Tanzania - National Agricultural and Livestock Extension Rehabilitation Project

Document of The World Bank FOR OFFICIAL USE ONLY Report No.: 17910 IMPLEMENTATION COMPLETION REPORT TANZANIA NATIONAL AGRICULTURAL AND LIVESTOCK EXTENSION REHABILITATION PROJECT (CR. 1994-TZ) May 26, 1998 Agriculture Operations Eastern and Southern Africa This document has a restricted distribution and may be used by recipients only in the perfornance of their official duties. Its contents may not otherwise be disclosed without World Bank authorization.  CURRENCY EQUIVALENTS Jun-96 623.3 Jun-95 604.8 Jun-94 516.2 Jun-93 395.0 Jun-92 300.0 Jun-91 227.0 Jun-90 193.2 Jun-89 145.0 FISCAL YEAR OF BORROWER Government of Tanzania July 1 - June 30 ABBREVIATIONS AND ACRONYMS ADF African Development Fund DIVEO Divisional Extension Officer IFAD International Fund for Agricultural Development FSR Farming Systems Research GOT Government of Tanzania MAC Ministry of Agriculture and Cooperatives NALERP National Agricultural and Livestock Extension Rehabilitation Project RALDO Regional Agriculture and Livestock Development Officer RELO Research Extension Liaison Officer SMS Subject Matter Specialist VEO Village Extension Officer Vice President Callisto Madavo Country Director James W. Adams Sector Manager Sushma Ganguly Team Leader Satish Kumar  IMPLEMENTATION COMPLETION REPORT FOR OFFICIAL USE ONLY TANZANIA NATIONAL AGRICULTURAL AND LIVESTOCK EXTENSION REHABILITATION PROJECT (NALERP) (Cr. 1994-TZ) TABLE OF CONTENTS PREFACE .......... EVALUATION SUMMARY .ii PART I: PROJECT IMPLEMENTATION ASSESSMENT .1 A. STATEMENT/EVALUATION OF OBJECTIVES .1 B. ACHIEVEMENT OF OBJECTIVES .2 General .2 Institutional Change .....................3....,......3 Management and Supervision ...3 Training ...4 Civil Works ...4 Technical Assistance ...4 Monitoring and Evaluation ...5 Research-Extension Collaboration .5 Women and Youth and Environment .6 Impact .6 Project Costs and Financing ................................................................... , . 7 C. MAJOR FACTORS AFFECTING THE PROJECT .8 Limits on Project Counterpart Funds .8 Poor Integration of Agricultural and Livestock Skills at Field LeveL .8 New Challenges, beyond Technical Matters, for which the Extension System was not Fully Prepared .8 D. SUSTAINABILITY .9 E. BANK PERFORMANCE .9 F. BORROWER PERFORMANCE .10 G. ASSESSMENT OF OUTCOME .10 H. FUTURE OPERATIONS .11 I. KEY LESSONS LEARNED .12 PART II: STATISTICAL TABLES ........................................................ 14 Table 1: Summary of Assessments ....................................................... 14 Table 2: Related Bank Loans/Credits .16 Table 3: Project Timetable .16 Table 4: IDA/Credit Disbursements: Cumulative Estimated and Actual .17 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 5: Key Indicators for Project Implementation ............................................. 17 Table 6: Key Indicators for Project Operation ............................................. 18 Table 7: Studies Included in Project ............................................. 18 Table 8A: Project Costs ............................................. 19 Table 8B: Project Financing ............................................. 19 Table 9: Economic Costs and Benefits ............................................. 20 Table 10: Status of Legal Covenants ............................................. 21 Table 11: Compliance with Operational Manual Statements ............................................. 22 Table 12: Bank Resources: Staff Inputs ............................................. 22 Table 13: Bank Resources: Missions ............................................. 23 Appendices 1. Mission's Aide Memoire 2. Government's Contribution to the ICR 3. Miscellaneous Tables 4. Yields and Production of Major Smallholder Crops  IMPLEMENTATION COMPLETION REPORT TANZANIA NATIONAL AGRICULTURAL AND LIVESTOCK EXTENSION REHABILITATION PROJECT (NALERP) (Cr. 1994-TZ) PREFACE This is the Implementation Completion Report (ICR) for the National Agricultural and Livestock Extension Rehabilitation Project (NALERP) in Tanzania, for which IDA Credit 1994-TZ in the amount of SDR 13.7 million (US$18.4 million equivalent) was approved on March 21, 1989, and became effective on July 6, 1989. The project was co-financed in parallel by the African Development Fund. The Mid-Term Review took place in June 1993 and the Credit was closed on March 31, 1997, as foreseen at appraisal. An amount of SDR 73,700 (0.5% of the credit) remained undisbursed, and is scheduled to be cancelled. The ADF loan has not closed, and the remaining civil works under the project will be completed by March, 1998. At closing of the ADF loan, it is expected that US$8.45 million of a total of US$8.8 million will be disbursed. This ICR was prepared by a mission from the FAO/World Bank Cooperative Programme, on behalf of the Africa Agriculture Operations Group 1 (AFTA1) of the World Bank. Preparation of the ICR commenced with the mission's visit to Tanzania from October 3 to 17, 1997. This report is based on material gathered from the project files and supervision reports, as well as discussions with Bank and Government of Tanzania (GOT) staff associated with the project. These included staff from the Ministry of Agriculture and Co-operatives (MAC), particularly the Department of Extension Services, and the Department of Research and Training within MAC. The mission visited two regions of Tanzania to assess the impact of the project in areas of different agricultural potential and farming systems. The Mission held discussions with Regional and District Agriculture and Livestock Development Officers, Regional Extension Officers, and visited farmers, farmer groups, District, Divisional and Village Extension Officers. The mission also visited two research institutions. The Government contributed to the preparation of the ICR through discussions on the mission's Aide-Memoire, and prepared the Government's "Implementation Completion Report".  - ii - IMPLEMENTATION COMPLETION REPORT TANZANIA NATIONAL AGRICULTURAL AND LIVESTOCK EXTENSION REHABILITATION PROJECT (NALERP) (Cr. 1994-TZ) EVALUATION SUMMARY Introduction 1. The NALERP was conceived in early 1987 at a time when the extension services in Tanzania had been fragmented between various parastatals and the Ministry, and the foodcrop research did not receive appropriate priority. Key aspects of extension - supervision, training, logistics, linkages with other knowledge and input systems - had suffered from lack of resources and an inefficient institutional structure. NALERP was also conceived at a time when the Bank was actively expanding its support to the training and visit (T&V) extension management system throughout the Africa Region. The project formed part of the Bank's development strategy for Tanzania which was to place emphasis on producer incentives to encourage production of basic staples for domestic consumption and cash crops for export. As part of this strategy, institutional improvements were being developed to improve service delivery with an increased attention to smallholders. The Bank's involvement was designed to support the Government of Tanzania's (GOT) objectives of food self-sufficiency, improving nutritional standards, restoring agricultural exports to former levels, and improving rural living standards. At the beginning of the preparation of the project, a pilot T&V project in five districts covering 50,000 farm families was initiated. The project was signed in May 1988, became effective in July 1989 and closed in March 1997. Project Objectives 2. The objective of the project was to enable the Government of Tanzania to provide the farming community with adequate extension messages to enhance agricultural production, productivity and rural incomes. The project represented the first phase, regarded as the rehabilitation phase, of a long- term process to establish an efficient, cost-effective and sustainable extension service. The project was to focus on a gradual but steady and well-planned merging of crop and livestock extension services into a multidisciplinary system, the strengthening of extension management, organization, coordination and supervision of activities, and the upgrading and/or provision of necessary human resources, physical infrastructure and logistic support. This was to be achieved through: (i) the establishment of a streamlined, adequately staffed organization and management structure; (ii) the introduction of an extension system incorporating the essential principles of the T&V management system, together with necessary training and supervision of activities; and (iii) strengthening of the Ministry's planning and implementation capacity, and assisting in the preparation of a second phase project, through both short and medium-term technical assistance. The cost of the project was estimated at US$30.4 million, financed by an IDA Credit (US$18.4 million), a loan from the African Development Fund (US$8.8  - iii - million) and the Government of Tanzania (US$3.2 million). The African Development Fund was responsible for the financing of the construction or upgrading of housing for Division staff, the rehabilitation and provision of facilities in four training institutes and the overseas training program. GOT was responsible for financing a share of the incremental recurrent costs of the project. 3. The strategy to achieve the objectives correctly envisaged a long-term program of which the project was the first rehabilitation phase. However, the objectives in the SAR were descriptive and no specific indicators for level of achievement that NALERP was to have reached, at project completion, towards longer-term program goals, were set out. Implementation Experience and Results 4. In general, the implementation of the project has been satisfactory. After a delay at the outset of the project which effectively put the start of implementation back by one year, implementation then proceeded as scheduled. The project was originally designed to cover 12 of the mainland Regions of Tanzania, but was expanded at the time of the mid-term review. At the time of termination, the project operated in 16 of the 20 mainland Regions, with the four Southern Highlands regions served by the Southern Highlands Extension and Rural Financial Services project. This project, funded by IFAD and GOT, also follows the T&V approach with the addition of a credit facility for farmers. Effectively therefore, the essential principles of the T&V management system were implemented nationally, but extension projects supported by other donors generally did not fully adopt the T&V approach. Achievement of Objectives 5. The project achieved the objective of revitalizing the extension service, building capacity to manage and make decisions on future directions, and giving the service a new work ethic. The overall numbers of the frontline extensionists declined due to retirements, termination of many under the civil services reform program and ban on recruitments imposed by the Government. The number of farm families in charge of a VEO over time typically increased, and reduced effective contact of the front-line extensionist with the farmers. The situation was later mitigated by the introduction of group approach under the revised extension methodology. 6. The ultimate objective of the project was enhanced agricultural production, productivity and rural incomes. There is clear evidence from the monitoring of results from adoption plots/units that increases in productivity have been achieved for adopters of the extension messages. This would translate into overall increases in production and incomes depending on the climate and season, scale of operations, continuation of improved practices and the relationships between revenues and input costs. Improved extension would also help to improve yields in poor seasons that would otherwise fall to a greater degree. The average yields in 1995-1997 of some smallholder crops like sorghum, maize and millet were higher than estimates made during appraisal. Cashew also performed well but for some other crops like rice and beans, the yields were lower than the appraisal estimates. Overall total production trends over the period of the project were mixed. Knowledge is but one input and variable in agricultural production, and whilst extension plays a clear role in awareness, it is not possible to estimate to what extent it alone influences adoption. The absence of baseline studies makes it even more difficult to gauge  - lv - the overall impact. Therefore, no robust conclusion can be drawn on the effect of the project alone on the national production of food and cash crops. Major Factors Affecting the Project 7. VEOs generally remained in the same villages to which they were originally posted and dealt with the same farmer clientele. The messages delivered by the Extension service were predominantly technical, simple and showed a response during the growing and harvesting seasons in which they were applied. Given an ineffective Extension service prior to project initiation, these messages were relevant early on in the project and generally were considered useful, at least by those in close contact with extension staff. It is not fully determinable whether farmers found at least some of the technical messages strongly relevant, given their repetitiveness and growing farmer awareness. The changing farmer need should drive the generation of technology and its dissemination. Both the research systems and extension in agriculture need to be more demand-driven than was the case during NALERP period. 