Tanzania - Forest Resources Management Project

Document of The World Bank FOR OFFICIAL USE ONLY Report No. 19624 IMPLEMENTATION COMPLETION REPORT TANZANIA F7OREST RESOURCES MANAGEMENT PROJECT (Credit 2335-TA) November 15, 1999 Environment Technical Group Africa Region |This document has a restricted distribution and may be used by recipients only in the performance of| |their of filcial duties. Its contents may not otherwise be disclosed without World Bank authorization.|  CURRENCY EQUIVALENTS (as of September 1999) Currency Unit Tanzanian Shillings US$1 = TSh 800 WEIGHTS AND MEASURES Metric System ABBREVIATIONS AND ACRONYMS CTB Central Tender Board DCA Development Credit Agreement DFO District Forest Officer DPC Deputy Project Coordinator ESMAP Energy Sector Management Assistance Program ESW Economic and Sector Work FBD Forestry and Beekeeping Division FRMP Forest Resources Management Project GIS Geographic Information System GOT Government of Tanzania GPS Global Positioning System IDA International Development Association IRA Institute for Resources Assessment JFM Joint Forest Management MLHSD Ministry of Lands and Human Settlements Development MNRT Ministry of Natural Resources and Tourism PC Project Coordinator PCC Project Coordination Committee PMO Prime Minister's Office PCU Project Coordination Unit RFEO Regional Forest Extension Officer RFO Regional Forest Officer RS Regional Surveyor SMD Survey and Mapping Division TA Technical Assistance TAFORI Tanzania Forestry Research Institute TANRIC Tanzania Natural Resources Information Center TFAP Tanzania Forestry Action Plan VEA Village Extension Agent TANZANIA'S FISCAL YEAR July 1 - June 30 Vice President: Callisto E. Madavo, AFRVP Country Director: James W. Adams, AFMTZ Sector Director: Hans Binswanger, AFTRE Team Leader: Peter Dewees, AFTEI  IMPLEMENTATION COMPLETION REPORT TANZANIA FOR OFFICL USE ONLY FOREST RESOURCES MANAGEMENT PROJECT (Cr. 2335-TA) Table of Contents PREFACE EVALUATION SUMMARY .....................................................................I PART I: PROJECT IMPLEMENTATION ASSESSMENT . .....................................................................I A. STATEMENT/EVALUATION OF OBJECTIVES ..........................................................I B. ACHIEVEMENT OF PROJECT OBJECTIVES .................................................................................3 C. IMPLEMENTATION RECORD AND MAJOR FACTORS AFFECTING THE PROJECT .................................................4 General ............ ....... ,.,.4 Project Implementation ......... 4 Strengthening Natural Resources Monitoring and Forest Policies .....................................................4 Strengthening Land Policy and Institutions . ....................................................6 Strengthening Regional and District Forest Services .................................................... 7 Organization and Management ............................................ 10 Technical Assistance and Training ............................................ I I Procurement ................... , , 11 Project Cost and Financing ................... 12 D. PROJECT SUSTAINABILITY .......................12 E. BANK PERFORMANCE ................. . 12 F. BORROWER PERFORMANCE ............................... 13 G. ASSESSMENT OF OUTCOME ........................... 1 3 I1. FUTURE OPERATIONS *fwo** ................ *P .j**e***j @* w*6******14 1. LESSONS LEARNED .............................. 14 PART 1I: STATISTICAL TABLES .............................................................................. 17 TABLE 1: SUMMARY OF ASSESSMENTS ............................................................................................ 17 TABLE 2: RELATED BANK LOANS/CREDITS. ......... . 18 TABLE 3: PROJECT TIMETABLE .......................................................................................... 19 TABLE 4: LOAN/CREDIT DiSBURSEMENTS: CUMULATIVE ESTIMATED AND ACTUAL (US$ THOUSANDS) ............. 19 TABLE 5: KEY INDICATORS FOR PROJECT IMPLEMENTATION . .............................................................................. 20 TABLE 6: KEY INDICATORS FOR PROJECT OPERATION ....................................... ..................................... 23 TABLE 7: ST UDIES INCLUDED IN PROJECT ............................................................................ 24 TABLE 8A: PROJECT COSTS ............................................................................ 25 TABLE 8B: PROJECT FINANCING .................................................. ....... 26 TABLE 9: ECONOMIC COSTS AND BENEFITS .................................................. 27 TABLE 10: STATUS OF LEGAL COVENANTS ................................................... 28 TABLE I 1: COMPLIANCE WITH OPERATIONAL MANUAL STATEMENTS .................................................. 30 TABLE 13: BANK RESOURCES: MISSIONS .................................................. 31 APPENDICES A. Aide Memoire B. IBRD Map This document has a restricted distribution and may be used by recipients only in the performance of their official duties. Its contents may not otherwise be disclosed without World Bank authorization.   IMPLEMENTATION COMPLETION REPORT TANZANIA FOREST RESOURCES MANAGEMENT PROJECT (Cr. 2335-TA) PREFACE This is the Implementation Completion Report (ICR) for the Forest Resources Management Project, for which the Credit 2335-TA, in the amount of SDR 13.2 million (US$18.3 million equivalent), was approved on February 11, 1992 and made effective on July 16, 1992. The credit was closed on June 30, 1999, one year after the original closing date. Nearly all of the funds committed under this credit (99.6 percent) were disbursed. The balance of funds undisbursed (around SDR 52,000) was cancelled on November 15, 1999. This ICR is based on an earlier draft prepared by the FAO/World Bank Cooperative Program, which carried out an ICR mission to Tanzania in March 1999.1 The ICR was finalized by Peter Dewees (AFTE I), and reviewed by Jean Roger Mercier (Acting Sector Manager, AFTE 1). The mission's Aide Memoire is included as Appendix A to this report. The Borrower, primarily through the Project Coordination Committee (PCC) and the Project Coordination Unit (PCU) arranged field visits and meetings, commented extensively on the mission's Aide Memoire, and provided its own assessment of the project and its achievements in a Project Implementation Overview Report, which can be found in the project files. Comments on the draft ICR have been solicited from Government. These were received and have been fully incorporated into this document. Several separate evaluation studies were commissioned of various project activities, and these, in addition to field visits and project documentation, were used in preparing this report. The FAO/CP mission was comprised of Messrs. Pietros Kidane (Mission Leader/Economist, FAO); Donald Sungusia (Agricultural Services Specialist, World Bank) and Richard Kaguamba (Forestry Specialist, World Bank).   IMPLEMENTATION COMPLETION REPORT TANZANIA FOREST RESOURCES MANAGEMENT PROJECT (Cr. 2335-TA) EVALUATION SUMMARY Introduction 1. During the 1970s and 1980s, there was a widespread perception in Tanzania that deforestation in some areas was accelerating at very high rates, primarily being driven by demands for woodfuel and increased demands for construction timber. In some regions, such as Mwanza, it was believed to have reached alarming levels, posing threats to the future production of wood and wood products, to household welfare, and to existing ecosystems. The perception was accurate, but tended to obscure more problematic causes of deforestation, primarily those associated with the need for agricultural land. 2. Because of the view that one of the driving causes of deforestation was the demand for woodfuel, in the mid-1980s, the Government of Tanzania (GOT) launched a study of woodfuel supply and demand. Government subsequently concluded, not surprisingly, that this approach would only deal with part of the problem. In order to take a more comprehensive and strategic view of the sector, it launched preparation of the Tanzania Forestry Action Plan (TFAP), which was completed in 1989, and which covered the period 1990/91-2007/08. The Tanzania Forest Resources Management Project (FRMP) was formulated as a first step in implementing the TFAP. 3. Consistent with the broad objectives outlined in the TFAP, FRMP's main objectives were to strengthen institutions in charge of the development of forest and land policies, as well as regional and district forestry services in selected areas. Project components included: (a) strengthening natural resources monitoring and forest policy, which focused on the production of national land use and natural resources maps, the establishment of a natural resources information center, improvement in royalty collection and the preparation of an assessment of best practices for forest activities, and of a review of long-term financing needs and options for the forestry sector; (b) supporting land policies and institutions, including activities related to village land demarcation, titling and registration; and (c) strengthening regional and district forest services in Mwanza and Tabora Regions through the provision of training, technical assistance, equipment and material to enable them to improve forest revenue collection, delivery of community and farm forestry extension, and the management of woodlands.  Evaluation Summary ii .~~~~~~~~~~~~~~~~~~~~~~~~~ 4. The project was appraised by IDA in two stages, in October 1990 and May 1991,2 and a credit of SDR 13.2 million, equivalent to US$18.3 million, was approved on February 11, 1992. The project was declared effective on July 16, 1992. Project Implementation Experience 5. Implementation of the project was initially slow. Project management failed to put in place the mechanisms for launching activities which were envisaged when the application for PPF funds was made, and, for example, took more than a year to put in place a full-time Project Coordinator (PC). Some preparatory tasks, such as the time consuming process of procuring goods and services, could have been launched earlier, but were instead carried out during the project implementation period. A shortage of counterpart funds also contributed to delays in the early years of the project, but this was resolved later when funds became available from the Ministry of Natural Resources and Tourism (MNRT), retained from royalties collected from the sale of timber and woodfuel from forest reserves. Needed staffing changes were later implemented which greatly facilitated project implementation. 