Public Disclosure Authorized Public Disclosure Authorized Public Disclosure Authorized

Document of

The World Bank

Report No: 29424 IMPLEMENTATION COMPLETION REPORT (IDA-28990) ON A CREDIT IN THE AMOUNT OF SDR 21.5 MILLION (US$31.1 MILLION EQUIVALENT) TO THE UNITED REPUBLIC OF TANZANIA FOR A NATIONAL AGRICULTURAL EXTENSION PROJECT II

Public Disclosure Authorized

June 22, 2004

CURRENCY EQUIVALENTS (Exchange Rate Effective November 2003) Currency Unit = Tanzanian Shilling (Tsh.) 1 Tsh. = US$ .001 US$ 1 = 1,018 Tsh. FISCAL YEAR July 1 - June 30 ABBREVIATIONS AND ACRONYMS ASDP BEO BMW BTS CAS CBO DALDO DCA DESC DMT FAMO FEPU FFS FRG GOT GOZ ICB ICR IDA IFAD IPPM IT KATI LGA LITI M&E MAC MAFS MANREC MATI MCM MOU MTS MTR MWLD NAEPII NALERP NCB NGO PMU

Agriculture Sector Development Program Block Extension Officer (Zanzibar) Bi-Monthly Workshop Bi-Monthly Training Sessions Country Assistance Strategy Community-Based Organization District Agriculture & Livestock Development Officer Development Credit Agreement District Extension Steering Committee District Management Team Farmer Motivator Farmers Education and Publicity Unit Farmer Field School Farmer Research Group Government of the United Republic of Tanzania Revolutionary Government of Zanzibar International Competitive Bidding Implementation Completion Report International Development Association International Fund for Agriculture Development Integrated Pest and Production Management Information Technology Kizimbani Agriculture Training Institute (Zanzibar) Local Government Authority Livestock Training Institute Monitoring and Evaluation Ministry of Agriculture and Cooperatives Ministry of Agriculture and Food Security Ministry of Agriculture, Natural Resources, Environment and Cooperatives in Zanzibar Ministry of Agriculture Training Institute Ministry of Cooperatives and Marketing Memorandum of Understanding Monthly Training Sessions Mid-Term Review Ministry of Water and Livestock Development National Agriculture Extension Project Phase II National Agriculture and Livestock Extension Rehabilitation Project National Competitive Bidding Non-Government Organization Project Management Unit

PORALG PSC RALDO SAR SDR SHERFSP SMS SUA T&V VEO WEO WDC ZCC

President’s Office Regional Administration and Local Government Project Steering Committee Regional Agriculture and Livestock Development Officer Staff Appraisal Report Special Drawing Right Southern Highlands Extension and Rural Financial Services Project Subject Matter Specialist Sokoine University of Agriculture Training and Visit Village Extension Officer Ward Extension Officer Ward Development Committee Zonal Communication Centre

Vice President: Country Director Sector Manager Task Team Leader/Task Manager:

Callisto Madavo Judy O'Connor Karen Mcconnell Brooks Mohammed Taqi Sharif

TANZANIA NATIONAL AGRICULTURAL EXTENSION PROJECT II

CONTENTS

1. Project Data 2. Principal Performance Ratings 3. Assessment of Development Objective and Design, and of Quality at Entry 4. Achievement of Objective and Outputs 5. Major Factors Affecting Implementation and Outcome 6. Sustainability 7. Bank and Borrower Performance 8. Lessons Learned 9. Partner Comments 10. Additional Information Annex 1. Key Performance Indicators/Log Frame Matrix Annex 2. Project Costs and Financing Annex 3. Economic Costs and Benefits Annex 4. Bank Inputs Annex 5. Ratings for Achievement of Objectives/Outputs of Components Annex 6. Ratings of Bank and Borrower Performance Annex 7. List of Supporting Documents Annex 8. Institutional Framework and Operations of Extension Services Annex 9. Experience of NAEP II Pilot Initiatives Component Annex 10. Footnotes Annex 11. 2003 Beneficiary Assessment - Summary of Results Annex 12. Executive Summary of Government of Tanzania ICR Annex 13. Summary of Revolutionary Government of Zanzibar ICR

Page No. 1 1 2 6 11 13 14 16 17 18 19 22 24 26 27 28 29 30 36 40 41 43 54

Project ID: P002753 Team Leader: Mohammed Taqi Sharif ICR Type: Core ICR

Project Name: NAT EXT PROJ PH.II TL Unit: AFTS1 Report Date: June 23, 2004

1. Project Data Name: NAT EXT PROJ PH.II Country/Department: TANZANIA

L/C/TF Number: IDA-28990 Region: Africa Regional Office

Sector/subsector: Sub-national government administration (82%); Agricultural extension and research (18%) Theme: Rural services and infrastructure (P); Decentralization (S); Participation and civic engagement (S); Other public sector governance (S) KEY DATES PCD: 09/14/1995 Appraisal: 02/09/1996 Approval: 07/11/1996

Original Effective: 10/08/1996 MTR: 12/21/1998 Closing: 12/31/2001

Borrower/Implementing Agency: Other Partners: STAFF Vice President: Country Director: Sector Manager/Director: Team Leader at ICR: ICR Primary Author:

Revised/Actual 10/08/1996 12/12/1999 12/31/2003

GOVT/MIN OF FINANCE

Current Callisto E. Madavo Judy M. O'Connor Karen Mcconnell Brooks Mohammed Taqi Sharif David Ivory (FAO-CP) and Malathi Jayawickrama

At Appraisal V.E. Kimberley Jaycox James W. Adams Sushma Ganguly Satish Kumar

2. Principal Performance Ratings (HS=Highly Satisfactory, S=Satisfactory, U=Unsatisfactory, HL=Highly Likely, L=Likely, UN=Unlikely, HUN=Highly Unlikely, HU=Highly Unsatisfactory, H=High, SU=Substantial, M=Modest, N=Negligible)

Outcome:

S

Sustainability:

L

Institutional Development Impact:

SU

Bank Performance:

U

Borrower Performance:

U

QAG (if available) Quality at Entry: Project at Risk at Any Time: Yes

ICR U

Bank Performance and Borrower Performance are considered 'marginally' unsatisfactory.

3. Assessment of Development Objective and Design, and of Quality at Entry 3.1 Original Objective:

Context. Agriculture accounts for about half of Tanzania’s GDP and exports, and hence its performance has a significant effect on output, income and poverty levels. Over the 1990s, agricultural trend growth was 3.6 percent, compared to average annual agricultural growth rates of 2.9 and 2.1 percent in the 1970s and 1980s (Tanzania, Accelerating Agricultural Growth and Poverty Reduction, Background CAS Note, February 2004). Agricultural exports grew at an annual rate of over 7 percent per year, although this rate has slowed in recent years due to declining world market prices. Food crop production has grown at about the rate of population growth (2 percent) and accounts for roughly 65 percent of agricultural GDP, with cash crops accounting for 10 percent. Tanzania has made significant progress towards the Government's target of sustained annual agricultural growth of 5 percent. The five-year moving average agricultural GDP growth rates were about 3.3 percent from 1991-2000, and 4.3 percent over 1999-2003. Five Year Average Agricultural GDP Growth Rates (%)

Agricultural GDP Growth Rate (%)

5.0

4.5

4.0

3.5

3.0

2.5

2.0 1991-5

1992-6

1993-7

1994-8

1995-9

1996-2000

1997-2001

1998-2002

1999-2003

Five Year Moving Average

Recent improvements in food crops have been attributed to technological change. NAEPII supported advances in several of these crops--maize, the most important (accounting for over 20 percent of total agricultural GDP) followed by rice/paddy, beans, cassava, sorghum, and wheat. Adoption of improved varieties, especially in maize, and better land management practices have increased overall yields substantially in recent years, to record annual average growth rates of around 4 percent for food crops. Yet, Tanzania's productivity levels for a variety of agricultural crops are much lower than those observed in other countries (Concept Note for the Tanzania Country Economic Memorandum, draft, June 10, 2004). An International Food Policy Research Institute review in 2000 showed that Tanzania has a strong comparative advantage in maize and rice/paddy, and all the traditional export crops (cashew, coffee, cotton, tea, and tobacco). Improved productivity, through both technical efficiency and technological change, will play a major role in increasing Tanzania’s future agricultural output and income, and reducing poverty. NAEPII was prepared and implemented in order to contribute to the ongoing process of agricultural growth.

-2-

Objectives. The project’s objective was to "continue to improve the delivery of extension services to smallholder farmers for increasing their incomes and productivity, while improving its relevance, sustainability and cost effectiveness" (SAR, June 5, 1996). NAEPII sought to achieve this by continuing the essential elements of the training and visit system (T&V) of agricultural extension introduced in earlier projects, but with emphasis on a participatory farmer group approach, supported by more knowledgeable extension staff and farmers, improved systems of information transfer to farmers, and pilot initiatives to enhance technology transfer, extension management, input supplies and farmer empowerment.1/ NAEPII’s design aimed to: (i) support efforts to reshape the role of the Ministry of Agriculture and Cooperatives (MAC), limiting its core functions at central level to policy formulation, regulation, extension and research, and to strengthen its extension service at the field level; (ii) increase participation and enhance farmers’ role in the generation and dissemination of technology; (iii) develop the human resource base through the training of farmers and extension staff; and (iv) increase agricultural growth and smallholder income. These objectives reflected the Government’s priorities, articulated in the National Agriculture and Livestock Policy, and the World Bank’s Country Assistance Strategy (CAS). Accelerating agricultural growth, with emphasis on agricultural research and extension, livestock health and natural resources management, is at the core of the Bank’s strategy to support growth through poverty reduction. 3.2 Revised Objective:

The Regional Administration and Local Government Act of July 1997 was enacted a year after project approval. Implementation responsibility for extension service delivery was transferred from MAC to the Local Authorities (LAs) covering regional administrations, district, municipal and urban councils. This resulted in the establishment of 119 autonomous district extension services and a reduced role for MAC. Following the Mid-Term Review (MTR) in December 1999, the Bank and the Government reached agreement on the need to amend the project’s objectives in line with the Government’s decentralization strategy. The development objective was revised to "continue to improve the delivery of extension services to smallholder farmers in the Borrowers territories by focusing on the enhancement of the technical and administrative capacity of the LGAs to provide such extension services”. This significantly changed the emphasis of the original objective, from developing an extension system that would directly result in increasing agricultural output and incomes of smallholders, to creating an institutional arrangement that would better facilitate extension services. MAC was subsequently restructured (in 2000) to form three separate ministries--Ministry of Agriculture and Food Security (MAFS), Ministry of Water and Livestock Development (MWLD), and the Ministry of Cooperatives and Marketing (MCM). In general, decentralization was viewed by the Bank as a positive move. The restructuring of MAC, however, was unhelpful to support the decentralization process, as central responsibility for, and contribution to, agriculture and livestock extension services was further split. The Development Credit Agreement (DCA) was amended on several occassions.2/ The project sought to achieve the revised objective through: (i) developing a Vision/Strategic Framework for Extension for the next decade, consistent with the changed institutional landscape; (ii) identifying the needs of selected LGAs (limited in geographic coverage) and building their administrative and technical capacity to effectively deliver extension services; (iii) enhancing training for staff of these LGAs, and farmers, on their roles in a participatory, demand-driven and decentralized extension delivery system; (iv) developing a participatory, demand-driven extension system that would particularly respond to smallholder needs through a variety of pilot initiatives; (v) ensuring the development of extension and technological packages appropriate to poor farmers and women; and (vi) strengthening and enhancing extension support for input supply and marketing, and organizing farmers to address these issues.

-3-

Original and Revised Project Scope. The original project focused on strengthening extension services of MAC in the 16 regions supported under the National Agriculture and Livestock Extension Rehabilitation Project (NALERP). The four remaining regions in mainland Tanzania were covered by the IFAD-supported Southern Highlands Extension and Rural Financial Services Project (SHERFSP). Following completion of SHERFSP in August 1999, NAEPII’s geographic scope was expanded to cover all 20 regions of mainland Tanzania. At restructuring in early 2000, the proposed focus of NAEPII was only on 56 districts (out of 119 districts in the 20 regions of mainland Tanzania) identified for Agriculture Extension Reform in the remaining time to project completion. In May 2000, the geographic scope was further expanded to include the islands of Zanzibar, to enable the Revolutionary Government of Zanzibar (GOZ) to complete critical activities left unfinished by the phased-out IFAD-funded Smallholder Support Project (1991-1997). 3.3 Original Components:

The original project had four components. (a) Institutional Strengthening (US$15.7 million). This component sought to continue reorganizing and strengthening the extension services of MAC headquarters and field services to strengthen districts’ roles and improve farmer linkages through: (i) increasing the number of full-time Subject Matter Specialists (SMSs) and deploying them to district and lower levels; (ii) establishing a Project Steering Committee (PSC) and District Extension Steering Committees (DESCs); (iii) retrenching unsuitable Village Extension Officers (VEOs); (iv) establishing an M&E system; and (v) training M&E staff. The Project also planned to provide vehicles, equipment and furniture, rehabilitation of some field offices, construction of a limited number of residential houses, and cover operating costs of vehicles, field supervision and maintenance. (b) Extension Education and Training (US$12.6 million). This component aimed to enhance the effectiveness of technology transfer, create more effective farmer-extension-research linkages, develop appropriate support to farming, and address gender issues in agriculture. NAEPII aimed to achieve this through: (i) Staff Training, including formal education at degree and diploma level, short courses, refresher courses, Bi-monthly Workshops (BMW) and Monthly Training Sessions (MTS) for field extension staff, and local and overseas study tours and training; (ii) Farmer Training, including residential courses, on-site training, farmer study tours, demonstrations and provision of extension materials for farmers; and (iii) Training Support, including establishing a Training Coordination Committee (TCC) to finalize annual training plans, review training curricula and progress, facilitate networking between training institutes, and give directions for overall management of training support. The Project planned to finance training costs, a needs assessment study, M&E operations, and training institutes, covering teaching aids, farm and office equipment, audio-visual aids, and civil works for rehabilitating selected facilities. (c) Communications Support (US$2.9 million). This component supported the government’s strategy to decentralize information management to the zonal level, to enable an effective response to area-specific information needs of local farming communities, through: (i) establishing six Zonal Communication Centers (ZCCs); (ii) reorganizing the MAC Field Extension and Publicity Unit (FEPU) in Dar es Salaam; (iii) a needs assessment study for communications development; (iv) overseas and local training for selected staff; and (v) technical assistance to train core ZCC staff. NAEPII planned to finance civil works, communication support equipment, vehicles, materials production, and field visits. (d) Pilot Initiatives (US$1.7 million). The Pilots attempted to improve extension management, technology transfer, farmers-extension-research linkages, participation of other providers in extension delivery, seeds

-4-

production and input supply, farmer empowerment and gender impact. A Government task force identified nine pilot initiatives at project appraisal, with additional initiatives to be identified during implementation. 3.4 Revised Components:

The restructured project retained the four original components. However, it reoriented activities and reallocated budgets between components, to emphasize capacity building of extension services in the LGAs under the first component (additional US$6.5 million allocated), and reduce support to extension education and training, and communications, under the second and third components, respectively. Pilot initiatives were expanded (an additional US$5.0 million). These changes applied to the 20 regions in mainland Tanzania. In Zanzibar, the change in project scope to include strengthening GOZ institutions required the allocation of a separate budget (US$2.5 million) from the 'unallocated' category of NAEPII. GOZ's counterpart fund contribution was US$100,000. While Zanzibar was not considered an additional component in the project DCA, it was considered separately, for both convenience and political reasons, as it is in this ICR. This component supported adaptive research on selected crops, improving human resource capacity of extension staff and farmers, developing a more participatory extension system and improving institutional facilities.

3.5 Quality at Entry:

The project was rightly predicated on the need to continue support for agricultural extension services initiated under NALERP, but it did not internalize major policy and strategy changes regarding decentralization. Although the SAR mentions that "Government initiatives under the civil service reform program are aimed at decentralizing responsibilities and functions to the districts" (SAR, page 8), the project was designed on the assumption that MAC would continue to control extension services from the center. Given what was known at the time, the project design should have: (i) embraced decentralization from the outset and been more proactive in the dialogue on decentralization; and (ii) incorporated some flexibility in terms of institutional arrangements and the flow of funds. Consequently, the project design became rapidly obsolete and required major restructuring after one year of implementation. The project design exhibits an element of internal contradiction, retaining overall commitment to the T&V system, but including features of participation and demand responsiveness suggesting doubts as to the soundness of the T&V approach. During the period of project preparation, considerable debate was occurring throughout the Bank and the developing world on the merits of T&V, in terms of impact, cost effectiveness and sustainability.3/ During the early period of implementation of the project, the debate was largely resolved through the completion of evaluative work by OED, documenting high cost and low sustainability of the T&V system. 4/ At the time of project design, however, senior and highly-placed advocates of the T&V system discouraged project teams from rejecting T&V in favor of less centralized and more demand-responsive approaches. Although the project files do not contain documentary evidence of the debate with regard to this particular project, it is likely that the overall intellectual climate within the Africa Region with regard to agricultural extension influenced project design, and led to retention of design elements that were subsequently dropped during implementation. The SAR does not include Key Performance Indicators (KPIs). The Memorandum of the President (MOP) has an abbreviated and incomplete set of KPIs. A well-defined set of KPIs should have been set out at preparation in order to guide the surveys conducted under NAEPII, especially the baseline and final assessments, in gathering information needed to monitor project outcomes. At the time NAEPII was approved, NALERP had been under implementation for eight years (although the ICR for NALERP was completed only in May 26, 1998). The Bank had ample opportunity to incorporate the lessons emerging

-5-

under the previous project. These lessons included: the costly aspects of the T&V system; and the importance of clearly identifying meaningful performance indicators. These lessons were not reflected in the design of NAEPII. Thus, overall quality at entry is considered unsatisfactory.