8. Whilst the training gave technical competence in both crops and livestock matters, some continued to emphasize their original background and were weak on the other aspects. Also, many VEOs lacked skills and knowledge on important topics such as farm budgeting, plant protection, farm economics, and land use. As the focus of groups for contact points was introduced, VEOs were not able to fully use the groups due to their lack of skills and knowledge in this area. These are the areas in which skills of the front-line extensionists need to be enhanced. 9. The effectiveness of the extension service declined in the second half of the project, due mainly to macro-economic conditions, from decreasing funds for regular activities, decreasing numbers of front- line staff, and an inability of the extension service to diagnose problems and change the messages in response to farmers needs and circumstances. Some of the vehicles, motorcycles and bicycles became unserviceable!!, and there had been a restriction on field activities and supervision by GOT since 1994 due to lack of operational funds. Although research and extension are priority areas for Government budgetary support in Tanzania, the extension activities did not attract sufficient funding. As a result, the numbers of farmers directly contacted by public extension service fell from 1994 onwards. Sustainability 10. The SAR considered sustainability from two viewpoints-institutional and financial. Institutional sustainability, which depends on reaching anticipated management strength following restructuring, commitment of all concerned and strong leadership in the MAC, was secured. The financial sustainability would be a question mark if it is viewed from the limited point of view of the MAC being able to finance the costs of extension with a continuously rising Government share and a declining IDA share, as the originally envisaged funding arrangement implied. This assumption did not materialize as the Government faced acute financial stringency which restricted funding to its development portfolio, including NALERP. However, viewed in the wider context of NALERP I' The project procured 169 4-WD vehicles, 892 motorcycles and 9,483 bicycles.  v - being phase one of a long term national extension program, whose activities are now being sustained under the follow-on phase 2 NAEP II, sustainability in the longer term can be considered likely. Project Costs 11. Total costs of the project up to March 31, 1997 were US$27.7 million and, including committed costs of civil works to be financed by ADF of US$1.3 million, total costs will be US$28.6 million as against an appraisal estimate of US$30.4 million. Savings in costs, particularly for vehicles, enabled the project to expand into four additional regions within the financing available. Total recurrent costs were estimated in the SAR at US$11.2 million. Actual recurrent costs of the project were US$7 million. Investment costs were higher by $2.3 million than that anticipated at appraisal. The total Government contribution to financing the project amounted to US$855,000, or approximately 3% of total project costs. Bank Performance 12. Bank supervision missions were frequent, with staff from the Resident Mission and the Nairobi Mission often participating. Generally supervision was satisfactory, but there should have been closer supervision of the M&E activities under the project. Africa Region staff assisted the Government in the mid-tern review of the project and worked with the Government on the preparation of NAEP II and the development of pilot initiatives that will be supported under that project. Borrower Performance 13. Generally, the performance of the Borrower has been satisfactory. The main covenants relating to timely implementation of the project were complied with although there was a delay at the outset. Legal covenants have been complied with, with the exception of those covering the agreements on staffing levels and the increasing budgetary allocations to cover annual costs as determined in the annual work plan. Accounting and reporting have been carried to a satisfactory standard. The Government has also prepared its contribution to the ICR. Project Outcome 14. The project achieved the objective of revitalizing the extension service, building capacity to manage and make decisions on future directions, and giving the service a new work ethic. Structured work schedules of frontline extensionists brought orderliness in contacting farmers, and through extension intermediation farmer-extension-research linkages improved. The extension methodology underwent a slow but measured change to respond to some emerging challenges like the introduction of group approach for a participatory consultation and maintaining outreach; and extension participating in FAO-supported food security initiative and Sasakawa Global-scheme on input supplies. Beneficiary assessment study results show that a high percentage of farmers (77%) reported beneficial contact with extension. Substantial increases in yield on adoptors farms were recorded. Project surveys show substantial yield increases in several crops, particularly on adopters  - vi - farms. However, the extension management of MOA remained monolithic with insignificant structured role for local Governments and other stakeholders in policy fornulation or program operation; the extension advice was generally top-down than demand-driven; and due to acute financial stringency affecting its entire development portfolio the Government could not provide adequate counterpart funding to this project as well, affecting field implementation. Therefore, while the development objectives were achieved, the outcome of the project is considered to be marginally satisfactory. Summary of Findings, Future Operations and Key Lessons Learned 15. Findings: The experience of T&V extension in Tanzania reflects that of similar projects in the Africa Region as reported by OED2/ * Whilst the project provided the necessary impetus to rehabilitate and build capacity within the extension service, to improve efficiency through the introduction of a unified extension system and to provide the much-needed resources, the project was negatively affected by a lack of adequate counterpart funds from Government for field activities including linkage promoting activities between farmers and the research/extension knowledge systems. This ultimately led to somewhat reduced impact. 16. Future Operations: As foreseen during the design of NALERP, a second-phase extension project, National Agricultural Extension Project Phase II (NAEP II) was prepared during 1995 and became effective in October, 1996. NAEP II incorporates lessons leaned from NALERP and includes measures to improve the sustainability and cost-effectiveness of the extension service. This project is financing the continuation of T&V activities in Tanzania with an emphasis on group contacts and farmer participation in farming systems-based extension messages. NAEP II will build on the experiences of NALERP and will pursue other extension approaches to ensure financial sustainability (e.g., cost-sharing and private sector and NGOs participation in crop-specific farming systems), participation in the knowledge generation process by the farmning community, and more cost-effective communication technology. 17. Key Lessons Learned: The key lessons learned from the project were as follows: In a long-term program, specific objectives, with indicators, for each of the phases in the program should be clearly identified at the start of each phase. In any case, indicators should be set out in order to objectively monitor project achievement. Simple technical messages are often initially relevant to farmers on first contact with an extension service but, after they choose either to adopt or not to adopt, messages need to address newly emerging needs of the various segments of farmers within the overall farming community. The current systems need to be strengthened to generate messages to satisfy farmer needs. The extension system needs to be more demand-driven. 21 "Agricultural Extension and Research - Achievements and Problems in National Systems", OED, June 1997.  - vii - Whilst project managers need to focus on the operational activities of an extension project - training and visiting regularly with impact messages and contact points - they should not lose sight of the overall objective of extension - outreach to most farm households in response to farmers' priorities. Excessive monitoring of project activities within the direct control of project management. (e.g., numbers and frequency of contacts, bi-monthly workshops, supervision visits) can detract attention from the more important evaluation of progress against overall development objectives. The project should have carried out a baseline survey at the beginning and later a survey at the conclusion to determine impacts (for example, increases in yields, farmer incomes and adoption rates). The involvement of farmers in designing and operating extension programs is a critical element in ensuring the continued technical relevance and usefulness of the service, which is essential for maintaining farmer demand and its technical sustainability. * The change to group-based extension was positive as ownership of the program was then transferred to the community. Extension staff should have skills and knowledge in managing groups, and participatory approaches. - In a country with large distances and poor road conditions, the cost of operating, maintaining and replacing a large fleet of vehicles is high and there is a risk that it will not receive priority in the Government's allocation of scarce funds. * In T&V extension systems, VEOs should be redeployed to new villages every two years in order to contact a new clientele who are unaware of the available simple technical messages. The Government does not consider this suggestion feasible, in view of costs involved in transferring VEOs every two years. If a Government finds difficulty in providing the required resources to support a T&V extension system, the operations should be scaled down to match the resources that the Government can reasonably be expected to allocate to extension in the longer term. * One of the main purposes of the ICR exercise is to analyse and document the lessons learned and assess the future operations of projects. An ICR exercise, completed after a follow-on project is already effective, can provide little input to the preparation and design of following projects, and an opportunity is thereby lost to fully incorporate the lessons learnt in further Bank assistance. In the context of a long-term program which envisages follow-on Bank assistance, the procedures for ICRs need to be altered to require the ICR to be completed and available at appraisal of succeeding operations.  IMPLEMENTATION COMPLETION REPORT TANZANIA NATIONAL AGRICULTURAL AND LIVESTOCK EXTENSION REHABILITATION PROJECT (NALERP) (Cr. 1994-TZ) PART I: PROJECT IMPLEMENTATION ASSESSMENT A. STATEMENT/EVALUATION OF OBJECTIVES 18. The NALERP was conceived in early 1987 at a time when the extension services in Tanzania had been fragmented between various parastatals and the Ministry, and the foodcrop research did not receive appropriate priority. Key aspects of extension - supervision, training, logistics, linkages with other knowledge and input systems - had suffered from lack of resources and an inefficient institutional structure. NALERP was also conceived at a time when the Bank was actively expanding its support to the training and visit (T&V) extension management system throughout the Africa Region. The project formed part of the Bank's development strategy for Tanzania which was to place emphasis on producer incentives to encourage production of basic staples for domestic consumption and cash crops for export. As part of this strategy, institutional improvements were being developed to improve service delivery with an increased attention to smallholders. The Bank's involvement was designed to support the Government of Tanzania's (GOT) objectives of food self-sufficiency, improving nutritional standards, restoring agricultural exports to former levels, and improving rural living standards. At the beginning of the preparation of the project, a pilot T&V project in five districts covering 50,000 farm families was initiated. 19. The objective of the project was to enable the Government of Tanzania to provide the farming community with adequate extension messages to enhance agricultural production, productivity and rural incomes. This was to be achieved through: (i) the establishment of a streamlined, adequately staffed organization and management structure; (ii) the introduction of an extension system based on the Training and Visit (T & V) methodology, together with necessary training and supervision of activities; and (iii) strengthening of the Ministry's planning and implementation capacity, and assisting in the preparation of a second phase project, through both short and medium-term technical assistance. 20. The project represented the first phase, regarded as the rehabilitation phase, of a long-term process to establish an efficient, cost-effective and sustainable extension service. The project was to focus on a gradual but steady and well-planned merging of crop and livestock extension services into a multidisciplinary system, the strengthening of extension management, organization, coordination and  - 2 - supervision of activities, and the upgrading and/or provision of necessary human resources, physical infrastructure and logistic support. 