6. A joint GOT and IDA mid-term review was carried out in March/April 1996. The MTR concluded that project implementation was generally satisfactory, and that the original project objectives continued to be valid. It further acknowledged progress which had been made in improving project implementation performance. Project Results 7. At completion, the project has succeeded in achieving its stated objectives. Institutions concerned with land and forest administration and management have been strengthened through the provision of technical assistance, training and the introduction of improved methods of work. Tanzania's capability for monitoring changes in woodland and forest cover has been greatly increased by the establishment of the Tanzania Natural Resources Information Center (TANRIC). A far reaching and innovative new Forest Policy has been adopted, in part based on inputs prepared with project finance. In Mwanza and Tabora Regions, local forestry institutions have been strengthened to provide better forestry extension services to communities and individuals. Some 9,600 private nurseries have been established in these Regions, and about 9.6 million seedlings have been planted. Over 11,000 farmers and school teachers have been trained in forestry matters, and around 10,500 improved wood stoves have been installed by farmers in their homes. The project has also developed and introduced a new system of forest revenue collection, demonstrating that royalty collection from forest reserves can be significantly increased. Coupled with a revenue retention scheme introduced during the project period, this has significantly improved the Forestry and Beekeeping Division's (FBD's) financial position. 8. The project has also piloted several initiatives which have sought to improve natural forest management, involving communities and individuals. The development of Joint Forest Management (JFM) in Tabora on a pilot basis, through which the benefits of woodland management are shared with communities living around reserved forests, has yielded useful 2 The two-stage appraisal process was undertaken because a number of concerns were identified during the October mission, including insufficient emphasis on priority setting in the TFAP, inadequacy of financial planning to ensure sustainability and replicability, lack of clarity with regard to seedling pricing policies, and insufficient progress in the development of policies associated with land management. These concerns were addressed in subsequent modifications to the project's design.  iii Evaluation Summary lessons. The project has also sought to strengthen traditional systems of woodland management (ngitiri) through a program of registration and extension, and this is proving to be a cost effective way of increasing local communities' and individual's responsibility for small blocks of woodland. The project's support for JFM and ngitiri management has important implications for reducing poverty associated with constrained access to and control over natural resources. 9. Important achievements have also been made in strengthening land policies. The project has supported the preparation of a new Land Policy, which was an outcome of a series of national and local consultations and which led, ultimately, to passage of new land legislation in early 1999. Land administration capacities have also been improved, both by investments in surveying and demarcation, and by computerizing lands records. The methods adopted by the project for surveying and demarcation have proven to be efficient and time saving, and consistent with the objectives of the broader land policy framework. The project has demarcated 3,560 villages and surveyed 1,820 villages, slightly more than what had been envisaged at appraisal. 10. Extensive private investments in nursery development, tree planting, and in improved wood stove technologies, facilitated by FRMP, are expected to yield long term benefits (though lessons in this area are difficult to assess because of weaknesses in the monitoring and evaluation of the project's extension approaches). In many areas, there is evidence that these activities are being adopted and extended without further project assistance. The development of other services supported by the project, for example, services with regard to natural resources information and land survey and titling, are likely to be sustainable, as new sources of revenues (coupled with revenue retention schemes) have enabled these service delivery institutions to reinvest and improve their capabilities. The sustainability of regional and district services is less certain. Budgetary resources to continue revenue collection activities initiated by the project,3 and for providing extension services are likely to be constrained. For these and other reasons, GOT is considering significant institutional reforms which should improve the capacity of local governments to manage forest resources. Lessons Learned 11. Lessons which have been leamed during project implementation include that: * the concept of strengthening traditional systems of woodland management (ngitiri) is fundamentally sound and has good replication potential throughout the country. The most important aspect of the program has been the support it has given to people to act to halt the disappearance of remaining woodland, and to bring as many pockets of woodland as possible under rehabilitation. This has been accomplished by working with villagers as the logical, long-term guardians of forest resources to protect and conserve these resources themselves, at little cost to government, and through socio-legal mechanisms which they themselves control and maintain according to their own interests. Potentially the most important poverty reducing impacts, as a result of the project, have been an outcome of the ngitiri program. 3 This constraint is not as circular as it appears. Revenues are remitted to FBD, which retains a proportion of them to cover some of its expenses. Most costs, however, are born by District Forest Officers, and by extension, the local District administration. Arguably, the linkage between a DFO's budget for revenue collection needs to be more explicitly tied to the revenues which are generated.  Evaluation Summary iv * Community involvement in Joint Forest Management of government forest reserves can lead to an improvement in the condition of reserves, and to a sharp reduction in the threats they face. The pilot approaches could, however, be strengthened particularly in regard to more fully transferring decision making responsibilities to the communities involved, rather than simply relying on a modified concession licensing process which leaves ultimate control in the hands of Government. * More generally, dry woodland management poses good opportunities because of the resilience of these woodlands in Tanzania to harvesting, the cost-effectiveness of relying on natural regeneration with minimal silvicultural treatments, and because of the few risks to the environment posed by sustainable management of dry woodland ecosystems. * Farmers are apt at adopting simple technologies that are based on the use of inexpensive local material, such as the improved wood stoves. Higher-tech approaches are not needed and are difficult to disseminate through extension. The woodstoves program has been important for women, primarily of reduced woodfuel demands which are an outcome of adoption, and the impacts on patterns of household resource allocation. * Community participation in the demarcation and surveying of village lands is essential and cost effective, as this reduces litigation among members to a minimum, leading to considerable savings in time and resources. * Project identification and preparation benefited from both the resources which were put into the effort, but also from the longer term iterative process of project design, and the linkages which were established with sectoral strategies, such as the Tanzania Forestry Action Plan. Indeed, experience with the project suggests that there can be significant project implementation benefits when project identification and preparation is bolstered by strategic studies, action-oriented research, and ESW. * Weaknesses in project performance were often linked with inadequacies in the prevailing institutional mechanisms for forest management in Tanzania. These place responsibility for forest management within a decentralized institutional structure, without adequate mechanisms for ensuring there is technical, administrative, and management oversight. Problems of governance in the sector are linked to these weaknesses, and future investments will likely be jeopardized in the absence of institutional reforms. The problem is widely recognized in Government, and the Ministry has made a commitment to launching a reform process which seeks to address these systemic weaknesses. * Technical assistance of the type financed by the project may be delivered more effectively by individual consultants, rather than firms. The grouping of consultancies into larger packages during the tendering process -- efficiency gains notwithstanding -- limited Government's flexibility in finding qualified individual consultants, and increased the complexities associated with technical supervision. Consequent deficiencies in performance were difficult, time consuming, and costly to rectify. * A Project Coordination Unit may facilitate implementation of project activities in the short term, but may have little overall impact on increasing institutional implementation capacity. In this instance, a PCU was justified on the basis of the need for coordination between multiple institutional partners, but provided little scope for the more general  v Evaluation Summary technical, administrative and management oversight needed in the sector, given its decentralized institutional structure. * Payment of project allowances may encourage project staff to work diligently. Given public spending constraints, these are not sustainable in the long-term. This is an issue which must be addressed systematically through the civil service reform process.   IMPLEMENTATION COMPLETION REPORT TANZANIA FORESTRY RESOURCES MANAGEMENT PROJECT (Cr. 2335-TA) PART I: PROJECT IMPLEMENTATION ASSESSMENT A. STATEMENT/EVALUATION OF OBJECTIVES 1. In the mid-1980s, the Joint World Bank/UNDP Energy Sector Management Assistance Program (ESMAP) was invited by the Government of Tanzania to assist it in identifying solutions to the problems posed by increasing demands for woodfuels. There was a widespread perception that these demands were contributing to deforestation, and a concern that the country's strategy towards forest resources management needed more fully to consider the scope for reducing these pressures on the one hand, while increasing supplies on the other. ESMAP prepared a series of proposals for investment in wood energy production and demand management, and submitted these to Government and to Bank management for review. 2. In the meantime, Government began to think more critically about the broader implications of its approach toward forest and woodland management. Nearly 40 percent of the country remains under forest or woodland cover, and woodfuels account for only a proportion of total use, other values being far greater in some respects for catchment protection, timber production, habitat protection, wild foods, and as a source of dry seasofi grazing for livestock. As a result, Government sought to develop a more comprehensive approach to the sector, and prepared, in 1989, the Tanzania Forestry Action Plan, 1990/91 to 2007/08 (TFAP). The TFAP aimed to tackle forestry issues in a more comprehensive manner, and considered a broader spectrum of issues of national importance, including forest policy, institutions, and the participation of the private sector and communities in tree planting and in forest resources management. 3. The Forest Resources Management Project (FRMP), prepared on behalf of Government and the Bank by the FAO/World Bank Cooperative Program, was formulated as a first step in implementing the TEAP, and had three specific objectives: (a) to strengthen institutions in charge of the development of forest policies and natural resources information; (b) to support the development of rural land policies, strengthen crucial organizations in charge of the implementation of these policies, and to support the implementation and monitoring of government's policy of village demarcation and titling; and (c) to strengthen regional and district forestry institutions in Mwanza and Tabora Regions. 