4. Achievement of Objective and Outputs 4.1 Outcome/achievement of objective:

Did NAEPII satisfactorily achieve its original objective of improving extension delivery to smallholder farmers for increasing their productivity and incomes? A comprehensive assessment of project outcome is difficult for several reasons including: (i) the weakness in KPIs noted above; (ii) the complexity in separating the effects of NAEPII and other variables on farmer productivity and incomes; and (iii) the lack of comparable information on the farmers' situation before and after the project in the four assessment surveys conducted under NAEPII -- Baseline survey (1996), Beneficiary survey (1999), Household asset surveys (2001) and the Final Beneficiary Assessment study (2003). These surveys evaluate different populations of farmers, districts and villages in successive rounds, and lack the information needed to evaluate outcomes in relation to objectives. Also, 2002-03 was a drought period, affecting overall national yields and the Final Beneficiary Assessment data. Another factor was the change in project objectives upon restructuring, which shifted emphasis to monitoring institutional capacity development rather than changes in agricultural production in the latter years. Given these limitations, a table of the main output and outcome indicators has been constructed in Annex 1, and gives an indication of project achievements. Almost without exception, the adopted technological packages as piloted by the micro projects led to more than doubling farm crop and livestock yields. The ultimate impact on incomes, however, depended on the market situations. In most cases, notably for seasonal and perishable goods (e.g. milk, vegetables and fruits), increased production resulted in lower producer prices. Nevertheless, the reduced prices were more than compensated by the increased volume, therefore leading to improved incomes. This emphasizes the need for extension agents to be able to advise on marketing as well as on production. The quality of inputs also affected farmer incomes (see Lessons from Experiences of NAEPII Micro Projects, and Beneficiary Assessment--Annexes 9 and 11 for details). What is not in dispute is the fact that where farmers opted to use improved technological packages, they experienced higher yields and better incomes, despite the setbacks mentioned above. Farmers gave evidence of how the increased disposable income had helped them to: Invest in improved shelter; • Invest in improved/local livestock and poultry; • Buy assets such as transistor radios, bicycles, ox-carts and ox-ploughs; • Afford to cater for school children needs (i.e.uniform, fees and books); • Pay for better medical services; • Get better balanced diets, thus reducing chances of illness in the family; • Experience fewer months of food deficit than before; and • Reduce debt levels and therefore lead a freer life. • Additionally some enterprises taken by women helped enhance their financial base and plan more effectively to improve their family livelihoods. Mainland Tanzania: Overall, the data suggest that the project enhanced delivery of extension services to participating farmers. 89% of the 388 farmers surveyed in 2003 were aware of local availability of extension services, 83% had obtained extension advice from VEOs/WEOs, 52% had received advice within the last month, and 77% rated the advice as excellent. Farmer adoption of extension advice (timely land

-6-

preparation and planting, use of improved seeds and proper spacing, manure application, weeding), and improved biological/environmental practices (post-harvest care, pesticide use) improved substantially between 1996 and 2003. There was wide variation, however, in the adoption rates of various technologies in cereal production. Although the awareness levels of improved practices were high, farmers were slow in adopting them as some technologies were not sufficiently financially attractive. The achievements under formal (24% of all field extension staff) and informal training (all staff received various forms of training) of extension staff and farmers (90,000 received some form of training) would also indicate significant capacity development of human resources. In the process of decentralization of extension services to LGAs and the emphasis given to develop a more demand-driven system, the project appears to have contributed to a more participatory agricultural extension service, where 45% of farmers were involved in training of fellow farmers, and 45% wanted this to continue. For Zanzibar, outcome assessments indicate that significant progress was made in building institutional capacity (improved facilities and mobility of extension staff, short course training for extension staff), empowering farmers (participatory training to identify and solve problems), and disseminating new agricultural technologies (65% of surveyed farmers adopted extension messages, and 62-74% of farmers adopted improved varieties of cassava, yams and sweet potatoes resulting in significant yield increases). The approach to supervision of NAEPII strengthened linkages with other donors, and between research and extension. Some joint missions were maintained with the Tanzania Agricultural Research Project II (TARP II). These were important in forging closer research-extension linkages (particularly after extension responsibility was removed from MAC). Joint contributions were established under some pilot initiatives, whereby other donors (i.e. FAO and DANIDA), local church organizations, NGOs and CBOs were involved in participatory extension activities. Based on the improved capacity to deliver extension services through training, and the adoption of extension advice and the resulting yield increases in the pilot areas, the project is rated satisfactory in meeting its original objectives. In addition to productivity improvements, the project also contributed to establishing a solid foundation on which to build an integrated approach to generation and dissemination of agricultural technology in Tanzania, and the foundation is supporting further work within the context of the Agricultural Sector Development Program (ASDP). Despite these obvious successes experienced by farmers engaged in the pilot initiatives, however, there remains a challenge to ensure diffusion of knowledge to the rest of the communities. This challenge is more to do with actual adoption rather than knowledge dissemination. Widespread adoption was limited due to other factors such as small market size, lack of input supplies, limited local marketing capacity, and lack of cash and credit. It is therefore important to address barriers to enhancing technological adoptions. These are mostly related to credit availability, marketing system, storage and transport infrastructure, all of which are addressed in the ASDP under preparation and Participatory Agricultural Development and Empowerment Project (PADEP). 4.2 Outputs by components:

Component 1: Institutional Strengthening (US$15.7 million) As part of the plan to reshape the role of the MAC, extension staff numbers, positions and locations were rationalized. About 900 of the 4,725 underqualified VEOs were retrenched under termination agreements, and many Subject Matter Specialists (SMSs) redeployed from MAC and the regional offices to districts. The transfer of SMSs, over a 3-year period, was slower than expected. An overall reduction in total extension staff occurred as the retrenched staff were not replaced, as planned. The project achieved most of the physical targets set for improving institutional facilities and purchase of equipment, vehicles, motorcycles and bicycles for district extension staff, with some targets being exceeded (motorcycles and

-7-

computers). The initial setup for administration and management of extension services was based on a line of command from MAC through regional offices, to extension staff deployed at district level. However, with devolvement of extension services to LGAs, overall responsibility for coordinating project implementation moved from MAC to President’s Office, Regional Administration and Local Government (PORALG), previously called Ministry of Regional Administration and Local Government. Extension staff were then placed under the administrative and management control of the district councils (DCs). The DCs, however, were not prepared for this new arrangement, and there was considerable delay in establishing a functional institutional framework for local extension services. At project completion, the institutional framework for extension services had been reasonably well established in the LGAs, as the two-year extension gave sufficient time for institutional development. A total of 5,835 extension workers are now deployed at both district (1,110) and at village levels (4,725) to cover about 10,470 villages, and management and operational structures have been established in the LGAs. However, staff employed occupy about 50% of the designated positions, based on the GOT plan to have one extension worker per village. As discussed in Annex 8, several factors continue to hinder the development of fully effective and efficient local extension services. These include: the continued staff deployment within districts according to the T&V system; the ineffectiveness of having a SMS cadre in district headquarters; lack of upgrading of educational skills of many staff; inadequacies in transport availability; and low staff morale linked to lack of promotional opportunities. A GOT Task Force on Agriculture Extension Reform prepared the draft Vision for Extension for 2010, which was discussed in two workshops in November 2000 and March 2001. The final report was distributed to key zonal and district stakeholders in June 2001. By project end, however, it was apparent that the full acceptance and absorption of the elements of, and implementation strategy for, this vision into the district, ward and village level had not been achieved. Although an M&E unit and some M&E capacity had been established under NALERP, it was not fully functional. NAEPII has supported the strengthening of the M&E Unit in MAFS by providing technical assistance, computers and motorcycles, staff training, and operational funds. The Unit has national responsibility for monitoring and evaluating extension performance, and has undertaken regular information collection on project implementation and farmer output, and the production of impact assessments of extension. However, local data collection was hindered as M&E units had not been established in the LGAs. Also, the limitations of the KPIs restricted the development of an effective M&E system. Establishing functional M&E units at the district level, developing an appropriate framework of indicators and M&E operations, and linking the district M&E units to the national M&E Unit remains a challenge. The project established a Zonal Research and Extension Liaison Office to improve communication linkages between research and extension. Functioning farmer-research-extension linkages have also been obtained through the establishment of joint zonal technical committees, regional workshops, agricultural shows, and formation of farmer groups. At the field level, this is attained through joint participation of researchers, extension agents and farmers in farmer field schools (FFS), on-farm research, and pilot initiatives. However, the separation of research (controlled by a national-level agency) from extension (in LGAs) has created a need to strengthen units and groups, such as the zonal research-extension liaison office, regional agriculture and livestock advisors, and ZCCs, to maintain effective research-extension linkages. Component 2: Extension Education and Training (US$12.6 million) Although the project achieved significant upgrading of educational levels and technical knowledge of extension staff, through the graduate (17) and post-graduate (54) training program for selected SMSs at Sokoine University of Agriculture, it did not fully achieve the targets, partly as a result of project -8-

restructuring. The plan to upgrade WEO and VEO certificate holders with a one-year upgrade course fell short of the initial targets (650 out of 800) due to reallocation of training funds to component 4. However, it benefited from the development of a new diploma curriculum and introduction of two-year diploma level (405) training at Ministry of Agriculture Training Institutes (MATIs) and Livestock Training Institutes (LITIs). The Government has indicated their plan to continue support for formal training to enable all staff to be better trained. Extension staff have also benefited greatly from a range of shorter-term formal and informal training. About 1,095 staff received informal training through study tours (special tours for VEOs and inter-regional and international tours). In addition, on-the-job training covered upgrading technical and practical skills (group working, participatory extension, communication and facilitation development). There remains a continuing need for technical training of extension staff. Regular technical training programs (BMWs and MTSs) halted part-way through the project, leaving a void in the transfer of technical information to WEOs and VEOs. Extension newsletters, started under NAEPII and discontinued due to lack of funds, also need to be reactivated. Regular specific needs-based and seasonal technical training for WEOs and VEOs needs to be introduced. Outcome indicators would suggest that the knowledge and skills of farmers have improved considerably. Large numbers of farmers benefited from residential (4,332) and on-site (37,179) training and from study tours (1,417). Extension messages and recommended technologies were also promoted to farmers through demonstrations (140,521) and media materials (print, radio and video) produced under the third component. Training institutions (Sokoine University of Agriculture, eight MATIs, four LITIs, the Kizambani Agriculture Training Institute--Zanzibar, and several Farmer Training Schools) received assistance to improve their capacity to provide appropriate training, through purchase of training aids and equipment and from rehabilitation of facilities. Component 3: Communications Support (US$2.9 million) The Farmers Education and Publicity Unit (FEPU) in Dar es Salaam was strengthened through: the purchase of equipment to improve the production of print media and in-house production of video and radio media; and the purchase of two media vans, public address systems and one vehicle, to enhance their capacity to disseminate information. While the target was also to strengthen six out of the seven ZCCs throughout the mainland (a ZCC was established for the southern highlands zone under SHERFSP), only two centers (Arusha and Morogoro) received significant assistance under NAEPII. This was a consequence of the agreement to place the ZCCs under the directors of the Zonal Research Stations as part of activities under the TARP II. Close collaboration between the FEPU and ZCCs was achieved in the collection of information and production of media materials. Numerous booklets, newsletters, brochures, radio programs (703) and video programs (2-3 per year) were produced. Three staff received specialized overseas training. However, a new era in information and IT transfer systems has made the FEPU video and audio production systems rather obsolete. Given the importance of communication of extension information in a decentralized extension service, where the ratio of clients to providers is high, there is a need to further upgrade and develop communications support at national and zonal level, develop IT networks to district offices, and establish the capacity of LGAs to reproduce local media materials for farmers. Component 4: Pilot Initiatives (US$1.7 million) A total of 35 pilot initiatives were conducted, to test ways of improving: (i) extension management and enhancing technology transfers and adoptions; (ii) farmer-extension-research linkages; (iii) the role of the private sector, CBOs and NGOs; and (iv) empowerment of farmers, including women. This component successfully achieved improvements in all the above aspects, and significantly increased agricultural production and incomes for participating farmers as noted above (and in Annex 9). Five initiatives piloted alternative extension management methods with CBOs/NGOs providing various extension services, -9-

using both farmer motivators (FAMOs) and extension workers, and the FFS approach as an alternative delivery system for technology transfer. These pilots demonstrated that a pluralistic approach to providing extension works well when public or alternative funds are available to NGOs/CBOs. Beyond the pilot initiatives, the FFS approach was piloted in 16 districts, where about 160 field schools were established, with another 200 FFSs planned for another 36 districts. Various stakeholders regard the FFS approach as a cost-effective method of extension, as it combines the virtues of farmer participation, responding to farmer demands, involving farmers in financial contribution, as well as creating an ideal link between farmers, extension and research. Some pilot initiatives encouraged the private sector to provide farm input supplies. The initial approach for an input supply guarantee scheme was discontinued after realizing it was not conducive to the spontaneous growth of a competitive market/trade. Instead, the project facilitated only those schemes that supported capacity building for the operators to use their own resources and provide credit at market rates. Farmers, individually and in groups/societies/cooperatives, were also successfully involved in the multiplication of foundation seeds for subsequent distribution and sale to other farmers. One initiative demonstrated that it was feasible for a grassroots institution to combine extension advice/training, and also provide farm inputs to their members. Several pilot initiatives transferred workable technological packages to farmers, such as mechanized farming (ox-drawn implements and power tillers), irrigation farming, water harvesting/dams, horticulture farming and animal health. Generally, these initiatives significantly increased crop and livestock production and net farm incomes. These results reinforce the importance of farmer group dynamics, integrated approaches in introducing new technologies, access to credit and savings mobilization, the timely supply of quality inputs, and assured markets for farm produce. Zanzibar Component Although project implementation in Zanzibar was not officially considered as a separate project component, the management, budget allocation, budget expenditure and project activities were autonomous under the GOZ. As noted above, this component achieved significant progress in building institutional capacity, empowering farmers, and disseminating new technologies. The civil works were achieved as planned. This included construction of seven residential staff houses, a student dormitory, the establishment of training farms, and procurement of vehicles and equipment. A total of 144 staff were trained in various disciplines. A total of 11,082 farmers were trained through well-established FFSs, Farmers Research Groups, field days, study tours, on-farm demonstrations and the provision of media materials. The project also supported crop research programs in banana, root crops and legumes. These initiatives increased crop production and farm incomes. The Zanzibar component also helped forge a closer working relationship between the agricultural staff of GOT and GOZ. 4.3 Net Present Value/Economic rate of return:

Extension is greatly concerned with the adoption of technology. Although many studies conduct ex-post evaluations of extension, they shed little light on the net benefits of extension. Research shows that even in ex-post evaluations, it requires a comparative approach along two dimensions -- 'before and after' the project, and 'with and without' the project. Most studies of developing countries evaluate extension through a simple before and after comparison of crop yields. They do not, however, evaluate the contribution of extension to increases in output, as it is almost impossible to separate the effects of other variables such as material inputs, soil quality and supply constraints that are likely to influence adoption and output. Since farmers are slow to adopt new technologies (compared to their awareness levels), increases in adoption rates observed during NAEPII may well have been a result of extension services provided under NALERP. Benefits of the institutional and training components are also difficult to quantify in monetary terms. The approach adopted in the SAR was to identify the "minimum incremental benefit stream" that

- 10 -

would justify the investment the government extension service, and then under an "incremental adoption" model, see whether the project would, under a reasonable set of assumptions, produce the desired minimum benefit stream. This ICR follows the incremental adoption analysis to recalculate the Economic Rate of Return (ERR) over a 20-year period from 1996, using actual/observed yields and area increases, prices and costs from 1996-2003 (Annex 3). If 100% of the observed yield increase (on the part of the 11% of farmers assumed influenced by the project) for each crop is attributed to the project, the benefit-cost ratio (B/C) at a discount rate of 12% is 1.529, and the internal rate of return (IRR) is 23%. If, however, we assume that NAEPII contributes to only 50% of the yield increase under each crop, then B/C and IRR decrease to 0.765 and 6% respectively. Thus, the IRR can be assumed to lie within the interval bounded by 6% and 23%. The Final Beneficiary Assessment attempts to capture, through proxies, the effects of NAEPII on variables that are difficult to quantify in monetary terms, and are excluded in the ERR calculation. These include the quality of housing, food security, family health and the ability to pay for medical care, and household expenditure levels. Overall, there have been some positive changes in indicators related to quality of housing, ownership of capital-intensive assets, and expenditure levels on household basic needs (Annex 11). 4.4 Financial rate of return:

Financial rate of return is not calculated separately. 4.5 Institutional development impact:

Given the Government's decision to decentralize, the ICR concludes that the project contributed toward establishing a working decentralized extension service, although the process of institutional change and strengthening is still under way. Institutionally, a functioning extension service has been established in all districts of mainland Tanzania, but appears not yet to have achieved full functionality due to delays in effectively transferring administrative (and hierarchical) responsibility from MAC (and thereafter 3 ministries) to the LGAs. In addition, the coordination and linkages between district extension services are weak. While there remains a need to improve the quality of extension services in the LGAs, staff are better trained, both in terms of their technical knowledge and their understanding and ability to undertake/provide a more farmer-responsive form of extension service. The decentralization process cast doubt on the appropriateness of the staffing and operational structure under T&V--its costly and hierarchical system of staff deployment and management, particularly given the small staffing cadre of each district service, and the anticipated influence of the private sector. ASDP, through its sector-wide framework for overseeing institutional reforms and investment priorities, is addressing these concerns and taking reforms a step further, through support at local and national level, and for cross-cutting and cross-sectoral issues. Annex 8 gives a comprehensive assessment of the project’s institutional development impact. Farmers are also more knowledgeable and have improved their abilities to analyze their farming situation in a participatory manner, and be more self-reliant in developing and improving agricultural production. Although only tested to a limited extent under the project, it would seem that a more pluralistic system of extension provision is possible, and farmers can become effective advocates for their own needs with regard to advisory services.

