Algeria - Agricultural Research and Pilot Extension Project

Document of The World Bank FOR OFFICIAL USE ONLY Report No. 17929 vIPLEMENTATION COMPLETION REPORT DEMOCRATIC AND POPULAR REPUBLIC OF ALGERIA AGRICULTURAL RESEARCH AND PILOT EXTENSION PROJECT (LOAN 3216-AL) June 1, 1998 Rural Development, Water, and Environment Department Middle East and North Africa Region This document has a restricted distribution and may be used by recipients only in the performance of their official duties. Its contents may not otherwise be disclosed without World Bank authorization.  CURRENCY EQUIVALENTS Currency Unit: Algerian Dinar (DA) Appraisal average (1990) US$ 1= DA 8.10 Intervening years average (1990-1996) US$ 1= DA 31.30 Completion year average (1997) US$ 1= DA 58.41 WEIGHTS AND MEASURES Metric System FISCAL YEAR OF BORROWER January 01 - December 31 ABBREVIATIONS CNPA Centre National de PEdagogie Agricole National Center for Agriculture Pedagogy CRRA Centre Regional de Recherche Agronomique Regional Agricultural Research Center EAC Entreprise Agricole Collective Farm Cooperative NA InstitutNational Agronomique National Institute for Agriculture INPV Institut National de Protection des Vegetaux National Institute for Crop Protection 1NRAA Institut National de la Recherche Agronomique d Algerie National histitute for Agricultural Research INVA Institut National de Vulgarisation Agricole National Institute for Agricultural Extension ITEBO Institut Technique de l 'Elevage Bovin et Ovin Technical Institute for Cattle and Sheep ITGC Institut Technique des Grandes Cultures Technical Institut for Field Crops MAP Ministere de I Agriculture et de la Peche Ministry for Agriculture and Fisheries OADA Organisation Arabe pour le Developpement Agricole Arab organization for Agricultural Development PBO Programmation et Budgetisation par Objectifs Programiing and Budgeting by Objectives SMS Specialiste Matitre Subject Matter Specialist T&V Formation et Visites Training and Visit System Vice President: Kemal Dervii Country Director: Christian Delvoie Acting Sector Director: Salah Darghouth Sector Leader: Mark D. Wilson Task Team Leader: Charles Ameur  FOR OFFICIAL USE ONLY IMPLEMENTATION COMPLETION REPORT ALGERIA AGRICULTURAL RESEARCH AND PILOT EXTENSION PROJECT (Loan 3216-AL) TABLE OF CONTENTS PREFACE ............................................i EVALUATION SUMMARY ............................................ ii PART I: PROJECT IMPLEMENTATION ASSESSMENT .....................................1 A. STATEMENT/EVALUATION OF OBJECTIVES ............................................. 1 B. ACHIEVEMENT OF PROJECT OBJECTIVES ..............................2..............2 C. MAJOR FACTORS AFFECTING THE PROJECT ............................................. 9 D. PROJECT SUSTAINABILITY .............................................9 E. BANK PERFORMANCE ............................................ 10 F. BORROWER PERFORMANCE ............................................ 10 G. ASSESSMENT OF OUTCOME ............................................ 11 H. FUTURE OPERATION ............................................ 12 I. KEY LESSONS LEARNT ...................... 12 PART H: STATISTICAL ANNEXES Table 1: Summary of Assessments Table 2: Related Bank Loans/Credits Table 3: Project Timetable Table 4: Loan Disbursements: Cumulative Estimated and Actual Table 5a: Key Indicators for Project Implementation: Research Component Table 5b: Key Indicators for Project Implementation: Extension Component Table 6: Key Indicators for Project Operation Table 7: Studies Included in Project Table 8a: Project Costs Table 8b: Project Financing Table 9: Economic Costs and Benefits Table 10: Status of Legal Covenants Table 11: Compliance with Operational Manual Statements Table 12: Bank Resources: Staff Inputs Table 13: Bank Resources: Missions 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.  APPENDICES A. Re-Allocation of Loan Funds Table: Re-Programming of Loan Tranches and Allotment of Unallocated Tranches B. INVA's Contribution C. Map  IMPLEMENTATION COMPLETION REPORT DEMOCRATIC AND POPULAR REPUBLIC OF ALGERIA AGRICULTURAL RESEARCH AND PILOT EXTENSION PROJECT (Loan 3216-AL) PREFACE This is the Implementation Completion Report (ICR) for the Agricultural Research and Pilot Extension Project, for which loan (3216-AL) in the amount of US$32.0 million was approved on June 5, 1990, signed on September 26, 1990, and made effective on June 14, 1991. The loan was closed on December 31, 1997, one year behind schedule. A total of US$ 12.56 million was canceled on May 16, 1996. Final disbursements took place on April 30, 1998. The loan account will be closed once the Bank has received documentation to justify expenditures made under the special account, at which time an estimated balance of US$10.0 million will be canceled. The present ICR was prepared by Mr. Charles Ameur, task team leader for this project. Owing to the turmoil in Algeria, no field visit was possible in the context of the ICR. The last supervision mission in the country was in September 1993. Three subsequent meetings took place respectively in Paris in June 1994, in Washington in November 1994, and in Tunis in June 1995. There were no further meetings outside the country. The ICR is therefore based on a review of the Staff Appraisal Report (SAR), the legal documents, the supervision reports, and on the project files of ECA/MNA Information Service Center. The Borrower contributed to the preparation of the ICR by drafting its own Completion Report. Mr. Mark D. Wilson, sector leader, MNSRE, reviewed the Bank final report. i  IMPLEMENTATION COMPLETION REPORT DEMOCRATIC AND POPULAR REPUBLIC OF ALGERIA AGRICULTURAL RESEARCH AND PILOT EXTENSION PROJECT (Loan 3216-AL) EVALUATION SUMMARY Introduction The Agricultural Research and Pilot Extension project was appraised on June 9, 1989, approved by the Board on June 5, 1990, becoming effective on June 14, 1991. At appraisal, project costs were estimated at US$74.9 million, toward which the Bank provided a loan of US$32.0 million. The Borrower was the Democratic and Popular Republic of Algeria. The implementing agencies were the National Institute for Agricultural Research of Algeria (INRAA); and the National Center for Agricultural Pedagogy (CNPA) which, during the course of the project, became the National Institute for Agricultural Extension (INVA). They were later joined by the National Institute for Agronomy (Institut National Agronomique d'Alger, INA). The original closing date of December 31, 1996 was extended to December 31, 1997. Because of significant delays, persistent problems experienced by the research component, and growing civil disturbances, the project was restructured during the mid-term review in 1995. At that time, funds for the research component were markedly reduced, from US$24.3 million to US$8.8 million. The extension component, however, had its share increased to US$8.64 million through the reallocation of US$1.0 million from the research component; these funds were used to purchase additional audio-visual equipment to reach more farmers, given the risks associated with field visits. Finally, at the Government's request, US$2.0 million was allocated to INA, which became the third beneficiary and implementation entity, on June 3, 1996. As a result of project downsizing, a total of US$12.56 million was canceled from the loan on May 14, 1996. The new and final loan amount was US$19.44 million. Final disbursements were to take place on April 30, 1998. At that time, there was a balance of US$10.8 million. Taking into consideration the US$12.56 million previously canceled from the loan, the estimated overall disbursement amounted to about US$10.0 million (31.2% of appraisal estimates and 51.5% of the formally revised forecast).' Project Objectives The objectives of the project were to establish strong and responsive research and extension services to increase agricultural production and productivity, reduce dependence on imported foodstuffs, and to raise farm income levels. To this end, the national agricultural research The above figures are not final since at the time this ICR was finalized, the Government had not yet fully justified some US$1.0 million from the Special Account. Also, FAO is to return to the Bank the unspent funds of some US$0.4 million for which it had entered into a contract with the Govermment to manage the extension overseas training program. ii  system was to be reformed and streamlined, and new approaches to agricultural extension were to be tested on a pilot basis. The project was conceived as a first five-year tranche, to be followed by additional Ban]k support in subsequent years. In research, the objectives were to: (a) strengthen the institutional structure in order to develop competent, well-motivated staff and improved coordination among institutes; (b) reorient resources toward clearly identified priorities through the introduction and testing of research programming by objectives; (c) introduce a system of resource allocation and monitoring, by research program. The extension