Uganda - Southwest Region Agricultural Rehabilitation Project

Document of The World Bank FOR OFFICIAL USE ONLY Report No. 16358 IMPLEMENTATION COMPLETION REPORT REPUBLIC OF UGANDA SOUTH WEST REGION AGRICULTURAL REHABILITATION PROJECT (CREDIT 1869-UG) March 4, 1997 Agriculture Operations Eastern and Southern Africa Africa Region This document has a restricted distribution and may be used by recipients only in the perfonnance of their official duties. Its contents may not otherwise be disclosed without World Bank authorization.  CURRENCY EQUIVALENTS At Appraisal: USh 60 = US$ I At Completion: USh 1,000 = US$ I ABBREVIATIONS AND ACRONYMS AEP Agricultural Extension Project ARTP Agricultural Research and Training Project ASIP Agricultural Sector Investment Program CP FAO/World Bank Cooperative Program CS Credit Scheme DFI District Farm Institute ERR Economic Rate of Return GDP Gross Domestic Product GOU Government of Uganda ICR Implementation Completion Report IDA International Development Association IFAD International Fund for Agricultural Development MAAIF Ministry of Agriculture, Animal Industry and Fisheries M&E Monitoring and Evaluation MOF Ministry of Finance MOLG Ministry of Local Government MTR Mid Term Review NARO National Agricultural Research Organization PCU Project Coordination Unit PY Project Year RFl Rural Financial Intermediary, RM Resident Mission SAR Staff Appraisal Report SOF Special Operation Facility SWUADA South West Uganda Agricultural Development Agency SWRARP South West Region Agricultural Rehabilitation Project TA Technical Assistance UCB Uganda Commercial Bank FISCAL YEAR OF BORROWER Government of Uganda I July - 30 June Vice President C. Madavo Country Director J. Adams Technical Manager S. Ganguly Task Team Leader T. Sharif  FOR OFFICIAL USE ONLY IMPLEMENTATION COMPLETION REPORT REPUBLIC OF UGANDA SOUTH WEST REGION AGRICULTURAL REHABILITATION PROJECT (Cr. 1869-UG) CONTENTS PREFACE .............................................. EVALUATION SUMMARY ............................................ ii Introduction ............................................ ii Project Objectives ............................................ ii Implementation Experience and Results ............................................ iii Future Operations and Key Lessons Learned ............................................ vi PART I: PROJECT IMPLEMENTATION ASSESSMENT .............................................1 A. STATEMENT/EVALUATION OF OBJECTIVES .............................................1 B. ACHIEVEMENT OF OBJECTIVES .............................................2 C. MAJOR FACTORS AFFECTING THE PROJECT ........6....................................6 D. PROJECT SUSTAINABILITY ............................................7 E. BANK PERFORMANCE ....8........................................8 F. BORROWER PERFORMANCE ............................................9 G. ASSESSMENT OF OUTCOME ............................................9 H. FUTURE OPERATION ............................................ 10 I. KEY LESSONS LEARNED ............................................ 10 PART II: STATISTICAL TABLES ............................................ 12 APPENDICES A. Aide Memoire B. Evaluation Summary of Government's Completion Report Map IBRD No.20814 FAO Map No. 1 FAO Map No. 2 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 l   IMPLEMENTATION COMPLETION REPORT REPUBLIC OF UGANDA SOUTH WEST REGION AGRICULTURAL REHABILITATION PROJECT (Cr. 1869-UG) PREFACE 1. This is the Implementation Completion Report (ICR) for the South West Region Agricultural Rehabilitation Project (SWRARP) in Uganda, for which Credit 1896-UG of SDR 7.6 million (US$10 million equivalent) was approved by the International Development Association (IDA) on February 9, 1988 and made effective on August 26, 1988. The project was co-financed by the International Fund for Agricultural Development (IFAD) for which a Loan of SDR 9.1 million (US$ 12 million equivalent) was approved on February 23, 1988. In addition, a grant of US$0.3 million was approved by IFAD from its Special Operation Facility (SOF). The project was administered by IDA. 2. The IDA Credit and IFAD Loans were closed on June 30, 1996, as compared with the original closing date of June 30, 1994. Some 93 percent of the IDA Credit and 89 percent of the IFAD Loan were disbursed, with the last disbursement taking place in July 1996 after the processing of the final withdrawal application. 3. The ICR was prepared by staff of the FAO/World Bank Co-operative Program (CP) on behalf of AFTAI/CD4 of the Africa Region. Contributions to the report were made by Taqi Sharif, Fred King, and Donna Criddle (AFTAI) and it was reviewed by Sushma Ganguly, Technical Manager, AFTAI. The ICR is based on information obtained from the project files and on the findings of an ICR mission' which visited Uganda in October 1996. The borrower prepared its own assessment of the project performance. A copy of the Aide Memoire of the Implementation Completion Mission and of the evaluation summary of Government's Completion Report are attached as Appendices A and B, respectively. Guy Evers (Mission Leader, Agronomist), Frances Maundrell (Agricultural Economist, Consultant) and El Houssine Bartali (Civil Engineer, Consultant).   - ii - IMPLEMENTATION COMPLETION REPORT REPUBLIC OF UGANDA SOUTH WEST REGION AGRICULTURAL REHABILITATION PROJECT (Cr. 1869-UG) EVALUATION SUMMARY Introduction 1. During the period 1980-1988, IDA supported the agricultural sector in Uganda by providing credits which focused on sector reconstruction, agricultural development, sugar, and forestry rehabilitation. The South West Region Agricultural Rehabilitation Project (SWRARP) quickly followed a similar Bank/IFAD funded Agricultural Development Project in Eastern Uganda, and was intended to form part of a program to revitalize agriculture in Uganda in the aftermath of many years of internal disruption. Project Objectives 2. At appraisal (1987), the Gross Domestic Product (GDP) per capita had fallen by 40 percent since 1970, with the ravages of war and insecurity having had a serious, negative impact on the rural economy. In the project area2, some 20 percent of the 2,700 km rural roads was considered to be impassable and 40 percent was passable only in the dry season. Marketing, input distribution channels, and agricultural support services had largely disintegrated. The project's objectives were to increase food production, incomes, and living standards of small farmers in the region. The project originally included the following components: (i) launching of an adaptive research program and reviving extension services; (ii) procurement and wholesale distribution of agricultural inputs; (iii) rehabilitation of 2,000 km of rural access roads; (iv) project management; and (v) monitoring and evaluation. Emphasis was given to rural road rehabilitation and input supply components, which together represented about 80 percent of total project costs. The project was co-financed by the Bank (US$10 million) and IFAD (US$12 million). A Special Operational Facility grant of US$0.3 million from IFAD was to be provided for start-up costs. 3. After appraisal, IFAD introduced further proposals for Community Development which were regarded as a separate component, although there was no change to IFAD's loan agreement. In 1992, a Pilot Credit Sub-component (under Community Development) was incorporated in an amendment of the IFAD loan agreement and, in 1994, following a request from the Government of Uganda (GOU), a new component for the construction of Small Valley Dams was introduced under IFAD's loan agreement. 4. The original project objectives and design were appropriate to the region's priority needs in the late 1980s. As both the national and local economies revived, the project design was rightly 2 At appraisal, the project area consisted of four districts (Mbarara, Bushenyi, Kabale and Rukungiri); two additional districts were created during implementation within the same area: Kisoro and Ntungamo.  - iii - adjusted from strict rehabilitation towards more long-term development objectives. The Community Development component and Pilot Credit Sub-component were designed to strengthen efforts aimed at core project c'bjectives by providing direct support to small farmers through groups. However, the addition of the Small Valley Dams component near the end of the project, diverted project efforts away from core project objectives rather than contributing to them. Implementation Experience and Results 5. Project costs are estimated at about US$26 million, or about 95 percent of Staff Appraisal Report (SAR) estimates. Investment costs were much lower than forecast (US$12.6 million as against US$19.3 million at appraisal) principally because of shortfalls in civil works, especially roads contracts and agricultural inputs. Recurrent costs were much higher than forecast (US$13.5 million as against US$8.1 million), mainly due to greater expenditure on operating costs attributed to the Project Coordination Unit (PCU), Technical Assistance (TA) and Overseas Training. 6. Project implementation was slow from credit effectiveness in August 1988 until the Mid- term Review (MTR) in 1992, with interventions limited to (a) adaptive research and extension in seven priority counties, (b) training, and (c) procurement and resale of some agricultural inputs. When project implementation took off in 19923, spreading over the entire region, GOU's Economic Recovery Program and the associated liberalization program had led to a rapid improvement in the economic conditions in the project area. By this time, the private sector was active and had its own access to foreign exchange, making the project's agricultural input supply component largely unnecessary. The component nevertheless had a positive impact by supplying inputs when they were sorely needed early in the project, as discussed below. At the MTR, agreements were reached on financial control and reporting, deployment of feeder road units, tender for the main roads' contract, introduction of performance allowances, and finalization of the design of the credit component. 7. The project contributed substantially to reviving agricultural research activities and restoring extension activities in the project area. Significant outcomes of this component, albeit limited to some zones of the project area, are the release and wide adoption of superior Irish potatoes, and the development and adoption of improved banana husbandry practices which will require minimal pesticide applications to effectively limit weevil attacks, and increase yields by 20 to 30 percent. The benefits of using improved seeds would, however, need further demonstrating to farmers, while seed production capacity would need to increase to meet expected demand. In 1994, this component's activities were handed over to regular research and extension programs, each being supported by a separate IDA-funded project (Credit 2446-UG and Credit 2424-UG). Rather than support extension and research activities in a fragmented manner, the rationale of the new projects was to support national programs. This was a positive step, but the transfer of extension and research activities reportedly resulted in some dilution of activities, due especially to the decentralization policy (implementation of extension activities has been transferred to the districts) and budget reductions. 8. Despite shortcomings, and being overtaken by events, the Agricultural Input Supply component had a positive impact by supplying inputs and facilitating the establishment of retailers in the region by providing them with stocks at a time when business was brisk. Private suppliers of / The project picked up in 1992 due to changes in Ugandan staff and had improved markedly even before the mid- term review which took place in May 1992.  - Iv - agricultural and veterinary inputs now operate in centers throughout the region. Overall, the value of items handled (including basic tools, sprayers, chemicals, wheelbarrows and bicycles) amounted to US$3.0 million, or 58 percent of SAR estimates. Expenditure on agricultural inputs was lower than forecast because: (i) the size of early orders was reduced due to lack of storage capacity; (ii) delays in procurement tended to have a domino effect which, together with difficulties in selling some stock items, limited the number of purchases which could be completed; (iii) it proved unnecessary for the project to import certain materials (e.g., construction materials and vehicles) which were readily available in the project area; and (iv) by the time the last purchases took place in 1993, the private sector was active and had its own access to foreign exchange. 9. The project's significant and most widespread economic impact has been generated by the rural access road rehabilitation program. Although implemented from 1992 onwards in an altered form4, the component improved some 1,550 km of rural roads to dry-weather standards (compared with the rehabilitation of 2,000 km to all-weather standards, as originally conceived at appraisal). It is estimated that around 175,000 rural families (about I million people), who live within easy reach of the roads have benefited from them. Surveys have confirmed that the opening up of previously impassable roads has generated an increase in general economic activity and agricultural production, with the share of the project area in the Kampala/Entebbe cooking-banana market increasing from 40 percent to 80 percent. Given adequate road maintenance, farmers in the project area will continue to benefit, through increased access to agricultural support services, easier transport and marketing. 10. Project Management was problematic in the first years of the project due to lack of experience of newly-appointed staff and high turnover of TA experts, which led to difficulties in initiating and coordinating project activities. The combination of both infrastructure and agricultural support components, totaling seven in all, under the PCU, which was located regionally in Mbarara, overstretched local management capacity. This led to liaison problems with the Ministry of Local Government (MOLG) during the first half of the project. Rapid turnover of staff also contributed to early implementation problems. Restructuring of the project and positive changes in senior project management in 1992, greatly improved project performance. Better collaboration with local authorities also resulted in enhanced political and administrative support from district administrations, and the Inter-ministerial Coordinating Committee played an increasingly effective role. 11. Overall, the Monitoring and Evaluation Unit performance fell short of its objectives. The unit carried out baseline, mid-term and final household surveys and thematic surveys which provide valuable information on project impact. However, due mainly to shortage of local staff (already redeployed on account of decentralization), reluctance to utilize available TA, and financial resource constraints, the final household survey was simplified, and does not provide quantitative data on agricultural production which could be compared with the results of the 1989 baseline survey. The unit did not carry out regular financial (cost) monitoring. Although recommended in the mid-term review, low priority was given to the monitoring of agricultural activities, cropping surveys were not 4/ Equipment procured under the project for maintenance was used by district units for spot improvement of 1,060 km from 1992. Road rehabilitation by contractors of 488 km took place between 1994 and 1996.  undertaken, and no analysis was carried out on the financial viability or impact on labor requirements of the recommended crop packages and income generating activities. 