India - Second National Agricultural Research Project

Document of The World Bank FOR OFFICIAL USE ONLY Report No. 16612 IMPLEMENTATION COMPLETION REPORT INDIA NATIONAL AGRICULTURAL RESEARCH PROJECT II (CREDIT 1631-IN) May 27, 1997 Agriculture and Water Operations Division Country Department II South Asia 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 = Indian Rupees (Rs.) SAR (Septembcr 1985) Rs. 12.00 = US$ 1.00 Completion Year (June 1966) Rs.35.08 = US$1.0 WEIGHTS AND MEASURES (Metric System) FISCAL YEAR OF BORROWER April I to March 31 ABBREVIATIONS AND ACRONYMS ADR Associate Director of Research ARIS Agricultural Research Information System CIAE Central Institute of Agricultural Engineering CIHNP Central Institute of Horticulture for Northern Plains CIRB Central Institute for Research on Buffaloes CIMMYT International Maize and Wheat Improvement Center DOA Department of Agriculture FAO Food and Agricultural Organization of the United Nations FTC Farmer Training Center GOI Government of India IARI Indian Agricultural Research Institute ICAR Indian Council for Agricultural Research ICRISAT International Crop Research Institute for the Semi-Arid Tropics IDA International Development Association IGFRI Indian Grassland and Fodder Research Institute ISNAR International Servicc for National Agricultural Research IRRI International Rice Research Institute KGK Krishi Gyan Kendras KVK Krishi Vigyan Kendra MOU Memorandum of Understanding NAARM National Academy of Agricultural Research Management NARP National Agricultural Research Project NBSS&LUP National Bureau for Soil Survey & Land Use Planning NGO Non-governmental Organization SAR Staff Appraisal Report SAU State Agricultural University SMS Subject Matter Specialist ZREAC Zonal Research and Extension Advisory Committee ZRS Zonal Research Station Vice President M. Nishimizu Director R. Drvsdale Division Chief/Manager S. Barghouti Staff Member Ashok Seth. Principal Agriculturist  FOR OFFICIAL USE ONLY IMPLEMENTATION COMPLETION REPORT INDIA NATIONAL AGRICULTURAL RESEARCH PROJECT II (Credit 1631-IN) Table of Contents PREFACE ..........................................................i EVALUATION SUMMARY ........................................................... ii PART I: PROJECT IMPLEMENTATION ASSESSMENT ..........................................................1 A. PROJECT OBJECTIVES ...........................................................1 B. ACHIEVEMENT OF PROJECT OBJECTIVES .......................................2.......2 Sector Policies ........................................................... 2 Phvsical Objectives ..................... .......................................3 Institutional Development . ...........................................................7 C. IMPLEMENTATION RECORD AND MAJOR FACTORS AFFECTING THE PROJECT. ........................................................... 12 Factors Not Generally Subject to Government Control .............. ............... 12 Factors Generally Subject to Government Control .................. .................. 13 Factors Generally Subject to Implementing Agency Control ........... ........... 13 D. PROJECT SUSTAINABILITY . .......................................................... 13 E. BANK PERFORMANCE .......................................................... 14 F. BORROWER PERFORMANCE .......................................................... 15 G. ASSESSMENT OF OUTCOME .... ...................................................... 15 H. FUTURE OPERATIONS ..................... ....................................... 15 I. LESSONS LEARNED ................................................. 16 PART II: STATISTICAL ANNEXES ............... ............................................. 19 Table 1. Summary of Assessments .......................................................... 19 Table 2. Related Bank Loans/Credits ........................................................... 21 Table 3. Project Timetable .......................................................... 22 Table 4. Loan/Credit Disbursements: Cumulative Estimated and Actual ....................................... 22 Table 5. Key Indicators for Project Implementation .......................................................... 23 Table 6. Key Indicators for Project Operation ........................................................... 23 Table 7. Studies Included in Project ........................................................... 24 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.  Table 8A. Project Costs .......................................... 25 Table 8B. Project Financing .......................................... 25 Table 9. Economic Costs and Benefits ......................................... 25 Table 10. Status of Legal Covenants .......................................... 26 Table 11. Compliance with Operational Manual Statements ................. ........................ 27 Table 12. Bank Resources: Staff Inputs .......................................... 27 Table 13. Bank Resources: Missions .......................................... 28 APPENDICES A. Sununary of Location, Implementation Period, Costs and Outcome of Sub-projects 32 B. Mission's Aide-Memoire 40 C. Borrower's Assessment and Comments on the ICR 54 . .  i IMPLEMENTATION COMPLETION REPORT INDIA NATIONAL AGRICULTURAL RESEARCH PROJECT II (Credit 1631-IN) PREFACE This is the Implementation Completion Report (ICR) for the National Agricultural Research Project II (NARP II) in India for which Credit 1631-IN in the amount of SDR69.6 million (US$72.1 niillion equivalent) was approved on February 25, 1986 and became effective on April 21, 1986. As part of the funds redeployment exercise, a total of SDR14.37 million (US$14.886 million equivalent) was cancelled, SDRI 1.8 million in December 1991, and SDR2.57 million in June 1993. The Credit was closed on June 30, 1996, after three one-year extensions of the original seven-year implementation period. Final disbursement from the Credit was made on December 13, 1996 at which time the revised credit amount of SDR55.23 million (US$76.67 million equivalent) was fully disbursed. The ICR draft was prepared by a mission from the FAO/World Bank Cooperative Program which visited India in June/July 1996. It was finalized by Mr. Ashok Seth and Ms. Teresita Estoque, Agriculture and Water Operations, Country Department 2, South Asia Regional Office, and reviewed by Mr. Shawki Barghouti, Division Chief, and Ms. Kazuko Uchimura, Project Adviser, SA2DR. Preparation of the ICR was based on material in the project files, field investigations and discussions with staff of the World Bank, the Indian Council of Agricultural Research (ICAR), the Indian Agricultural Research Institute, and State Agricultural Universities (SAUs), who were associated with the project. Representatives of the SAUs in Andhra Pradesh (AP), Himachal Pradesh (BP) and Madhya Pradesh (MP) contributed to the preparation of the ICR by arranging field visits and meetings with field staff and farmers involved with the project. Additional information was obtained from meetings with representatives of the State Departments of Agriculture of AP, HP and MP, from individual Impact Assessment Reports from each SAU, a workshop on project outcome held on July 3, 1996 with several Vice Chancellors and Directors of Research of most SAUs, an implementation completion report prepared by ICAR, and from comments on the mission's Aide-Memoire. These are available in the Asia Information Center. ICAR's comments on this ICR and the summary of Borrower's Assessment of the project are in Appendix C.  ii IMPLEMENTATION COMPLETION REPORT INDIA NATIONAL AGRICULTURAL RESEARCH PROJECT II (Credit 1631-IN) Evaluation Summary Introduction 1. The first National Agricultural Research Project (NARP) (Credit 855-IN), implemented from 1978 to 1985, was designed to decentralize research to various geographical areas of the major states of India, by strengthening the capability of the State Agricultural Universities (SAUs) to conduct research in defined agro-ecological zones, and to give special emplhasis to cereals, pulses and oil seeds under rainfed and mixed farming conditions. Project Objectives 2. The general objective of the seven-year National Agricultural Research Project II (NARP II) was to intensify and continue the process initiated under the first NARP for the strengthening of the capability of SAUs to conduct location-specific research in the main agro- ecological zones, and to extend its scope to include the priority agricultural research needs of each zone under rainfed and irrigated conditions. The activities to be supported included completion of unfinished NARP sub-projects, expansion into new areas of research, comprising irrigated farming, animal-drawn equipment, horticulture and commercial field crops, agro-forestry and animal nutrition, and strengthening facilities at regional research stations for training subject- matter specialists (SMSs) of state Departments of Agriculture (DOA). After a mid-term review in 1989, an additional objective was the development of computer facilities for analysis and processing of research information and administrative functions at research stations, and a national computerized agricultural research information system. The project objective of continuing support for the establishment of a decentralized research system was important to the concept of restructuring research to meet the highly diversified needs of the farming systems throughout India, and to focus research on solving local problems and exploiting opportunities to increase agricultural production, with a strong emphasis on adaptive research. Implementation Experience and Results 2. Achievement of Objectives. Overall, the project has had a satisfactory outcome in developing a decentralized agricultural research system in 17 major states of India by supporting the development of a network of research stations under 27 SAUs.  iii 3. Development of Facilities. The project completed the establishment/upgrading of 120 zonal research stations (ZRSs) and 222 sub-stations, through the construction of 1,736 buildings and miscellaneous civil works, positioning of 995 incremental scientific staff and 954 incremental support staff, and provision of a considerable amount of equipment and vehicles to support a substantial program of adaptive research under some 220 sub-projects (SPs). The project had an important element of human resource development, including the provision of training facilities and training equipment; and loca, and overseas training courses for more than 1,400 scientists. 