component was designed to: (a) test on a pilct basis approaches, organization, and methods for the provision of agricultural extension services to the new private voluntary farm cooperatives (EACs) and to individual farmers (EAIs). The goal was to ensure that the services met the needs of each beneficiary and that strong flows of information existed among farmers, research, extension, and training. Implementation Experience and Results The project made a slow start, due mainly to the tedious process of putting in place the- needed institutional reforms. These included vesting INRAA with the necessary powers and authority to coordinate the national applied research program as well as the training of research staff and the use of technical assistance. Also, defining and adopting a statute for INRAA and for the researchers took some three years. Extension, for its part, was operating in an unsettled rural sector, testing new approaches to informing farmers of improved techniques during a time when socialist farms were gradually being replaced by collective farms that evolved into private, individual farms with precarious property rights. Additionally, the country's increasing turmoil became a major disrupting factor for the project from 1994 onwards. As a result, few of the project objectives have been achieved and the overall outcome is highly unsatisfactory. Research Component. From the beginning, INRAA experienced serious challenges to its role of coordinating the activities of the various technical research institutes (ITGC, INPV, ITEBO). These institutes had little or no tradition of working together, let alone of sharing programs, researchers, and equipment. This behavior had clearly been anticipated in the SAR. It had been thought, however, that the risk would be mitigated by the publication of INRAA's new statutes and by the achievement of formal collaboration agreements between INRAA and the research institutes. The adoption and publication of a statute covering researchers took more than two years. This, among other factors, led to delays in recruiting much-needed personnel. In addition, none of the five regional agricultural research centers (CRRA) were created. By April 1994, INRAA had seen its fourth Director General in as many years. These problems were compounded by the fact that the security situation in the country was steadily deteriorating during the period of the project. In mid-1995, the research component was restructured to more closely complement the activities that INRAA would be able to implement over the remaining two years of the project, taking into account a possible request by the Government for a one-year extension of the closing date. Meanwhile, sufficient funds were earmarked to allow for the collaboration of those technical research institutes willing to collaborate with INRAA. At the same time, a review of the initial selection of regional research centers resulted in two new sites being added in the southern part ouF the country and four of the five previous such centers being dropped to include the steppe, the south and the coastal areas. Security reasons were also of high concern in the final selection of sites. iii  In regard to the major objectives, the strengthening of the institutional structure to develop a competent, well-motivated staff and improve coordination among the research institutes was only partially achieved, and at a late stage. Because 1NRAA was forced to spend far too much time on institutional matters out of its control, the recruitment of staff was delayed and the procurement of much-needed equipment was appallingly slow. It should be noted, however, that although INRAA could have procured goods at an earlier stage, it opted instead, to its credit, to assess needs carefully prior to disbursement. The reports reviewed for this ICR also indicate that INRAA introduced and tested the programning and budgeting by objective approach (PBO) on a limited scale. This approach applied mainly to its own activities, since the research institutes did not cooperate in the project until very late in the implementation cycle. INRAA also claims to have implemented an evaluation system for its research activities. Finally, by October 1997, the allocation of staff and of equipment to the various regional centers was still at the programming stage since the centers remained to be officially established. Extension Component CNPA, which became INVA in 1995, played a leading role in improving extension organization and methods on a pilot basis in six of the country's 44 wilayate. Despite experiencing disruptive changes in management, with three managing directors in as many years, INVA was able to develop monitoring methodologies, implement staff training, forge somewhat stronger links mainly with the technical research institutes, and expand the audience for extension messages through mass media communication. The disbursement of the Bank loan to finance the supply of equipment to both the center and local extension teams, the construction and rehabilitation of facilities, and the hiring of technical assistance went according to plan - albeit with significant delays, initially due to institutional issues and later for security reasons. INVA had a good picture of the farming community and types of farmers by 1993, following a survey conceived and carried out jointly with INRAA. As a result, it was able to begin to reach farmers through groups. However, that approach was abandoned in 1995, because visits to the countryside became increasingly hazardous and group meetings in the field were banned. INVA thus began to reach farmers through mass media, for which it was extremely well equipped as a result of the project. INVA also fonmulated a statute for extension personnel that was adopted by the authorities in 1996. INVA introduced such inmovations as transferring extension staff and funds to Chambers of Agriculture to manage on a pilot basis. In addition, INVA had an active training program - both local and overseas -- especially in the final years. Finally, it devised, albeit two years behind schedule, an improved monitoring and evaluation system with clear indicators. Research-Extension Linkages. Overall research-extension links remained weak throughout the project. INRAA was unable to address this problem because it was entangled with institutional issues for the first five years of the project. Even after the institutional issues were resolved, IRAA did little to address these linkages, partly because the research institutes were late in coming up with an inventory of their respective staff, means, programs, and results. lNRAA did not receive the above information until March 1996. In addition, open house days at the research stations came too late in the project implementation cycle, and there do not appear to have been definite instances when research met extension to try to solve the problems of local farmers. Researchers also did not systematically contribute to the development of extension material. Another problem was that INRAA had no direct involvement in the research development teams. Thus it was left to the technical research institutes, over which INRAA had no authority, to iv  establish an inventory of results from experimental trials, and transfer these results to extension and to farmers. According to extension, little of this actually happened. Major Factors Affecting the Project and Sustainability Economic problems in Algeria had some repercussions on implementation in the early years of the project. The staff recruitment programs at both INRAA and INVA were curtailed because of budgetary constraints, uncompetitive salaries to attract experienced personnel, and the lack of sufficient openings on the part of the Ministry for Agriculture and Fisheries (MAP). Most delays that hampered progress of the research component derived from external factors out of INRAA's control. These included: (a) an initial high turnover of general managers; (b) the Govermment's at times acute budgetary constraints; (c) lengthy delays in addressing key institutional issues such as the statutes of INRAA and of researchers, the redeployment of equipment and the reallocation of staff among research institutes, the creation of various research committees, and establishment of the envisaged five CRRAs; (d) a lack of collaboration between INRAA and the technical research institutes, which led to delays in implementing the programming and budgeting by objective approach; (e) the ever deteriorating security situation, which was a source of delays in the procurement and allocation of vehicles and equipment as well as in the hiring of technical assistance; and (f) restricted mobility, which made it difficult for extension to fully test differing approaches, a key task prior to considering a national coverage. The sustainability of the project as a whole is unlikely. 