Tunisia - Northwest Agricultural Production Project

Document of The World Bank FOR OFFCLAL USE ONLY ReportNo. 12817 PROJECT COMPLETION REPORT REPUBLIC OF TUNISIA NORTHWEST AGRICULTURAL PRODUCTION PROJECT (LOAN 2502-TUN) MARCH 9, 1994 Agriculture Operations Division Maghreb Department Middle East and North Africa Regional Office This document has a restricted distribution and mav be used bv recipients only in the performance of their official duties. Its contents may not otherwise be disclosed without World Bank authorization.  PROJECT COMPLETION REPORT REPUBLIC OF TUNISIA NORTHWEST AGRICULTURAL PRODUCTION PROJECT (LOAN 2502-TUN) CURRENCY EQUIVALENTS Currency Unit Tunisian Dinar (TD) Appraisal Year US$1.00 = 0.77 TD Implementation average US$1.00 = 0.86 TD Completion Year average US$1.00 = 0.92 TD WEIGHTS AND MEASURES Metric System GLOSSARY OF ABBREVIATIONS ASAL Agricultural Sector Adjustment Loan (Pret a l'Ajustement du Secteur Agricole) AFVA Agricultural Extension and Training Agency (Agence de la Vulgarisation et de la Formation Agricoles) BNT National Bank of Tunisia (Banque Nationale de Tunisie) BNA Banque Nationale Agricole (National Agricultural Bank, BNT's new name following merger with the Banque Nationale de D6veloppement agricole, BNDA, in 1989) CCDA Central Agricultural Development Commission (Commissariat Central de Developpement agricole) CCGC Central Cooperative for Field Crops (Coop6rative Centrale des Grandes Cultures) CRA Local Extension Unit (Centre de Rayonnement Agricole) CRGR Rural Engineering Research Center (Centre de Recherche du Genie rural) COCEBLE Central Cooperative for Wheat (Coop6rative Centrale du Ble)  FOR OMCIAL USE ONLY THE WORLD BANK Washington, D.C. 20433 U.S.A. Office of Director-General March 9, 1994 Operations Evaluation MEMORANDUM TO THE EXECUTIVE DIRECTORS AND THE PRESIDENT SUBJECT: Project Completion Report on Tunisia Northwest Agricultural Production Proiect (Loan 2502-TUN) Attached is the Project Completion Report on Tunisia - NW Agricultural Production Project (Loan 2502-TUN). Parts I and IlIl were prepared by the Middle East and North Africa Regional Office. Part 11 was prepared by the Borrower. The project aimed to get large, rainfed, wheat-cum-livestock farms to adopt improvements already locally field-tested by agricultural research. T&V-style extension was to be the main tool. Also part of the package were improvements to input supply, marketing, milk processing, credit for on-farm investments, road building, and further research. The project was designed for implementation by existing governmental organizations with coordination by an implementation unit in the agriculture ministry. During the implementation period, the agriculture ministry, including its extension and research units, was repeatedly reorganized. Competition was introduced into marketing and other functions which were public monopolies when the project was designed. The project was not adjusted accordingly. Until late in the period, supervision focussed on disbursements and construction rather than on project impact on agricultural production and farmers' incomes. The Borrower lost interest in the by-then inappropriate project. Government cut its projected financing 68%. 61 % of the Bank loan was canceled. Belgian support for the successful farm management center, however, expanded. Except for roads, there is little evidence of impact or reason to infer impact. Extension workers didn't visit farmers much. Record-keeping was poor and monitoring nonexistent. There is little to sustain. Institutional development has taken a different course from that envisaged by the project. The PCR is frank and informative. Part II provides technical corrections, and emphasizes the negative project impact of institutional changes, weaknesses in staffing and in Bank supervision, importance of monitoring and evaluation, and utility of lessons for the subsequent Research and Extension Project. No comments were received from the cofinancier. No audit is planned. Attachment 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.   FOR OMCAL USE ONLY PROJECT COMPLETION REPORT REPUBLIC OF TJNISIA NORTHWEST AGRICULTURAL PRODUCTION PROJECT (LOAN 2502-TUN) TABLE OF CONTENTS Page No. PREFACE ..................... EVALUATION SUMMARY .............................................. iii PART I - PROJECT REVIEW FROM THE BANK'S PERSPECTIVE 1. Project Identity .1 2. Project Background. 1 3. Project Objectives and Description. 2 4. Project Design. 4 5. Project Implementation. 7 6. Project Results .14 7. Project Sustainability .15 8. Bank Performance .15 9. Conclusions and Lessons Learned .16 10. Borrower Performance ..................................... 18 11. Project Documentation and Data ........................... 18 PART II - PROJECT REVIEW FROM BORROWER'S PERSPECTIVE ........... 19 PART III - STATISTICAL INFORMATION ............................. 24 1. Related Bank Loans .25 2. Project Timetable .26 3. Loan Disbursements .27 4. Project Costs .29 5. Project Financing .30 6. Status of Legal Covenants .31 7. Missions .33 8. Staff Inputs .34 9. Project Results .35 This document has a restricted distribution and may be used by recipients only in the performance of their official duties. Its contents may not otherwise be disclosed without World Bank authorization.   PROJECT COMPLETION REPORT REPUBLIC OF TUNISIA NORTHWEST AGRICULTURAL PRODUCTION PROJECT (LOAN 2502-TUN) PREFACE This is the Project Completion Report (PCR) for the Northwest Agricultural Production Project, for which Loan 2502-TUN in the amount of US$15 million was approved on March 14, 1985. The Loan became effective on June 27, 1986 and was closed six months behind schedule on December 31, 1991. A sum of US$6.2 million was cancelled from the Loan on January 24, 1989 and a further US$2.9 million was cancelled at Loan closing on May 7, 1992. Actual disbursements totalled US$5.9 million or 39% of the originally approved Loan amount. This PCR (Parts I and III) was prepared by the Middle East and North Africa Regional Office on the basis of a country visit in May 1992, a draft completion report and progress reports prepared by the Borrower, the Staff Appraisal Report, the Loan Agreement and a study of the files.   - iii - PROJECT COMPLETION REPORT REPUBLIC OF TUNISIA NORTHWEST AGRICULTURAL PRODUCTION PROJECT (LOAN 2502-TUN) EVALUATION SUMMARY Obiectives and Description 1. The main objective of the project was to assist Government to increase production of cereals and livestock products on private farms and production cooperatives in the plains of the four northwestern provinces of B6ja, Jendouba, El Kef and Siliana, to provide for the rapidly increasing demand for food produce in Tunisia. Increases in production and productivity were largely to be brought about by improved technologies which had been sucessfully tested in the area (replacing some fallow with forage and integrating livestock) and the application of more relevant techniques (such as fertilizer and weed control strategies aimed at the entire, integrated farming system). This was to be achieved mainly through the reorganization and strengthening of extension and research services in the project area and improving supportive infrastructure, provision of training, technical assistance, studies, livestock services, farm machinery testing and credit for on-farm investments and input resale centers. 2. The project was intended as the first stage of a future national program for research and extension. A new unified extension service based on the T&V principles was to be tested under which all extension messages would emanate from extension agents located in rural extension centers; extension deputies at regional level were to coordinate the extension system from subject-matter specialists to field extension agents; and a project implementation unit in the CCDA (Central Commission for Agricultural Development, located at the Ministry of Agriculture in Tunis) was to oversee and coordinate the project's various components and executing entities. There were to be no formal institutional links between research and extension but research staff would be expected to participate in field trials and demonstration plots.  - iv - 3. In addition to the regular extension activities concerned with small and medium farmers, the project was to support, with Belgian grant financing, the establishment of a farm management center to advise large (over 100 ha) farmers and production cooperatives (UCP). Implementation Experience and Results 4. The project was implemented during a period of considerable economic and institutional changes in Tunisia: price and policy reforms under two ASALs, reforms of the credit system under the ASALs and the Fourth Agricultural Credit Project, and five reshuffles within the Ministry of Agriculture which included, starting in 1989, a reorganization of agricultural extension and research, culminating in 1991 with a research and extension project. The original design and set-up of the project was no longer valid following these changes. As a result, a number of components, most notably, extension and credit, could not be implemented as foreseen and the situation became aggravated by the declining commitment of the Borrower. 5. The project did not, as originally intended, serve as a model to help direct the preparation of a National Research and Extension Project. Planning of the extension service was not based on a prior diagnosis of farmers' varying needs and constraints. The principle of a single line of command from central and regional management to local extension agents was not adhered to as the project tried to superimpose its own extension agenda on the existing national extension program. In addition to the organizational problems which worsened with each reshuffle within the Ministry of Agriculture (para. 4.03 and 5.05), the effectiveness of the extension program was hampered by lack of well-motivated and trained staff, the absence of links with research, difficult mobilization of the credit component, due to the existence of other competing lines of credit, and lack of transportation and other means. 6. Consequently, the extension agents had only occasional, irregular contacts with a limited number of farmers, and contact groups were not usually formed; little attempt was made to review whole farms on an agro-economic basis to help farmers increase their financial income through further investments, particularly in livestock; feedback from farmers was not sought and, therefore, the extension agents did not know the impact of their recommendations on farming systems and behavior. 