5. Major Factors Affecting Implementation and Outcome 5.1 Factors outside the control of government or implementing agency:

A significant outside factor that affected project outcome was the severe deterioration in the terms of trade

- 11 -

for agriculture, particularly export crops. Coffee, cashew, cotton, tobacco and tea account for 80 percent of Tanzania's agricultural export revenues. Export unit values fell sharply since 1995, especially for cotton and coffee. Export earnings of all crops, which increased from the early- to the mid-1990s, due to the initial impact of marketing liberalization and rising international prices, declined after 1995, partly due to falling international commodity prices (Tanzania's Coffee Sector, June 2003). 5.2 Factors generally subject to government control:

The devolution of extension services to the LGAs within one year of project start-up necessitated restructuring of the project, but had a positive impact on the performance of extension. The institutional restructuring of central line ministries responsible for agricultural extension services likewise necessitated changes in the project, but did not contribute on balance to more effective management of extension or of the sector. The latter further divided and added complexity to extension service support for agriculture and livestock. There also appeared to be a general lack of support given, or urgency shown, by the Government to complete the transfer and establishment of a fully-operational devolved extension service. Additionally, the Government did not release counterpart funds to the project from mid-1999 to end-2000. Another factor that affected project outcome was the ban imposed by the GOT (up to 2002), for national food security reasons, on the export of foodgrains to neighboring countries experiencing periodic large food deficits. This denied farmers the opportunity of selling agricultural produce at favorable prices. 5.3 Factors generally subject to implementing agency control:

The original institutional arrangement for implementing the project was through a PSC placed within MAC and DESCs in each district. This operated until 1999/2000. After restructuring, it was planned that the roles of the DESCs would be taken over by District Management Teams (DMTs) and Ward Development Committees (WDCs) under the Local Government Reform Program. It is apparent that the DMTs, which were responsible for all sector activities (and not only agriculture), have not yet effectively assumed the roles of the DESCs, and therefore left a vacuum in adequate provision of management guidance and oversight to project implementation in the districts. With devolvement of responsibility for extension services, MAC no longer had administrative control of field extension staff and their activities. At the same time, the LGAs had no institutional arrangement in place to assume responsibility for extension services or to implement NAEPII. Significant delays in setting up the necessary institutional arrangements for the decentralized operation of extension services by the LGAs (80 LGAs at that time), led to restricted or even stalled project activities and problems with project fund disbursements. Furthermore, although the original project design was to transfer all SMSs to field positions, slow transfers of staff (mostly due to personal reasons) and low staff morale within this cadre, also contributed adversely to project implementation. Eventually (November 1998), MAC provided detailed guidelines for management of agricultural extension for the district and municipal councils, but at this time MAC could neither issue administrative instructions to the DCs to implement NAEPII, nor establish financial arrangements for funding the DCs for implementing the project. The key factors that allowed the project to achieve several of the revised targets and meet objectives were that the responsibility for many of the activities (education, training, communications and M&E) remained with the PSC and PMU (later to become a Project Coordinating Unit) in MAFS. Over the extended life of the project, as administrative authority decentralized, the project was able to continue to work through MAFS to provide capacity building and training for agents in the districts. In addition, the pilots became relatively more important after the restructuring and contributed to meeting the objectives. The inclusion of the Zanzibar component led to the establishment of a separate steering committee (Project

- 12 -

Technical Committee) and Project Management Unit (PMU) in MANREC, without disruption or adverse effects on the institutional changes on the mainland and restructuring of project objectives. 5.4 Costs and financing:

Project costs. The total project cost, including contingencies, was estimated at US$32.94 million (Tsh. 22.07 billion equivalent) over five years. Foreign costs were estimated at US$10.7 million. The project was effective on October 8, 1996, with closure originally scheduled for December 31, 2001. The project closed on December 31, 2003, after two one-year extensions. Project disbursements continued until end-April 2004. Actual total project cost over approximately 7 years was US$30.61 million, representing 92.9% of the original total cost estimate. Although 98.4 percent of the Credit was expended, the lower-than-predicted total cost is due to changes in the exchange rate between the SDR and US dollar. Project Financing. An IDA Credit of SDR21.5 million (US$31.1 million equivalent) was approved on July 11, 1996, with counterpart funds of US$1.8 million to be provided by GOT. With the introduction of the Zanzibar component, a further US$0.1 million was to be provided by GOZ. IDA credit disbursements totaled SDR21.15 million (US$28.24 million equivalent) on April 30, 2004, the last date for approval of credit disbursement, leaving SDR350,415.0 undisbursed. These funds will be cancelled. Total counterpart contributions were US$2.25 million and US$0.12 million from GOT and GOZ respectively.

6. Sustainability 6.1 Rationale for sustainability rating:

Institutional sustainability. The LGAs appreciate the importance of agricultural development to the rural and national economy, and in discussions with executive directors, all expressed continuing financial and operational support for agricultural extension services from their own budget resources.5/ According to the Local Government Act, extension is a mandatory function of the LGAs. In addition, GOT has verbally assured its commitment to completing the training program for extension staff (as about 1,140 certificate and 2,000 diploma holders did not receive training under NAEPII), so that eventually the majority of staff will receive diploma or degree level training. The line ministries (MAFS and MWLD) are also committed to support extension activities through provision of development budget funding to the LGAs. While these ministries are institutionally separated from the decentralized agricultural extension services, they can exert a significant influence on agricultural development priorities and activities for districts and regions, through development budget allocations to the LGAs. There is reasonable assurance that the decentralized institutional arrangement and operation of extension services supported under NAEPII will be sustained, since the ASDP incorporates continued commitment to decentralized and demand-driven extension services. The Zanzibar component supported and strengthened an existing integrated research and extension system, and therefore remains sustainable. Representatives from Zanzibar are participating in design of the ASDP.

Extension system sustainability. The sustainability of the extension system also seems likely. While there are measures that need to be taken to improve management and operational functionality (Annex 8), the more demand-driven system of extension achieved, whereby farmers are trained and take leadership in defining extension needs and promotion to fellow farmers, is expected to continue. The greater involvement of other extension providers (CBOs/NGOs and private sector), although still embryonic, appears likely to be more widely adopted. The expectation is that farmers will become more self-sufficient in financing extension services. Because the majority of farmers still practice subsistence agriculture, there remains a great need for the public sector to continue financing extension services, although service delivery will not be restricted to the public sector. 6.2 Transition arrangement to regular operations:

- 13 -

Considerable variation exists between districts in their staffing levels, facilities, and priorities to support agricultural development. While all districts received some support under NAEPII, only 56 districts received significant support. Generally, it seems that the LGAs intend to continue to finance the basic operations of extension services. It is unclear, however, to what extent they will be able to finance further development of physical facilities, transport needs and field operations to an adequate standard for effective operations of district services. As indicated previously, the intention of the central government is to provide additional development funds to districts to support extension operations, and to continue responsibility to complete the upgrading of extension staff qualifications through formal education. Future Development. The draft proposal for a new program (Agriculture Services Support Program) under the Agriculture Sector Development Program that will continue to support extension, research and other agricultural services is well advanced. The findings of this ICR will be relevant to the further formulation of this program.

7. Bank and Borrower Performance Bank 7.1 Lending:

The project was formulated in cooperation with the Government during the latter stages of NALERP. As discussed in section 3.6, however, the quality at entry is considered unsatisfactory. The Bank’s contribution to project preparation is also considered unsatisfactory. Decentralization of extension services to the LGAs was initiated in 1997 with the Bank supporting this policy initiative. Over about an 18-month period during 1998-99, implementation had stalled due to slow redeployment of extension staff and slow establishment of the necessary institutional arrangements to set up effective extension services in the LGAs. Implementation progress was first rated unsatisfactory in October 1998 (and continued to be rated unsatisfactory by the Bank until December 2000). Not until early 2000, however, following a one-year delayed MTR in December 1999, was the project restructured/reoriented. The project should have been restructured at an earlier date. Although the project clearly offered valuable training and capacity building for the agents newly devolved to the district level and lower, the project did not appear to offer appreciable assistance to PORALG and LGAs in rapidly assuming their new roles. In hindsight, one can argue that the new project design should have better reflected the new decentralized institutional arrangements for extension services in the LGAs with transferred responsibility to PORALG, and been focused on building institutional capacity in PORALG and the LGAs. Hindsight cannot tell us, however, whether alternative approaches would have been more effective. 7.2 Supervision:

The Bank fielded thirteen regular supervision missions during the 7-year period. Missions were regular, although project records show only two missions between December 1998 and July 2000, a very critical juncture for the project. During these decisive stages of devolution of extension services and associated institutional changes, the Bank should have been more pro-active in dealing with the problems of implementation. The reports of the supervision missions generally provided a comprehensive review of implementation progress and clearly indicated what follow up actions were required in future implementation activities. Bank performance in implementation of the restructured project is considered satisfactory, although, over the entire life of the project, Bank performance in supervision is unsatisfactory. 7.3 Overall Bank performance:

Overall Bank performance is rated (marginally) unsatisfactory, in light of poor quality at entry, unnecessary delay in redesigning the project, and possible missed opportunities in the redesign of the

- 14 -

project. Satisfactory performance after the restructuring contributed to achievement of the objectives. Borrower 7.4 Preparation:

The Borrower was closely involved with project formulation over an 18-month period, during the latter stages of NALERP, which was important to ownership and commitment to NAEPII. As discussed in sections 3.6 and 5.2, the borrower was fully aware of the impending decentralization, and shares in responsibility for flawed design. 7.5 Government implementation performance:

As indicated previously, several government actions significantly affected implementation. The legislation of the Regional Administration and Local Authorities Act in 1997 effectively created 119 autonomous district extension services from a single, centrally-administered, national extension service. Overall responsibility for coordinating project implementation was supposed to move to PORALG, but didn’t. At this time, the capacity within MRALG and the LGAs was weak, and there was an absence of a Vision/Strategic Framework for Extension in a decentralized extension system. Furthermore, the Government decided to continue to undertake extension activities financed under NAEPII in all districts, rather than concentrate efforts on the 56 priority districts designated as Agriculture Extension Reform districts, as recommended at the MTR and by the Bank. The Government requested that the geographic scope of the project be expanded to include four more regions on the mainland, and to include Zanzibar. Implemented at a time of major institutional changes, these decisions may have spread project resources further, and had implications on project outcome. In 2001, MAC was restructured into three new ministries, whereby the central responsibilities for oversight of agricultural extension were split between the MAFS and MWLD, which increased the complexity of support and coordination. The restructuring of ministries that followed the election of a new Government enhanced coordination problems in the sector. It is also apparent that the Government should have been more proactive in assisting the transitional process. Therefore, government implementation performance is considered unsatisfactory. 7.6 Implementing Agency:

At project effectiveness, extension services were unified under a single agency, MAC. The Government established the institutional arrangements for management and implementation of the project under a PSC, overseeing a Project Management Unit in MAC, which was linked to DESCs in the districts. Funds were routed from the PMU to the districts through the regional offices, and the Regional Agriculture and Livestock Development Officer (RALDO) coordinated the activities of extension staff in the districts. Performance was satisfactory in the first year or so of implementation. Through the middle years, overall implementation performance was weak. Although responsibility for activities crucial to success of the project moved to the aegis of PORALG with decentralization, much of the responsibility for implementation of project activities (education, training, communications and M&E) remained with the PSC and PMU in MAC, with a representative from PORALG included in the PSC. Implementation activities at the district level, faced major problems. After decentralization, the districts received funds directly to cover normal recurrent expenditure for extension operations, and the Zonal Research and Extension Officer was the main link between MAC headquarters and the districts. While MAC assisted with preparation of new job descriptions for agricultural sector staff in the LGAs, and provided financial and accounting guidelines to the LGAs, the restructuring process was very slow. Transferring implementation responsibility at the local level from the dissolved DESCs to the DMTs, and establishing an effective institutional and financial arrangement in each district council for disbursements of

- 15 -

funds for project activities encountered substantial difficulties. Clearly, the long period required for this transition adversely affected project implementation over a period of about two years. Although MAC attempted to assist in the transition process, the overall result was generally unsatisfactory during this time. The latter part of project implementation experienced satisfactory disbursement flows to the districts, institutional arrangements for extension services were established, and many of the revised activity targets were achieved following the extensions, despite the continuing limitations in project management by the DMTs. The IDA credit was also nearly expended, and the Borrower had provided more than its scheduled contribution. The Borrower partially or completely complied with the covenant agreements. Under the Zanzibar component, the separate steering committee (Project Technical Committee) and PMU established in MANREC functioned satisfactorily in implementing the agreed activities, and despite its separate autonomy, there appeared to be good coordination with implementation management on the mainland. 7.7 Overall Borrower performance:

Borrower performance is judged to be (marginally) unsatisfactory, based on the borrower's shared responsibility for flawed quality at entry, lagged provision of counterpart funding for a period, counter-productive restructuring of the agricultural ministries, and slow implementation of institutional changes needed to make decentralization work. The generally satisfactory performance of the Borrower in supervision after the restructuring allowed the project to meet its objectives, but is not sufficient grounds to support a satisfactory rating overall.

8. Lessons Learned The experience with NAEP II offers lessons with regard to project formulation and implementation generally, and with regard to agricultural extension specifically. Key lessons are presented below. Management of the World Bank must protect the intellectual space for staff members to disagree professionally, and to act on their best professional judgment. NAEP II was designed during a period of active debate about the future direction of support for agricultural extension. The Bank had historically made a major investment in the Training and Visit system, but as early as 1989 doubts were beginning to be expressed that the system was effective and affordable for poor countries (World Bank Technical Paper #87, 1989). Empirical evaluative work launched by OED yielded convincing evidence in 1999 that in fact the T&V system was of questionable efficacy and doubtful sustainability (see Footnotes 3 & 4 in Annex 10). During the period prior to the issuance of the OED report, an atmosphere prevailed in which junior staff who had doubts about the T&V system could reasonably fear for their careers if they questioned the strong advocacy for the system by a small number of highly-placed senior staff. The result was a supression of internal professional debate to the detriment of the quality of project design and the interests of our clients. [This lesson is the opinion of the Sector Manager, and not necessarily that of the primary authors of the ICR.] Key performance indicators should be defined during project preparation and monitored during implementation. Documentation of impact and results is needed as a management tool during implementation, and in order to secure continued public support for extension. Design of follow-on projects should reflect lessons of prior projects. Lessons need not be those formally noted in a completed ICR, but they should be incorporated in project design.

- 16 -

Prompt and determined restructuring in response to changes in the institutional framework and country circumstance can keep a project on track to achieve its objectives. While restructuring was a positive and important initiative, project outcome, especially institutional development, would have been more substantial if restructuring had been earlier and the design better reflected the new institutional arrangement. Allocation of uncommitted funds, but for specific use, can enhance the flexibility of the project to respond to changing conditions. Pilot initiatives were used successfully to focus on the development and adoption of suitable technologies in local areas, as well as to create a multiplier effect in surrounding areas, and to emphasize a participatory approach to development. The decentralization of extension services to LGAs, combined with a move toward a participatory demand-driven public and private extension system, improves the effectiveness and affordability of services. The reform of extension services designed initially on the model of centralized T&V systems follows a general pattern, and includes steps of creating a vision, aligning key stakeholders around it, defining increased roles for private providers of service, reduced roles for the public sector, and changes in staffing and flows of funds. Chronic underfunding of T&V systems may not have been detrimental to aggregate economic performance. Once the systems have been reformed, however, and have greater potential to deliver results, levels of funding should be adequate to operate them at the capacity commensurate with design. The system of advisory services under the Agricultural Sector Development Program will have to operate on a larger scale in the future in order to have an adequate impact on productivity. Continued public funding for agricultural extension is appropriate when the information conveyed has features of a public good, and when the users of information are too poor to purchase optimal quantities through private transactions. Public funding should be distinguished from public provision of service. The project outcome emphasizes the importance of increasing the reach of extension services through the creation of networks of supporting providers. This involves encouraging better trained and more independent farmer groups, farmer motivators, NGOs and private sector groups to provide extension services. Effective production and communication of agricultural information is an important means to transfer knowledge for production. Developing communication networks and extension materials requires greater emphasis on acquiring modern information technologies, and enhancing the network of IT services throughout relevant Government ministries and LGAs, in order to create rapid, on-line access to extension materials at local levels. Information conveyed through the agricultural extension system should enhance economic as well as technical decisions. Household surveys assessing project benefits to farmers indicate that adoption rates of technologies were much lower than awareness rates of farmers, highlighting the importance of markets, marketing, input and output prices, credit and transport infrastructure as important determinants of technology adoption. Interventions to improve the performance of agricultural markets increase returns to investment in agricultural research and extension. Advising farmers on how best to identify economic opportunities at a point in time will enhance their ability to use technical information.