4. Project components comprised: Strengthening Natural Resources Monitoring and Forest Policies, through: (i) production of national land use and natural resource maps; (ii) development of a natural resource information center; (iii) improvement and monitoring of the royalty collection system; (iv) identification of best practices for forest activities; (v) preparation  Part 1. Implementation Assessment 2 of an action plan for the long-term financing of the forestry subsector; and (vi) improvement of accommodations for the Forest and Beekeeping Division (FBD) main offices. * Strengthening Land Policies and Institutions, by supporting: (i) the development of land policies; and (ii) village demarcation, titling and registration. * Strengthening Regional and District Forest Services in Mwanza and Tabora. In Mwanza, activities were to include: (i) survey and planning of the extension program; (ii) district forest service reorganization; (iii) development and distribution of extension materials; (iv) community and farm forestry extension; (v) local research and development; (vi) revenue collection and sharing; and (vii) village woodland and forest reserve management. In Tabora, a one-year preparatory phase was to include: (i) vegetation mapping; (ii) reconnaissance surveys/inventories; and (iii) regional forest resource planning. A second phase, which was to last five years, was proposed and was to include: (i) forest reserve management; (ii) village woodland management; (iii) applied research; (iv) community and farm forestry extension; (v) revenue collection; and (vi) training. 5. The project, to be implemented over a period of six years, was estimated to cost TSh 6.1 billion, or US$20.93 million, with a foreign exchange component of 68 percent. About 88% of the project costs were to be financed by IDA, 9% by GOT and about 3% by another donor.4 6. The management of the project was to be vested with the Ministry of Natural Resources and Tourism (MNRT), the Ministry of Lands and Human Settlements Development (MLHSD), and the two Regions. MNRT, through its Forestry and Beekeeping Division (FBD), was to be responsible for strengthening forest services, and MLHSD was to be responsible for land policy, village demarcation and registration activities. Tabora and Mwanza Region governments were to be responsible for implementing forestry and land components in the field. Project coordination was to be carried out by a high level Project Coordination Committee (PCC) whose members would be largely from the two Ministries and two Regions, and from the Prime Minister's Office (PMO). A Project Coordinator (PC), to be appointed by MNRT, and a Deputy Project Coordinator (DPC), to be appointed by MLHSD, were to form the secretariat of the PCC. Three tiers of responsibility for implementation were envisaged for the forestry component: national level under FBD, regional level under the Regional Forest Officer (RFO); and district level under the District Forest Officer (DFO). Overall implementation responsibility for the lands component was to be vested with the Survey and Mapping Division (SMD), but the Regional Surveyor (RS), who reports to SMD, was to be responsible for the lands component in each regions. 7. The project was approved by IDA on February 11, 1992, and a Development Credit Agreement (DCA), which committed funds totaling SDR 13.2 million, or US$18.3 million, was signed on April 24, 1992. The Credit was declared effective on July 16, 1992. 4 As indicated in the Staff Appraisal Report, it was expected that NORAD would co-finance the project by providing US$0.7 million to support the Natural Resources Information Centre. This co-financing did not materialise and was not considered during the approval and signing of the credit.  3 Part 1. Implementation Assessment 8. The long period which elapsed from the time the potential for an investment was first identified (1984) until approval (1992) is worth commenting upon. More than anything, the elapsed time shows that both the Bank and Government were keen on getting the process right, and were not willing to move forward with the investment until the parameters were clear. Arguably - and despite the high costs incurred of identification, preparation, and appraisal - the reticence to move forward with a project any earlier ensured a stronger level of ownership and implementation impact. Indeed, this is one of the few forestry projects in the Africa region, prepared during the 1980s, where intended implementation outcomes and development impacts have been largely achieved. B. ACHIEVEMENT OF PROJECT OBJECTIVES 9. The project, as it was designed, sought to combine a strong capacity-building program, with pilot and full-scale development activities. Capacity building focused on strengthening the institutions charged with forest and land management and administration by improving the basis for policy development. The project also focused in the first instance on improving implementation capacity through pilot-scale activities (in particular, Joint Forest Management, community-based woodland management, revenue collection, and forestry extension), and where appropriate, undertaking wider-scale activities (in particular, land surveying and demarcation). At completion, the project had succeeded in meeting most of its stated objectives. 10. The policy process, for example, has become more fully informed by access to better information. The project supported the Tanzania Natural Resources Information Center (TANRIC) which, among other things, has fully-mapped Tanzania's vegetation cover. A series of policy studies also sought to provide better information during the preparation of new land and forest policies. These included a review of best practices in the forestry sector, an assessment of long-term financing in forestry, and a series of papers on land tenure and administration. 11. An excellent Forest Policy paper has been produced by FBD, indicating in part its improved capability for analyzing complex forestry related issues, and defining future sectoral goals and needs.5 The project has also improved the technical competence of FBD, particularly in the areas of monitoring royalty collection. By introducing a new system of royalty collection, regional forest services have demonstrated that royalty collection can be increased markedly, by closing some loopholes in the system. In addition, they have acquired valuable experience in forestry extension. 12. The project has contributed significantly to land policy formulation and improved land administration by financing studies used in the policy formulation process, and by providing physical and technical inputs for land surveying and demarcation. Project inputs have helped to clarify issues related to land matters, and enhanced the development of new land legislation. During preparation of the land policy, institutions involved in land administration and management were substantially strengthened by virtue of their participation in the policy preparation exercise. In addition, the technological know-how of the institutions concerned in land demarcation and survey, as well as in land registration and titling, has been stepped up. In 5 The Forestry and Beekeeping Division noted, in its review of the draft ICR that, "The good quality of the Forest Policy paper is not the contribution of FRMP alone but also other stakeholders who were consulted and the consultants who were recruited to assist in the policy preparation."  Part 1. Implementation Assessment 4 particular, as a result of the acquisition of Global Positioning System (GPS) technology, training, and improvements in mobility, the project has significantly raised productivity in land demarcation and surveying. C. IMPLEMENTATION RECORD AND MAJOR FACTORS AFFECTING THE PROJECT General 13. In order to expedite project execution, IDA approved a Project Preparation Facility (PPF) advance (US$650,000) to finance preparatory tasks for project implementation, including the recruitment of staff, procurement of equipment, material and civil works, as well as for planning of activities and preparation of tendering documents. However, the PPF was not used, mainly due to communication difficulties between various GOT departments. This meant that preparatory tasks had to be carried out during the time envisaged for implementation, hence delaying project start-up. The PCC and the PC were in place before the credit was negotiated, although the post of PC was vacated soon after the project was declared effective, following his promotion as Director of FBD. A new PC was appointed only in August 1993. Implementation performance was impaired in the first year by shortage of funds, staff, office facilities and delays in opening project accounts in the Regions. In addition, the then-Director of FBD gave limited institutional and logistical support to the project, and arguably hampered implementation by delaying, among other things, the tender processes for the proposed new HQ building, and the introduction of the pilot revenue collection scheme. 14. Slight improvements in implementation were evident in the second year, following the recruitment of some staff and the procurement of a few vehicles. However, significant progress was recorded only in the third year, when the recruitment of consultants was virtually completed, and procurement of vehicles and other equipment and material was stepped up.6 By this time, most serious implementation problems were resolved, except those related to the shortage of GOT counterpart funds, particularly with regard to forestry activities. Major funding crises were avoided mainly due to IDA's agreement to continue financing 100% of operational costs until 60% of this cost category was disbursed, and due to increased use of FBD funds retained from revenue collection. 15. In March/April 1996, a mid-term review was carried out for the project by a joint GOT and IDA team. This review confirmed the validity of the original project objectives, and expressed its satisfaction about the project's progress. The original project closing date of June 30, 1998 was extended twice: first to March 31, 1999, and then June 30, 1999. At completion, the credit was fully disbursed. Project Implementation Strengthening Natural Resources Monitoring and Forest Policies 16. Land Use and Natural Resources Mapping. In mid 1997, the project had successfully produced the country-wide maps envisaged under the project, albeit with about two and half 6 Significantly, a new Director of FBD was appointed around this time, and provided an important boost to project implementation.  5 Part 1. Implementation Assessment years delay from the original plans. The 64 map sheets produced by the project were based on the most recent satellite imagery, and, among other things, they show general classes of forest cover and agricultural and settlement areas. They also provide baseline information on the different uses of land, and this has facilitated the tasks of assessing natural resources and preparing forestry development plans. Another use of the maps has been for regular monitoring of forest resources and changes in land use. The maps are widely used by different governmental and non-governmental organizations. The PCC had originally intended that they should be produced in the facilities of SMD, and the project financed improvements in SMD's printing facilities. In spite of these improvements, the maps were produced abroad because it was more cost-effective. 