7. As a result of these shortcomings, as well as three successive years of drought from 1987-89, the Project fell far short of its objectives in having a direct effect on farmer incomes and productivity increases. Given the absence of data (para. 6), no economic rate of return calculation was attempted for the project. On the other hand, the Beja farm management center which catered to the extension needs of the large farmers was successful and the Belgian Government has extended its support to the center for another year.  v 8. In addition, even though most of the physical targets were also not achieved due to endless procurement problems, the project has made some improvements in the services to farmers through the construction of extension centers, roads, the provision of research and livestock services, vehicles and equipment. By the end of the Project, only $5.9 million was disbursed (401 of the Loan) mainly due to the low utilization of the large credit component. The problems encountered with the extension approach of the Project have provided valuable lessons which have been incorporated in the design of the National Agricultural Research and Extension Project. Sustainability 9. In order to ensure continuity, effective from Project closing, the research and extension activities initiated by this Project were brought under the umbrella of the Agricultural Research and Extension Project which will continue to strengthen research and extension through provision of facilities, equipment, vehicles, staff and training. The creation of lead agencies, AFVA and IRESA, to manage and coordinate research and extension at the center, with planning, decision making and implementation delegated at field level, will ensure the effectiveness and sustainability of research and extension services. Emphasis will be placed on improving the quality of extension staff, who would be trained to work with groups, and to use mass media and other modern communication methods to reach a wider audience, while reducing costs. Strong links would be developed between research and extension who would conduct diagnostic surveys to better target farmers and determine how best to satisfy their information needs. As farmers increase profits and income with the assistance of extension staff trained in farm management and economics, they would be expected to form cooperatives or associations which could, in the long run, gradually take over responsibilities for extension. This would be in line with Government strategy to progressively withdraw from the provision of extension services in favor of the cooperative and private sectors. Findings and Lessons Learned 10. The project was too ambitious in scope and became impossible to implement once the institutional structure fell apart. Given that weaknesses in design became apparent from the start, the Bank should have recommended an early restructuring of the project and cancellation of the credit component. Bank supervision missions focused mainly on procurement and construction of physical infrastructure and failed to address at an early stage important issues of non-compliance with audit covenants and lack of a project monitoring and evaluation system.  - vi - 11. The lessons listed below have been taken into account in follow-on projects, in particular, the Fourth Agricultural Credit Project (completed) and Agricultural Research and Extension Project (ongoing), and the proposed NW Mountainous Areas Development project (participatory approach and simplicity of design and organizational set-up) and Agricultural Sector Investment Loan (monitoring and evaluation system). Finally, roads components are, in recent years, being implemented systematically by the Ministry of Equipment, and increased attention is being given generally in the Bank to ensuring effective supervision of projects. a) project design should not be over-complex and close cooperation and coordination within and among the various concerned agencies of the Borrower is necessary for effective implementation of multi-component Projects. b) extension personnel should be trained to take account of socioeconomic factors as well as technical considerations in dealing with farmers. In addition, money can be saved and the coverage of scarce staff extended by taking advantage of media and other modern communication methods. c) a strong linkage mechanism between extension and research institutions has to be established and the availability of appropriate technology ensured. d) there is a need to develop a participatory approach to planning which promotes analysis, programming and evaluation by farmers, supported by researchers and extension staff. e) an appropriate project monitoring and evaluation system has to be in place from project start-up, to assess project progress and impact, and take early corrective measures. f) once again, experience has shown that best practices call for not including specific credit components in agricultural development projects, but to rely on the rural banking system to finance credit needs under the projects. In particular, it is counterproductive to have credit components targeted to a limited number of beneficiaries in a given area with terms, conditions and lending criteria differing widely from the terms and conditions of the lead credit system of the country, creating distortions in the credit distribution and recovery system. This lesson has been incorporated in follow-on agriculture projects, for example, the Gabes Irrigation Project which met its credit needs through the Fourth Agricultural Credit Project. g) the Ministry of Equipment should be given responsibility for road components in all sectors. h) the successful implementation of a complex project requires not only continuity of staff but the inclusion of a diversity of skills in the supervision team.  PROJECT COMPLETION REPORT REPUBLIC OF TUNISIA NORTHWEST AGRICULTURAL PRODUCTION PROJECT (LOAN 2502-TUN) PART I: PROJECT REVIEW FROM BANK'S PERSPECTIVE 1. Proiect Identity Project Name: Northwest Agricultural Production Project Loan No. 2502-TUN RVP Unit Middle East and North Africa Region Country : Tunisia Sector Agriculture 2. Proiect Background 2.01 The Project was identified in mid-1983 by a joint Bank/FAO-CP mission following publication of the Bank Agricultural Sector Survey of Tunisia in September 1982. The Project aimed at contributing to the priority identified in the sector study of improving production from Tunisia's extensive rainfed areas which have considerable unfulfilled potential. Investments in the past had mainly favored irrigated agriculture. As the possibility for developing new irrigation areas diminished, increased attention was focused on the development of rainfed agriculture particularly for grain and livestock to provide for the rapidly increasing demand for food produce in Tunisia. Under the Technical Assistance Project (Loan 2197-TUN approved in 1982), several strategies were being developed to increase efficient agricultural production. The Project was to constitute a pilot operation to test the nationwide implementation of master plans for research and extension, farm input distribution, farm mechanization, milk collection (prepared separately from the Technical Assistance project). The project was prepared by the Department of Planning, Statistics and Economic Analysis of the Ministry of Agriculture with the assistance of FAO/CP. The project was appraised by the Bank in January/February 1984, approved by the Board in March 1985, signed in April 1985, and became effective in June 1986. It closed at the end of December 1991, six months behind schedule. 2.02 Policy and Sector Context: In the early 1980's, the need for strengthening the agricultural sector was demonstrated by the decreasing relative importance of the sector within the overall economy and the increasing dependence on imports as a result of the growth in food demand fueled by high Government investment, sharp increases in wages and salaries, and liberal credit policies. Between 1960 and 1988, agriculture's contribution to GDP had decreased from 24 to 13*, employment in the sector had dropped from 54 to 30' and the proportion of agricultural to all exports from 60 to 8%. While these decreases in the significance of the agricultural sector in the economy were largely due to the growth of other sectors, the rate of decline indicated the need for revitalization of the primary  - 2 - production sector. The sector's overall performance had also not matched the continued growth of public investment in agriculture, particularly in the decade between 1975 and 1985. The Project was prepared in 1985 and its implementation started in 1986. This period concided with the launching of the Government's Medium Term Economic Adjustment Program supported by six adjustment loans between 1986 and 1991. The reform program aimed at making the economy more outward oriented, more competitive and increasingly based on the private sector. The Government concentrated its initial sector level reform efforts in agriculture with the development of a Medium Term Agricultural Sector Adjustment Program or MTASAP supported by two Agriculture Sector Adjustment Loans (ASAL) in 1986 and 1989. The ASALs have focused on prices and input subsidies, liberalization of imports and exports of key agricultural products, marketing in two principal subsectors (cereals and edible oils), and strengthening of essential Government support services (including credit, livestock services and research and extension). The NW Agricultural Production Project was part of the Government's strategy to achieve production increases by making more effective use of existing resources. However, it was designed at the end of the VIth Plan, i.e. during a transition period prior to the economic transformation of Tunisia which began under the VIIth Plan. Several of the project components could not be implemented as foreseen because the nature of the issues and constraints they were designed to address had changed. The Government's commitment to the project waned given these circumstances and the disruptions caused by the Ministry of Agriculture's program of decentralization in 1989, which included a reform of the organization of agricultural extension, followed, in 1991, by a major restructuring of research and extension supported by an Agricultural Research and Extension Project. 3. Proiect Obiectives and Description 3.01 Proiect Obiectives: The main objective of the Project was to assist Tunisia to increase production of cereals (mainly wheat) and livestock products (chiefly milk and mutton/lamb) on private farms and UCPs in the plains of the four northwestern provinces of Beja, Jendouba, El Kef, and Siliana. The project was to focus on 18 delegations out of 33 in the area and target larger farmers with 10 ha or more in contrast to the Northwest Rural Development Project (Loan 1997-TUN) which was, at the time, assisting poorer farmers in the mountainous zones of the provinces. Since about 60% of Tunisia's total food imports are comprised of wheat, feedgrains, milk and meat, additional Project production was expected to assist the balance of payments. Farm income was also expected to increase. These objectives were to be achieved by strengthening extension and research in the area and improving supportive infrastructure. 3.02 Project Description: The Project consisted of the following components: (a) introduction of an efficient extension service in the project area to be managed along the lines of the Training and Visit (T&V) system, including (i) construction of 18 new and expansion of 15 existing extension centers (CTVs) and provision of equipment,  vehicles and incremental staff for the CTVs; and (ii) establishment of a farm management center to advise large farmers and UCPs; (b) rehabilitation and provision of equipment for one animal production and two field crop research stations; (c) construction and provision of equipment, vehicles, incremental staff, operating and maintenance expenses for three animal health centers, ten dairy cattle artificial insemination posts, one milk collection center and three animal purchase centers; (d) on-farm testing of new types of farm machinery; (e) upgrading of approximately 110 km of rural roads and repairs, culverts and fords for about another 20 km; (f) provision of local and overseas training for extension staff and a researcher; an advisor for the farm management center; consultant services for extension management, farm machinery testing, research station development, and implementation of a monitoring and evaluation system; and studies of the treatment of straw to improve its value as animal feed, and of the development of service cooperatives in the project area; (g) credit investments for: (i) on-farm development, including the construction and improvement of dairies, cowsheds and sheepfolds, the purchase of tractors, implements, cows, rams, ewes and seasonal inputs, and the establishment of pastures; and (ii) provision of 15 new input resale centers (to supply fertilizer, selected seed, pesticides, herbicides), seven of these centers for the Government's National Cereal Board (OC) and eight for service cooperatives (COCEBLE and CCGC). 