4. Research Programs. Research staff at the ZRSs are well motivated and have developed research programs relevant to farmers' needs. The comparative isolation of the stations has necessitated strong support, within resource limits, from the senior research staff at the SAUs, as well as the need to provide opportunities for staff training and ready access to up-to-date scientific information. The zonal research-extension advisory committees and joint visits with extension staff to farmers' fields have ensured that the research staff are more familiar with farmers' problems on the major crops. On the other hand, the combination of weak research programs at the SAUs, and sparse participation in meetings by the line agencies for horticulture, animal husbandry, forestry and fisheries, has resulted in an insufficient flow of new technologies for these sub-sectors. More basic research on ten key topics was supported under a basic research sub-projects undertaken jointly by ICAR institutes, SAUs and one private institution. This has performed an important function in underpinning other programs of applied and adaptive research. 5. Linkages. The project was designed to improve the linkages between research and extension, as well as to promote collaboration and linkages among SAUs. While the linkages between the ZRSs and state DOAs were initially generally satisfactory, there is more recent concern over the cessation or frequent delay of regular research-extension-farmer meetings and scheduled training programs for agricultural extension staff, and lack of significant inputs from the other line departments of the State Governments into the research programs. Initiatives to improve collaboration between SAUs with common research goals have included a series of meetings on research for zones that extend across state boundaries, particularly the tribal areas, and a series of regional meetings to review progress in the implementation of the project, and to discuss common problems. While these meetings provided useful forum for review of implementation progress, they did not result in significant shift in focus of program to "common" problems. 6. Management of Research. A major responsibility of SAU research management is to set research priorities for each state and its zonal components and to develop a research strategy and plan to implement these priorities. While there has been a considerable amount of informal identification of priorities for research, by way of zonal status reports and research reviews of the SAUs and zonal consultations, there is no formal system for setting priorities and re-assessing the present allocation of staff and financial resources to meet these priorities. The  iv establishment of a computerized Agricultural Research and Information System (ARIS) will help this process, as well as provide management, scientific literature and natural and physical resources information. The design of the system is complete and the computer hardware and software ordered would be progressively installed in selected locations throughout the system. 7. Project Sustainability. While sustainability of the decentralized research system is likely, it will require continuing financial support from ICAR and increased support from State Governments. SAUs are presently considering possibilities for income generation but will also need to review the costs of the present system which, in some cases, presently supports too many research stations. In addition, the staff of the ZRSs, particularly in more remote areas, will need strong professional and social support from SAUs, and promotion of reward systems that recognize contributions to increased farm productivity. 8. Project Costs, Financing Arrangements and Implementation Timetables. Of the total IDA Credit of SDR69.6 million (US$72.1 million), SDR55.23 million (US$71.91 million) has been disbursed. A total of SDR14.37 million (US$14.886 million) of the Credit was cancelled during the funds redeployement exercise in December 1991 and May 1993. The Credit was closed on June 30, 1996 after three one-year extensions of the original seven-year implementation period. Final disbursement from the Credit was made on December 13, 1996 at which time the revised credit amount was fully disbursed. Total project cost is estimated at Rs.2,274 million (US$97.31 million), 73 percent above the original estimated project cost of Rs. 1,331.12 million (US$110.93) due to a 183 percent devaluation of the Indian rupee against the US dollar, but is 12 percent below the original estimated costs in dollar terms. Indicative financing estimates for sub- projects were provided in the SAR. Whilst plans were prepared and approved for each sub- project during project implementation, there were no explicit timetables for achievement of physical targets, and expenditure was considerably delayed, resulting in the need for three annual extensions of the project. 9. Key Factors Affecting Achievement of Major Objectives. The SAUs and ICAR gave continued high priority and attention to implementation of the project, and much was achieved in translating the concept of a decentralized research system into a functioning organization. Early supervision reports cited many cases of delays -- staffing, civil works and slow release of funds -- which, because of the large number of institutions involved, collectively added up to substantial problems. 10. Bank and Borrower Performance. The Bank undertook some 200 supervision missions during implementation of the project. Supervision missions were mostly staffed from the Bank's Resident Mission in India which provided more continuous contact and an ability to offer helpful and constructive advice as problems arose in this complex project. The Borrower encountered many problems in the earlier stages of the project, but as these were gradually overcome, the performance improved with a satisfactory, though delayed, completion.  v 11. Project Outcome. The project outcome is considered satisfactory as it achieved most of its major objectives. Those objectives not completed are anticipated to be attained in the near future. Summary of Findings, Future Operations and Lessons Learned 12. Important Findings. Restructuring a research organization that is national in scope and involves many autonomous research organizations is a complex undertaking, requiring associated changes in management systems and in the focus and attitude of scientists in developing innovative research for solving problems and fulfilling needs of local farmers and farming communities. Managing such changes in organization and research goals is an intricate process in which the SAUs, and particularly the Project Unit, had to learn through experience. A more explicit recognition of the difficulties experienced under the first NARP and the need for greater managerial support in the design of NARP II, might have avoided some of the implementation problems and delays. 13. Plans for Future Project Operations and Sustainability. The Borrower's Assessment (ICAR, June 1996) of the project, while generally indicating future commitment and needs, does not provide a specific time-bound plan for future development. However, from numerous discussions, it was clear that the SAUs remain committed to maintain the decentralized research system established. There is universal agreement that physical facilities are now generally adequate to meet research objectives. Future emphasis would be given to selective upgrading of knowledge and skills of scientists, and ensuring an attractive work environment, particularly at the more remote research stations, developing communications, computerization, media materials and information dissemination, increasing support for research on animal nutrition, fisheries, horticulture, silviculture and farm mechanisation, strengthening the weak research-extension linkages through an increase in the number of "front-line" on-farm trials and strengthening the Krishi Vigyan Kendras (KVKs) and Krishi Gyan Kendras (KGKs). 