21. The strategy to achieve the objectives correctly envisaged a long-term program of which the project was the first rehabilitation phase. However, the objectives in the SAR were descriptive and no specific indicators for level of achievement that NALERP was to have reached, at project completion, towards longer-term program goals, were set out. No time frame was given for the longer-term goals. B. ACHIEVEMIENT OF OBJECTIVES General 22. The project has satisfactorily achieved its development objective of rehabilitating the extension services provided to farmers and livestock keepers. The project re-invigorated the field, district and region extension staff who previously had no meaningful work program and only limited financial and logistic resources to carry out their work. At the time of termination, the project operated in 16 of the 20 mainland Regions of Tanzania having been included in the following order. 1989 Tabora, Singida, Dodoma 1990 Kilimanjaro, Kagera, Arusha 1991 Shinyanga, Mwanza, Mara 1992 Morogoro, Lindi, Mtwara 1993 Kigoma 1994 Tanga, Dar es Salaam, Coast 23. The four Southern Highlands regions of Ruvuma, Rukwa, Mbeya and Iringa are served by the Southern Highlands Extension and Rural Financial Services project. The project, funded by IFAD and GOT, also follows the T&V approach with the addition of a credit facility for farmers. The project contributed to coordination of other donor activities in extension, but there was limited adoption of the strict T&V approach in other projects. Other projects in extension implemented outside the MAC did not unify their approach to that of NALERP, nor did they coordinate their programmes. 24. At the height of the project 3,628 VEOs were responsible for 3,680 villages, 2,145,000 farmers (assuming 583 farmers per village) and 591 farmers per VEO against project targets of 4,113 VEOs, 4,800 villages and 2,800,000 farmers at a ratio of 681 farmers per VEO. 25. The effectiveness of the extension service declined in the second half of the project, due mainly to macro-economic conditions, from decreasing funds for regular activities, decreasing numbers of front-line staff, and an inability of the extension service to diagnose problems and change the messages in response to farmers needs and circumstances. Some of the vehicles, motorcycles and bicycles became unserviceable, and there had been a restriction on field activities and supervision by GOT since 1994 due to lack of operational funds. The project had a program of periodically replacing the unserviceable vehicles, and this was done during the project period . Although research and extension are priority areas for Government budgetary support in Tanzania, the extension activities did not attract sufficient funding.  -3 - As a result, the numbers of farmers contacted fell from 1994 onwards. At the conclusion of the project, the estimated number of farmers contacted by the extension service in 1996/97 was 1.5 million. 26. The Ministry modified the Extension methodology in response to operational experience and funding constraints. Within the first 12 months of project implementation, VEO training sessions were reduced in frequency from fortnightly to monthly and SMS training from monthly to bi-monthly. As a result, more time was available for extension, costs were reduced, and there was no loss of staff competence in the messages they were disseminating to farmers. 27. The major change has been away from contact farmers established for points of regular contact with VEOs, to VEOs working with groups, in many cases already established for other purposes. The pilot T&V project established to test the effectiveness of this methodology were based on the Ujamaa concept and were already established as part of an administrative set-up, or for collective agricultural production or political purpose. NALERP then attempted to set up contact farmer groups through the direct actions of VEOs. These new groups were sometimes formed on an incorrect perception of purpose - the members believed that they would obtain credit or inputs in addition to extension. This was observed in 1991 and questions of sustainability of the contact farmer groups newly set up for extension was questioned at that time. By the time of mid term review, the emphasis was on working with farmer groups that either existed already within the communities, or were established in harmony with community practices. At NALERP completion, the use of groups by the extension service was still evolving. One of the factors responsible for lack of better performance in this area was that the staff had little or no skills or knowledge on group formation and dynamics. Institutional Change 28. The Ministry amalgamated the Departments of Agriculture and Livestock under a Commissioner who reports to the Principal Secretary. Extension is under an Assistant Commissioner (AC) reporting to the Commissioner, who is supported technically by Assistant Commissioners for Plant Protection, Livestock Development, Crop Development and Irrigation. At the regional level, unification was achieved through a Regional Agricultural and Livestock Development Officer (RALDO) supported by specialists and a Regional Extension Officer (REO), a structure which is repeated at the district level. At the divisional (i.e., sub-district) level an Extension Officer, usually holding a Diploma, supervises 4-6 Village Extension Officers (VEOs). 29. The institutional changes associated with unification, in terms of structure, management and administration have been successfully implemented and are perceived as being a positive development at all levels in the organization. At the front-line, although retrained, a significant number of VEOs, particularly those who formerly were Livestock Extension Officers, emphasized their previous area of specialization. This is not surprising given a training period of only five months and a natural tendency to feel more confident in the discipline in which one was first trained and has the longest experience. Management and Supervision 30. Implementing the T&V system established a strong management system and management ethic within the Extension service. Regular visit schedules to contact farmers and groups are maintained and monitored through reporting procedures and field visits. Supervision at all levels however was adversely affected by lack of operational funds, and supervision by headquarters, regional, district and divisional staff were 34%, 50%, 42% and 54% respectively of target levels.  - 4 - 31. The Mid-Term Review, seen by management as an opportunity to take stock of progress and problems, was preceded by discussions and interviews with extension staff, trainers, farners, researchers, NGOs and others in 8 of the 13 regions of the project. The results of the survey guided discussion at the mid term review and enabled informed decisions to be taken on a wide range of issues including visit schedules, supervision, adoption plots, message content and the use of groups as an alternative to individuals as contact points for extension messages. These examples demonstrate an increased managerial capacity of MAC to plan, implement and supervise a national extension programme, and this increased capacity is attributable to the project. Training 32. Generalist front-line officers have been created by re-training through courses of five months duration as opposed to ten months as anticipated at appraisal. At project initiation, approximately 40% of front-line staff were untrained Field Assistants. No training was provided at this level and most of the Field Assistants were retrenched. 33. The VEO retraining program was slow up to the time of the mid-term review at which time only 18% of the target had been reached. The situation improved after mid term review and certificate- level retraining of 1,446 extension officers, 69% of the target, or 50% of the total extension field force, was completed. Whilst the training gave some technical competence in both crops and livestock matters, VEOs continued to emphasize their original background and were weak on the other aspects. Also, VEO lacked skills and knowledge on important topics such as farm budgeting, plant protection, farm economics, and land use. The VEO did improve the delivery of information on markets and prices. The training of trainers program was 54% completed. Other higher-level, overseas training and overseas study tours were carried out satisfactorily. Civil Works 34. The civil works component for housing, financed by ADB, was not satisfactorily implemented. The initial assumption that village labour would be volunteered to assist in the construction of 300 Divisional Extension Officers (DIVEO) houses did not eventuate. A reduced building program was then planned (141 houses) with higher unit costs, but this then fell behind due to delays and failure of contractors to perform as required. At the time of closing of the IDA credit, a total of 67 houses had been completed but contracts for the remaining houses is in force and construction is underway. The DIVEO housing program has been incorporated into NAEP II. The rehabilitation of the Training Institutes was satisfactorily completed. Technical Assistance 35. The long-term technical assistance team to support project implementation started work during 1990, and finished in 1992 with the exception of the Financial Advisor who remained throughout the duration of the project. The experiences of the technical assistance financed by the project were mixed. In general, the long-term consultants in Agricultural Extension Methodology, Finance and Procurement proved to be satisfactory, but the Monitoring and Evaluation Specialist's terms of reference were not completed at the end the assignment. This had important implications on the performance of the Monitoring and Evaluation component of the project. The short-term consultancies proved to be effective with the exception of the consultancy on cotton research which did not result in developing appropriate recommendations for farmers.  -5 - Monitoring and Evaluation 36. Implementation of this component was negatively affected by the consultant not completing the task of developing and implementing systems for M & E of project activities and impact. The SAR specified that baseline surveys would be undertaken in the first and second years of the project and impact evaluations of T&V in the regions at the commencement of years two and three. These surveys were not undertaken. The Overall Evaluation System specified in the SAR was not installed. Although introduced only in September 1994, quantitative goals are now established prior to the commencement of the growing season. The M&E staff in the Ministry did make progress in regular reporting of activities from field level through a system of data gathering and analysis, and coordinated three yield/adoption surveys. Research-Extension Collaboration 37. Formal links between Research and Extension were built into the project design. These included the posts of Research-Extension Liaison Officers (RELOs), pre-season workshops, bi-monthly workshops for the training of SMSs and two annual workshops to be held at zonal research stations. The Farming Systems Research Programme was expected to provide additional opportunities for collaboration within the context of establishing and responding to priority needs of rural communities and RELOs were expected to be seconded to the Farming Systems Research (FSR) team to facilitate this work. 38. Limited improvements in working relationships were achieved during the project period. RELOs were not appointed as MAC management decided that the REO could perform this task. At best, the calendar of seasonal impact points selected by extension officers were sent to the zonal research station for their commnents. The monthly workshops, which were changed to bi-monthly early in project implementation, did provide a forum for effective interaction between extension and research. With NALERP funds, researchers generally attended as resource persons. However, research priorities were not reviewed at these workshops, but this role has now been taken over by Zonal Committee Meetings which set zonal research priorities. 39. At the mid term review, the FSR programme was found as being not fully successful in establishing closer relationships. Some improvements however occurred after 1993. The FSR programme has mapped sub-zones and increased its use of PRA approaches in identifying problems. Inevitably, the subsequent research plots are scattered and difficult and expensive to establish and monitor. Some research stations have entered into fonnal agreements with Extension for assistance to implement the research programme. Nevertheless, a check conducted by the ICR mission in one district in August 1997, showed that some extension messages have remained largely unchanged during the project period, although the economic analysis of fertilizer application was added after the mid term review. The messages concerned picking and grading of cotton, storage of paddy, transplanting, fertilizing, weeding and harvesting of vegetables, whereas farmer priority problems identified at the same time were credit, input supply, marketing, labour constraints and soil fertility. 