uars were completed under this component. 8. Road construction was hampered by periodic fuel shortages. Some of the design criteria, especially those of the drift crossings were faulty. Nevertheless, the project constructed 332 km of road and 62 km of track. At completion, the roads were generally in fair condition with the exception of the drift crossings. Maintenance targets were not reached by a wide margin. 9. The Veterinar component established 14 veterinary dispensaries, nine drug sales points, and carried out an average of 292,500 vaccinations annually from 1982 to 1987. 10. The Building program and Agricultural Research and Training program were completed as scheduled. The MAU was established but the monitoring and evaluation section never functioned effectively because of the early resignation of personnel. PMU's role in agricultural extension was reduced to one of liaison and coordination while the actual extension functions were transferred to another ongoing IDA-financed project, the Agricultural, Farm Management, Extension and Training Project (AFMET, Cr.905-SO). This modification was made at the MTR when the lack of coordination between the two projects was identified as an issue. 11. The component to strengthen the National Monitoring and Evaluation Facility was successfully implemented, despite d delayed start in personnel recruitment. Its work program included an initial inventory of all ongoing agricultural projects, training programs on basic M&E procedures, improving progress reporting for the Bay Region Project, and data collection on selected projects for planning, control and program evaluation purposes. Project Results 12. The project achieved most of its physical targets as redefined at MTR, and several benefits had been noted. However, at completion, the project economic rate of return was reestimated at 5 percent, compared with the appraisal estimate of 20 percent. Due to the lack of adequate data at both appraisal and completion, the estimation of project benefits is subject to a wide margin of error.  v - Sustainability 13. Sustainability was not considered an issue at appraisal. Given the lasting civil war in the country and the massive destruction that has taken place, it is likely that little remains of the project. Lessons Learned 14. The lessons learned are summarized as follows: (a) Project organization is a major issue in integrated rural development projects. This project created a separate project unit that drew together staff and resources from the participating ministries. The PMU took a while to develop into an effective institution. In the end, one could say that this arrangement enhanced implementation in the short-term but the impact on institution-building was negative. Its overstaffing and cost overruns were a drain on the Government and when the project ended, staff would have difficulty in re-integrating into the regular system; (b) There was a lack of coordination with an ongoing IDA-financed project aimed at strengthening the national agricultural extension organization. Care should be taken, particularly in the design of integrated regional projects, that they do not conflict with national systems; (c) the Mid-Term Review was an innovative feature (now widely adopted across the Bank) which resulted in a redesigned project and improved implementation. However, it failed to tackle the root of the problem, which was the economic justification for the project; (d) given the lack of knowledge about the region at preparation and appraisal, a weakness the Bank admitted to at ihe outset, project design was overly ambitious and complex. Attention of supervision staff and the PMU was spread thin over many components while some essential features, such as monitoring and evaluation and cost recovery were not given enough attention. These complexities were reinforced by the involvement of multiple donors, causing delays in project start-up and the loss of implementation momentum; (e) although this was designed as a first phase of a long-term program, greater consideration could have been given to the question of sustainability. The Staff Appraisal Report did not discuss this issue since it expected support to continue in a follow-up project.  A J  PROJECT COMPLETION REPORT SOMALIA BAY REGION AGRICULTURAL DEVELOPMENT PROJECT (CREDIT 972-SO) PART I - PROJECT REVIEW FROM BANK'S PERSPECTIVE A. Qjrect Identity 1. Project Name: Bay Region Agricultural Development Project Credit No: 972-SO RVP Unit: Africa, Country Department II Country: Somalia Sector: Agriculture B. Background 2. Somalia was heavily dependent on livestock and farming for its economic development but with the exception of commercially produced bananas and sugarcane, production tended to be subsistence oriented. At the time of project formulation, only 0.7 million ha of Somalia's total land area were cultivated-540,000 ha of this was rainfed and the balance was under controlled or uncontrolled irrigation. About 195,000 families were engaged in crop production and 90% of the cultivated area was used to grow sorghum, maize and sesame. 3. One of Somalia's economic objectives was self-sufficiency in major food commodities through an expansion of the area under cultivation and gains in productivity. The Government planned to achieve this objective through increased technical assistance to agriculture, soil conservation, better water management and irrigation intensification, and progressive transition from nomadism to settled farming. Irrigation development was intended as a shield against future weather vagaries, while range conservation and management would protect livestock against prolonged drought and land degradation. The livestock development strategy emphasized disease control measures, improvement of the marketing system and improvement of herd productivity through better range and animal management. 