14. Lessons to be drawn from the implementation and outcome of this project include: (a) The concept of a decentralized research system, started under NARP and carried on under NARP II, based on a uniform system where research has become the responsibility of ZRSs of SAUs conducting adaptive research in local agro-ecological zones, is sound and its continued development is essential if the diverse farming systems in India, particularly the rainfed systems, are to benefit from new technologies. (b) Generally, most ZRSs, but not sub-stations, have adequate facilities and sufficient numbers of scientific staff. However, the nature of the farming systems and their great diversity makes it necessary to have more multi- disciplinary teams working in an inter-disciplinary context with more emphasis placed on financial viability of technologies, marketing potential, diversification (broad-basing) of farm production and relevance of  vi technologies to farmers circumstances. The SAUs need to give higher priority to horticulture, livestock, fisheries and agro-forestry research. (c) The project design had an important element of flexibility in that individual subprojects were prepared and approved jointly by SAUs and ICAR, and while indicative average costings were provided for each subproject, actual costs of subprojects were need-based. In practice this meant that GOI provided 100 percent financing to the SAUs for approved subprojects and funds were advanced six months ahead of expenditure. This allowed greater support to be given to resource poor states and remote or backward areas, leading to a more comprehensive decentralized research system. Successful implementation of subprojects, however, required more explicit timetables for expenditure and physical targets to be met and closer monitoring and supervision by the implementing agency. (d) The establishment of a productive and creative research system is a long term process, in which the present system is at a relatively early stage of development. The further advancement of ZRSs will depend on adequate financial support to operate and maintain facilities and research programs, improved management of research resources (better priority allocation, planning, sharing of common resources, monitoring and evaluation) and continuing improvement in the research capacity of the scientific staff (often isolated from the mainstream of scientific activity), requiring specific attention to training, reward systems and strong technical support from the SAUs and ICAR. (e) Financial sustainability of the research system developed is presently jeopardized in some states by the indicated inability of some State Governments to fully fulfill the terms of their MOUs to financially support the system after the closure of the project. Lack of funds will particularly adversely affect research operations, maintenance of facilities and equipment and capital replacement. Although some SAUs have considered possible revenue earning opportunities to improve financial viability, these will always be limited. It is essential for SAUs to use their financial, physical and staffing resources more efficiently, limit the number of sub- stations, and improve cooperation and collaboration in research programs among states with similar agro-ecological zones. (f) Despite the emphasis given to establishing and strengthening research- extension-farmer linkages in the many research and extension projects funded by the World Bank in India, these linkages still remain weak. The establishment of formal linkages (zonal workshops, training for SMSs of extension departments, joint field days, etc.) between research and the line agencies of state departments (agriculture, forestry, livestock and fisheries),  vii has not, except for crop production, engendered strong cooperation, information exchange and collaboration in technology generation and transfer to farmers. It is essential that the extension departments of the SAUs and state governments and the research agencies (SAUs and ICAR) develop novel and more effective approaches to improving collaboration. (g) The project's implementation benefited from strong expertise in agricultural research of the Bank's Resident Mission staff who supervised the project.   1 IMPLEMENTATION COMPLETION REPORT INDIA NATIONAL AGRICULTURAL RESEARCH PROJECT II (Credit 1631-IN) PART I: PROJECT IMPLEMENTATION ASSESSMENT A. PROJECT OBJECTIVES 1. The first National Agricultural Research Project (NARP), implemented from 1978 to 1985, was designed to decentralize research to various geographic areas of the states, by strengthening the capability of the SAUs to conduct research in defined agro-ecological zones, in support of reformed state-wide agricultural extension projects (15 IDA-financed extension projects in 17 states between 1977 and 1995). This was to be brought about through improvements to existing university-based agricultural research and concentrated on research oriented towards the needs of the main agro-ecological zones, with special emphasis on cereals, pulses and oilseeds, under rainfed and mixed farming conditions. 2. The general objective of the Second National Agricultural Research Project (NARP II) was to intensify and continue the strengthening of the capability of State Agricultural Universities (SAUs) to conduct location-specific research in the main agro-ecological zones, initiated under the first NARP (Credit 855-IN), and extend project scope to include the priority agricultural research needs of each zone under rainfed and irrigated conditions. Specifically, NARP II aimed to: (a) complete the unfinished NARP subprojects (those that had not completed five years of operation); (b) expand the scope of NARP to cater to the priority needs of each agro-ecological zone, including: (i) new areas of research, such as irrigated farming, animal-drawn farm implements, horticulture and commercial field crops, agro-forestry and animal nutrition; (ii) developing field programs for field testing and refining of research results; and (iii) strengthening of training facilities at Regional Research Stations for training Subject Matter Specialists (SMSs) of the state Departments of Agriculture (DOA); (c) provide funds for individual research activities on priority topics emerging or identified during project implementation and not covered under (b) above; (d) strengthen research administration of two SAUs (in the states of West Bengal and Jammu & Kashmir) not supported under NARP; and (e) continue funding of the Project Unit established in the Indian Council of Agricultural Research (ICAR) under NARP, including provision for training of research scientists. NARP II became effective in April 1986 and was originally scheduled to be completed by June 30, 1993, but was given three annual extensions (to June 30, 1996) to allow full utilization of the Credit, and to complete the original and revised  2 project objectives. NARP II covered 17 major states of India and was implemented by 26 SAUs plus one Veterinary and Animal Sciences University (in the state of Tamil Nadu), covering 120 agro-ecological zones. 3. Consequent to a mid-term review in May 1989, approval for project restructuring was not given until June 1991, and appraisal of new project proposals did not occur until July 1992. Additional project objectives introduced following project restructuring were: establishment of a national computerized agricultural research infortnation system; establishment of computer facilities for scientists and administration units at all SAU research stations; initiation of selected basic research programs at ICAR institutes and SAUs to support regional programs; and rehabilitation of research stations damaged by a cyclone in the State of Andhra Pradesh. 4. The project concept supported the Bank's country strategy aimed at assisting India to achieve self-sufficiency in foodgrains, and to increase farm incomes in an equitable manner. It also responded well to the Borrower's circumstances and the development priority for the agricultural sector oriented towards the need to develop locally specific technologies to improve and diversify agricultural productivity given the ecological diversity of the country. Overall, the timing of the project, its scope and the relative size of the sub-projects were appropriate to the need to complete the decentralization of the agricultural research system initiated under NARP. Many of the risks identified in the Staff Appraisal Report (SAR), such as irrelevance and/or non-adoption of research findings, institutional constraints in project implementation and inadequate State Government funding, all occurred to a certain extent, but collectively only had a moderately adverse effect on project outcome. B. ACHIEVEMENT OF PROJECT OBJECTIVES Sector Policies 5. The project responded to a request from the Government of India (GOI) to the International Development Association (IDA) to provide further assistance to establish an effective and efficient decentralized research system in India to support the national goal of self-sufficiency of foodgrains (production of 168-170 metric tons by the terminal year of the Seventh Five Year Plan - 1990), which would require an annual growth rate of 2.7 percent. In fact, by 1990 total foodgrain production was 194 metric tons and self-sufficiency was achieved and the annual growth rate for the decade 1984-94, was 3.0 percent. The project has contributed substantially to the achievement of these sector objectives through the development of locally-applicable, improved technologies, although the precise extent of this contribution cannot be estimated (paras 32 and 33).  