Mozambique - Agricultural Rehabilitation and Development Project (ARDP)

Document of The World Bank FOR OFFICIAL USE ONLY Report No: 19416 IMPLEMENTATION COMPLETION REPORT MOZAMBIQUE AGRICULTURAL REHABILITATION AND DEVELOPMENT PROJECT (ARDP) (Cr. 2175-MOZ) June 9, 1999 Rural Development Operations Eastern and Southern Africa This document has a restricted distribution and may be used by recipients only in the performance of their official duties. Its contents may not otherwise be disclosed without World Bank authorization.  CURRENCY EQUIVALENTS Currency Unit = Metical (MT); pl. Meticais At Appraisal US$1 = MT 935 (July 1990) At Completion US$1 = MT 11,270 (June 1997) FISCAL YEAR OF THE BORROWER Government of the Republic of Mozambique January 1 - December 31 ABBREVIATIONS AND ACRONYMS ADB - African Development Bank AFTR1 - Africa Technical Division Rural Development Group 1 ARDP - Agricultural Rehabilitation and Development Project ASRDP - Agricultural Services Rehabilitation and Development Project BPD - People's Development Bank (Banco Popular de Desenvolvimento) CCADR Agricultural Credit and Rural Development Fund CM - State Cashew Enterprise (Cajui de Mocambique) CNAA - Valuation and Sales Commission CRP - Cashew Rehabilitation Project DEA - Department of Agricultural Economics of MAP DNDR - National Directorate for Rural Development of MAP DPAP - Provincial Directorate of Agriculture and Fisheries GOM Government of Mozambique ICR - Implementation Completion Report INIA - National Agricultural Research Institute of MAP MAP Ministry of Agriculture and Fisheries NCRP - National Cashew Research Plan PROAGRI - Agricultural Sector Public Expenditure Program SAR - Staff Appraisal Report SEC - State Secretariat for Cashew SEHA - State Secretariat for Agricultural Hydraulics SPC - Provincial Service for Cashew UREA - Agricultural Enterprise Restructuring Unit Vice President Callisto E. Madavo Director Phyllis R. Pomerantz Sector Manager Sushma Ganguly Team Leader Miriam E. Brandao  FOR OFFICIAL USE ONLY IMPLEMENTATION COMPLETION REPORT MOZAMBIQUE AGRICULTURAL REHABILITATION AND DEVELOPMENT PROJECT (ARDP) (Cr. 2175-MOZ) Table of Contents Preface ...............................................................i Evaluation Summary ............................................................... ii PART I: PROJECT IMPLEMENTATION ASSESSMENT A. STATEMENT/EVALUATION OF OBJECTIVES .......................................................I B. ACHIEVEMENT OF OBJECTIVES ............................................................... . 2. C. MAJOR FACTORS AFFECTING THE PROJECT ..................................................... S D. SUSTAINABILITY ............................................................... 10 E. BANK PERFORMANCE ............................................................... 10 F. BORROWER PERFORMANCE ............................................................... 11 G. ASSESSMENT OF OUTCOME ............................................................... 11 H. FUTURE OPERATIONS ............................................................... 12 I. KEY LESSONS LEARNED ............................................................... 12 PART II: STATISTICAL TABLES Table 1: Summary of Assessments ............................................................... 14 Table 2: Related Bank Loans/Credits ............................................................... 16 Table 3: Project Timetable ............................................................... 16 Table 4: Credit Disbursements: Cumulative Estimated and Actual ................ 16 Table 5: Key Indicators for Project Implementation ......................................... 17 Table 6: Key Indicators for Project Operation .................................................. 17 Table 7: Studies Included in Project ............................................................... 17 Table 8A: Project Costs ............................................................... 18 Table 8B: Project Financing ............................................................... 18 Table 9: Economic Costs and Benefits ............................................................... 18 Table 10: Status of Legal Covenants ............................................................... 19 Table 11: Compliance with Operational Manual Statements ............................. 24 Table 12: Bank Resources: Staff Inputs ............................................................... 24 Table 13: Bank Resources: Missions ............................................................... 25 APPENDICES A. Mission's Aide-Memoire B. Government's Contribution to ICR Map No. IBRD 29996 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.  IMPLEMENTATION COM:PLETION REPORT MOZAMBIQUE AGRICULTURAL REHABILITATION AND DEVELOPMENT PROJECT (ARDP) (Cr. 2175-MOZ) Preface This is the Implementation Completion Report (ICR) for the Agricultural Rehabilitation and Development Project (ARDP) in Mozambique, for which IDA Credit 2175-MOZ in the amount of SDR 11.9 million (US$ 15.4 million equivalent) was approved on September 6, 1990, and became effective on April 30, 1991. This ICR was prepared by a mission from the FAO/World Bank Cooperative Program and subsequently revised and finalized by Miriam Brandao. The report was reviewed by Sushma Ganguly, Technical Manager, Rural Development 1 (Eastern and Southern Africa). Preparation of the ICR commenced with the mission's visit to Mozambique May 5-19, 1998. This report is based on material gathered from the project files and supervision reports, as well as discussions with Bank and Government of Mozambique (GOM) staff associated with the project. These included staff from the Project Management Unit (established to coordinate the project's implementation), and Government of Mozambique staff associated with the project, including the State Secretariat for Cashew, the Directorate of Economics, the National Directorate of Agricultural Extension, the National Agricultural Research Institute, and the People's Development Bank. Information was also gained from field visits to the provinces of Inhambane and Gaza, where discussions were held with staff of Provincial Directorates of Agriculture and Fisheries (DPAPs) and the Provincial Services for Cashew. The mission visited the Nhacutse, Chizavane and Inhamussa nurseries, the Nhacoongo Experimental Station and cashew growing areas in the above provinces, as well as the Ricatla Research Station in Maputo province. The Government contributed to the preparation of the ICR through discussions on the mission's Aide-Memoire. Government's own contribution to the ICR is provided on Appendix B.  ii IMPLEMENTATION COMPLETION REPORT MOZAMBIQUE AGRICULTURAL REHABILITATION AND DEVELOPMENT PROJECT (ARDP) (Cr. 2175-MOZ) Evaluation Summary Introduction 1. The cashew industry, the largest in the world in production terms and the single highest export earner for Mozambique in the 1970s, had fallen to about 10% of former production levels by the mid-1980s. As an important source of income for over a million smallholder farmers, the potential impact of rehabilitating the industry on farm families as well as on factory employment and on the country's trade balance, was high. 2. The project grew out of analysis undertaken jointly by the Bank and the Government of Mozambique (GOM) in 1986 in four sectors: cashew, cotton, forestry and irrigation. The Agricultural Rehabilitation and Development Project (ARDP) aimed to reverse the declining trends in the cashew subsector through a package of initiatives to support the production and processing of cashew. In addition, it provided funds to undertake some exploratory studies that would form the basis of further planning, particularly in irrigation. The ARDP was viewed as the first phase of a long-term rehabilitation and development program. Statement/Evaluation of Objectives 3. The main objective of the project was to support the Government's efforts to assist the family subsector to reverse the decline in cashewnut production and exportation through selective rehabilitation of physical infrastructure and services, the provision of technical assistance, and the implementation of policy reforms in pricing and marketing. The project was also to contribute to the Government of Mozambique's (GOM's) stated policy of restructuring state and commercial agricultural enterprises, and support basic studies and surveys which were to strengthen and guide development of the agricultural sector once the security situation had improved. These objectives were to be achieved through three main project interventions: (a) rehabilitation and development of the cashew subsector; (b) assistance to state and commercial agricultural enterprises; and (c) specific studies and surveys. Implementation Experience and Results 4. General. The project suffered significant start-up delays. Following effectiveness in 1991, little implementation progress was made until 1993. Since this was the first IDA operation in the agriculture sector since independence, there was no experience with Bank operations and a general unawareness of procedures. The security situation in the provinces, while improving, impeded field work and it was understandable that significant time would  iii be required to rebuild GovernmLent services. Staff were poorly paid and hence poorly motivated. 