12. The Community Development component, including the Pilot Credit Sub-component (PCS), has provided support to over 500 farmer groups and, through groups, to individuals, with some 65 percent women members. The Groups identified by community development staff were also used as channels for extension and seed production activities. The component has had a localized impact on incomes and nutrition through its promotion of diversified income-generating activities; but possibly its greatest impact has been the fostering, through training, of local management capacity to maintain simple books of account for Group activities, including for the PCS. 13. The PCS was, overall, a success in that it has played an important role in fostering enterprise and demonstrating income-generating activities in rural areas. Against an original allocation of US$ 414,000, a total of US$ 1.01 million was disbursed largely using reflows, as a result of repayment rates exceeding 80 percent. The performance of the project team is commendable considering the low repayment rates achieved under similar Credit schemes in other projects. It is also notable that the PCS was a program of Ugandan design, implemented by determined and concerted efforts of many different local authorities. While the administrative costs of PCS were relatively high, it can nevertheless serve as a model for the future expansion of Credit to farmers in the rural areas, given the present reluctance of commercial banks accredited under ongoing Credit programs to open rural branches and onlend to the farmers. Following a proposed GOU policy for lending to farmers, an NGO, the South West Uganda Agricultural Development Agency (SWUADA) was set up in early 1996 to, inter alia, provide credit to farmers in the project area by registering as a rural financial intermediary (RFI), and receive assistance for building its technical and managerial capacity. SWUADA had sought seed capital money from GOU, but on account of delays in the implementation of the capacity building program envisaged under an ongoing credit program, no material progress has been made in this regard. 14. The small valley dams/tanks component aimed first at rehabilitating existing and then constructing new, small valley dams and/or tanks. Its implementation performance has been less than satisfactory. Of the 15 potential sites considered, five were selected and engineering designs were prepared, but one had to be dropped due to time and budget limitations. However, no cost-benefit or environmental impact studies were made, contractors could not meet their deadlines, and one of them failed to comply with the construction specifications. As of October 1996, progress of work ranged from 61 percent to 93 percent toward completion of the four dams. This component was adversely affected by time constraints and that the same project staff, supervision consultant and contractors also had to complete concurrently the rural road component. 15. Although the Bank's performance at appraisal was generally satisfactory, more attention should have been given to: (i) the capacity of local staff to implement a multi-sectoral project; (ii) the complexity of local tendering procedures; (iii) the estimates of the time required to establish logistics for project implementation and inputs procurement; and (iv) the costing of the rural road component. From mid-1988 to mid-1996, there were a total of 15 supervision missions which sought to address the key issues affecting the project's performance by taking these up with the Government, and directing their efforts towards facilitating project implementation. Continuity of  - vi - task management in the latter half of the project, combined with participation of Resident Mission (RM) staff in supervision, as part of a team, contributed to achievement of project objectives. In addition, the support provided by the RM in screening of disbursement and procurement documents was greatly appreciated by project management. Until around 1992, Borrower performance was deficient, especially due to poor project management, inadequate counterpart funding, and failure to comply with the covenants on accounts and audit. From 1992, the above deficiencies were rectified, and the Borrower's performance and commitment to the project until completion were commendable. 16. Sustainability issues relate principally to assuring continued road maintenance, seed production and credit. It is noteworthy that GOU has, out of its own resources, provided funding in FY97 to complete ongoing project activities, including funds for feeder roads' improvement and maintenance. This is in line with GOU's development priorities. This funding, will, however, not be adequate to purchase additional roads maintenance equipment sufficient for the six districts. The benefits of using clean seeds, especially for potatoes, would need further demonstrating to farmers, and seed production capacity should substantially increase to meet expected demand, and therefore sustain production. The pilot credit operation could have been continued through SWUADA as an RFI if GOU had accepted the recommendation of the last supervision mission, whilst remaining within the framework of its proposed capacity building program for RFIs under existing credit programs. Because of delays in the operationalization of the modalities agreed between GOU and the Bank for providing credit to farmers through RFIs, including the capacity building of RFIs, project sustainability for the above activities is considered uncertain. Adaptive Research and Extension has been merged with national programs supported by IDA projects, and the sustainability of this component now relies on the performance of these two projects. Private sector arrangements, which appear in place throughout the region, have successfully taken over the supply of agricultural inputs. 17. Combining with the Government's economic recovery program, SWRARP has thus made a significant contribution to improving small farmer incomes and welfare. The SAR indicated that the Economic Rate of Return (ERR), over a 20-year life, was expected to be in excess of 15 percent. Since continuing benefits will depend greatly on the level of road maintenance, and this is as yet uncertain, the project's ERR has been reworked over the period to the year 2000. The revised 5 calculations show an ERR of about 17 percent . Designed as a rehabilitation, rather than a development project, SWRARP has largely achieved its objectives and is rated satisfactory. Future Operations and Key Lessons Learned 18. Although the IFAD and IDA co-financing credits were closed by mid-1996, SWRARP activities will continue during FY97, solely under GOU financing. The operations to be undertaken by the project during FY97 include road rehabilitation and maintenance, completion of the small valley dams component, follow up of credit activities, and handing over project activities to their respective agencies by June 1997. However, because the project's monitoring and evaluation section failed to measure key indicators adequately and because it is impossible to clearly separate the project's impact from other changes in the economic environment, the estimate is far from precise.  - vii - 19. Key lessons learned are as follows: (a) The project recorded impressive achievements in activities such as the road spot improvement program; the pilot credit scheme; and the community development scheme. This is largely the result of a strong commitment of the Government (at central and district levels), which provided the project management team with good logistical support and allowances. (b) Continuity of Bank staff for task management, particularly in the second half of the project, and close liaison with and involvement of Resident Mission staff in supervision, as part of a team, contributed to achieving project objectives. (c) The setting up of a screening service in the Resident Mission, to initially review withdrawal applications for disbursements and tender/bid documents for proposed procurements, towards the end of the project, was greatly appreciated by project management. (d) More attention needs to be devoted to devising satisfactory arrangements for procurement and making contingency plans at appraisal. Where rapid start-up of project activities is expected to be crucial to the impact of the project (as in the case of agricultural input supply and road rehabilitation components), the Bank's appraisal team should include an experienced project manager who might, in addition to defining implementation schedules, also set in motion the procurement process by arranging the preparation of draft tender documents. (e) It is recognized that the addition of the Small Valley Dams component towards the end of the project was based on a decision at the highest level of Government. Generally, however, when borrowers and lenders agree to add a sizeable component towards the end of the project, the Bank should make a full assessment of the capacity of project management and of potential contractors to undertake the additional works. (f) In a project such as SWRARP, Technical Assistance was required to play a crucial role. In order to limit the risks of excessive turnover of key TA staff, such as the financial controller, every effort should be made by Government to choose suitable candidates for key technical assistance posts. (g) The adaptive research program gave priority to the screening of exotic, newly introduced crop varieties. In addition to screening exotic crop varieties, consideration should be given to local varieties, to their improvement through proper seed production, and to the possibility for their incorporation in local breeding programs. (h) To yield significant impact, the release of superior crop varieties should be accompanied by sustainable seed production systems, especially simple, local farmer-based seed production, as successfully experienced by the project for potatoes.  IMPLEMENTATION COMPLETION REPORT REPUBLIC OF UGANDA SOUTH WEST REGION AGRICULTURAL REHABILITATION PROJECT (Cr. 1869-UG) PART I: PROJECT IMPLEMENTATION ASSESSMENT A. STATEMENT/EVALUATION OF OBJECTIVES 1. The South West Region Agricultural Rehabilitation Project (SWRARP) quickly followed a similar Bank/IFAD-funded Agricultural Development Project in eastern Uganda and was intended to form part of a program to revitalize agriculture in Uganda in the aftermath of many years of internal disruption. The project's objectives were to increase food production, incomes and living standards of small farmers in 4 (later divided to become 6) districts6 of the Southwest. The SAR affirmed that smallholder farming families of the region were poor and should be potential beneficiaries of the project. As formulated in the SAR, the project included the following components: (a) Adaptive Research and Extension for rehabilitating two District Farm Institutes (DFI) and improving the facilities of research sub-stations; and for strengthening adaptive research and extension through provision of vehicles, equipment, technical assistance, training and operating costs. (b) Agricultural Input Supply for the procurement and sale of small, priority agricultural inputs; for the procurement of vehicles, material or equipment to strengthen marketing infrastructure; and for establishing a warehouse at Mbarara. (c) Rural Access Roads, including a program of rehabilitation and spot repairs on 2,000 km of rural roads; and the provision of plant, equipment, incremental operating costs and in-service training for staff of the Ministry of Local Government (MOLG). (d) Project Management to enable the Project Coordination Unit (PCU) to coordinate implementation and administer a Community Development Fund. (e) Monitoring and Evaluation (M&E) to bring together information on project implementation and evaluate impact. 2. During implementation, IFAD further elaborated proposals for Community Development in order to give greater support to group formation and training. This was then regarded as a separate component, although there was no change to IFAD's loan agreement. Farm credit was originally intended to be provided through separate on-going programs, but when this proved to be unworkable 6 At appraisal, the project area consisted of four districts (Mbarara, Bushenyi, Kabale and Rukungiri); two additional districts were created during implementation within the same area: Kisoro and Ntungamo.  - 2 - for the assisted groups, a Pilot Credit Sub-component (under Community Development) was incorporated into an amendment of the IFAD loan agreement in mid-1992. Following a request from GOU at the highest level in 1994, an agreement was reached to include the construction of a number of small valley dams within IFAD's loan agreement. Thus, the project finally implemented 7 components. 3. At project appraisal, Gross Domestic Product (GDP) per capita had fallen by 40 percent since 1970, and war and insecurity had devastated the rural economy. In the project area, some 20 percent of the 2,700 km rural road network was considered to be impassable and 40 percent was passable only in the dry season. Marketing, input distribution channels, and agricultural support services had largely disintegrated. Project design was thus directed mainly towards rehabilitation of key infrastructure (roads - 48 percent of project costs) and restoration of supply of agricultural inputs (30 percent of project costs), with smaller allocations for the Adaptive Research and Extension component and community development. 4. The original project objectives and design were appropriate and responsive to the region's priority needs in the late 1980s. However, the Economic Recovery Program launched in 1987 and the associated Liberalization Program led to a rapid improvement in the economic conditions in the project area. As both the national and local economies revived, project design was rightly adjusted from strict rehabilitation towards more long-term development objectives. The Community Development component and Pilot Credit Sub-component were designed to strengthen efforts aimed at core project-objectives by providing direct support to small farmers through groups. Inclusion of the Small Valley Dam component extended the scope of the project to livestock sector development. Although this directed support to areas of the region which would otherwise have benefited only marginally from other project interventions, the addition of this component near the end of the project appears to have diverted project efforts away from core project objectives rather than contributing to them. The combination of both infrastructure and agricultural support components, totaling 7 in all, under a single Project Coordination Unit under the Ministry of Agriculture, Animal Industry and Fisheries (MAAIF), which was located regionally at Mbarara, overstretched local management capacity. This led to liaison problems with the Ministry of Local Government (MOLG), which was implementing the Roads component, during the first half of the project period. B. ACHIEVEMENT OF OBJECTIVES 5. Overview. The project became effective in August 1988, but implementation was slow until the MTR in 1992. Interventions were limited to adaptive research and extension (in 7 priority counties), (Map 1) training, and the procurement and resale of some agricultural inputs. From 1992, the project took off, with a substantial part of activities being achieved during the last three years before completion. As of 1993/94, the Adaptive Research and Extension Component was transferred to two other IDA-funded projects (para 22). 