3 Physical Objectives 6. Sub-projects. The project was to finance the spillover costs of 90 subprojects (12 administrative, 76 research and 2 training subprojects) initiated under NARP until they completed five years of operation; in fact, 14 administrative, 74 research (including one each to assist the Indian Agricultural Research Institute [IARI] and National Academy of Agricultural Research Management [NAARM]) and 4 basic research subprojects were completed under NARP II. In addition, proposed new subprojects to be financed under NARP 11 were to be of three kinds: (a) supplemental subprojects for each Zonal Research Stations (ZRSs) already established under NARP (82) and designed to broaden their research activities/scope; (b) new subprojects to establish 39 more ZRSs; and (c) new administrative subprojects (2); in fact 86 supplemental subprojects, 37 new subprojects and 5 administrative subprojects were completed under NARP II. Following project restructuring, 10 additional subprojects were initiated to further support library development, complete civil works, purchase additional vehicles and equipment, purchase computers, establish an agricultural research information system (ARIS), support new basic research initiatives in selected topics, and provide overseas training for selected scientists and research managers. 7. An innovative project design feature was for detailed proposals for new subprojects to be prepared during the first three years of the project. Initiation of subprojects followed approval at different times during project implementation but all were financed for only 5 years (refer Appendix A), with the SAUs expected to continue financing beyond that time. Sanctioning of new subprojects was considerably delayed (only 50 percent were approved by original planned date of September 1988) because of the slow pace of preparation, appraisal and approval by the Inter- disciplinary Scientific Panel (ISP) and Project Funding Committee (PFC), and weak financial and technical capability of certain SAUs. Only the recently launched subprojects concerned with basic research and the establishment of ARIS have not been completed. The detailed implementation assessment of expenditure and performance of each subproject is given in Appendix A. Overall, a minority of subprojects were assessed as having no significant problems with implementation, but the majority had moderate problems with staffing, civil works, extension linkages, technical and management aspects, while a small number had major or severe problems. The definition of more explicit timetables for implementation of each subproject and closer monitoring and evaluation (M&E) by the ICAR Project Unit would have avoided many of the problems encountered. The subprojects supported the development of not only the 120 ZRSs but also of 222 sub-stations. 8. Incremental Staff. The majority of additional staff recruited consequent to the upgrading and expansion of the zonal research system of the SAUs were employed during NARP. Under NARP II, a further 995 scientific positions and 954 supporting staff positions were sanctioned to provide sufficient manpower for additional stations upgraded under the project. About 26 percent of scientific posts and 29 percent of support posts sanctioned remained vacant at  4 project closure, with more vacancies at sub-stations. Some ZRSs have never had a complete staff complement due to their remote locations and lack of specialized staff. SAUs have indicated they would maintain staff numbers and absorb additional salary transfers to SAU budgets following project completion. 9. Civil Works. A program of construction of new facilities for research stations in 82 agro-ecological zones was initiated under the first NARP but were incomplete at the closure of the project. The civil works program under NARP II financed the completion of all but one of these incomplete facilities as well as new facilities proposed under NARP II/; some minor variation in covered floor area was required to accommodate cost escalation. Total cost was about Rs.740 million (32 percent of total project costs compared with the original estimate of Rs.319 million which was 24 percent of the original estimated project cost). While facilities inspected by the ICR mission in Himachal Pradesh were of good quality, those in Madhya Pradesh and Andhra Pradesh seemed to have suffered from poor quality construction and poor maintenance. Water damage to buildings has been aggravated by the flat roof and poor drainage design used. A special subproject to rebuild two research stations severely damaged by a tropical cyclone in Andhra Pradesh was successfully implemented, in association with the Andhra Pradesh Cyclone Emergency Reconstruction Project (Credit 2179-IN/ Loan 3260-IN). 10. Equipment and Vehicles. About Rs.800 million (36 percent of total project costs) has been spent on equipment and vehicles across the 120 ZRSs and some ICAR institutes to ensure reasonable laboratory, field and office equipment, including computers, as well as mobility for research staff to undertake field research. Considerable delays occurred in purchase of vehicles with only 19 out of 120 vehicles purchased by the end of 1993. Delays also occurred under the restructured subprojects to provide additional vehicles, due to delays in the sanctioning of funds to ZRSs. 11. Operational. Research operational costs were about Rs.300 million (13 percent of total project cost) and were a little higher than originally estimated. It was apparent that maintenance expenditure was minimal. It has been indicated from discussions with SAUs that, following project completion, budget allocation from the state governments would be expected to become insufficient for maintenance and capital replacement and proper functioning of the research programs in many SAUs. 12. Training Facility Development. As part of the upgrading/establishment of facilities at SAUs, training facilities and hostels were constructed at each ZRS. These facilities were used for training the Subject Matter Specialists (SMSs) of the state DOAs. The research staff of the ZRSs " A total of 1,736 new facilities, including: laboratory cum office buildings (290), godowns (37), training halls (40), training hostels (41), scientific and supporting staff residences (869), other laboratories (67), sheds (167), stores (84) and miscellaneous civil works (139).  5 provided the training, while funding for the monthly training program was provided under the NAEPs. The national level training sub-project provided funds for training SAU scientists (para. 16) and the development/upgrading of physical training facilities at six ICAR institutes . Key training equipment was also provided. The total cost of the subproject is estimated at Rs.55 million. 13. Computerization and Development of Information Systems. The national agricultural research system in India consists of more than 25,000 scientists, working at more than 80 large institutes, 27 state universities and over 400 zonal stations and centers. Support for the development of a national Agricultural Research Information System (ARIS) to link all of these research agencies was given under a new subproject following project restructuring. Development of the system concept was a joint effort between ICAR and ISNAR (project-financed technical assistance), and was designed to cover information for management (research projects, human, financial and physical resources), scientific information (i.e. library documentation) and information on natural and physical resources. Such a system aims to improve information management, support stronger research planning, monitoring and evaluation, and thus improve research relevance and quality. 14. Five years have elapsed since the original concept of the system was proposed by the GOI in 1991. Its present non-functioning is a reflection of both its complexity and the delays in decisions at each step in the development process. ISNAR was not selected to develop the system concept/design until May 1992, but payments from ICAR were delayed until November 1993. Subsequently, the hardware design of the system, including the equipment specifications and quantities, was prepared by the National Center for Software Technology, as consultants. Bids were opened in August, 1995 and World Bank approval for the purchases given in February, 1996. The equipment includes 56 Unix Servers (mini-computers) for the 27 State universities, ICAR institutes and bureau, and 230 LAN servers (desk-top computers) for each of these, as well as the zonal research stations. The remainder of the equipment includes work stations, printers, modems and air-conditioners. 15. When installed, the system will be centralized in ICAR with connectivity to SAUs. Local area networks will be established at SAUs, using a fiber-optic system and they will be inter- connected using terrestrial and satellite communication systems. The total amount committed for establishing ARIS was Rs.263 million, while the total expenditure under the project at 31 March 1996 was Rs. 130.1 million. High priority must now be given to the successful establishment of a functioning and useful system, as designed. This should be done before there is any further consideration of expanding or making additional large investments in the system under a new project. The development of a successful system will depend on the commitment of managers and " National Bureau for Soil Survey and Land Use Planning, Indian Grassland and Fodder Research Institute, Central Institute of Horticulture for Northern Plains, Central Institute of Agricultural Engineering, Central Institute for research on Buffaloes and the Indian Agricultural Research Institute.  6 scientists at all levels and can be assisted by the development of a series of indicators for assessing the degree and quality of implementation and use, as well as extensive computer training courses for all potential users. 