40. Initiatives were taken by the Ministry to improve Research-Extension collaboration. The Commissioner of Agriculture and Livestock Development and the Commissioner of Research and Training in 1992 agreed to provide resources to improve linkages in defined areas. The 1992 policy statement and implementation guidelines addressed the issue, a Zonal Technical Committees were formed, Extension was regularly represented at the national and zonal levels of research committees, and donor  - 6 - coordinating committees formed in some regions to remove duplication in extension and research activities. These initiatives however appear to have been of limited benefit. For the future the Government could consider a more likely strategy for improving collaboration, namely, by steadily increasing formalization of roles, responsibilities, funding arrangements and lines of communication within the framework of FSR as this method would involve close collaboration with farming communities, by both research and extension systems. Women and Youth and Environment 41. NALERP data were not disaggregated with respect to gender. A limited study (300 farmers in six regions) of contact farmers/ members of contact groups conducted in 1995 gives an indication of the female participation in the project. The study revealed that 39% of contact farmers were women, 43% of the membership of contact groups were female, and adoption rates in both crops and livestock components were higher amongst women than men. These figures indicate that, at least at the interface between the extension system and farming communities, the project has had some success in reaching women farmers. The result is significant in the light of the fact that only 12% of VEOs and approximately 20% of SMSs are women. In contrast, of the 879 farmers who received specialist training under the project, only 185 were women. The mid term review observed that whilst women and youth groups are common, few VEOs used these groups as a contact groups for extension. In view of the fact that extension messages contained advice on soil conservation, optimal use of pesticides, farm forestry, improved fallows, afforestation activities (agro-forestry, fuelwood production), composting, rationalized use of fertilizers, and water harvesting the environment impact of the project should be regarded as beneficial. Impact 42. The first attempt to evaluate impact was carried out in 1995 as a part of the preparation for the proposed second phase of support to Tanzania's Agricultural Extension System ("An Evaluation of NALERP Phase I by Beneficiaries", Sokoine University, 1996). In this survey, a "random sample" of 50 farmners was taken in each of six districts in the villages where VEOs were resident (in 70% of the total villages under the extension program). Most respondents (77.7%) reported that they were visited by the extension agent every two weeks and those visited irregularly (22.3%) were divided equally among those who were contact farmers and belonged to a contact group, and those who were not contact farmers. Extension was thus able to reach a large majority of farmers in the target area. 43. Yield measurements for adopters were taken from adoption plots. The plots were demonstrating a package of recommendations and were monitored by the VEO during the growing season for their demonstration effect. Under these circumstances it would be reasonable to assume a yield gap between these plots and the fields of others who adopted. A survey based on a random sample of adopters which excluded the adoption plot would have more accurately represented yield increases on the farms of those farmers who had accepted extension messages. 44. The surveys conducted by the project to assess yield impact indicated high yield increases (e.g., maize yields doubled in the Coastal, Western and Central Zones, and were up 33% in the Northern  -7 - Zone). However, an analysis of national yield data3_/ for the major smallholder crops indicates that yields for maize, millet and cassava are 3% to 4% above pre-project levels, yields for sorghum and beans are equal to, and yields for coffee and rice about 15% below pre-project levels. The main exception is cashew where yields for 1995/97 are up over 40% on 1987/98 levels. The national production experience over the life of the project has been mixed. Total production of millet and sorghum is over 20% higher than before NALERP, rice 3% higher, but production of coffee and maize declined by 10%, cassava 14% and beans 25%. However, it needs to be noted that crop yield data collected by different sources displayed considerable variability. 45. There is evidence from the monitoring of results from adoption plots/units that increases in productivity have been achieved mainly for adopters of the extension messages. This would translate into overall increases in production and incomes depending on the climate and season, scale of operations, continuation of improved practices and the relationships between revenues and input costs. Improved extension would also help to improve yields in poor seasons that would otherwise fall to a greater degree. Knowledge is but one input and variable in agricultural production, and whilst extension plays a clear role in awareness, it is not possible to estimate to what extent it alone influences adoption. The absence of baseline studies makes it even more difficult to gauge the overall impact. However, there is considerable anecdotal evidence of yield increased reported by farmers and these were due to several factors such as adoption of extension messages, better inputs, use of farm implements and better access to markets. Therefore, no robust conclusion can be drawn on the impact of the project alone on the national production of food and cash crops. Project Costs and Financing 46. Total costs of the project up to 31 March, 1997 were US$27.7 million and, including committed costs of civil works to be financed by ADF of US$1.3 million, total costs will be US$28.6 million as against an appraisal estimate of US$30.4 million. Savings in unit costs, particularly for vehicles, enabled the project to expand into four additional regions within the budgeted costs and financing plan. Total recurrent costs were estimated at US$11.2 million. Actual recurrent costs of the project were US$7 million. Investment costs were higher by $2.3 million than that anticipated at appraisal. 47. The SAR identified a level of Government financing of the project equal to US$3.2 million, representing its share of recurrent operating costs (33% of local costs). This represented approximately 10% of total project costs. Goods and services, including vehicles and equipment procured under the project were exempted from taxes and duties. Therefore the total Government contribution to financing the project comprised the actual Treasury releases applied to counterpart funding of the project. This amounted to US$855,000, or approximately 3% of total project costs. This was made up of $161,000 being the Government's 5% share of the recurrent costs over $3.05 million (SDR 2.2 million) up to US$6.3 million, and 100% thereafter up to the final total of US$7 million (figures are approximate). 3' FAOSTAT - World Agricultural Information Centre - an analysis of three-year moving averages of yields and total production of major smallholder crops.  -8- C. MAJOR FACTORS AFFECTING THE PROJECT Limits on Project Counterpart Funds 48. The fiscal analysis in the SAR set out the cash flow of the Extension service which would depend on the annually increasing allocations from the Government's recurrent budget. The project was designed to place increasing responsibilities on Government to fund recurrent costs and the requirement to make these increases was written into the Development Credit Agreement as a covenant. Originally, the Bank was to finance operating costs on a sliding scale, starting at 100% reducing to 55% up to a certain limit, then nil. Following a request from the Ministry of Finance in 1994 to alter the financing percentages on the basis of a tight budgetary situation, the Bank agreed to finance 95% of recurrent costs up to the ceiling set out in the Credit Agreement. 49. The Bank's supervision missions reported that, from 1994, inadequate funding of recurrent costs from Government sources affected the project's operations particularly in relation to supervision at field levels. Repairs, maintenance and operations of vehicles were also affected. The Ministry of Finance in July, 1994 requested the Bank to consider a reduction in the counterpart funds required to be committed to the project on the basis of prevailing economic situation and tight budgetary constraints. Whilst the request was for the Bank to finance 100% of recurrent operating costs, the Bank gave its agreement to amend the Development Credit Agreement to finance 95% of incremental operating expenses. 50. Above the limit of IDA financing of recurrent costs (US$6.3 million) the costs were borne totally by Government. Project activities were scaled according to expected availability of funds from Government. As mentioned above, total recurrent costs were some $4 million short of that anticipated at appraisal. This had an impact on the scale and frequency of project activities. The Government's total contribution to the project's costs out of $29 million was $855,000, about 3%. This places the project in non-compliance with OP 6.30. Poor Integration of Agricultural and Livestock Skills at Field Level 51. Whilst the training gave some technical competence in both crops and livestock matters, some VEOs continued to emphasize their original background and were weak on the other aspects. Also, VEO lacked skills and knowledge on important topics such as farm budgeting, plant protection, farm economics, and land use. As the focus of groups for contact points was introduced, VEOs were not able to fully use the groups due to lack of adequate skills and knowledge in this area. These areas represent the future training needs for the field extensionists. New Challenges, beyond Technical Matters, for which the Extension System was not Fully Prepared 52. The messages delivered by the Extension service were predominantly technical, simple and showed a response during the growing and harvesting season in which they were applied. Given an ineffective Extension service prior to project initiation, these messages were strongly relevant early on in the project and generally were considered useful by the farmers. It is not fully determinable whether farmers found at least some of the technical messages strongly relevant, given their repetitiveness and growing farmer awareness. The changing farmer need should drive the generation of technology and its dissemination. Both the research systems and extension in agriculture need to be more demand-driven than was the case during NALERP period.  9- D. SUSTAINABILITY 53. The SAR considered sustainability of the project from two viewpoints - institutional and financial. Institutional sustainability which would depend on reaching anticipated management strength following re-structuring, commitment from all involved and a strong leadership within MAC, was achieved. Financial sustainability was expected to be achieved through a combination of (i) cost reductions due to the merger of agriculture and livestock extension, and the unification of extension programs financed by different donors; and (ii) an expected 4% growth in real terms in budget allocations to extension. The SAR estimated that the Ministry "would easily be in a position to sustain activities under the project". Government Budget allocations to MAC increased until 1992 in real terms and as a percentage of Agricultural GDP over 1988 levels. The allocation then fell below the 1988 levels in real terms and from 1.2% to 0.7% of the GDP. The financing of the project was evidently designed to place more demands on resources from within Government over the course of the project. Despite recognition in the SAR of the danger that eroding budget allocations could affect project implementation and full realization of its objectives this somewhat optimistic assumption was made, which in fact did not materialize. 54. We have also taken into account that the Bank and the Government have had an understanding of having a program approach in funding extension. In the SAR, the NALERP was deemed as " the first phase of a long -term process to establish an efficient, cost-effective, and sustainable --extension service on the Tanzanian mainland". Following this approach a second phase extension project was approved by the Bank (effective October 1996) and is currently under implementation. The activities started under the NALERP are being sustained under the phase 2 project NAEP II. E. BANK PERFORMANCE 55. The Bank's performance during the design stages was satisfactory, with the Bank providing guidance to the National Task Force on Extension during 1987. The Bank also assisted the Ministry to implement the pilot project, based on T&V, covering some 50,000 families in five districts. The pilot T&V project gave an insight into problems that later were to appear again during NALERP. The SAR identified the problems as: (i) lack of funds thereby reducing the frequency of regular training sessions and workshops; (ii) ineffective contacts between research, extension and farmers; (iii) inadequate supervision of activities. As mentioned above, the pilot addressed already-established farner groups, and not groups established by VEOs. The SAR admits that it was too early to evaluate the pilot performance and impact, but yet concluded (based mainly on results from other countries) that T&V was the correct approach as it was an efficient management tool, developed stronger and more efficient linkages with the research community, and induced research to develop farmer oriented programs. The design process should have included a comprehensive evaluation of the pilot phase. 