9. Partner Comments (a) Borrower/implementing agency:

The Executive Summary of the Government of Tanzania's ICR and a Summary of the Zanzibar ICR

- 17 -

provided by the Revolutionary Government of Zanzibar are in Annexes 12 and 13. (b) Cofinanciers: (c) Other partners (NGOs/private sector):

10. Additional Information

- 18 -

Annex 1. Key Performance Indicators/Log Frame Matrix Outcome / Impact Indicators: Indicator/Matrix Impact of Extension Services

1

Projected in last PSR (Not projected)

- Availability of extension services - Obtained extension advice from EOs - Method of contact with EOs - Type of extension agent contacted - Quality of advice received - Frequency of contact with EO - Farmer to farmer training

Actual/Latest Estimate Final Beneficiary Assessment Study - 2003 (388 farmers in 11 districts sampled) 88.9% of farmers surveyed 81.7% of farmers surveyed Direct (29%), Group (27%), Direct and group (32%) EO (82.6%); Fellow farmer (7.3%) Rated excellent advice (77.3%) Within 1 mth.(52.3%); 2-4 mths (24.4%) 45% of farmers and 45% wish to continue as contact farmers

- Farmers’ Sources of extension information

Radio & TV (62%); Newspapers/magazines (48%); books/booklets (30%), media vans (11%)

Farmer Adoption (%) of Extension Advice for Cereal Crops at Farm Level - Timely land preparation - Timely planting - Use of improved seeds - Use of proper spacing - Fertilizer/ Manure applications - Timely weeding

Baseline Beneficiary Beneficiary study (1996) study (1999) study (2003) 60% 65% 77% 62 66 76 29 48 55 47 45 72 23 (1998) 24 60 70 (1998) 69 86 (Adoption rates (%) of technologies for cereals)

Farmer Adoption (%) of Improved Biological/Environmental Practices - Biological Control/PM - Agroforestry practices - Crop rotation/intercropping - Soil Erosion Control - Post-Harvest care - Pesticide Use

Adoption of Improved Technologies (%) 1998/99 2002/03 7 26 19 31 42 43 18 26 33 63 20 48 (Final Beneficiary Assessment, 2003)

Adoption of Improved Livestock Practices - Improved housing - Parasite & disease control - Improved cattle feeds - Pasture management

41% of pastoralists 67% of pastoralists 39% of pastoralists 31% of pastoralists

Change in Yields of Major Commodities Maize (kg/ha) Paddy (kg/ha) Sorghum (kg/ha Beans (kg/ha) Cassava (kg/ha) Cotton (kg/ha) Coffee kg/ha) Milk (litres/cow/day) eggs(eggs/hen/year)

(Not Projected)

- 19 -

Baseline Beneficiary Beneficiary study (1996) study (1999) study (2003) 505 588 500 271 644 567 289 259 226 175 235 332 743 263 330 148 275 4 4.142 62 245 (2003 was drought year)

Output Indicators: Indicator/Matrix

1

Projected in last PSR

Actual/Latest Estimate

MAINLAND 1. Extension Education and Training Farmers: Residential training On-site training Farmers study tours Farmers benefitting from training

Initial Plan: 10,000; Revised: 6,000 Initial Plan: 75,000; Revised 45,000 Initial Plan: 5,000; Revised: 4,000 Revised Plan: 54119

Actual: 5,604 (Govt. ICR) 37,179 (actual in last PSR, Govt. ICR) Actual: 1,807 (Govt ICR) Actual: 44450

Staff Degree and Diploma Programs: BSc degree MSc degree Retraining Diploma holders Diploma in General Agriculture

Initial and Revised Plan: 40 Initial Plan: 80; Revised Plan: 64 Initial and Revised Plan: 800 Revised Plan: 400

Actual: 20 (Govt ICR) Actual: 54 Actual: 650 (PSR and Govt ICR) Actual: 405

Staff Study Tours Study tours by outstanding VEOs Inter-regional tours International study tours No. of Bi-monthly Training Sessions No. Quarterly Training Sessions

Initial Plan: 360; Revised Plan: 288 Initial Plan: 1440; Revised Plan: 1152 Revised Plan: 15 Initial Plan: 16224; Revised Plan: 8954 Initial Plan: 384, Revised: 224

Actual: 216 Actual: 864 (last PSR and Govt ICR) Actual: 15 Actual: 5,444 Actual: 168

Revised Plan: 240

Actual: 240

Revised Plan: 55

Actual: 55

Revised Plan: 1,122 Revised Plan: 300

Actual: 1,100 Actual: 300

2. Communications/Media - FEPU Radio Programs Farmers magazine (issues) Extension News (issues) Booklets (issues – different titles) Leaflets (issues – different titles) Posters (issues – different titles)

Initial Plan 624: Revised: 767 Initial Plan: 42 Initial Plan: 42 Initial Plan: 14 Initial Plan: 54 Initial Plan: 42

Actual: 1040 Actual: 38 Actual: 37 Actual: 12 Actual: 39 Actual: 31

3. Communications/Media - ZCCs Posters Leaflets Booklets Video programs Radio Programs

Initial Plan: 48 Initial Plan: 96 Initial Plan: 24 Initial Plan: 1464 Initial Plan: 42

Actual: 69 Actual: 219 Actual: 1 Actual: 624 Actual: 192

Short Courses for Staff TOT on Participatory Approaches (2wks) DSMS on farm management & participatory approaches Frontline staff on farm management & participatory approaches Computer courses

4. Pilot Initiatives I. Extension management & farmer empowerment II. Role of private sector in extension services provision III. Technology transfer initiatives

5. Civil Works New staff Houses Arusha Morogoro Mwanza Shinyanga Tabora

Actual: 6 (with FFS approach in 16 districts) Actual: 8: (mostly on seed production) Actual: 21 (irrigation, small dams, mechanization, crop production, animal health)

Initial Plan: 19 Initial Plan: 20 Initial Plan: 28 Initial Plan: 24 Initial Plan: 22

Actual: 19 Actual: 20 Actual: 28 Actual: 24 Actual: 22 Actual: 4 Actual: 8 Actual: 2

District Extension Offices - new - rehabilitated ZCC (rehabilitated)

- 20 -

ZANZIBAR 1. Extension Staff Training Diploma training Certificate training Overseas training Refresher course BEO Refresher course SMS 2. Farmer Training On-farm demonstrations/trials (140) Field Days Study tours Residential courses 3. Media Extension Materials Newsletters Radio programs Television programs 4. Civil Works New staff houses Training dormitory Other works 1

Initial Plan: 19 Initial Plan: 83 Initial Plan: 12 Initial Plan: 40 Initial Plan: 30

Actual: 54 Actual: 76 Actual: 12 Actual:164 Actual: 30

Initial Plan: 1916 Initial Plan: 1315 Initial Plan: 1073 Initial Plan: 7250

Actual: 2035 (M-1070; F-965) Actual: 1057 (M-618; F-439) Actual: 1358 (M-819; F-539) Actual: 6632 (M-2710; F-3322)

Initial Plan: Initial Plan: 150 Initial Plan: 144

Actual: 2000 Actual: 93 Actual: 80

Initial Plan: 7 Initial Plan: 1 Initial Plan: 8

Actual: 7 Actual: 1 Actual: 8

End of project

- 21 -

Annex 2. Project Costs and Financing Project Cost by Component (in US$ million equivalent) Appraisal Estimate US$ million

Component

Actual/Latest Estimate US$ million

1. Institutional Strengthening 2. Extension Training 3. Communications Support 4. Pilot Initiatives

13.74 10.56 2.58 1.49

17.10 9.40 2.30 1.80

Total Baseline Cost

28.37 0.63 3.95 32.95 32.95

30.60

Physical Contingencies Price Contingencies

Total Project Costs Total Financing Required

Percentage of Appraisal

30.60 30.60

Project Costs by Procurement Arrangements (Appraisal Estimate) (US$ million equivalent)

Expenditure Category

1. Civil Works 2. Vehicles 3. Plant & Equipment 4. Training Studies & Technical Assistance 5. Pilot Initiatives 6. Recurrent Operating Costs Total

Total Cost

Procurement Method 1/ ICB

NCB

Other 2/

0.000 (0.000) 2.729 (2.729) 0.207 (0.207) 0.000 (0.000) 0.000 (0.000) 0.000 (0.000) 2.936 (2.937)

0.932 (0.838) 0.691 (0.691) 1.200 (1.164) 0.000 (0.000) 0.897 (0.897) 0.000 (0.000) 3.719 (3.590)

0.233 (0.210) 0.000 (0.000) 0.386 (0.374) 14.753 (14.753) 0.897 (0.897) 10.018 (8.339) 26.286 (26.286)

- 22 -

1.165 (1.048) 3.420 (3.420) 1.793 (1.745) 14.753 (14.753) 1.795 (1.795) 10.018 (8.339) 32.942 (31.100)

124 89 89 121

Project Costs by Procurement Arrangements (Actual/Latest Estimate) (US$ million equivalent)

Expenditure Category ICB 1. Civil Works

Procurement Method 1/ NCB Other 2/

0.000 (0.000) 0.000 (0.000) 0.176 (0.176) 0.000 (0.000) 0.000 (0.000) 0.000 (0.000) 0.176 (0.176)

2. Vehicles 3. Plant & Equipment 4. Training Studies & Technical Assistance 5. Pilot Initiatives 6. Recurrent Operating Costs Total

0.740 (0.740) 2.143 (2.143) 0.963 (0.963) 0.000 (0.000) 0.000 (0.000) 0.000 (0.000) 3.846 (3.846)

Total Cost

0.214 (0.196) 0.667 (0.661) 0.309 (0.306) 11.355 (11.355) 1.744 (1.744) 12.299 (9.958) 26.588 (24.220)

0.954 (0.936) 2.810 (2.804) 1.448 (1.445) 11.355 (11.355) 1.744 (1.744) 12.299 (9.958) 30.610 (28.242)

1/ Figures in parenthesis are the amounts to be financed by the Bank Loan. All costs include contingencies. 2/ Includes civil works and goods to be procured through national shopping, consulting services, services of contracted staff of the project management office, training, technical assistance services, and incremental operating costs related to (i) managing the project, and (ii) re-lending project funds to local government units.

Project Financing by Component (in US$ million equivalent)

Appraisal Estimate Bank Govt. Total 13.9 1.7 15.6

Actual/Latest Estimate Bank Govt. Total 14.175 2.078 16.253

Percentage of Appraisal Bank Govt. 102.0 122.2

12.6 2.8

0.0 0.1

12.6 2.9

10.101 2.222

0.000 0.290

10.101 2.512

80.2 79.4

4. Pilot Initiatives

1.8

0.0

1.9

1.744

0.000

1.744

96.9

Total Project Costs

31.1

1.8

32.9

28.242

2.368

30.610

98.4

Component 1. Institutional Strengthening 2. Extension Education & Training 3. Communications Support

- 23 -

290.0

131.7

Annex 3. Economic Costs and Benefits As the SAR pointed out, the economic impact of an agricultural extension project is diffuse. While agricultural extension is an important complement to the other factors that influence the choice of alternative techniques by farmers, it is difficult to isolate the impact of NAEPII from other variables that determine farmer adoption of more efficient production methods, and yield and production changes. The approach adopted in the SAR was to identify the "minimum incremental benefit stream" that would justify the investment the government extension service, and then under an "incremental adoption" model, see whether the project would, under a reasonable set of assumptions, produce the desired minimum benefit stream. A detailed description of the basic assumptions about financial and economic prices, costs, target crops and adoption rates is contained in Annex 4 of the SAR. Under the "minimum national impact" approach, the SAR noted that to obtain a rate of return of 25 percent or a benefit-cost ratio of 2.1 (at a discount rate of 12 percent, Tanzania's opportunity cost of capital--OCC) on the proposed off-farm investment in government extension services, it was sufficient for net on-farm benefits generated from selected crops and livestock to grow at a rate of 0.45 percent per annum, over the 20-year economic life of the project. The selected crops were maize, sorghum, millet, paddy, cotton, coffee, and tobacco (covering an aggregate of 70 percent of the cultivated area), and the livestock were traditional and improved cattle. The benefits derive from an average annual yield increase of 0.15 percent and an average annual increase of on-farm costs of 0.07 percent. The "incremental adoption" analysis showed that the above rate of return and benefit-cost ratio could result from the intervention of an extension service that influenced 11 percent of the farmers cultivating the selected crops (and raising the selected animals), raising their yields between 10 and 22 percent over a five-year period. Since these crops represented about 70 percent of the area cultivated, this implied that the extension service would bring an additional 270,000 farmers to use improved production techniques. This ICR follows the incremental adoption analysis to recalculate the Economic Rate of Return (ERR) over a 20-year period from 1996, using actual/observed yields and area increases, prices and costs from 1996-2003. Several assumptions/caveats need to be noted: (i) the assessment is carried out in 2003 prices, over the 20-year period; using the assumptions in the SAR on export and import parity prices, at the farmgate; actual project costs were used as off-farm costs, assuming recurrent costs of maintaining the improved level of service over the next 12 years; (ii) actual yields from the 2003 household surveys of pilot areas were not used, as these surveys emphasized assessing changes in income, and covered only small groups of farmers for each crop. Also, as 2003 was a drought year, there were wide fluctuations in the pilot survey yield data. Therefore national data were used for yields, production and area under cultivation from 1996-2002. Yields were increased at the observed rate for each crop (with increases ranging from 1-3 percent per annum) over the next 12 years; (iii) recent farm budget data could not be obtained. The ICR used the crop budgets presented in the SAR, but adjusted the labor inputs and other variable costs to reflect increases in the part-time rural wage rate, and the Dar Consumer Price Index; (iv) data on fertilizer use was incomplete. Since crop budgets from 1996 were used, the ICR does not reflect increases in the price of fertilizer. Discussions with farmers and extension workers during the ICR mission confirmed that farmers maybe achieving the observed yields using less fertilizer and more manure--a cost-effective technique, given the high fertilizer prices following the removal of the fertilizer - 24 -

subsidy; (v) this analysis assumed area increases for most crops, following observed trends from 1996-2003, whereas the SAR had assumed constant area under cultivation for each crop over the 20 year period. (vi) as the information on livestock and milk production of indigenous cattle and improved cattle could not be obtained, this analysis uses the same information as in the SAR. If all observed yield increases for each crop are attributed to extension services, the benefit-cost ratio at 12% is 1.529 and the internal rate of return (IRR) is 23%. If, however, we assume that extension contributes only to a 50% of the yield increases under each crop, then the benefit-cost ratio and the IRR decrease to 0.765 and 6% respectively. Assuming 100% of yield increase due to extension: Net Present Value of on-farm benefits at 12%: Net Present Value of off-farm Costs at 12%: Benefit Cost Ratio (at 12 %): Internal Rate of Return (IRR):

38,284.3 Tsh. ml. 25,034.5 Tsh. ml. 1.529 23%

- 25 -

Annex 4. Bank Inputs (a) Missions: Stage of Project Cycle Month/Year Identification/Preparation Sept/Oct - 1995

Count

No. of Persons and Specialty (e.g. 2 Economists, 1 FMS, etc.) Specialty

Performance Rating Implementation Development Progress Objective

8

Agriculture Services

12/06/1996 07/11/1997 03/13/1998 12/03/1998

6 2 1 5

S S S U

S S S S

06/14/1999 12/21/1999-MTR

4 8

U U

U U

07/14/2000 11/21/2000 07/20/2001

1 1 4

U S S

U S S

03/29/2002

4

U

S

10/01/2002

4

S

S

06/01/2003

5

S

S

12/04/2003

2

TTL (1); RESEARCH (4); TTL (I); AG. SERVICES (1); TTL (1) TTL (1); FIN. MGMT. (1); ASSISTANT(1); EXTENSION (1); SOIL FERTILITY (1) TTL (1); ECONOMIST (3) TTL (1); AG. RESEARCH AND SERVICES (3); ECON (1); INSTI (1); FINANCE (1); COMMU.(1); TTL (1) TTL (1) TTL (1); AG./ECONOMIST (2); PROC SPECIALIST (1) TTL (1); AGRICULTURIST (1); RURAL DEV. (1); PROC. (1) TTL (1); OPERATIONS (1) ; AGRICUL. (1); RURAL DEV. (1) TTL (1); AGRICULTURIST (1); RURAL DEVELOPMENT (1); M&E (1); PROCURE. (1) TTL (1); TEAM MEMBER (1)

S

S

April/May

1

Agriculturist

S

S

Supervision

ICR

(b) Staff: Stage of Project Cycle Identification/Preparation Supervision ICR Total

Actual/Latest Estimate No. Staff weeks US$ ('000) 56.4 167.4 142.5 610.3 198.9

- 26 -

777.7

Annex 5. Ratings for Achievement of Objectives/Outputs of Components (H=High, SU=Substantial, M=Modest, N=Negligible, NA=Not Applicable)

Macro policies Sector Policies Physical Financial Institutional Development Environmental

Rating H SU H SU H SU H SU H SU H SU

M M M M M M

N N N N N N

NA NA NA NA NA NA

H H H H H H

M M M M M M

N N N N N N

NA NA NA NA NA NA

Social Poverty Reduction Gender Other (Please specify) Private sector development Public sector management Other (Please specify)

SU SU SU SU SU SU

- 27 -

Annex 6. Ratings of Bank and Borrower Performance (HS=Highly Satisfactory, S=Satisfactory, U=Unsatisfactory, HU=Highly Unsatisfactory)

6.1 Bank performance Lending Supervision Overall

Rating HS HS HS

S S S

U U U

HU HU HU

U U U U

HU HU HU HU

Bank performance is considered 'marginally' unsatisfactory. 6.2 Borrower performance Preparation Government implementation performance Implementation agency performance Overall

Rating HS HS HS HS

S S S S

Overall Bank Performance and Borrower Performance are considered 'marginally' unsatisfactory. Since 'marginal' ratings are not allowed/accepted in ICRs, Bank and Borrower Performance are rated Unsatisfactory.