3.03 Project Management: The Ministry of Agriculture's Central Commission for Agricultural Development (CCDA), located at the Ministry's headquarters in Tunis, was thought to be the organization best suited to coordinate activities of the separate agencies involved in each component and was given primary responsibility for the project. The head of the CCDA, i.e., the project manager, was to chair a Project coordination committee, responsible for periodic liaison with the various entities involved in project execution. Day-to-day management was to be handled by a deputy project manager (agricultural engineer) leading a project implementation unit (PIU) within the CCDA, including a chief accountant and an agricultural economist. Field management of the project was to be handled by the regional branches of the CCDA--the regional commissions for agricultural development (CRDA)-- located in the four provinces. New staff in each CRDA were to include a deputy responsible for extension and for project management and execution, and  an accountant. The deputy was to coordinate the extension system from subject-matter specialists to field extension agents. Subject-matter specialists from the crop production, animal production, the Department of Assistance to Small and Medium Size Farmers (DAPMEA), and the National Livestock and Range Agency (OEP), located at the CRDA level, were to develop extension messages, help train extension agents and answer specialized questions. For farmer contact, one multipurpose extension agent was to deal with groups of about 250 farmers in place of the old system under which farmers were contacted by several crop specific agents. 3.04 Proiect Cost and Financing: At appraisal, total project cost (including taxes and duties of about US$9.7 million) was estimated at US$37.9 million, of which $16.8 million was in foreign exchange. The Bank Loan of US$15 million was to finance 100* of the foreign exchange cost of the project, excluding seasonal inputs (US$1.6 million) which were to be financed by the National Bank of Tunisia (BNT), and technical assistance for the farm management center to be financed on grant terms by Belgium (about US$0.2 million). The balance of funds was to be provided by the Government, BNT and Project beneficiaries. The Bank Loan was thus to finance 40% of the total Project cost. It was to be made to the Government and utilized by the Ministry of Agriculture (for extension, research, farm machinery testing, Project management and training, technical assistance and studies), by the Ministry of Equipment (for rural roads), and passed on as a grant to the OEP for livestock services and under a subsidiary loan agreement to BNT for credit. 4. Proiect Design 4.01 The Staff Appraisal Report (SAR) did not anticipate any major project risks due to the fact that the technology of cereals/livestock integration to be introduced under the project had been tested at an experimental and demonstration farm, operated with the assistance of the Belgian aid program since 1969, and that some farmers adopting recommended cultural practices in the area had obtained substantial yield increases for field crops and animal products. In addition, the project was to focus on larger farmers who were receptive to advice. 4.02 The Project risk of low adoption rates of improved technology by farmers was, however, underestimated. The SAR failed to identify the absence of a formal linkage mechanism between research and extension as a project risk. With the benefit of hindsight, it is now clear that the timing of project appraisal (1984) was unfortunate, since significant policy and institutional reforms took place between 1986 and 1991, which considerably affected the Project set-up and its intended objectives. The project design was also weakened due to the over-emphasis given initially to resolving complex policy issues (milk and meat marketing and fertilizer distribution margins), delays in preparation of master plans, the top down management approach adopted and the inclusion of a credit component which ran contrary to ASAL reforms.  - 5 - 4.03 Prolect ManaQement: The project was supposed to be managed by a stable team of experienced and competent staff headed by a director entirely devoted to the project. However, the team appointed at the CCDA (a project manager, deputy project manager, agricultural economist and accountant) did not have sufficient authority (they were placed at the same level in the hierarchy as the technical directorates), were not assigned exclusively to the project, and did not have the means (staff, vehicles, telephone) required to supervise the multiple institutions responsible for project components. Their effectiveness was hampered by repeated staff changes caused by 4 reorganizations of the Ministry of Agriculture 1 and lack of an extension coordinator (extension consultants were not hired as covenanted). The first deputy project manager, who was an extension specialist, left towards the beginning of the project and was replaced by an agro-economist. The agricultural economist and accountant left in 1990 and were never replaced. Initially, there were meetings of the coordination committee to bring together the various technical units, but these were not pursued. In addition, the CCDA was subjected to the bureaucratic restrictions of government ministries in Tunisia. It had to obtain approval of the national or agricultural procurement committee for procurement of goods and appointment of consultants, and was often unable to procure vehicles and other equipment because of import restrictions. The lack of a coordinating entity for the project led to enormous implementation difficulties. 4.04 Master plans: While the input distribution and milk collection and distribution studies were completed by appraisal of the project or shortly thereafter, there were substantial delays in starting the preparation of national master plans for research and extension and for mechanization because of difficulties in recruitment of consultants. Therefore, the project proposals dealing with research and extension and farm mechanization did not benefit from the first conclusions of the master plan studies, and the poor design of these components (particularly research) reflects the lack of thorough preparation. 4.05 Policy Issues: During project preparation, it was felt that the project's objectives in increasing meat and dairy production could not be achieved without improved milk and meat marketing facilities. As private and 1/ - In 1987, an extension directorate was created in a new Ministry of Agriculture and Food Processing while the extension component of the project was being managed by the CCDA in the Ministry of Agriculture. - In 1988, the Ministry of Agriculture and Food Processing was dissolved and its functions transferred to the Ministry of Agriculture. - In 1989, under the Ministry of Agriculture's program of decentralization, the CCDA was abolished and the PIU relocated to the Directorate General of Plant Production. - In 1990-91, the Agency for Agricultural Extension and Training (AFVA) and the Institution of Agricultural Research and Higher Education (IRESA) were created to streamline research and extension activities at national level.  - 6 - cooperative initiative was not providing adequate services, the project would, through the concerned public agencies, supply the necessary infrastructure. The marketing facilities (milk collection and input resale centers) would subsequently be operated by a service cooperative. Second, during negotiations, the Government agreed to take all necessary measures in order to ensure that adequate incentives would be given to private dairy processing plants and meat merchants who were faced with unfair competition from the public sector and to improve the timeliness of input distribution (fertilizers) to farmers. Targeted objectives were to be pursued under the ASALs. 4.06 Under ASAL-II, meat and powdered milk imports were liberalized. The Government also eliminated meat subsidies on its own. The consumer price of milk remains heavily subsidized and fixed, although in January 1993, through the First Structural Adjustment Loan (SAL-I) and the Economic and Financial Reforms Support Loan (EFRSL), the subsidy has been extended from only reconstituted milk to include local processed milk as well, improving the competitive position of local milk. Liberalized fertilizer margins have led to greater private activity in the market, increasing reliability of supply for farmers. 4.07 Government parastatals (the OC and OEP) started to withdraw from input and marketing activities in 1986 before the necessary incentives had been put in place to attract private sector involvement. The milk collection center was built under the project but not equipped and the animal resale centers had to be cancelled because no cooperative or private enterprises could be found to manage them in the absence of the public sector. During the project, major milk industrial plants continued to have every incentive to buy imported milk powder to produce reconstituted milk because it benefited from a large subsidy and sold well to consumers. They were thus not inspired to develop a local milk collection network. The failure of meat production activities is more of an enigma as producer prices were liberalized in 1985 and rose rapidly after that (at least until import liberalization in 1992). Moreover, the Government does not intervene to purchase meat in the rest of the country and its lack of participation in the project area should not have had an important impact. The region was traditionally cereal and/or mutton producing, and farmers were not familiar with technical aspects of beef production; under the project, moreover, extension messages focused on crops, to the detriment of livestock, and credit was not available as needed. These factors may explain farmers' inability to compete while other regions managed under similar circumstances. 4.08 Project preparation missions also failed to take into account the negative impact on milk and meat production as the Government embarked on its import liberalization program and farmers had to face world prices and lower subsidies. 4.09 Credit: Initially (1985), Bank financing for credit had been excluded from the Loan on the grounds that the credit needs of the project could be met out of the Third Agricultural Credit Project (Loan 1885-TUN), a future fourth Line of Credit to BNT, or out of Government's subsidized credit  - 7 - program (FOSDA). Another reason for the Bank not to finance credit was the slow progress on agricultural credit issues (interest rates, recoveries) under the Third Agricultural Credit Project at the time. During negotiations of the NW Agriculture Production Project, the Government committed itself to further progress on these issues. On this basis, it was agreed to include credit financing under the Loan. Later, under the Fourth Agricultural Credit Project (Loan 2865-TUN), good progress was made with policy reforms, including interest rate adjustment to market rates and phasing out Government subsidized credit (FOSDA). However, the credit component as designed under the Project contained substantial weaknesses in lending procedures and credit approval process. First, BNT had not been invited to the negotiations of the Loan; thus, the subsidiary agreement signed by Government in October 1986, had to be amended in December 1987, to take into account the revised terms and conditions being worked out under the Fourth Agricultural Credit Project. BNT was also not given primary responsibility for lending decisions, as the credit scheme was in the hands of Government, and it was expected to administer it as any Government subsidized credit scheme which the Bank and Government had agreed to phase out. For this reason, and because credit funds were available under the Fourth Agricultural Credit Project under improved terms and conditions consistent with ongoing reforms, it had little incentive to use the credit funds of the project. As reforms proceeded, there was no more justification for a separate credit component for the project. A decision to cancel about half of the component was, however, only made in January 1989 (para 5.17). With regard to the input resale centers, cooperatives or the private sector could not be found to manage them in the absence of the public sector which had withdrawn (para. 5.17) and those directed credit funds were also not used. 5. Pro-ect Imvlementation 5.01 Due to the time taken to recruit consultants to establish the project accounting system, Loan effectiveness was delayed by about a year. Other problems affecting project start-up included delays in preparation of the sub-loan agreement for credit with BNT and rural road studies, and administrative procedures which slowed procurement progress and selection of consultants. The project never recovered from these initial delays. This was due to weak organization and management, a national freeze on staff recruitment, import restrictions which continued to affect procurement of vehicles and equipment, and slow and inadequate mobilization of the credit component (para. 5.17). These problems, combined with a devaluation in the Dinar, led to a drastic reduction in total project costs (from $38 million to $10 million). In January 1989, a first cancellation of US$6.2 million ($4.5 million from the credit component) took place. The project should have been restructured at the time and the credit component cancelled. An additional US$2.9 million was cancelled at Loan closing, i.e. a total of $9.1 million. In Part II, the Borrower mentions (para. 5) that the Bank refused to honor a few disbursement applications. This was because they (three applications) were received after the closing date.  - 8 - 5.02 Extension Organization: A major weakness for the extension program was the lack of an extension coordinator at central level (para. 4.03 and 5.05). 5.03 At regional level, the deputies who were to coordinate extension were not assigned full-time to the project, since they were also the regional representatives of the plant production department, did not coordinate their activities with those of the separate livestock department and, except perhaps for the deputy in Beja, were occupied by other projects in the area. They were not successful in providing extension agents with the support of a multidisciplinary team and forging links with research. 5.04 At field level, the extension agents were involved in other projects and in a variety of administrative and data collection activities which took over 70% of their time. They visited farmers occasionally whenever they had the time and the means of transport. 5.05 The need to strengthen extension coordination at the central level became important when an extension directorate was formed in a separate Ministry of Agriculture Production and Food Processing in 1987, while the extension component of the project was being managed by the CCDA in the Ministry of Agriculture. Although this new ministry was dissolved in 1988, extension coordination continued to be an issue because of the departure in 1988 of the deputy director who was also an extension specialist. Bank supervision missions made repeated recommendations to appoint, as covenanted in the Loan Agreement, a long-term extension consultant in the project execution unit to assist with extension organization at field level. However, project management recruited instead, towards the end of the project, a short- term extension consultant for training of national staff on extension organization, group communications and skills gap analysis. In 1989, following a reorganization of the Ministry of Agriculture, the CCDA was abolished and the project implementation unit was relocated to the Directorate General of Plant Production which was involved in the preparation of the National Research and Extension Project. This new decentralized arrangement worked better because CRDA extension staff were from the Plant Production Directorate and their commitment to the project grew with their increased responsibility. Finally, in order to further streamline research and extension services, the Agency for Agricultural Extension and Training (AFVA) and the Institution of Agricultural Research and Higher Education (IRESA) were created in 1990 at the central level of the Ministry of Agriculture to help in the planning, control, monitoring and evaluation of research and extension. 5.06 The Extension Approach: The 15 CTVs which existed in the project area were to serve as pilot units where the new extension methodology was to be introduced and tested. As a first step, the CTVs were to make a survey of the area which they covered and select a small number of contact farmers eligible for support under the project (minimum farm size 10 ha). In consultation with the farmers, and with the help of CRDA specialists, farm development plans were to be established and requirements for on-farm improvements mainly for integration of livestock (stables, sheds, etc.) and short-term credit identified. Special credit agents in the CRDA were to  - 9 - assist farmers with preparation of credit requests and direct them to the appropriate credit institution. Practical demonstrations were to be conducted on these model farms and this was expected to attract the participation of neighboring farmers. The contact groups thus formed were to be visited by the extension agents every two weeks. 5.07 In reality, the extension methodology consisted of setting up annually 1 or 2 demonstration plots in 1 or 2 farms with input supply provided by the Government. As was the custom during annual fairs organized by Government, various farmers from the districts participated in these demonstrations about twice during the season seeking advice on harvesting, weed control and fertilizer application. However, demonstrating cultural techniques in a large plot in one or two areas once or twice a year limited the benefits to few farmers. There was little participation of neighbors and contact groups were usually not formed. The system of supplying farmers with seeds, pesticides and sometimes machinery, labor, was not effective since the farmer contributed little besides his land and was likely to be passive to the whole operation. Due to the lack of interaction with farmers, extension agents did not know how much of the technology was adopted and reasons for non-adoption. In addition, little attempt was made to review whole farms on an agro-economic basis to help farmers increase their financial income through further investments, particularly in livestock. 5.08 Thus, the T&V approach whereby only a few farmers were contacted with the expectation that the message would be spread around, did not work. The difficulties in mobilizing the credit component (para. 4.09 and 5.17) and lack of links with research (para 5.18) contributed in large part to this poor performance. 5.09 In addition to the regular extension activities concerned with small farmers, the project included the establishment in Beja of a farm management center to advise large (over 100 ha) farmers and UCPs. This activity was financed by a Belgian Government grant. 5.10 The Beja farm management center was established in November 1986 by the ex-project coordinator for the Belgian technical assistance team in charge of the Fretissa project. The Beja center selected farms based on size (100 ha at least) and interested in technical progress and in livestock/crop integration. The tunisian team consists of eight staff, of which a secretary, a driver, a technician and 5 scientists. The Belgian team consists of 4 of which an agronomist and head of team, a farm machinery, plant production and livestock specialist. In addition to the experts, the grant financed vehicles, computer hardware and software, 5 short term and 4 long term fellowships. Given the wide area covered by the farm management center, separate small advisory units headed by two agro-economists were established in the other governorates, while the main team concentrated on Beja. The Beja center also provided equipment and vehicles to the other 3 centers out of the budget previously allocated for cooperative establishment (para. 5.25). 5.11 During the course of the Project, the Beja center assisted about 75 farms of which most were cooperative farms (UCP). Few of the larger  - 10 - private farmers requested advice as they were reluctant to release cost and benefit data. The center assisted farmers with improved farm planning, budgeting, farm machinery operation and maintenance, and advice on cultural techniques to maximize production (rotating cereals with other crops, better farming practices and integrating crop and livestock activities). The team also assisted with the preparation of credit requests to BNT. Given the high demand for its services, the Belgian Government agreed to extend its support of the farm management center until December 1992. The 1991 annual report of the farm management center mentions that livestock activities suffer from the lack of health care, knowledge of techniques, high costs, since most of the purebred animals are imported, and from the fact that livestock requires continuous work while well-off farmers are often absentee farmers. 5.12 Extension Infrastructure and Means: To allow the mobility of extension staff, the project was to provide 58 cars to staff in the extension centers (CTV), the project coordinators at CRDA level and the Beja farm management center, as well as 4 minibuses for the transport of farmers. Due to procurement difficulties, by the end of the Project, 38 cars and one minibus were purchased. In addition, 10 tricycles were purchased which, when tested by the CRDAs, were not found to be suitable for the road conditions in rural areas. In Beja, the extension agents had the use of the car 2 or 3 days a week but the other CTVs had much less access to transport. Furthermore, some of the project vehicles were used by other agencies for non-project activities and there were gasoline restrictions. As a result, farmers were not visited regularly and often did not know the date scheduled for the extension agent's visits. The lack of transport had a negative influence on the effectiveness of the extension service as a whole. 5.13 Given that certain CTVs were constructed by other projects, only 12 (instead of 18) new extension centers were constructed and 13 (instead of 15) existing ones were expanded. Some of these extension centers are located in rural areas and are called CRAs (Centres de Rayonnement Agricole) to distinguish them from CTVs which are located at the sub-regional or district level and supervise 3 to 6 CRAs. 5.14 The CTVs were also provided as foreseen with audiovisual material and agricultural equipment for demonstration plots. In addition, although not originally included in the project, the 4 regional soil and conservation units were provided with equipment in order to help extension staff with demonstrations to farmers of the benefits of contour strip farming as a means to reduce erosion, increase moisture retention and increase yields. A total of 18,000 ha were treated under the project. 5.15 Extension traininq: The local training of extension agents (45) took place in accordance with the project mainly at the Sidi Thabet agricultural training center. The first training sessions were on agricultural development, planning and credit. Further training sessions included the physical environment, farm management and budgeting, mechanization, communications techniques, extension methodology, farmers' attitudes & objectives. However, most of the training was irrelevant and  - 11 - academic since it did not result from diagnosis of farmers' actual problems and constraints at the field level. 5.16 The overseas training for supervisors and subject-matter specialists did not take place given the difficulties at the time in obtaining approval for foreign technical assistance in Tunisia. 