17. Natural Resource Information Center. In line with project objectives, the Tanzania Natural Resources Information Center (TANRIC) was established by the Institute for Resources Assessment (IRA) of the University of Dar es Salaam. The project provided technical assistance, computers and other office equipment, and funds for purchasing satellite imagery, maps and for ground survey staff. Training in geographic information system (GIS) was also provided under the project. TANRIC now maintains a considerable information base on natural resources and has capabilities for collecting and analyzing data and delivering the latest infornation to customers. The demand for its services is increasing, and its business has enabled it to become nearly self-supporting. 18. Improvement and Monitoring of Royalty Collection Systems. The project proposed to develop a revenue reporting system, introduce a pilot phase of revised revenue collection method, undertake a feasibility study of assessing fee collection from woodfuel used for tobacco curing, and evaluate during the mid-term review the success of the new system. Because of the delay in recruiting consultants, a new revenue system was introduced only in 1996, initially in the four pilot regions, i.e. Tabora, Shinyanga, Singida and Mwanza, and later extended to all the regions. The system consisted of registering business ventures in harvesting, processing and trading of forest products, as well as establishing checkpoints and stepping up forest patrolling activities. 19. Collection of revenue levels improved soon after the introduction of the new system in the four pilot regions from about TSh 90.72 million in 1996 to TSh 277.86 million and TSh 299.17 million for 1997 and 1998, respectively. The increase of revenue in Mwanza was a major jump from TSh 8.70 million in 1996 to TSh 103.0 million in 1998. The contrast between the high collection of revenues of Mwanza, which has few forest resources, and the low collection of Tabora, which has heavy woodland cover, is explained by failure of the collection efforts in Tabora. Mwanza was collecting revenues mainly from wood that was coming from Tabora. In spite of the overall 3-fold increase of revenue, this amount still represents a small portion of the potential revenue which could be collected country-wide. 20. An evaluation of the pilot revenue collection system was carried out at the project's completion. The evaluation attributed weaknesses in the system to deficiencies in data collection and monitoring, poor patrolling, corruption, and an inadequate institutional structure for managing and supervising field staff. The evaluation report also suggested that clear instructions were not provided to DFOs and RFOs, and that this has resulted in divergent, rather than standardized approaches, being taken in the field. With regard to spot checking of timber in transit, the evaluation report observed the difficulty of accomplishing this task effectively,  Part I. Implementation Assessment 6 because of insufficient information about the timber being transported in terms of number of logs, size, quality and species. 21. Some of the difficulties in revenue collection have arisen due to the current institutional structure which is characterized primarily by organizational fragmentation. Although central government forest reserves are under the jurisdiction of FBD, they are managed by the RFOs and DFOs, who are employed by the Regional and District Administrations, respectively. FBD has only limited technical and administrative oversight. DFOs, who are responsible for revenue collection, are not subject to supervision by FBD, and may instead give priority to district matters. FBD shares part of the revenues it collects with the RFOs, but not directly with DFOs. In order to obviate these difficulties, MNRT is considering significant institutional reforms which will address this and other institutional constraints to the management of Tanzania's woodland resources. 22. Studies on Best Practices for Forest Activities and Long-Term Financing of Forestry. FBD recruited consultants to undertake these two studies. The study on best practices provided useful and innovative ideas on many issues, including the involvement of local communities in the management of public lands and forest reserves, and of the private sector in the management of plantations. The findings and recommendations of this study have, in many respects, been incorporated into the new Forest Policy and have been reflected in the wider policy framework. FRMP financed studies, workshops and seminars as an input into the development of the new Forest Policy,7 and these also contributed to improved capacity for management in the sector. The study on long-term financing has provided useful information on the valuation of forest products and recommended a change of the current valuation system. The study indicated that self-financing of future forestry operations would be guaranteed, if effective collection of the revenues could be assured. 23. Improvement of Infrastructure. The project proposed to resolve FBD's problem of limited office space by constructing offices on a government site in Dar es Salaam. Although an initial design for the structure was developed before project appraisal, and architects were appointed at the end of 1993, FBD did not succeed in having its new headquarters built. Last minute changes in site plans and excessive delays in advertising tenders did not allow tenders to be finalized or awarded by the end of July 1997 (the latest deadline agreed with IDA). Consequently, the construction of the building not undertaken, and funds were reallocated to other project components. Strengthening Land Policy and Institutions 24. Land Policy Development. Land related studies envisaged under the project were modified in mid-1993 to include support for the process of land policy formulation. Consultants were recruited to assist this important GOT policy fornulation exercise, which was completed with the adoption of a National Land Policy by Cabinet in April 1994. The Land Policy eventually led to the passage by Parliament of new legislation in February 1999 (the Land Act and the Village Land Act). 25. Lands Records Storage and Management. The project also assisted MLHSD in improving the organization of land records and cadastral information system. Following a 7 The preparation of the Policy itself was led by FBD, with donor support from GTZ.  7 Part 1. Implementation Assessment review of the poor state of land records storage and management, a plan was drawn up to save the critical infornation contained in the deteriorating records by storage in a computerized data base. To this end, the project developed a system of computerized data management, and provided staff training, technical assistance, computers and other office equipment. At completion, the project succeeded in introducing a system for the safe storage and easy use of land records. About 45.000 records, or some 18% of the total, have been computerized. 26. Village Demarcation, Titling and Registration. It had been envisaged that demarcation, survey and titling of village lands would take place in five regions. Plans were later extended to cover eleven regions. Project activities commenced with demarcation, and progress accelerated after the mid-term review of the project, when staff mobility improved. Demarcation activities were undertaken with the involvement of the villagers, a process which was an important project accomplishment. Demarcation of village lands was completed in Dodoma, Morogoro, Singida, Tabora, Shinyanga, Mwanza, Mara, Ruvuma and Kagera regions. Small areas of Arusha and Kilimanjaro regions remain undemarcated, pending settlement of disputed boundaries. Final surveying commenced only in August 1998 due to the delayed arrival of the necessary equipment, particularly the GPS units. Once the equipment was in hand, however, the process of surveying progressed rapidly. At completion, the project has demarcated about 3,560 villages and surveyed some 1,820 villages. Following the adoption of the new land policy, the original plan to issue village land title to the village administrations was changed, and village councils are provided with an administrative certificate (rather than title) describing their boundaries, together with the boundary maps, of their respective villages. Strengthening Regional and District Forest Services 27. Forestry Extension. The project proposed to develop innovative approaches for forestry extension, and to establish high quality baseline information for an effective monitoring and evaluation system. Technical assistance inputs were expected to assist in accomplishing this task. These did not materialize as intended, because the consultants who were contracted to assist the regions failed to focus on developing alternative extension approaches. What emerged was a traditional forestry extension approach that simply promoted tree planting, with little understanding of the social and economic factors which influence household decision-making in this regard, or any appreciation about how tree and woodland management could be incorporated into land-use practices. 28. Having said this, the project did build on earlier experiences in Mwanza by providing support for Village Extension Agents (VEAs) who had been designated by villagers as focal points for receiving technical messages and training in nursery and tree planting techniques. The main topics covered by forestry extension were tree planting, woodland management, improved stoves and Joint Forest Management (JFM) of Government forest reserves. Tree planting by individuals, groups and institutions was by far the most common extension activity. Extension messages and training in nursery techniques, species selection, tree planting and tending were, therefore, given to a large number of individuals and to an appreciable number of institutions, particularly schools. Some 9,600 nurseries of different sizes have been established in both regions, and about 9.6 million seedlings have been produced - about 104% of appraisal target. The project contacted about 37,600 households and provided training to over 11,000 farners and school teachers. In addition to delivery of extension messages, tree planting was promoted through the provision of tree seeds, polythene tubes and pesticides free of charge -- a practice  Part 1. Implementation Assessment 8 which is not sustainable. Ultimately, however, the impact of the extension program cannot be fully assessed due to the failure of consultants to assist in establishing a monitoring and evaluation system. This is as pressing a need as ever, because of the importance of understanding the incentives of the private sector to respond to perceived needs for tree planting. 