4. The 4 million ha of land in the Bay Region contained nearly 60% (320,000 ha) of Somalia's existing rainfed agriculture. It had a further 370,000 ha of cultivable rangeland, and was considered to have the greatest potential of all regions for a rapid increase in the area of rainfed cultivation and crop yield. The remainder of the Region (3.3 million ha) was rangeland which supported nearly 900,000 heads, or 9% of the country's livestock. The Bay Region had a long tradition of arable farming and was one of Somalia's major producers of sorghum and to a lesser extent of cowpeas and mungbeans. Average rainfall was between 400 and 500 mm. Accordingly, the Bay Region was identified as a suitable area for investment in rainfed farming and livestock production. Its natural resources and potential for dryland farming fitted the Government's priorities and provided the best opportunity for immediate action. Several production constraints in the Region were identified in the late 1970s. The lack of permanent water for human and animal use during the dry seasons was a major problem. Due to unsatisfactory road condition, substantial quantities of surplus grain were not finding their way to the urban centers and were wasted. The practice of shifting cultivation coupled with a complete dichotomy between crop and animal production needed to be replaced by a more  - 2 - permanent, sedentary farming system. Finally, the region lacked the qualified manpower needed for its development. C. Project Objectives and Description 5. PrQiect Objectives. The project was to lay the groundwork for increasing crop and livestock production in the Bay Region by addressing the above deficiencies. Over a period of six years, it was to improve the potable water supply, upgrade the road network, support the development of a farming system that would preserve the productivity of the land, train Somalis and help build institutions capable of carrying on agricultural development in the Bay Region. 6. Proiect components included: (1) Improving the supply of potable water in the Region for human and livestock consumption. (2) Upgrading and maintaining all roads in the Bay Region to give all-season access to areas where agricultural production was presently or potentially significant. (3) Intensification of the Farming System through: (i) introducing simple improved production practices and initiating a program of adaptive research and field demonstration trials aimed at crop production, animal husbandry and range management; (ii) creating four Pilot Agricultural Development Units (PADU) to integrate crop and livestock activities and test the technical and institutional means of integrating farm and range activities; (iii) strengthening the veterinary service in the region; (iv) strengthening the agricultural extension service in the region; (v) producing quality seeds; (vi) testing means to improve on-farm storage of grains and fodder; and (vii) supporting the production and diffusion of improved farm tools and animal drawn implements. (4) Strengthening National Monitoring and Evaluation Facility (NMEF). (5) Providing technical assistance and training. D. Project Design and Organization 7. The project was conceived as the first phase of a long term development program. It was to provide the infrastructure, institutional strength and information base, funds and facilities required to prepare a second phase project that would focus more directly on increasing agricultural production. Because of the many technical and institut.onal questions that remained unanswered, the first phase project was to be implemented in two three-year stages. The first was to concentrate on procuring project goods, staffing, training, gathering information, extending assistance to farmers, and initiating an applied research program to find solutions to already identified farming problems. At the end of this first stage, the Government of Somalia (GOS) was to carry out a mid-term review (MTR). Disbursements of credit funds following this review were contingent upon agreement between IDA and GOS to continue with the implementation of each component as originally defined or as modified during the MTR. This was an innovative feature which has since become widely used in Bank- financed projects.  8. The project was the first integrated agricultural development project in Somalia. At that time experience with agricultural projects in the country was limited and the appraisal team necessarily had to make many assumptions as to the desirable project scope, the country's implementation capacity and the physical and managerial difficulties inherent in executing a complex project. Since the project involved several ministries, it was decided that a Project Management Unit (PMU) be established at the regional center of Baidoa to implement the project on behalf of the Ministry of Agriculture. It was to work closely with the regional units of the Ministries involved and coordinate activities with other development activities undertaken by the regional government. The risk that this would not work -- due to limited cooperation of the concerned agencies in seconding personnel and resources to the PMU - was recognized at appraisal. Such difficulties did arise during project implementation, particularly in the initial stages. Towards the end, the PMU was operating more smoothly but several issues remained. These included: cost overruns for the PMU because of overstaffing and salary supplements, the failure to involve local government agencies and communities, and the loss of built up institutional capacity if the unit were to disband when the project ended. In the short run, such a unit was beneficial to project implementation