5.17 Credit: In addition to the lack of incentives for BNT to use the credit component of the project (para. 4.09), credit demand by farmers for on- farm investments was low because of the drought of 1988/89 which affected their ability to repay past loans and their willingness/ability to invest, particularly for farm machinery. Despite this slowdown in the rural economy, some cooperatives and large farmers in the project area in good standing with BNT were able to obtain credit under the normal BNT terms and conditions. Some small and medium farmers also benefitted from the FOSDA credit being phased out. In addition, the credit funds directed to the input resale centers were not used. This was because of the OC's withdrawal from input distribution in conformity with ASAL-I and the indebtedness of the cooperatives CCGC and COCEBLE which made them ineligible at the time for credit by BNT or any other bank. As of January 1989, as part of the larger cancellation of $6.2 million (para. 5.01), loan funds allocated to the credit component were reduced from the original $8 million to $3.5 million. Disbursement of the credit component finally started to pick up when the Bank approved in November 1990 BNA's 2 request to utilize the undisbursed balance of funds for credit to refinance loans made for the period 1989 to 1991 to creditworthy farmers eligible under the Project. By the end of the project, a total of about $2.7 million was disbursed for credit, i.e. about a third of the original allocation. Most of the credit was used for purchase of agricultural and irrigation equipment. 5.18 Research: In the design of the Project, formal links between research and extension were not established. There were no contacts between research and extension and no use of joint field trials as foreseen in the SAR. Research priorities continued to be decided at the central level and were not tailored to farmers' constraints and needs. This was a major flaw of the extension system of the Project. 5.19 The Project was to provide equipment and facilities for the research stations in the area, in particular for the animal production research station of the National Institute for Agricultural Research (INRAT) at El Afareg in Beja. The project rehabilitated the El Afareg station by providing a laboratory building and equipment for crop and pasture experiments, constructing two small houses for technicians, farm equipment for general farm operations (tractors and attachments), and constructing a well for irrigation and cattle. Unfortunately, the rehabilitation of El Afareg station was not completed by the closing date. There remains construction of a stable and milking shed which were started in 1990; the lodgings lack water and electricity; and although there is a lab with equipment, there are no 2/ BNT's new name following a merger with BNDA in 1989.  - 12 - scientists to conduct the experiments because of the difficulty to attract people to live in isolated rural areas. Due to lack of staff and operating funds, the head of El Afareg station was unable to conduct any experimental work. 5.20 Farm Machinery Testing: In order to ensure that Project farmers and UCPs are advised of the most appropriate farm machinery to maximimize productivity, the Project was to set up a farm machinery testing facility, provide consultants to help choose the type and source of farm implements to be tested, devise the methodology for on-farm comparative testing and supervise the implementation of the testing program. The Project was also to provide for the purchase of the implements to be tested and small equipment required for the testing program. Given the difficulty in recruiting consultants, a list of agricultural testing equipment was finally prepared in collaboration with CRGR (Centre de Recherche du Genie Rural) by a consultant recruited by the Bank for the appraisal of the Research and Extension Project. However, due to procurement problems, only a part of the equipment, consisting of a moldboard plow, a disc harrow, chiesel, and 2 cultivators, was purchased in 1991 for the CRGR at the research station of El Afareg. This was unfortunate since CRGR had previously done a lot of work and had many prototypes ready to be made and tested. The experimental work on farm machinery and implements with particular emphasis on tillage, seeding and fertilizer application will be done under the Research and Extension Project using a multidisciplinary approach, involving agronomists, soil scientists as well as farm machinery specialists. 5.21 Livestock services: As foreseen, the three animal health centers at Dahmani, Testour and Teboursouk were constructed. Only a part of the equipment and 8 of 10 vehicles planned for 8 existing dairy cattle artificial insemination posts were bought and made available to the centers. The milk collection center was built but not equipped and the animal purchase centers were not constructed due to the disengagement of the OEP from management of marketing centers, and the problems to find suitable cooperative or private enterprises to manage these centers, given their lack of profitability at the time. 5.22 Roads: Under the Project, 110 km of feeder roads were to be upgraded but this program was reduced to 92 km (39 km Beja, 19 km Jendouba and 34 km Le Kef) due to financing of 18 km of roads in Siliana from other sources. The rural roads component was delayed by several factors. The roads study commissioned by the Ministry of Agriculture in 1988 had to be amended because it did not include an economic evaluation, as required in the Loan Agreement, and was then rejected twice by the procurement committee on the grounds that it was the Ministry of Equipment's responsibility. A technico- economic study was finally prepared and the contract was approved in September 1989. Works started in December 1989 but were delayed by land expropriation issues, which had not been taken into account, and by poor weather conditions, particularly in 1990/91. Unfortunately, the poor weather conditions persisted throughout the period of the extension of the closing date of the Loan, and thus, by the end of the Project in December 1991, works were not completed but had stopped. While roads in El Kef were practically completed, only about 70W  - 13 - of road rehabilitation works in B6ja, and 80W in Jendouba, had been executed. The Ministry of Equipment is expected to complete the works. 5.23 Training, Technical Assistance and Studies: Due to the preference given in Tunisia for local technical assistance, in August 1989, the Bank approved an amendment of the Loan Agreement whereby consulting services were to be financed at 100% (instead of 80% local and 100% foreign). However, of the original $500,000 allocation for technical assistance, only $94,000 was disbursed. The low use of consultant funds is a result of the overall poor implementation and design of the project. 5.24 The funding for the local training of extension staff and the farm management center was used as foreseen. The head of El Afareg research station also underwent training for six months at a research station development and management training course in Mexico. However, on his return, this was not followed as planned by a consultancy on establishment of support stations in Tunisia. In addition, a major project weakness was the lack of a long-term extension management consultant to help make the extension service operational. 5.25 Four studies were done under the Project. Two were for the rural roads technico-economic studies (para. 5.22), the third was to set up an accounting system for the Project (para. 5.27) and the fourth, not originally included, was for preparation of the Research and Extension Project. A study of the treatment of straw to improve its value as animal feed was dropped because it had been carried out by INRAT, and the study to assist with the development of service cooperatives in the Project area was to be undertaken by the Ministry of Agriculture with its own funding. Some computer equipment was purchased for the implementation of a monitoring and evaluation system which was never established (para. 5.26). 5.26 Monitoring and Evaluation: The agricultural economist in the PIU was to monitor the physical execution of the project and achievement of production and policy objectives, according to key indicators outlined in the technical support volume of the SAR. Due to the departure of the agricultural economist in 1988 and lack of a replacement, no consistent monitoring and evaluation system was developed. The impact of extension could not be measured because the pre-project situation had not been assessed and data had not been collected in the field. The semi-annual reports submitted by the Borrower were on the status of physical construction, extension training and the financial accounts of the Project. 5.27 Accounting System: The design of a satisfactory project accounting system was a condition of effectiveness of the Loan. After long delays, an accounting firm was selected and financed a study to set up a system. However, the accounting system was not set up as foreseen due to the lack of computer equipment, and overlapping responsibilities of the accountants in the CRDAs and the chief accountant in the PIU, who left in 1990 before the first audit mission took place. A Bank mission in 1990 following up on lack of audit compliance helped the PIU to organize their accounts (no separate accounts were kept for the Project) and the first consolidated audit report  - 14 - was prepared in 1991 covering the years 1988-1990. The audit report for 1991 accounts was provided in early 1993. 6. Project Results 6.01 The project did not, as originally intended, serve as a model to help direct the preparation of a National Research and Extension Project. Planning of the extension service was not based on a prior diagnosis of farmers' varying needs and constraints. The principle of a single line of command from central and regional management to local extension agents was not adhered to as the project tried to superimpose its own extension agenda on the existing national extension program. In addition to the organizational problems which worsened with each reshuffle within the Ministry of Agriculture (para. 4.03 and 5.05), the effectiveness of the extension program was hampered by lack of well-motivated and trained staff, the absence of links with research, difficulty to mobilize the credit component, and insufficient financial means. 6.02 Consequently, the extension agents had only occasional, irregular contacts with a limited nuxmber of farmers, and contact groups were not usually formed; little attempt was made to review whole farms on an agro-economic basis to help farmers increase their financial income through further investments, particularly in livestock; feedback from farmers was not sought and, therefore, the extension agents did not know the impact of their recommendations on farming systems and behavior. 6.03 As a result of these shortcomings as well as three successive years of drought from 1987-89, the Project fell far short of its objectives in having a direct effect on farmer incomes and productivity increases (para. 6.05). On the other hand, the Beja farm management center which catered to the extension needs of the large farmers was successful and the Belgian Government has extended its support to the center for another year. 6.04 In addition, even though most of the physical targets were also not achieved due to endless procurement problems, the project has made some improvements in the services to farmers through the construction of extension centers, roads, the provision of research and livestock services, vehicles and equipment. By the end of the Project, only $5.9 million was disbursed (40W of the Loan) mainly due to the low utilization of the large credit component. The problems encountered with the extension approach of the Project have provided valuable lessons which have been incorporated in the design of the National Agricultural Research and Extension Project. 