56. Bank performance in portfolio management after the project became effective was satisfactory. Bank supervision missions were frequent, with staff from the Resident Mission and the Nairobi Bank Regional Mission participating. Bank staff assisted the Government in the mid-term review  -10- of the project and worked jointly with the Government on the preparation of NAEP II and the development of pilot initiatives that will be supported under that project. F. BORROWER PERFORMANCE 57. Generally, the performance of the Borrower has been satisfactory. The main covenants relating to timely implementation of the project were complied with although there was a delay at the outset. Legal covenants monitored by the Bank's supervision staff have been complied with. Accounting and reporting have been carried to a satisfactory standard through the Technical Assistance to the Department financed by the credit. The Government has also prepared its contribution to the ICR. 58. However, two covenants not monitored by the Bank in the Development Credit Agreement are worthy of mention. Section 3.04 required the Government to maintain the establishment of the extension service at the level agreed with the Bank as at September 1989, with staff leaving being replaced with diploma/certificate holders. The field staffing levels under T&V have a significant impact at the level of outreach achievable and, as mentioned above, substantial numbers of staff have been retrenched under the rationalization of MAC. Therefore, strictly speaking, this covenant was partially complied with. Indeed it was not possible for the Government to comply, given the impact of later Bank- supported adjustment operations on staffing levels within the public service. 59. There was also a covenant, Section 3.05, that concerned budgetary allocations to extension. The Government was to annually increase budgetary allocations over the level provided in 1987/88 fiscal year to cover annual costs as agreed in the annual work plan. This was not complied with. G. ASSESSMENT OF OUTCOME 60. The project achieved the objective of revitalizing extension service, building capacity to manage and make decision on future directions, and giving the service a new work ethic. The successfully executed training program helped improve knowledge base and skills of the extension service. Structured work schedules of the frontline extensionists brought orderliness in contacting farmers and in extension delivery. The two-way traffic between the researchers and the farmers in resolving farmer production issues through extension intermediation got well established, and linkages between farmers, extension and research improved. Substantial yield increases were recorded on adopters fields. Over the project period the extension methodology underwent a slow but measured change to respond to emerging challenges such as the down-sizing of civil service and consequently of the front-line extensionists. The program adopted the group approach in place of the earlier contact farmer approach to maintain its outreach and to make consultative process more participatory. The extension service participated in and helped carry out several pilot initiatives to respond to farmer needs - the FAO's food security pilot and Sasakawa Global's input supply initiative were supported and partly executed at the field level by the extension service. The lessons learned have been factored into the Phase 2 extension project.  61. However, there were several emerging demands and challenges in its long implementation period that the project did not address adequately. It also faced implementation constraints due mainly to acute financial stringency faced by the Government. A central role for farmers in technology generation and dissemination was not envisaged in project design and therefore extension advice continued to be more top-down than demand-driven. The group approach in contacting farmers only partially addressed this deficiency. The extension management arrangement continued to be largely monolithic with MOA hierarchy remaining in charge, from policy formulation at the headquarters level to implementation at the village level. Local Governments, other stakeholder agencies like NGOs and farmer organizations did not have much structured contact with MOA implementers. Though private sector and NGOs started emerging as providers of extension in different parts of the country, the project did not take note of these developments. The project implementation in the field suffered from government's inability to find adequate counterpart resources to fund project activities (almost all development projects faced this issue including this project). Therefore, while the development objectives were achieved, the outcome of the project is considered marginally satisfactory for the above reasons. 62. These lessons of NALERP have been appropriately incorporated in the follow-on project namely the National Agricultural Extension Project Phase 2 (NAEP 2). The phase 2 project encourages pluralism in providing extension services. Local authorities and other stakeholders including NGOs are represented on the district extension coordination committees responsible for program planning and implementation oversight. The project provides for undertaking pilot initiatives in the areas of extension management, extension methodology, pluralism and other related issues articulated by the beneficiaries. Impact assessments through surveys and beneficiary assessments have been planned and would provide reliable indicators for evaluating project impact. H. FUTURE OPERATIONS 63. A review of NALERP was undertaken in 1995 as part of the preparation for NAEP II which then addressed further modifications to the T&V approach that would finally be part of the NAEP II design. These were to (i) address issues in financial sustainability through collaboration with NGOs, and through the transfer of responsibility for commodity-focused farming systems to private agro-industries; (ii) farmer participation in the generation and dissemination of technology; and (iii) use of communication technology in extension. In addition, the research programs now being undertaken have incorporated farming systems research with environmental and production sustainability as key objectives. 64. The operations of NALERP have been taken over by NAEP II, which will follow the T&V approach but will continue with the emphasis on group-based contact points and test through pilots several new initiatives in extension methodologies and extension management. NAEP II is operational in the same 16 regions as NALERP and the IFAD Southern Highlands project in the remaining four regions of mainland Tanzania iis expected to be extended and adopt similar alterations in emphasis.  - 12 - I. KEY LESSONS LEARNED 65. The key lessons learned from the project were as follows: * In a long-term program, specific objectives, with indicators, for each of the phases in the program should be clearly identified at the start of each phase. In any case, indicators should be set out in order to objectively monitor project achievement. * The pilot scheme provided insufficient input to project design since inadequate time was allowed for experience to be generated and an evaluation to be made. However, later the lessons learned from fertilizer experiments were suitably incorporated into the extension messages under the program. * Simple technical messages are often initially relevant to farmers on initial contacts with an extension service. Later, the messages need to address newly emerging issues raised by the farmers. In order to appropriately respond to these issues the extension system should be more demand-driven. Whereas project activities within the direct control of project management (e.g., numbers and frequency of contacts, bi-monthly workshops, supervision visits) were adequately monitored, the impacts were not sufficiently studied. The project did not carry out initial base-line surveys to later measure project impacts. The involvement of farmers in designing and operating extension programs is a critical element in ensuring the continued technical relevance and usefulness of the service, which is essential for maintaining farmer demand and its technical sustainability. * The change to group-based extension was positive as ownership of the program was then transferred to the community. Extension staff should have skills and knowledge in managing groups, and participatory approaches. * Government commitment is essential for T&V projects to have a successful outcome, both in terms of staffing numbers and funding from Government budgetary resources. If a Government finds difficulty in providing the required resources to support a T&V extension system, the operations should be scaled down or redesigned to match the resources that the Government can reasonably be expected to allocate to extension in the longer term. * One of the main purposes of the ICR exercise is to analyse and document the lessons learned and assess the future operations of projects. An ICR exercise, completed after a follow-on project is already effective, can provide little input to the preparation and design of following projects, and an opportunity is thereby lost to fully incorporate the lessons learnt in further Bank assistance. In the context of a long-term program which envisages follow-on Bank assistance, the procedures for  - 13 - ICRs need to be altered to require the ICR to be completed and available at appraisal of succeeding operations.  - 14- PART II: STATISTICAL TABLES Table 1: Summary of Assessments A. Achievement of objectives Substantial Partial Negligible Not (/) (1) (1) Applicable (/) Macro policies El :I 0 El Sector policies t1 El Financial objectives El F7i 0I [I] Institutional development E El El Physical objectives n7 F71 Eo E Poverty reduction 7 l E E Gender issues El E1 El El Other social objectives El Fl El E Environmental objectives EF F] El E Public sector management El El E E Private sector development El EZ E L Other (specify) El El El El B. Project sustainability Likely ll Unlikely Uncertain 1/ In the longer termn.  - 15 - Highly C. Bank performance Satisfactory Satisfactory Deficient ('/) (1) (1) Identification [ Preparation assistance E E Appraisal i: Supervision 7 El Highly D. Borrower performance satisfactory Satisfactory Deficient (V') (/) (1) Preparation L 7 F Implementation Covenant compliance L F E Operation (if applicable) LI L E Highly Highly E. Assessment of outcome satisfactory Satisfactory-4J Unsatisfactory Unsatisfacto (5/) (a/) (a/) ~~~~ry (V) /I Fgl Ltfc "Marginally Satisfactory.  - 16- Table 2: Related Bank Loans/Credits Loan/credit title Purpose Year of approval Status Preceding operations 1. National Rehabilitation of ag. 1989 Closing 31 Agricultural and Research services December 1997. Livestock Research Project Following operations 1. National Phase II of extension 1996 Ongoing Agricultural Extension rehabilitation. Project II 2. Agriculture Sector Support to rationalizing 1993 Ongoing Management Project MAC, strengthening core functions, sector expenditure review. Table 3: Project Timetable Steps in project cycle Date planned | Date Actual/ | | Latest Estimate Identification (Executive Project Summary) March 1988 Preparation March-December 1988 Appraisal April 15, 1988 Negotiations November 1988 November 28, 1988 Board presentation February 1989 March 21, 1989 Signing 6 March 1989 May 8, 1989 Effectiveness May-June 1989 July 6, 1989 Midterm Review May-June 1992 June 1993 Project completion 31 March 1997 March31, 1997 Loan closing 31 March 1997 March 31, 1997  -17- Table 4: IDA Credit Disbursements: Cumulative Estimated and Actual (US$ million) FY FY FY FY FY FY FY FY FY 89 90 91 92 93 94 95 96 97 Appraisal estimate 3.0 5.7 8.5 11.7 14.1 16.0 17.6 18.4 Actual 0.0 2.1 4.9 6.6 9.0 12.0 14.5 18.2 19.3 Actual as % of estimate 0% 36% 57% 57% 64% 75% 83% 99% Date of final disbursement July 31, 1997 Table 5: Key Indicators for Project Implementation I. Key implementation indicators in SAR/ Estimated targets1/ Actual (% achievement) President's Report 1. Farmers covered by extension program 2.8m 2.14m (75%) 2. Fanners contact groups 237,996 92,262 (38%) (regularly contacted) 3. Adoption of crop 85% 67% recommendations 4. Adoption of livestock 65% 43% recommendations 5. Monthly training sessions 4,944 77% 6. Bi-monthly workshops 480 65% 7. Adoption plots 221,320 140,909 (64%) 8. Supervision Div/Dis/ 32,268/4,404/900/434 54%/42%/50%34% Region/HQ trips 1/ Targets based on best farmers performance and fixed by MOA during implementation.  - 18 - Table 6: Key Indicators for Project Operation I. Key operating indicators in SAR/President's Report Estimated Actual Not Applicable Not Applicable Table 7: Studies Included in Project Purpose as defined Study at appraisal/redefined Status Impact of study 1. Needs Assessment Not defined at appraisal Complete Input to NAEP II on Extension design Communication (Tanzania Food and Nutrition Centre) 2. Training Needs Not defined at appraisal Complete Input to NAEP 11 Assessment (Dr Ann design Stroud, Dr S Lugeye) 3. Benchmark survey Not defined at appraisal -- Input to NAEP II for NAEP II (Dr A design Temu, Sokoine University) 4. An assessment of Not defined at appraisal Input to NAEP II NALERP by dsg Beneficiaries (Dr N Molel, Sokoine University)  - 19- Table 8A: Project Costs Appraisal estimate Actual/latest estimate (US$M) (US$M) Local Foreign Total Local Foreign Total Costs costs Costs!' Costs../ Item 1. Extension Organization 6.0 15.8 21.8 - - 21.6 and Management 2. Staff and Farmers 3.6 2.6 6.2 - - 4.9 Training 3. Technical Assistance - 2.4 2.4 - - 2.1 TOTAL 9.6 20.8 30.4 - - 28.6_/ a/ Detailed distribution between local and foreign cost figure not available. b/ Actual cost is below appraisal estimate because of shortage in counterpart funding and reduction in Govermnent's share of operating costs. Table 8B: Project Financing Appraisal estimate (US$M) Actual/latest estimate (US$M) Local Foreign Total Local Foreign Total Costs Costs costs al costs a/ Source IDA 4.4 14.0 18.4 - - 19.3_ African Development Fund 2.0 6.8 8.8 - - 8.5 Domestic contribution 3.2 - 3.2 - - 0.8 TOTAL 9.6 20.8 30.4 - - 28.6 a/ Detailed distribution between local and foreign cost figure not available. b/ Actual disbursements exceeded appraisal estimates due to depreciation of the US$ against the SDR.  - 20 - Table 9: Economic Costs and Benefits No economic rate of return for the project was calculated at appraisal. The economic benefits that were to result were in the form of additional production from farmers adopting improved technology. The SAR estimated that incremental production of 1,800 tons of paddy, 99,000 tons of maize, 3,000 tons of cashew and 31,000 tons of seed cotton would be produced as a direct result of the project, which would relate to an economic benefit of US $30 million in 1988/89 real terms. As mentioned in the main text, an analysis of national yield data for the major smallholder crops indicates that yields for maize, millet and cassava are 3% to 4% above pre-project levels, yields for sorghum and beans are equal to, and yields for with coffee and rice about 15% below pre-project levels. The main exception is cashew where yields for 1995/97 are up over 40% on 1987/98 levels. Overall national production of millet and sorghum is over 20% higher than before NALERP, rice 3% higher, but production of coffee and maize has declined by 10%, cassava 14% and beans 25%. Cashew production has risen from 18,000 tons to 82,000 tons over the project period and whilst extension had a part to play in this increased production, the liberalization of marketing and the affordability and availability of the chemicals to contain powdery mildew disease had probably a greater effect.. There is no evidence that the potential yield improvements that could have been obtained through adopting new technology have been translated into national figures. No robust conclusions on the economic benefits of the project in terms of incremental production can be drawn.  -21 - Table 10: Status of Legal Covenants Original Revised Covenant Present fulfilment fulfil- Description of Agreement Section type status date ment covenant Comments date DCA 3.01 (b) 05 C NA The borrower shall cause the With an exception of time implementation of the project in loss in the initial period accordance with the program in Schedule resulting in late (delayed) 4 commence-ment by one year, implementation has progressed as scheduled. Complied with. DCA 3.03 05 C NA The borrower shall establish and maintain With an exception of.time an Extension Unit headed by Assistant loss in the initial period Commissioner for Extension Services resulting in late (delayed) commence-ment by one year, complied with. DCA 3.04 05 CP NA The Borrower shall (a) until September (a) and (b) complied with; 30 1989 maintain its present (c) not complied with. establishment for the Extension Service, Partial compliance (b) by September 30, 1989, furnish to the Association the revised establishment for the Extension Service in accordance with MALD's new structure, and (c) thereafter maintain the establishment at that level, with the recruitment of extension staff from fresh diploma/certificate holders being offset by attrition. DCA 3.05 05 NC' NA The Borrower shall annually increase the Not complied with. budgetary allocations for agricultural and livestock extension over the levels provided in 1987/88 fiscal year, to cover annual costs as agreed in the annual work plan. DCA 4.01(b) 05 C NA Furnish Audit report within 9 months In compliance. after the financial year end. DCA 4.01(c) 01 C NA Maintain satisfactory records of SOE. In compliance. Separate audit for Special Account and for expenditures on the basis of SOEs. Status: C = Complies with. CD = Complied with after delay. NC = Not complied with. CP = Complied with partially. NA = Not available.  - 22 - Table 11: Compliance with Operational Manual Statements Statement number and title Describe and comment on lack of compliance Basically, there was compliance with the applicable Operational Manual Statements. Table 12: Bank Resources: Staff Inputs Stage of Planned Revised Actual 1/ project cycle _ l Weeks US$ Weeks US$ Weeks US$ Preparation to NA NA 28.9 71.9 Appraisal l Appraisal NA NA 22.4 56.4 Negotiations through NA NA 4.3 11.4 Board approval Supervision NA NA 235.1 543.6 Completion 8.6 35.8 TOTAL 299.3 719.1 1/ Including travel costs (direct costs). NA = Not available--not recorded in the FACT system.  - 23 - Table 13: Bank Resources: Missions Performance rating Number Specialized Implemen Develop- Stage of Month! Of Days in Staff skills -tation ment Types of project cycle year Persons field Represented a! status objectives b/ Problems c/ Through appraisal . Appraisal through Board approval Supervision Sep-89 2 8 Ext NR NR EF 66 Feb-90 2 10 Ext, Serv S S Mktg, Cr 44 May-91 4 14 Ext, Ser, Coops, S S Staff, M&E Train C4 Dec-91 2 14 Ext, Ser S S Staff, Fin MTR Jun-93 2 10 Ext, Serv S S Fin Supervision Nov-93 5 5 Ext, Ser, Fin, Proc S S Fin Oct-95 2 18 Ser S S Fin Mar-96 2 17 Ser, Ext S S Fin Sep-96 2 5 Ser, Ext S S Fin Dec-96 3 14 Ser, Ext S S Fin Feb-97 2 15 Ser, Ext S S Fin Completion5/ Oct-97 2 14 Econ, Ext I I Abbreviations a/ c/ Ser - Agricultural Services Specialist EF - Conditions for Effectiveness and Implementation Ext - Agricultural Extension Specialist start-up Coops - Cooperatives Specialist Mktg - Marketing Train - Training Specialist Cr - Credit Proc - Procurement Specialist Staff- Staffing of the extension service Econ - Economist Fin - Financial Management, counterpart funds. b/ S - Satisfactory 5' Undertaken by FAO/CP.  APPENDIX 1 TANZANIA NATIONAL AGRICULTURAL AND LIVESTOCK EXTENSION REHABILITATION PROJECT (CR. 1994-TZ) IMPLEMENTATION COMPLETION MISSION AIDE MEMOIRE  AIDE MEMOIRE CONTENTS Project Objectives and Components ........................... 25 Implementation Timing ........................... 26 Achievement of Objectives ........................... 26 Implementation Experience and Results ........................... 27 Financial and Economic Aspects ........................... 32 Bank Performance ........................... 33 Borrower Performance ........................... 34 Major Factors Affecting Project ........................... 34 Future Operations ........................... 34 Sustainability ........................... 35  FOOD AND AGRICULTURE ORGANIZATION OF THE UNITED NATIONS Investment Centre Division TANZANIA: National Agricultural and Livestock Extension Rehabilitation Project (NALERP - IDA Credit 1994-TA; ADF Loan LVS/88/22) Implementation Completion Report - Aide-memoire Introduction 1. A missionl from the FAO Investment Centre visited Tanzania from 3 to 17 October, 1997 to prepare the Implementation Completion Report for the National Agricultural and Livestock Extension Rehabilitation Project (NALERP) under the FAO/World Bank Cooperative Program. The mission worked closely with officials from the Ministry of Agriculture and Co-operatives (MAC), particularly the Department of Extension Services, and had discussions with the Department of Research and Training within MAC. The mission visited two regions of Tanzania to assess the impact of the project in areas of different agricultural potential and fanning systems. The Mission held discussions with Regional and District Agriculture and Livestock Development Officers, Regional Extension Officers, and visited farmers, farmer groups, District, Divisional and Village Extension Officers in Morogoro and Kilombero Districts in Morogoro Region and Magu District in Mwanza Region. The mission also visited the Katrin Research Centre and Ukiriguru Research Institute. 2. This aide-memoire was discussed at a wrap-up meeting, chaired by the Commissioner of Agriculture and Livestock Development which was attended by staff of the Ministry who were involved in the implementation of the project. 3. The mission would like to record its appreciation for the cooperation and assistance shown by all officers and staff of the Govemment with whom the mission has been working. Project Objectives and Components 4. The objective of the project was to enable the Govemment of Tanzania to provide the farming community with adequate extension messages to enhance agricultural production, productivity and rural incomes. The project represented the first phase, regarded as the rehabilitation phase, of a long-term process to establish an efficient, cost-effective and sustainable extension service. The project was to focus on a gradual but steady and well-planned merging of crop and livestock extension services into a multidisciplinary system, the strengthening of extension management, organization, coordination and supervision of activities, and the upgrading and/or provision of necessary human resources, physical infrastructure and logistic support. 'Tim Lamrock (Economist/Mission Leader - TCIR) and Brian Scoullar, (Extension Specialist - Consultant) 25  5. This was to be achieved through: (i) the establishment of a streamlined, adequately staffed organization and management structure; (ii) the introduction of an extension system based on the Training and Visit (T & V) methodology, together with necessary training and supervision of activities; and (iii) strengthening of the Ministry's planning and implementation capacity, and assisting in the preparation of a second phase project, through both short and medium-term technical assistance. 6. At the wrap.-up meeting, the Govemment viewed the project's objectives as the rehabilitation of the extension service in order to set a base for a systematic extension delivery mechanism. 7. The Mission considers that the objectives were appropriate but were not clearly stated in the SAR, particularly as to the level of achievement that NALERP was to have reached, at project completion, towards the longer-term goal of establishing an efficient, cost-effective and sustainable extension service. Implementation Timing 8. The credit ifor SDR 13.7 million (US$ 18.4 million equivalent) was signed in May, 1989 and became effective in July, 1989. The project was co-financed by the African Development Fund (ADF loan LVS/88/22) which was responsible for the financing of the construction or upgrading of housing for Division staff, the rehabilitation and provision of facilities in four training institutes and the overseas training program. 9. The IDA credit closed on schedule on 31 March, 1997 with the credit fully disbursed. The ADF loan has not closed, and the remaining civil works under the project will be completed by March, 1998. At closing of the ADF loan, it is expected that US$ 8.45 million of a total of US$ 8.8 million will be disbursed. Achievement of Objectives 10. The project has satisfactorily achieved its development objective of rehabilitating the extension services provided to farmers and livestock keepers. The project re-invigorated the field, district and region extension staff who previously had no meaningful work program and only limited financial and logistic resources to carry out their work. The project has had some impact on agricultural production, productivity and rural incomes (but see para 0). 11. The project suffered in the second half of its life, due to mainly macro-economic conditions, from decreasing funds for regular activities, decreasing numbers of front-line staff, and an inability of the extension service to diagnose problems and change the messages in response to farmers needs and circumstances. Project activities are now being supported by National Agricultural Extension Project Phase II (NAEP II) which was effective in October, 1996. 26  Implementation Experience and Results General 12. In general, the implementation of the project has been satisfactory. After a delay at the outset of the project which effectively put the implementation schedule back by one year, implementation has proceeded as scheduled. At the time of termination, the project operated in 16 of the 20 mainland Regions having been included in the following order. 1989: Tabora, Singida, Dodoma 1990: Kilimanjaro, Kagera, Arusha 1991: Shinyanga, Mwanza, Mara 1992: Mtwara, Morogoro, Lindi 1993: Kigoma 1994: Tanga, Dar es Salaam, Coast 13. The four Southern Highlands regions of Ruvuma, Rukwa, Mbeya and Iringa are served by the Southern Highlands Extension and Rural Financial Services project. The project, funded by IFAD and GOT, also follows the T&V approach with the addition of a credit facility for farmers. 14. At the height of the project 3,319 VEOs were responsible for 3,680 villages, 2,145,000 farm families (assuming 583 farm fanilies per village) and 646 farm families per VEO against project targets of 4,113 VEOs, 4,800 villages and 2,800,000 farm families at a ratio of 681 farm families per VEO. 15. The project contributed to coordination of other donor activities in extension, but there was limited adoption of the strict T&V methodology in other projects. Other projects in extension implemented outside the MAC did not unify their approach to that of NALERP, nor did they coordinate their programmes. 16. The second phase of Bank support to extension, NAEP II, was prepared and appraised during the course of NALERP. Institutional change 17. The Ministry has amalgamated the Departments of Agriculture and Livestock under a Commissioner who reports to the Principal Secretary. Extension is under an Assistant Commissioner (AC) who is supported technically by ACs for Plant Protection, Livestock Development, Crop Development and Irrigation. At the regional level, unification was achieved through a Regional Agricultural and Livestock Development Officer (RALDO) supported by specialists and a Regional Extension Officer (REO), a structure which is repeated at the district level. At the divisional (ie subdistrict) level an Extension Officer, usually holding a Diploma, supervises 8 - 10 Village Extension Officers (VEOs). 