but it also had a negative impact on institution-building. The unit was isolated from the rest of the regional government and was a drain on GOS resources. Other organizational models had been tried out in rural development projects. In some, the implementation responsibility was given to local branches of ministries. The result was a lack of coordination with other ministries which led to confusion and inertia. It appears that in order for such a structure to work, it is preferable to have a strong decentralized system of government which empowers local authorities to coordinate and mobilize ministerial staff in the region. This model is being used in ongoing rural development projects in Kenya. Their outcome remains to be seen. 9. Another shortcoming in design was the lack of coordination with the Agricultural, Farm Management, Extension and Training Project (AFMET, Cr. 905-SO), the first agricultural extension project in the country. Its preparation and appraisal also coincided with the Bay Region project. However, no clear guidelines were established at appraisal concerning the detailed, day-to-day control of the agricultural extension service in the Bay Region under the two projects. Such precise definition and linkage was important since the extension project was expected to take to farmers the findings of the research station in Bonka, the development of which was part of the Bay Region project. This problem was identified at the MTR, and the project was redesigned to improve coordination and linkage with the AFMET project. E. Proiect Implementation 10. The project had a difficult start due to the complicated financing arrangement. Four donors sponsored the project. IDA and IFAD were on a joint basis; the African Development Fund financed on a parallel basis the road component while USAID covered the costs of the greater part of the well drilling program, parts of the operation of the research station and the cost of consultants for agricultural research in Bonka and much of the training. Procurement of goods and services for the USAID financed portion was tied to U.S. suppliers. GOS was expected to contribute the equivalent of US$ 5.6 million, or about 12% of total project costs. One of the conditions of effectiveness of the IDA/IFAD contribution was that all other co-financiers' contribution had to be secured. Consequently, although the project was appraised in 1978, it did not come into effect until September 12, 1980. 11. After the project became effective, the initial implementation was slow due to the combined effects of delays in the recruitment of USAID financed technical assistance experts and the reluctance of the Water Development Agency (WDA), Ministry of Public Works and National Range Agency, to  - 4 - transfer Bay Region staff, equipment and facilities to the PMU. Until the end of 1981, the PMU had not been organized, staffed and equipped. 12. Mid-Term Review and Project Redesign. The Government and cofinanciers carried out a mid-term Review in late 1983. The review team found that sufficient efforts had not yet been made during the last three years to improve agricultural production in the Bay Region over the long term. The funds spent up till the review, about USS 22 million, had been for equipment (water supply, roads and the PMU). Implementation of the agricultural improvement schemes was confined to improvement of veterinary services, which had made modest but satisfactory progress. A natural resources survey and mapping and a comprehensive socio-economic study carried out in the first stage also provided new information that made the original project assumptions on agricultural production questionable (see para. 14). 13. Despite the shortcomings, it was agreed that the project had indeed filled a void caused by insufficient infrastructure and less than adequate production support services in the region. The Government and cofinanciers agreed to proceed with the second stage but at a reduced scope. While based on the original project objectives, the project was redesigned to concentrate resources on: (a) trying to find ways to reduce the risks of rainfed crop farming in the region by means of operationally oriented agricultural research and trials; (b) consolidating the veterinary services; (c) providing reliable water supply where the population density was greatest and most vulnerable to frequent droughts; and (d) rebuilding and maintaining the economically indispensable roads of Bay Region without adding significantly to the existing road network. 14. Two components were dropped: the PADUs and the associated range managemert. The absence of large tracts of land (indicated in the survey findings), plus the difficulties inherent in changing local grazing methods and customs, precluded the allocation of a sizeable range area to any one group of farmers to establish four PADUs in which to test integrated range and crop production management trials. PMU's role in agricultural extension was also reduced, leaving most of the functions to AFMET. 15. Progress of Each Component. The water suplies program encountered problems due to the late arrival of equipment and the absence of identified aquifers and the high incidence of boreholes that were unproductive or had poor quality water. The cost per well was also much higher than expected. At the mid-term review, it was decided that drilling should be confined to the limestone plateau, and if further exploratory drilling was successful, to the limestone depression. Additional financing was provided by USAID under the Comprehensive Groundwater Development Project. The unit cost was to be reduced by using light weight equipment. At the end of 1987, a total of 116 wells were drilled. The idea of providing windmill pumps proved unsuccessful: based on experience elsewhere in Somalia and in other developing countries, it was likely that the problems with windmills would continue unless an entirely new approach was taken towards their maintenance and operation. Uars 1/ were to be an alternative source of water, either where groundwater was not available, or where a uar would provide a cheaper and/or more reliable supply. A total of 38 uars were built and some of them were equipped with hand pumps. 