1NRAA faces an uncertain future as long as the enabling legislation (Loi Programme) is not passed and Government's commitment to research remains in doubt. The links between research and extension remain weak, which calls into question the validity of the research institutes' programs to address farmers' actual problems and constraints. Bank and Borrower Performance The Bank performed satisfactorily in its responsibilities of project preparation and launching. Appraisal failed to anticipate: (i) the difficulty for INRAA to rapidly adopt and implement the programming and budgeting by objective approach; and (ii) the lack of cooperationi by the technical research institutes. Because of the Bank Mission's inability to go to the field, for security reasons, and to properly assess project implementation, supervision has been rated deficient. Identification and preparation were carried out with assistance of the FAO/World Banlc Cooperative program. Staff continuity throughout the project lifespan and an in-depth mid-term review provided appropriate guidance during implementation. A well-attended workshop including research, extension and higher education staff was launched at project inception. The Banlc strongly supported, from the onset, the need to move quickly to improve working relationships among the various research institutes while institutional issues were being addressed. The Banlc also recommended as early as April 1992 that a strategy for extension be defined to allow for a gradual divesting of the public sector and a corresponding transfer, over time, of responsibilities to the private sector. Because of the country's increasing unrest, supervision missions in the field became impossible by end-September 1993. Meetings between the Bank and project management took place on three occasions respectively in Paris, Washington, and Tunis. There were no further meetings from June 1995 till the project closed in December 1997, the Government insisting that Bank missions travel to Algeria. No supervision was carried out in the field. This situation definitively had a negative effect on project implementation. Programs slipped; disbursements v  experienced long delays; progress reports, especially from research, were late; and above all it was very difficult to maintain a regular dialogue between project management and the Bank. Project Outcome The improved managerial and technical capacities gained by 1NRAA during project implementation constitute a definite asset. This, however, took an inordinate amount of time and energy, to the detriment of concrete activities that would have ensured a solid medium and long- term research program likely to benefit rural communities. It was only toward 1996 that INRAA began to receive some support from the technical research institutes and that the programming and budgeting by objective approach was initiated. Although somewhat strengthened, 1NRAA still lacks full collaboration from the technical research institutes, and the future of research remains unclear unless the enabling legislation is passed. Also, the acquisition of equipment was delayed, which was another limiting factor on the implementation of the anticipated research programs. As to the extension component, it was reorganized, strengthened and properly equipped within two to three years of project inception. However, the lack of sufficient results from research considerably slowed its ability to transfer technology and knowledge to farmers. In addition, the country's increasing turmoil became an ever more limiting factor to progress at the field level. Because of the above constraints and severe limitations, the overall project outcome is highly unsatisfactory. The project has so far not generated the anticipated agricultural production increases. Future Operations Some key institutional issues remain to be solved as previously indicated. Also, and unless the Loi programme passes the Parliament, the future of research will remain uncertain. INVA would need to firm up a strategy before any expansion nationwide of the extension component could be considered. Although further phases were anticipated in the SAR, a follow-up project is very much in doubt given the country's current budgetary constraints. Key Lessons Learned The key lessons learnt from project implementation include the following: (a) major institutional issues should be solved as a condition for negotiations -- this would have prevented INRAA from spending precious time away from project implementation; (b) programming and budgeting by objective is a complex approach that took INRAA nearly four years to put in place. Thus, the local context, capacity, and limitations of implementing agencies should be carefully assessed and analyzed during preparation; (c) when research embarks on complex restructuring and adoption of a fundamentally different approach, fast disbursement should not be expected from project onset; (d) modifying the approach to extension, especially at the national level, requires having a vision and defining a strategy at an early stage, together with extensive and effective participation by farmers. For lack of both at project onset, it took INVA more time than anticipated to play the vital role of testing various approaches to extension on a pilot basis; (e) given the country and the diversity of farmers, it is most likely that the Training and Visit (T&V) approach to extension will prove inadequate on a national scale; and (f) when elaborating a strategy, decisionmakers must decide whether extension should remain the sole responsibility of the public sector or should be a mix of public and private interventions. They should also decide the level of farmers' contributions to the cost of extension, and the amount of funds the Government is ready to commit to extension on a regular and long-term basis. vi  IMPLEMENTATION COMPLETION REPORT DEMOCRATIC AND POPULAR REPUBLIC OF ALGERIA AGRICULTURAL RESEARCH AND PILOT EXTENSION PROJECT (Loan 3216-AL) PART I: PROJECT IMPLEMENTATION ASSESSMENT A. STATEMENT AND EVALUATION OF OBJECTIVES 1. The overall objective of the project was to establish strong and responsive research and extension services in support of Government efforts to increase agricultural production and reduce the country's dependence on imported foodstuffs. It was conceived as a first tranche, possibly to be followed by additional Bank support, based on results. The project aimed at reorganizing and strengthening the National Institute for Agricultural Research (INRAA) and the National hIstitute for Agricultural Extension (INVA). It also aimed to improve linkages among research, extension, training, and farmers to ensure that research work was relevant to farmers' needs, and that priority technical recommendations would be quickly and effectively transmitted to beneficiaries. 2. To achieve these objectives, the project consisted of a program of investments associated with a series of institutional reforns, staff appointments, targeted training, and technical assistance, and the provision of funds to cover related operational costs. The research component included: (a) strengthening INRAA's capacity to coordinate research programs and carry out basic research; (b) reorienting resources toward clearly identified priorities through the introduction and testing of research programming by objectives; (c) implementing through INRAA and other technical research institutes (ITGC, INPV, and ITEBO) five priority research programs on cereals, legumes, fodder crops, red meat production, and cereal/livestock systems in the rainfed and irrigated subsectors; (d) introducing a system of resource allocation and monitoring by research program; (e) upgrading research quality by appointing staff and providing training, evaluation, and better facilities; and (f) recruiting technical assistance. The extension component included: (a) testing approaches, organization, and methods for the provision of agricultural extension services to the farm cooperatives (EACs) and individual farmers in 20 dairate located in 6 wilayate; (b) strengthening of five extension training centers (including one in methodology and four in technical aspects); (c) strengthening the National Center for Agricultural Training (CNPA) through the provision of vehicles and the production of written and audio-visual material, thus expanding the audience via media communication and increased staff mobility; (d) providing for local and overseas training; (e) recruiting technical assistance; and (f) setting up a monitoring and evaluation system. Institutional arrangements for coordinating research and extension programs were also to be improved. 3. At appraisal, the project cost was estimated at US$74.5 million. Financing included a Bank loan of US$32.0 million and Government financing of US$42.5 million, US$11.1 million of which came from taxes and duties. The costs of the research and extension components were estimated at US$35.9 million and US$19.2 million, respectively.  B. ACHIEVEMENT OF PROJECT OBJECTIVES General 4. The overall project objectives were not achieved, especially for the research component. 