18. The institutional changes associated with unification, in termns of structure, management and administration have been successfully implemented and are perceived as being a positive development at all levels in the organization. At the front-line, although retrained, a significant number of VEOs, particularly those who formerly were Livestock Extension Officers, preferred their previous area of specialization. This is not surprising given a training period of only five months and a natural tendency to feel more confident in 27  the discipline in which one was first trained and has the longest experience. Management and Supervision 19. Implementing the T&V system established a strong management system and management ethic within the Extension service. Regular visit schedules to contact farmers and groups are maintained and monitored through reporting procedures and field visits. Supervision at all levels however was adversely affected by lack of operational funds, and supervision by headquarters, regional, district and divisional staff were 34%, 50%, 42% and 54% respectively of target levels. Mid-term Review 20. The Mid-Term Review, seen by management as an opportunity to take stock of progress and problems, was preceded by discussions and interviews with extension staff, trainers, farmers, researchers, NGOs and others in 8 of the 13 regions of the project. The results of the survey guided discussion at the MTR and enabled informed decisions to be taken on a wide range of issues including visit schedules, supervision, adoption plots, message content and the use of groups as an alternative to individuals as contact points for extension messages. These examples demonstrate an increased managerial capacity of MAC to plan, implement and supervise a national extension programme, and this increased capacity is attributable to the project. Training Program 21. Generalist front-line officers have been created by re-training through courses of five months duration as opposed to ten months as anticipated at appraisal. At project initiation, approximately 40% of front-line staff were untrained Field Assistants. No training was provided at this level and most of the Field Assistants have been retrenched. 22. The VEO retraining program was slow up to the time of the mid-term review at which time only 18% of the target had been reached. The situation improved after MTR and certificate-level retraining was completed by June, 1997. The training of trainers program was 54% completed. Other higher-level, overseas training and overseas study tours were carried out satisfactorily. Civil Works 23. The civil works component for DivEO housing, financed by ADB, was not satisfactorily implemented. The initial assumption that village labour would be volunteered to assist in the construction of 300 DivEO houses did not eventuate. A reduced building program was then planned (141 houses) with higher unit costs, but this then fell behind due to delays and failure of contractors to perform as required. At the time of closing of the IDA credit, a total of 67 houses had been completed but contracts for the remaining houses is in force and construction is underway. The DivEO housing program has been incorporated into NAEP II. The rehabilitation of the Training Institutes was satisfactorily completed. Technical Assistance 24. The long-term technical assistance team to support project implementation started work during 1990, and finished in 1992 with the exception of the Financial Advisor who remained throughout the 28  duration of the project. The experiences of the technical assistance financed by the project were mixed. In general, the long-term consultants in Agricultural Extension Methodology, Finance and Procurement proved to be satisfactory, but the Monitoring and Evaluation Specialist's terms of reference were not completed at the end the assignment. This had important implications on the performance of the Monitoring and Evaluation component of the project. The short-term consultancies proved to be effective with the exception of the consultancy on cotton research which did not result in developing appropriate recommendations for farmers. Monitoring and Evaluation 25. Implementation of this component was negatively affected by the consultant not completing the task of developing and implementing systems for M & E of project activities and impact. The SAR specified that baseline surveys would be undertaken in the first and second years of the project and impact evaluations of T&V in the regions at the commencement of years two and three. These surveys were not undertaken. The Overall Evaluation System specified in the SAR was not installed. Although introduced only in September 1994, quantitative goals are now established prior to the commencement of the growing season. Research - Extension Collaboration 26. Formal links between Research and Extension were built into the project design. These included the posts of Research-Extension Liaison Officers (RELOs), pre-season workshops, bi-monthly workshops for the training of SMSs and two annual workshops to be held at zonal research stations. The Farming Systems Research Programme was expected to provide additional opportunities for collaboration within the context of establishing and responding to priority needs of rural communities and RELOs were expected to be seconded to the FSR team to facilitate this work. 27. Limited improvements in working relationships were achieved during the project period. RELOs were not appointed as management decided that the REO could perform this task. No pre-season workshops were held. At best, the seasonal impact points selected by extension officers were sent to the zonal research station for their comments. The monthly workshops, which were changed to bi-monthly early in project implementation, did provide a forum for effective interaction between extension and research. With NALERP funds, researchers generally attended as resource persons. However, research priorities were not reviewed at these workshops, but this role has now been taken over by Zonal Committee Meetings which set zonal research priorities. 28. At the MTR the FSR programme was also judged as being unsuccessful in establishing closer relationships. Some improvements however have occurred since 1993. The FSR programme has mapped sub-zones and increased its use of PRA approaches in identifying problems. Inevitably, the subsequent research plots are scattered and difficult and expensive to establish and monitor. Some research stations have entered into formnal agreements with Extension for assistance to implement the research programme. TORs have been agreed and time allocated in Extension workplans for collaboration at district and field levels. These arrangements are currently rated as useful by both Research and Extension staff in the zone in which it was observed by the mission. 29. Initiatives were taken by the Ministry to improve Research-Extension collaboration. The CALD and the CRT in 1992 agreed to provide resources to improve linkages in defined areas. The 1992 policy statement and implementation guidelines addressed the issue, a Zonal Technical Committees have been 29  formed, Extension is represented at the national and zonal levels of research committees and donor coordinating committees have been formed in some regions to remove duplication in extension and research activities. These initiatives however appear to have been of limited benefit. The most likely strategy for improving collaboration is a steadily increasing formalization of roles, responsibilities, funding arrangements and lines of communication within the framework of FSR because this method involves close collaboration with farming communities, by both research and extension systems. Extension - Farmer Contact 30. For at least half of the project duration, VEOs measured contact by counting the number of farmers who attended each extension activity; ie demonstrations on the fields of Contact Farmers, with Contact Groups and at field days. All farmers, including contact farmers, who attended each activity were counted as a new contact, even though they may have participated previously, and therefore the figures represent more a measure of contact frequency and not the number of farmers contacted. Moreover no adjustments were made in the figures for more than one adult attending from one farm household. This methodological problem was not raised at the MTR and no reference to it was seen in the supervision reports of the Bank which were available to the mission. 31. The Ministry became aware of this discrepancy and took steps to rectify it during the second half of project implementation. The M&E Unit introduced a listing of farmers by name and by farm household as the means of establishing the degree of farmer contact. For various reasons, acceptance of this method within the Service was not complete by project termination. 32. It is therefore difficult to arrive at a figure for extension contact with farm households in the project area, but the figure of 2,140,000 probably overstates the actual numbers contacted. Adoption 33. The first attempt to evaluate impact was carried out in 1995 as part of the preparation for the proposed second phase of support to Tanzania's Agricultural Extension System. Undertaken by a consultant for the Ministry and implemented through the M&E Unit of NALERP, a "random sample" of 50 farmers was taken in each of six districts. Most respondents (77.7%) reported that they were visited by the extension agent every two weeks and half of the remainder also identified themselves as Contact Fanners. The subsequent report of the Ministry indicated that the farmers were selected from villages in which a VEO was residing. In 1995 approximately 30% of villages did not have a resident VEO. The survey therefore was an analysis of the impact of the T&V system on awareness and adoption amongst Contact Farmers from villages in which a VISO resided and not of the farming population as a whole in the six districts studied. As a result,the Ministry report gives wider credence with respect to contact, awareness and adoption than is justified. 34. Field visits by the mission revealed that the study was not an isolated incidence of this error. Discussion with Ministry staff revealed that adoption figures quoted were for farmers with whom the Extension service had contact and not from the farming population as a whole. Visits by the mission to villages which did not have a resident VEO gave an initial indication that extension- farner contact and adoption was lower than in those where a VEO resided. Insufficient time was available however to test further this initial impression. 30  Yields 35. Yield measurements for adopters were taken from adoption plots. The plots were demonstrating a package of recommendations and were monitored by the VEO during the growing season for their demonstration effect. Under these circumstances it would be reasonable to assume a yield gap between these plots and the fields of others who adopted. A survey based on a random sample of adopters which excluded the adoption plot would have more accurately represented yield increases on the farms of those farmers who had accepted extension messages. Modifications to the Basic T& V during Project Implementation. 36. The Ministry modified the Extension methodology in response to operational experience and funding constraints. Within the first 12 months of project implementation, VEO training sessions were reduced in frequency from fortnightly to monthly and SMS training from monthly to bimonthly. As a result, more time was available for extension, costs were reduced, and there was no loss of staff competence in the messages they were disseminating to farmers. 37. The major change has been away from contact farmers established for points of regular contact with VEOs, to VEOs working with groups, in many cases already established for other purposes. The pilot T&V project established to test the effectiveness of this methodology targeted 6,500 groups covering about 50,000 families. These groups were based on the Ujamaa concept and were already established as part of an administrative set-up, or for collective agricultural production or political purpose. NALERP then attempted to set up contact farmer groups through the direct actions of VEOs. These new groups were often formed on an incorrect perception of purpose - the members believed that they would obtain credit or inputs in addition to extension. This was observed in 1991 and questions of sustainability of the contact farmer groups newly set up for extension was questioned at that time. By the time of MTR, the emphasis was on working with farmer groups that either existed already within the communities, or were established in harmony with community practices. Relevance and Feasibility of Extension Messages 38. The messages delivered by the Extension service were predominantly technical, simple and showed a response during the growing season in which they were applied. Given an ineffective Extension service prior to project initiation, these messages were relevant early on in the project and generally were considered useful, at least by those in close contact with extension staff. As demand was satisfied (or could not be adopted through resource constraints) farmers' needs changed. In resource-poor areas farmers were concerned with environmental degradation and a lack of capital. In resource-rich areas, farm families were looking to intensify and diversify their farming enterprises. In both cases farmers' needs for new knowledge and skills increased in complexity, cut across sectors and to some extent embraced a longer time horizon. Generally, however, extension messages did not change, and therefore no longer were relevant. 