29. Despite the weak support provided for the development of new approaches toward forestry extension, the project did support two particular sets of initiatives which were relatively unconventional. The first involved building on traditional practices of setting aside tracts of land for later or emergency use. Originally a concept which was applied to grazing areas, ngitiri, the project extended the idea to residual pockets of woodlands. The second approach was an effort to share the benefits and responsibilities for the management of Government Forest Reserves with communities through what was called Joint Forest Management (JFM). An independent evaluation of the project's ngitiri and JFM activities was carried out prior to the ICR mission. 30. One of the main observations made about the ngitiri program (which began on a modest scale in Mwanza, and then was adopted in Tabora) was that it did not go far enough in reconstructing and modernizing the concept. This is partly because the approach only received halting support from the technical advisers who were engaged to provide assistance in this area. In some areas, ngitiri were simply registered, and there was inattention given to: where and how authority over common land ngitiri is lodged; to guiding villagers about protection strategies; and to encouraging villagers to exploit the considerable powers they possess to regulate access, use and management actions in general. The evaluation also noted the large discrepancies in the areas reported to be under ngitiri management, and suggested that hectarage was, in any event, not an appropriate measure of success. Rather, the reviewer noted, the value of ngitiri is in the support it gives to ordinary people to act to halt the disappearance of woodland, and to bring as many pockets of woodland as possible under rehabilitation. Dismissing ngitiri as insignificant because they are small (as the technical advisers seemed to have done) is unwarranted, simply because they are highly significant to their owners. The evaluation concluded that the ngitiri concept is fundamentally sound, and has good replication potential throughout the country. The most positive aspects of the ngitiri program are those which targeted villagers as the logical, long-term guardians of forest resources, and which work with them to protect and conserve these resources themselves, at little cost to government, and through socio-legal mechanisms which they themselves control and maintain according to their own interests. 31. Joint Forest Management was an approach adopted on a pilot basis in the Urumwa Forest Reserve, a 13,000 ha reserve 15 km or so from Tabora. In practice, 7 user groups have entered into contractual arrangements with Government which provide them with exclusive harvesting rights of specified small coupe areas in return for protecting these and other areas from illegal use. In addition, each household within each user group is supposed to pay an annual fee of TSh 500. The evaluation concluded on the one hand, that community involvement in Urumwa had led to an improvement in the condition of the Reserve, and to a sharp reduction in the threats it faces, but on the other, that (like the ngitiri effort), the approach taken didn't go far enough, particularly in regard to more fully transferring decision making responsibilities to the communities involved.8 The evaluation also noted that there had been a number of very positive s The development of a genuine partnership between Government and communities has been challenging. For example, the first contracts in Urumwa were prepared by Government, with little input from communities, and they covered only one year. The contracts did not have a regulatory regime that specified the rights and duties of the participants in the arrangement. They were also not supported by management plans which incorporated  9 Part 1. Implementation Assessment shifts during implementation in the narrow approach originally proposed, and that this suggested some interesting opportunities for expanding upon a significantly modified approach in future. 32. An important project achievement, not foreseen at appraisal, was the introduction of improved stoves. The wood stoves which have been introduced by the project are simple to construct and made from local material. At completion, some 10,500 stoves had been constructed and installed in homes. The extension method is based on showing women how to install stoves by actually constructing them, and letting each person to replicate the process separately or in a team. This approach has had a domino effect, as the trained women are showing others how to construct the stoves wherever they go. Until now, the main emphasis was on promoting wood stoves, but improved charcoal stoves, manufactured by women, are also being introduced by the project. Studies carried out by project staff indicate that the improved stoves are more woodfuel efficient, and a saving of about 50 percent of woodfuel is obtained compared to the traditional stoves. 33. In retrospect, this diversified set of approaches toward land, woodland, and woodfuel management have likely been far more effective than any single approach, combining community-based woodland management, with tree planting, and woodfuel demand management. This was a constructive strategy, which shows promise for being effective in mitigating the impacts of increased population growth on natural habitats. The project's primary impacts associated with gender have been an outcome of its support for woodstoves development, and for ngitiris and Joint Forest Management. 34. Similarly, the most significant impacts on rural poverty, which are outcomes of the project, are associated with its support for the introduction of systems of ngitiri management, Joint Forest Management, and woodstoves. This integration into the project of these approaches which strengthened the control over natural resource assets by communities, and individuals, and which reduced, so clearly, household dependence on woodfuels, is likely to have a profound long term impact on natural environments, and on the households which are so dependent on these resources for their sustainable livelihood. 35. Forest Reserve Management. Appraisal expectations of preparing detailed forest management plans for priority forest reserves were never met. A considerable amount of work was carried out in Geita, but this has been undermined because of mining interests. These interests were identified during appraisal, but were never effectively managed during implementation. The proposed inventory for priority forest reserves in Tabora was not completed in time, hence management plans could not be prepared before the departure of the technical assistance personnel. There is, however, general agreement that implementation of the plans would have been problematic in any event because of encroachment and illegal timber harvesting and charcoal production. 36. The problem of encroachment was not evident at project preparation or at appraisal, and its development during the project implementation period coincided with the many political and economic reforms which took place in these years. The situation has been exacerbated by low levels of protection due to shortages of forest guards and by a lack of interest in dealing with the problem at the district-level, particularly by district forestry staff, who are supposed to be the technical, procedural and legal elements acceptable to all parties. These deficiencies have been widely noted, and form the basis for modifications in the approach.  Part ]. Implementation Assessment 10 front line officers for the management and protection of forest reserves. Indeed, the problem of encroachment captures the difficulty of working at the local level through a fragmented, decentralized, institutional structure to protect assets which are in the national interest. Despite Government's efforts to deal with encroachment through other non-project mechanisms,9 problems still persist. Illegal logging and charcoal production continue in many regions. If anything, the problems of encroachment suggest considerable latitude for greatly strengthening the involvement of communities in the direct control and management over forest and woodland resources. 37. The project had carried out aerial photography of some forest reserves of Tabora as an input to the management plans, and these are currently with TANRIC. It is hoped that these will be useful for preparing management plans for some forest reserves in the future. In the forest reserves of Tabora region, the project cleared around 940 km of boundaries (which had deteriorated since the early 1970s), and these have served to separate the reserves from other types of land use and ownership. Boundary patrolling by bicycle and motorcycle was also reintroduced during the project period. Under the project, applied research in miombo woodland management is being carried out by the Tanzania Forestry Research Institute (TAFORI), and it is hoped that important technical advice for managing woodlands will emerge in the future. Although the research started only in 1996, some experiments, focusing on regeneration characteristics of clearfelled woodlands, and understorey burning as a management tool, appear to offer promising results. Organization and Management 38. When the project resolved its start-up problems, the management went into its full strength and assumed its expected role in facilitating the work of the implementation agencies. The PCC met regularly to review and approve project work plans and to resolve eventual issues. The PC provided continuous support to regional project staff, and facilitated linkages with other concerned bodies. The DPC has also provided similar support to project staff implementing the land component. Weak institutional support during the earliest stages of project implementation became much less of an issue with the appointment of a new director of FBD. 39. Implementation plans were prepared based on the existing GOT organizational structure, which was characterized by a fragmented line of institutional responsibility. Implementation of the components regarding natural resources monitoring and forest policies were vested with FBD/PCU. However, in the field, the RFOs were given the responsibility for implementing the project at the regional level and for coordinating and monitoring district level activities, while DFOs were in charge of the actual implementation of project activities in the districts. Even the flow of funds to the field offices was different i.e. through FBD/PCU to the regional accounts, and through PMO/PCU to the district accounts. Since the regions and the districts have their respective programs in their areas of jurisdiction, and since they are answerable to the commissioners in the regions and districts, the RFOs and DFOs tended to implement the project along their habitual independent manners. In this process, FBD felt neglected and with little ownership for the project, although it was still expected to have overall responsibility for 9 During the project period, around 1,200 illegally encroaching people were evicted from forest reserves in Tabora Region, but this was undertaken without the involvement of project staff or funds. Large numbers are also believed to have moved from reserved areas voluntarily because of a public campaign against illegal encroachment, and following clarification of boundaries.  