6. Adaptive Research and Extension. The adaptive research sub-component was designed to address small farmer needs, with emphasis on the region's major food crops. In 1989, research priorities were identified by the project, by conducting farmers' surveys, as lack of improved varieties, pests and diseases, declining soil fertility and inferior agricultural practices. The project carried out on-station and on-farm, trials, with major areas of emphasis being the introduction of exotic germplasm, screening and release of promising varieties of Irish and sweet potatoes, beans, maize, wheat and short term upland rice. However, for these crops, little emphasis was given to  - 3 - improving seed quality of and/or regenerating suitable local varieties, and to improving crop management. The project also improved local technologies for banana weevil control and soil conservation. The Extension Sub-component was intended to revive the extension services in the project area. The project provided in-service training and logistic support and field allowances for staff, and equipment for the DFIs located in Bushenyi and Kabale Districts. The project also enabled extension staff to train direct project beneficiaries (i.e., farmer groups) and other farmers, through their regular activities. Some 470 farmer day courses were organized by the project, mainly at DFIs. However, the two DFIs have not yet been rehabilitated (now to be financed under another project, para 22), and appear to be largely under-utilized. 7. The project contributed substantially to reviving agricultural research activities and restoring extension services in the project area. The salient outcome of this component is the release and wide adoption of superior Irish potato varieties in Zone 1 (Map 1) and, in Zone 2, the development and adoption of improved banana husbandry practices requiring minimal pesticide applications and reported to effectively limit weevil attacks, while increasing yields'. The lack of seed production capacity and related development initiatives in the project area have generally limited the adoption of the released improved varieties, except for Irish potatoes, for which the project supported seed production by contracting groups or individual farmers. The benefits of using clean seeds, and the need to renew potato seed regularly has, however, to be further demonstrated to farmers and seed production capacity should substantially increase to meet expected demand. 8. Agricultural Input Supply. The SAR identified lack of inputs, and the scarcity of foreign exchange to procure them, as a key constraint to agricultural production in the region, and proposed the importation of about US$5.6 million worth of inputs for wholesale distribution over 5 years (1989-1993). The total value of inputs procured between 1990 and 1993 was US$3.0 million, equivalent to 58 percent of SAR estimates. However, the Mbarara warehouse could not be constructed in time to receive the first purchases, and it was never built, as alternative solutions became available (renting storage space). Expenditure on agricultural inputs was lower than forecast because: (i) the size of early orders was reduced due to lack of storage capacity; (ii) delays in procurement tended to have a domino effect which, together with difficulties in selling stocks, limited the number of purchases which could be completed; (iii) it proved unnecessary for the project to import certain materials (e.g., construction materials and vehicles) which were readily available in the project area; and (iv) by the time the last purchases took place in 1993, the private sector was active and had its own access to foreign exchange. Despite these shortcomings, and being overtaken by events, the Agricultural Input Supply Component did, nevertheless, have a positive impact by supplying inputs when they were sorely needed early in the project. It also facilitated the establishment of retailers in the region by providing them with stocks at a time when business was brisk. 9. Rural Access Road Rehabilitation. The initial objective was to rehabilitate and spot repair 2,000 Iam, using contractors over a 3-year period, in order to permit year-round access. Substantial delays were incurred in selecting consultants to identify the roads and sections to be rehabilitated, 7 It has been assumed that, for farmers living within easy reach of project roads, the rate of adoption of banana and potato technologies is in the order of 30 percent, yield increases 20 to 30 percent, and area increases 10 percent to 20 percent. 8 Including basic tools, sprayers, vegetables seeds, chemicals, wheelbarrows and bicycles.  - 4 - and to conduct the feasibility study (para 24). The consultants found that, because costs had been underestimated at appraisal and further road deterioration had taken place in the interim, the target for rehabilitation would have to be reduced from 2,000 km to around 500 km of first priority roads. Moreover, these roads could only be restored to dry weather status. The partial rehabilitation of first priority roads was undertaken by two contractors between mid-1994 and mid-1996, focusing mainly on impassable sections. At project completion, 488 km of contract roads had been rehabilitated (Map 1). 10. The road maintenance plant and equipment were acquired in early 1992. This equipment was initially distributed in four units, one for each of the initial project districts. The equipment included: motorgraders, wheel loaders, dump trucks, vibratory rollers, vehicles and assorted equipment for 4 mechanical workshops. The total cost amounted to about US$2.2 million. In order to reduce the shortfall resulting from the target reduction and to speed-up progress, it was agreed between GOU and IDA in early 1992 to use the road maintenance equipment for spot improvement of 1,000 km of feeder roads. Although project equipment was used for work for which it was not originally intended, and therefore depreciated faster than anticipated, it contributed substantially to the improvement of the rural road network in the project area. As of October 1996, some 1,060 km of roads had been spot improved (Map 1). Despite the reduction in road standard and the reduced length of roads improved (about 1,550 vs 2,000 km), the component appears to have substantially achieved its objectives. 11. Community Development. The objective of this component was to promote and train groups, giving emphasis to women group members. The number of groups in the project area had risen from 218 in Project Year (PY) I to well over 500 by the end of the project, with some 65 percent women members. Many received training in day and residential courses at DFIs through which both technical and managerial skills were imparted. Over 70 groups have been assisted through grants for building materials, inputs and equipment. In addition, selected members of 15 women's groups received grants of a heifer for zero-grazing, whose female descendants are in turn being donated to other group members. Beneficiary groups are, in general, spread throughout the region although, from the mission's visits, it would appear that many of the heifers were donated to the more prosperous members of groups near towns (Map 2). 12. Pilot Credit Scheme. The revolving credit fund (US$414,000) started in late 1993. Funds have been lent to groups for on-lending to members, generally for 6 months, but sometimes up to a 12-month term. Using reflows, 373 loans totaling US$ 1.01 million have been made to 241 groups and around 3,000 to 4,000 individual group members, of whom more than 40 percent are women. Beneficiary groups have tended to be clustered in a small number of locations (Map 2). Most loans have been used to finance labor for crop production, especially banana rehabilitation, but also for trading, livestock enterprises, and other income-generating enterprises such as beekeeping and mushroom growing. The value of individual loans commonly ranges between US$100 and US$500 with a maximum of US$1000. Interest rates have varied between 17 percent and 26 percent p.a. and were generally positive in real terms. The pilot Credit Scheme (CS) must, overall, be considered a success in that it has played an important role in fostering enterprise and demonstrating income- generating activities in rural areas. The project team has also shown that it is possible to achieve repayment rates exceeding 80 percent, rarely achieved among small farmers in Uganda. It is notable that this was a program of Ugandan design, implemented by determined and concerted efforts of many different local authorities. While the administrative costs of CS were relatively high, it can nevertheless serve as a model for the future provision of credit to farmers, given the present  - 5 - reluctance of commercial banks accredited under ongoing credit programs to open rural branches and onlend to farmers. 13. Monitoring and Evaluation (M&E). Overall, the M&E unit performance fell short of its objectives. The unit carried out baseline, mid-term and final household surveys and thematic surveys which provide valuable information on project impact. However, due mainly to shortage of local staff (already redeployed), reluctance to utilize available TA and financial resource constraints, the final household survey was simplified, and does not provide quantitative data on agricultural production which could be compared with the results of the 1989 baseline survey. The unit did not carry out regular financial (cost) monitoring. Although recommended in the mid-term review, low priority was given to the monitoring of agricultural activities, cropping surveys were not undertaken, and no analysis was carried out on the financial viability or impact on labor requirements of the recommended crop packages and income generating activities. 14. Small Valley Dams/Tank. This component aimed first at rehabilitating existing and then constructing new, small valley dams and/or tanks. The valley dams and tanks are intended to reduce water shortages for both livestock and human consumption and hence reduce nomadism. Of the 15 potential sites considered, 4 valley dams and 1 valley tank were selected and engineering designs were prepared, but one had to be dropped due to time and budget limitations. However, no cost- benefit or environmental impact studies were made. Contractors could not meet the deadline, and one of them failed to comply with the construction specifications. As of October 1996, progress of work ranged from 61 percent to 93 percent toward completion. This component was adversely affected by the time constraint imposed by its late inclusion in the project relative to the closure of the project. This was exacerbated by the fact that the same project staff, supervision consultant and contractors also had to complete concurrently the rural road component. 15. Financial and Economic Objectives. Actual project costs are estimated at about US$26 million, or about 95 percent of SAR estimates (Part II, Table 8). Investment costs were lower than forecast (US$12.6 million as against US$19.3 million at appraisal), principally because of shortfalls in civil works, especially roads contracts (55 percent of SAR estimates) and the agricultural inputs component (45 percent of SAR estimates). Recurrent costs were higher (US$13.5 million as against US$8.1 million), mainly due to greater expenditure on operating costs attributed to the PCU (US$3.2 million as against US$0.2 million), Technical Assistance (US$4.0 million as against US$3.3 million) and Overseas Training (US$1.0 million as against less than US$ 0.1 million). 16. There were severe shortfalls in the Government's counterpart contribution up until 1992. Total counterpart contribution has, however, amounted to the equivalent of about US$3.9 million (compared to US$5.1 million at appraisal, including taxes) of which 40 percent was from the Special Inputs Fund, 14 percent from MAAIF, and 44 percent from MOLG (US$0.9 million USAID contribution and US$0.8 million MOLG budget). Had the input supply component been implemented earlier and in full, it would have generated significantly higher GOU revenues and could have eased the shortage of counterpart funds in the project's early years. 17. The SAR indicated that the economic rate of return (ERR) of the adaptive research and extension and road components over a 20-year life was expected to be in excess of 15 percent. Benefits were estimated for transport cost savings and predicted yield increases over virtually the  -6 - whole range of food crops. Project benefits have now been re-assessed9 based on: (i) the numbers of farmers living within easy reach of SWRARP improved roads (estimated at 175,000); (ii) estimates on yield and cropped area increases; and (iii) the benefits from income-generating activities, additional trade and employment, and transport cost savings. The revised calculations show an ERR of roughly 17 percent (see Part II, Table 9). C. MAJOR FACTORS AFFECTING THE PROJECT 18. General. The project's implementation has coincided with a long period of peace and political stability which allowed free movement of people and goods. The Economic Recovery Program launched in 1987, and the associated liberalization program, resulted in sustained GDP growth ranging from 3 percent to 8.5 percent per year, and led to a rapid improvement in economic conditions in the project area. 19. Project Management. During the first years, the lack of experience of newly-appointed staff in project management led to difficulties in initiating and coordinating project activities. Rapid turnover of key TA staff (para 21) also contributed to the early implementation problems which might have been solved earlier, if GOU and the Bank had taken prompt, decisive action. Given the initial difficulties of transport and telecommunication, the PCU in Mbarara was isolated from MAAIF (Entebbe), and MOLG, the Ministry of Finance (MOF) and the Bank's office in Kampala. This resulted in substantial delays in project implementation, especially for roads (para 24). Restructuring of the project at the time of the MTR (May 1992) and positive changes in senior project management in 1992, greatly contributed to overall project achievements. 20. Counterpart Funding. At appraisal, assurances were given that Governnent would provide sufficient local funds to support project execution. Counterpart funding was to have been boosted by use of revenues from sales of agricultural inputs under the project. However, Government budgetary allocations were not fulfilled and funds from input sales were not used until 1992 because of delays, misunderstandings and administrative problems. Demand for local funds for project execution in the early years was much higher than anticipated because disbursements were held up by procedural problems associated with lack of proper accounting by project staff and the unfamiliarity of project management with Bank procedures. 21. Technical Assistance. Some 340 person-months of long and short term TA were financed under the project. The main fields of assistance included civil engineering, financial control, input supply, adaptive research and extension. Due to failure to fulfill his mandate, the financial controller had to be replaced twice over less than three years, which contributed to the initial difficulties in setting up an acceptable financial system and in keeping proper accounts. Towards the end of the project, Project Management considered that there was no need to use the available balance of 5 person-months TA to support M&E activities. The absence of external assistance during that period may, however, have contributed to the problems reported on the last surveys carried out by the M&E unit (para 13). 9~~~~~~~~~~~~~~~~~~~ 9 However, because the project's M&E unit failed to measure key indicators adequately and because it is impossible to clearly separate the project's impact from other changes in the economic environment, the estimate is far from precise.  - 7 - 22. Transfer of Adaptive Research and Extension Activities. During 1993/94, following MAAIF restructuring and the approval of an IDA-funded Agricultural Research and Training Project (ARTP) and an Agricultural Extension Project (AEP), project research and extension activities were transferred to the then newly-created National Agricultural Research Organization (NARO) and to AEP, respectively. On-station/on-farm research has been pursued in the project area for Irish potatoes, wheat, banana and agro-forestry/soil conservation (funded by NARO and donors). The transfer of extension activities to the districts, which assumed responsibility for implementation of activities following decentralization, and budgetary constraints has reportedly resulted in some dilution of activities, especially in respect to staff allowances and mobility. 