- 28 -

Annex 7. List of Supporting Documents Staff Appraisal Report United Republic of Tanzania, National Agricultural Extension Project – Phase II, Report No. 15117-TA, June 5, 1996, Agriculture and Environment Operations Division, Eastern Africa Department, Africa Region. Aide memoire/Supervision reports Supervision Mission (December 1996) – Aide memoire IDA Review Mission (June/July 1997), Aide memoire IDA Supervision Mission (March 1998), Aide memoire IDA Review Mission – November/December 1998, Aide memoire Supervision Mission (June 1999) – Aide memoire Mid-Term Review Mission (December 1999) – Final Aide Memoire Supervision Mission (July 2000) – Aide memoire Supervision Mission (November 2000) – Final Aide memoire Supervision Mission (July 2001) – Aide memoire Supervision Mission (March 2002) – Draft Aide memoire Supervision Mission (October 2002) – Final Aide memoire Supervision Mission (June 2003) – Aide memoire Supervision Mission (December 2003) – Final Aide memoire Studies 1. National Agricultural Extension Project II (NAEPII), Baseline Survey Report of 1996, Final Report, Ministry of Agriculture and Cooperatives, United Republic of Tanzania, 1996. 2. Needs Assessment for Training, October 1996 3. Needs Assessment for Communications Support, October 1996 4. Pilot Initiatives, October 1996 5. Vision/Strategic Framework for Extension 2010, 2001 6. Capacity Building Needs Assessment Study, 2001 7. Revision of Curricula of the Diploma Level Course in General Agriculture, 2001 8. Household Assets Survey of Phase I and Phase II Districts under Local Government Reform, Ministry of Agriculture and Food Security, United Republic of Tanzania, 2001. 9. Study on the Institutional Arrangements and Capacity Assessment of Public Sector Agriculture Extension, Draft Report, Ministry of Agriculture and Food Security, United Republic of Tanzania, October 2003. 10. Final Beneficiary Assessment Study, National Agricultural Extension Project (NAEPII), Ministry of Agriculture and Food Security, United Republic of Tanzania, September 2003 11. Lessons from Experiences of NAEP II Micro projects, Agrisystems (EA) Ltd , March 2004

- 29 -

Additional Annex 8. Institutional Framework and Operations of Extension Services Mainland Tanzania and Zanzibar have very different institutional arrangements. Therefore, this review of the past and present institutional arrangement for agriculture extension services, and the implications for future development of these services considers them separately. Mainland Tanzania Institutional Considerations. At the commencement of NAEPII, agricultural research and extension services for agriculture and livestock, were under the control of a single ministry, MAC. A large number of extension staff were located in Dar es Salaam with other extension officers located at regional, district and village locations. Under the original project design, there was to be some rationalization of extension services through staff redeployment and retrenchment. Decentralization of extension services was to occur through deployment of Subject Matter Specialists (SMSs) from the center to the district and division level. In addition, there was to be a retrenchment of nearly 1,000 extension staff that were deemed to be unsuitably qualified and not retrainable. Significant institutional changes occurred in the early years of project implementation that were not apparently considered at project appraisal. With the introduction of the Regional Administration Act No. 19 of 1997 and subsequent amendments of this Act in 1999 and 2000, line ministries became responsible for policy formulation, setting the sector regulatory framework, issuing technical guidelines, determining minimum standards for service delivery, sectoral M&E, training and research. Also, a new ministry was established--Ministry of Regional Administration and Local Government. This subsequently became the President’s Office, Regional Administration and Local Government (PORALG). Under the new arrangement, PORALG and the Regional Secretariats were to guide and supervise the planning and management of the district administrations and provide the interface between activities and investments planned and managed by the central government and the districts. Responsibility for extension services was devolved completely to the local government authorities (LGAs). This meant that all the extension staff located centrally were to be redeployed to the LGAs, with only a small core of extension staff retained within MAFS and MWLD. However, with restructuring, the overall number of extension staff has decreased, as the nearly 900 staff retrenched were not replaced. In October 2003, there were 5,835 extension staff that were deployed at both district headquarters (1,110) and at ward/village levels (4,725) to provide service to about 10,470 villages. Thus only about 50% of the designate positions (based on one extension officer per village) are filled. Further, in February 2001, MAC was restructured to create three ministries--the Ministry of Agriculture and Food Security (MAFS), the Ministry of Water and Livestock Development (MWLD) and the Ministry of Cooperatives and Marketing (MCM). Although MCM has no role in providing extension services, this further separated national responsibility for agriculture and livestock. The process of decentralization has been slow. Effective decentralization was hampered by many factors including, reluctance by extension staff to accept the new administrative arrangements, reluctance to relocate, disparate allegiance between LGAs and line ministries, insufficient allocation of operational funds by LGAs, lack of promotional opportunities, and difficulty in recruiting additional qualified staff. Rigidities within the system also slowed the process, as vested interests and an unwillingness to accept change acted against reform measures. The government aims to streamline and improve performance within the civil service through measures to raise the level of service packages to compete with the private sector. (The Client Service Charter is a tool

- 30 -

introduced by the Government for improving service delivery.) The introduction of reforms under the Public Service Reform Program aims to improve recruitment, retention and motivation of Government staff at all levels. However, sectors such as agriculture are a particularly challenging, given the size of the staff, its predominantly non-urban location, and its specialist skill requirements. Establishment of an effective decentralized service also depends on the successful adjustment from central to locally-planned agricultural development, and improved and locally-customized information systems for monitoring and reporting on services. The weak capacity in many LGAs has slowed the reform process, as responsibility for government services has been transferred, while resource mobilization and financial and administrative powers has remained largely under the control of central Government. Other bureaucratic impediments include the lack of a well conceived change-management strategy, conflicting provision in sector and local government legislation and lack of uniform standards of service. Physical resources available to extension staff, and their ability to travel easily throughout the area of responsibility affect the operation of extension services. NAEPII recognized this, and provided improved physical facilities, equipment, vehicles, motorbikes and bicycles. With decentralization to LGAs, adequate physical resources become even more important. The government has to have a needs-based plan for, and commitment to, continue capital investment in local facilities to enable the decentralized extension services to operate efficiently. The restructured project design did not sufficiently address the total change in administrative responsibility. It needed to focus support on establishing/strengthening the new administrative structures of PORALG and the individual LGAs. Extension System. The extension system introduced under NALERP was based on the T&V system of extension services. This system is very staff intensive, strongly hierarchical in staff deployment and management, oriented to training and information provision, and thereby provides a top-down method of extension delivery to selected (contact) farmers. T&V is expensive to operate. Although the system was modified under NAEPII to focus more on encouraging and developing a demand-driven and participatory extension system, several problems remain. The financial ability of most LGAs to support mainstream extension services is limited, and it is highly likely that the government will not be able to significantly increase the size of the extension service. Furthermore, it is likely that the public sector will gradually, but increasingly, limit its role in financing the provision of public goods and services. Thus Pilot Initiatives outcomes are critical to further deepening reforms, and modifying and developing an alternative system of extension delivery that is linked to and partially replaces the public system. The Pilot Initiatives under NAEPII demonstrated that: • NGOs, CBOs and the private sector can be important players in providing an effective extension services; • Farmer Field Schools (FFSs), as a methodology for technology transfer and farmer training, is a promising, cost-effective method for client-oriented extension services, which responds to farmer demands, empowers them through group training and participation, and provides effective farmers-research-extension links; • the establishment of grassroots organizations and their ability o mobilize savings is critical for the sustainability of community-based initiatives, including paying for service providers and sharing the costs of running FFSs. Education and Training. GOT aims is to have a cadre of more knowledgeable and better educated

- 31 -

agricultural extension staff. GOT's target is to have senior extension staff and SMSs trained to university level, and for VEOs and WEOs to have certificate and diploma level training. A major thrust of NAEPII was to upgrade the knowledge and skills of extension staff through formal and informal education and training. This included undergraduate and post-graduate programs at Sokoine University of Agriculture, diploma and certificate level training at Ministry of Agriculture Training Institutes and Livestock Training Institutes, and overseas, national and regional study tours. Training sponsored under NAEPII supplemented other formal and informal training of extension staff sponsored directly by the Government. The type of training and educational levels required is an issue. The government’s training policy aims to have staff with very specialized training. To this end , Government encouraged Sokoine University of Agriculture (SUA) to offer very specialized courses at undergraduate and post-graduate levels. Such specialization at undergraduate level seems inappropriate for extension staff. Extension staff should be generalists. At the most, there could be some general specialization (agriculture, livestock or extension) in the final year of the undergraduate course. SUA, which trains many graduates who seek employment in the private sector, has found that employers in the private sector do not want such specialized graduates, and therefore have moderated their curricula to be more generalized within the specialist degrees. Continued technical training is another issue, because of the decision taken to stop in-service technical training for extension staff. The T&V system started off with regular fortnightly training for extension staff. This subsequently changed to monthly training and then bi-monthly training under NAEPII. During the implementation of NAEPII the bi-monthly training was discontinued. This has left a void in technical training for extension staff. An alternative means of regular technical training for extension staff is required to keep abreast of new technologies and to provide a forum where technological issues can be discussed. An alternative is to introduce seasonal and needs-based technical training programs for district extension staff. The technical newsletter initiated under NAEPII was discontinued due to lack of funding when NAEPII closed. This needs to be reinstated, but with a more technical format, as it can provide an important means of communicating technical information from research, and a forum for extension staff to communicate ideas and views to research staff. Planning of Extension Services. The development of annual district plans for extension activities is essential, both as a means of defining demand-driven extension services requirements and the necessary budget support for identified activities. These need to be presented to district councils as part of their overall annual planning requirements. Extension staff at village and ward level should formulate plans based on participatory analyses of farmers needs at local levels, and these should be aggregated into an overall plan for the district. Discussions at selected districts revealed that this is only done in some districts, and not using a bottom–up approach. Research-Extension-Farmer Linkages. NAEPII sought to strengthen research-extension-farmer linkages through: (i) close coordination between the extension service and research, where staff research and extension staff meet regularly (Zonal Technical Committees, Zonal and Regional Agriculture Committees, Zonal and national Agriculture Shows, national Coordinating Committee, Regional Extension Coordinating Committee, Pre-seasonal/annual zonal workshops in each agro-climatic zone); (ii) establishment of a research-extension liaison office and officer in the zonal research centers; (iii) establishment of a detailed schedule of regular coordination and contacts between research and extension workers; (iv) adoption of farmer group approach to collaboratively define and solve problems; (v) adoption of a farming systems approach to research, with funds available from the TARPII to supplement the activities at farm level under NAEPII; and (vi) involving farmers or farmer representatives on committees deciding priorities, strategies and programs in research. Under NAEPII, the devolvement of extension services from MAC (that was previously responsible for both national extension and research), made it more difficult to - 32 -

maintain good research-extension-farmer linkages. Despite this institutional separation, these linkages seem to have been reasonably effectively maintained, through MOUs between Zonal Research Directors and District Executive Directors to reestablish/strengthen research-extension-farmer linkages. The farmer group approach, which has been strongly reinforced through the FFS approach, involvement of on-farm research with extension, the more inclusive approach of extension services provided by NGOs and CBOs and many of the Pilot Initiatives under NAEPII, have all also supported and strengthened these linkages. Communications. NAEPII correctly identified the need to strengthen the capacity of government services to generate relevant agricultural information and communicate such media effectively to farmers. Unfortunately, the design of activities for this sub-component pre-dated the enormous development of information technologies (ITs) that have occurred in the last five years. However, despite this limitation, there were significant developments in capacity to improve media (print, radio and video) production and dissemination through the Farmer Education and Publicity Unit (FEPU) and the sub-units (Zonal Communication Centres - ZCCs). Under NAEPII, ZCCs were only established in two zones with support given under TARPII responsible for establishing ZCCs in other zones. While the FEPU was better able to produce print media and in-house radio and video media, and also developed a greater capacity to disseminate media materials, the privatization of radio and television has increased charges for radio and television time, thus limiting use due to the budget. There is a need to further strengthen and extend these services. Under a decentralized system, these responsibilities should be retained at the central level. Modern IT equipment and methodologies should be introduced, and an information network developed to the district level. Monitoring and Evaluation (M&E). Assistance from NAEPII for the M&E Unit included staff training, some equipment and operating costs. The Unit produced a range of monitoring and evaluation reports, including quarterly reports monitoring progress on project implementation, annual reports on farm production (three major crops and three livestock indicators for each zone), annual impact assessment of extension messages, quarterly progress reports on communications outputs and bi-annual assessments on progress on research-extension-farmer linkages. The unit was also responsible for a number of other specific studies. However, relevance of reports to project M&E was limited by the inadequate identification of KPIs. Field extension workers gave informal assistance in collecting data. Under a decentralized system of extension, there is a vital need to put in place a formal arrangement for district extension services to monitor a set of indicators pertinent to their performance (activities and outcomes) that can be used in planning and managing district extension services. These activities should be integrated with those of the national M&E Unit, which would have its own priorities for information collection and evaluation, to that can be used for national policy and planning. Non-Public Service Extension Providers. Three Pilot Initiatives tested the use of CBOs and NGOs as alternative extension service providers. All three initiatives were able to mobilize farmers and improve agriculture and livestock production through introduction of selected technologies. In each case, they successfully identified and trained farmer motivators to support the initiatives, and established a farm input buying center under one initiative. However, as all initiatives were supported by donor funds, some doubt exists on the viability of this approach where external or public funds are not available. Future Development of Extension Services. From the experiences of NAEPII and the extension system that has now developed in mainland Tanzania, the ICR concludes that the further development of extension services in Tanzania requires: (i) removal of designated positions and functions for SMSs to become senior extension officers, and redefinition of WEOs and VEOs as extension officers; (ii) a change in organizational structure of the district extension service, from a hierarchical to flat structure whereby there is a small core of extension management staff at the district level with all other extension staff, including - 33 -

ex-SMSs, WEOs and VEOs, deployed as small extension groups at suitable locations within a district; (iii) commitment by the district councils to develop appropriate extension and transport facilities and sufficient operating costs to each of the extension groups in the district, (iv) a changing role of extension agents from advisor to facilitator; (v) establishing/strengthening linkages between district extension services within regions, between districts and zonal research centres, and between districts and communication centres; (vi) development of an M&E unit at district level with linkage to national M&E unit in MAFS; (vii) increasing control of services by farmers through cost sharing, farmers’ fora and advocacy; (viii) increasing use of contracted services; and (ix) a focus on knowledge provision, as well as technical advice. To enable the extension officers to carry out these functions their skills and knowledge have to be further broadened and enhanced in extension methodologies, facilitation techniques, problem identification and problem-solving methodologies, and to source and supply technological knowledge and information to clients. Zanzibar Institutional Considerations. Throughout project implementation, the extension services in Zanzibar remained merged with agricultural research activities under the umbrella of the Commission of Agriculture, Research and Extension, under MANREC. Administratively, the two islands of Zanzibar comprise 5 regions and 10 districts with 105 blocks, with each block assigned an extension worker covering 800-1,200 farmers. There are two main research stations, with sub-stations, and one training institute for both extension staff and farmer training. Extension System. A modified T&V extension system is used for provision of extension services. Abandoning the T&V system in Zanzibar should be considered, for the same reasons discussed above. Education and Training. Staff training was provided through refresher courses for Block Extension Officers (BEOs) and SMSs, and exchange visits and specific short/long training courses at institutions on the mainland and through limited overseas training. Formal training to certificate level was given to new recruits to the extension service to replace old untrained staff. Further needs-based assessments are required for continuing upgrading of extension staff skills and knowledge, and the need to retain/abolish the SMS cadre should be debated. Research-Extension-Farmer Linkages. Research-extension-farmer linkages were developed through the establishment and management of FFSs and establishment of Farmers Research Groups (FRGs). Through the FFS, researchers, extensionists and farmers collaborated in identifying, designing, planning, implementing and evaluating demonstrated technological packages. The FRGs were used to test on-station developed technologies under farmers’conditions and with their involvement. In addition, various workshops and technical meetings were held for researchers, extensionists, farmers and other stakeholders. The inclusion of both research and extension services within the one agency in Zanzibar naturally allows closer linkages than on the mainland. Communications. The agriculture Information Section was assisted in producing newsletters, leaflets, radio and television programs. Consideration needs to be given to linking this unit with corresponding mainland units to improve the production of extension media materials in Zanzibar. Monitoring & Evaluation. The M&E Unit collected project information and made periodic assessment of implementation progress of planned activities. Routine data collection was complemented by annual crop-based surveys. A range of monthly, quarterly, bi-annual and annual reports were prepared. A review should be undertaken to assess the appropriateness of information collected for M&E purposes.