16. Staff Training (ICAR and SAUs). Three types of national and international training programs for scientists and research managers were supported by the project. Some 1,318 SAU scientists undertook 82 one-month training courses in six ICAR institutes (para. 12) from August 1989 to September 1995 to broaden their knowledge and skills in soil survey and land use planning, fodder and forestry, horticulture, farm implements, animal nutrition and water management. The training programs suffered from poor trainee participation from SAUs, slow implementation of the civil works programme and equipment purchases (para. 12) and disagreements in regard to level of grants for boarding, lodging and travel. 17. Technical assistance for conducting the overseas training program was provided by FAO. A total of 87 young scientists from SAU research stations undertook three-month overseas training at foreign national and international research centers in areas of biotechnology, crop hybrid breeding, integrated pest management, nutrient management systems, water management, energy management and integrated systems from April 1995 to June 1996. In addition, 33 senior research managers from ICAR and SAUs undertook 15-day study tours to visit international and national research centres from January 1994 to April 1996. Advantage was not taken to utilize the considerable resources of NAARM for research management training. All study fellows were required to produce reports on their training. Total cost of the international subproject is estimated at Rs.39 million for training of young scientists and research managers. 18. Computer training has been given to 202 staff from ICAR headquarters and a total of 276 staff from ICAR institutes, SAUs and other institutions. Training has been in computer and software use, networking, communications and trouble shooting for scientists and for staff associated with library, financial, personnel and information management. ICAR accepts that this level of training is completely inadequate considering the thousands of potential users of the ARIS. Accordingly, provisions have been made in the Ninth Plan of ICAR to increase the numbers of trainees as well as the range of computer-related training programs. 19. Total Project Costs. Of the total IDA Credit of SDR69.6 million (US$72.1 million equivalent), a total of SDR14.37 million (US$14.886 million equivalent) of the Credit was cancelled (in December 1991 and June 1993). With the final disbursement on December 13, 1996, the revised credit amount of SDR55.23 million (US$76.67 million equivalent) was fully disbursed. 20. Total project cost at March 1996 is estimated at Rs.2,274 million (US$97.31 million), which is 73 percent above the original estimated project cost of Rs. 1,331.12 million (US$110.93 million), due to a 183 percent devaluation of the Rupee against the Dollar, and is 12 percent below  7 the original estimated project cost in Dollar terms. The Government of India (GOI) financed about 25 percent and 22 percent of the total project costs between 1986 to 1993 and 1994 to 1996, respectively, compared with the loan agreement of 65 percent reimbursement for the first 7 years and 90 percent for the last three years (i.e. 35 percent and 10 percent GOI financing, respectively). Disbursements were delayed throughout the duration of the project with only 75 percent of the Credit (in US$ equivalents) disbursed by the original date of Credit closure (June 30, 1993). A special feature of the financing mechanism that was extremely beneficial to the SAUs was that the GOI provided 100 percent financing to the SAUs for approved subprojects and funds were advanced 6 months ahead of expenditure. 21. It is concluded that by closure, most of the proposed physical facilities (laboratories, offices, training facilities, and staff residences), vehicles and equipment (laboratory, field, office and training) and additional staff were in place and proposed training programs completed. Thus the physical objective of establishing adequate research facilities with an adequate number of suitably trained staff in each agro-ecological zone in India was substantially achieved. Institutional Development 22. Development of Zonal Research Capacity. Inspite of the problems with delays in implementation of the civil works program, the project has been successful in staffing the zonal stations and in providing sufficient infrastructure and equipment to carry on a substantive program of adaptive research. The system of developing subprojects (guidelines provided in SAR and followed by ICAR) ensured an equitable availability of funds to each ZRS, depending on their needs. At the small sample of stations visited by the ICR preparation mission, the staff were enthusiastic and committed to their job and were keen to demonstrate their research. The field visits and workshop (July 1996) attended by SAU vice-chancellors and directors of research confirmed that there had been a successful restructuring (decentralization) of the research system. This restructuring has: (a) improved the system's ability to respond to clients' needs; (b) improved the quantity of adaptable technologies; (c) improved to some extent the linkages with other organizations, particularly state agricultural extension departments; and (d) responded to staff interests by improving their work and social conditions. However, some major underlying concerns were expressed by scientists about future funds for their work, lack of mobility and sometimes lack of support from SAU headquarters. 23. Each ZRS has between 20 and 30 research staff, covering the major disciplines in the biological and physical sciences. However, there is also an important role for the few social scientists positioned at the ZRSs. They are required to work with the other researchers on diagnosis of farmers' needs, to ensure that the research has a farming systems perspective. They should also be involved with the extension system in developing an understanding of how best to commnunicate  8 with farmers. Most states are now preparing a zone-wise package of practices for farmers based on the outcome of adaptive research in zones, in place of existing state-wide recommendations. 24. Relevance and Quality of Zonal Research. Under NARP, 120 agro-ecological zones agro-ecological zones were defined in 17 states in which climate, topography, soils and land use have many common features' . Detailed "Status Reports" on each of these zones were started under NARP and completed under NARP II, but some as late as 1993. The detailed guidelines and a few model reports prepared by ICAR for the SAUs ensured that there was a degree of uniformity in reporting formats and that information contained provided a sound basis for developing research programs on major crops and cropping systems in local areas. Some of the reports have been updated but all reports should be periodically revised to ensure accuracy of the agricultural data as farming systems change. A major deficiency in definition of agro-ecological zones was that they were state based and therefore their boundaries were defined by political rather than purely natural factors. Thus agro-ecological zones and the Status Reports prepared for them, do not cross state boundaries which creates zonal duplication in adjacent states (it is suggested by the Project Unit that in using only agro-ecological criteria for definition of zonal boundaries, only about 65-70 discrete agro-ecological zones actually exist in India and therefore technically about 50% of the ZRSs are unnecessary). There is considerable merit, therefore, in consolidating information on natural and physical resources and research priorities, plans, programs and output for the same agro-ecological zones across state boundaries. When ARIS is developed, it could be used to consolidate and analyze such information according to natural boundaries. Such a system can be used to both identify research gaps, promote coordination, cooperation and complementarity in research, and avoid duplication. 25. A standard procedure for preparation and approval of research programs for ZRSs has been established to ensure relevance of research. Research program approval follows a sequential path of endorsement from bi-annual meetings of zonal research and extension advisory committees (ZREAC), the SAU, and finally at the state level by the State Program Planning and Implementation Committee and ICAR. While the system of consultation, through regular meetings of the ZREACs and organization of joint field visits by research and agricultural extension staff, has ensured that research at the zonal stations is relevant to the crop problems in the zone, there has been insufficient research emphasis on other aspects of the local farming systems. Good attendance and participation by staff of the state Department of Agriculture (DOA) at the SAU regional group meetings has provided a valuable input in defining research priorities. In contrast, attendance and participation by the other line departments -- horticulture, animal husbandry, forestry and fisheries - - and feedback from them has been poor. At present the research programs emphasize crops, and are dominated by a disciplinary approach -- mostly varietal improvement, pest and disease control, and fertilizer treatments. Future emphasis must be given to research with a farming systems " A description of the 120 agro-ecological zones and their production potentials and constraints is given in: Saxena, A.P. (1989) "Strategies for agricultural research and development - a zonal approach", ICAR, Delhi.  