5. The main period of project activity took place during 1994 and 1995 when notable progress was made in the extension activities and technical assistance. Most of the civil works on the nurseries took place during 1995. However, the availability of counterpart funds became an impediment to project activities in 1994/95 and continued during the remainder of the project's life. 6. Two major policy changes early in the project forced the redesign of the components for the assistance to the cashew processing industry. First, with the change in Bank policy (effective in 1992) not to assist state-owned enterprises, the credit component for factory rehabilitation and the technical assistance component for the State Cashew Enterprise (CM) were frozen. Second, with the privatization strategy adopted by the Government, the role of the Agricultural Enterprise Restructuring Unit (UTREA) was redesigned to provide a technical asset valuation function as an input into the privatization process. Therefore, the project's activities in supporting the restrucluring of state and commercial enterprises were re-cast. 7. As a proportion of budget allocations, extension was the highest cost activity to be supported by the project. The activities centred on the development of an extension network in the provinces where previously none existed, and training and equipping field extension workers to advise on an integrated package of cashew and food crops technology messages. Designed to be a pilot program, the project was able to establish networks in 13 districts in Inhambane and Gaza provinces. Some 160 extension workers were hired and trained by the project, reaching in six years 80% of its target of 200 extensionists. It was expected that each extension agent would provide service to between 500 and 600 farm families when the project was fully established. But, the maximum coverage achieved was 274 families per extensionist, with the average remaining at around 125 families. Factors such as the poor road conditions, the long distances covered, and lack of appropriate means contributed to this low ratio. But the main constraint to the achievement of the obj ectives of this component was the scarcity and unpredictability of counterpart funds and, later, salaries for extension staff. The extension service is considered to have been reasonably successful in its training and capacity building activities in relation to food crops; it was less successful in cashew, where only a small proportion of farmers were successfully taught to graft. The project's extension activities were taken over by the ASRDP when the ARDP closed. 8. The research component involved support to the preparation of the National Cashew Research Plan (NCRP) and support (in the form of vehicles, equipment and materials) to the country's clonal orchards at the Ricatla Research Station. General research objectives, appropriate to address the technical problems in the industry, were set out. However, the budget allocation to this component was very small, representing less than 2% of total component costs. The reason for this low support was that this component was only meant to complement activities being carried out under a project funded by the African Development Bank (ADB) in the northern part of Mozambique (which covered clonal propagation). Assurances were obtained by IDA during negotiations that the NCRP would be formulated and initiated by the second project year, yet the draft Plan was never adopted by the GOM. Because of this low budget allocation, the research component stalled while the components with budgets (extension, nurseries, credit) were able to go ahead.  iv 9. The cashew nursery program aimed to establish 14,000 ha of new stock for the family subsector and 1,600 ha for commercial plantations by the eighth project year. Three project nurseries and one experimental station were to be rehabilitated for the purpose of producing grafted trees, and more than 23 buildings were either built or rehabilitated by the project. But, the target production rates for the component were unrealistic considering that: (i) the nurseries had to be rehabilitated and staffed, (ii) there was no previous experience at this scale of operation in the country, and (iii) the techniques for grafting cashew require high skills. Retarding the progress in the nursery component was the failure to achieve satisfactory success rates in vegetative propagation methods. In addition, problems with contractors resulted in long delays which led to two nurseries not being rehabilitated. And, the irrigation systems envisioned did not become operational during the life of the project due to technical difficulties and poor management. 10. The investment credit component failed totally as no loans were advanced. Initially, there were delays in starting this component because the participation agreement between the Bank of Mozambique and BPD was not signed. Later, support to factory rehabilitation through this component became contrary to Bank policy (1992). The main obstacle for potential borrowers of production and marketing credit was the requirement to provide adequate collateral to the People's Development Bank (BPD) since it was to take the full commercial risk. Attempts by BPD to negotiate guarantees or indemnity arrangements with the Ministry of Finance were unsuccessful. Some 20 proposals were put forward by beneficiaries but were all rejected by BPD on the above grounds. In 1993, the Bank proposed that the project provide credit for farmers' groups. A survey to assess this was undertaken and action was taken by BPD to promote the group approach. In 1995, with still no credit advanced, the Bank proposed that NGOs be involved in the formation of farmers' groups for credit for planting new trees. The group approach failed, not only on the issue of adequate security, but also due to the wariness of farmers to form new financially-based groups of any kind, stemming from their negative experiences of groups and cooperatives during socialist times. 11. Institutional Development. The support to CM was cancelled due to the policy changes mentioned above. In the State Secretariat for Cashew (SEC), local and international consultants provided valuable assistance in financial control, monitoring and evaluation and pricing policy analysis. With moves to close the SEC, this capacity has been transferred to the Department of Economics (DE) in the Ministry of Agriculture and Fisheries (MAP). 12. Agricultural Enterprise Restructuring. Following the policy changes conceming state-owned enterprises mentioned above, the role of the UREA was changed to that of evaluating agricultural state enterprises and clearing legal issues. Starting effective implementation in 1993, 71 enterprises were to be assessed over three years. A review of UREA's performance and constraints undertaken by the Bank in February of 1995, found that the Unit had not met the achievement levels and had cost more than a similar unit in the Ministry of Industry and Energy. The privatization process for agricultural enterprises was facing greater legal issues (related to land tenure) than other sector enterprises as well as design impediments in the privatization process itself. The Bank, therefore, decided to discontinue support to the Unit. The Unit was closed and its capacity was absorbed into the Valuation and Sales Commission (CNAA) in the Ministry of Agriculture and Fisheries. 13. Other Surveys and Studies. The National Cashew Tree Population Survey, and The National Cashew Development Master Plan were prepared to provide a longer-term planning  v perspective, as well as integrating other activities in the subsector. The Irrigation Master Plan concentrated on the river baslins in the southem and central regions of the country and was used as base data for the preparation of a small-scale irrigation project to be financed by the ADB, and the FAO Special program for Food Security. Both activities are now included in the Agricultural Sector Public Expenditures Program (PROAGRI). Achievement of Objectives 14. For the cashew component, the Staff Appraisal Report (SAR) indicated that expected benefit targets of the project were annual incremental production of 10,000 tons of cashew from project beneficiaries in the project areas located inInhambane and Gaza provinces. Project beneficiaries were expected to be about 50,000 smallscale family farmers and 100 new commercial producers of cashew by the eighth year of the project. At the time of project closure, project beneficiaries were about 16,000 smallholder farmers (32% of target) and no commercial producers. Production of raw nuts in the project area was variable over the project period, attributable to many influences resulting from the project itself and from extemal factors. It is not possible to identify the production changes attributable to each of the main influencing factors: drought, improving security and higher farm gate prices brought about by market reform in the industry. The project made a start to restore effective agricultural extension services in the two provinces by hiring, training and equipping some 160 extension workers from 1992-95. However, the planting program, the key to the development strategy behind the extension and nursery components, which were the main investments made under the project, was unsuccessful and did not contribute to increased production. Also, no commercial ifarms were established. The proj ect therefore failed to achieve its production objectives. 15. Due to major changes of both Government and Bank policy concerning support to state-owned enterprises, the component to support these was drastically redesigned and reduced. In .effect, these changes removed the obj ective of supporting the restructuring of state and commercial enterprises from the proj ect. The results of the specific studies contributed to the understanding and guidance of the development of the sector, particularly on the work done on irrigation and the cashew tree survey. Project Technical Design 16. The planting program proposed by the project, using existing genetic material, was appropriate as a short-term measure to start the recovery process. However, it needed to be implemented in conjunction with a longer-term strategy for genetic improvement both for improved productivity and nut quality, and for oidium resistance. In addition, the project design should have also recognized the immediate production losses from oidium, and included a short term strategy againstoidium to bridge the intervening period. Instead of this two-pronged approach, the project relied on obtaining a long-term strategy from theCashew Research Master Plan, which it proposed to finance. Two versions of this Plan were eventually prepared, but the first was not available until the fifth year of the project, and neither version has been formally adopted by Government. 