1NRAA spent too much time on institutional issues, most of which were out of its control. Some of these issues were still being debated by end 1997. While waiting for the institutional matters to be resolved, INRAA postponed the procurement of much- needed equipment and vehicles. Thle recruitment of researchers was also delayed because of the lack of openings and of funds. Procurement was delayed as well in mid-1995 while INRAA waited for the necessary decrees to be adopted regarding its status and that of its researchers. INA, which came on board in mid-1996, also experienced significant delays on procurement matters. As the project came to a close, therefore, the research component had acquired only a small portion of the equipment initially anticipated, and vital collaboration between INRAA and the technical research institutes - although said to be improving in 1996 - remained informal and unsatisfactory. The regional research centers (CRRAs) were never officially created. The programming by objectives stalled in 1994 for lack of collaboration on the part of the technical research institutes. The training program for researchers started slowly and never recovered, although short-term overseas training intensified toward project end, relying on both Bank funds and bilateral aid. The technical assistance, which had been barely drawn upon in the project's early years, came to a grinding halt as the country experienced ever-increasing turmoil from 1994 onwards. 5. The extension component, which faced far fewer institutional problems than the research component, was operational in 1993. Extension was reorganized under the project and acquired most of the equipment, vehicles, and buildings envisaged at appraisal. However, links between research and extension remained weak, with research results not being transferred to farmers, especially after field visits to rural communities became risky and group gatherings were bannecl. INVA focused much effort on training its staff, and also significantly improved its mass media capacity as an alternative to field visits. Although a monitoring and evaluation system had been designed at the center, it was not properly utilized at the wilayate level. Thus, it remains difficult to state to what extent farners' productivity and income increased as a result of the project. Specific achievements and shortcomings of some of the project components are detailed below. Research 6. Institutions. It took longer than anticipated to reform agricultural research. Although INRAA was given responsibility in 1986 for basic agricultural research, for coordinating applied research at the national level, and for coordinating training of research staff and the use of technical assistance, its status still needed to be amended to confirm this coordinating role. This role was at the heart of a controversy between 1NRAA and the technical research institute's (lTEBO, ITGC and INPV). The institutes had gained much autonomy in the 1970s at a time when INRAA's role was severely reduced, and they were reluctant to relinquish their autonomy under the project. As a result, there is still little cross-subsector coordination of research. 7. Under the project, INRAA was to bring together researchers from the various research and higher education institutions by obtaining similar status for both and by involving them in preparation of the national priority programming process. However, the necessary interministerial decrees relating to the status of INRAA and the researchers were not signed until 1995 and 1996, respectively. Most of the delay stemmed from differing interpretations of the proposed status by 2  INRAA and the technical research institutes. Until the new status became effective it was difficult to recruit additional researchers, let alone retain those with much experience. By the end of the project some progress had been made in aligning the salaries of the research staff with those of the research scientists, but since this is not an automatic process, it requires regular and time- consuming updating. 8. By end-1996, INRAA was still dealing with important institutional issues. In addition to the question of its status, these issues included the promulgation of texts to create the CRRAs and render them operational, the preparation of texts for the creation of associated research laboratories, and the establishment of committees for the Scientific Council. In addition, INRAA and the technical research institutes could not agree to collaborate and share responsibilities. TNRAA was supposed to carry out basic research and coordinate research programs nationwide, while the institutes were to engage in adaptive research, and support seed and planting material production. Instead, the institutes continued to conduct their own experiments and to deal directly with international donors. 9. The initial high turnover of senior management was an additional source of delays, with INRAA having four directors general in as many years. Varying policies, visions, and priorities definitely took their toll on project implementation. Also, it was difficult to recruit researchers until 1995, as envisaged at appraisal, first because of INRAA's lack of status and later because of the country's financial constraints. Of the anticipated 98 researchers to be recruited during the project, only 56 were actually hired, and no more openings were created. As a result, newcomers were hired only to fill existing vacancies. 10. Programming and Budgeting. In addition to institutional strengthening, programming and budgeting by objective (PBO) was another key feature of the research component. One purpose of the research was to help prepare and implement medium and long-term programs through PBO by making optimum use of existing researchers and of means. Furthermore, it was intended to associate scientific personnel from the various technical research institutes and from higher education, and to join their technical and financial capacities for the purpose of coordinating research activities. The multidisciplinary teams were to work using joint laboratories and experimental research stations, each entity contributing its own human and financial resources. At the basic level, the identification, monitoring and evaluation of the programs, subprograms and research activities were to be overseen by Regional Program Committees, to be located at the CRRAs. At the central level, a National Program Committee was to be based within MAP. Although the committees were created, their members were never officially designated. 11. TNRAA's request, in 1994, to establish an inventory of researchers and laboratory equipment received no support from the technical research institutes; nor did the institutes respond to INRAA's proposed canvas for the programming by objectives. INRAA had called in 1994 for an inventory of the research stations, of their scientific and technical potential, and of their research and extension themes. In view of these problems, and since the CRRAs never became operational, it is difficult to believe that programming and budgeting has had any significant impact on the way research was conducted. 12. At the time of the mid-term review in 1995, cooperation between INRAA and the technical research institutes was minimal and unsatisfactory. The project was therefore restructured and markedly downsized. All research subcomponents, including activities, programs, and procurement, were reduced in scope. Under the restructured project, INRAA was to pursue its own 3  programs, with funds set aside for possible cooperation with those institutes that might elect to collaborate at a later stage. In addition, procurement of equipment and vehicles was reduced, preference was given to prefabricated buildings rather than to new constructions, and priority was given to in-country rather than overseas training. As a result, the research component was reduced from US$24.3 million to US$8.8 million, including contingencies. 13. The physical downsizing of the research component was specifically tailored to INRAA's implementation capacity given past problems, constraints and limited collaboration with the technical research institutes. Another outcome of the mid-term review was to reduce the number of CRRAs from five to three, with two of the CRRAs (Adrar and Touggourt) located in the southern part of the country, where a more secure environment existed. Yet until 1995, it was still unclear whether the CRRAs would be under INRAA or ITGC. More troublesome was the fact that the CRRAs still had no legal status as the project was coming to an end in December 1997. Although it had been envisaged at appraisal that each CRRA would be responsible for the experimental farms in its region, this never materialized for the above stated reasons. 14. While INRAA claims to have implemented some monitoring and evaluation of its ovwn research programs, the progress reports show that the system remains weak. Some monitoring of inputs, such as research staff, program budgets, and equipment, is evident. But when it comes to evaluation, there are still insufficient links among initial problems and constraints, the selection of specific research themes, and the final objective of any given research program. Toward the end of the project, INRAA resumed publication of its Revue de la Recherche Agronomique de I INRAA. This was a commendable decision that no doubt is contributing to opening up INRAA's programs to the international research community. 