6.05 Crop and Livestock Production: The main effect expected from the Project was an increase in crop and livestock production and farmer incomes. However, the pre-project situation was not known (no diagnostic surveys carried out) and agricultural statistics in the project area were not monitored, and thus it is impossible to measure any impact of the Project on agricultural productivity or incomes. Based on data published by the Agricultural Statistics Division of the Ministry of Agriculture, production of major foodgrains and livestock products in the 4 governorates of the Project  - 15 - has increased between 1986 and 1991, mainly due to higher yields. However, the Project was meant to cover only a part of the governorates and thus no comparison with SAR projected increases in production and yields could be made. In addition, good results during the last two years of the Project can mainly be attributed to favorable climatic conditions compared to four years of drought from 1986 to 1989, and it is virtually impossible to separate what the Project did from the CRDAs' overall agricultural development program for the NW area at the time. Due to the absence of data, no economic rate of return calculation was attempted for the Project. 7. Prolect Sustainability 7.01 In order to ensure continuity, effective from Project closing, the research and extension activities initiated by this Project were brought under the umbrella of the Agricultural Research and Extension Project which will continue the program initiated of strengthening research and extension through provision of facilities, equipment, vehicles, staff and training. Experimental work on various types of farm machinery and implements which could not be implemented would take place under the Research and Extension Project. Research would continue its work on forage crops, livestock and field crops complementarity and increased attention would be given to socio- economic aspects in order to provide policy makers, extension staff and the farming community with cost and benefit data on proposed new techniques and methods. The new decentralized extension structure with a lean management team at the center and planning, decision making and implementation at field level will ensure its effectiveness and sustainability. Extension staff would be trained to work with groups and to use mass media and other modern communication methods to reach a wider audience while reducing costs. By conducting diagnostic surveys, research and extension staff would identify the various categories of farmers and determine how best to satisfy their information needs. In future, as farmers are better able to manage their assets and make a profit, they could start to become organized as cooperatives or associations and assume increasing responsibility for extension. This would be in line with Government's strategy to increasingly transfer responsibility for extension to the private and cooperative sectors, leaving the public sector to focus on small subsistence farmers in remote areas which are likely to continue to need assistance. 8. Bank Performance 8.01 As mentioned earlier, the Project had some serious design flaws (the credit component, in particular). In hindsight, some of the risks which were to occur to the Project could have been foreseen and avoided by scheduling the Project at a later date once a dialogue on pricing and policies was established with Government (after ASAL-I or II). Although the organizational problems were mostly out of its control, the Bank did not follow up early enough on some important issues, for example, non-compliance with audit covenants and lack of a project monitoring and evaluation system. The serious shortcomings of the project were not brought to management attention until 1991 when a new mission rated the project a "3". After appraisal, a mission composed of an economist and financial analyst was  - 16 - launched to assist with Project start-up and a total of 16 supervision missions were led mainly by an agronomist with minimal participation of livestock, extension or financial staff, and this may have contributed to the poor performance of those components. The missions concerned themselves mainly with disbursement, construction of physical infrastructure and staffing. No attempt was made to assess the project's impact on production and farmers' incomes, disregarding the analysis of project impact. It was only in the last two years of the Project, in the context of the preparation of the Research and Extension Project, that a serious attempt was made to render the extension service more effective by improvements proposed in extension organization and training. The Bank's decision to extend the Loan closing date by six months did not allow completion of the rural roads because of poor weather conditions in winter of 1991. 9. Conclusions and Lessons Learned 9.01 The project was too ambitious in scope and became impossible to implement once the institutional structure fell apart. Given that weaknesses in design became apparent from the start, the Bank should have recommended an early restructuring of the project and cancellation of the credit component. 9.02 The lessons which can be derived from the project are the following: 1) project design should not be over-complex and close cooperation and coordination within and among the various concerned agencies of the Borrower is necessary for effective implementation of multi-component Projects; 2) more effective use of extension personnel will be possible only if they are trained to take account of socioeconomic factors as well as technical considerations in dealing with farmers; 3) in addition, money can be saved and the coverage of scarce staff extended by taking advantage of media and other modern communication methods; 4) a strong linkage mechanism between extension and research institutions has to be established and the availability of appropriate technology ensured; 5) there is a need to develop a participatory approach to planning which promotes analysis, programming and evaluation by farmers, supported by researchers and extension staff. If this is done by homogeneous groups of farmers, it should result in research and recommendations more relevant to the needs of all important sub-categories of farmers; 6) an appropriate project monitoring and evaluation system has to be in place from project start-up, to assess project progress and impact, and take early corrective measures;  - 17 - 7) experience has shown that best practices call for not including specific credit components in agricultural development projects, but to rely on the rural banking system to finance credit needs under the projects. In particular, it is counterproductive to have credit components targeted to a limited number of beneficiaries in a given area with terms, conditions and lending criteria differing widely from the terms and conditions of the lead credit system of the country, creating distortions in the credit distribution and recovery system. This lesson has been incorporated in follow-on agriculture projects, for example the Gabes Irrigation Project which met its credit needs through the Fourth Agricultural Credit Project. 8) the Ministry of Equipment should be given responsibility for road components in all sectors. 9) the successful implementation of a complex project requires not only continuity of staff but the inclusion of a diversity of skills in the supervision team. 9.03 The above lessons have been incorporated in projects which are ongoing or under preparation: the NW Mountainous Areas Development Project for lessons 1 and 5 (participative approach), the Agricultural Research and Extension Project for lessons 2 to 6, the Fourth Agricultural Credit Project for lesson 7. The lack of monitoring and evaluation has plagued the majority of agriculture projects under implementation. The Agricultural Sector Investment Loan proposes to include a study to establish an integrated monitoring and evaluation system for the 1994-1996 investment program of the Ministry of Agriculture. Finally, roads components are, in recent years, being implemented systematically by the Ministry of Equipment (lesson 8) and increased attention is being given generally in the Bank to ensuring effective supervision of projects (lesson 9). 9.04 Follow-up on Lessons Learnt: As a result of the reorganization of the Ministry of Agriculture in 1989, the newly created CRDAs were given a large degree of administrative, financial and executing autonomy. This will ensure that development projects in future will have a more decentralized management and procurement system which should help avoid the bottlenecks encountered by line agencies of the Ministry of Agriculture. 9.05 Following reforms initiated under the ASAL, the Research and Extension Project was prepared in 1990 to further rationalize research and extension at national, regional and local levels. It proposes to cover the first five-year phase of a ten-year research and extension program. The major improvements proposed or undertaken under the Project are the following. First, lead agencies to manage and coordinate research and extension, IRESA and AFVA, were established. A major weakness under the NW Production project, which was the lack of an extension unit at CRDA level providing farmers and extension agents with the support of a multidisciplinary team, would be corrected through the creation of central extension support units (DVPA) in the 15 governorates concerned by the Research and Extention Project. Each DVPA would have three arrondissements, covering, respectively, plant production, animal production and financing and credit. In addition, in order  - 18 - to relieve extension agents of non-extension related duties, the supervisory staff (CTV) would be given more of an administrative role with responsibilities for helping extension agents organize their work and providing them with technical support. The extension agents would be given the single function of transmitting technical information to and from farmers. Strong links between extension and research would be developed so that farmers' problems would be taken to research for solution, appropriate research programs developed, and appropriate technologies taken from research to farmers. Following the skill gaps analysis conducted under the NW Production Project, only highly motivated extension staff with the required qualifications would be recruited and trained. An effective monitoring and evaluation system would be established and staff trained in its use. 9.06 With regard to agricultural credit, the Fourth Agricultural Credit Project with BNT/BNA was implemented in a satisfactory manner from 1988 to 1991 and it addressed institutional and financial improvements at the level of BNT/BNA, the lead bank for agriculture in Tunisia. The PCR for this credit project was issued in June 1992. 10. Borrower Performance 10.01 The Borrower's performance during project implementation was not always satisfactory. The lack of a coordinating entity with sufficient authority, including control over budgetary resources, led to important delays and shortfalls in project implementation. As the project began to be implemented, and it became apparent that its organization and a number of its objectives were in conflict with the Ministry of Agriculture's decentralization program and reforms underway under the ASALs and the Fourth Credit Project, the Government gave limited priority to the Project which aggravated already existing procurement and recruitment problems. The Borrower's compliance with loan covenants was not very good. The requirement to establish permanent posts for extension and key staff in the project implementation unit was not met and a first audit report was submitted only towards the end of the Project. Monitoring and evaluation was also not covered. 11. Project Documentation and Data 11.01 The implementation volume of the SAR is voluminous and discusses in detail pricing and marketing constraints for cereals and livestock. While this was useful background information for the preparation of the ASAL-I, it is apparent from reading the SAR that the Project components were hastily prepared and it is not surprising that at appraisal and during implementation a number of components had to be reduced and/or could not be implemented as foreseen. A major flaw was that the implementation volume included so many monitoring indicators that it was difficult to isolate which were the most important ones to use in order to follow up on progress with the extension system and adjust it if necessary. No data was thus collected to monitor the impact of extension. Semi-annual progress reports on the physical implementation of the project were, however, submitted to the supervision missions.  - 19 - PROJECT COMPLETION REPORT REPUBLIC OF TUNISIA NORTHWEST AGRICULTURAL PRODUCTION PROJECT (LOAN 2502-TUN) Part II: PROJECT REVIEW FROM BORROWER'S PERSPECTIVE  - 20 - PROJECT COMPLETION REPORT REPUBLIC OF TUNISIA NORTHWEST AGRICULTURAL PRODUCTION PROJECT (LOAN 2502-TUN) Part II: PROJECT REVIEW FROM THE BORROWER'S PERSPECTIVE (Translated from the original French version) This is Part II of the Completion Report of the Northwest Agricultural Production Project which constitutes the Borrower's perspective. This Part II is based on the data of a preliminary report prepared by the Borrower and submitted to the World Bank in May 1992, as well as on the Parts I and III prepared by the World Bank and submitted to the Borrower on April 23, 1993. This Part II has been prepared according to the format recommended by the Bank and includes six paragraphs as follows: 1. Confirmation of factual data contained in Part III The information presented in Part III of the report corresponds with that of the preliminary report and reflects the Project's real situation at completion. However, the following remarks are made in relation to Table 6: Status of Legal Covenants: Section 3.01 (b) Annual budget and financing plans were submitted to the Bank regularly and in accordance with the Loan Agreement. Those budgets were allocated to the Project as planned, except for the budget related to the acquisition of vehicles for the extension component, which was reduced after restrictions imposed on vehicle procurement at national level. Section 3.04 (c) The coordinating committee was established in accordance with the Loan Agreement and held regular meetings. However, these meetings were reduced following project management changes. Section 3.05: Employment of management, accounting, and agricultural Consultants The following actions were taken in this context: - organization of a six-month training course for the head of a research station in Mexico (CIMMYT Center) on "Research Stations Development"; - a Belgian consultant has been recruited under the Belgian Technical Assistance to set up the operating methodology of the management center, follow-up on actions taken and give technical and management guidance for the center's activities ; - a study on the establishment of an accounting system for the Project was undertaken by an accounting firm;  - 21 - - Since September 1987, steps were undertaken to recruit an extension management consultant. However, this did not materialize and Project management recruited instead an expert in rural communications who prepared a training program and implemented a training course for 45 heads of local extension centers. This course dealt mainly with communication techniques and extension activities planning. Section 3.06: Management of the Milk Collection Center The milk collection center was built but the World Bank did not give its approval to equip it because of lack of interest of cooperatives or the private sector to manage it. Section 3.11 The implementation of the accounting system which had been designed by an accounting firm required the acquisition of the necessary means, in particular, the applications software. However, the software was not purchased given that project management estimated that its cost was too high. 2. Remarks on the analysis contained in Part I Para 4.03 - Proiect Management The Project was supposed to be managed by a stable team of experienced and competent staff; this team was supposed to be assisted by a coordinating committee composed of the various institutions concerned with the Project. However, due to the lack of an action plan defining the role and responsibilities of each participant, this committee did not have the efficiency required. Para 4.07 The parastatal agencies (Cereals Agency and National Livestock and Pasture Agency) did not in fact start to withdraw from input distribution and marketing until after 1986. Their withdrawal is not yet completed and the Cereals Agency continues, for example, to distribute fertilizers where needed, until such time as the activity can be handed over to the private sector. Para 4.08 In order to mitigate the negative impact of liberalization of imported milk and meat on local production, mititgating measures have been taken by the Government such as the imposition of tariffs as of 1991 on powdered milk and meat imports. Para 5.07 In fact, several demonstration plots and information days are organized annually at the level of local extension centers, which cover crop and livestock production. Para 5.08 The "Training and Visits" method was certainly not fully adopted by all the heads of CRAs. However, thanks in large part to the training program implemented by the Project, some of the heads of CRAs which have the necessary means were able to apply this method successfully.  - 22 - With the establishment of the Agricultural Extension and Training Agency, this method will be used by the other CRAs. Para 5.15: Extension Training The expert in rural communications prepared a training program for the extension agents of the Project, which takes into account the training deficiencies and is best adapted to the extension work. This training was implemented under the form of workshops and practical sessions attended by all extension agents working in the Project area. Para 6.02 An effort has been made by the Beja farm management center to set up development plans for the large farms concerned by the Project. This task has required the participation of a multidisciplinary team at engineer level, and cannot be implemented by the heads of CRAs. 3. Bank Performance While Bank supervision missions were regular and followed the Project from appraisal to completion, they concerned themselves mainly with disbursements and construction of physical infrastructure rather than with the assessment of the Project impact on agricultural production and farmers' incomes. In effect, it would have been beneficial for supervision missions to participate more efficiently in defining actions to be undertaken annually while insisting on the continuous monitoring and evaluation of Project activities. Moroever, for improved efficiency, the profile of experts participating in the missions should have corresponded to the different specialties required to carry out the Project. 4. Borrower Performance The Project was implemented during a period of profound organizational and institutional changes at the level of the Ministry of Agriculture, which seriously hampered its execution. Nevertheless, the Project was able to realize most of the physical actions foreseen in the Loan Agreement (construction of CRAs, rehabilitation of rural roads, construction of animal health and milk collection centers, etc..) as well as training activities and certain studies. The lessons learned proved very useful for the appraisal of the Research and Extension Project. The lessons to be retained for the future may concern the following aspects: - for reasons of convenience and management flexibility, specific actions such as rehabilitation of rural roads or construction and equipment of livestock and research facilities should be undertaken directly by the concerned services; credit should be granted directly to beneficiaries;  - 23 - - the Borrower should provide appropriate training to the technicians in charge of Project execution, should ensure continuity of staff and a proper match of their skills with those required of the Project; - the establishment of a monitoring and evaluation system should be required of all projects. 5. Relationship between the Bank and the Borrower during Project Execution Relations between the Bank and the Borrower were always friendly and cooperative. However, the Bank refused to accept a few disbursement applications, as well as the equipment of the milk collection center. 6. Cofinancier's Performance Under the NWAP Project (Beja Management Center), the "Agence Generale de Cooperation et de Developpement Belge" sends a Belgian consultant every six months to follow and assess the activities of the Project. These consultancies should be allowed for Tunisian consultants who could make a useful contribution to the smooth implementation of the Project and should not remain the sole right of Belgian consultants.  - 24 - PROJECT COMPLETION REPORT REPUBLIC OF TUNISIA NORTHWEST AGRICULTURAL PRODUCTION PROJECT (LOAN 2502-TUN) Part III: STATISTICAL INFORMAT1ON  TABLE 1 : RELATED RANK LOAS PROJECT YEAR OF PURPOSE LOAN ANoCET STATUS APPROVAL Third Highways 1978 To help Govt. fmptement a nationwide rurat development S 32 mitlion Loan closed 03/31/86. PCR (1601-TUN) program (improvements in rural roads, extension services of which issued 04/88. and technical assistance to the Ministries of Public S 6.7 milLion Works and Agriculture). cancelled Third Agricultural 6/80 To increase agricultural production and incomes of small S 30 miltion Closed 12/31/86. PCR Credit Project and medium farmers as well as improve the system of issued 06/91. (Loan 1885-TUN) agricultural credit distribution by strengthening the agricultural credit operations of BNT. Northwest Rural 1981 First phase of a 15 year program to support rural S 24 million Closed 12/31/89. PCR Development Project development in the depressed Northwest region of of which issued 10/91. (Loan 1997-TUN) Tunisia. S 7.6 miltion cancelled Fifth Highway Project 1982 Helped Goverrnment continue to provide farmers with S 35.5 mitlion Closed 10/31/91. PCR (Loan 2108-TUN) improved roads, extension centers and farm credit. of which due 06/93. $ 4 million canceltled Technical Assistance 1982 To assist Tunisian Govt. in increasing the number of S 5 million Loan closed 06/30/88. PCR Un (2197-TUN) projects for the 6th Development Plan (1982-86) and in issued 12/90. improving their quality in the areas of agriculture, industry/exports and energy. Agricultural Sector 1986 Supported the first phase of implementation of Govt.'s $ 150 million Loan closed 06/30/89. PCR Adjustment Loan (ASAL-1) program of agriculture sector reforms, within the issued 01/91. (2754-TUN) framework of a general macro-economic restructuring program. Fourth Agriculturat 1987 Promoted development of a financially sound rural credit S 30 miLLion Loan closed 11/91. PCR Credit Project system and financed credit to agriculture in support issued 06/92. (2865-TUN) of the Govt.'s Mediun-Term Adjustment Program (1987-91) formulated in connection with ASAL-1. Second AgricuLtural 1989 Supported the second phase of implementation of Govt.'s S 84 million Under inplementation. Sector Adjustment medium-term agriculture sector adjustment program, Loan (ASAL-I1) within the framework of a general macro-economic (3078-TUN) restructuring program. Agricultural Research 1990 Supports reorganization and strengthening of research S 17 million Under implementation. and Extension Project and extension (3217-TUN)  -26 - Table 2: PROJECTTIMETABLE PLANNED ACTUAL Identification 6883 Preparatlon Beginning 9/83 Preparation End 12/83 Appraisal Mission 1184 Follow-up Appraisl 5/84 Loan Negotiations U8/4 1/14 - 1/118/5 Board Approval 3/14/85 Loan Signature 4/85 4/25/85 Loan Effectiveness 8/28/85 e/27/J8 Loan Closing 6/30/91 12/31/91 Project Completion 6/30/90 12131/91  - 27 - Table 3: LOAN DISBURSEMENTS A. CUMULATIVE ESTIMATED AND ACTUAL DISBURSEMENTS (USS Million) IBRD FISCAL YEAR APPRAISAL ESTIMATE ACTUAL % OF APPRAISAL SEMESTER ENDING ESTIMATE FY 86 Dec. 1985 0.30 0.00 June 1988 0.45 0.00 Dec. 1986 0.75 0.00 June 1987 1.00 0.41 41% FY 88 Dec. 1987 1.50 0.00 June 1G88 2.00 0.45 23%6 FY89O Dec. 1988 2.00 0.22 11% June 1988 2.00 0.52 26% FY 90 Dec. 1989 2.00 0.43 22% June 190 1.50 0.37 26% FY 91 Dec. 1900 1.00 0.75 75% June 1901 0.50 2.10 420% FY 92 Dec. 1991 0.00 0.28 June 1902 0.00 0.30 TOTAL 15.00 5.83 30%  - 28 - Table 3: LOAN DISBURSEMENTS B. PLANNED AND ACTUAL DISBURSEMENTS BY CATEGORY PLANNED REVISED 1/ ACTUAL ACTUAL AS % USS MILLION US$ MILLION USS MILLION OF REVISED Civil Works 3.6 3.6 2.25 63% Goods 1.7 1.2 0.78 63% Credit under Part G (i) 8.0 3.5 2.73 78% and G (iii) of the Project Consultants' Services 0.5 0.3 0.004 31% Initial Deposit in 0.3 2/ 0.0 0.0 Special Account Unallocated 0.0 0.2 0.0 TOTAL 15.0 8.8 5.834 66% 1/ Original loan amount was reduced by US$6.2 m. as of January 24, 1989. 