17. Therefore, the project's technical design left major gaps in the strategy to restore the industry. Had the planting program been implemented successfully, some production  vi recovery would have resulted but the trees would have had the same nut quality and susceptibility to disease and pests as before. The project made only passing reference to research and did not assist the Government to address the long-term technical improvements for the industry. The project's rather limited short-term strategy resulted in lost opportunities to make a start to restore, not only the production of raw nuts, but the international competitiveness of the domestic processing industry. Project Cost and Financing 18. The eight-year project was appraised on September 10, 1989, signed in September 24, 1990 and became effective on April 30, 1991. The project costs were estimated at US$ 17.8 million, with the IDA Credit financing US$ 15.4 million. As part of a Bank portfolio restructuring exercise, steps were taken during 1996 to redesign the activities and credit agreements of both the ARDP and the Agricultural Services Rehabilitation and Development Project (ASRDP). The undisbursed balance of the ARDP Credit was cancelled and its extension, research and nursery activities were transferred to the ASRDP. The restructuring was approved by the Bank in May 1997 and the ARDP credit was closed on June 6, 1997. An amount of SDR 7.5 million of the original SDR 11.9 million (63%) was cancelled. Following qualified audit reports of project accounts in late 1996, the Bank placed conditions on further disbursements under the project and started the process of recovery of the funds advanced to the Special Account. The final audit of the project's accounts was received in mid-1998, at which time preparations were begun for carrying out this ICR. Major Factors Affecting the Project 19. The outcome of the project was influenced by following main factors: the inexperience of Mozambican staff with Bank procedures; the technical design of the project; the complex implementation responsibilities; the scarcity and unpredictability of counterpart funds; the failure to achieve satisfactory success rates in grafting; the incomplete rehabilitation works at the specified nurseries; and the inability of potential borrowers of the credit to provide adequate collateral. Project Sustainability 20. The extension, research and nursery sub-components of the ARDP were transferred to the IDA-funded ASRDP. Gaps that ensued during the transition process were covered by the Govemment, which provided basic operational funds. Under the PROAGRI, a 15-year program which is starting in 1999, agricultural services are given high priority and represent some 39% of program costs. Therefore, under the PROAGRI, it is considered likely that these services will be sustained. Proj ect sustainability is unlikely without the PROAGRI. Bank Performance 21. For the reasons mentioned above concerning the technical design of the project, and the complexity of its implementation arrangements, the Bank's performance in identification and preparation assistance is rated as unsatisfactory. During the supervision of the project,  vii the performance of the Bank in assisting the Government to address implementation issues was considered deficient to marginally satisfactory. There seems to have been a substantial disconnect in performance ratings with respect to the development objectives (DO) during the early stages as well as throughout the latter years. Although few of the project's objectives were being attained, the project was consistently rated as either highly satisfactory or satisfactory. With regard to implementation status, the rating was satisfactory for most of the life of the project, which does not help to explain why some components were being cancelled and others were being completely redesigned. On the other hand, the project received a lot of attention and support from the Bank throughout its implementation period. There was a strong presence of agriculturists and extension/agricultural services specialists in Bank supervision teams, contributing to the progress made in these areas during 1994/95. And, in the end, the supervision team also showed a high level of proactivity by recommending the cancellation of the project two years ahead of schedule once it became apparent that the proj ect would not achieve its development objectives. Borrower Performance 22. Although this was the first 'Bank project in agriculture following years of insecurity and was consequently a new exercise for Mozambique staff attached to the project, the continued poor implementation performance cannot be attributed to this factor. The Government's performance in coordinating activities, taking decisions and maintaining the project as a priority is considered unsatisfactory. In addition, the lack of counterpart funds was a major factor negatively affecting the performance of the agricultural services component of the project. And theire is still the outstanding issue of the use of project funds for the payment of ineligible expenditures, which is not yet resolved. Compliance with legal covenants and audits was generally satisfactory. Project Outcome 23. While the project made progress towards the reestablishment of extension networks in selected districts, and made valuable contributions tosubsector analysis and planning through the studies and surveys, the project achieved minimal progress towards the primary objective of reversing the decline in production in the cashew subsector and is rated as having an unsatisfactory outcome. Summary of Findings, Future Operations and Key Lessons Learned 24. Findings. The main emphasis of the proj ect was the restoration of extension (and other agricultural) services in two areas badly affected by internal insecurity that were cashew production areas. The technical base for cashew had also eroded, and the project relied heavily on a program of propagation of existing genetic material through grafting in rehabilitated nurseries. The project design underestimated the difficulties in achieving success in grafting cashew, and also underestimated the importance of research for genetic improvement as the basis for future competitiveness of the industry. Poor coordination, low morale, unavailability of adequate counterpart funds and the failure to complete the rehabilitation works contributed to the failure of the project to achieve its objectives.  viii 25. Future Operations. As mentioned above, there has been little progress attributable to the project in restoring the subsector. Initial delays and a shortened project life have meant that the nurseries will need further funds and technical assistance if they are to make a contribution to the production of grafted seedlings. But support to the development of a cashew production strategy as well as the provision of extension services in the two provinces is continuing under the ASRDP. With the policy of Govemment to abolish the " Secretariats," a new Cashew Institute has been created which will have the main responsibility to guide the development of the industry. But, the Govemment will continue to have a strong role in the provision of agricultural services to the smallholder sector, and considering the importance of the cashew subsector, renewed attention is planned for its development under the PROAGRI. 26. Lessons Learned. The key lessons to be leamed from the project are: * Experience has shown the important role that macroeconomic and agricultural policies play in influencing agricultural performance. In countries where the appropriate macroeconomic and agricultural policy environments are not in place, projects should provide the platform for the Bank to support policy reforms and the fine-tuning of policy regimes. The continued disbursement of project funds should be linked to the effective implementation and maintenance of the policy framework, without which project success cannot be guaranteed. * The technical aspects of restoring this type of subsector must be given sufficient priority in design and implementation to address short-, medium- and long-term production constraints. Careful consideration must be paid to technical assistance, training and research needs which are commensurate with the country's overall institutional capacity. * Careful consideration should be given to ensuring that project objectives and design are based on operational capacity and skills levels of the institutions concerned. Where institutional capacity is weak, project objectives should be modest and project design should be simple. * As the example of the research component indicates, actions/activities fundamental for project development and results should be incorporated in the project design, where accountability is more easily ensured. * Projects are designed to respond to conditions prevalent at a specific time and place, therefore, context is very important. This project was prepared when Mozambique was still at war, and implemented during the early years of the peace process. In its desire to support and promote peace, the Bank needs to carefully consider the prevailing policy, institutional and legal environments, resisting the inclination to create a "special case" mentality, where some of the rules are bent to suit the context. * When it becomes apparent that government funds are not available, as indicated by the continued lack of counterpart funding, Bank project should be  ix scaled back to a sustainable size. In the particular case of agricultural projects, where delayed funding at crucial times can set activities back one whole agricultural season, the provision of counterpart funds on time and on budget is essential to ensure the achievement of results. * When it became apparent that project objectives would not be achieved, the Bank took steps to transfer essential components to another ongoing credit, cancelled remaining funds, and closed the project two.years ahead of schedule. This type of action indicates a very strong level of proactivity on the part of Bank staff and management and should be encouraged.  