23. Staff Versus Farmer Training. Although the Training budget in the SAR was limited, an ambitious training program abroad was set-up for project staff and district extension staff. The project supported more than 40 short courses and 18 MSc degree courses, of which seven were in extension/education, four in agronomy/ crop protection and seven in economics/management. This led to rather significant short- to long-term absences of senior staff in charge of project and extension activities. Moreover, investment in staff training greatly exceeded those in farmer training, the latter suffering from the delays in restoring DFI capacity and, towards the end of the project, the lack of funds at the district extension offices. 24. Delays in the Rural Road Component. From the beginning of project implementation, this component suffered from chronic delays (para 7). The contract for road rehabilitation started in May 1994, was delayed by almost four years, and was due to be completed by the end of the project (June 1996). A number of factors contributed to the delays in appointing the consultant engineers, the supervision engineers and the contractors. These include: (i) inadequate cost estimate at appraisal resulting in the need to modify the component activities, and having it endorsed by the relevant national authorities and IDA; (ii) poor communication between the project in Mbarara, MAAIF in Entebbe, and MOLG and MOF in Kampala; (iii) unfamiliarity of project management with MOLG and IDA procedures, leading to the need to re-formulate proposals and tenders; (iv) lack of counterpart funds and/or delays in releasing them; and (v) limited capacity of available road contractors who were requested to complete work under pressure, and in some cases being also contracted for other project works (small valley dams). Initial delays also resulted in further road deterioration, which meant further adjustment to the component design and costing. In addition, one contractor could not fulfill his contract, which was partially handed over to a second contractor. In undertaking the additional work, the second contractor was obliged to subcontract to other local contractors. Part of the additional work was undertaken by project management under force account. D. PROJECT SUSTAINABILITY 25. Adaptive Research and Extension, one of the field components, has been merged with national programs supported by IDA projects (ARTP and AEP). Therefore, the sustainability of this component relies on the performance of these two projects. In this respect, priority should be given to developing farmer-based seed production for both local and recently released varieties. Agricultural Input Supply was intended to be taken over by the private sector. This took place spontaneously during the course of the project and satisfactory private sector arrangements appear to be in place. Private suppliers of agricultural and veterinary inputs operate in centers throughout the region.  - 8 - 26. Maintenance of roads rehabilitated by contractors and by the project is imperative, but only a few roads have undergone routine maintenance so far. GOU now gives the highest priority to road maintenance, and increased its budget allocation for rural roads almost four-fold in FY97. However, this is still below the country's requirements, and districts are not able to match MOLG allocations as is required. Road maintenance equipment purchased under the project for 4 district units has been spread over the 6 district units (see para 1). The Government is in the process of procuring additional equipment worth US$ 4 million for the whole country, of which some will be for the project area, but this will still not raise the districts' maintenance capabilities to the level originally intended by the Bank. Other measures which would contribute to improved road maintenance, especially in Zone 1 (Map 1), include limiting land slide incidence by persuading farmers to adopt soil conservation practices, limiting axle loads of lorries, and rehabilitating trunk roads. The shortage of equipment and insufficient local funding may hamper the sustainability of part of the rehabilitation undertaken by the project. 27. Sustainability of small valley dams/tank will first depend on the contractor's performance to complete the on-going works. It will also depend on the degree of organization of Water Users' Associations (WUAs) and how well they manage to raise funds from members and support from local authorities to maintain/protect the site environment in general and the dam structure and related equipment in particular. However, WUAs have not yet demonstrated their capacity to manage their new assets. 28. The project was asked to terminate its credit activities from mid-February 1996, and only loan recovery activities have been pursued. Funds recovered are to be transferred by SWRARP to the Bank of Uganda. In January 1996, the South West Uganda Agricultural Development Agency (SWUADA) was established as an NGO. Among its intended activities, SWUADA will provide credit to its members, along the lines of the project credit component. The last supervision mission recommended in July 1996 that GOU consider giving some seed capital (Ush. 100 million) to SWUADA, and based on legal safeguards, consider it for inclusion in the Rural Financial Intermediary (RFI) Capacity Building Program, proposed under the IDA-supported Cotton Subsector Development Project. In light of proposed GOU policy for providing credit to farmers through RFIs, SWUADA would become a financial intermediary between commercial banks and farmers. If SWUADA is to successfully take over the credit scheme in a sustainable manner, it must: (i) reduce the cost of loan management and maintain high recovery rates (para 12); (ii) be registered as a financial intermediary under the awaited new GOU policy; and (iii) receive assistance for building its technical and managerial capacity. On account of delays in the operationalization of the capacity building program of RFIs, however, no progress has been made in this regard. In light of above factors, sustainability of all project activities is judged as uncertain at this stage. E. BANK PERFORMANCE 29. Although the Bank's performance at appraisal was generally satisfactory, more attention should have been given to: (i) the capacity of local staff to implement a multi-sectoral project; (ii) the complexity of local tendering procedures; (iii) the estimates of the time required to establish logistics for project implementation and input procurement; and (iv) the costing of the rural road component. From mid-1988 to mid-1996, there were a total of 15 supervision missions including the mid-term review (1992). During the initial years, there was a relatively high turnover of Bank staff working on the project, which had four task managers. Continuity of task management in the latter half, combined with involvement of Resident Mission (RM) staff in the supervision missions, as part  - 9 - of a team (the role of task manager was alternated between headquarter and resident mission staff for a year or so), contributed to the achievement of project objectives. The supervision missions sought to address the key issues affecting the project's performance by taking these up with the Government, and directing their efforts towards facilitating project implementation. In particular, the Bank's prompt action on agreeing on a variation order on road contracts in the last year was crucial to the progress of that component. The performance of the minor project components might, however, have been improved if the Bank had provided more technical supervision. The support of the Uganda RM was greatly appreciated by project management, especially the instigation of a financial screening service for claims and, towards the end of the project, assistance with procurement procedures. 30. Around 45% of all disbursements occurred during FY95 and FY96, corresponding to the project extension periods. Although two extensions had already been approved, at the end of FY96, the project was operating at full steam to complete, under time pressure, the lately added small valley dam component and the rural road rehabilitation program. At the end of the project, it was also necessary to further support the pilot credit scheme until the new GOU policy on rural credit became effective. However, rather than considering an exceptional third extension (the credit was not exhausted), the Bank and IFAD decided to close the loans. F. BORROWER PERFORMANCE 31. Borrower performance can broadly be divided into two periods. Until around 1992, Borrower performance was deficient, especially in that: (i) it appointed an inexperienced project management team which failed to initiate project activities; (ii) there was poor coordination between the project team and the various institutions concerned (MAAIF, MOLG and MOF); (iii) it only met, and with delays, a marginal part of its counterpart funding commitment; and (iv) it did not comply with the covenants on accounts and audit. From 1992, the above deficiencies were gradually rectified, and the Borrower's performance and commitment to the project, particularly following agreements reached with the Bank's MTR mission on financial control and management, deployment of feeder road units, tender for main roads' contract, performance allowances, and finalization of the design of the credit component, until completion were commendable. In particular, the project recorded major achievements under the following activities: (i) the rural road spot improvement program; (ii) the pilot rural credit scheme; and (iii) the Community Development scheme. Better collaboration with local authorities also resulted in enhanced political and administrative support from the district, and the Inter-ministerial Coordinating Committee played an increasingly effective role. G. ASSESSMENT OF OUTCOME 32. It is difficult to assess the impact of SWRARP alone in the project area because of the overall economic changes during the period, complementary development interventions, and because of the shortage of data. Also, a large part of the road rehabilitation program has been completed only recently and some of its potential impact has not yet been realized. The mission believes, however, that the project's overall impact has been substantial and positive. Mission estimates suggest that at least 175,000 rural families, equivalent to around one million people benefited from project activities. The road program has already generated an increase in general economic activity and agricultural production through opening previously impassable roads to traffic and reducing transport costs. This has been verified by surveys showing increased traffic volumes  - 10- on both feeder and trunk roads in the area. It is also supported by a survey carried out by a research team of NARO'0 claiming that, over the last 10 years, the share of the project area in the Kampala/Entebbe cooking-banana market increased from 40 percent to 80 percent. Together with improved access to markets, improvements in crop productivity resulting from the Adaptive Research and Extension component (especially of bananas and Irish potatoes) have boosted farmers' incomes and food availability almost region-wide. This has been confirmed in the final household survey, measured by the increased number of assets acquired by project beneficiaries. The Community Development and Credit component has had a more localized impact on income and nutrition through its promotion of diversified enterprises and by fostering local management capacity. It may, however, have had the effect of further concentrating, rather than spreading, the distribution of resources in some rural communities. H. FUTURE OPERATION 33. Although the IFAD Loan and IDA Credit were closed by mid-1996, SWRARP activities will continue during FY97, solely under GOU financing. This is commendable and reflects the GOU's interest in completing ongoing project activities. The Government's decision to extend its support was also made in the hope that a follow-on projectl proposal would be considered by the Bank and IFAD. Eventually, the proposal was not considered, as the Bank and IFAD are now moving towards other lending instruments such as the Agriculture Sector Management Project and possibly an Agricultural Sector Investment Program (ASIP), rather than individual projects. 34. The operations to be undertaken by the project during FY97 concern mainly the following: (i) additional feeder road rehabilitation (spot improvement of 300 km), maintenance of project- rehabilitated roads, and follow-up of the work to be undertaken under the defects liability period for the contract roads; (ii) completion of the small valley dam/tank component; and (iii) loan balance collection (USh 350 million), credit group training, and provision of training and managerial assistance to SWUADA. Since the PCU is planned to be dismantled by June 1997, future operations will also be directed towards handing over project activities to their respective agencies. I. KEY LESSONS LEARNED 35. Key lessons learned are as follows: (a) The project recorded impressive achievements in activities such as the road spot improvement program; the pilot credit scheme; and the community development scheme. This is largely the result of a strong commitment of the Government (at both central and local levels), which provided the project management team with good logistical support and allowances. Communication from the Banana-based Cropping Systems Research program of NARO. t Project proposal for the South-Western Uganda Agricultural Development Project (SWUADP), prepared by the Agricultural Planning Department of MAAIF (June 1996).  (b) Continuity of Bank staff for task management, particularly in the second half of the project, and close liaison with and involvement of Resident Mission staff in supervision, as part of a team, contributed to achieving project objectives. (c) The setting up of a screening service in the Resident Mission, to initially review withdrawal applications for disbursements and tender/bid documents for proposed procurements, towards the end of the project, was greatly appreciated by project management. (d) More attention needs to be devoted to devising satisfactory logistical arrangements and making contingency plans at appraisal. Where rapid start-up of project activities is expected to be crucial to the impact of the project (as in the case of agricultural input supply and road rehabilitation components), the Bank's appraisal team should include an experienced project manager who might, in addition defining implementation schedules, also set in motion the procurement process by arranging the preparation of draft tender documents. (e) It is recognized that the addition of the Small Valley Dams component towards the end of the project was based on a decision at the highest level of Government. Generally, however, when borrowers and lenders agree to add a sizeable component such as the Small Valley Dams component towards the end of the project, the Bank should make a full assessment of the capacity of project management and of potential contractors to undertake the additional works. (f) In a project such as SWRARP, Technical Assistance was required to play a crucial role. In order to limit the risks of excessive turnover of key, expatriate TA staff, such as the financial controller, every effort should be made by Government to choose suitable candidates for key technical assistance posts. (g) The adaptive research program gave priority to the screening of exotic, newly introduced crop varieties. In addition to screening exotic crop varieties, consideration should be given to local varieties, to their improvement through proper seed production, and to the possibility for their incorporation in local breeding programs. (h) To yield significant impact, the release of superior crop varieties should be accompanied by sustainable seed production systems, especially simple, local farmer-based seed production, as successfully experienced by the project for potatoes..  - 12- PART II: STATISTICAL TABLES Table 1: Summary of Assessments A. Achievement ofobjectives SubstaPtial a Negligible NotApplicable O) O~(1) (O)O Macro policies Li ED [ Sector policies Li i [ E Financial objectives [z] Li Ei Li Institutional development Ei Li EJ] Physical objectives FI] Li Li Poverty reduction Li i L i Gender issues Li Li Li Other social objectives L]i ED Environmental objectives Li i i E Public sector management Li Li Li Private sector development Li 1?] Li Li Other (specify) Li i L L B. Jroject shinnbiiLX Likldy Unlikely Uncerain O) ~(I O~,) C. Rnk sats&cW Satsn Deficient (1/) (1) (I) Identification L Li Preparation assistance i [ Appra Li Li Supervision L [ L  - 13 - D. Borrower performance satisfacto= Satisfactory Deficien (1) (O) (1) Preparation Z E ] Implementation D T E Covenant compliance L] E1 Opeation (if applicable) [EI 7I Ew HWihlv E. Assessmentofoutcome satisfa Satisfact Unsatisfactory uatbsa D/ D/) ) O [E O Ol  - 14 - Table 2: Related Bank Loans/Credits Loan/Credit Title Purpose Year of Status Approval Preceding Operations 0983-UG First Reconstruction Program 1980 Completed 1252-UG Second Reconstruction Program 1982 Completed 1474-UG Third Reconstruction Program 1984 Completed 1328-UG Agricultural Rehabilitation 1983 Completed Project 1434-UG Second Technical Assistance 1983 Completed 1539-UG Agricultural Development Project 1985 Completed 1824-UG Forestry Rehabilitation 1987 Completed 1893-UG Sugar Rehabilitation 1988 Completed Following Effectiveness of SWRARP 2176-UG Livestock Services 1990 Ongoing 2190-UG Agricultural Sector Adjustment 1990 Ongoing Credit 2362-UG Northern Reconstruction 1992 Ongoing 2424-UG Agricultural Extension 1992 Ongoing 2446-UG Agricultural Research & Training 1993 Ongoing 2609-UG Cotton Subsector Development 1994 Ongoing 2845-UG Agricultural Sector Management 1996 Awaiting effectiveness  - 15 - Table 3: Project Timetable Date actuaV Steps in project cycle Date planned Latest estimate Identificationa/ Jan/Feb 1984 Preparation"' March/April 1985 Pre-appraisal Oct/Nov 1986 Appraisal May/June 1987 May/June 1987 Negotiations Nov. 1987 Nov 1987 Board presentation Dec. 1987 Jan 1988 Signing Feb 1988 Feb 1988 Effectiveness May 1988 Aug 1988 Project completion Dec. 1993 Feb 1996 Credit closing June 1994 June 1996 a/ Carried out by IFAD. bI Carried out by the FAO Investment Center on behalf of IFAD.  -16- Table 4: Loan/Credit Disbursements: Cumulative Estimated and Actual (US$ million)1/ 87/88 88/89 89/90 90/91 91/92 92/93 93/94 94/95 95/96 96/97 Cumulative SAR Estimate 0.8 3.6 8.1 13.3 18.0 21.1 22.3 IDA 0.1 1.0 3.0 5.5 7.9 9.5 10.0 IFAD2' 0.7 2.6 5.1 7.8 10.1 11.6 12.3 Actual3' IDA 0 0 0.10 0.40 2.08 3.46 3.87 6.53 9.98 10.09 IFAD 0 0 0.65 1.67 3.03 5.13 8.09 10.03 11.28 11.53 SOF (Grant) 0.19 0.22 0.22 0.22 0.29 0.29 0.29 0.29 TOTAL 0.00 0.00 0.95 2.29 5.33 8.81 12.25 16.85 21.55 21.90 0% 0% 12% 17% 30% 42% 55% 76% 97% 98% " Small discrepancies due to use of differing exchange rates and rounding. 21 Including SOF Grant (USS 0.3 million). 3/ Date of final disbursement: July 19, 1996.  - 17- Table 5. Key Indicators for Project Implementation PART l: Key Implementation Indicators as per SAR INDICATOR Unit Estimated Actual FINANCIAL INDICATORS Disbursement by Donor IDA Credit SDR M 7.6 7.1 IFAD loan SDR M 9.1 7.8 IFAD Special Operation Facility (grant) US$ M 0.3 0.3 alue of Agricultural inputs procured US $ M 5.6 3. alue of Sales versus Procured % 100 80 PHYSICAL INDICATORS Staff housing units constructed No 4 3 Office units constructed: SWRARP Mbarara Office Block m2 0 250 DAO Rukungiri Office Block m2 80 80 Storage units constructed Mbarara Warehouse m2 1500 0 Kalengyere Store m2 60 60 Rubare Store m2 80 60 Training units (DFIs) Renovated No 2 0 Training unit (DFIs) Re-equipped No 2 2 ehicles Staff vehicles Procured No 30 42 Staff Motorcycles Procured No 34 51 Staff Bicycles Procured No 242 7 Rural Access Roads Rehabilitation Rehabilitation to Year Round Access Condition Km 2000 0 Rehabilitation to dry weather conditions (contractors) Km 0 488 Spot improvement by SWRARP Direct Labor Units Km 0 1060  - 18 - INDICATOR Unit Estimated Actual Procurement of Road Maintenance Plants & Equipment SWRARP Procured plants and Equipment: Motorgraders 120 HP No 8 8 Wheel loaders 80 HP No 4 4 Dumptrucks 8 ton No 16 16 4WD pickups No 8 8 Vibro roller No 4 4 Motor cycles 125 cc No 4 4 Bicycles No 40 0 ILO procured plants and Equipment: Crane truck No 0 1 Self loader No 0 1 Water loader No 0 1 Vibro roller No 0 5 Dozer No 0 2 Fuel bowser No 0 1 TECHNICAL ASSISTANCE Project Expediter Pers-M 3 3 Financial Controller Pers-M 36 42 Input Supply Manager Pers-M 36 46 Adaptive Research Specialist Pers-M 36 42 Extension & Training Specialist Pers-M 36 36 Monitoring & Evaluation Advisor Pers-M 24 19 Input Procurement Agent (contracted to UCB) Months 60 60 Input Procurement Agent (intern. consulting firm) Pers-M 0 12 Rural Road Inventory (intern. consulting firm, contract) Months 7 5 Supervision of Feeder Roads (intern. consulting firm) Pers-M na 47 Design of Project Buildings (intern. consulting firm) Pers-M 4 4 ADAPTIVE RESEARCH AND EXTENSION ACTIVITIES Number of Extension Staff No 244 292 Proportion of female extension staff % 10 14 Number of trials established No 291 Extension Packages produced No 5 gricultural Extension Training: Farmer training (day courses held until 1994) No 473 Farmers Trained (until 1994) Pers-days - 16152 Technology Adoption Rate (over total farm population) % 8 n.a.  - 19- Table 5. Key Indicators for Project Implementation PART II: Modified Key Implementation Indicators INDICATOR Unit Estimate Actual TECHNICAL ASSISTANCE Design of Credit sub-component Pers-M 6 6 Expediter of Credit sub-component Pers-M 12 18 Design of Small Valley Dam/Tank Component Pers-M na 2 TRAINING Skills acquired by District Extension Staff MSc/MA degree No 0 10 Diploma No 0 2 Short Courses Abroad No 0 25 Skills acquired by Project Staff MSc/MA degree No 0 8 Diploma No 0 2 Short Course Abroad No 0 15 PILOT CREDIT SCHEME Amount to be disbursed USh M 394 394 Recovery Rate (excluding on-going phases IX-XI) % 100 76-100 Number of Women % 30 43 SMALL VALLEY DAMS/TANK Dam Rehabilitation No 1 1 Dam Construction No 3 2 Tank Construction No 1 1 Table 6: Key Indicators for Project Operation L Key Operating Indicators in SAR/President's Report Estimated Actual not applicable  - 20 - Table 7: Studies Included in the Project |. Study Done By Purpose Status Impact 1.1 Baseline and SWRARP establish a benchmark for monitoring conducted in 1990; final report as per M&E purposes Household Surveys and evaluation (M&E), and design issued in 1992 research & extension activities 1.2 Mid-term SWRARP monitoring and evaluation conducted and report issued in 1992 used for M&E and mid-term review Household Surveys 1.3 Final SWRARP project impact assessment conducted in 1995; report issued in limited due to methodology changes, Household Surveys 1996 qualitative information on project impact 2. Roads Inventory Study Consulting produce roads inventory and conducted from Dec 89 to April 90; re-design component; reduce SAR Company prepare tender documents report submitted early 1991 target for contract rehabilitation from 2,000 to less than 500 km 3. Road Traffic Surveys SWRARP assess changes in traffic and seven surveys were carried out during identification of key feeder roads, and commodities transported project implementation sources and outlets of commodities 4. Inputs-Market Surveys SWRARP accumulate market data on inputs, ten surveys conducted throughout better understanding of market agric. produce and household goods project life; reports issued changes and project impact 5. Privatization of Consulting advise on methodology and timing for carried out and report issued in 1992 recommendations found inadequate Input Sales Company privatization 6. Cropping Surveys - collect quantitative data on crop not undertaken NIL prod. for assessing project impact 7. Extension Staff Survey SWRARP assess extension staff resources two surveys, done in 1989 and 1992 adjustment of staff allocation and and status throughout the project training requirements 8. Beneficiary Contact SWRARP determine the diffusion of carried out and report issued in 1994 useful information on project impact Survey extension messages on farmers' groups 9. Adaptive Research and SWRARP assess the impact of adaptive Done and report issued in 1995 limited indications on adoption rates Extension Assessment research and extension activities of project-promoted technologies 10. Impact Assessment SWRARP assess the effect of credit on completed and report issued in 1995 identification of mechanisms for of Credit beneficiaries effective credit delivery 11. Valley Dam Consultant identify sites for valley dam/tank 13 sites surveyed and engineer tender document for valley dams/tank Feasibility Study rehabilitation or construction prepared for 5 sites construction/rehabilitation Note: Studies No 1 to 7 as per SAR, and studies No 8 to 11 are additional.  - 21 - Table 8A: Project Costs (US$ million) SAR Actual Investment Costs Local Foreign Total Local Foreign Total A. Civil Works 3.9 3.7 7.6 2.6 3.0 5.6 B. Vehicles 0.1 2.1 2.2 0.2 2.5 2.7 C. Equipment 0.1 0.5 0.6 0.5 0.5 D. Agricultural Inputs 1.9 4.0 5.9 0.3 2.8 3.1 DD. Community Development Fund/Credit' - - - 0.6 - 0.6 Sub-total Base Investment Costs 6.0 10.3 16.3 3.7 8.8 12.5 Physical Contingencies 0.8 1.2 2.0 - - - Price Contingencies 0.4 0.6 1.0 - - - Sub-Total (incL contingencies) 7.2 12.1 19.3 3.7 8.8 12.5 Recurrent Costs AA. Operating Costs - Roads2 - - - 1.7 0.6 2.3 A. Operating Costs - Vehicles 1.2 1.7 2.9 0.6 0.7 1.3 B. Other Operating Costs 0.4 0.1 0.5 2.7 0.3 3.0 C. Salaries and Allowances 0.7 - 0.7 1.9 - 1.9 D. Technical Assistance 0.1 2.7 2.8 0.2 3.8 4.0 DD. Overseas Training' - - - - 1.0 1.0 Sub-Total Base Recurrent Costs 2.4 4.5 6.9 7.1 6.4 13.5 Physical Contingencies 0.2 0.5 0.7 - - - Price Contingencies 0.2 0.3 0.5 - - - Sub-Total (incl. contingencies) 2.8 5.3 8.1 7.1 6.4 13.5 TOTAL PROJECT COSTS 9.9 17.5 27.4 10.8 15.2 26.0 " Not included at appraisal. 21 For spot improvements. At appraisal this was to have been done by civil work contract, and operating costs were for maintenance.  - 22 - Table 8B: Project Financing (US$ million) SAR Actual Local Foreign Total Local' Foreign Totall IDA I. Civil Works IA road rehabilitation 2.01 3.07 5.08 n.a.3 n.a. 5.08 IB Other civil works 0.38 0.38 0.76 n.a. n.a. 0.51 II. Vehicles & Equipment 0.19 2.61 2.80 n.a. n.a. 3.61 Consultancy (road supervision) - - - n.a. n.a. 0.89 Unallocated 0.54 0.82 1.36 - Total 3.12 6.88 10.00 n.a. n.a. 10.09 IFAD III Agricultural Inputs 1.25 3.96 5.21 n.a. n.a. 2.68 IV. Operating Costs 0.16 1.88 2.04 n.a. n.a. 4.04 V. Consultant Services, Studies & Training 0.09 2.58 2.67 n.a. n.a. 4.12 Credit - - - n.a. n.a. 0.41 Valley Dams - - - n.a. n.a. 0.28 Unallocated 0.79 1.29 2.08 - Total 2.29 9.71 12.00 n.a. n.a. 11.53 SOF Project Start-up 0.06 0.24 0.30 n.a. n.a. 0.29 GOU2 4.09 1.01 5.1 n.a. n.a. 4.14 of which: MAAIF Budget 0.79 MOLG Budget 0.69 USAID (through MOLG) 0.98 Project Input Revenues 1.68 Project Total 9.57 17.83 27.40 n.a. n.a. 26.05 Values are not strictly comparable to SAR Costab values because: (a) Costab values include the tax elements of imported inputs. Taxes on imported inputs were neither actually paid nor recorded in project accounts. Total costs including national taxes would be higher than the above figures. (b) Costab includes local and foreign cost components for each item. Project accounts do not make any such distinction. An actual breakdown is therefore not available. 2/ SAR included US$ 3 million taxes. 3/ n.a.: data not available.  - 23 - Table 9: Economic Costs and Benefits The SAR indicated that the economic rate of return (ERR) over a 20 year life, of the adaptive research and extension and road components was expected to be in excess of 15%. Benefits were estimated on transport cost savings and predicted yield increases over virtually the whole range of food crops. Project benefits have now been re-assessed based on: (a) the numbers of farmers living within easy reach of SWRARP improved roads; (b) yield increases of 25% for bananas and potatoes by 20% of farmers in these areas; (c) area increases of 10% (bananas), 20% (potatoes) and 5% (other crops) on farms within easy reach of improved roads; (d) typical prices for produce as observed over the project life in local market survey; and (e) an estimate of other benefits from increased milk production and marketing, trade in other produce, additional employment and passenger and other transport cost savings in total equivalent to 10% of estimated benefits from the above annual crops and bananas. Since continuing benefits will depend greatly on the level of road maintenance, and this is as yet uncertain, the project's ERR has been reworked over the period to the year 2000 (3 years after Government support to the project, as such, is due to stop, and a total of 12 years). The annual value of incremental production for years 1997-2000 is estimated at US$ 6.7 million for bananas, US$ 0.8 million for potatoes and US$ 2.1 million for other crops. The annual total net incremental benefit for the same years is estimated at US$ 8 million. All project costs were included in the revised calculation, except for community development, credit and valley dams (for which no direct benefits were included either) and overseas training and monitoring and evaluation which produced no identifiable economic benefit. The revised calculations show an ERR of about 17%. However, because the project's monitoring and evaluation unit failed to measure key indicators adequately and because it is impossible clearly to separate the project's impact from other changes in the economic environment, the estimate is far from precise. Detailed assumptions and basis for the economic analysis are available in the project file.  - 24 - Table 10: Status of Legal Covenants Uganda SWRARP Original Revised Covenant Present fulfillment fulfillment Description of Agreement Section type status date date covenant Comments C 1869 4.01 (b) I CD Continuous NA Borrower to have records and accounts Until the fourth year, audited and a certified the project faced copy of the audit major difficulties to report furnished to set up a financial IDA not later than six system and keep months after the end of proper accounts. the fiscal year Chronic delays in submitting satisfactory 4.01 (c) I CD Continuous NA Borrower to maintain audited SOE and records of SOE, retain project accounts were them for examination noted during the first by IDA, and have half of the project. them audited annually After the MTRh the project generally complied with the two covenants Status: C - Complied with CD - Compliance after Delay NC - Not Complied with SOON Compliance expected in Reasonably Short Time NYD - Not Yet Due Table 11: Compliance with Operational Manual Statements Statement Number and Title I Describe and Comment on Lack of Compliance No deviation from applicable Operational Manual Statements observed H~~~~~~~~~~~~~~~~.  - 25 - Table 12: Bank Resources: Staff Inputs 1/ Stage of Planned Revised Actual project cycle Weeks US$ Weeks US$ Weeks US$ (000) (000) (000) Preparation to na 1/ na na na na na appraisal Appraisal na na na na na na Negotiations through na na na na na na Board approval Supervision na na na na na na Completion Report 2/ 2/ 2/ TOTAL 123.9 226.5 118.9 257.5 I/ Data on Bank Resources are not available prior to supervision of the project due to internal Bank reorganizations. 