1] Part 1. Implementation Assessment implementation. Early during implementation, FBD expressed its misgivings about the independent mindedness of the RFOs and about its own limited involvement in implementation, because even the staff it seconded to the regions, i.e. the RFEOs, were answerable to the RFOs. Project management tried to resolve the issue by requesting the RFOs to keep FBD informed on all project related matters. The RFOs, in turn, complained about the independent-mindedness of the DFOs, although they were meant to coordinate and monitor district level activities. Such problems did not occur with regard to the land component, because this was being implemented by the Director of SMD through the RS. 40. The diffused line of command did not facilitate the emergence of a clear organizational set-up for implementing the forestry component. Despite these difficulties, the ad hoc organization and management system put in place by the project (i.e. the PCC and PCU), has been successful in implementing the complex development program under FRMP. However, the wisdom of establishing a management system that artificially tended to centralize activities that have been long decentralized, is questionable, although has been justified as a necessary mechanism to coordinate the activities of two ministries and multiple other players. Still, it points to the weakness of the overall institutional structure for forest management in Tanzania within which the project operated. Technical Assistance and Training 41. The project has provided a strong technical assistance input. In addition to short-term individual consultancies, five firms have provided medium to long-term technical assistance in the fields of natural resources mapping, resource information, improvement of royalty collection, extension forestry, natural resources management, land policy development, improvement of land records and registration and design of FBD headquarters. Excluding the architectural consultancy, a total of 662 person months (385 expatriate and 277 local) of technical assistance was provided by these firms. Performance was, on balance, satisfactory, though one firm was especially deficient in the areas of forestry extension, natural forest management, and monitoring and evaluation. This experience has raised concerns about the need for stronger technical supervision, which FBD argues it can only provide when individuals, rather than firms, are involved. A large training program was also provided under the project, and this has helped implementation in the field. Most of the training focused on improving the operational capabilities of staff in the fields of forestry extension and management and on demarcation and survey of land. The training program consisted primarily of short-term courses and study tours. Procurement 42. Procurement was carried out by the PCU and by the Central Tender Board (CTB). The procedure followed was to deliver the tendering documents prepared by PCU to CTB, after a "no objection" was given by IDA. After being reviewed, the tenders were advertised by CTB and bids were directly received and opened by it. The bids were then sent to the PCU for evaluation. Following the reviewing of the evaluation made by the PCU, and having satisfied itself of the result, an approval for awarding the contract was given by CTB to the PCU. This system has worked out quite well in most cases, although some delays were encountered in few cases due to CTB's numerous queries. The only major procurement problem faced by the project resulted in the failure to construct the FBD headquarters.. FBD management submitted the evaluations only in November 1997, about three months after the last deadline set by IDA for forwarding the  Part 1. Implementation Assessment 12 contract. CTB's approval had been obtained in April 1998, which was too late to commence and finish construction by the project's closing date. Delays in the procurement of GPS units were also experienced due to several misunderstandings resulting from a lack of clarity about IDA procurement procedures for reviewing contract modifications. 43. With regard to the procurement of consultants, the advantages of seeking tenders from firns to provide the several packages of services needed were sometimes offset by the difficulties of supervising staff provided by these firms. FBD has expressed a decided reticence to rely, in future operations, on firms to provide TA support, and may seek instead to recruit individual consultants to provide closely supervised advisory services. The perception is very much that firms provided higher cost services, and produce lower quality outputs. Nonetheless, it is also clear that FBD needs to monitor performance more closely, and insist on the delivery of acceptable outputs and services from both individuals and firms. Project Cost and Financing 44. Total project costs at completion amounted to US$19.40 million or TSh 7,369 million. The IDA credit of SDR 13.2 million was nearly fully disbursed, and over the life of the project, the value of the credit increased from US$18.3 million to US$ 18.8 million, due to the depreciation of the US$ in relation to the SDR. The balance, amounting to about US$0.55 million or 4% of total costs, was financed by GOT. About 66% of the costs were expended on the forestry component, and the remainder 34% on the lands component. Undisbursed project funds, totaling around SDR 52,000, were cancelled in November, 1999. D. PROJECT SUSTAINABILITY 45. Activities supported by the project which are now privately operated, such as private nursery development, wood stoves, and the tending of on-farm trees, will be sustainable. Activities under TANRIC and the lands component are also likely to be sustainable, because services for natural resources information, or for land surveying and titling are being provided against payment for services rendered. Some service delivery institutions will be allowed to retain funds obtained from customers to finance their operations. Sustainability of continuing the services provided by the field forestry services, will, however, depend on the level of support that these will receive from GOT. Based on past experience regarding GOT's difficulties in providing adequate budget, it is likely that funding will be a constraint. This may put their sustainability in doubt, although FBD's stand in financing forestry development has improved markedly with the increase in revenue collection. Indeed, preliminary estimates suggest that rates of revenue collection have continued to increase since4 the project's closing. The current proposal for establishing a separate executive agency, with complete operational autonomy to be responsible for the management of the forest reserves, may create opportunities for higher levels of revenue collection and better management of the forests. E. BANK PERFORMANCE 46. Overall, the performance of IDA has been satisfactory, both during project preparation and appraisal, as well as throughout project supervision. By being involved in the identification and preparation stages of the project - as an executing agency of ESMAP and as a participant in the FAO preparation mission - IDA has kept track of the project progress, and contributed to its  13 Part 1. Implementation Assessment formulation. As one of the main players in the preparation of the TFAP, it has also succeeded in bringing about strong linkages between the project and the TFAP. Appraisal of the project, which was carried out in two phases, was adequate. The different project components and activities were clearly described and sufficiently detailed. In retrospect, it can be argued that preference should have been given to designing a project organization and management within the existing GOT structure, instead of creating an ad hoc project coordination unit. IDA undertook 10 full supervision missions, which were generally carried out efficiently. The missions' professional mix was well balanced between the forest and lands components. The supervision missions provided useful advice both to GOT and IDA for the resolution of implementation problems. An example of such is the recommendation to IDA to keep disbursement against operating costs at a higher level in order to give chance to GOT to sort out its funding problems, which at that time were constrained by a particularly tight budget. One noteworthy weakness in supervision was the failure by the initial missions to work with the implementing agencies to put in place more effective mechanisms for using PPF funds. F. BORROWER PERFORMANCE 47. The Borrower's performance was generally satisfactory. During the formulation of the project, GOT took its time until it was satisfied that the project fitted well within the TFAP. This has given the project an opportunity to benefit from new forestry development approaches that highlighted the importance of policy and institutional matters and people's participation in forestry development. During the implementation period, GOT gave strong political support to the project through the PCC and its high level members. This GOT stance was an important asset to the project, as it was clearly perceived by staff at all levels as a reflection of firm government commitment to the project in particular, and to the development and management of natural resources in general. Some deficiencies in project implementation were experienced during the initial years, viz., non-use of PPF, delayed replacement of the PC and inadequate counterpart funding. These were, however, later resolved. with changes in senior staff, and their effects nullified towards the end of project implementation. The only difficulty that remained throughout the project period was related to the construction of FBD headquarters. The failure to finalize the necessary preparations and studies for the FBD headquarters was a major shortcoming and a missed opportunity for the Division in improving its working conditions and operational efficiency. G. ASSESSMENT OF OUTCOME 48. The project has met nearly all of its objectives as it has strengthened the institutions responsible for administering and managing the country's forests and land, and hence contributed towards improved management of natural resources. The project has assisted in a significant manner in the formulation of both the Forest and Land Policies. The Land Policy has led to the enactment of new land legislation. An important achievement of the lands component was also the strengthening of the departments engaged in land surveying, recording and registration by improving their working capabilities through the provision of equipment, training and the introduction of modern working methods, such as using the GPS technology. The survey of village lands will be useful for land use planning and for land administration at the village level. Within the forestry component, the promotion of private nurseries has enabled the increased production of tree seedlings. The tree seedlings produced and planted under the project are estimated to cover about 3,500 ha of land, and this will positively contribute to the  Port 1. Implementation Assessment 14 future supply of construction poles and woodfuel. The stoves are already helping families to reduce their woodfuel consumption due to their improved efficiency. The economic rate of return of the tree planting and wood stoves program has been estimated at 12%, which is adequate and in line with appraisal findings. One of the most tangible project results has been the marked increase in royalty collection, and the demonstration that this could be further improved. The project was not successful in constructing new, and badly needed Headquarter facilities for FBD. 49. The project has also strengthened the forestry services in the two regions through training and improved mobility, and are now capable of undertaking much larger programs with more confidence, given their experience under the project. The introduction of ngitiri and JFM has stimulated people to think in terms of participatory management of the country's forest resources, and is expected to direct future development efforts to more cost effective and sustainable management systems. The project has also contributed to reducing the rate of encroachment in forest reserves in Tabora by establishing boundaries and by supporting the physical presence of patrolling Forest Officers. The phenomena of generalized encroachment on forest reserves coincided with the implementation of the project, and it is believed by practitioners in Tanzania Lhat the incidence of encroachment in Tabora would have been far more dramatic if the project was not being implemented there. H. FUTURE OPERATIONS 50. During the final year of implementation, each of the implementing agencies prepared an operational plan which outlined activities which would be undertaken after the project's closing date. GOT has expressed its commitment to providing necessary funds to FBD and to regional and district forest services to continue the activities initiated by the project. It is unlikely that forestry staff will receive incentive payments provided under FRMP, and the impact of this on staff performance could be potentially unfavorable. This is a systemic problem, evident wherever mechanisms are put in place to increase project effectiveness in the short term, without ad.dressing the wider framework in the context of civil service reform. 