IMPLEMENTATION COMPLETION REPORT MOZAMBIQUE AGRICULTURAL REHABILITATION AND DEVELOPMENT PROJECT (ARDP) (Cr. 2175-MOZ) PART I: PROJECT IMPLEMENTATION ASSESSMENT A. STATEMENT/EVALUATION OF OBJECTIVES I. The main objective of the project was to support the Government's efforts to assist the family subsector to reverse the decline in cashewnut production and exportation through selective rehabilitation of physical infrastructure and services, the provision of technical assistance, and the implementation of policy reforms in pricing and marketing. The project was also to contribute to the Government of Mozambique's (GOM's) stated policy of restructuring state and commercial agricultural enterprises, and support basic studies and surveys which were to strengthen and guide development of the agricultural sector once the security situation had improved. These objectives were to be achieved through three main project interventions: (a) rehabilitation and development of the cashew subsector; (b) assistance to state and commercial agricultural enterprises; and (c) specific studies and surveys. 2. Cashew subsector rehabilitation and development covered the provision of agricultural services such as extension, applied research, production of new stock, and marketing; investment credit for commercial production, marketing, and the rehabilitation of three cashewnut processing factories; institutional development; and evaluative surveys. Assistance to state and commercial agricultural enterprises was designed to complement ongoing institution-building activities supported by other donors, and took the form of advisory services provided to small- and medium-size state and commercial enterprises in the areas of: developing restructuring policies and strategies; management, administration, and production planning; conducting specific studies and surveys; and personnel training. The Studies and Surveys component was intended to strengthen agricultural sector management capacities, policy making, and the formulation of medium- and long-term development strategies. This component included an Irrigation Development Master Plan, a Cashew Tree Population Survey, a Cashew Development Master Plan, a Survey of Factors Affecting the Agricultural Productivity of Women, and an Organizational Development Study for the State Cashew Enterprise (CM). 3. The project was appraised in August 1990, signed in September 1990 and became effective in April 1991. The project costs were estimated at US$ 17.8 million, with the IDA Credit financing US$ 15.4 rnillion. Following qualified audit reports of project accounts in late 1996, the Bank placed conditions on further disbursements under the project and started the process of recovery of the funds advanced to the Special Account. As part of a Bank portfolio restructuring exercise, steps were taken during 1996 to redesign the activities and credit agreements of both ARDP and the Agricultural Services Rehabilitation and  2 Development Project (ASRDP) with the purpose of cancelling theundisbursed balance of the ARDP Credit, transferring the extension, research and nursery activities to the ASRDP, and then closing the ARDP Credit. The restructuring was approved by the Bank in May 1997 and the ARDP credit was closed on June 6, 1997, two years ahead of schedule. An amount of SDR 7.5 million of the original SDR 11.9 million (63%) was cancelled. 4. The cashew industry, once the largest in the world in production terms and the single highest export earner for Mozamlbique in the 1970s, had fallen to about 10% of former production levels by the mid-1980s. The potential impact of rehabilitating the industry on farm families returning to their farms after years of insecurity, on factory employment and on the country's trade balance, was high. The objectives were well placed within the context of the economic rehabilitation priorities for the agricultural sector that were identified in the Agricultural Sector Survey completed as a forerunner exercise to the preparation of the project. The project was consistent with activities to be financed underIDA's Second and Third Rehabilitation Credits, and other IDA-supported projects in the industrial sector. The project's objectives also complemented those of the Cashew Rehabilitation Project (CRP) in northern Mozambique, supported by the African Development Bank (ADB). 5. The statement of objectives in themselves gave no clear indicators of project performance which was to be ac:hieved. However, for the cashew component, the SAR indicated that the expected benefit targets of the project were annual incremental production of 10,000 tons of cashew, 1,200 tons of maize and 1,000 tons of beans from project beneficiaries in the project areas located in Inhambane and Gaza provinces. The project beneficiaries were to be about 50,000 small-scale family farmers and 100 new commercial producers of cashew. B. ACHIEVEMENT OF OBJECTIVES Cashew Subsector 6. Production of raw nuts in the project area was variable over the project period, attributable to many influences resulting both from the project itself and from extemal factors. Production (tons) Province 90/91 91/92 92/93 93/94 94/95 95/96 Inhambane 3,283 3,206 915 38 5,710 8,912 Gaza 7,065 7,184 1,707 124 8,365 10,231 Total 10,348 10.,390 2,622 162 16,075 19,143 7. Drought during 1992 to 1.994, particularly severe in the southern regions of Mozambique, was a major influence on production levels during the project. In addition, market reforms had a significant impact. At the end of the project, while the marketing of cashew had not been fully liberalized, major steps had been taken by the Government in this direction. A reduced export tax on raw nuts to protect the processing industry remained  3 (down from 26% in 1994/95 to 14% in 1996/97). Starting in 1994, an improved share of export values was reflected in farm gate prices and this obviously had an impact on collection and management. Another contributing factor leading to increased collections of nuts was the improving security situation which enabled farmers to return to their farms and permitted access to the trees for harvest and better management. 8. It is not possible to identify the production changes attributable to each of the above factors. While the project financed technical assistance to give guidance on pricing and marketing policy, specific reforms were not conditionalities of the projectd. The project made a start to restore effective agricultural services in the two provinces. However, the planting program, the key to the development strategy behind the extension and nursery components which were the main investments made under the project, was unsuccessful and did not contribute to increased production. Also, no commercial farms were established. The project therefore failed to achieve its production obj ectives. 9. Due to major changes of both Government and Bank policy concerning support to state-owned enterprises, the component to support these was drastically redesigned and reduced. In effect, these changes removed the objective of supporting the restructuring of state and commercial enterprises from the proj ect. 10. The results of the specific studies contributed to the understanding and guidance of the development of the sector, particularly on the work done on irrigation and the cashew tree survey. C. IMPLEMENTATION RECORD AND MAJOR FACTORS AFFECTING THE PROJECT General 11. Two major policy changes early in the project forced the redesign of the components for the assistance to the cashew processing industry. First, with the change in Bank policy beginning in 1992 not to assist state-owned enterprises, the credit component for factory rehabilitation and the technical assistance component for CM were frozen. Second, with the privatization strategy adopted by the Government, the role of the Agricultural Enterprise Restructuring Unit (UTREA) was redesigned to provide a technical asset valuation function as an input into the privatization process. Therefore, the project's activities in supporting the restructuring of state and commercial enterprises were re-cast. 12. Although the SAR stated that the project was designed in a simple but effective fashion to ensure simplicity in implementation, the implementation of the project was complex in that it involved many implementing agencies for project activities. The overall implementing responsibility lay with the Ministry of Agriculture, specifically with the State Secretariat for Cashew (SEC). The responsibility for the cashew subsector components was to be assumed by a component coordinating unit established in the office of the Secretary of State for Cashew. Others involved were the Provincial Services for Cashew (SPCs), the National Agricultural Research Institute (INIA), the National Directorate for Rural Development From 1995, the issue of processing industry protection became a major point of discussion between the Government, the Bank, and the private sector participants in the industry.  4 (DNDR), the Provincial DirectoraLtes of Agriculture (DPAs); the State Cashew Enterprise (CM), -the Bank of Mozambique (BM), the People's Development Bank (BDP), the Agricultural Enterprise Restructuring Unit (UREA) to be established in DEA, and the State Secretariat for Agricultural Hydraulics (SEHA). The mechanisms of coordinating many agencies proved to be difficult. 13. The project suffered significant start-up delays. Following effectiveness in 1991, little implementation progress was made until 1993. This can be attributed to several factors. Since this was the first IDA operation in the agriculture sector since independence, there was no experience with Bank operations and a general unawareness of procedures. The security situation in the provinces, while improving, impeded field work. While farm families were returning to their previous farms, time was needed for them to build confidence that the peace would hold permanently, an essential element of their taking a long-term view on investing effort and money into on-farm rehabilitation. After such a protracted period of insecurity, it was understandable that significant time would be required to rebuild Government services. Staff were poorly paid and hence poorly motivated. This was exacerbated by the fact that staff engaged by the project were on higher salaries, leading to resentment between coodinators and implementers on the ground. 