2/ Excluding FAO inputs, these figures are 5.0 weeks and US$ 10,500.  - 26 - Table 13: Bank Resources: Missions Pprfnrmn Psiting Stage of project cycle Month/ Number Days Specialized Implementation Development Types of problems Identification Preparation 3-4/85 7 23 A, CE, E, HE, LE, - - Pre-appraisal 10-11/86 5 20 A, CE, E, R/E - - Appraisal 5-6/87 6 20 A, IS, M&E - - Negotiations 1-2/88 2a/ 7 A, - - Supervision 2 7-8/89 3 17 Al CDd/ 3 2 CF, OS, PM, PR, T 3 3/90 5 7 A, Ad/, CDd/ 3 2 CF, PM 4 9-10/90 4 5 E, FA, PR, R/E 3 2 CF, PM, PR 5 4/91 5 5 CD, E, FA,R/E 3 2 CF, PDI, PM, R/E, 6 9/91 5 7 E, FA, PR, RD, R/E 3U 2 AFC, CF, PM, RD 7 11/91 1 8cI FA, d/ AFC, RD 8 2/92 1 22c/ FA e C/ AFC, RD 9f/ 5/92 5 13 DI, FA, FA, PR, RD 2 2 AFC, FC, RD 10 10-11/93 4 12 AI, CE, E, FA S S ,AFC, RD 11 4-5/94 3 14 E, FA, RD S S PR, RD 12 10/94 3 10 E, FA, RD S S CS, PDI, RD 13 4/95 3 6 E, FA, RD S S PR 14 12/95 2 14 FA, RD S S RD, VD 15 6/96 2 14 FA, RD S S PR Completion 10/96 3 14 A, E, CE - - at A = AgTiculturist; CD Community Development; CE = Civil Engineer; DI = Disbursement; E = Economist; FA = Financial Analyst; IS = Institution Specialist; R/E = Research - Extension; PR = Procurement Specialist; RD = Roads Engineer; R/E = Research - Extension b/ AFC = Accounts, Audit and Financial control ; CF = Counterpart funding ; CS = Credit scheme ; OS = Office space; PDI = Pricing and Distribution of Inputs PM = Project Management; PR = Procurement ; RD = Feeder roads component; RTA = Recruitment of TA; T = Transport; VD = Valley dams. I = Non significant problems; 2 = Moderate problems; 3 = Major problems (appropriate actions are being taken to address those problems); S = Satisfactory. d/ Funded by IFAD No form 590 prepared Midterm Review.  IMPLEMENTATION COMPLETION REPORT REPUBLIC OF UGANDA SOUTH WEST REGION AGRICULTURAL REHABILITATION PROJECT (CR 1869-UG) APPENDIX A AIDE MEMOIRE   2 IMPLEMENTATION COMPLETION REPORT REPUBLIC OF UGANDA SOUTH WEST REGION AGRICULTURAL REHABILITATION PROJECT (CR 1869-UG) APPENDIX A AIDE MEMOIRE Introduction 1. A mission1/ from the FAO Investment Centre (FAO/IC)/World Bank (WB) Cooperative Program visited Uganda from 7 to 21 October 1996 to prepare the Implementation Completion Report (ICR) for the above project. The mission held discussions with relevant officials from the Ministry of Finance (MOF), the Ministry of Local Government (MOLG) and the Ministry of Agriculture, Animal Industry and Fisheries (MAAIF). The mission worked closely with the Senior Management and staff of SWRARP. Field visits were made to the six project districtsl/ and discussions held with senior Representatives of Local Governments, SWRARP, MAAIF, farmers and group leaders. 2. The mission is grateful to the SWRARP team, ministry officials and farmers for their assistance in gathering the necessary information and organizing successful field visits. Preliminary impressions and findings are presented below, and are subject to modifications following further detailed analysis. Objectives, Components and Time-Frame 3. The project's objectives were to increase food production, incomes and living standards of small farmers in the six districts of the South West of Uganda. Initially, at project appraisal (SAR), SWRARP included five components: (a) Rural Access Roads (US$ 13.3 m) including (i) a program of rehabilitation and spot repairs on 2,000 km of rural access roads; and (ii) provision of plant, equipment, incremental operating costs and in-service training for MOLG. (b) Agricultural Input Supply (US$ 8.1 m) for (i) the procurement and sale of small, high volume inputs; (ii) vehicles, material or equipment to strengthen marketing infrastructure; and (iii) establishment of staff and warehouse at Mbarara. 131 Mr Guy Evers, Mission Leader/Agronomist (FAO TCIR), Ms Frances Maundrell, Agricultural Economist (Consultant) and Mr El Houssine Bartali, Civil Engineer (Consultant). 4' At appraisal, the project area consisted of four districts (Mbarara, Bushenyi, Kabale and Rukungiri), and two districts were created during implementation within the same area: Kisoro and Ntungamo.  3 (c) Adaptive Research and Extension (US$ 3.9 m) for (i) rehabilitation of two District Farm Institutes (DFI) and improvement of facilities at research sub-stations; and (ii) strengthening of adaptive research and extension through provision of vehicles, equipment, technical assistance, training and operating costs. (d) Monitoring and Evaluation (M&E) (US$ 0.8 m) to bring together information on project implementation and evaluate impact. (e) Project Management (US$1.3 m) to enable the Project Coordination Unit (PCU) to coordinate implementation and administer a Community Development Fund. 4. The project was identified by IFAD in 1984, prepared by FAO/IC in 1985 and appraised by WB in 1987. The IDA Credit and IFAD loan became effective in 1988. A Mid-Term Review (MTR) was launched in 1992. There were eight Bank supervision missions before the MTR, and six afterwards. The project initial closing date was set at mid-1994, and was eventually extended to mid-1996. Implementation Context and Evolving Project Design 5. By 1988, when SWRARP started, GDP per capita had fallen by 40 percent since 1970 and the ravages of war and insecurity had reduced the rural economy to a parlous condition. Thus, project design was directed mainly towards rehabilitation of key infrastructure (roads - 48 percent of project costs) and restoration of supply of agricultural inputs (30 percent of project costs) with smaller allocations for adaptive research, extension and community development. The SAR affirmed that all smallholder farming families of the region were poor and should be potential beneficiaries of the project. The Economic Recovery Program launched in 1987, and the associated liberalization program, led to a rapid improvement in economic conditions in the project area. The Government of Uganda (GOU) did, however, recognize that its readjustment policies were tending to result in growing disparities in incomes and later took steps to direct assistance towards the poorest members of the community which were most seriously affected. 6. Project implementation was slow until 1992, with interventions limited to adaptive research and extension in 7 priority counties and training, together with the procurement and resale of some agricultural inputs. Project management was restructured in the early 1990s and when project implementation took off in 1992, spreading to cover the entire region, the design of the project had been reoriented towards a capacity-building and more long-term development approach, in keeping with the changed economic and administrative context then prevailing. Thus, at the end of its life, the project had changed shape with the addition of two components and adjustment of funding allocations. 7. The SAR provided for a US$ 95,000 Community Development Fund, to be administered under the Project Management Unit and financed by IFAD. Subsequently, IFAD further elaborated proposals for Community Development to give group formation and development greater prominence and appears to have regarded it as a separate component, although no direct change in the Loan Agreement resulted. 8. Small farmer credit, having been proposed on a number of occasions during the project cycle, was finally included in the form of a pilot credit subcomponent (under Community Development) in an amendment of the IFAD loan agreement in mid 1992. Originally allocated SDR 500,000 (apparently to cover operating costs as well as a US$500,000 credit fund), the final allocation - designated for a credit fund only - was fixed at SDR 300,000 (US$414,000) in June 1994.  4 9. Following a request from GOU in 1994, an agreement was reached to include the construction of a number of valley dams within IFAD's funding. The last amendment of the IFAD Loan agreement (June 1994) thus included a new Category V for Valley Dams (SDR 380,000) which are now considered to be the project's seventh component. Project Costs and Financing 10. SWRARP provisionally estimates total project costs at US$ 25.1 million compared with the SAR estimate of US$27.4 million. Given the changes in project structure mentioned above and the lack of proper books of account before 1992, it is not possible at this point to give a breakdown of costs by component to compare with the estimates made at appraisal. To illustrate the application of funds, the table below instead shows disbursements against the planned allocation of funds among categories of expenditure. Loan/Grant and Category Loan Allocation Disbursement 1988 1994/96 IDA (Credit) (finally amended 1996) SDR million Category I A - Civil Works (Roads) 3.85 3.85 3.46 Category I B - Civil Works (Other) 0.58 0.42 0.36 Category II - Vehicles & Equipment 2.12 2.82 2.62 Category III - Consultant Services - 0.51 0.61 Unallocated 1.05 - TOTAL 7.60 7.60 7.06 IFAD Special Operation Fund (SOF, grant) 0.30 0.29 IFAD (loan) (finally amended 1994) Category I - Agricultural Inputs 3.66 3.00 1.91 (incl. Community Development Fund) Category ll - Operating Costs 1.54 2.14 2.70 Category III - Consultancies, Training 2.31 3.28 3.10 and Studies"l Category IV - Credit - 0.30 0.25 Category V - Valley Dams - 0.38 0.19 Unallocated 1.59 TOTAL 9.10 9.10 8.15 5' In the SAR Implementation Volume overseas training was allocated US$ 75,000 among miscellaneous operating costs of the Adaptive Research and Extension component. This was intended to fund short term study by adaptive research staff. Quite early in the project a decision appears to have been made to devote considerably more resources to overseas training. At the time no amendments were made to the structure or Loan agreements of the project to accommodate this decision, but in its Loan amendment of mid 1994, IFAD substantially increased its allocation under Category III (Consultants, Training and Studies) mainly, apparently, to reflect sums already spent and planned to be spent on this item.  5 11. Government counterpart contribution has amounted to the equivalent of about US$ 3.9 m of which 42 percent from inputs revenues, 14 percent from MAAIF, and 44 percent from MOLG (US$ 0.9 million USAID contribution and US$0.8 million MOLG budget). Implementation Experience 12. Rural Access Road Rehabilitation. The rural access roads network in the project area was estimated at appraisal at 2,700 km of which 20 percent were impassable due to collapsed bridges and swamps and 40 percent were passable only in the dry season. The initial objectives were to rehabilitate and spot repair 2,000 km, using contractors over a 3 year period, in order to permit year round access, and to support MOLG's road maintenance program by the provision of equipment, training and operating costs. 13. Substantial delays were incurred in selecting an international consulting firm to identify the roads and sections to be rehabilitated and conducting the study. The consultants found that, because costs had been underestimated at appraisal and more deterioration had taken place in the interim, the target for rehabilitation would have to be reduced from the initial 2,000 km to only 488 km of first priority roads in order to keep costs within the original funding allocation. Moreover, these 488 km could only be restored to dry weather status. It was not until the end of 1993 that a supervising consultant was selected to review initial design and supervise works. 14. The partial rehabilitation of first priority roads was undertaken by two contractors between mid-1994 and mid-1996, focusing mainly on impassable sections. One contracting firm which was engaged on contracts for other clients in Uganda, could not fulfil its contract, which was partially handed over to the second contractor. The delays by the first contractor entailed additional payments to the supervising consultant. In undertaking the additional work, the second contractor was obliged to subcontract to other local contractors. As of October 1996, nearly 488 km of contract road rehabilitation have been completed. However, an estimated 5 km in Kisoro district was not completed by the second contractor, due to time constraints. In Ntungamo district, SWRARP direct labor units (under a sub-contract to the second contractor) are still in the process of installing about 60 percent of the required culverts, which were delivered late. About half of the contract roads are still under the liability defects period which, for the last works, will expire in March 1997. 15. In order to reduce the shortfall in the road component resulting from the target reduction, and to speed progress, it was decided to use the project-funded road equipment and plant for spot improvement of 1,000 km of feeder roads. Equipment recovered from a former ILO road rehabilitation project was also used. As of October 1996, some 1020 km ol roads have been spot improved. 16. Staff training under the component consisted of on-the-job training for 6 projectlMOLG engineers with the consultant on road inventories and condition surveys; an MSc in road engineering in the UK for the Project Engineer, and local training for projectlMOLG staff in site management, record keeping, job costing and inventory control. 17. Agricultural Inputs. The SAR identified lack of inputs, and the scarcity of foreign exchange to procure them, as a key constraint to agricultural production in the region and proposed the importation of about US$6.5 m worth of inputs for wholesale over 5 years (1989-1993), starting with US$1.1 m in 1989 and rising to US$1.7 m in 1993. Midway in this program a study was to be conducted to determine if and how the project's wholesale business should be privatized.  6 18. In fact, implementation of the component was delayed by non-availability of storage space and hitches in procurement procedures. The first two buys arrived over 1990, 1991 and early 1992 when inputs were in short supply, but the consignments were limited in size (together US$1.12 m) and, because they were priced below prevailing prices in surrounding areas there was some leakage of inputs from the region. By the time the third and fourth buys (together US$1.70 m) took place in 1993, the private sector was active and had its own access to foreign exchange. Large fluctuations in the exchange rate between contract and delivery of SWRARP consignments meant that turnover was slow and the project had to price some inputs at a loss in order to sell them. Materials for rehabilitation of marketing infrastructure were not imported because they were found to be available on the local market. The warehouse at Mbarara, proposed to be built and completed in time to receive the first buy, was never built and the recommendations of the privatization study were found to be too difficult to be implemented. 19. The total value of inputs procured under the project was US$2.85 million and the funds generated by sales to date are USh 2.42 b, 68 percent of which has been ploughed back into the project as GOU counterpart funding. Had the input supply component been implemented in full, it would have generated significantly higher GOU revenues. 20. Despite its shortcomings, and being overtaken by events later on, the component did nevertheless have a positive impact by supplying inputs when they were sorely needed early in the project and it facilitated the establishment of retailers in the region by providing them with stocks at a time when business was brisk. 