- 34 -

Non-public Service Providers. Collaboration was established with NGOs and several donor-funded development projects to join efforts in agriculture sector development. These initiatives should be continued. Adaptive Research. The adaptive research sub-component focused on: (i) on-station and on-farm production, multiplication and distribution of disease free planting materials; (ii) on-farm adaptive research trials; (iii) germplasm collection and maintenance; and (iv) soil fertility enhancement. The on-farm multiplication of planting materials was very instructive to farmers in appreciating the multiplication process, as well as in observing the agronomic performance of different plant materials. On-farm adaptive research was conducted on cassava, yam, mango and sweet potato, and significantly higher yielding clones and varieties were selected. Soil fertility initiatives were concerned with introducing measures to deal with problems of low soil fertility and land degradation, and use of cover crops. Impact assessments indicated high levels of adoption of improved varieties and higher yields in selected crops with reasonable adoption of extension messages (65% of farmers). In relation to sourcing information, radio was easily the most important media, and farmer to farmer training was also an important means of information transfer. The close collaboration between research, extension and farmers needs to be continued and would benefit from closer relationships with research and extension services on the mainland.

- 35 -

Additional Annex 9. Experience of NAEPII Pilot Initiatives Component (The examples below are from "Lessons from Experiences of NAEPII Micro Projects", NAEPII, February 2004.) The Ministry approved 35 pilot projects (from 43 submissions) in three main categories: (i) extension management and farmer empowerment (6 projects, including FFS approach); (ii) role of private sector in extension services provision (8 including those on seed production); and (iii) technology transfer initiatives (21 projects). Twenty-three of these projects were externally evaluated by personnel from MAFS and Line Ministries at the end of 2003, to provide insights into their performance. In looking at the NAEP contribution, it is important to keep in mind that other factors, internal and external to the communities, may have facilitated or hindered the impact of interventions. These are mentioned whenever possible. The review focuses on the impact of interventions on: (a) productivity and production; (ii) markets, household income and livelihoods; and (c) other non-tangible benefits realized in the course of implementing the initiatives. Examples of Impact on Productivity Changes Intervention Area Farmer Field School Concept for Extension Management

Land Management and Mechanization

Before Project Situation

After Project Situation

Little knowledge of the technique among target farmers. Yields of maize below 4 bags/acre and paddy below 6 bags/acre. No FAMOs trained.

The use of IPPM in paddy increased by at least 34%, but mostly doubled.

Little knowledge and application of the techniques. Under-utilization of animal power. Crop yields relatively low.

-Use of animal-drawn implements up to 100% among group members

Yield of Paddy: doubled to over 12 bags/acres. Maize: three-fold increase to over 12 bags/acre 30 IPPM trainers produced as FFS facilitators and 750 FAMOs trained to reach other farmers on IPPM-FFS

Note -Some 12 FFS initiated by farmers on their own. -Up-scaled to 16 districts

Singida: MpambaaMATMP; Mbarali

-Increase in farmed area per family -In Singida, yields from mechanized farms increased by 63% (maize), 146% (sunflower), and 48% (sorghum).

-Mtwara MRIP; Water harvesting Paddy yields less than 3 Paddy yield increased to 6 bags (transplant and irrigation bags/acre. Poor condition only); 10-13 bags/acre (transplant, irrigated and Tandahimba-Chiu mo CBSTII; micro-dams and water fertilized) (Mtwara and Tandahimba) management

Adoption of high Lack of technology and value crops and less farming of high value horticulture for crops like mushrooms, extra-income tomatoes, onions and fruits. Livestock and poultry health and feed management

Farmers adopted cultivation of mushrooms and vanilla for income generation. More families with home gardens and fruit trees. Youth and women groups into gardening.

-Low milk yield, small - Chicken mortality significantly reduced (Songea eggs; light live and and Bagamoyo-from 90 to 20%) carcass weight of poultry -Livestock mortality down to 2% from 10% and livestock (Shinyanga, Monduli): -Livestock conception rate

- 36 -

Sumbawanga (mushroom) and Kagera (vanilla)

-Songea (Mwatrans) and Bagamoyo and Monduli CBESP; Dodoma RDTDC; Hai PHCAF;

-Conception rates up to 90% and calving rate up Kasulu ILCTCT low 50% and calving rate by 20%. about 60%. -Milk yield per local cow increased four times (to - High morbidity and 1.6 lts/cow) mortality rates due to poor hygiene and disease/pest -Morbidity from tick-borne diseases down from 35 attacks to 15% in Dodoma. Cleared trypanosomiasis in target areas (Kasulu) General improved agronomic practices

Low use of improved agronomic practices, resulting to low yields i.e. banana 8t/acre (Rungwe); paddy 25bags/acre (Mkindo)

-Increased adoption rate of improved agronomic practices among 80% of group members; FYM used by 50%, mulching by 65% and organic farming by 80% (KADEA), banana increased to 12t/acre (Rungwe), paddy up to 36bags/acre (Mkindo)

KADEA-AESFC P; Rungwe CBESIPP; Morogoro MRPTI

Abbreviations: IPPM=Integrated Production and Pest Management; CBESP=Community Based Extension Services for Pastoralists; MRIP=Mahurunga Rice Irrigation Project-Mtwara; CBSTII= Chiumo Basin Smallholder Traditional Irrigation Improvement; KADEA-CBAESF= Community Based Agricultural Extension and Sustainable Food and Cash Crops Production;

Dodoma RDTDC=Rehabilitation of Dips, Tanks and Disease Control; MATMP=Mpambaa Animal Traction MP; CBESIPP= Community Based Extension Services for Piggery Production; MRPTI=Morogoro Rice Production Technology Improvement; ILCTCTC=Kasulu Integrated Low Cost Techniques to Tsetse Control.

Examples of Effect on Income and Household Livelihoods Effects on:

Before Project Situation

After Project Situation

Marketing

Poor prices and lack of reliable markets for agro-products

- Formed/Strengthened Cooperatives, SACCOS and WUAs (Mbarali MPSIS and Manyoni, Sumbawanga for mushrooms) to produce and market crops - Started Crop Banks for store produce and sell when prices improve (KADEA)

Incomes

- Poor marketing and prices resulted in low and erratic income from farming - Low productivity also meant less surplus for sale - Poor crop husbandry led to poor quality products/low livestock weigh, and lower income - Lack of grassroots MFI meant no access to credit

General - Low farm productivity due to Livelihoods poor farming techniques, basic farm tools, poor breeds, and ignorance, leading to low levels and reliable income sources -poor housing and clothing;

-Savings per livestock group accounts grew up to TSh.0.75 mil (Bagamoyo CBESP), mechanization groups’ income also increased from both crops and hire services of implements (ox-plough, ox-carts, magoye-ripper) (Singida MATMP, Masasi-Chipuputa, and Mbarali MPSIS); Kongwa charge 3-5,000/acre cultivation; -Income from sale of fruits and vegetables (Igunga PCMRFV); -Livestock and poultry live weight improved, hence better selling prices -Members able to borrow thousands of shillings from their societies (Mbarali); -Improved nutritional status from vegetables and fruits (Igunga) -Reduced human labor due to use of draught power, and proximity to water sources (wells and dams) -Improved housing due to extra income earned -Improved health status due to cleaner water, balanced diets, and adequate food intake

- 37 -

-food insecurity, poor health -Improved knowledge and better understanding of their environment and coping strategies; -improved ecology and better micro-climate due to planted forests, agro-forestry and soil-water conservation; -Understanding of civic rights and responsibilities. Women and Unequal intra-household youth distribution of resources and income from jointly-invested efforts.

Incomes increased following sales of surplus milk, vegetables, and fruits. In e.g. Shinyanga women groups: household milk sales 7-10lts/day (Tsh.60,000 pm).

Food Security

Food security improved: e.g. surplus milk (produced up to 14 l/day) (e.g. Shinyanga WGET);

Women groups and SACCOS formed to offer unbiased services Institutionalized bias on access Women in committees and other group leadership to knowledge and skills in favor responsibilities of men Recurring food deficits and insecurity due to poor production and low incomes to cope with the shortages Imbalanced diets due to lack of food varieties to provide all the needed nutrients all the time

Food deficit months reduced by 1 month (e.g. Same CAEFS); Improved nutrition from balanced diets (e.g. Igunga PCMRFV)

WGET= Shinyanga Women Group Extension Training Micro-Project; MP=Micro-project; PCMRFV=Production and Consumption of Micro-nutrient rich fruits and vegetables; MPSIS Mbarali Mechanization Programme in Smallholder Irrigation Scheme; WUA=Water Users Association; SACCOS=Savings and Credit Cooperatives Societies; MATMP=Singida Mpambaa Animal Traction MP.

Examples of non-tangible benefits from pilot interventions Effects on:

Before Project Situation

After Project Situation

Institutional strengthening

Poorly-trained and inadequately-equipped field extension workers, farmer motivators, farmer field schools, and weaker community-based grassroots institutions

-Provided offices and transport to Extension Officers and FAMOs to work more efficiently (e.g. Rombo) -Oxenisation a priority in district plans (e.g. Kongwa) -Linking LGAs with other extension service providers and CBOs/NGOs in the district -Strengthened FFS and linkages with LGAs and CBOs -MFIs and other associations created at the grassroots

Community -Less organized community Empowerment efforts to prioritize and solve and problems Organization -Top-down approach in forming groups -Less-informed communities on available potentials to solve problems -Less knowledgeable on group

-Down-top approach in decisions to form groups and identifying areas of capacity building needed (e.g. KADEA AESFCP) -Dipping committees established and functioning (Dodoma RDTDC, Monduli and Bagamoyo CBESP) -Pastoral groups graduate to cooperatives (Bagamoyo) -Farmer groups and associations formed -WUAs and SACCOS formed and/or strengthened

- 38 -

dynamics, leadership skills and planning

-Farmers more confident with FAMOs services compared to FAEO (Kagera) -Farmers taking responsibility for the welfare of FAMOs

Networking: Vertical and Horizontal Linkages

-Little recognition by LGAs on the role of CBOs/NGOs;

-District Councils provide material support to CB Extension MP (Kagera); and dipping material (Dodoma)

-Less networking with other farmer groups within and outside district

-Networking with stakeholders beyond the district (e.g. Songea-Hanga village);

-Top-down initiatives in forming grassroots institutions Gender Less concrete strategies to mainstreaming implement national policies on women’s economic empowerment at grassroots

-Neighbor farmers to NAEPII CB Initiatives copy improved technologies (Mtwara Mahurunga RIP; Dodoma RDTDC) -Farmer groups graduating into associations and cooperatives (KADEA) -Reduced women labor burden due to mechanization -New technologies empowered women economically. Example: half of the group members are women (Kongwa oxen mechanization); more women than men interested in minor crops (mushroom in Sumbawanga) -Women empowered with leadership skills

Sustainability of farmer empowerment

-Top down approach denied communities sense of ownership -Poor mobilization and top-down perpetuated resource dependence syndrome.

Pluralism (Private Sector Role)

Single (public service) system of extension management and service delivery, with little recognition on the role of other stakeholders (non-government actors)

High level of awareness and appreciation of the importance of group action and pooled resources to solve community problems. Contribute in kind and cash in running the grassroots institutions. Established and democratically run institutions ensure stable leadership. Contributed and well-managed funds to ensure confidence by members.

- Small-scale farmers successfully engaged in improved seed bulking (e.g. 20 tons/year-Igunga and Karagwe SSFS) for sale to other farmers - Input supply system using private sector stockists and suppliers established (Dodoma DFICS); - Input suppliers also engaged in crop marketing (Dodoma) - Loan repayment for inputs almost 100% (Dodoma);

Note: Dodoma RDTDC=Rehabilitation of Dips, Tanks and Disease Control; FAEO= Field Agricultural Extension Officer; CSPDPA= Igunga Community-based Seed Production in Drought-prone Areas; DFICS= Dodoma Development of Farmer Inputs Credit Scheme; Karagwe FSFS=Seed Security for Food Security; CBESP=Community Based Extension Services for Pastoralists.

- 39 -

Additional Annex 10. Footnotes 1/ A pilot agriculture extension project was introduced in five regions in 1986, and expanded to sixteen regions under the National Agricultural and Livestock Extension Rehabilitation Project (NALERP) from 1989 to 1996. 2/ The DCA was amended in May 2000 (following the MTR) and in November 2001 (following changes in various ministries). LEGAF also prepared an Amended and Restated DCA, which incorporated earlier amendments as well. The amended DCA was not presented to the Bank’s Board. 3/ Michael Baxter, Roger Slade and John Howell, "Aid and Agricultural Extension--Evidence from the World Bank and Other Donors", World Bank Technical Paper #87, 1989. 4/ Madhur Gautam and Jock R. Anderson, "Reconsidering the Evidence on the Returns to T&V Extension in Kenya", Operations Evaluation Department, World Bank, April 1999. 5/ The Government ICR, undertaken by an independent consulting company, indicated some reservations that District Councils will allocate a sufficient budget for adequate extension services delivery, particularly with the removal of the crop and livestock cess from 2003/2004 FY.

- 40 -

Additional Annex 11. Final Beneficiary Survey, 2003 - Summary of Results Awareness and application of acquired knowledge The study found a large gap between knowledge and its application by farmers. In 1999, 17% of farmers surveyed confessed they were knowledgeable on the various improved practices for land husbandry. In 2003, this had risen to 26%. Less than 55% of the farmers used improved seeds of one crop or another. The practice is that they would use improved seeds in crops with a higher market price potential, and continue with local seeds for other crops, or in part of the farm for same crop. The proportion of farmers using improved seeds and applying manure/fertilizer increased from 48% and 23% in 1998, to 55% and 60% in 2003, respectively. However, there is still a big challenge in encouraging farmers to use agro-forestry, soil erosion control measures and integrated pest management, which are still practiced by a very small proportion of the farmers. Crop production The most common cereals grown by farmers are maize, paddy and sorghum. Cassava and banana are also important crops. These are grown for both food and selling for cash. More than 50% of the farmers grew maize as the main crop. Non-food cash crops are coffee, sunflower, cotton, onions, tomatoes and citrus. There has been a gradual increase in crop yields in normal weather years. However, yields in 2002, used in this survey, were lower than average due to country-wide drought. According to the interviewed farmers in maize-suitable districts, they usually obtain an average yield of 1,200 kg per hectare, which is an improvement compared to an average yield of 600-700 kg per hectare reported in 1999. The average level of yields attained by most farmers is still very low compared to the potential levels using very basic crop husbandry practices. Although there is a significant proportion of farmers attempting to use various technological packages, the impact on overall production is limited, due to the fact that: A small proportion of farmers apply improved practices; • Even those applying improved practices tend to selectively apply to parts of the farm, (or, in the • case of livestock, retain old unproductive breeds while keeping a few improved cattle or goats like a show case to other farmers). Livestock production Although on average there are 3 cattle per household, there is one 1 dairy cow for each 4 households. Ownership of livestock is also highly skewed, whereby only 27% of households had cattle. Less than 15% of the respondents possessed at least one improved dairy cattle breed. Milk production per cow, from both improved and local breeds, is also still low, especially for local breeds, bringing the daily average to 0.9 litres per local cow and 5 litres per improved cow. Food Security Improvement Proxies Despite the impressive high level of awareness among farmers on the required husbandry practices and yield-boosting technologies, many constraints continue to prevent widespread adoptions and application of such technologies for higher production levels. Consequently, slightly more than 50% of the households experienced apparent seasonal deficit of main staples ranging from 3 to 6 months each year. Most of the foodstuff consumed is starch, with more than 61% and 55% the households having protein foods as part of lunch or supper, respectively. - 41 -

Adequate and balanced food intake has a positive bearing on the health status of an individual. The survey indicated that 60% of the households had at least one member who fell sick and had to seek medical help during the previous 12 months. Only 15% of them could not afford formal treatment due to lack of money. Income Improvement Proxies According to this study, there have been some positive changes in indicators related to quality of housing, ownership of capital-intensive assets, expenditure levels on household basic needs. The proportion of households owning production tools such as ox-ploughs increased from 11% in 2001 to 17% in 2003. Those owning bicycles and transistor radios increased from 55% and 52% in 2001, to 65% and 82% in 2003, respectively. As for the quality of housing, in 2001, 49% of the respondents had houses with iron-sheet roofs. In 2003, this was 64%. Expenditure on non-food items such as education and farm implements also increased. The percent increases in these categories between 2001 and 2003 were 8.2 and 12.6 respectively. A better measure of increased purchasing power would have been quantities purchased at cost adjusted by the inflation rate (unit costs at real prices). Either way, it could be argued that if prices go up and farmers continue spending the same amount or more, it implies a state of increased farmer incomes, which could be attributed to successes in raising productivity and/or total production. Trend of adoption rates (%) for improved technologies 100% 90% 80% 70% 60% 50% 40% 30% 20% 10% 0%

1998/99

1999/00

2000/01

2001/02

2002/03

Land Preparation

63%

65%

68%

71%

77%

Improved Seeds

48%

48%

50%

50%

55%

Plant Spacing

41%

45%

55%

59%

72%

Weeding

70%

69%

76%

66%

86%

Manure Application

23%

24%

21%

22%

60%

- 42 -

Additional Annex 12. Executive Summary of Government of Tanzania ICR

EXECUTIVE SUMMARY I.