14. The main period of project activity took place during 1994 and 1995 when significant progress was made in the extension activities and technical assistance. Most of the civil works on the nurseries took place dLuring 1995. However, the availability of counterpart funds became an impediment to project activities in 1994/95 and continued during the remainder of the project's life. 15. The amount originally allocated for civil works in the project was SDR 160,000 (US$250,000 equivalent). The amount actually disbursed under this category was SDR 462,000 (US$ 650,000 equivalent). While the construction of the houses for extension workers was completed, a few buildings at the nurseries and the experimental station were not completed. None of the irrigation systems was completed. Cashew Subsector Rehabilitation Extension 16. As a proportion of budget allocations, extension was the highest cost activity to be supported by the project. The activities centred on the development of an extension network in the provinces where previously none existed, and training and equipping field extension workers to advise on an integrated package of cashew and food crops technology messages. Designed to be a pilot program, the project achieved in establishing networks inMassinga, Morrumbene, Homoine, Panda, Jangamo, Inharrime, Inhambane and Zavala districts in Inhambane and Xai-Xai, Manjacaze, Chibuto, Bilene and Chokwe in Gaza, introducing a methodology based on the training and visit system covering both food crops and cashew. 17. The extension system was designed to be multi-disciplinary and involved the cross training.of extension workers. The training program got under way in 1993/94, providing training on various themes to 1,762 participants (extension staff and farmers) via 83 training courses. Some 160 extension workers were hired and trained by the project, reaching 80% of its target of 200 extensionists. There were cashew specialists from the Provincial Service for  5 Cashew (SCP) to provide technical back-up. Some start-up coordination problems existed between SCP and DPA staff. It was expected that each extensionist would provide service to between 500 and 600 families when the program was fully established. But, by project closure the maximum coverage achieved was 274 families perextensionist, with the -average actually remaining at around 125 families. Factors such as the poor road conditions, the long distances to be covered, and lack of appropriate means contributed to this low ratio. 18. The extension messages on food crops were simple, production oriented techniques on planting, spacing, fertilizer application and pest management. For cashew, the main messages included selection criteria of trees, grafting techniques andoidium. There were some cultural practices but most farmers were already aware of the need for bush clearing. Only a small proportion of farmers were successfully taught to graft, not surprising as the technique was not mastered in Mozambique until the introduction of softwood grafting in 1995. It was also proposed that the farmers organize communal nurseries to produce their grafted trees, but very few of these were actually formed. 19. The main constraint to the full achievement of the objectives of this component was the scarcity and unpredictability of counterpart funds and, later, salaries for extension staff, although Gaza suffered less in this area. Farmers took considerable time to build up trust in the Field Extension Workers and confidence in the lasting peace. In circumstances where visits and field activities were postponed or cancelled due to lack of funds, this process was set back. The cessation of funds from the project starting in 1996 resulted in a reduction in extension staff of about 50% and a run down of the service. With delays in salary payments and insufficient operational funds, staff morale was low. 20. The extension service is considered to have been reasonably successful in its training and capacity building activities in relation to food crops; it was less successful in cashew. These functions were taken over by ASRDP when the project closed. Research 21. The research component involved the preparation of the National Cashew Research Plan (NCRP) and support (in the form of vehicles, equipment and materials) to the country's clonal orchards at the Ricatla Research Station. General research objectives, appropriate to address the technical problems in the industry, were set out and included: identification of pests and diseases; screening of oidium tolerant trees; development of vegetative propagation; improvement in cultural practices; development of integrated pest management methods. However, the budget allocation to this component was very small, representing only 2% of total component costs. The reason for this low support was that this component was only meant to complement activities being carried out under a project funded by the African Development Bank (ADB) in the northern part of Mozambique (which covered clonal propagation). Because of this low budget allocation, the research component stalled while the components with budgets (extension, nurseries, credit) were able to go ahead 22. Assurances were obtained by IDA during negotiations that the NCRP would be formulated in the first project year and initiated by the second project year. But, there were considerable delays in starting the formulation of the NCRP. In September 1993, a draft version of this plan was produced by an external consultant (Milhiero). Following the completion of the draft NCRP report it remained for INIA and SEC to jointly review it and  6 provide an approved final version to the Bank. By the closure of the project, a final version had not been agreed on or adopted by the GOM. 23. The government had previously allocated all research functions to INIA, however, this organization had virtually no specific research expertise in cashew. All available information on cashew was located in SEC, but these resources did not include any technical expertise. Clearly, the production of the final version of NCRP would, at the very minimum, require the close collaboration of SEC and MNIA, and poor coordination of activities was a factor that affected this and other project activities. 24. In addition the commencement of the Cashew Rehabilitation Project (CRP) in 1991, funded by the ADB, may have contributed to some lack of urgency in pursuing the research activities- 'after all, some things were happening'. The CRP project had similar (but more limited) research objectives but the scope of operations was restricted to northern provinces. Although the CRP started in 1991, little progress was obtained before 1995 because of difficulties in achieving acceptable success rates in vegetative propagatioi. Cashew Nurseries 25. Cashew nurseries were to be set up in order to enable smallscale farmers to replant or clone graft new stock and commercial farmers to set up smallscale plantations. The target for the family subsector was to establish 14,000 ha of new stock by the eighth project year. Commercial plantations were expected to be set up on 1,600 ha, at the rate of 300 ha per year for the first four years of the project, thereafter tapering off to about 100 ha per year. Project nurseries were to be set up at the rate of one per year, for the first three years of the project. They were also to assist small farmers with grafting training. Three nurseries and one experimental station were to be rehablilitated (Nhacoongo and Rio das Pedras - later changed to Inhamussa - in Inhambane province and Chizavane and Nhacutse in Gaza province). The target rates for the nursery program were unrealistic considering that: (i) the nurseries had to be rehabilitated and staffed, (ii) there was no experience at this scale of operation in the country, and (iii) the techniques for grafting cashew require high skills. 26. The project did carry out much of its program of construction and rehabilitation of the nurseries and experimental station. More than 23 buildings were either built or rehabilitated in the Nhacoongo Experimental Station and in the Nhacutse Nursery. Multiple problems with contractors resulted in long delays, and in the end there was no time to initiate the works intended for the other nurseries. In addition, the irrigation systems did not become operational during the life of the proj ect due to technical difficulties and poor management. 27. A major factor retarding progress in the nursery component was the failure to achieve satisfactory success rates in vegetative propagation methods. It appears that the grafting issue was only resolved in 1995 by assistance from the CRP project in Nampula when the soft wood grafting technique was introduced. One other factor that affected the performance of the nurseries was that they were also significantly involved with other crops. It is possible that attention to these may have detracted from the cashew effort. 2 Cashew is far more difficult to graft than other tree crops such as citrus, mango, avocado and rubber.  7 Investment Credit 28. The project was to extend investment credit for on-farm development, marketing and factory rehabilitation on commercial terms through participating financial institutions (commercial banks). The People's Development Bank (BPD), the major lender to the rural sector, was to be the manager of the line of credit. 