9 perspective which will often involve multi-disciplinary teams. The mission saw little evidence of such teams being in operation. One important objective of the project was to focus research on finding solutions to farmers' problems (i.e. demand-driven research), Thus the zonal research teams have been under considerable pressure to provide quick answers to the current problems of the farmers. However, it is equally important that future emphasis be given to exploring new opportunities to increase agricultural productivity, economic efficiency and farmer incomes, This may require more applied or strategic research in which collaboration with scientists at the SAU centers or ICAR institutes should be encouraged. 26. Several factors have affected the quality of research at ZRSs. While laboratory facilities are generally adequate, if limited, there needs to be a more planned approach to their use to integrate and facilitate the needs of different programs and disciplines for laboratory services and thus ensure greater efficiency. For example, crop, pasture, forages and agro-forestry research may require similar soil and plant analyses, but are undertaken by different laboratories. Although the project supported library development (equipment, books and journal purchases) at each ZRS, facilities developed are quite modest. Good library facilities are essential for research information and the ZRSs need to be better supported by abstract and document delivery services from the SAU or ICAR. When ARIS is operational, one of its important functions will be the supply of scientific informnation for the ZRS scientists but the management must ensure they make full use of its services. General computerization of activities at research stations and connectivity to internet/email services are also important. Supervision by experienced scientists is also a major factor in ensuring research quality hence the need for strong and continuous interaction between scientists at research stations and staff of the SAU center and ICAR institutes. 27. Basic Research. Ten research programs have been supported under a basic subproject, following project restructuring, proposed by an FAO/CP preparation mission in October 1991 and a World Bank appraisal mission in July 1992. The 10 research programs (field crop biotechnology, brassica cell biology, rice-wheat cropping systems, horticultural crops biotechnology, livestock biotechnology, livestock diseases, freshwater and brackish water aquaculture, resource characterization of rainfed farming systems, crop stress physiology and integrated pest management) are being jointly undertaken in 10 centers and 35 subcenters located in ICAR institutes, SAUs and one private institution. The programs are undertaking strategic rather than basic research, in as much as they are attempting to generate new knowledge about certain specific problems or to use sophisticated techniques like gene transfer to improve certain characteristics of plants. In this they perform a very important function in supporting advances in applied and adaptive research. Such strategic research will continue to be an important and probably growing facet of the research system, requiring continued and growing support from the system's own resources, as these are rapidly evolving research fields, and frequent updating of equipment and familiarization of staff with the changing research scenarios. Because the research system in India has relatively limited finances, expansion of strategic research needs to be carefully  10 planned around the likely availability of long-term resources rather than on possible future project funding. 28. Management of Research. The development of an efficient and effective decentralized research system depends to a large extent on managerial capability to prioritise research topics, plan appropriate research programs and monitor and evaluate research implementation and outcome. The project aimed to improve the capabilities of research managers through sponsoring study tours (para. 17) and through the development of a management information system (paras 13-15). However, the benefit to the project was not fully realized because: (a) management training was primarily focused on Directors of Research from SAUs (not Associate Directors of Research who manage the ZRSs), and a number of research directors from ICAR; (b) the study tours were undertaken during 1995-96, towards the end of the extension period of the project; and (c) ARIS is not yet functional. The project design gave no emphasis to management training. It is an essential ingredient to the development and functioning of an efficient and effective research system. The considerable resources of the NAARM at Hyderabad for providing a wide range of management training were not fully utilized.' 29. Research Linkages. Formal research linkages among SAUs and other research agencies appeared limited under the project. This was partly the fault of project design (artificial agro-ecological zone boundaries established within states and no facility to fund upstream research emanating from adaptive research in local areas) which did not ensure close collaboration among SAUs in adjacent states with common agro-ecological zones, or between SAUs and ICAR institutes and international research centers. Two very useful meetings on common agricultural problems in tribal areas of adjacent zones were held during the project, but this did not result in closer coordination, cooperation and collaboration in research between SAUs in the same or adjacent states. 30. Research-Extension Linkages. Improving extension linkages has been a major thrust of the NARP II, and the zonal station concept was designed so that research scientists would interact closely with the extension agents of the State development departments, NGOs and the farmers at zonal levels. The formal and informal mechanisms for such interactions were established under NARP and NARP II (state technical committees, ZREACs and monthly workshops and joint field trials, field days and farm visits). Generally the linkages with the state DOAs have been satisfactory, though there is now concern over the cessation or frequent delay of regular meetings and scheduled training programs for extension staff (some states have discontinued the regular training for SMSs). The lack of collaboration between research and other state line agencies involved with horticulture, livestock, forestry and fisheries has been a recurring concern during project implementation as these agencies generally showed little interest and ICAR disagrees with this statement (refer to Appendix Comments from Borrower).  11 commitment to the research programs, even though the project was designed to cover their interests. Although many of the Krishi Vigyan Kendras (KVKs) and Krishi Gyan Kendras (KGKs) mandated to transfer technologies to farmers are managed by SAUs, the project design did not give adequate attention to this important linkage in the process of technology transfer. l 31. It is concluded that, by project completion, the research system in India had been successfully decentralized with the establishment of an extensive network of ZRSs, in association with the SAUs, throughout the country. Within this research system, substantive improvements in the institutional capacity of the majority of ZRSs to undertake programs of adaptive research had been achieved. Relevance and quality of research had also been improved, although there was scope for greater improvement. Improvements in management of research and the development of effective functional linkages with other government agencies concerned with agricultural development were however much less than anticipated and therefore improvements to public sector management are only assessed partially. Project Benefits 32. Research System. NARP II and its predecessor NARP, were designed to bring about a major transformation in the structure of the research system. This was a large undertaking, both in terms of institution-building and in the transfer of much of the focus of research to a "grass roots" level with a strong emphasis on adaptive research. Such a change was particularly necessary to serve the needs of rainfed fanning where the farming systems are generally more complicated, more diverse over short distances and more risk-prone than irrigated farming where previous research had much of its impact. The tangible benefits of such a process are the development of the research stations, the provision of housing and laboratories, the creation of a cadre of scientific staff covering a wide range of disciplines, and the flow of new information and new technology. Equally important are the intangible benefits, such as changes in attitude to the needs of farmers, the ability to conceptualize innovative, problem-solving approaches to these needs, and the ability to develop research programs with a farming systems perspective. 33. Impact of Research on Agricultural Development. The ICAR took the initiative in planning and introducing impact studies into the project and provided funding and guidance for their publication. Twenty-one SAUs have produced these reports to date and these summarize the nature of the agro-climatic zones for which the SAU is responsible, the research undertaken and the output of the research system in terms of new varieties, pest and disease control techniques, fertilizer recommendations, new implements and new husbandry methods. To estimate impact, various time periods have been used as a comparison; some reports take the period 1981/82- 1991/92 while others compare pre-NARP, post-NARP and NARP II (up to 1991/92). From data ICAR disagrees with this statement (refer to Appendix Comments from Borrower).  