Background to NAEPII

1. The National Agricultural Extension Programme Phase II (NAEP II) was implemented in 20 of Mainland Tanzania (1996/97 to December 2003) and all regions of Tanzania Island (Zanzibar) (from 2000). The approach of NAEP II is basically a modification of the T&V system, seeking to achieve its development objective of increased farmer income and enhanced food security by developing a participatory, demand-driven extension system; developing appropriate technological packages for low-income farmers; strengthening the role of farmer-based organisations in input supply and enhancing the participation of farmers and extension staff in training programmes. 2. The project’s development objectives contribute to Government of Tanzania’s efforts to achieve sustainable and equitable economic growth as part of poverty alleviation strategies. It is operated within the general framework of rural development and agricultural sectors development strategies already initiated and with ongoing programmes such as the Agricultural Sector Development Programme. NAEPII contribution to improving demand-driven extension services delivery systems and linkages with technology development was effected within the context of empowering communities through their local governments under the Local Government Reform Programme. 3. This report revisits the goals and expected outputs of NAEPII in terms of physical achievements, impact of the interventions and sustainability, factors that affected implementation, lessons learned and recommendations. NAEPII Components 4. The programme had originally four main components, with a total budget of US$32.9, including Government contribution of US$1.8 million. The budget was later in 2000 revised to include Zanzibar Islands, such that the revised budget allowed for US$2.5m to be allocated for Zanzibar out of IDA’s budget of US$31.1 m. Tanzania Mainland a) Institutional strengthening for the reorganization and rationalization of the extension services; originally budgeted at US$15.7m. b) Extension education and training for staff, farmers and training institutions; originally budgeted at US$12.6m. c) Communications support for decentralizing information management; originally budgeted at US$2.9m. d) Pilot initiatives to improve extension management, technology transfer to farmers, farmer-extension-research linkages, pluralistic extension delivery methods, farmer empowerment, and gender impact. This was budgeted at US$ 1.7m. Government of Zanzibar e)

Zanzibar Component of NAEPII: Originally not included in the budget but later allocated US$ 2.5

- 43 -

m obtained from savings or cuts from various components for Mainland Tanzania as proposed in the SAR. GOZ contribution was US$0.225m. After revision, the amount of funds available for Tanzania Mainland for the four components was less by US$2.5m. Funds utilisation 5. By the end of November 2003, actual IDA disbursement for NAEPII in Mainland Tanzania was US$ 26,176,198 (97.6% of the budget), while US$1,963,695 had been disbursed for NAEPII in Zanzibar Islands (78.5%) utilisation. The contribution by Mainland Tanzania reached US$1.816 million (67% of the budget), while that of Zanzibar Government was US$0.125 million (78.6% of the budget). 6. Almost all of the planned procurement of goods (e.g. vehicles and civil works) and services in both Tanzania Mainland and Zanzibar were successfully undertaken as planned, despite some delays in some of the items due to either technical specifications or tendering procedures. II.

COMPONENT IMPLEMENTATION PROGRESS

Key Areas of Project Success: 7. Despite the administrative setbacks occasioned by the dynamics of decentralisation of local government systems and their roles in running and supporting extension services in the country, the project on the whole managed to record significant progress in building the capacity of extension agents, strengthening of extension delivery system and collaboration with other institutions, including research, academic/training institutions and the private sector. Some notable achievements are summarised below: Key area of achievement

Collaborators

1. Development of Extension Vision 2010 and Plan of Implementation Plan by 2000

Key Line Ministries, Local Governments

2. Developed a special curriculum for a Diploma in General Agriculture, Higher which was adopted by training institutions by 2001 Institutions

Learning

3. Prepared modules for PRA which were used as standard tools for training Higher and application in the field by 2001 Institutions

Learning

4. On-farm trials, bulking and distribution of disease-resistant maize seeds for TARPII and Uyole the Southern Highlands zone to combat maize blight which was contained Agricultural Research through collaborative efforts. Centre 5. Initiated multiplication of foundation seeds using villagers organised in Research Stations and groups before distributing to other farmers for planting Villagers and District Leadership 6. Farmers empowered to determine research agenda through FRGs and FFS, Research institutions thereby enabling resource commitment for solving farmer-identified problems and Farmers 7. Built the capacity of Information and Communication centres and offices to Training institutions effectively gather, repackage and disseminate extension information and private sector

- 44 -

8. Built the capacity of the majority of extension workers and officers through Higher Learning refresher courses and upgrading of professional qualifications (undergraduate Institutions and LGAs and post-graduate studies). Under-qualified and untrainable staff were retrenched. 9. Sponsored extension workers and officers on study tours and short courses Higher locally and overseas e.g. Brazil (on zero tillage techniques), Philippines (on Institutions PRA techniques), United Kingdom (Radio programming and social marketing).

Learning

10. Constructed houses and offices for Extension Workers at Ward, Division Private Sector and and district levels District Leadership 11. Facilitated extension officers and frontline extension workers with Private sector extension kits and transport facilities (motorcycles and bicycles)

8.

Below is a synopsis of what has been achieved under each component of the project.

Institutional Strengthening 9. The output on extension institutions strengthening was characterised by activities aimed at creating participatory and consultative decision making mechanism through management committees; streamlining the quality and deployment of trained manpower, strengthening linkages between farmers, extension agents and researchers, and establishing a cost effective monitoring and evaluation system. 10. Management organs for NAEPII were established and operated in both the Tanzania Mainland and the Islands of Zanzibar. At the national level, they included the NAEPII Project Steering Committee (PSC), whose equivalent in the Islands was the Zanzibar Project Technical Committee (ZPTC), which co-opted some members of the NAEPII Project Steering Committee. There was also a Project Monitoring Unit (PMU) Liaison Office in Pemba to coordinate NAEPII operations in the sister island. 11. In the Mainland, below the PSC were the District Extension Steering Committees (DESCs) which operated until 1999/2000 when their roles were supposed to be taken by the District Management Teams (DMT) and Ward Development Committees (WDC) under the Local Government Reform Programme (LGRP). Unfortunately the DMTs have failed to effectively play the role of DESCs and therefore left a vacuum in the districts in terms of providing management guidance. 12. In realising the role of Local Government Authorities in sustainable management of extension services, deliberate efforts have been taken to involve, at the national level the PO-RALG in the management of NAEPII, including its active participation in the NSC. At the local level, Council personnel have been included in training and study tours sponsored by the project. However, despite this support, there is little evidence to date, except for a few districts, to indicate that LGAs are taking a pro-active role to financially support the management and delivery of extension services. 13. Rationalisation of qualified and competent staff at different levels went on well during the first three years such that all SMSs were placed under the District Extension Office. However, plans to deploy SMSs at the Division level were revised following the decentralisation process where the Ward took a more important role than before. However, 900 personnel assessed as under-qualified for professional upgrading were retrenched to give way for a new crop of trainable personnel. 14. NAEPII resources were also used in both the Mainland and Zanzibar to facilitate enhanced collaboration with different stakeholders through meetings, workshops, study tours and exchange of germplasm for different crops. External collaborating institutions included KARI, IITA, SARRNET and CIP.

- 45 -

15. Farmer-based grassroots institutions such as Farmers Research Groups (in Zanzibar) and Farmer Field Schools (Mainland) were encouraged through trainings and granting demonstration materials. Support activities in this area were carried out more or less as per plans. 16. Also established and supported as part of institutional strengthening were Monitoring and Evaluation Units at Ministries’ HQs in Dar es Salaam (MAFS) and Zanzibar (MANREC). However District support for M&E routines in the Mainland were somehow disrupted following the dissolution of DESCs in 1999. Despite the hiccups, it was still possible to train SMSs and other professionals on computer use and also involve them in routine reporting and undertaking of various studies and surveys. 17. In order to ensure that piloted extension methodologies, anchored on the principles of participation and demand-driven system, MAFS prepared a National Extension Vision 2010, which incorporates those cardinal principles. Various stakeholders, both national and district-based participated in its deliberation. Extension Education and Training 18. The training plans for staff and farmers, involving short and long courses, workshops and study tours, were guided by results from a Training Needs Assessment. A Training Coordination Committee (TCC) in the Mainland was in-charge of the training plans, and included members from other line ministries’ departments such as Adult Education and Community Development. 19. Besides providing the needed knowledge and skills to extension staff, the sponsored trainings in both Tanzania Mainland and Zanzibar gave a boost to the local training institutions in terms of enhancing their technical competence as centres of excellence and utilising their capacities to the full. For example, 18% of the registered M.Sc students at Sokoine University of Agriculture between 1996 and 2000 were sponsored by NAEPII. The trained manpower was also readily available for consultation by the private sector. Training in Tanzania Mainland 20. Extension personnel were trained at four different levels: refresher courses, certificate courses, diploma and first degree courses, and advanced diploma or master’s degree level. The staff also attended Bi-monthly and Quarterly Trainings, which involved more than 5,600 sessions, attaining 61-75% of the target. 21. About 20 and 54 staff attended graduate (bachelor) and post-graduate (masters) courses, which were 50% and 84% of the targets, respectively. Another 405 and 650 were sponsored for Diplomas in Agriculture and Diploma level refresher courses, which were 91% and 81% of the targets, respectively. Short courses, which involved more than 1,690 women and men, were conducted almost according to set targets. Likewise, 75% of the planned study tours, involving more than 1,070 people, were conducted. 22. As for farmers, more than 44,450 farmers (82% of the target) participated in various training programmes in the form of residential courses, one-day on-site trainings and study tours. Training in Zanzibar 23. Kizimbani Agricultural Training Institute (KATI) was earmarked to offer certificate level courses to about 123 Block Extension Officers (BEOs). Another batch of 164 BEOs and 29 SMSs attended refresher courses. There were also short and long courses for extension and research officers locally and overseas. These courses were conducted more-or-less according to plans. 24. More than 3,300 women and 2,700 men attended residential courses (82% of the target). Others (1,943 women and 2,507 men) were also facilitated to attend on-farm demonstrations, field days and study tours. The attendance was above target. Communication Support

- 46 -

25. NAEPII took lessons from the experiences of NALERP in deciding to facilitate the establishment of Zone Communication Centres (ZCCs) based on the seven agro-ecological zones. The main reason was to have information centres which responded to area-specific needs and requirements of farmers. 26. By the end of NAEPII, plans to equip the ZCCs were undertaken as planned in 3 zones (including one inherited from SHERFS Project) and partially fulfilled in 4 zones. The delays in evaluating the experiences from the initial 2 pilot zones (Eastern and Northern) postponed commencement of interventions in the other zones until 1999 when partial support was started. Staff training, involving more than 18 professionals from FEPU and ZCCs, was provided locally and abroad in script preparation, journalism and communication skills. 27. Plans by FEPU to produce magazines, newsletters, booklets and posters/leaflets were attained by between 74% and 90% of the target. As for the ZCCs, their plans were fulfilled by between 65% and 67% for newsprint and 42% for video titles shows. 28. There were 19 different video titles produced in-house and using hired professionals covering various topics. FEPU also produced and aired a total of 1,040 radio programmes. However, the use of TV was limited to Dar-es-salaam and Morogoro only. 29. In Zanzibar, all the planned number of newsletters were printed and distributed (2,000 copies), but aired 62% and 56% of the planned radio and TV programmes, respectively. Pilot Initiatives 30. More than 25 micro-projects were initiated in Tanzania Mainland, with the aim of improving extension management, technology transfer methods, enhancing inter-linkages between farmers, extension agents and researchers, encouraging alternative extension service providers, and piloting on community-based seed production and input supply. 31. In developing alternative extension services delivery methodologies, more than 9 micro projects were initiated and supported, where Farmer Field Schools (FFSs), Farmer Motivators (FAMOs) and Paravets took centre-stage in piloting farmer-based extension delivery system. Several Farmer Groups, CBOs and NGOs have been formed and registered and successfully operating in different districts. 32. Some initiatives under NGOs and CBOs, including church-based NGOs, were tested as alternative extension service providers. Typical examples are Mogabiri Farm Extension Centre, MWATRANS Farmers Association in Songea and Catholic Diocese in Shinyanga. The initiatives have shown that it is feasible to develop farmer-based institutions as alternative service providers, given the right formulation and guidance in the early stages of formation. 33. Among the micro-projects initiated to test effective ways of technology transfer include on-farm testing on (i) fertilisers and manure application (ii) water harvesting and irrigation farming (iii) oxen-mechanisation (iv) livestock and poultry disease control by dipping and vaccination (v) Luciana production and utilisation (vi) mushroom production and (vii) vanilla production. The initiatives created chances for farmers to test not only the productivity advantages, but marketing potentials and investing in alternative sources of income generation from the surplus income. The activities also acted as demonstration centres for other farmers to learn and emulate. 34. The various micro-projects had the parallel advantage of mainstreaming gender concerns as part of economically and technologically empowering men and women in rural communities. In all the undertakings there was provision for ensuring minimum participation of women and youth. 35. Although there was no formal budget for the equivalent of pilot initiatives in Zanzibar, but in essence Farmer Extension Groups and Farmer Research Groups played a similar roles as farmers participated in testing various technologies and exchanged ideas and experiences with other stakeholders. Adaptive Research (Zanzibar) 36. Adaptive research initiatives gave a chance for farmers to participate in seeking solutions for profitable farming. The initiatives, undertaken both on-farm and on-research stations focussed on: - 47 -

a. multiplication and distribution of disease-free quality planting materials (banana and tubers-yams, potatoes, cassava) on a total of 25.5 ha of small experimental plots in Unguja and Pemba. FFSs, FRGs and individual farmers were later on involved in multiplication of healthy materials in 21 sites. b. On-farm adaptive research trials were undertaken for screening for the best among four local yam varieties, evaluating six local and two improved cassava varieties for yield and pest resistance, top-working on 600 mango plants for export quality, and screening the best among local and improved sweet potatoes. The results were jointly evaluated between farmers and researchers and recommendations made. c. Germplasm collection and maintenance of genetic resources for local and improved food plants. Germplasm accessions were obtained for cassava (52 local and 6 improved), sweet potatoes (40 local and 11 improved), yams (30 local and 4 improved from IITA) and banana (37). d. An action plan for Soil Fertility Initiatives was formulated and approved; and also Cover Crop Seed Bulking programmes were established in both Pemba and Unguja (2 acres each). III.

EFFECTS AND IMPACT OF NAEPII INTERVENTIONS

Institutional Strengthening 37. By encouraging a group approach in interactions for extension delivery, NAEPII was in essence contributing to the nurturing of basic building blocks for any grassroots institutions. Farmer Extension Groups and Farmer Research Groups and Farmer Field Schools have the potential to be transformed into other economic and social groups such as Producer Groups, Crop and/or Input Marketing Groups or Associations, Primary Cooperatives and Input Marketing Groups. 38. Group members were taught on group dynamics, rules and procedures and management skills. This must have contributed to infusing a democratic culture, which is an essential part of good governance and transparency in the civil society. There are reports that group leaders and FAMOs for example, are already recognised as part of the village leadership and are invited in village meetings. 39. Skills are taught on keeping records of group information (production and meetings). These could be used individually as skills for profitable farming and other business enterprise. 40. At the national level, the Steering Committees drew members from different departments. This encouraged inter-ministerial collaboration in finding solutions to cross-sectoral constraints, and having a coordinated approach in mobilising resources and interventions for other problems as well. The co-option of NAEPII National Steering Committee members in the Zanzibar Project Steering Committee contributed in cultivating joint decision making between GOT and GOZ, thus strengthening the Union. 41. The project supported linkage activities through meetings, workshops and exchange visits, helped inter-institutional knowledge exchange and improvement. 42. Monitoring and Evaluation reports provided the required feedback to NAEPII management which was the basis for critical decisions on activity implementation and offering policy directions. 43. It was during NAEPII when GOT realised the need to come up with a Vision and Extension Framework Tanzania Mainland. The future for extension management, which is demand-driven and participatory, poses a challenge to LGAs to look into ways of supporting exchange services. 44. The trained manpower under NAEPII added to the pool of professionals in the country, which could be sourced by other public and private bodies to enhance production and processing, all part of poverty alleviation efforts. Changes in farmers’ knowledge, attitudes and skills 45. Various activities supported by NAEPII helped to impact on farmers’ knowledge, attitude and skills on improved farm management practices. These included individual and group trainings, on-farm trials, FFS, FRGs, field days and demonstrations, newspapers/magazines, radio and electronic educational programmes and news. The main source of knowledge is still through messages from extension agents, mainly in group interactions. - 48 -

46. Several studies conducted during the course of NAEPII implementation indicate that there has been a steady increase in the interactions between farmers and extension agents and/or extension information sources, such that awareness and knowledge on proper improved practices for key crops and livestock is greater than 90% among farmers. 47. However, whereas there is success in infusing knowledge among farmers, there is still a large gap in putting that knowledge into practice, and therefore getting the right skills, by many farmers as illustrated below. Adoption rates for improved practices and yield changes 48. It would appear that even among various techniques necessary for improving productivity of crops and livestock, farmers have a tendency to adopt some faster than others. Even though, the proportion of adopters is far less than those aware of the technologies. 49. Among the practices which NAEPII appear to have made the largest impact in adoptions over years is the importance of plant spacing and use of improved seeds. Though improved seed is also selectively adopted depending on the profitability of the crop, farmers practising proper spacing and using improved seeds (of one crop or another) increased from 41% and 48% in 1998/99 to 72% and 55% in 2002/03, respectively. This indicates that even when using local seeds, they use proper spacing. 50. As expected, even though adoptions have been selective, those practising improved technologies have managed to substantially increase crop and livestock yields over time, with annual fluctuations due to vagaries of weather. Paddy appear to have made the largest gain in yield, mostly due to the adoption of water harvesting techniques, use of seedbeds before transplanting, proper plant spacing and use of improved varieties. According to 2000 survey data, paddy yield had increased by 184%, and that for beans by 139% and maize and sorghum by 33% and 63%, respectively compared to baseline year of 1996. The yield results for 2002/03 were slightly lower because of draught which affected most districts. 51. Results from sample farmers in Zanzibar also indicate the same trend, where adoption for improved seed/planting material is highest for cassava and yam varieties (64% used improved varieties), followed by potatoes (60%). By applying FYM, practising farmers have doubled cassava and banana yields to 17t/ha and 15t/ha, respectively, and tripled potato yields to 25t/ha. The reasons given for non-adoption of improved technologies such as use of FYM and improved planting material include costs associated with the technology (e.g. bulkiness of FYM and transport costs) and labour supply constraints. Changes in farmer’s income and livelihoods 52. Results from Beneficiary studies for 1999 and 2003 indicate that more than 95% of farmers relied on both crop and livestock husbandry as their primary sources of income. Changes in income levels can better be explained by looking at indicators related to quality of housing, ownership of capital- intensive assets and expenditure levels in household needs. Time series studies for the period of project implementation show that the proportion of household owning quality houses (with iron sheets) increased fro 49% in 2001 to 64% in 2003. Those owning productive tools such as ox-ploughs increased by 11% in 2001 to 17% in 2003. Those owning bicycles and transistor radios increased from 55% and 52% in 2001 to 65% and 82% in 2003, respectively (NAEPII Beneficiary assessment study, 2003). IV.