29. This component failed totally as no loans were advanced. Initially, there were delays in starting this component because the participation agreement between the Bank of Mozambique and BDP was not signed. The terms of this agreement were the subject of protracted negotiations between the two agencies. In addition, as mentioned above, support to factory rehabilitation through this component became contrary to Bank policy after 1992, and no loans were made for this purpose. The main obstacle for potential borrowers of production and marketing credit was the requirement to provide adequate collateral to BPD since it was to take the full commercial risk. Attempts by BPD to negotiate guarantee or indemnity arrangements with the Ministry of Finance were not successful. Some 20 proposals were put forward to BPD but were all rejected on these grounds. 30. The design of the credit component was based on unsupported or incorrect assumptions. The project assumed that a new group of commercial farmers growing cashew would emerge during the first four years of the project. For this to happen, investors would have needed to have waited for the reforms in marketing and pricing to take effect (financial retums to producers under price fixing were initially too low to justify new investment), to have proved planting material available, and to provide security to BPD. There was no knowledge of commercial-scale cashew production at the time. Increases in producer prices began to take effect only during 1995 and by the time the project activities were suspended, land tenure, ownership and transfer issues remained uncertain following the war. Land, being state owned, was not suitable security and other collateral was required by BPD. 31. The production base for the cashew industry was the small-scale family farm and one group of the project's beneficiaries was identified as 50,000 small-scale farmers that would require investment credit for planting new trees to achieve increased production. Small farmers were not catered for under the credit component, but the project relied on the operations of the Agricultural Credit and Rural Development Fund (CCADR) to provide the necessary smallholder credit. This fund, established in 1988 and administered by BPD, was designed to provide subsidized credit for rural investment by veteran soldiers, returnees, cooperatives and family producers, producers of cereals and others affected by natural calamities. However, by June 1989, over 70% of the portfolio had been directed towards state enterprises and other politically-influenced lending. During the remainder of 1989 and 1990, the fund was fully disbursed and the repayment experience was unsatisfactory. Therefore, by the beginning of the project, this was unavailable for small cashew farmers, as the project design had anticipated. In 1993, the Bank proposed that the project provide credit for farmers' groups along the lines of the credit component of ASRDP. A survey to assess this was undertaken and action was taken by BDP to promote the group approach. In 1995, with still no credit advanced, the Bank proposed that NGOs be involved in the formation of farmers' groups for credit for planting new trees. The group approach failed, not only on the issue of adequate security, but also due to the wariness of farmers, stemming from their negative experiences of groups and cooperatives during socialist times, to form new financially-based groups of any kind. Underlying these factors, the nurseries did not produce planting material for sale that could have been financed under the credit scheme.  8 Institutional Development 32. The support to the State Cashew Enterprise (CM) was cancelled due to the policy changes mentioned above. In SEC, local and intemational consultants provided valuable assistance in financial control, monitoring and evaluation and pricing policy analysis. With moves to close the SEC, this capacity has been transferred to the Directorate of Economics of the MAP. Agricultural Enterprise Restructuring 33. Following the policy changes conceming state-owned enterprises mentioned above, the role of the UREA was changed to that of evaluating agricultural state enterprises and clearing legal issues. Starting effective operations in 1993, 71 enterprises were to be assessed over three years. Dealing with small to medium scale enterprises, the results of the work were to be passed on to the Valuation and Sales Commission (CNAA) within the Ministry of Agriculture. Larger enterprises were handled directly by the Unit for Enterprise Restructuring in the Ministry of Finance. 34. A review of UREA's performance and constraints undertaken by the Bank in February, 1995 found that the Unit had not met the achievement levels (as defined by the number of enterprises assessed), and had cost more than a similar unit in the Ministry of Industry and Energy. The privatization process for agricultural enterprises was facing greater legal issues (related to land tenure) than other sector enterprises as well as design impediments in the privatization process itself. The Bank therefore decided to discontinue support to the Unit. The Unit was closed and its capacity was absorbed into the CNAA. Other Surveys and Studies 35. Through the Cashew Subsector component, the project financed two studies. The National Cashew Tree Population Survey, completed in 1994, was the first such study undertaken since before independence, and provided valuable data on the age condition and location of all cashew plantings. This is an important document and has been used, and still being referred to, in subsector planning. 36. The National Cashew Development Master Plan was prepared to provide a longer-term planning perspective, as well as integrating other activities in the subsector. There were two versions3 but at the time of closing of the project, a master plan had not been formally adopted by Government. 37. Irrigation Master Plan. This major study concentrated on the river basins in the southern and central regions of the coumtry. This was in response to the higher incidence of drought experienced there, and were the areas that were more accessible during the time of the study, 1993. In the light of policy changes towards a market-based economy following the study, the investment recommendaLtions need to be revisited. The increased participation and responsibility of users for operation and maintenance also needs to be incorporated. The 3 Olivares - "Cashew Production Development Strategy - 1996-2005", 1995 and Mann "Cashew Development Master Plan 1996-2006".  9 study was used as base data for the preparation of a small-scale irrigation project to be financed by ADB, and the FAO Special Programme for Food Security. Both activities are now included in PROAGRI. Technical Aspects of Project Design 38. During over 15 years of insecurity, no replanting of trees took place. A significant number of the trees had been completely abandoned or lost through fire. Reduced numbers and the low productivity of the remaining ageing trees were considered to the main problems to be addressed through a major replanting program. The strategy was to propagate new planting material from a selection of mother trees, both in nurseries, and by way of an extension program to teach farmers to identify better performing mother trees and to graft scions from these trees onto seedlings. Genetic improvement was not considered to be a priority (because it was being pursued under the ADB-funded project) and therefore the project placed low priority on research. The activities of the project included the formulation of a research strategy for the longer term, but there was no budget for the consequent research activities. In addition, since it was known at project preparation that the technical base for the industry had been virtually lost, the project provided little in the way of technical assistance, which only amounted to several short-term inputs totalling six months by a nursery management specialist. This was later extended to 16 months. The Cashew Research Master Plan identified the need for long-term external TA but at the close of the project, no action had been taken on implementing the plan. 39. At the start of the project in 1990, there was little useful data available on the cashew selections in Mozambique and the project target was to establish 15,600 ha of new stock (some 1.5 million new trees) in eight years. In these circumstances, the only real option possible would have been an initial mother tree selection based on a single assessment of high total yield and possibly large nut size. This type of assessment could lead to erroneous conclusions - total yield is as much related to environment, tree condition, and management as it is to genetics, while performance in one year could vary for a number of reasons. In addition, nut size is usually inversely related to kernel size and kernel recovery rates, and selection on nut size only could lead to reductions in these characteristics. The highly desirable characteristics, large kernel size, high kernel recovery rates and high yield, are exhibited in a very small proportion of the gene pool, and it is these characteristics that are crucial for improving the financial results for both farners and the processing industry. This can only be achieved by a scientific program of genetic selection. 40. Oidium was recognized as a factor in Mozambique (and Tanzania) and therefore addressing this problem in both the short and long term was important. The approach taken in Tanzania was to supply chemicals (sulphur) and spraying equipment. There are, however, implications for soil acidity in this approach. This project did not address the problems of oidium, either through short-term measures (chemical treatment) or through longer-term measures (genetic selection for resistance). The research component contemplated the screening of oidium tolerant trees, but, as previously stated, this program was not funded and could not be implemented. 4 An improvement in kernel size from W320 to W240 together with improvement in kemel recovery from say 22% to 28% would improve crop unit values to farmers by 30% and to processors by 30 to 35% in higher values and lower unit processing costs.  