12 generated on percentage adoption, which obviously varies with commodities, within zones and with technologies, additional yields (average yields with the new practice less average yields without it), benefits, and hence the impact, are calculated. Attribution of yield increase to research alone, as a few reports do, ignores the increases in inputs; the all-India use of fertilizers more than doubled between 1980/81 and 1990/91 while agricultural consumption of electricity (farm machinery, irrigation pumps, etc.) increased more than threefold. Nevertheless, the concept of impact studies is excellent and is to be encouraged, but clearly there needs to be a more rigorous approach to the question of measuring impact including a more substantial economic input and analysis. Furthermore such an approach should supply information, not only on what adoption took place but also on why it took place, thus defining more clearly the opportunities and conditions for the spread of new technology. Finally it should be emphasized that the production of credible figures for the impact of research is very important from the point of view of enhancing financial support for research, particularly from the State governments. Eventually such data should also provide the basis for calculating project benefits. 34. Economic Benefit. Although the SAR stated that no attempt would be made to quantify the benefits expected from the project, it was estimated that an average annual increase in net returns of Rs.3.6/ha for agricultural land serviced by a ZRC (about 1.4 million ha) would result in an economic rate of return (ERR) of 20 percent for an average research subproject for the time period 1985/2010, based on total investment and operating costs. In fact, average net retums per ha, at constant prices, have increased much more than these estimates. However, it is not possible to determine the real impact of project research activities on farm beneficiaries because of concurrent improvements in other factors of production, such as enlarged irrigated area and increased availability of fertilizers, and the direct influences from other agencies involved in promoting agricultural production (government extension, NGOs and private agricultural sector companies). C. IMPLEMENTATION RECORD AND MAJOR FACTORS AFFECTING THE PROJECT Factors not Generally Subject to Government Control 35. There were delays in the implementation of the civil works program due to difficulties in purchasing land and escalation in prices of building materials resulting in applications by contractors for revisions in contract payments and construction delays. The cyclone in Andhra Pradesh also necessitated additional funding for reconstruction of badly damaged research stations in that State.  13 Factors Generally Subject to Government Control 36. At the GOI level, delays in permission for new vehicle purchases and fund release by the Ministry of Finance adversely affected project implementation. At the state government level there were a number of factors affecting project implementation, including: (a) lack of funding for operational (contingency) purposes; (b) very slow process for selection and acquisition of land; (c) State Seed Corporations unable to multiply and deliver new varieties in sufficient quantities; (d) poor interaction between the line agencies, other than agriculture, with the zonal research teams; and (e) non-transfer of agricultural research responsibilities from DOA to SAU in one state (West Bengal - resulting in one year suspension of project support) and a number of state development departments have still retained some research capacity (e.g. in Assam, Rajasthan, Uttar Pradesh and West Bengal), although there was agreement by the states to devolve (under the project) research responsibilities to the SAUs. Factors Generally Subject to Implementing Agency Control 37. Factors within ICAR control which adversely affected project implementation and outcome included: (a) delayed release of funds to FAO for the training programs; (b) delayed release of funds to ISNAR for preparation of ARIS concept/plan; (c) delay in release of funds to SAUs; (d) poor quality construction in some buildings and glasshouses at the ZRSs due to inadequate supervision by the architects and engineers especially recruited to the Project Unit; (e) initial delays in staffing the Project Unit (only adequately staffed from 1991) which adversely affected preparation of new subprojects and M&E of approved subprojects; and (f) inadequate use of the 210 man-months of technical assistance provided to assist the Project Unit in subproject appraisal, implementation and evaluation of subprojects and Status Report preparation because the daily remuneration of Rs.200 was regarded by consultants as inadequate. 38. Factors within the control of SAUs which affected project implementation and effectiveness and research output included: (a) slow or non-release of funds to the ZRSs; (b) poor quality management in several SAUs; (c) slow progress and inadequate supervision of contractors on civil works, including glasshouse design and construction; (d) a strong tendency for SAUs to employ their own graduates rather than externally qualified graduates that could introduce new ideas and methodologies; and (e) lack of adequately qualified staff in certain areas of research, such as socio-economics, irrigation agronomy, farming systems and integrated pest management. D. PROJECT SUSTAINABILITY 39. The SAUs and State Governments are committed to maintaining the research system developed, and generally effective institutional/managerial capacity has been established. The sustainability of the decentralized research system established under the SAUs is generally  14 considered likely, although for certain states it is uncertain. However, most SAUs will require continued financial support from ICAR and increased support from State Governments. The main sustainability issues are the need for adequate future budgets and a research environment which provides adequate research facilities, incentives and support for scientists to remain at ZRSs and sub-stations. In some states, it is apparent that too many research stations have been established (for example, Himachal Pradesh is a small state with some 25 research stations), leading to ineffective research programs because of limited core staff size at many research stations and unnecessary duplication of research between nearby research stations. Furthermore, it is expected that not all of the 120 ZRSs will continue at full strength after project completion. In addition, some apprehension exists in the SAUs that there will be insufficient budget available to ensure adequate operations of research and maintenance of facilities (in 1994 it was estimated that up to 35 percent of completed SPs have some or serious constraints in operating funds), but contingency plans to establish alternative sources of funding (revolving funds, income generating production activities, consultancy services, etc.) are being considered by some SAUs to partially meet budget deficits. ICAR is also considering possibilities for providing future support to the SAUs (competitive grants, revolving funds and training funds). E. BANK PERFORMANCE 40. The need for a second phase project was essential to complete the decentralization of the research system in India, but the Bank failed to provide sufficient guidance to the ICAR in its preparation and were deficient in their initial appraisal of the project. This contributed to the need for a subsequent substantial revision of the project during internal Bank processing to ensure that the project scope, concept and additional components were appropriate. While this review and revision resulted in a significantly better project concept and design, the process was unnecessarily protracted (one year from first Project Brief to green cover SAR) because of deficiencies (lack of coordination and time-tabling of review) in project processing. By contrast, Bank supervision was highly satisfactory. The Bank undertook some 200 supervision missions, involving some 785 staff weeks during implementation of the project. This represents an extremely high level of supervision but is a result of the extremely complex nature of the project and the availability of the Bank's resident mission staff. The Bank maintained a close collaborative relationship with the ICAR and SAUs and made a significant contribution to project achievements. Bank supervision reports indicated that the missions undertook regular, detailed reviews of implementation progress of the project and, while indicating deficiencies in implementation, gave helpful and constructive advice to overcome such problems. The Bank also provided considerable help in the completion of the preparation of Status Reports for each agro-ecological zone. In mid-I 990, the Bank concentrated its efforts on helping underperforming states (Assam, Bihar, West Bengal, Madhya Pradesh and Uttar Pradesh). With very few exceptions, supervision missions were staffed from the Bank's Resident Mission in India and local consultants, which provided more continuous contact and  15 assistance in implementation. There has been universal appreciation by SAUs and ICAR of the assistance given by the Bank in the implementation of this project. F. BORROWER PERFORMANCE 41. The SAUs and ICAR gave continued high priority and attention to implementation of the project as proposed. Much has been achieved in