15. Research Programs. According to INRAA, a new approach to programming was devised following the mid-term review in 1995. This approach focused on regional priorities such as, inter alia, the rehabilitation of olive plantations, soil salinity, the rehabilitation of date plantations under oasis conditions and/or under pressure irrigation, and milk production. The programs were said to be screened by scientific committees at various levels to ensure that they were feasible and that both budget and staff requirements were appropriate. Meanwhile, INRAA's reports stressed that the production systems based on cereals and livestock were given higlh priority. It is not clear from INRAA's reports the extent to which the technical research institutes contributed to these programs, which were somewhat different from those identified at appraisal. 16. The actual level of farmers' participation in selecting priority research programs at the- regional level is difficult to ascertain, mainly because INRAA - although it engaged in some monitoring of its research programs and activities - did not have a proper evaluation system. Research themes were not justified in a systematic fashion that highlighted the magnitude of local farmers' problems and constraints, or the resulting yields, productivity, and incomes. There were- no clear indications as to what the research was attempting to improve; nor were there any reference points or projections. 17. The project failed to strengthen the necessary links between research and extension, due to the fact that the technical research institutes, which were to ensure such links, failed to fully collaborate with either INRAA or to a lesser extent with INVA. This lack of linkage was corroborated by the Arab Organization for Agriculture Development (OADA) consultants, who stated in August 1996 that the research component had not started (aside from dealing with institutional issues). Neither farmers nor extension benefited much from research. At the field level, 4  technical committees had been established in each of the six wilaya where technicians, specialists, and farmers were supposed to meet on a regular basis. Nor is there much evidence of any formal connection between research and extension at the regional level. Moreover, there is no mention of any collaboration between INRAA and INVA in either of the two entities' progress reports over the life of the project. In spite of the above and according to INVA management links between extension and research existed at: (i) the local (wilaya) level where personnel from the technical research institutes would attend twice quarterly meetings of the wilaya technical committees; and (ii) the national level where staff from the institutes would attend bi-monthly coordination meetings organized by INVA and including the central directorate for training, for research and extension within MAP as well as wilaya project managers and heads of training centers from the project area. 18. Strengthening of Research Stations. To implement the five research priority programs envisaged at appraisal, the project was to finance: (a) the strengthening of 1NRAA to handle basic upstream research including in phytogenetics, soil studies and analysis, and bioclimatology; (b) development of the Mahdi Boualem research center; (c) establishment of a central documentation service and a backup service for maintenance of laboratory equipment; (d) strengthening of three technical institutes (ITGC, ITEBO, and INPV); and (e) upgrading of five experimental farms to CRRA level covering the main agroecological zones. These plans implied an increase in the numbers and training of researchers and technicians, and the provision of vehicles, equipment, housing, laboratories, workshops, greenhouses, and operating costs. 19. From the progress reports, it appears that INRAA did somewhat strengthen its ability to carry out basic fundamental research on the topics envisaged at appraisal. This was made possible through the recruitment/redeployment of 56 researchers out of 98 planned; acquisition of about 50 percent of the laboratory equipment anticipated; procurement of 46 small cars, one minibus, 14 four-wheel drive vehicles, and 8 pickups; procurement of laboratory equipment, computers, and printing material; construction of offices (ready made units); and provision of operating funds. As the project was coming to an end and important institutional issues were still unsolved, the Bank upheld its non-objection to the sizable construction program at the Mahdi-Boualem research center. The construction would have taken more than eighteen months to complete and the bidding documents did not reach the Bank until in June 1997. 20. As indicated above, the technical research institutes were allocated very little means as a result of their failure to collaborate with or submit clear programs to INRAA. While there were only three unofficial CRRAs at mid-term review instead of the five anticipated at appraisal, two of which were to be located in the South of the country, the CRRA program was never implemented. 21. Following the mid-term review and the ensuing restructuring and downsizing of the project, the Government requested that the Bank reallocate some US$2.0 million to a new beneficiary, the Institut National Agronomique (INA). The agreement to reallocate these funds was signed on June 3, 1996. INA had been closely associated with INRAA in past years and was a source of researchers as well. The funds were to enable INA to acquire much-needed laboratory and pedagogical equipment. Because of lengthy and cumbersome procurement procedures, INA got only a third of the equipment initially planned. Monitoring and evaluation of that subcomponent was particularly weak. There was neither a report on the equipment procured by INA nor on the type of research to be carried out. 5  22. Staff Development and Technical Assistance. One of the project's initial achievements was to obtain similar status for researchers and higher education personnel. However, TNRAA, failed to recruit 42 of the 98 researchers anticipated at appraisal, and there was no indication in the reports of where or on what grounds the personnel recruited were deployed. Nor was there any mention of the critical masses of researchers constituted for the various research programs. 23. Training received much attention from 1NRAA management from the beginning of the project. Some 26 researchers received scholarships, mainly from France, and took post-graduate degrees overseas. This was close to the 30 that had been envisaged at appraisal. An increasing percentage of trainees, however, chose not to return to their country as the civil strife worsened. This led INRAA to opt for in-country training, complemented by short visits of the intemational trainers to Algeria. This option did not work either, however, as the trainers were reluctant to come to the country. INRAA then reverted to short-term training courses in various countries, with the training program managed by FAO. This limited program appears to have been successful. 24. The technical assistance subcomponent, which was to include a plant breeder and a soil fertility expert for three years, and an agroclimatologist for two years, was never started for security reasons. Also, the project barely took advantage of the consultants envisaged for training, management, documentation, and computerization, since, again, foreigners were reluctant to workc in the country. 25. INRAA was slow in preparing and providing regular progress reports, which were to have been submitted biannually. The backlog was such that activities for 1995, 1996, and the first semester of 1997 were finally reported on in one batch. The result was an overwhelming amount of raw data with little analysis and synthesis. It is not before September 1997 that a concise summary was made available. Extension 26. Institutions. Responsibility for implementing the extension component was initially vested in CNPA, a national body specialized in agricultural training. In this respect, the Government's objective was to establish a coherent system for the transfer of improved technology to individualL farmers as well as the cooperatives (EACs). Because of Algeria's limited experience witht agricultural extension, this component was to be implemented as a pilot to test approaches thalt could eventually be repLicated nationwide, leading to a simplified, more direct, and strengthened system of extension. This has been achieved to a certain degree. Definite improvements in. extension are noticeable, especially in staff organization, programming of extension activities, more focused training, increased technical support of field agents, more communication with farmers through the mass media, and a more proactive monitoring and evaluation system. INVA (the successor to CNPA), at the center, is fully staffed, including 10 senior and 4 support personnel. At the regional level, all the staff envisaged at appraisal have been recruited/redeployed. While extension appears to be well structured at the national level, it is part of various entities at lower levels (DSA at the wilaya level, subdivision at the daira level, and the communal delegation at the commune level). This makes it more difficult to properly coordinate, monitor, and evaluate extension from the center. More importantly, it tends to dilute extension among several other activities, with no real sense of priority. Thus, a clearly defined extension service or division needs to be considered at the wilaya and daira levels, a proposal that had not been envisaged at appraisal. Another concem is that extension personnel continue to spend no more than 30 percent of their time on extension, having to deal with many other aspects none directly linked with their mandate. 