10 41. Therefore, the planting program using existing genetic material was appropriate as a short-term measure to start the recovery process. However, it needed to be implemented in conjunction with a longer-term strategy for genetic improvement both for improved productivity and nut quality, and for oidium resistance. In addition, the project design should have also recognized the immediate production losses from oidium, and included a short- term strategy against oidium to bridge the intervening period. 42. Therefore, the conclusion must be drawn that the project's technical design left major gaps in the strategy to restore the industry. Had the planting program been implemented successfully, some production recovery would have resulted but the trees would have had the same nut quality and susceptibility to disease and pests as before. The project made only passing reference to research and did not assist the Government to address the long-term technical basis for the industry (except by funding the preparation of the Cashew Master Plan which was not conclusive on this matter and was never adopted by the GOM). The project's rather limited short-term strategy resulted in lost opportunities to make a start to restore, not only the production of raw nuts, but the international competitiveness of the domestic processing industry. D. SUSTAINABILITY 43. The extension, research and nursery sub-components of the ARDP were transferred to the IDA-funded ASRDP. Gaps that ensued during the transition process were covered by the Government, which provided basic operational funds. Under the PROAGRI, a 15-year program starting in 1999, agricultural services are given high priority and represent some 39% of program costs. Therefore, under the PROAGRI, it is considered likely that these services will be sustained. Without the PROAGRI, sustainability is unlikely. E. BANK PERFORMANCE 44. For the reasons mentioned above concerning the technical design of the project, especially its short-term focus to a long-term problem, and the complexity of its implementation arrangements, the Bank's performance in identification and preparation assistance is rated as unsatisfactory. The design of the investment credit component was also based on unsupported or incorrect assumptions. The project assumed that a new group of commercial farmers growing cashew would emerge during the first four years of the project. For this to happen, investors would have had to have waited for the reforms in marketing and pricing to take effect (financial returns to producers under price fixing were initially too low to justify new investment), to have proved planting material available, and to provide security to BPD. There was no knowledge of commercial-scale cashew production at the time. Increases in producer prices began to take effect only during 1995, and at the time the project's activities were suspended, land tenure, ownership and transfer issues remained uncertain. Land, being state owned, was not suitable security and other collateral was required by BPD. 45. During the supervision of the project, the performance of the Bank in assisting the Government to address implementation issues was considered deficient to marginally satisfactory. On the positive side, the project received a lot of attention and support from the Bank throughout its implementation period. There was a strong presence of agriculturists and extension/agricultural services specialists in Bank supervision teams, contributing to the progress made in these areas during 1994/95. The Bank averaged 2.8 supervision missions  11 per year, with 1-6 people spending an average of 16 days in the field. The project's task management was changed only once during the supervision period, providing consistency and continuity for the Government. Supervision missions were usually composed of technical specialists who sought to provide solutions to long-standing problems. This is evidenced by the fact that in addition to the task manager (first an agriculturist and then an economist) there were credit and/or financial management specialists (three missions), extension specialists (five missions), and others (specialists in agriculture, organizational management, land management, environment, and public service development) who participated in missions. Specialized knowledge of research and cashew issues was provided by agriculturists experienced in research systems and tree crops. And, in the end, the supervision team also showed a high level of proactivity by recommending the cancellation of the project two years ahead of schedule once it became apparent that the project would not achieve its development objectives. On the other hand, there seems to have been a substantial disconnect in performance ratings with respect to the development objectives (DO) during the early stages as well as throughout the latter project years. Although few of the project's objectives were being attained, the project was consistently rated as either highly satisfactory or satisfactory. There was also a disconnect with regard to implementation status. Its rating was satisfactory for most of the life of the project, and this contradicts the fact that out of necessity (including poor performance) some components were being cancelled and others were being completely redesigned. Also, a stronger technical presence of cashew grafting and research specialists could have made a great difference in the overall direction of the project, correcting some important design flaws. F. BORROWER PERFORMANCE 46. Although this was the first Bank project in agriculture following years of insecurity and was consequently a new exercise for Mozambique staff attached to the project, the continued poor implementation performance cannot be attributed to this factor. The Government's performance in coordinating activities, taking decisions and maintaining the project as a priority is considered unsatisfactory. In addition, the lack of counterpart funds was a major factor negatively affecting the performance of the agricultural services component of the project. Compliance with legal covenants and audits was generally satisfactory. G. ASSESSMENT OF OUTCOME 47. The positive outcomes of the proj ect were: * The establishment of extension networks in selected districts that made a start to restoring Government agricultural services after the war; * The contributions made to sector and subsector planning by the surveys and studies; * The contribution that the project made to the analysis behind the reforms of the processing and marketing of cashew. 48. However, the following factors heavily influenced the assessment of the outcome of the project:  12 * poor project design; * delayed and limited implementation progress; * incomplete civil works; * non-achievement of production objectives; * limited progress in tackling the short- and long-term development constraints that affect both producers and processors. 49. The project achieved minimal progress towards the primary objectives of reversing the decline in production in the cashew/-subsector and is rated as having an unsatisfactory outcome. H. FUTURE OPERATIONS 50. As mentioned above, there has been little progress attributable to the project in restoring the subsector. Initial delays and a shortened project life have meant that the nurseries will need further funds and technical assistance if they are to make a contribution to the production of grafted seedlings. But support to the development of a cashew production strategy as well as the provision of extension services in the two provinces is continuing under the ASRDP. With the policy of Government to abolish the "Secretariats," a new Cashew Institute has been created which will have the main responsibility to guide the development of the industry. But, the Government will continue to have a strong role in the provision of agricultural services to the smallholder sector, and considering the importance of the cashew subsector, renewed attention is planned for its development under the PROAGRI. I. KEY LESSONS LEARNED 51. The key lessons to be learned from the project are: * Experience has shown the important role that macroeconomic and agricultural policies play in influencing agricultural performance. In countries where the appropriate macroeconomic and agricultural policy environments are not in place, projects should provide the platform for the Bank to support policy reforms and the fine-tuning of policy regimes. The continued disbursement of project funds should be linked to the effective implementation and maintenance of the policy framework, without which project success cannot be guaranteed. * The technical aspects of restoring this type of subsector must be given sufficient priority in design and implementation to address short-, medium- and long-term production constraints. Careful consideration must be paid to technical assistance, training and research needs which are commensurate with the country's overall institutional capacity.  13 * Careful consideration should be given to ensuring that project objectives and design are based on operational capacity and skills levels of the institutions concerned. Where institutional capacity is weak, project objectives should be modest and project design should be simple. * As the example of the research component indicates, actions/activities fundamental for project development and results should be incorporated in the project design, where accountability is more easily ensured. * Projects are designed to respond to conditions prevalent at a specific time and place, therefore, context is very important. This project was prepared when Mozambique was still at war, and implemented during the early years of the peace process. In its desire to support and promote peace, the Bank needs to carefully consider the prevailing policy, institutional and legal environments, resisting the inclination to create a "special case" mentality, where some of the rules are bent to suit the context. * When it becomes apparent that government funds are not available, as indicated by the continued lack of counterpart funding, Bank project should be scaled back to a sustainable size. In the particular case of agricultural projects, where delayed funding at crucial times can set activities back one whole agricultural season, the provision of counterpart funds on time and on budget is essential to ensure the achievement of results. * When it became apparent that project objectives would not be achieved, the Bank took steps to transfer essential components to another ongoing credit, cancelled remaining funds, and closed the credit two years ahead of schedule. This type of action indicates a very strong level of proactivity on the part of Bank staff and management and should be encouraged.  