2/ By amendment of June 22, 1987, the initial deposit of US$0.3 was transferred to the unallocated category and the special account was increased to US$0.8 m. to be used by all categories of the Loan.  29 - Table 4: PROJECT COSTS (USS MILLION) ORIGINAL ACTUAL VARIATION (2-1) LOCAL FOREIGN TOTAL LOCAL FOREIGN TOTAL (1) ( t2) Extension 1.7 1.2 2.9 1.16 1.09 2.24 -0.6 Research 0.3 0.4 0.7 0.20 0.18 0.38 -0.32 Livestock Services 1.1 0.7 1.8 0.12 0.11 0.23 -1.57 Farm Machinery Testing 0.2 0.2 0.4 0.007 0.007 0.01 -0.38 Rural Roads 2.4 1.9 4.3 2.06 1.40 3.46 -0.84 Project Implementation Unit 0.2 0.0 0.2 0.042 0.041 0.08 -0.11 Training, TA, Studies 0.5 1.3 1.8 0.022 0.18 0.20 -1.58 Agricultural Credit 8.2 6.6 14.8 0.95 2.73 3.68 -11.12 Contingencies 6.5 4.5 11.1 0.032 0.032 -1 1.06 TOTAL 21.1 16.8 38.0 4.58 6.74 10.32 -27.7 NOTE: Local expenditures were converted into US Dollars with the exchange rate of US$1 *0.9 TD.  _30 - Table 5: PROJECT FINANCING SOURCE PLANNED ACTUAL SAR World Bank 15.0 5.8 Belgium 0.3 ' 2.0 BNA 7.0 0.95 Beneficiaries 3.2 - Government 12.5 3.52 TOTAL 38.0 12.27 The Loan Agreement was amended on March 7, 1987 to reflect an Increase In the amount of the Belgian grant to FB77,350,000 (about US$2 million).  TABLE 6: STATLIS OF LEGAL COVENANTS Covenant Loan Subject Deadline for Status Agreeamt Compliance Section 3.01(b) Submission of approved annual budget and every year Submitted, but financing inadequate due to lack of financing plans. Borrower commitment. Section 3.02(a) Adoption and implementation of action plan on 06/30/85 Series of measures implemented on interest rates and agricultural credit credit recovery under the ASALs and Fourth Agricultural Credit. Section 3.02(b) Sub-loan agreement between borrower and BNT for 11/85 Amendment to agreement signed by Government and BNT on-lending of loans proceeds to finance short- in December 1987 did not solve pending issues. term, mediun and Long-term credit. Section 3.02(c) Sub-Loans to sub-borrowers to be in accordance Eligibility criteria defined, but no specific credit with satisfactory eligibility criteria and with action plan was ever put into place because credit action plan referred to in Section conflicting with on-going rural credit reforms in 3.02(a) above. country. Section 3.03 Borrower to appoint and maintain adequate Not complied with due to overlapping extension staff based on staffing plan submitted responsibilities of extension agents. to Bank. Section 3.04(a) Appointment of Central Connissioner as Project CCDA was abolished in 1989 and Director of Field Manager and provision of adequate resources and Crops appointed as Project Manager. In 1990, lead staff to the CCDA for the project. agencies for extension and research, AFVA and IRESA, were created. Section 3.04(b) Maintain Project Inplementation Unit (PIU) with Effectiveness Done initially, but staff not permanent and not staff including Deputy Project Manager, Chief assigned exclusively to the project. Accountant and Agricultural Economist Section 3.04(c) Establishment of coordinating committee 10/85 Done, however, meetings were held infrequently. Section 3.04(d) Appointment of regional commissioners as project 10/85 Done, however, regional commissioners had managers. overlapping responsibilities. Section 3.04(e) Appointment of deputy managers responsible for 10/85 Not done as foreseen because function of deputy extension and project management and of assigned to regional representatives of plant accountants at CRDA level. production department. Accountants at CRDA worked on other ongoing projects in the area (see Section 4.02(a) and (b)). Section 3.05 Efployment of management, accounting, and Partially done. However, a major problem was that agricultural consultants. extension consultants to assist management with extension program not recruited.  Covenant Lo Sibject Deadline for tus Agreaent I Conpliance Section 3.06 Agreement with private or cooperative entity for Milk collecting center built, but not equipped due leasing and management of milk collecting to lack of interest of cooperatives or private center, maximunm five years after center becomes sector to manage it. operational. Section 3.07 DGPC to furnish Bank with plans for upgrading 06/30/85 Initial study done at effectiveness did not include rurat roads in next year including feasibility and April 30 calcuLation of economic benefits. The study had to study and selection criteria (including an ERR) each year be redone and this contributed to the long delay in satisfactory to the Bank. thereafter roads execution. Section 3.08 DGPC to maintain roads in the Project area. At all times Roads component not completed by Loan closing. Ministry of Equipment to complete and maintain. Section 3.09 Borrower to take all necessary measures to Dealt with under ASAL I and 11 and later under the ensure adequate incentives for milk, meat and SAL I and EFRSL. fertilizer marketing on a least-cost basis to the Borrower. Section 3.10 Milk collecting price margins to cover 0 & N Not relevant since animal buying and milk collection cost of collecting centers; animal buying center centers not supported under project. price margins to cover 0 & M cost from the start and to cover investment cost within 5 years from start of operations. Section 3.11 Measures to start installation of project 03/86 Although accounting firm recruited by effectiveness, accounting system with a view that it be fully accounting system not set up as foreseen due to lack operational by 03/86. of computer and full-time staff. Section 3.13(c) Semi-annual progress reporting within two months every 6 months Done except that there was no information on after each semester. progress with extension and its impact on farmers. Section 3.13(e) Borrower to furnish completion report. Draft report on physical achievements of project submitted to completion mission. H Section 4.02(a) Borrow to maintain adequate accounts for the At all times Not complied with - separate accounts discontinued and (b) project including separate accounts for during implementation by CRDA. This led to the long statements of expenditure. delay in producing a first consolidated audit report D and to the difficulty of the final audit mission to cy give an unqualified opinion on the financial situation of the project. 0 section 4.02(c) Borrower to furnish audit reports of annual Yearly First consolidated report for 1987-90 accounts o accounts and special account within six months provided November 1991. Audit report for 1991 at the end of each year including separate accounts received March 1993. accounts by auditors on withdrawals based on statements of expenditure. OL _ AwOc  - 33 - Table 7: MISSIONS STAGE OF MONTH/ NO. OF DAYS IN SPECIALIZATIONS RATINGS NATURE OF PROJECT CYCLE YEAR STAFF FIELD REPRESENTATED PROBLEMS Identification 07/83 Preparation 09183 3 12 Economist Agronomist Livestock Specialist 10/83 1 6 Research Specialist 09/83 4 23 FAO/CP Final Preparation 11/83 1 20 Economist 11/83 1 9 FAO/CP Appraisal 01/84 2 22 Economist Agronomist 01/84 2 10 Economist Loan Officer 01/84 1 8 Economist 01/84 1 6 Livestock Specialist 02/84 1 10 Consultant Follow-up Appraisal 05/84 2 10 Economist Financial Analyst Supervision 02/85 3 10 2 Economists Agronomist Supervision 12/85 1 20 Financial Analyst Supervision 04/88 1 19 Agronomist Supervision 08/86 2 9 Agronomist 2 Management Economist Supervision 12/86 1 2 Agronomist Supervision 03/87 1 10 Agronomist 2 Management Supervision 02/88 2 12 Agronomist 3 Management Res. & Ext. Specialist Supervision 04/88 1 11 a/ Consultant Supervision 08/88 2 11 Agronomist 2 Management Supervision 11/88 1 5 Agronomist 2 Management Dev. Impact Supervision 01/89 3 7 b/ Division Chief of Technical Department Financial Analyst Credit Specialist Supervision 04/89 2 16 Financial Analyst 2 Management Extension Specialist Dev. Impact Supervision 0o/89 1 2 Extension Specialist Supervision 10/89 1 9 Extension Specialist Supervision 02/90 1 14 Agronomist 3 Management 1 9 Livestock Specialist Dev. Impact 1 7 Extension Specialist TA, Financial, Procurement Supervision 11/90 2 6 Agronomist Economist Completion 05/92 1 5 Operations Analyst I a/ of which partly spent on supervision of BNT4. b/ of which spent on general credit issues and PANO credit component.  - 34 - Table 8: STAFF INPUTS PREPARA- APPRAI- NEGOTIA- LOAN PROJECT SUPER- TION SAL TIONS PROCESSING ARIS ADMINISTRATION VISION COFINANCING PCR TOTAL FY 84 81.6 47.3 0.2 135.1 FY 85 7.3 16.2 9.4 0.3 14.8 48.0 FY 86 0.9 4.4 0.1 5.4 FY 87 0.7 7.4 0.1 8.2 FY8 8 .8 8.8 FY 88 5.3 5.3 FY90 22.3 15.6 37.9 FY 91 7.7 7.7 FY 92 1.9 5.4 7.3 TOTAL 103.9 54.6 16.2 15.6 0.0 1.9 65.9 0.2 5.4 263.7  - 35 - Table 9: Project Results KEY INDICATORS ITEM UNIT SAR ACTUAL TARGET ACHIEVEMENT A. Staff Reauirements 1. Headquarters No 3 Professional Staff 3 staff not permanently assigned (Project Implementation Unit) to project. Agroeconomist and accountant left in 1990 and were never replaced. 2. CRDA deputies for extension + No 1 deputy + 1 accountant in Deputies and accountants not accountants each CRDA assigned exclusively to the project with negative consequences. 3. Staff at Sub-regional extension No 66 of which 15 were already in 16 centers (CTV)' place at beunning of project. 4. Staff at Local Extension Centers No 48 (CRA)_ B. Civil Works 1. Extension Centers No 25 (a) BOja 6 5 (b) Jendouba 9 4 (c) El Kef 9 8 (d) Siliana 9 8 2. Construction of Animal Health No Centers (a) Bqia 2 2 (b) Jendouba (c) El Kef 1 1 (d) Silinna 3. Construction of Mk Collection No 1 I Center 1 (El Kef) 1 (El Kef)2 4. Rural Roads' km 110 92 (a) Beja 39 39 (b) Jendouba 19 19 (c) El Kef 34 34 (d) Siliana 18 C. Vehicle and Eauioment Procurement 1. Vehicles for Extension Centen No _ B (a) Bqia 11 9 (b) Jendouba 14 7 (c) El Kef 13 8 (d) S;liana 11 14 2. Vehicles for artifial inseminations No centers 10 (a) BDja 2 1 (b) Jendouba 2 2 c) El Kef 3 3 (d) Siliana 3 2  -36 - Table 9 (Continued) 3. Audiovisual equipment for No 1 set for each CRDA 1 set for each CRDA extension centers 4. Office equipment for extension No 1 set for each CRDA 1 set for each CRDA centers 5. Office furniture for extension No 1 set for each CRDA 1 set for each CRDA centers 6. Agricultural demonstration No 1 set for each CRDA 1 set for eacb CRDA equipment for extension centers 7. Equipment for conservation works No 1 set for each CRDA 1 set for each CRDA D. Credit for: 8.0 million (a) On Farm Investments USS 6.6 million Total loans given by BNT were S3.2m TD of which $ 2.7m disbursed from Bank loan. (b) Input Resale Centers USS 1.4 million 0 E. Consultants4 US$ 05 million 0.09 miion '/ The SAR only foresaw the creation of extension centers at sub-regional level. This is the staffing situation at the end of the project. It is unclear as to when the extension agents were phased in. Ratio of on avenge 1 extension agent to 420 farmers was achieved by the end of the project versus 1 to 250 foreseen at appraisal. Under the Reseamch and Extension Project, the ratio is 1 to 400. 2, Constructed but not equipped. 'I By the end of the project, only roads in El Kef were completed. The Siliana roads were financed from other sources. 70% of roads were completed in B"ja, and 80% in Jendouba. 'I None of the studies foreseen were done except for accounting study and roads feasibility study. Oversems training also did not take place except for training of head of El Afareg station.