6  27. On April 1, 1995, CNPA became the National Institute for Agricultural Extension (ENVA), by Decree No. 99.95 signed by the Prime Minister. This indicates the priority given to some degree to extension by the authorities. At the mid-term review it was felt that the extension component was performing well. INVA had acquired most of the hardware anticipated at appraisal. The training program was being implemented in earnest. Much attention was being given to developing the institute's outreach capacity through the strengthening of mass media. Also, a real effort was being put into monitoring and evaluation at the Central level. Disbursement was proceeding at a satisfactory pace. INVA requested some additional funds for audio-visual equipment, to which the Bank agreed. Total project funds for the extension component were thus increased by US$1.0 million to US$8.64 million. 28. As foreseen at appraisal, the extension component consisted of the introduction of an improved organization in 170 communes controlled by 20 dairate located in the six wilayate of Ain-Defla, Blida, Tipaza, Medea, Bouira, and Tizi-Ouzou, surrounding the capital Algiers. The project area covers some 452,000 ha of arable land and comprises about 160,000 farmers, of which some 50,000 were to be reached. Staffing at the regional level included 191 field extension agents, 20 supervisors, 18 bureau chiefs, 25 specialists, 12 surveyors, 6 local coordinators, 12 trainers, and 15 support staff. The promulgation in April 1996 of a specific statute covering the extension agents definitely boosted the staffs morale. By project end INVA had acquired 189 vehicles, as well as audio-visual, office, and printing equipment. It had also completed the construction program, financed local and overseas training for farmers and staff, contracted with FAO to manage the training program, and recruited the anticipated consultants (OADA) for technical assistance envisaged at appraisal. 29. In a departure from the project's institutional objective to test approaches to extension that could eventually be used nationwide in a second phase, INVA stuck instead to a somewhat modified Training and Visit (T&V) extension system. There was much more emphasis on mass- media than in the typical T&V system. In addition, strict ratios required each extension agent to meet 5 to 10 contact farmers per month, who in turn would reach 10 to 15 neighbors. This contact farmer concept has been abandoned in many countries, and nothing in INVA's progress reports indicates whether the practice was working well. In addition to contact farmers, extension reached farrmers through visits to individual farmers and groups indoors, demonstration plots and field days when authorize, mass media, and fairs. The major results so far include: (a) a better structured extension system; (b) a more bottom-up approach; (c) strong emphasis on regular training; and (d) increased awareness and more systematic use of monitoring and evaluation as a management tool. In spite of its varied approach, however, INVA seems to have adhered until early 1997 to only one extension system, and there is no physical or economic evidence that the approach will work nationwide. In fact, based on worldwide experience, a single approach is very difficult to replicate throughout a diversified country such as Algeria. As to the late transfer of extension agents to some chambers of agriculture there is no indication either in INVA or the consultants reports as to whether the operation has or not been successful. 30. The adoption of a single system until late in the project points to the lack of a clear strategy for extension, as suggested by Bank missions in 1994. There were no proposals as to how 7  extension could evolve over time within the Algerian context.1 Indeed, it is crucial to consider: (a) whether extension should remain the sole responsibility of the public sector, or should instead rely on a mix of public and of private interventions; (b) the type of farmers to be reached, their proportion nationwide, and the criteria for their selection; (c) whether farmers should contribute to the cost of extension, and how their contributions would be collected; (d) whether farmers targeted for extension should be periodically replaced with newcomers; and (e) the amount of funds the Government is ready to commit to extension on a regular basis. 31. Staff Development. Training received a great deal of emphasis from the beginning of the project. Basic extension training for all 191 field agents was completed by November 1994. In support of the training program, five training centers were strengthened and included at Am Taya, Ain Bessem, Bougara, Medea, and Mechtras. The initial training was subsequently replaced by more technically specific types of courses. However, the further training of Subject Matter Specialists (SMSs), trainers, and senior staff from the center was delayed due to the difficulty of recruiting technical assistance, and INVA finally asked FAO to manage the overseas training program. The courses were held in France and the Netherlands, and focused on extension methodologies and programming. There is no indication in INVA's progress reports of the impacit of the training on staff performance. Nor is there any infonnation about the selection of training sites, the courses offered, or the people sent. 32. Audio-visual and Mass Media Support. INVA put a great deal of emphasis on audio- visual support for the farmers. The audio-visual department was well equipped under the project: and had competent staff. In addition, the mass media were regularly used to reach more farmers al: a time when traveling the countryside involved major risks. Production of mass extension materials; was significantly increased, and by 1997 included 317 TV spots, 3,917 radio broadcasts, numerous slides, photos and posters, and no less than 45,000 brochures and pamphlets. The radio broadcasts were aired on two national stations, as well as on two regional and 17 local stations. Aji time was split among vegetable production (55.4 percent), animal production (24.8 percent), andL environment, irrigation and pesticide use (21.8 percent). It appears from a survey conducted by INVA in 1995, that mass media support was much appreciated by farmers, although here again, INVA failed to analyze the impact of this important subcomponent in its progress reports. There was also no indication as to who selected the themes, on what basis, and whether research was involved at any stage. 33. Technical Assistance. The project had envisaged the hiring of short-term consultants to design extension methodology, as well as for planning, monitoring and evaluation, and project Although neither recorded in INVA's latest progress report for the period January-June 1997 nor in its contribution to this ICR, it would appear that a national extension system was recently submitted by INVA to MAP. The proposed lean and flexible extension system would also have been widely distributed to all the partners involved in extension including the central directorates at MAP, the institutes, the regional agricultural services, the national and the regional Chambers of agriculture, and the private sector including input suppliers. Furthermore, a survey would be in progress involving some 2,000 farmers to get their views on the pilot extension system as well as on the various approaches to extension tested so far. (Some of the suggestions presented by INVA in its contribution to the ICR, on institutional aspects and on methodologies, would have gained being further developed). Based on the outcome of the above consultations, INVA would be in a better position to define within the coming year a strategy for extension that would contribute to the selecting of a new extension system adapted to the new organization of the agricultural sector. 