14 'PART II: STATISTICAL TABLES Table 1: Summary of Assessments A. Achievement of objectives Substantial Partial Negligible Not Applicable Macro policies F I F F Sector policies C 57 7 F Financial objectives F 7 5 F Institutional development L LI L I Physical objectives -I FLI FI Poverty reduction I LI L I Gender issues LI LI F Other social objectives LI I FI Environmental objectives LI L I L Public sector management L -I 1 0 Private sector development I LI 5 I Other (specify) LI LI L  15 B. Project sustainability Likely Unlikely5 Uncertain (1) (/) (4) Highly C. Bank performance Satisfactory Satisfactory Deficient (1) (I) (1) Identification E 7 - Preparation assistance 7 L T Appraisal L g T Supervision 6 o F Highly D. Borrower performance Satisfactory Satisfactory Deficient (/) (1) (1) Preparation F I [ Implementation H C T Covenant compliance T E Operation (if applicable) Highly Highly E. Assessment of outcome Satisfactory Satisfactory Unsatisfactory Unsatisfactory (V) V ) (10(/ Sustainability becomes likely under the PROAGRI. 6 Because of the extensive attention paid to the project by supervision teams, as well as the strong level of proactivity shown by the team in recommending the cancellation of the project ahead of schedule, Bank performance in supervision assistance could be considered marginally satisfactory.  16 Table 2: Related Bank Loans/Credits Loan/credit title Purpose Year of Status approval Following operations 1. Agricultural Services Support the development of 1992 Ongoing Rehabilitation and agricultural services, Development Project (Cr. particularly in relation to the 2337-MOZ) cotton-subsector. Table 3: Project Timetable Steps in project cycle Date planned Date actual/ latest estimate Identification (Executive Project Summary) September 5, 1986 Preparation December, 1992 August 1, 1989 Appraisal September 10, 1989 Negotiations April, 1990 May 29, 1990 Board presentation June, 1990 September 6, 1990 Signing September 24, 1990 Effectiveness April 30, 1991 Midterm review (if applicable) March, 1995 No midterm review Project completion June 30, 1999 June 6, 1997 Loan closing June 30, 1999 June 6, 1997 Table 4: Loan/Credit Disbursements: Cumulative Estimated and Actual (US$ millions) FY91 FY92 FY93 FY94 FY95 FY96 FY97 FY98 Appraisal estimate 2.90 6.30 8.70 11.10 12.80 14.00 14.80 15.40 Actual 1.00 1.91 2.87 3.96 5.81 6.13 6.29 - Actual as % of estimate 34% 30% 33% 36% 45% 44% 43% Date of final disbursement May 6, 1997  17 Table 5: Key Indicators for Project Implementation I. Key implementation indicators in SAR/ Estimated Actual President's Report None Specified Table 6: Key Indicators for Project Operation I. Key operating indicators in Estimated Actual SAR/President's Report Not Applicable Not Applicable Not Applicable Table 7: Studies Included in Project Purpose as defined Study at appraisal/redefined' Status Impact of study l. Irrigation To elaborate water resource Completed Substantial - has Master Plan management studies for the main served as base data river basins in central and southern for subsequent Mozambique, identify major investments constraints in irrigation development, and formulate implementation policies. 2. National To provide reliable data on the age, Completed Substantial - has Cashew Tree condition and location of cashew provided valuable Population plantings and assess the present and data and is still Survey future potential of the subsector. being used in subsector planning 3. National To develop a medium- and long-term Completed Minimal - never Cashew development plan, identify major adopted by GOM Research Plan difficulties and constraints, and (NCRP) formulate policies needed for long- term development, especially regarding pricing, processing, marketing & institutional aspects. 4. National To develop a longer-term planning Completed Minimal - two Cashew perspective, as well as integrating versions produced Development other activities in the subsector. and neither Master Plan adopted by GOM  18 Table 8A: Project Costs Appraisal estimate (US$M) Actual/latest estimate (UJS$M) Local Foreign Total Local Foreign Total Costs Costs, Costs costs Item Cashew 4.27 8.82 13.09 0.36 3.79 4.15 Subsector Rehabilitation Agricultural 0.55 2.94 3.49 0.07 1.14 1.21 Enterprise Restructuring Studies and 0.28 0.94 1.22 0.00 1.32 1.32 Surveys I TOTAL 5.10 12.70 17.80 0.43 6.25 6.68 Table 8B: Project Financing Appraisal estimate (US$M) Actual/latest estimate (US$M) Local Foreign Total Local Foreign Total Costs Costs Costs Costs Source IDA 2.7 12.7 15.4 0.60 5.69 6.29 Government 1.2 - 1.2 0.38 N/a 0.39 Participating Banks 0.7 - 0.7 0 0 0 Credit Beneficiaries 0.5 - 0.5 0 0 0 TOTAL 5.1 12.7 17.8 0.98 5.69 6.68 Table 9: Economic Costs and Benefits The SAR calculated financial returns for improved cashew cultivation on both family and commercial farms, and for investment in factory rehabilitation. The cashew cultivation practices assumed purchased grafted seedlings and fertilizer as the basis for increases in production, which was valued to produce intemal rates of returns in various cost/price scenarios. Since the project supported no factory rehabilitation and failed to produce grafted seedlings in any significant quantity, a recalculation of the financial and economic rates of return from the project was not considered warranted.  19 Table 10: Status of Legal Covenants Mozambique Agricultural Rehabilitation and Development Project Agreement DCA Credit Section Covenant Present Original Revised Description of covenant Comments No: IDA Ttnnt Fulfllbnent 21750 Type status foiluet run Date Date 2.02 (b) C The borrower shall open and None maintain in dollars a special account in a commercial bank on terms and conditions satisfactory tot he Association and in accordance with the provisions of Schedule 5. 2.04 C Payment of Commitment None Charges on the principal amount ofthe Credit to the Association as of June 30 of each year. 2.05 C Payment of Service Charge on None the principal amount of the Credit to the association. 3.01 (a) C Commitment to the Objectives None ofthe Project with respect to funds, facilities and management (staff) as set forth in Schedule 2 ofthe DCA. 3.01 (b) CP To cany out the Project in Activities under accordance with the Part D of the Implementation Program set Implementation forth in Schedule 6 to the DCA. Prograrn have been suspended due to privatization. 3.01(c) C Out of the proceeds of the Credit None for Part A.2 of the Project, relend to BM an amount equivalent to SDR 2,300,000 under the Subsidiary Administration Agreement to be entered into between the Borrower and BM on terms and conditions which shall have been approved by the Association and as set forth in Schedule 4 to this DCA. 3.02 (a) C Except as the Association shall None otherwise agree, procurement of goods, works a.nd consultants' services to be financed out of the proceeds of the Credit shall be govemed by the provision so Schedule 3 to the DCA. 3.02(b) C Procuremern of plant protection None chemical materials under Part A.2 of the Project shall be reviewed and approved by the Association.  20 Agreement DCA Credit Section Covenant Present Original Revised Description of covenant Comments No: IDA Aflmn uUnn 21750 Type status fulfillment Fnlf ent Date Date 3.03 (a) NC 12/31/1990 Legally establish and maintain Establishment of the Sales and Trade Unit within Sales and Trade CM in accordance with the Unit suspended due provisions of Part D of Schedule to privatization 6 to the DCA. 3.03 (b) (i) NC 12/31/1990 Select a Marketing Economist Selection suspended and Economist specialized in - see above. food processing to be assigned to Sales and Trade Unit. 3.03 (b) (ii) C 12/31/1990 Select a Senior Pricing Policy Complied with. Analysis Officer within SEC with the functions and responsibilities referred to under Part E of Schedule 6. 3.03 (b) (iii) C 12/31/1990 Select AERUs None Economist/Agricultural Economist, Management Specialist and Information Systems/Accounting Specialist, with qualifications and experience satisfactory to the Association and in accordance with TOR agreed with the Association. 3.03 (c) C 12/31/1990 AERUto fumishto the None association, for its review and comments, the TOR for the formulation of the policy and strategic investment plan referred to under Part B. 1 of the Project. 3.04 (a) CP 12/31/1993 Carry out and complete the Delays due to studies and surveys referred to intemal political under Parts A.4 and C of the difficulties in the Project in accordance with TOR privatization agreed with the Association and process and funmish to the Association, for its management review, the recommendations of problems in AERU. said studies and surveys. 3.04(b) C 12/31/1993 Carry out said recommendations, None taking into consideration the views provided by the Association. 3.05 (a) CP 03/31/1995 Conduct, with the association, a In-depth review mid-term review to monitor carried out in July progress achieved in Project 1995. implementation and meeting Project objectives, and update Project timetables and cost estimates. 3.05 (b) CP Carry out those Amrendment to the recommendations arising out of DCA. said review in accordance with an action program and a timnetable ageed with the Association.  21 |Agreeinent DCA Credit Section Covenant Present OgWnal Revised Descrhption of covenant Coendents 21750 Type status in Ps(i, b ytA d Date Date | ~~~~3.06 CP 03/31/1919