21. Adaptive Research. This sub-component was designed to address small farmer needs through an adaptive research program with emphasis on the region's major food crops. In 1989, research priorities were identified through a project study as lack of improved varieties1/, pests and diseases, soil fertility, and agricultural practices. The major areas of emphasis have been the screening and release of promising varieties of Irish and sweet potatoes, beans, maize, wheat and short term upland rice, technologies for banana weevil control and soil conservation. The major outcome of these activities are: (i) the wide adoption of one Irish potato variety (Victoria); and (ii) the development and partial adoption of improved banana husbandry practices (sanitation, trapping, mulching and soil/water conservation) reported to limit weevil attacks, while significantly increasing yields. 22. Concerning legumes and cereals, the lack of seed production capacity and related development initiatives in the project area, have limited the adoption of the released improved varieties. For Irish potatoes, the project, in conjunction with Kalengyere research station, developed a seed production strategy through contracts with groups or individual farmers. The benefits of using clean seeds, and the need to renew potato seeds regularly should, however, be further demonstrated to farmers, and seed production capacity should be substantially increased to meet expected demand. In June 1994, following MAAIF restructuring and the approval of an IDA-funded Agricultural Research and Training Project (ARTP), project research was transferred to the then newly created National Agricultural Research Organization (NARO). On-station/on-farm research is being pursued in the project area for Irish potatoes, wheat, banana and agro-forestry/soil conservation (funded by NARO and donors). 23. Extension and Training. This sub-component was intended to revive the depleted extension services in the project area. In particular, the main objectives were to provide in-service training and logistic support for staff, ensure linkage with research, and restore the capacity of Bushenyi and Kachwekano DFIs 16/ The distinction between improved seeds and varieties was not made.  7 to provide farner and staff training, and equip each with a media-centre. In-service training, mobility and staff allowances enabled extension staff to train project direct beneficiaries (i.e. farmer groups) and other farmers through their regular activities. The project released five technical packages for use by extension. Some project districts were used for testing the unified extension system with the training and visit approach. During 1993/94, all extension activities were handed over to the IDA-funded Agricultural Extension Project (AEP). Some 470 farmer day courses were organized by the project, mainly at DFIs. However, the two DFIs have not yet been rehabilitated (now to be financed under AEP), and appear to be largely under-utilized. 24. In addition to local training, an ambitious staff training program abroad was set-up, resulting in more than 40 short courses and 18 Msc degree courses, of which 7 were in extension/education, 4 in agronomy/crop protection and 7 in economics/ management. Consequently, investment on staff training greatly exceeded those in farmers training, the latter suffering from the delays in restoring DFI capacity, and the lack of funds of the district extension offices. 25. Community Development. Implementation of the Community Development component was adversely affected by inconsistency in its formulation and financing between the SAR, various reformulations by IFAD and IFAD loan agreements. Fortunately, the project set about promoting and training groups with the resources available to it, giving emphasis to women group members. The number of groups in the project area had risen from 218 in PY 1 to well over 500 by the end of the project with some 65 percent women members. Many have received training in day and residential courses through which both technical and administrative skills have been built. As a result, management capacity and confidence, especially of group leaders, is now substantial. 26. Through the Community Development Fund (finally received in March 1993) and funds from the Special Operations Fund (received in 1994), over 70 groups have been assisted. Twenty-six received grants of building materials and funds to employ technicians for completion of stores, otherwise largely built with the groups' own efforts. Some of these 26 groups and around 50 others received inputs and/or equipment. In addition, selected members of 15 women's groups received grants of a heifer for zero- grazing whose female descendants are in turn being donated to other group members. Beneficiary groups are in general spread throughout the region although, from the mission's visits, it would appear that many of the heifers were donated to the more prosperous members of groups near towns. 27. Credit. The credit fund was released in instalments of US$50,000 to US$90,000 from early 1994 onwards and has been lent to groups for on-lending to members, generally for 6 months, but sometimes up to a 12 month term. With recycling, 373 loans have been made to 241 groups and around 2700 individual group members of whom some 43 percent are women. Beneficiary groups have tended to be clustered in a small number of locations in each district. Most loans have been used to finance labor for crop production, especially banana rehabilitation, but also for trading, livestock enterprises, other income generating enterprises such as beekeeping and mushroom growing, and even for land purchase. The value of individual loans commonly ranges between US$100 and US$500, with a maximum of US$1000. Interest rates have varied between 17 percent and 26 percent. 28. SWRARP has achieved commendable recovery rates for its loans (until recently over 80 percent ) by (a) frequent visits by credit officers to groups and where necessary to individual members (b) efforts of local authorities including LC and group leaders and (c) rewarding groups which repay promptly with the offer of repeat and increased loans. The recovery rate is all the more remarkable because, according to SWRARP's own assessment, one third of Phase I and II beneficiaries interviewed claimed not to have made a profit from their loans. However, the pilot credit operation under SWRARP may not be  8 replicable in its current form because of the high costs of training and supervising groups and their members. From its visits, the mission has also noted that credit recipients tend to be the most educated and capable minority of group members and that the scheme may therefore tend to exacerbate rather than mitigate rural income disparities. Nevertheless, it has played an important role in fostering enterprise and demonstrating income generating activities in rural areas and it should be maintained on its present scale if proper institutional arrangements can be put in place. 29. Valley Dams. This component aimed first to rehabilitate 8 existing dams and construct 3 new valley dams and/or tanks. Of the 15 potential sites considered, 4 valley dams and one valley tank were selected, but one had to be dropped due to time and budget limitations. The dams are located in Rwemikoma (rehabilitation), Kantaganya-Engali (new), Nyarubungo-Katerera (new), all in Mbarara district, and Kibutamo (new valley tank) in Ntungamo district. The fifth dam in Mpangamushanju (Mbarara district) will now be proposed for construction under the Livestock Services Project. 30. As of October 1996, progress of work is: Rwemikoma (61 percent ), Kantaganya-Engali (76 percent), Nyarubungo-Katerera (93 percent) and Kibutamo (92 percent). The supervision consultant should see that the contractors complete all remaining work, particularly to ensure adequate embankment strength and that spillways will function effectively (especially at Rwemikoma dam). 31. The valley dams and tanks are intended to reduce water shortages for both livestock and humans and hence reduce nomadism. However, no cost-benefit nor environmental impact studies were made. Both planning and construction appear to have been adversely affected by the time constraint imposed by the late inclusion of the component and relatively early closure of the project. The performance of one contractor has been mediocre. Sustainability will basically depend on the degree of organization of water users' associations and how well they manage to raise funds from members and support from local authorities to maintain/protect the site environment in general and the dam structure and related equipment in particular. 32. Monitoring and Evaluation. The M&E unit carried out a baseline, mid-term and final household survey and 7 thematic surveys which provide valuable information on project impact. Due mainly to staff and resource constraints, the final household survey was, however, simplified (sample and questionnaire), and does not provide quantitative data on agricultural production which could be compared with the results of the 1989 baseline survey. The unit did not carry out regular financial (cost) monitoring. Although recommended in the MTR, low priority was given to the monitoring of agricultural activities, and no analysis was carried out on the financial viability or impact on labor demand of the recommended crop packages and income generating activities (e.g., mushroom, poultry and heifer). 33. Management Support and Technical Assistance (TA). During the first years of implementation, there were difficulties in coordinating activities and managing the project. Adequate office accommodation was also not available until 1992. Restructuring of the project and changes in senior management before the MTR greatly contributed to overall project achievements. Better collaboration with local authorities also resulted in enhanced political and administrative support from the districts and the Inter-ministerial Coordinating Committee (ICC) played an increasingly effective role. 34. Some 350 man-months of long and short term TA were hired by the project. The main fields of assistance included civil engineering, project management, research and extension. Rapid turnover of key TA staff at the beginning of the project contributed to the early implementation problems which might have been solved earlier, had GOU and the Bank taken prompter decisive action.  9 Lender and Borrower Performance 35. Both Borrowers and Lenders contributed to the problems which hampered the early implementation of the project as well as to the resolution of the difficulties around the time of the Mid- Term Review. At appraisal, local tender procedures appear not to have been fully understood, and there were over-optimistic estimations of the time required to establish logistics for project implementation and input procurement (combined with the lack of contingency plans in case of delay). The lack of detailed implementation guidelines for project management was also an important omission. During project implementation, lack of GOU counterpart funding, delays by participating institutions and project management problems (including failure to adhere to established GOU procedures) hindered progress. There were also procedural delays by the Bank. Bank supervision was helpful in highlighting and solving project issues with GOU authorities. In particular, the Bank's prompt action on agreeing a variation order on roads contracts was crucial to the progress of the component. The performance of the minor project components might, however, have been improved if the Bank had provided more technical supervision. Both lenders exercised good understanding and flexibility in adapting financing arrangements to GOU and changing project needs. The support of the Uganda WB office was greatly appreciated by project management - especially the institution of a financial screening service for claims and, towards the end of the project, assistance with procurement procedures. Project Impact 36. It is impossible to quantify the impact SWRARP alone has had in the project area because of the large and very generalized economic changes over the period, complementary development interventions, and because of the shortage of data. Also, a large part of the road rehabilitation program has only been completed very recently and some of its potential impact has perhaps not yet been realized. The mission believes, however, that SWRARP's overall impact has been substantial and positive. The road program has already generated an increase in general economic activity and agricultural production through opening previously impassable roads to traffic and reducing transport costs. This has been verified by surveys showing increased traffic volumes on both feeder and trunk roads in the area. Together with improved access to markets, improvements in productivity of bananas and Irish potatoes resulting from the Adaptive Research and Extension component have boosted incomes and food availability almost region- wide. The Community Development and Credit component has had a more localized impact on incomes and nutrition through its promotion of diversified enterprises and has established foci of local management capacity. It may, however, have had the effect of further concentrating, rather than spreading, the distribution of resources in some rural communities. Project Sustainability 37. Of the four field components of SWRARP, one - Adaptive Research and Extension - has been merged with national programs supported by WB projects (ARTP and AEP). This has reportedly resulted in some dilution of activities in the project area but is in line with prevailing Government and WB policy. Agricultural input supply was intended in the SAR to be taken over by the private sector. Although not entirely in line with the pattern of wholesale dealing or on the scale envisaged, this did, in fact, take place spontaneously during the course of the project and satisfactory private sector arrangements do appear to be in place: private suppliers of agricultural and veterinary inputs operate in centres throughout the region. 38. Maintenance of both contract and spot improved roads is imperative but only a few roads have undergone routine maintenance so far. GOU now gives the highest priority to road maintenance and has increased its budget allocation for rural roads almost four fold in FY97. However, this is still below the  10 country's requirements and districts are not able to match MOLG allocations as is required. Road maintenance equipment purchased under SWRARP for 4 districts has been spread over the 6 districts of the project area. Government is in the process of procuring additional equipment worth US$ 4 m for the whole country, of which some will be for the project area, but this will still not raise the districts' maintenance capabilities to the level originally intended by the Bank under the project. Other measures which would contribute to road maintenance include persuading farmers not to cultivate their fields up to the road edges, limiting axle loads of lorries, and rehabilitating trunk roads. 39. The project was directed to terminate its credit activities from mid-February 1996, and only loan recovery activities have been pursued. Funds recovered are to be transferred by SWRARP to the Bank of Uganda. In January 1996, the South West Uganda Agricultural Development Agency (SWUADA) was established as an NGO. It embraces the SWRARP districts and three other westem districts. Among its intended activities, SWUADA will provide credit to its members, along the lines of the SWRARP credit component. If SWUADA is to successfully take over the credit scheme, it must concentrate its efforts on this undertaking in SWRARP districts only, and receive assistance for building its technical and managerial capacity. Future Operations 40. Although IFAD and IDA loans were closed by mid-1996, SWRARP activities will continue during FY97, solely under GOU financing. This is commendable and reflects the GOU's interest in completing project activities, such as feeder road rehabilitation and maintenance, and credit group training. The project is planned to be dismantled by mid-1997, and part of the future operations should be directed towards handing over project activities to relevant organizations. The mission confirms that it discussed the FY97 project operations with the project management, of which an outline is attached to this Aide Memoire. Lessons Learned 41. When borrower and lenders agree to add a significant component at the end of the project (e.g. small valley dams), appropriate additional time should be given to (i) allow the preparation of feasibility studies including economic and environmental aspects; and (ii) avoid having to transfer project staff resources from other on-going activities. In these circumstances, it is especially important that only competent contractors are engaged to undertake the work. 42. In addition to screening of new/exotic crop varieties, consideration should be given to local varieties, especially to the reasons for their adaptation to prevailing farning systems, and to their improvement through proper seed production (and possibly mass selection), and better crop husbandry practices. 