8  management. Because of the security situation, however, it is not until 1995 that a team from the Arab Organization for Agricultural Development (OADA) was finally put together. It started work in Algeria in 1996. The overall performance of the consultants has been mixed. With few exceptions, practical advice was mired in theory. There were great expectations that the experts would come up with clear and pertinent recommendations on extension, monitoring and evaluation, and training. This did not happen. 34. Monitoring and Evaluation. INVA has, over time, gradually improved its ability to monitor and, to a lesser extent, evaluate project activities. Progress reports were regularly provided, albeit with the deficiencies noted above. INVA also conducted, in 1994, an interesting socioeconomic survey in all six wilaya involved in the extension component. INVA's analytical capacity, though, remained insufficient. Also, according to the OADA consultants, there was a reluctance by the various levels of the hierarchy to cross check data at the field level. This left the door open to both involuntary and deliberate errors. Most of all, there was a lack of reference points, which made proper evaluation of the project impact nearly impossible. Although INVA did conduct a few surveys on production and yields, they were not systematic and lacked an economic perspective. Also, there was seldom any reference to the rate of farmers' adoption and/or rejection of extension messages. Finally, there was too little follow-up to strengthen the weak links between extension and research. C. MAJOR FACTORS AFFECTING THE PROJECT 35. Some institutional issues were solved during the project lifetime, including the updating of INRAA staff's status to that of higher education personnel, preparation of texts to create the CRRAs and render them operational, preparation of texts for the creation of associated research laboratories, and establishment of committees for the Scientific Council. But a few important issues remain, including the lack of collaboration between the technical research institutes and INRAA, and the uncertainty as to whether the enabling legislation for research, which could make or break the future of agricultural research in the country, will be passed by Parliament. Also, the text creating the CRRAs were never signed. Finally, it would appear that the implementation of programming and budgeting by objective, a new and complex approach that required more time than anticipated, did not reach the critical budgeting stage. Close follow-up and careful management will be needed if the approach is to survive. 36. Budgetary constraints affected both the research and extension components to varying degrees, especially with regard to recurrent expenditures. Together with the lack of flexibility in administrative procedures, this was a major factor in the slow recruitment of new staff, particularly researchers. As to INVA, it adopted a modified version of the T&V approach. There was word at one stage that some responsibility for extension had been transferred on a trial basis to five wilaya, but there was no mention of such move in any of INVA's progress reports. D. PROJECT SUSTAINABILITY 37. Sustainability of the project as a whole under the country's current civil strife is unlikely. In addition, the very future of research at the national level depends on whether or not the enabling legislation is passed by the Parliament. As to the project research component, investment has been extremely low. Despite the many delays and difficulties, however, INRAA was able to introduce 9  the programming and budgeting by objective approach. If successful, this will be one of INRAA's main legacies. This improvement, however, was not extended to the technical research institutes. Considering INRAA's present research programs, that organization has returned to its true mandate, which centers on basic research, but it still remains for INRAA to coordinate adaptive research - which is key to extension - with the technical research institutes. But given the lack of cooperation between these entities, and the few pieces of equipment procured and allocated to the technical institutes, the impact of the research component will be minimal. 38. Some aspects of the extension component are sustainable, whereas others are not. It is likely that the T&V extension system will be increasingly difficult to sustain over time and on an,y large scale. Thus it is important for INVA to actively seek alternative approaches, which will require adapting extension to various categories of farmers and involving the private sector, as well as relying on the chambers of agriculture. The training provided under the project will continue to have a positive effect for as long as current staff remains in their positions. Finally, the significant investments in audio-visual equipment and production are likely to last for years to come. In fact, given the current hazardous situation prevailing in the countryside, radio and TV messages are becoming the essential means for reaching farmers and technicians alike. E. BANK PERFORMANCE 39. In agreement with the Government of Algeria, project identification and preparation were undertaken with assistance of the FAO/World Bank Cooperative Programn. Bank performance at identification and preparation was satisfactory. However, at appraisal, Bank staff underestimated the difficulties involved in implementing the research component. Some complex institutional issues required much more time than anticipated to be solved, and others were never resolved due to unfavorable circumstances. Six supervision missions were carried out in Algeria and three outside the country. In addition, a mid-term review took place in Tunis in June 1995. Continuity was optimum, with the same task manager leading all supervision missions from 1991 until closing in December 1997. At the mid-term review, the Bank and the Government tried to salvage the project, especially the research component, which was downsized to fit INRAA's implementation capacity. The Bank responded promptly to the Government's request to extend loan disbursemenit by one year to December 31, 1997. But the Government's insistence at the end of 1995 that projecit management could no longer meet Bank missions outside the country severely hurt the good ancl efficient working relationship that had existed. The situation became highly detrimental to the project. With hindsight, the Bank should have taken a firmer stance earlier on and suspendecl disbursements. In November 1997, the Bank declined an official request to extend the project closing date by another year. F. BORROWER PERFORMANCE 40. The Borrower's performance was satisfactory during project identification throughl appraisal. During implementation, the commitment of both INRAA and INVA to carry out the: activities defined in the SAR remained strong. The odds, however, rapidly tumed against the: research component. There were four directors in as many years. INRAA spent precious time pushing for a solution to institutional issues over which it had no control. Budgetary constraints remained constant. The decision to defer procurement of much-needed equipment, in order to check: facilities and equipment specifications against the precise identification of research priority 10  programs and to make an accurate inventory of existing equipment, was commendable. But this approach backfired, since INRAA ended up not having enough time to acquire further equipment before the closing date. The significant downsizing of the project following the mid-term review did not generate the anticipated improvements. Cooperation between INRMAA and the technical research institutes remained minimal. Necessary decisions to make the CRRAs operational were never taken by the Ministry of Agriculture, thus preventing the regionalization of research. INRAA finally resorted to medium-term research programs that it felt were achievable. The training of researchers started late and an insufficient number of trainees was sent overseas; in addition, many decided not to return once trained. The technical assistance program barely got off the ground, essentially for security reasons. Progress reports were irregular for nearly three years, followed by an avalanche of repetitive information that was not sufficiently analyzed or synthesized. Most audit reports were provided on time and were of reasonable quality. INRAA used an insignificant amount of the funds allocated to research. 