1/ Excavated depressions used to store surface water run-off.  16. BQ; construction encountered several problems. Progress was hampered by periodic fuel shortages. Some of the design criteria, especially those of the drift crossings were faulty. In fact, given the relatively flat topography of the project area, it was open to question whether the best way of taking rainwater across the road was indeed by drifts and not by some other means. Nevertheless, the project constructed an extensive road and motor track network. A total of 332 km of roads and about 62 km of tracks were built against a revised target of about 300 km of roads and 160 km of tracks for the entire project period. Road maintenance, however, was not implemented as well. At completion, the roads built by the project were generally in fair condition with the exception of the drift crossings, but maintenance targets were not reached by a wide margin. For instance, the 1987 achievement was only 102 km while until the middle of 1988 maintenance was carried out only on 26 km. The target at appraisal was to have about 560 km. of gravel roads under regular mnaintenance by the end of the project. This was revised to 345 km at the mid-term review. 17. The Building program was completed as scheduled. The only snag was that in designing the foundation of the office block on the Bonka Research Station and in one of the veterinary clinics, insufficient allowance was made by the consultants for the demands of foundations on shrinking clay. This was remedied at the consultants' expense. At completion the program constructed the following: project headquarters and staff housing (23 houses and two office blocks in Baidoa and six houses in Mogadishu for the NMEF); facilities for the research station in Bonka and for the seed farm; 14 veterinary dispensaries, workshops, and garages for equipment. i8. After initial difficulties, the PMU was organized and functioned as planned, with a few notable exceptions: (i) the monitoring and evaluation section was not really established because of the early resignation and departure of the personnel; (ii) agricultural extension became the function of AFMET upon its establishment and only a liaison and coordinating role was retained for the PMU. The PMU was divided between Mogadishu, where it was necessary to maintain a liaison office for administrative purposes, and Baidoa, where project activities were controlled and supervised. Despite salary supplements, the project had problems retaining staff because of difficult conditions in the project area. 19. The progress of the Veterinary Component was good. The program established 14 veterinary dispensaries, established nine veterinary drug sales points and took over the administration of marketing facilities in two centers. Three vaccination teams carried out the following number of average annual vaccinations from 1982 to 1987: Rinderpest 80,000 Blackquarter 120,000 Anthrax 21,000 Haemorrhagic Septicaemia 12,000 Contagious Bovine Lumpy skin disease 55,000 Pleuropneumonia 4,500 Total 292,500 Annually an average of about 80,000 cattle, 131,000 camels and 95,000 sheep and goats were treated against trypanosomiasis, endo- and ecto-parasites and other diseases. 20. However, there were several issues concerning the effectiveness and sustainability of this component. First, the 'cold-chain' facilities, which ensured that vaccines were stored at the proper temperature until used, did not function perfectly. This cast some doubts on the effectiveness of some of the vaccinations. Another question was whether the extent of the vaccinations was sufficient to  - 6 - ensure protection of the entire population. Furthermore, as the diagnostic laboratory had been cut out of the project at the MTR, the staff in the field were able only to respond to, rather than anticipate, disease outbreaks. Finally, in determining the benefits and sustainability of the treatments, the question must be asked about the training and competence of the lower-grade staff. There appeared to be at least the possibility that it was the owners who approached them, with their own diagnosis of the nature of the disease, rather than the staff determining, by examination, what the problems were. The complementary activities to promote improved livestock production were not developed. 21. Agricultural Research and Training. An overseas university fielded a research team to pursue an adaptive research program and to train their Somali counterparts. The disciplines provided were: (i) plant breeding, with particular attention to sorghum, mungbeans and cowpeas; (ii) agronomic investigation into the feasibility of alternative crops; (iii) soil fertility; (iv) pest control; (v) farming systems studies. The university also provided an Operations Manager for the research station and the seed farm, and a Research Adviser. Under IDA financing the team was from time to time augmented by expertise to investigate animal traction, water conservation and retention studies, and a soil scientist. 22. The Bonka Research Station undertook a broad range of research trials aimed at understanding the principal constraints to crop production in the Region. In a detailed farmers survey undertaken in the 1987 Gu season, farmers identified sorghum stalkborer and ticks as the major constraints to increased crop and livestock production. Research priorities were also given to improving the level of soil moisture and fertility and overcoming labor constraints in weeding. 