31  Financial and Economic Aspects. Project Costs 39. Total costs of the project up to 31 March, 1997 were US$ 27.7 million and, including committed costs of civil works to be financed by ADF of US$ 1.3 million, total costs will be US$28.6 million as against an appraisal estimate of US$ 30.4 million. Savings in costs, particularly for vehicles, enabled the project to expand into four additional regions within the budgeted costs and financing plan. Table 1: Project Costs and Financing (US$ ,000) Financina Sourcel IDA AUF (;UI lotal SAH Estimate Component Extension Organization and Management 13,580.0 7,130.0 855.0 21,565.0 22,109.0 Training 3,600.9 1,320.0 4,920.9 6,053J Technical Assistance 2.096.4 2.096A 2.238.0 Total 19,2773 8,45011 855.0 28,5823 30,400.0 40. Total recurrent costs were estimated at US$ 11.2 million. Actual recurrent costs of the project were US$ 7 million. Investment costs were higher by $2.3 million than that anticipated at appraisal. 41. The SAR identified a level of Government financing of the project equal to US$ 3.2 million, representing its share of recurrent operating costs (33% of local costs). This represented approximately 10% of total project costs. Goods and services, including vehicles and equipment procured under the project were exempted from taxes and duties. Therefore the total Government contribution to financing the project comprised the actual Treasury releases applied to counterpart funding of the project. This amounted to US$ 855,000, or approximately 3% of total project costs. This was made up of $161,000 being the Government's 5% share of the recurrent costs over $3.05 million (SDR 2.2 million) up to US$ 6.3 million, and 100% thereafter up to the final total of US$ 7 million (figures are approximate). Counterpart Financing 42. The Bank's supervision missions reported that, from 1994, inadequate funding of recurrent costs from Government sources affected the project's operations particularly in relation to supervision at all levels. The Ministry of Finance in July, 1994 requested the Bank to consider a reduction in the counterpart funds required to be committed to the project on the basis of prevailing economic situation and tight budgetary constraints. Whilst the request was for the Bank to finance 100% of recurrent operating costs, the Bank gave its agreement to amend the Development Credit Agreement to finance 95% of incremental operating expenses, as opposed to a sliding scale of 95%, 80% and 55%. 43. Above the limit of IDA financing of recurrent costs (US$ 6.3 million) the costs were borne totally by Government. Project activities were scaled according to expected availability of funds from Government. As mentioned above, total recurrent costs were some $4 million short of that anticipated at appraisal. This had an impact on the scale and frequency of project activities. 32  44. A review of the statement of Government counterpart funds released to the project indicates that in March, 1994, a Treasury release of TSh 176 million was effected which brought the Government's total counterpart funds for that year to TSh 198 million (US$ 438,700 equivalent). Treasury releases in the remaining stages of the project amounted to TSh 189 million (US$ 326,000 equivalent). This amount was used to pre-finance activities under NALERP. When the project closed on 31 March, 1997, a substantial portion of these releases for counterpart funds remained unutilized and an amount of TSh 290 million (US$ 480,000 equivalent) was transferred to NAEP II accounts as Government pre-financing of NAEP II. Benefits 45. The ultimate objective of the project was enhanced agricultural production, productivity and rural incomes. Knowledge is but one input and variable in agricultural production. The mission feels that it would be unreliable to draw national or regional conclusions on project benefits based on production increases as it is difficult to identify benefits that derive from extension alone. 46. The absence of baseline studies makes it even more difficult to gauge the overall impact. There is clear evidence from the monitoring of results from adoption plots/units compared with non-adopters that, although the quantum may be overstated, (as discussed in para 35) significant increases in productivity have been achieved for adopters. This would translate into overall increases in production and incomes depending on the scale of operations, continuation of improved practices and the relationships between revenues and input costs. The extent to which this is the case cannot be determined from available data. Bank Performance Project Design 47. The pilot T&V project gave an insight into problems that later were to appear again during NALERP. The SAR identified the problems as: (i) lack of funds thereby reducing the frequency of regular training sessions and workshops; (ii) ineffective contacts between research, extension and farmers; (iii) inadequate supervision of activities. As mentioned above, the pilot addressed already-established farmer groups, and not groups established by VEOs. The SAR admits that it was too early to evaluate the pilot performance and impact, but yet concluded (based mainly on results from other countries) that T&V was the correct approach as it was an efficient management tool, developed stronger and more efficient linkages with the research community, and induced research to develop farmer oriented programs. 48. The mission feels that the design process should have included a comprehensive evaluation of the pilot phase. Supervision 49. Bank supervision missions were frequent, with staff from the Resident Mission and the Nairobi Mission often participating. Generally supervision was satisfactory, but the Bank staff should have recognized the imprecision in the method of recording of farmer contacts. This overstated the coverage of the extension service which was being monitored as a key performance indicator. There also should have been closer supervision of the M&E activities under the project. 33  Borrower Performance. 50. Generally, the performance of the Borrower has been satisfactory. The main covenants relating to timely implementation of the project were complied with although there was a delay at the outset. Legal covenants monitored by the Bank's supervision staff have been complied with. Accounting and reporting have been carried to a satisfactory standard through the Technical Assistance to the Department financed by the credit. The Government has also prepared its contribution to the ICR. 51. However, two covenants not monitored by the Bank in the Development Credit Agreement are worthy of mention. Section 3.04 required the Government to maintain the establishment of the extension service at the level agreed with the Bank as at September 1989, with staff leaving being replaced with diploma/certificate holders. The field staffing levels under T&V have a significant impact at the level of outreach achievable and, as mentioned above, substantial numbers of staff have been retrenched under the rationalization of MAC. Therefore, strictly speaking, this covenant has not been complied with. Indeed it was not possible for the Government to comply, given the impact of later Bank-supported adjustment operations on staffing levels within the public service. 52. There was also a covenant, Section 3.05, that concerned budgetary allocations to extension. The Government was to annually increase budgetary allocations over the level provided in 1987/88 fiscal year to cover annual costs as agreed in the annual work plan. This was not complied with (see discussion below). Major Factors Affecting the Project 53. The major factors affecting the project were: X Retrenchment of front-line staff, X (tGradual reduction of contact activity over time; X Limits on project funds; X Poor integration of agricultural and livestock skills at field level through cross-training; * New challenges, beyond technical matters, for which extension staff were fully prepared. Future Operations 54. A review of NALERP was undertaken in 1995 as part of the preparation for NAEP II which then addressed further modifications to the T&V approach that would finally be part of the NAEP II design. These were to (i) address issues in financial sustainability through collaboration with NGOs, and through the transfer of responsibility for commodity-focused farming systems to private agro-industries; (ii) farmer participation in the generation and dissemination of technology; and (iii) use of communication technology in extension. In addition, the research programs now being undertaken have incorporated farming systems research with environmental and production sustainability as key objectives. 55. The operations of NALERP have been taken over by NAEP 11, which will follow the T&V approach but will continue with the emphasis on group-based contact points. NAEP II is operational in the same 16 regions as NALERP and the IFAD Southern Highlands project in the remaining four regions of mainland Tanzania is expected to be extended and adopt similar alterations in emphasis. 34  Sustainability 56. The SAR considered sustainability of the project from two viewpoints - institutional and financial. 57. Institutional sustainability would depend on total commitment from all involved and a strong leadership within MAC. The mission considers that the project has achieved the anticipated management strength. 58. Financial sustainability was expected to be achieved through a combination of (i) cost reductions due to the merger of agriculture and livestock extension, and the unification of extension programs financed by different donors; and (ii) an expected 4% growth in real terms in budget allocations to extension. The SAR estimated that the Ministry "would easily be in a position to sustain activities under the project". The financing of the project was designed to place more demands on resources from within Government over the course of the project to ensure that the Government financed the recurrent costs of the project in full in its eighth year2. 59. The following table gives the Government Budget allocations to MAC. The figures indicate that the budget for agriculture increased until 1992 in real terms and as a percentage of Agricultural GDP over 1988 levels. It since has fallen to below 1988 levels in real terms and from 1.2% to 0.7% of Agricultural GDP. Table 2: Government Budget Allocations to Ministry of Agriculture Year Current % of Agrcultural Constant 1987 ending June TSh million GDP TSh million 1988 2,429 1.2% 2,429 1989 2,469 0.8% 1,678 1990 4,021 1.1% 2,425 1991 5,808 1.3% 3,010 1992 6,289 1.1% 2,855 1993 7,402 1.0% 2,765 1994 7,865 0.8% 2,310 1995 8,337 0.7% 2,053 1996 10,952 0.7% 2,164 60. The SAR recognized the danger of eroding budget allocations that would undermine the main project objectives of establishing a technically viable and financially sound extension system. The activities under NALERP will now be sustained under NAEP II which is also designed to increase Government commitment to extension by requiring increasing budget allocations. Follow-up 2Assurances to this effect were made by Government at credit negotiations. 35  61. On return to Rome, the mission will prepare a draft ICR to be forwarded to IDA by mid-November, 1997. This will include further analysis, including impact and cost-effectiveness of the Project and details of the lessons learned. Dar es Salaam 17th October, 1997 36  APPENDIX 2 TANZANIA NATIONAL AGRICULTURAL AND LIVESTOCK EXTENSION REHABILITATION PROJECT (CR. 1994-TZ) GOVERNMENT CONTRIBUTION TO THE ICR  THE UNITED REPUBLIC OF TANZANLA I MINISTRY OF AGRICULTURE & COOPERATIVES NAT1ONAL AGRICULTURAL & LIVESTOCK EXTENSION REHABILITATION PROJEICT j <~~~~~~.' NAL_EItI..1' IMPLEIVMENTATION COMPFLETIN' REPORT 1 ( ,k' : l' / ,:0 .  TABLE OF CONTENTS: 1.0 Introduction ............................ - > 2.0 Objectives and Summary of Major achievement 1........ 3.0 Organization and Managernent ................................. 3.1 Procurement ......................................... 10 3.2 Performance of Technical Assistance ... 1 1 3.3 Civil works ......................... ............. 14 4.0 Extension Methodology and Technology .................. 15 4.1 T&V Field Implementation . .- 15 4.2 The mid-term Review . .17 -4.3 T&V Tools and Pillars ..18 4.4 NALERP Final Review . .22 5.0 Staff and Farmers Training .- 24 6.0 Research Extension Farmers/Linkage ...1 6.1 Field Implementation and MTR findings 3-1 6.2 Indigenous Technical knowledue 32 6.3 Adoption/Observation Plots . . .32 6.4 Development of Impact Points . . .33 e.5 Field Days. Tours and Agricultural Shows .3 6.6 Professional Societies . . .34 6.7 NALERP Completion Review . . .34 7.0 Coordination of Extension Oriented Activities. . 8.0 Mionitoring & Evaluation. 3(3 9.0 Beneficiaries ....38 Table I: Key Performance Indicators. 8 Tables: Table 2: Procurement status as on March. 1997 13 Table 3: Status of Civil works at NALERP completion 14 Annexes: Annex I .43 Annex 1.. 4 Annex Ill .....5.. , , , ...... 45  1.0. Introduictioni Agricultural extension has undergo nemany changes overithe years since indeperndence. The most important change took place in 1972 when the development oriented functions of the government including extension were transfered to regional administrations. In 1983 the Ministry of Agriculture was re-centralized in the hope that its functions would be discharged more efficiently. However, actual implementation of extension services was below expectations. The extension services were deficient aid faced the following major _.. .problemnslweaknesses. , ,_ . . _, - * nadequate and fragmented organizational arrangement (e.g. crop and livestock divisions had parallel extension services); * Weak linkages with other knowledge systerns and inputs suppliers; * Lack of logistical support (e.g. shortage of transport, training, working gear. funds etc); and