fulfilling the original project concept of establishing and developing functional research facilities in local agro-ecological zones, and improving research and technology development for local farming systems. Overall, ICAR, through the Project Unit, was strongly supportive in project implementation and supervision. However, there were serious early difficulties (para. 37) and inadequate staffing and poor functioning of the Project Unit, which adversely affected the early stages of project implementation. There was considerable variation between states and SAUs in their management of and commitment to the project. There were several factors (para. 38), both within and outside the control of the SAUs, which adversely affected their performance. However, to their credit, most SAUs were able to satisfactorily resolve these problems with time, but the early delays and problems encountered ultimately contributed to the necessity for three annual extensions of the project to fully achieve the original project objectives. G. ASSESSMENT OF OUTCOME 42. Overall, the project mostly achieved its objectives by the end of the three-year extension period, and its outcome is assessed as satisfactory. The majority of the special subprojects designed to improve or establish ZRSs throughout India were completed successfully by the end of the project, with negligible or only moderate problems being encountered in the achievement of the physical targets set (paras 7-12). This resulted in significant improvements in the research facilities and capacity of ZRSs to undertake research (para 22) and generally improved research quality and management. Thus, the objective of establishing a functional decentralized system of research associated with SAUs that is responsive to research needs in local areas was substantially achieved. The establishment of functional and effective linkages between research and extension has been less successful, especially with line departments of the state governments, other than Agriculture. Despite this weakness there has been a significant uptake by farmers of new technologies developed by SAUs, and there have been significant gains in agricultural productivity. H. FUTURE OPERATIONS 43. The Borrower's Assessment of the project (ICAR, June 1996, Appendix C), while generally indicating future commitment and needs, does not provide a specific time-bound plan for future development. However, from numerous discussions, it was clear that the SAUs remain committed to maintain the decentralized research system established and strengthened under NARP  16 and NARP II. There is universal agreement that physical facilities are now generally adequate to meet research objectives. Future emphasis would be given by the SAUs to selective upgrading of knowledge and skills of scientists and ensuring an attractive work environment, particularly at the more remote research stations. Some SAUs have a policy of complete residential accommodation for all staff at research stations. Emphasis will also be given to developing communications, computerization, media materials and information dissemination capabilities, as well as increasing support for research on animal nutrition, fisheries, horticulture, silviculture, and farm mechanization. All SAUs have expressed a desire to strengthen the weak research-extension linkages that have developed following the termination of the National Agricultural Extension Projects and consequent widespread abandonment of the T&V system. Increased numbers of "front-line" on-farm trials and strengthening of Krishi Vigyan Kendras (KVKs) for dissemination of technologies to farmers are also envisaged to improve technology transfer. 44. In a more holistic sense, however, what is equally imperative is a need to concentrate on strengthening the technology generation and transfer process, with particular emphasis on increasing the effectiveness of technical and technological information exchange between research, extension and farmers to ensure research is demand-drive, that technologies developed are appropriate to local needs and resources circumstances, and that technical information is readily available to all. It is essential in a follow-up project, such as the National Agricultural Technology Project presently being prepared for funding by the World Bank, that it would have the objective of responding to needs for improvement in defined cropping systems or thematic topics (soil and water conservation and utilization), integrated nutrient use efficiency, natural resource management, etc.), whereby all elements of the research-extension-farmer system contribute to the one overall goal. Such a program approach may involve basic and applied research by ICAR institutes, applied research and on-farrn trials by SAUs, district extension programs by DOAs (including demonstrations, training, technical information, etc.), and training programs for farmers at KVKs and Farmer Training Centers. I. LESSONS LEARNED 45. There are a number of lessons to be drawn from the implementation and outcome of this project, including: (a) The concept of a decentralized research system, started under NARP and carried on under NARP 1I, based on a uniform system where research has become the responsibility of ZRSs of SAUs conducting adaptive research in local agro-ecological zones, is sound and its continued development is essential if the diverse farming systems in India, particularly the rainfed systems, are to benefit from new technologies.  17 (b) Generally, most ZRSs, but not sub-stations, have adequate facilities and sufficient numbers of scientific staff. However, the nature of the farming systems and their great diversity makes it necessary to have more multi- disciplinary teams working in an inter-disciplinary context with more emphasis placed on financial viability of technologies, marketing potential, diversification (broad-basing) of farm production and relevance of technologies to farmers circumstances. The SAUs need to give higher priority to horticulture, livestock, fisheries and agro-forestry research. (c) The project design had an important element of flexibility in that individual subprojects were prepared and approved jointly by SAUs and ICAR, and while indicative average costings were provided for each subproject, actual costs of subprojects were need-based. In practice this meant that GOI provided 100 percent financing to the SAUs for approved subprojects and funds were advanced six months ahead of expenditure. This allowed greater support to be given to resource poor states and remote or backward areas, leading to a more comprehensive decentralized research system. Successful implementation of subprojects, however, required more explicit timetables for expenditure and physical targets to be met and closer monitoring and supervision by the implementing agency. (d) The establishment of a productive and creative research system is a long term process, in which the present system is at a relatively early stage of development. The further advancement of ZRSs will depend on adequate financial support to operate and maintain facilities and research programs, improved management of research resources (better priority allocation, planning, sharing of common resources, monitoring and evaluation) and continuing improvement in the research capacity of the scientific staff (often isolated from the mainstream of scientific activity), requiring specific attention to training, reward systems and strong technical support from the SAUs and ICAR. (e) Financial sustainability of the research system developed is presently jeopardized in some states by the indicated inability of some State Governments to fully fulfill the terms of their MOUs to financially support the system after the closure of the project. Lack of funds will particularly adversely affect research operations, maintenance of facilities and equipment and capital replacement. Although some SAUs have considered possible revenue earning opportunities to improve financial viability, these will always be limited. It is essential for SAUs to use their financial, physical  18 and staffing resources more efficiently, limit the number of sub-stations, and improve cooperation and collaboration in research programs among states with similar agro-ecological zones. (f) Despite the emphasis given to establishing and strengthening research- extension-farmer linkages in the many research and extension projects funded by the World Bank in India, these linkages still remain weak. The establishment of formal linkages (zonal workshops, training for SMSs of extension departments, joint field days, etc.) between research and the line agencies of state departments (agriculture, forestry, livestock and fisheries), has not, except for crop production, engendered strong cooperation, information exchange and collaboration in technology generation and transfer to farmers. It is essential that the extension departments of the SAUs and state governments and the research agencies (SAUs and ICAR) develop novel and more effective approaches to improving collaboration. (g) The project's implementation benefited from strong expertise in agricultural research of the Bank's Resident Mission staff who supervised the project.  19 PART II. STATISTICAL ANNEXES Table 1: Summary of Assessments A. Achievement of obiectives Substantial Partial Negligible Not Applicable Macro policies L 1 II[ Sector policies K] E71 E] K Financial objectives I K] EJ [I Institutional development F1 EII [K F] Physical objectives LI E] LI] K] Poverty reduction ] EI] K] Gender issues n] o7 E7 L