23. The results of crop trials were mixed. The efforts on sorghum development were unsuccessful. At appraisal the expectation was that a sorghum variety which would out yield the local variety and also assure greater safety in adverse years would be quickly identified. However, after much time and effort, it was determined that the local variety was very homogenous and therefore selection was unlikely to yield results. The conclusion was that any varietal improvement would necessarily come only after long-term efforts of introducing other desirable traits into the local germplasm, without losing any of the existing ones. Attempts to introduce improved pulses varieties were successful. A new mungbean introduced from Taiwan was reported to be increasingly accepted by growers, although, as shown by the consultant's survey, mungbeans did not yet appear to be extensively traded. Similar achievements were expected with cowpeas. Adaptive research on other crops did not yield very positive conclusions. 24. Regarding soil fertility, there was conclusive evidence that the application of triple superphosphate (TSP) gives considerable yield increases of all crops. The effects of TSP were for the first time extensively demonstrated to farmers in the Gu of 1988. Its widespread application would depend on availability, the creation of an effective distribution organization and on the effectiveness of agricultural extension. The effects of pest control were demonstrated and progress was made on controlling storage pests, which was a significant step in the preservation and improved marketability of grains. 25. At completion, most of the training programs planned for research workers were completed and all trainees were due to resume their posts. Further expatriate technical assistance that would be desirable for the formulation of research program in some fields and also their effective management was to be provided under the second AFMET project (Cr. 1794-SO). The designers of that project foresaw the need for external research support for the Bonka station after the termination of the  contract with USAID and appropriate provisions were made to bridge the gap. The AFMET n project included financing for consultants until October 1992. 26. The MAE unit was established in 1983 but never functioned effectively because of the early resignation of the personnel, who were not satisfied with the working conditions and career prospects that the project offered. This failure to assemble key production indicators made project impact and economic reevaluation difficult. 27. The component to strengthen the National Monitoring and Evaluation Facility was successfully implemented, despite a delayed start in personnel recruitment. It was headed by a Somali team leader, supported by four expatriate M&E specialists and other local staff. They occupied new offices built under the project. Its work program included an initial inventory of all ongoing agricultural projects, training programs on basic M&E procedures, improving progress reporting for the Bay Region Project, and data collection on selected projects for planning, control and program evaluation purposes. 28. Project Costs. Cost overruns were foreseen during the first project year due to delays in effectiveness and higher than expected domestic inflation of 80 percent that year. At the MTR, project costs were revised upwards by 14 percent. The increase was mainly for the water supply component, and the financing gap was to be filled by USAID through an extension of its Comprehensive Groundwater Development Project. However, due to the civil war in the country, actual costs could not be obtained. F. ProiecLReslti 29. The project achieved most of its physical targets as redefined at the MTR. It constructed 332 km of roads, completed 38 uars and 116 wells, of which 74 found exploitable amounts of water. All the buildings were completed. The PMU and the research station were staffed and the research organization effectively organized. Training of research workers, road engineers and technicians, and veterinarians, was completed as planned. 30. At the time of preparation of the completion report, security condition in Somalia prevented the Bank from visiting the country to assess the economic rate of return (ERR). An evaluation carried out by a team of consultants hired under the project in 1988 reestimated the ERR at around 5 percent, compared with the appraisal estimate of 20 percent. The consultant report pointed out that although annual plantings of sorghum in the Bay Region rose from 193,500 ha in 1982 to just over 300,000 ha in 1987, much of the increase could not be attributed to the project's interventions. The liberalization of the market, giving farmers the freedom to sell to others than the Agricultural Development Corporation, had a greater effect. The consultant carried out two calculations of the ERR: (a) if 35 percent of the area expansion was attributed to the project, the ERR would be 2 percent; (b) if 60 percent was attributed to the project, the ERR would be 5 percent. In both cases, the returns to the project were unsatisfactory. The consultant cited a number of factors for the discrepancy from the appraisal estimate: increases in on-farm operating costs, project investment costs and a reduction in the price of commodities relative to manufactured goods and services since appraisal. Also, with the increase in knowledge about the project area, some of the assumptions used were different from those at appraisal. Because of the failure of the M&E unit to produce data, the consultants had to conduct their own survey of villages in the project area. Their report warned that "even those benefits which it has been possible to estimate are subject to a wide margin of error."  - 8 - 31. Despite the difficulties in quantifying project results, several benefits had been noted. For examnple, prior to water supply development, farmers often left their villages in the dry season. This resulted in late planting and reduction In area planted as well as yields. The availability of water resulted in timely planting and an increase in the area planted. The road network facilitated transportation and marketing of cereals. G. Project Sustainability 32. The SAR stated that at the end of the project period, continued Government support would be necessary since the project was only the first phase in a long-term development effort in the Region. Sustainability was not considered an issue at the time. Nonetheless, the appraisal mission did require that cost recovery from boreholes provided by the project be carried out in accordance with a study on water charges. There is no indication in the Bank's files that the study was done, nor is there mention of cost recovery. At the end of the project, it was envisaged that unless external financing continued, project works would not be maintained and the supply of veterinary drugs and other imported inputs would be interrupted. The outbreak of the civil war further eroded project benefits. 33. In 1988, GOS hired consultants to prepare the second phase project. The proposal turned out to be similar to the completed project. The projected ERR was 12 percent although there was no high value crop that could carry the investment cost. While discussions were still going on about the second project, the civil war broke out. Given the massive destruction that has taken place, it is likely that little remains of the project. H. Bank Performance 34. The project was prepared and appraised without adequate information on all the important socio-economic aspects. During the project period, a regional study and socio-economic study were carried out, with results leading to the elimination and modification of some of the project components. In retrospect, these studies should have been done during project preparation, especially when the project's success hinged on acceptance by the rural population, and not only on the technical nature of the project components. 35. Most of the supervision missions were helpful and constructive, and working relations were satisfactory. An average of 20 staff weeks per year of supervision were allocated to the project (1: 1 ratio compared to preparation/appraisal), which is quite reasonable. However, because of the complexity of the project, the attention of supervision staff was spread over many components, with the result that some essential features, such as monitoring and evaluation and cost recovery were neglected. Although supervision missions had consistently recognized that the project's development impact would be marginal, nothing was done to address this issue. It was also raised at the MTR, but no recalculation of the ERR was done. 1. Borrower Performance 36. Tle project was a complex development effort involving six ministries and four financiers. Difficulties in coordinating the project activities and the slow transfer of personnel from various government agencies to the project caused delay in project implementation. Furthermore, it was unfortunate that the Somali Petroleum Agency refused permission for the project to import fuel directly despite representations by the project General Manager, the MOA and the donors.  - 9 - Consequently fuel had to be purchased with local currency, causing constraints to the road and water components. J. Consulting Services 37. The performance of research consultants was satisfactory. However, the technical competence of some consultants charged with physical implementation appeared rather weak. For example, the first group charged with the design of the buildings failed to ensure adequate site preparation on shrinking clay subsoils. The drift crossings had to be rebuilt to a better design. The consultants charged with the design and supervision of uar construction had either made design mistakes or did not exercise sufficient supervision of the implementing contractors. K. Lni Learned 38. The lessons learned are summarized as follows: (a) Project organization is a major issue in integrated rural development projects. This project created a separate project unit that drew together staff and resources from the participating ministries. The PMU took a while to develop into an effective institution. In the end, one could say that this arrangement enhanced implementation in the short-term but the impact on institution-building was negative. Its overstaffing and cost overruns were a drain on the Government and when the project ended, staff would have difficulty in re-integrating into the regular system; (b) There was a lack of coordination with an ongoing IDA-financed project aimed at strengthening the national agricultural extension organization. Care should be taken, particularly in the design of integrated regional projects, that they do not conflict with national systems; (c) the Mid-Term Review was an innovative feature (now widely adopted across the Bank) which resulted in a redesigned project and improved implementation. However, it failed to tackle the root of the problem, which was the economic justification for the project; (d) given the lack of knowledge about the region at preparation and appraisal, a weakness the Bank admitted to at the outset, project design was overly ambitious and complex. Attention of supervision staff and the PMU was spread thin over many components while some essential features, such as monitoring and evaluation and cost recovery were not given enough attention. These complexities were reinforced by the involvement of multiple donors, causing delays in project start-up and the loss of implementation momentum; (e) although this was designed as a first phase of a long-term program, greater consideration could have been given to the question of sustainability. The SAR did not discuss this issue since it expected support to continue in a follow-up project.  - to - PART Il: PROJECT REVIEW FROM BORROWER'S PERSPECTIVE Due to the civil war in Somalia, Parts I and III cannot be sent to the Borrower for comment.  - I1 - PART III STATISTICAL INFORMATION Table 1. Related Bank Loans and/or Credits Loan/Ciedit Title Purpose Year of Status Approval Agricultural Extension To establish a solid basis for the 1979 Completed and Farm Management progressive build up of operational Training Project (Cr.905; and efficient extension and farm US$ 10.5 million) management in Somalia. Central Rangelands To consolidate and improve 1979 Completed Development Project rangeland and livestock production (Cr.906; US$ 8 million) in the Central Rangelands; improve pastoralist' incomes; and encourage the gradual concentration of pastoral communities. Northwest Region To increase farm incomes by 1976 Completed Agricultural Development addressing the crucial issues of sever Project soil erosion, inadequate extension (Cr.635; USS 10 million) service, poor agricultural practices, poor seed quality, and inadequate and uneven distribution of water facilities for both human and livestock. Second Northwest Region To increase income of rainfed 1985 Suspended due Agricultural Development farmers; increase fruit and vegetable to outbreak of Project (Cr.1538; US$ production; and strengthen the local civil war. 10.6 million) institutional capacity . Semi-Mechanized Rainfed To develop an improved rainfed 1987 Suspended due Agricultural Pilot Project agricultural technology for Somalia's to outbreak of (Cr. 1774; US$ 13.4 semi-arid zones and put in place a civil war. million) technical and organizational system as part of a large strategy which includes irrigation, livestock, health and range development project.  - 12- Table 2. Project Time Table Time Date Planned DateRevised Date Actual Identification - Nov./Dec. 1976 (Project Brief) Preparation - - June 1977 Preappraisal - - Dec. 1977 Appraisal Mission July 1977 - June/July 1978 Post-Appraisal - - Dec.1978 Mission Credit Negotiation - - Nov. 6-9, 1979 Board Approval - - Dec. 27, 1979 Credit Signature - - Mar. 26, 1980 Credit - Sept.12, 1980 Effectiveness Project Completion March 1986 Sept. 1987 Sept. 1988 Credit Closing Mar. 31, 1987 Mar. 31,1988 March 31, 1989  - 13 - Table 3. Credit Disbursement (S Million) Appraisal Estimated Actual as % of IDA Fiscal Ye Cumulative Actual Cumulative Estimated 1981 0.4 0 0 1982 1.74 0.13 7.5 1983 5.46 1.04 19 1984 7.86 1.77 23 1985 9.54 2.54 27 1986 11.22 4.26 38 1987 12 7.01 58 1988 8.96 75 1989 10.20 85 1990 11.53 96 1991 11.91 99  - 14 - Table 4. Project Implementation Indicators Origyinal Actua Water Supply Vehicles & Equipments Rigs 3 2 Trucks 4 Compressor 2 Concrete Mixer I Water Tanker 2 Fuel Tanker 1 4WD Pick-up 7 4 4WD Station Wagon I Pump 18 Bulldozer 1 Trippers 2 Flat Bed 3 Generators 6 Tech. Asst. Hydrogeologist 1 2 Water Engineer I I Drilling Superintendent I Chief Driller 2 2 Chief Mechanic I I Chemist I Access Roads Vehicles & Equipments Trucks 32 5 4Wd Pick-up 10 5 4Wd Station Wagon 2 3 Trailers 10 1 Bull Dozer 6 Grader 8 3 Wheeled Loaders 5 3 Rollers 9 3 Water Tanker 3 Mobile Workshop 2 Compressors 2 Generator 2 Tech. Asst. Senior Highway Engineer I I Mechanical Engineer I I Plant Superintendent 1 I Mechanical Superintendent I I  - 15 - Agr. Research Tech. Asst. Agr. Research Officer 1 1 Extension Specialist 1 1 Range Management Officer I Senior Agriculturalist 1 Agriculturalists 3 Technicians 5 Assistants 4 Librarian 1 Meteorologist I Seed Manager I I Plant Protection Officer 4 Vehicles & Equipments Pick-up 14 Truck 2 Tractor 7 Cultivator 1 Planter I Ridger I Cult 4 Air Seeder I Off Set Disc 1 Leader 1 Combine Header I Red Weeder 1 Veterinary Service Tech.Asst. Vet. Officer I I Other Staff 71 84 Dispensaries 13 14 Vehicle & Equipment Lorry I Motorcycle 4 Pick-up 7 Spray Race Engine 2 PMU Tech. Asst. Project Director I I Tech. Manager I I Financial Controller I I Buildings (for all components) $4.2 million Houses 29 Office Blocks 2  - 16 - Table 5. Project Cost and Financing A. Proeect Costs (US$ million) Project Component Appraisal Estimate Revised 1/ Local Foreign Total Water Supply 1.4 3.5 4.9 22.6 Access Roads 2.2 7.0 9.2 11.7 Adaptive Research 0.4 0.5 0.9 5.2 Seed Farm 0.6 0.9 1.5 Range Service 0.9 0.7 1.6 0.1 Veterinary Service 0.9 1. 2.0 1.7 Tech. Asst. and Trainining 0.6 10.9 11.5 PMU 0.7 0.6 1.3 5.6 National Monitoring Evaluation Unit 0.3 0.6 0.9 2.0 Unallocated 2.1 Physical Contingencies 0.4 1.4 1.8 Price Contingencies 2.2 7.2 9.4 Total Project Costs 10.6 34.4 45.0 51.0 1/ Project costs were revised at the mid-term review, but the categories used were different from those at appraisal. Adaptive research and seed farm were combined into one category, and technical assistance and training were incorporated into the other components. Actual project costs are not available in the files. Due to the security situation in Somalia, a PCR mission cannot go to the field to collect data.  - 17 - B. Project Financing (US$ million) Sources Planned Revised Final IDA 1. Buildings 2.5 2.5 2.8 2. Vehicles and equipment 0.2 1.7 1.4 3. Local salaries 1.1 0.4 0.4 4. Materials and supplies 3.3 1.8 2.4 5. Tech.asst., training and cons.sers. 3.6 4.8 4.9 6. Unallocated 1.3 0.8 Total 12.0 12.0 11.9 IFAD 8.0 7.4 7.7 USAID 10.5 18.3 10.7 ADF 8.9 8.2 8.6 GOS 5.6 5.1 n.a. TOTAL 45.0 51.0  - 18 - Table 6. Project Results A. Direct Benefits Indicators Planned Revised Actual |Water Supply Wells 110 116 Uars 4 38 Roads (kim) 558 300 332 Tracks (klm) 250 160 62 Veterinary Annual Vaccinations 292,500 (1982-1987) No.of Livestock Treated Per Annum Cattle (head) 80,000 Camels (head) 131,000 Sheep & Goats(head) 95,000 Veterinary Dispensaries 13 14 Buildings $4.2 million Houses 29 Office Blocks ___ _2 No. of Somali Staff Trained Abroad 30 49 B. Sorifhum Production in the Bay Begion* Year Area ('000 ha) Production ('000 tons) Yield (kg/ha) 1982 193.5 84.2 435 1983 106.0 25.0 236 1984 199.5 80.2 402 1985 150.6 71.4 474 186 220.4 151.5 687 1987 300.1 165.8 552 * Source: National Food Early Warning System of the Ministry of Agriculture.  - 19 - C. Economc Impat Appraisal Estimate Restimated at Completion Economic Rate of Return 20 percent 5 percent D. Studies Title Purpose Impact Farming Systems Study: To collect comprehensive Provided valuable Questionnaire for Semi data on Bay Region farmers' information for identifying Yearly Interviews production activities. development in the region.  - 20 - Table 7. Status of Covenan Section/Agreement Status of Compliance 4.01(b) GOS to ensure coordination of all development Little cooperation from Ministries activities in Bay Region. until mid-1983. 4.02 Establish and maintain a PMU. Compliance, with delay in staffing and equipping it. Monitoring and evaluation unit not functional. 4.03 Establish and Maintain a Bay Region Development Committee established, but Coordination Committee. meets infrequently. 4.04 GOS to strengthen the Monitoring and Evaluation Compliance. Recruitment delayed. capability of the Ministry of National Planning. 4.05(a) To employ construction and procurement consultants Contract signed one year late. by 4/30/1980. 4.05(b) Employment of a hydrologist, a research Composition of team altered due to and agronomist, and a livestock specialist to prepare changes in project design. Agreed 4.12 work plan for phase n by 2/28/1983. date was not met. 4.07(a) GOS to submit plans, specifications, reports, Compliance, with dc'ay. contract documents and work and procurement schedules for the project. 4.07(b) To maintain records and procedures for A fixed assets register was not and Monitoring and Evaluation, to submit maintained. 4.07(c) progress reports. 4.09 To establish So.Sh. 5 million revolving fund. Compliance. 4.10 Ministry of Naticnal Planning to prepare, Postponed. Satisfactory work program not later than 2/28/1981, a three-year was prepared and approved in work plan for NMEF. 12/1982. 4.11 GOS to submit for IDA approval, an overseas Date was extended until 3/31/1981. training program by 6/30/1980. Complied on 3/31/1983. 4.13 Appointment of PMU and other technical Compliance, with delay. assistance staff.  Table 8. Use of Bank Resources A. Staff Inputs (Weeks) Stage of Project Cycle 1977 1978 1979 18 1981 1982 1983 1984 98 1986 1987 1988 1989 Total Preappraisal 83.2 22.4 105.6 Appraisal 69.2 8.3 77.5 Negotiation 2.7 2.7 Supervision 0.9 18.9 12.8 26.5 37.3 26.4 18.2 14.3 11.1 6.2 172.6 Total 83.2 22.4 69.2 11.9 18.9 12.8 26.5 37.3 26.4 18.2 14.3 11.1 6.2 358.4  B. Missions Se of rsiect MonthYlear Number of Persons Speialization Perfoman Types of Problems c/ Cycle Reresente a/ RatinLy Status b/ Idaificaton 11/1976 3 AE, A, GS Pe-Appraisal 12/1977 1 E Appasa 06/1978 6 AE, H, FA, LS, CE, E Supervision 1 07/1980 3 AE, A, E I M Supavision2 03/1981 4 AE, A, E, MES I M Supevision3 11/1981 2 AE, A 2 M, F, O Supervision4 06/1982 1 AE 2 M, F, O SupervisinS 12/1982 2 AE, A 3 M, F, T, O Supervision 6 06/1983 2 E, E Supwviso7 08/1983 3 AE, A, CE 3 M, F, T, O Supervision8 04/1984 5 AE, A, IS, FA 3 M, F, T Supervision 9 12/1984 1 E Supervision 10 02/1985 3 AE, A, CE 3 M, T, F Supervision 11 06/1985 2 AE, IE 3 M, T Supervision 12 02/1986 1 AG 3 M,T,O  Stage of Project Month/Year Number of Persons Specialization Performance Types of Problems c/ Cycle Represented a/ Rating Status b/ Supervision 13 10/1986 1 A 2 M, I Supervision 14 04/1987 1 A 2 M, I Supervision 15 12/1987 3 A, A, E 2 M, I Supervision 16 05/1988 2 E, H 2 M, I Supervision 17 01/1989 2 E, A 2 M, I a/ A=Agronomist; AE=Agricultural Economist; AG=Agriculturalist, CE= Civil Engineer; E=Economist; FA=Financial Analyst; GS=Ground Water Specialist; H=Hydrologist; IE=Irrigation Engineer; LS=Livestock Specialist; MES=Monitoring and Evaluation Specialist. b/ 1 =No Significant Problem; 2= Moderate Problems; 3=Major Problems. c/ F=Financial; M=Managerial; T=Technical; I=Impact; O=Other  I  IBRD 13870 AUGUST 1070 r- \ rEOPLE'S DEMOCRATIC 4 REPUBLIC OF YEMEN DJ OU TI e Adodo Bososo + Moit 1+ ( / \ Berberoa,- | J rigovo \ d 7t; _ -q ( ~~~~~~~~~~~~~~~~~~~~~Loz Douc <; J1 >\Nobcd,d UfoS \ t )/ Tug Weoh Bur \o Gorl!i teee l e,sh0 Bender Bedle SOMALIA 70. BAY REGION AGRICULTURAL DEVELOPMENt PQ- 8- PRtOJECTLOCATION PROJ(CT LOCATION BITUMINODUS SRFACED ROADS GRAVELIEARTH SURFACED ROADS t MAJOR PORTS ' yo 4 DOMESTIC AIRFIELDS NTERNIATIONAL AIRFIELOS RIVERS . DISTRICT 8OUNDARIES / REGION BOUNDARIES A - - INTERNATIONAL SUNOAMIES 0 _ . > erX a X~~~~~~~~~~~~~s Mareb \ - < ~~~~~Beletwe'n; jEl Bur > _j ~~~Hodu / o 4.- ~~4" ~ ~~ 7 /'', W a lo /\ / Gorvne; Burt ( _ _ 0 50 100 150 200 250 300 j \ VA ,)sdA | F 7,,Bo-k- KILOMETEIt5 j t | O NS')'1 ;/ .' i a ~ ~~~Jowhor \\; V s gt GADIS~~~DIHU Brovo \ e Xi zi s i r ~~~~~~~~ ~~~~~~~IND,A'^A OCEAN 44' 48i   IBRD 14245 BAY REGION AGRICULTURAL DEVELOPMENT PROJECT ROADS - PHYSIOGRAPHY fORMER EXISTING PROJECT IBRD PROJECT .... =CLASS I ROADS CLASS It ROADS CLASS III ROADS LIMESTONE PLATEAU LIMESTONE DEPRESSION A OLR IN BASEMENT COMPLEX CULTIVABLE LAND IDENTIFIED PROBABLE CULTIVABLE LAND W MAJOR TOWNS o OTHER TOWNS ESCARPMENT DISTRICT BOUNDARIES \ =oe REGION BOUNDARIES - --INTERNATIONAL BOUNDARIES KILOMETERS 9 510 0 MILES P .00 AUGUST 1979