41. The extension component experienced less serious institutional issues, which had to do mainly with delays in opening INVA's bank account. But extension was affected to varying degrees by a number of external factors, especially during the project's first two years. INVA had acquired most of the equipment envisaged by end of 1993. It produced a first strategy for extension and played a leading role in having the Ministry of Agriculture adopt a status for extension workers. INVA had three directors in as many years. Although disruptive, however, the effect of this was far less damaging than for research. The training program accelerated during the last two years and a significant number of staff was sent overseas for short-term courses. The technical assistance subcomponent did not start until 1995 with the hiring of OADA experts, at a time when the security situation was further deteriorating. INVA produced regular progress reports that, while they improved over time, lacked proper analytical perspective. INVA used some three-quarters of the funds it had been allocated under the project. G. ASSESSMENT OF OUTCOME 42. The project has, although to a limited extent, contributed to strengthening 1NRAA and improving its structure and working capacity. The allocation of additional resources in infrastructure, equipment and scientific personnel was minimal due to both delays and institutional issues. The lack of collaboration with the technical research institutes prevented much-needed coordination of research at the national level by INRAA. The latter focused much attention on the programming and budgeting by objective approach, from which the technical research institutes had more or less excluded themselves. The overall links between extension and research remained weak with INVA having some limited working relationship mainly with the technical research institutes. The training program, although reduced from initial projections, had a beneficial effect on research team members working on priority programs. 43. Extension is better organized. Beneficiaries are said to participate more actively in defining priorities and extension messages. Staff are much better trained and more mobile, and their activities are programmed and monitored. Mass media communications are used more widely and effectively. Because of the weakness of the evaluation system, measuring the project benefits is not possible. New technologies are said to have been successfully transferred to farmers, mainly by extension, which would have resulted in increased production and higher yields in cereals, potatoes, vegetables, fodder crops, and fruit trees. However, such assessments are not properly documented for lack of referentials. 11  44. Overall, very limited progress has been made toward achieving the project's original objectives in research. The SAR's objectives now appear to have been overly ambitious for a five- year implementation period. Clearly, appraisers had not fully appreciated the complexity of programming and budgeting by objective, given the lack of experience with introducing these procedures in other countries. Also, much of INRAA's attention was, for several years, diverted to addressing main institutional issues. In regard to INVA, the implementation objectives were less ambitious and have to a large extent been achieved. The real impact will need to be assessed through better monitoring and evaluation over time. Meanwhile, much remains to be done to devise a clear strategy for extension - not envisaged at appraisal - that would lead to feasible and cost- effective approaches that could be implemented in a subsequent phase. H. FUTURE OPERATIONS 45. Although further phases had been anticipated in the SAR, a follow-up project is very much in doubt given the country's current budgetary constraints. In addition, research and extension do not appear to be the Government's immediate priorities, particularly since tangible results are difficult to assess, especially for research. Also, the ongoing civil strife mitigates against traveling in the countryside, meeting farmers' groups, conducting demonstration plots and field days involving large gatherings, and working at regional research stations. 46. No future phase should be considered unless the institutional issues that have plagued research in past years are solved. Similarly, further involvement in extension should be conditional on the preparation of a strategy with clear goals, involving farmers' communities and all potential stakeholders at the onset. Finally, should the Government still consider agriculture a development priority, the many challenges of the sector would be best addressed through an Agricultural Support Services Project that reflects Algeria's rapidly evolving agricultural sector. I. KEY LESSONS LEARNT 47. The key lessons learnt from project implementation include the following: (a) Major institutional issues should be solved as a condition for negotiations - this would have prevented INRAA from spending precious time away from project implementation. (b) Programming and budgeting by objective is a complex approach that took INRAA nearly four year to adopt properly. Thus, the local context, capacity, and limitations should be carefully assessed and analyzed before planning such an approach. (c) When research embarks on complex restructuring and on the adoption of a fundamentally different approach, fast disbursement should not be expected from the project onset. (d) Modifying the approach to extension, especially at the national level, requires having a vision and defining a strategy at an early stage, together with extensive 12  and effective participation by farmers. For lack of both at project onset, it took INVA more time thn expected to play the vital role envisioned, of testing various approaches to extension on a pilot basis. (e) Given the diversity of the country and the farmers, it is likely that the T&V approach to extension, by itself, would prove inadequate on a national scale. (f) When elaborating a strategy, decisionmakers must decide whether extension should remain the sole responsibility of the public sector or whether it should be a mix of public and private interventions. They must also detennine the level of farmers' contributions to the cost of extension, and the amount of funds the Government will commit to extension on a regular and long-term basis. 13  PART I: STATISTICAL ANNEXES Table 1: Summary of Assessments A. Achievement of objectives Substantial Partial Negliible Not Anplicable Macro policies Sector policies [l iii] Financial objectives I [I] institutional development I ] L |l] Physical objectives I LIII] Poverty reduction ] L i [Iii] Gender issues [I I I II [I I Other social objectives Environmental objectives |I ] I Public sector management r ] [ [1111] Private sector development r I Other (specify) [ B. Project ssainilitv Likelv Unlikely Uncertain (/) (/) (v) W W LIII  Highly C. Bank Derformance satisfactor Satisfactor Deficient V)~~~V Identification [i ] [l] Preparation assistance l_l_____ Appraisal [ Supervision | 7 1 I-iRblv D. Borrower performance satisfactorv Satisfactory Deficient (v) (1) (8/) Preparation [ Implementation ] l Covenant compliance Li Operation (if applicable) l Highl i&l E. Assessment of outcome satisfactorv Satisfactorv Unsatisfactony unsatisfactorv IZI Lii [11 11 LIII] Note: Overall rating is highly unsatisfactory based on the many constraints and difficulties experienced by INRAA. The latter's main achievement was institutional, however its actual impact was barely expressed during the project lifespan. Extension faced less difficulties. It managed to reorganize extension and to propose a national extension system. Its impact which can only be properly assessed in the medium to long term was starting to show as the project came to a close.  Table 2: Related Bank Loans/Credits Loan/credit title Purpose Year of Status approval 1. Sahara Regional Rehabilitation of southern oasis 1991 Ongoing Development and on-farm development, (Loan 3405-AL) supported by institution strengthening, technical assistance and training. 2. Forestry & Watershed Loan Develop a long-term action 1992 Ongoing (Loan 3487-AL) program to better protect/manage Algeria's natural resources. Institution l_____________________ strengthening. 3. Cadastre Strengthening the institution 1991 Ongoing (Loan 3488-AL) implementing the cadastre and support for increasing registration for both rural and urban land. 4. El Kala National Park and Mitigate the degradation of the 1993 Ongoing Wetland Complex biodiversity within the El Kala Management (GEF) National Park and Wetland (Grant TF-28641) Complex. 5. Rural Employment Loan Generate and stimulate 1997 Ongoing (Loan 4143-AL) sustainable rural employment in rural areas of Northwest Al Table 3: Project Timetable Steps in project cycle Date planned Date actual/Latest estimate Identification 05/15/87 05/15/87 Preparation 05/26/89 05/26/89 Appraisal 06/09/89 06/09/89 Negotiations 03/26/90 03/26/90 Board presentation 06/05/90 06/05/90 Signing 09/26/90 09/26/90 Effectiveness 06/14/91 06/14/91 Mid-term review 06/26/95 06/26/95 Project completion 06/30/96 06/30/96 Loan closing 12/31/96 12/31/97  Table 4: Loan Disbursements: Cumulative Estimated and Actual (US$ million) FY 90 FY 91 FY 92 FY 93 FY 94 FY 95 FY 96 FY97 Appraisal Estimate 2.0 3.5 8.5 15.0 23.0 29.0 32.0 19.44a) Actual 2.0 2.0 2.3 4.9 4.9 7.1 8.5 10.0 Actual as % of estimate 100% 57% 23% 32% 21% 24.5% 43.7% 51% Date of final disbursementb) 30 April 1998 Last cancellation May 14, 1996 12.56 ') Although the final disbursement date had been set at April 30, 1998, the total amount disbursed so far is estimated at about US$10.0 million (