SUSTAINABILITY OF THE ACHIEVEMENTS

Institutional Strengthening 53. Project organs at the national level, such as the Steering Committee can only be sustained if all stakeholders contributing to the Committee appreciate the project tasks as common challenges. However, the dedication of DMTs at District levels was lower than that of the dissolved DESCs, due to the multi-sectoral responsibility of DMTs. 54. The creation of grassroots farmer-based institutions to coordinate both demand for, and provision of extension services, appear to be the key to long terms sustainability of a demand-driven extension - 49 -

services. However, the development of this system will depend much on the appreciation and political of civic leaders at the ward and district levels. At the moment, there is little evidence that Local Councils are willing to mobilise and allocate resources to adequately extension services delivery. 55. Farmer groups, including those operated as FFS and FRGs have institutionalised leadership structures and election process, including collection and management of funds. These are core features for good governance which contributes to sustainability. However, there will be still need to continue providing technical and management backstopping to such groups before they mature and transform into viable economic groups. 56. However, the system of voluntary FAMOs and Paravets is doubted by some if it can be sustainable due to the long term opportunity cost of time foregone by the volunteers. The LGAs should play an active role in partially offsetting the costs of supporting the FAMOs and Paravets from members’ funds. 57. The collaborative meetings with other stakeholders which were held under NAEPII depended on the project’s budget. No plans were made to create self-sustaining networks. It’s doubtful if such collaborative meetings will continue beyond NAEPII. 58. One of the common pitfalls in most Monitoring and Evaluation systems is the erroneous assumption of taking M&E tasks as exogenous to the normal duties of extension managers, SMSs and frontline staff. As such personnel take no keen interest in understanding the systems and what key indicators to be collected and reported. It’s always left to one or two people to deal with M&E system. The system of approaching beneficiaries only during case studies to “gather data” is least sustainable compared to the benefits of involving beneficiaries and designing a participatory self-monitoring system. 59. The participation of key stakeholders in the development of key features of Vision 2010 and Extension Framework ensures that it can be continuously revised through regular consultations in the course of implementing its strategic interventions. Extension Education and Training 60. The sustainability of the trained manpower can be viewed from two angles: that of maintaining the staff as government of employees or self-sustaining in private sectors contracts or self-employment as farmers or advisors. With the budgetary constraints of the LGA, and indeed the Central Government, it is doubtful that they could be sustainably employed at well motivating packages compared to the private sector, provided infrastructure and market conditions are improved to allows for a spontaneous growth of agriculture production and agro-processing. 61. The training institutions had an opportunity to improve on their infrastructure and teaching syllabi, thanks to NAEPII. Lecturers and tutors got chances to supervise field-tested extension professionals, an exchange which is desirable in upgrading their professional competence. It is hoped that this experience will be passed on to other students joining the institutions in the future. 62. More than 50 officers were sponsored for post-graduate studies (M.Sc), which involved research work as part of postgraduate theses. The results of from these scholars could selectively form agenda for further actions or implementation by government. However, there was no forum at the ministry level to receive these recommendations and debate their suitability for further action. 63. The trained manpower has knowledge and skills which could be tapped by the private sector. Information and Communication Support 64. The Information Centres/ZCCs are operated as government departments, providing highly subsidised services, whose sustainability depends on the budgetary allocation. At the moment, the LGAs are still less eager to contribute towards the operation of Zone Research Centres, citing lack of financial resources. 65. Historical dependence on government-owned media made it possible to have subsidised information print media and radio programme. However, with the advent of privately owned newspapers and radio/TV, the historical budget could not sustain reaching out to these alternative channels. Indeed, even - 50 -

government-owned radio and TV stations now charge commercial rates, squeezing even further resources for information dissemination. 66. However, it is expected that the trained manpower provides a useful professional pool to be sourced by NGOs and the private sector involved in education and information dissemination. Pilot Initiatives for Extension Methodologies 67. Most of the initiatives, though quite successful, are centred on donor-funded projects, raising doubts on their sustainability without the subsidy provided in their undertakings. 68. The profitability and sustainability of seed multiplication and input supply ventures depends much on the number of customers for the inputs. Moreover, the viability of improved seeds produced from foundation seeds depends on quality control back-up supposed to be provided by the Crop Inspectorate units at the district level. The efficiency of such services will depend on the commitment of LGAs in financing agriculture. 69. For the farmer-based extension service providers, their viability and sustainability will depend on the willingness of farmers to pay for services provided by professional extension agents, FAMOs and Paravets. 70. FAMOs and Paravets were provided with loans for bicycles as an incentive for their voluntary services. This system is likely to sustain if LGAs recognise their roles and allocate funds for their support. Adaptive Research 71. Since extension services are still subsidised, their important roles of Research Stations in spearheading R&D will continue to depend on government budgetary allocations. However, the long term solution is to facilitate a private-sector led high value agricultural production, agro-processing and exportation, which could then afford to pay for the costs of R&D. The cases for mangoes and spices in Zanzibar are pointers towards this direction. 72. As with the cases of pilot initiatives in the Mainland, the success in technological adoptions depends on the potentials for profitable farming. This in turn is depended on type of enterprise promoted and market potentials. V.

LESSONS LEARNED

Institutional Strengthening 73. Institutional guidance is likely to be better provided by task-specific organs, like was the case with DESCs, compared to multi-sectoral supervising organs like the DMT. 74. Despite the noble-goal of decentralising extension services delivery to the districts, there is an appreciation-gap among civic leaders on the importance of extension services to agricultural productivity and income generation. The excuse given is lack of funds to finance to extension, despite agriculture providing the bulk of district wealth/revenue. 75. There were opportunities which were not adequately exploited in transforming some of the stakeholders’ meetings involving people with common interests into regular fora or networks for exchanging information and views on extension, research and farmer-extension-research linkages, marketing and pressure groups for input supply improvements. 76. Though it is formally acceptable and institutionalised, and tends to work given resource availability, a top-down M&E system is not sustainable for development process monitoring. 77. It’s good practice to involve all stakeholders in developing important policy documents such as the Vision and Extension Framework, 2010. However, the process would have been more complete if more consultations were done at the grassroots level. Extension Education and Training 78. The trained manpower, including FAMOs and Paravets provide important human resource pool not only to government and projects, but to NGOs and private sector operators. - 51 -

79. Given the opportunity and resources, local training institutions can be transformed to centres of excellence. However, their sustainability still depends on government and project funding to sponsor students. 80. There was a lost chance in ensuring that research findings by Masters-degree students were adequately synthesized for possible action plans. Information and Communication Support 81. Farmers have over time diversified their sources of information for improved agricultural production. There is an increasing role of private FM stations and newspapers in providing cocktails of educational and entertainment dossier. 82. There is a need to seek for more budgetary allocations so as to diversify information outlets and repackage information to appeal to a new generation of audience. Pilot Initiatives for Extension Methodologies 83. Although most farmers are knowledgeable, adoption of improved technological packages is constrained by, among others, structural and infrastructural deficiencies characterised by poor accessibility to rural areas, leading to poor and untimely input supply system and uncompetitive producer prices; lack of affordable credit facilities; and ignorance on enterprises with highest comparative advantage in different agro-ecological zones. 84. Though most of the pilot initiatives seemed to produce promising results, there is still a feeling that they require more time to mature and graduate into self-sustaining grass-roots farmer-led initiatives. 85. Despite the promising indicative results from the micro-projects, it is still doubtful if LGAs would go a long way in supporting them in the absence of projects which are currently collaborating with the District Councils. Adaptive Research 86. The choice of enterprises to be tested for adoption is not based on sound cost-benefit analysis and comparative advantages of different agro-ecological zones. 87. Some of the promising enterprise which farmers could easily adopt and grow profitably, lack a clearly defined marketing system and support policies. 88. The use of FFS and FRGs have strengthened linkages between farmers, extension agents and researchers, and therefore easing the learning and adoption processes. 89. Use of participatory methodologies has empowered farmers in the decision-making processes. VI.

RECOMMENDATIONS

90.

Drawing from discussions above, the following can be recommended:

Institutional Strengthening 91. Managements organs should be created with more specific responsibilities, rather than depending on those charged with multiple or cross-sectoral tasks. 92. More efforts should continue to raise awareness among district leaders on the importance of, and their role in promoting a demand-driven extension services funded from local resources and diversified to spread the costs of delivery. 93. Some efforts should also be taken to learn lessons from experiences within the country where cost-sharing extension service delivery has been practices by NGOs and projects. 94. A participatory, beneficiary self-assessment M&E system should be encouraged in future interventions. Moreover, managers and technical personnel should be re-oriented on the basics of an effective participatory M&E system. 95. The Government should include milestones for regular stakeholder reviews of the implementation - 52 -

of strategies under the Extension Framework 2010. 96. All factors remaining constant, an effective extension system with positive results from acquired knowledge and skills, will ultimately depend on reliable market for inputs and products, which calls active collaboration of all stakeholders. That is why LGAs will have to take more centre stage in agricultural management and development. 97. The Government of Zanzibar might consider preparing its own Extension Vision and Framework for the Islands. 98. In financing arrangements, there was concern that the size of Special Account was not big enough to allow for a faster disbursement of funds. Extension Education and Training 99. The supported training institutions should be assisted to come with long-term student recruitment and/or re-training plans from the private sector, instead of depending on government-sponsored students. 100. Future extension materials should be more integrated, bringing in other life skills and messages such as HIV/AIDS, gender mainstreaming, marketing and enterprise management. 101. In future, there should be a forum for receiving and synthesising research findings from sponsored students in higher learning institutions. 102. Professionals should also be given skills of running private enterprise or NGOs so that they are ready to provide their expertise upon retirement. Information and Communication Support 103. More budgetary allocation to Communication/Information Centres to take advantage of the alternative mass media outputs for information and education dissemination. 104. Possibility of multiplying for commercial sales some interesting video titles should be explored. Pilot Initiatives for Extension Methodologies 105. More investment of resources in opening up rural areas for ease of accessibility to allow cost effective delivery of inputs and more competitive procurement of produce. 106. The use of FAMOs and Paravets as part of national extension delivery system should be considered. Meanwhile, an informal incentive system should be supported by beneficiaries so as to motivate the volunteers. 107. Efforts to be taken to develop village-based input supply systems, starting with farmer economic groups and private shop owners. 108. The potentials and benefits from a private-operated extension system are enormous. More work at policy level should be done to encourage that system. Adaptive Research 109. More work to be done in advising farmers on the comparative advantage of different enterprises from a cost-benefit angle based on evolving market conditions. This way, efforts on on-farm research and trials will be cost effectively invested in the most profitable enterprises by ranking. The recommendations above will work better if there is coordination in weather information gathering, forecasting, and crop/livestock monitoring early warning.

- 53 -

Additional Annex 13. Summary of Revolutionary Government of Zanzibar ICR This document represents the terminal report of the National Agricultural Extension Project- Phase II (NAEP II), Zanzibar Component which was implemented from May, 2000 to December, 2003. The Project was originally planned to close in December, 2002 but IDA granted one year extension to complete planned activities. A total of US $ 2.5 million was earmarked by the IDA towards implementing the project in Zanzibar. As counterpart fund contribution, the Government of Zanzibar was required to provide an equivalent of US $ 100,000 for the entire project period. The main objective of the project was to support the Government’s objective of increasing food production and farm incomes through building the capacity of the institutions responsible for generation and dissemination of agricultural technologies and promotion of participatory research and extension in agriculture. Emphasis was targeted on resource poor farmers, including women. The major food crops which include cassava, sweet potatoes, yams, and pulses were the candidate crops for this project. Basically, IDA support through NAEP II was to enable the Government of Zanzibar to complete critical activities left behind unfinished by the phased out IFAD- funded Smallholder Support Project (1991 – 1997) in areas of adaptive research, institutional strengthening, and extension activities. As of November, 2003, the project had managed to spend a total amount of US $ 2.0 million (79% of the allocated funds) from IDA and the Government of Zanzibar had during the period released US $ 125, 030 (125% of the commitment) towards implementing the project. Achievement of the project include completion of planned civil works at Kizimbani and Matangatuani, which include construction of seven residential staff houses, a student dormitory (to accommodate 50 students), establishment of training farms (poultry and cow sheds, fencing paddocks and rehabilitation of a milking parlour and cattle dip). Besides civil work construction, a number of vehicles, equipment and goods were procured as planned. In the course of project implementation, a total of 144 staff were trained in various disciplines both locally and regionally. The purpose was to equip the staff with adequate skills and knowledge to better serve the farming communities. Farmers training was also among priority activities of the project. During the project implementation period a total of 11,082 were trained (male 5,217 and female 5,265). Farmers were trained through well established Farmers Field Schools, Farmers Research Groups and were also trained through organised residential courses, Farmers Field Days, Study tours, On- farm demonstrations and through mass media. The impact of training has been significant in terms of farmers empowerment and capacity building and increase in agricultural productivity. The Project also supported food crops research programme encompasses work on banana, root crops (cassava, sweet potatoes and yams) and legumes (mostly cowpeas and pigeon peas). The research work was conducted both on- station and on- farm. With on- farm research, farmers were involved in all stages of the research work, right from problem identification to the evaluation of the research results. The aspects looked at with the on-going research programme include screening of local and introduced varieties of candidate crops for their tolerance to pests and diseases, yields and their responses to farm yard manure applications. Results from the on –farm trials have shown that three improved varieties of cassava clones (ZNZ/98/036, ZNZ/98/031 and ZNZ/98/084) are relatively tolerant to pests and diseases and therefore ranked highly by farmers. Farmers also showed preference to local varieties of Boma and Kibiriti Mweusi which were also found to be less susceptible to pests and diseases infestation. Furthermore, results from on- farm trials have identified the yam variety Tona to be superior over other screened varieties (Majoka, Mwendachi, and Zambarau) due to its yield potential. The Variety Tona

- 54 -

recorded average yield of 40.5 t/ha compared to Majoka (31.0 t/ha), Mwendachi (19.0 t/ha), and Zambarau (17.0 t/ha). With sweet potatoes, among the seven varieties that were evaluated (TIB-4, Salyboro, Zapallo, Tainung 65, W- 151, SPN/O and Shangazi), three varieties were ranked highly by farmers. The varieties that were found superior over the others are Zapallo (15.6 t/ha) SPN/O (13.0 t/ha) In addition, the adaptive research sub- component maintained germplasms for all candidate crops. These germplasms were maintained at Kizimbani and Matangatuani Agricultural Stations and were meant to serve as a source of planting materials for future research programme. On assessing the impact of project intervention, out of 112 interviewed cassava growers, 72 percent were found to have already started adopting the use of recommended varieties of Boma and Kibiriti Mweusi. Furthermore, it was found out that yield of cassava has increased two times more for farmers using recommended varieties (17.0 t/ha) compared to average yields of cassava of farmers still not using recommended varieties (8.0 t/ha). For those farmers using recommended varieties and organic manure, recorded yields are on average of 20.0 t/ha. Average yields of yams under farmers conditions are 25.0 t/ha for farmers using the recommended variety of Tona and organic manure and only 8.0 t/ha for farmers using other yams varieties with no manure. With sweet potatoes, 60 percent of the sampled farmers (n = 42) were found to have adopted the use of recommended varieties (SPN/O and Zapallo). Yield records show that farmers who use recommended varieties obtain on average 13.0 t/ha as compared to the average yields obtained by farmers using not recommended varieties (6.0 t/ha). Following the achievement made through NAEP II and from the results of field experimentation and demonstration, it is recommended that adaptive research and extension activities be sustained to verify findings and disseminate innovations to farmers in order to have wider economic impact on the farming communities.

- 55 -

- 56 -