43. To yield significant impact, the release of superior varieties should be accompanied with sustainable seed production systems. 44. More attention needs to be devoted to devising satisfactory logistical arrangements and contingency plans at appraisal. Where rapid start-up of project activities is expected to be crucial to the impact of the project, the appraisal team should include an experienced project administrator who might, in addition to investigating conditions on the ground and defining implementation schedules, also set in motion the procurement process by drawing up draft tender documents.  11 45. Project baseline and terminal surveys should be contracted out to external specialist organizations, so as to allow project M&E units to concentrate on routine M&E activities, reduce overhead costs, permit independent evaluation of project impact and build centres of technical excellence in the conduct and analysis of surveys and other forms of socio-economic investigation. 46. All avenues should be pursued to optimize the choice of candidates for key technical assistance posts. It is often the practice to require companies to distinguish between their own staff members and contracted consultants in their bids. They might further be required to include, in individual CVs, the employment record of the candidates concerned and/or otherwise demonstrate how familiar they are with the performance of the same. Perhaps penalty clauses for consultants should be tightened. (Kampala, 21 October 1996)  IMPLEMENTATION COMPLETION REPORT REPUBLIC OF UGANDA SOUTH WEST REGION AGRICULTURAL REHABILITATION PROJECT (CR 1869-UG) APPENDIX B EVALUATION SUMMARY OF GOVERNMENT'S COMPLETION REPORT   REPUBLIC OF UGANDA SOUTH WEST REGION AGRICULTURAL DEVELOPMENT PROJECT IFAD LOAN SRS-010 UG - IDA CREDIT 1869-UG PROJECT COMPLETION REPORT EVALUATION SUMMARY Project Objectives:- The objectives of the South West Region Agricultural Rehabilitation Project (SWRARP) were to increase food production, farm incomes and standards of living of small farmers in the region. The objectives were to be achieved by the provision of basic farm inputs, support for adaptive research and extension and relevant institutions and rehabilitation-of rural feeder roads. The revitalization of small scale agricultural production was viewed as a means of boosting production to ensure food security, generate marketable surpluses and increase per capita income. The main constraints were lack of inputs, inefficient or weak agricultural research and extension services and lack of access, or very high transport costs which prohibited access, to produce markets. The Project Strategy was therefore to achieve the above objectives through the provision of inputs, adoption of improved technologies from adaptive research, through extension and rehabilitation and spot improvement of feeder roads. Four economically disadvantaged districts, which later increased to six, with redemarcation of district boundaries, were targeted. Implementation Experience The Loan and Credit became effective in August 1988, however, the project took off a year later in September 1989 when all counter part and TA staff were in post. This time lost, coupled with unsatisfactory performance of some TAs, delays in procurement and inadequacy of counterpart funds up to mid 1992, led to the closing date of the project being extended twice, involving a total of 20 months. The project closed at the end of February 1996. It is estimated to have cost US $ 23.445 million which is less than the appraisal estimate of US $ 27.4 million. Of this GOU contribution is US $ 4.74 million equivalent. The procurement of inputs was considerably delayed due to an inappropriate procurement arrangement which bypassed CTB, inadequate knowledge of ICB procedures by UCB and project staff, lengthy and bureaucratic delays in approving lists of items to be procured as well as delays in approving and award of tenders. The  2 above constraints were inherited by the Input Supply Unit after the contract with UCB terminated in June 1993. As a result a US $ 1.0 million order of Inputs initiated in 1992 was cancelled, as it was caught by the February 1996 deadline. This left the same amount unspent. The project incurred losses and low sales due to foreign suppliers who on some occasions delivered goods differing from those specified both in quality, quantity and physical specifications. This was occasioned by an international firm contracted by GOU to carry out preshipment inspection, rendering services in an unsatisfactory manner. The adaptive research on-station activities were limited to 2 sites namely, a research station and a research farm, excluding 2 sites at the 2 DFIs. This was due to insufficient counterpart funds. On-farm trials were carried out on ,selected contact farmers holdings in seven priority counties throughout the region. There was a shift from contact farmers to contact groups after Midterm Review to facilitate faster adoption. A store was erected at each of the research institutions. Institutional capacity building was supported by staff training abroad and within the country. This was in the form of short courses and study tours and longterm courses leading to M.Sc degrees. A total of 40 short courses were attended and 17 staff received higher degrees. The project also assisted with transport, operating funds and office equipment for extension related activities. A media resource centre for production of extension messages, transport, classroom and dormitory furniture were provided at each of the 2 DFIs. Each District acquired a road maintenance Unit made of up road equipment. The project was restructured in 1991 to introduce a women organisation unit and credit scheme and a Dam/Valley tank component in 1994. Under the project's Women Organisation Unit, female extension staff were accorded transport and demonstration equipment for teaching women. The Credit Scheme has generated keen interest and there is greater demand for loans than can be met. Project Results and Impact The overall impact of the project can be assessed through various impact studies as well as the baseline, midterm and terminal household surveys. The aim of the project was to increase food production, marketed surpluses, incomes and standards of living. It is evident that the project has contributed positively to agricultural growth in the area of food crop production and marketing. Through -a revitalized and motivated extension service farmers have been trained in improved husbandry practices for irish potato , banana and wheat production. Adaptive Research has replaced low yielding varieties of potatoes with yields of 2 - 8 tons per hactare, with improved ones yielding 15 - 30 tons per hectare. Three new varieties viz Kisoro, Kabale and Victoria have bean released by the National Variety Release Committee and adoption rates are very  3 high. However demand for improved seed is still very high and distribution problematic, largely depending on individual initiatives. The sustainability of this impact depends on research releasing improved seed every 3 - 4 years. The banana weevil which was threatening to wipe out the industry has been brought under control using cultural methods - which is a low cost sustainable approach. Improved varieties of wheat were released to farmers. However, wheat production is constrained by inappropriate largely monopolistic marketing. The market is not guaranteed and therefore unreliable and prices very low. Upland rice was introduced in Rukungiri where adoption has been very high and 60 tons are produced in one subcounty of Nyamirama per year. Further expansion of production is hampered by lack of mechanised milling facilities. Input Supplies have supplemented improved varieties and husbandry to step up production, arising from, increased yields per unit area (ie productivity) as well as increased acreage. Inputs are widely available and prices more stable compared to the preproject situation. Ownership of hand tools by households has increased as well as use of chemical inputs. In terms of capacity building, the institutions associated with the project have benefitted from the training programme, transport and furniture and equipment provided by the project. The project, however did not carry out any Civil Works on the DFIs other than demolishing one dormitory and making bills of quantities and technical drawings for the dormitory and other works at both DFIs. Inspite of road rehabilitation starting rather late, evidence from traffic surveys reveals a sudden increase in traffic (most of it lorries and pickups), reduction in vehicle operating costs and average journey times. A lot of produce is finding its way to markets and producer prices have generally improved. Ownership of bicycles, a major means of transport has greatly increased as well as bicycle traffic. At grassroot level, about 500 groups were mobilised and trained and many are engaged in income generating activities. On the impact of credit, information available from beneficiaries reveals that their incomes have improved as a result of which they are sending their children to school and their purchasing power has generally improved. Overall, the households produce and sell more and are more food secure than previously. Family nutrition has also largely improved through training offered to women's' groups and incorporation of milk in diet, which is from zerograzed heifers. Livestock ownership is an important indicator of wealth and as such it is one of the proxy indicators of income. The terminal household survey reveals that the number of  4 and 1995. The surveys reveal that households on the whole consume more meat, own more radios and time pieces than in 1990. On standards of living, there is evidence that households own and live in more metal roofed houses than when the project started. Based on the above performance and impact, it is in order to state that the project has performed quite satisfactorily in pursuit of its objectives. However it is not possible to say how much of the impact can be attributed to SWRARP as other projects were implemented concurrently with SWRARP.. These include a highway roads project, an ILO/UNCDF funded feeder road project and South West Integrated (Health and Water) Project (SWIP), and Credit schemes (Entandikwa, PAP, NGOs). Other than the Impact from agricultural components (adaptive research, extension and input supply) where SWRARP was the sole or main actor, the impact accruing from improved transportation, community group mobilization and credit is attributed to all the players, not SWRARP alone. Sustainability: The main risk to the sustainability of activities supported by SWRARP is the shortage and/or untimely release of funds by the successor organisations to implementors. This is true of research and extension. Feeder Road maintenance is largely the responsibility of District Administration. Judging from the amount of resources allocated to roads by districts under decentralization, certainly roads are not given the priority they deserve and are likely to deteriorate once again. However, some feeder roads will be maintained under the ADB-funded project in the short run. Input supply is being carried on by the private sector although there is need for government to institute regulatory and quality controls. On credit, farmers groups have formed an agency (SWUADA) to act as a rural financial intermediary under the new national credit policy. However, there is uncertainty as to whether banks in the project area are willing to take on the role of Participating Commercial Banks (PCBs). Findings and Lessons Learnt The main findings and lessons learnt from implementing this project are of a design and operational nature.  5 Design Considerations: (i) Induction Prior to Implementation Commencement A workshop should be held at the start of the project involving all key project personnel to be involved in the implementation and donor personnel who supervise the project. The main purpose of the workshop should be to ensure that: (a) all staff understand the roles of each component of the project to minimise digression and possible role conflicts; (b) all staff understand the legal requirements of the project including assembly of data for purposes of monitoring and evaluating the project during and after implementation; and (c) everybody understands the financial implication of strictly adhering to the approved work programmes and budgets. (ii) Flexibility of Project design Project design can be influenced by changes in government policies and modification of strategy to better realise objectives. It is therefore important that sufficient flexibility be inbuilt at design stage to allow for later changes. (iii) Proper Scheduling of Civil Works In integrated projects where" civil works such as roads, buildings, water reservoirs are to be constructed to facilitate other project activities, it is crucial that they start ahead of other activities so as not to delay them. This will enhance maximisation of beneficial impact on beneficiaries. - (iv) Women tailored programmes An important lesson from SWRARP is that women have a big role to play in development provided programmes are tailored to meet their needs. (v) Cross Project Transfer of Activities Experience has shown that activities initiated under one project should be completed by the same project as they risk suffering from further delays if transferred to another project or not be implemented at all.  6 (vii) Role of Local leaders Local leaders namely LC Officials and Chiefs can play an important role in project implementation if sensitised. Their good will in mobilising beneficiaries and enforcing bye-laws and regulations can make the difference between a successful and disastrous project. This is especially so in recovery of loans in a credit scheme. (viii) After sales service It is important that in future projects of a similar nature, involving procurement of plant/equipment under ICB, the firms or their agent should be evaluated for their ability to offer sound after-sales services. (ix) Capacity considerations in Contract award The capacity of some contractors, notably the local ones may not stand up to the magnitude of work in major civil works. As such, selection should be guided by the contractors capacity to undertake the works. (x) Road maintenance to avoid economic wastage Road rehabilitation is a very costly venture which should be entered into only if maintenance is guaranteed. (xi) Credit Sustainability The failure of UCB to renew its agency agreement with SWRARP on cost grounds and Cooperative's Bank refusal to participate as PCB may be signal that they are not ready to participate in the new credit delivery mechanism. It is important that GOU does not leave rural people without proper alternative credit sources as arrangement are put in place. May 1996  IBRD 20814 UGANDA \ 33 Z SOUTHWEST REGION - / D AGRICULTURAL REHABILITATION PROJECT PROJECT AREA _ \i ta 7 & a/ KoobonaJ Project Area / MOYD KITGUM / N) Mcor Roods J