51. Among other things, Government has made its intention clear to follow up on the institutional reforms proposed in FRMP, and to establish a new autonomous forest agency with greater responsibilities for service delivery than the current institutional structure. The challenge will be to support a substantive change in the rationale for a publicly funded forest service, from one which is fundamentally oriented towards regulation and control, to one which is oriented t1P7-oard the provision of services for rural communities. A future Bank-financed operation is in preparation, which will seek to develop some of the lessons from FRMP, and which will Mcorporate biodiversity conservation objectives. 52. With regard to the lands component, future operations will continue under the normal business undertaking of the SMD, as this institution is expected to receive sufficient funds from its Ministry, because of increased revenue collection. I. LESSONS LEARNED 53 Lessons which have been learned during project implementation include that:  15 Part 1. Implementation Assessment * the concept of strengthening traditional systems of woodland management (ngitiri) is fundamentally sound and has good replication potential throughout the country. The most important aspect of the program has been the support it has given to people to act to halt the disappearance of remaining woodland, and to bring as many pockets of woodland as possible under rehabilitation. This has been accomplished by working with villagers as the logical, long-term guardians of forest resources to protect and conserve these resources themselves, at little cost to government, and through socio-legal mechanisms which they themselves control and maintain according to their own interests. * Community involvement in Joint Forest Management of government forest reserves can lead to an improvement in the condition of reserves, and to a sharp reduction in the threats they face. The pilot approaches could, however, be strengthened particularly in regard to more fully transferring decision making responsibilities to the communities involved. * More generally, dry woodland management poses good opportunities because of the resilience of the woodland to harvesting, the cost-effectiveness of relying on natural regeneration with minimal silvicultural treatments, and because of the few risks to the environment posed by sustainable management of dry woodland ecosystems. * Farmers are apt at adopting simple technologies that are based on the use of inexpensive local material, such as the improved wood stoves. Higher-tech approaches are not needed and are difficult to disseminate through extension. * Community participation in the demarcation and surveying of village lands is essential and cost effective, as this reduces litigation among members to a minimum, leading to considerable savings in time and resources. * Project identification and preparation benefited from both the resources which were put inito the effort, but also from the longer term iterative process of project design, and the linkages which were established with sectoral strategies, such as the Tanzania Forestry Action Plan. Indeed, experience with the project suggests that there can be significant project implementation benefits when project identification and preparation is bolstered by strategic studies, action-oriented research, and ESW. * Weaknesses in project performance were often linked with inadequacies in the prevailing institutional mechanisms for forest management in Tanzania. These place responsibility for forest management within a decentralized institutional structure, without adequate mechanisms for ensuring there is technical, administrative, and management oversight. Problems of governance in the sector are linked to these weaknesses, and future investments will likely be jeopardized in the absence of institutional reforms. The problem is widely recognized in Government, and the Ministry has made a commitment to launching a reform process which seeks to address these systemic weaknesses. * Technical assistance of the type financed by the project may be delivered more effectively if individual consultants are selected on the basis of professional competence and experience. The packaging of consultancies into larger packages during the tendering process limited Government's flexibility in finding qualified individual consultants, though some efficiencies were gained. However, consequent deficiencies in performance were difficult, time  Part 1. Implementation Assessment 16 consuming, and costly to rectify. Technical supervision of advisory services needs to be strengthened and supported. * A Project Coordination Unit may facilitate implementation of project activities, but may have little overall impact on increasing institutional capacity. In this instance, a PCU was justified on the basis of the need for coordination between multiple institutional partners, but provided little scope for the technical oversight needed in the sector. * Payment of project allowances may encourage project staff to work diligently; but, given public spending constraints, is not sustainable in the long-term.  PART II: STATISTICAL TABLES Table 1: Summary of Assessments A. Achievement of objectives Substantial Partial Negligible Not Applicable Macro policies E El El Sector policies [E El El Financial objectives [7] El El El Institutional development El FEl El Physical objectives El El El Poverty reduction E[K] El El Gender issues E El El Other social objectives E [ l E Environmental objectives F E E E Public sector management E El El Private sector development El E El Other (specify) E E E E B. Project sustainability Likely Unlikely Uncertain Highly C. Bank performance satisfactory Satisfactory Deficient Identification E El Preparation assistance E [ E Appraisal E El Supervision E El]  Part 2. Statistical Tables 18 Highly D. Borrower performance satisfactory Satisfactory Deficient Preparation E m E Implementation 2 m F Covenant compliance F Operation (if applicable) Highly Highly E. Assessment of outcome satisfactory Satisfactory Unsatisfactory unsatisfactory Table 2: Related Bank Loans/Credits Loan/credit title Purpose Year of approval Status Preceding operations 1. Sao Hills Forestry Establish industrial 1976 Completed Project plantations 2. Sao Hills Forestry Expand industrial 1982 Completed Project, Phase II plantations Following operations 1. Agricultural Research Strengthen agricultural 1998 On-going research, including agro- forestry  19 Part 2. Statistical Tables Table 3: Project Timetable Steps in project cycle Date planned D Date actual/ l l latest estimate Identification 1984 - 1988 " 1984-88 Preparation March 1990 March 1990 Appraisal September/October 1990 September/October 1990 Post-Appraisal May 1991 May 1991 Negotiations December 9 to 12, 1991 December 9 to 12, 1991 Board presentation February 11, 1992 February 11, 1992 Signing April 24, 1992 April 24, 1992 Effectiveness July 16, 1992 July 16, 1992 Mid-term review March 14 to April 12, 1996 March 14 to April 12, 1996 Project completion March 1999 March 1999 Loan closing June 30, 1998 June 30, 1999 l ____________________ March 31, 1999 1 This period was not continuous. The project identification process commenced in late 1984 when GOT carried out a woodfuel study with the assistance of ESMAP. More in-depth studies were commissioned by ESMAP in 1986, and completed in 1988, with the preparation of a project document that was entitled Woodfuel/Forestry Project. This document became one of the main sources of information during the formulation of FRMP. Table 4: Loan/Credit Disbursements: Cumulative Estimated and Actual (US$ thousands) FY 93 FY94 FY95 FY 96 FY97 FY 98 FY 99 Appraisal estimate 4,156 9,103 11,847 14,803 16,742 18,330 - Actual 1,000 1,710 5,840 9,720 14,210 16,730 18,620 Actualas%ofestimate 24 19 49 66 85 91 102 Date of final disbursement " Due to the depreciation of the US$ vis-A-vis the SDR, disbursement expressed in US$ is over 100% of what had been estimated at appraisal.  Part 2. Statistical Tables 20 Table 5: Key Indicators for Project Implementation Key Implementation Indicators Unit [ Estimated 1 Actual A. Natural Resources Monitoring/Forestry Production of natural resource 64 map sheets To be produced by Produced in 1997 maps covering the country 1995 Establishment of a natural One full-fledged To be established by Completed in 1998 resource information centre information centre 1998 Introduction of new royalty Introduction of new To be introduced by Introduced in 1996 collection system Procedures 1995 Preparation of a new forest policy A new policy Not foreseen at Accomplished in 1998 approved by GOT appraisal Undertake a study on best practice A complete study To be completed by Completed in 1995 for forestry 1995 Undertake a study on long-term A complete study To be completed by Completed in 1996 financing for forestry 1995 Construction of FBD headquarters One complex To be constructed by Not constructed 1995 Purchase of 4 x 4 wheel drive Number Not specified at 7 Station Wagon vehicle (for FBD, appraisal PCU and TANRIC) Purchase of computers (for FBD, Number Not specified at 22 PCU and TANRIC) appraisal B. Strengthening Land Policy and Institutions Prepare a land Policy A new policy To be completed by Completed in 1994 approved by GOT 1993 Village demarcation Number of villages 1,800 3,563 Village survey - do - 1,800 1,818 Village land titling and - do - 1,800 Not required due to registration change in titling policy Study on pastoral and common A complete study To be completed by Completed in 1994 property regimes 1994 Study on land markets A complete study To be completed by Completed in 1994 1994 Study on regional land A complete study To be completed by Completed in 1994 registration/titling 1994 Long-term training M. Sc. geoinformatics Number of staff Not specified at 3 appraisal M. Sc. photogrammetry -do- -do- I Diploma in geoinformatics -do- -do- 5  21 Part 2. Statistical Tables Key Implementation Indicators Unit Estimated Actual Short-term training Surveyors training in project Number of staff Not specified at 20 planning, management and appraisal reporting l Courses on computer application - do - -do- 36 in office administration and planning Use of survey equipment -do- -do- 10 Computer basics for general staff -do- -do- 24 Computer basics for executives -do- -do- 13 Trouble shooting on survey -do- -do- 4 hardware Data base management basics -do- -do- 12 Survey processing and drafting -do- -do- 9 using LISCAD Computer aided drafting using -do- -do- 13 AUTOCAD Arclnfo/ArcView applications in -do- -do- 12 land management Seminars/study tours Overseas seminar in mapping Number of staff Not specified at 2 appraisal Overseas conference on land -do- -do- 2 valuation Study tour on land information -do- -do- 12 systems Purchase of major equipment 4x4 wheel drive station wagon Number 33 40 and pick-up vehicles Bicycles -do- 140 50 Personal computers -do- Not specified at 14 appraisal GPS units -do- -do- 7 Survey instrumernts -do- -do- 5 Camping equipment -do- -do- 20 C. Strengthening Regional and District Forest Services 1. Mwanza Preparation of regional forest dev. Complete plans To be completed by Completed in 1996 Plans 1993 Preparation of district forest dev. Complete plans To be completed Completed for all plans by1996 districts in 1996 Staff trained in extension Number 63 75 Staff trained in management -do- 6 1  Part 2. Statistical Tables 22 Key Implementation Indicators | Unit | Estimated | Actual Teachers trained -do- Not specified 198 Farmers trained Number Not specified at 2,780 appraisal Household covered by extension -do- 1,800 31,680 Social groups covered by -do- 144 883 extension Schools covered by extension - do - 160 1,032 Woodland management - do - Not foreseen at 28,195 appraisal Forest reserve under joint ha Not foreseen at 47,000 management appraisal Nurseries established Number 2,124 9,461 Seedlings produced - do - 7,100,000 7,537,000 Wood stoves installed - do - Not foreseen at 10,850 appraisal 4x4 wheel drive station wagon - do - 13 17 and pick-up vehicles Trucks - do - Not foreseen at I appraisal Motorcycles - do - 28 9 Bicycles - do - 80 75 Computer - do - Not specified at I appraisal 2. Tabora Preparation of regional forest dev. Complete Plans To be completed by Completed in 1996 Plans 1993 Preparation of district forest dev. Complete plans To be completed Completed for all Plans by1996 districts in 1996 Staff trained in extension Number 20 15 Staff trained in management - do - 4 4 Teachers and tabacco-leaf - do - 310 310 technicians trained Farners trained - do - 8,000 8,000 Household covered by extension - do - 720 5,990 Groups covered by extension - do - 48 89 Schools covered by extension no 70 390 Registration of village forest ha Not foreseen at 8,270 reserves appraisal Forest reserve under joint ha Not foreseen at 13,200 management appraisal Forest reserve boundary cleared km Not specified at 941 appraisal Nurseries established Number 540 164  23 Part 2. Statistical Tables Key Implementation Indicators | Unit I Estimated Actual Seedlings produced Number 2,000,000 2,027,000 Vegetation land use map ha 1,500,000 1,000,000 production Reconnaissance inventory survey Number of forest 24 24 for timber and woodfuel reserves Management of priority forest reserves Boundary clearing Km Not specified at 940 appraisal l Boundary ploughing Km Not specified at 239 appraisal Preparation of management Number of forest 27 0 plans reserves Woodland management ha Not foreseen at 29,190 appraisal Villages covered under Number of villages 10 107 woodland management Forest reserves under joint ha Not foreseen at 13,200 management appraisal Villages contracted under joint Village Not foreseen at 7 management appraisal Establishment of plots for Plot 12-24 7 applied research Purchase of major equipment Trucks Number Not foreseen at 2 appraisal Motorcycles - do - 39 13 Bicycles - do - 45 40 Computers - do - Not foreseen at 2 appraisal Table 6: Key Indicators for Project Operation I. Key operating indicators in SAR/President's Report Estimated Actual Not applicable  Part 2. Statistical Tables 24 Table 7: Studies Included in Project Study Purpose as defined at Status Impact of study appraisal/redefined 1. Royalty Collection and Increase revenue by improving Completed in The new system was Monitoring System royalty collection system 1995 introduced, and resulted in increased revenues 2. Best Practices in Forestry Identify forestry practices that Completed in Findings used in the can improve the management 1995 preparation of the new of forest resources Forestry Policy 3. Long Term Financing of Indicate the best alternatives Completed in Findings used in the Forestry Activities for long-term financing of the 1996 preparation of the new forestry sub-sector Forestry Policy 4. Evaluation of Revenue Evaluate the new system Completed in Findings expected to be Collection and Monitoring introduced under FRMP 1999 used in the follow-up System in Singida, Tabora, development programs Shinyanga and Mwanza Regions 5. Evaluation of Extension Evaluate the improvement of Completed in Findings expected to be Services in Mwanza and services carried out under 1999 used in the follow-up Tabora Regions FRMP development programs 6. Evaluation of Project Evaluate project activities Completed in Findings expected to be Experience with Joint Forest introduced by FRMP 1999 used in the follow-up Management and Ngitiri development programns Registration 7. Land Market Review of the land market and Completed in Findings used in the the process of land transfer in 1994 preparation of the Land order to recommend Policy improvements 8. Pastoralists Land Use and Study the implication of Completed in Findings used in the Common Property Regimes pastoral and agriculture land 1994 preparation of the Land uses, and articulate major Policy issues 9. Land Registration Review of present land Completed in Findings used in the registration system in order to 1994 preparation of the Land recommend improvements Policy  25 Part 2. Statistical Tables Table 8A: Project Costs Appraisal estimate (US$M) Actual/latest estimate (US$M) Item Local costs Foreign Total Total costs A. Natural Resources Monitoring and Forest Policies Land Use/Natural Resources Mapping 276.5 1,268.4 1,544.9 1,708.0 Resource Information Centre 28.1 595.3 623.4 648.0 Royalty Collection Activities 105.0 415.9 520.9 614.0 Forest Activities: Best Practices Study 63.0 89.7 152.7 77.0 Long-term Forestry Financing Study 31.5 60.4 91.9 Improvement of Infrastructure 622.8 366.3 989.1 652.0 Project Coordination Office 115.7 313.1 428.9 2,662.0 Sub-total (A) 1,242.7 3,109.2 4,351.9 6,361.0 B. Institutions Strengthening: Land Policy Land Policy Development 379.3 1,265.7 1,645.0 1,659.0 Village Demarcation 1,592.3 1,484.7 3,077.0 3,388.0 Village Titling and Registration 249.1 592.6 841.6 92.0 Deputy Project Coordinator 110.7 172.0 282.7 1,506.0 Sub-total (B) 2,331.4 3,515.0 5,846.4 6,645.0 C. Regional and District Forest Services Mwanza Regional Forestry Program 784.2 2,086.5 2,870.7 2,383.0 Tabora Regional Forestry Program 1,105.3 3,238.8 4,344.2 3,953.0 Tabora Research Program 262.3 302.5 564.8 54.0 Sub-total (C) 2,151.8 5,627.8 7,779.6 6,390.0 Total Base Costs 5,725.9 12,252.0 17,977.0 Physical Contingencies 149.9 250.2 400.1 Price Contingencies 912.0 1644.6 2,556.6 Total Project Costs 6,787.8 14,146.7 20,934.5 19,396.0  Part 2. Statistical Tables 26 Table 8B: Project Financing Source Appraisal estimate (US$ '000) Actual/latest estimate (USS '000) Local costs Foreign costs Total Total IDA 4.8 13.5 18.3 18.6 NORAD - 0.7 0.7 l GOT 1.9 0 1.9 0.8 TOTAL 6.7 14.2 20.9 19.4  27 Part 2. Statistical Tables Table 9: Economic Costs and Benefits General The project proposed to support GOT in addressing broad land and natural resources management issues, and in promoting the involvement of local people in the management of forests. As such, major project achievements have been in policy development, institution-building and improvement of natural resources information base. These benefits are by their nature not quantifiable, but were found to be essential for justifying the project. Project Benefits At project appraisal, quantifiable benefits were assessed from two small components, i.e. improved forest management and forestry extension. At completion, however, only benefits from forestry extension can be considered for quantification, since the forest management envisaged at appraisal did not materialize as originally intended. Benefits from forestry extension are based on increased woodfuel production, as the result of the tree seedlings distributed by the project, as well as from woodfuel savings obtained through the introduction of improved stoves. Comparative studies carried out by the project show that the improved stoves have a saving of about 50% on woodfuel consumption, compared with the traditional ones. Assuming an average family consumption of 5 kg per day, the savings of 2.5 kg have been assessed as project benefits and used in the economic analysis. The project had distributed about 9.4 million seedlings of different species (out a total production of 9.6 million), which are equivalent to a plantation area of about 3,520 ha, assuming a plant population of 1,600 seedlings per ha and a survival rate of 60%. Eucalyptus is the most prevalent species distributed, hence it has been taken as a representative plant, and the analyses are based on one ha eucalyptus model. Although some of the trees planted under the project would be sold as poles, the bulk of the production would be used as woodfuel. Hence the valuation of produce was based on woodfuel. Woodfuel production has been calculated over 21 years, three rotations of seven years, and were estimated on the basis of yields of 18 m3 per year per ha for the first rotation, and 20 m3 per year per ha for the following two coppice rotations. Project Costs Since the economic analysis is carried out for woodfuel and stoves, the costs related to the regional and district forestry services in Mwanza and Tabora have been included in the cost stream, although these comprised expenditures not related to extension such as costs related to improve revenue collection. These costs were adjusted to 1999 prices using the consumer price index for the country. The investment and operation and maintenance costs used in the one ha model are based on 1999 prices. Market and Prices Most of the tree plantings have taken place in Mwanza Region, where woodfuel demand is high. Hence, project output would contribute positively to the regional supply, albeit in a limited manner. The price of woodfuel has been increasing in Mwanza Region, particularly in the urban area. The average market price now varies between TSh 1,700-2,500 per stack of 15 kg, and many households are using kerosene as a substitute or a supplementing source of energy. In the analysis, the economic value of woodfuel has been expressed in terms of Kerosene, and this has been calculated at TSh 23,310 per m3. Economic Rate of Return Based on the assumptions made above, an economic rate of return has been calculated for the project over 27 years. This yielded a return of 12%, which is the same as estimated at appraisal for the extension component.  Part 2. Statistical Tables 28 Table 10: Status of Legal Covenants _ _ i l l |Revised l_______ - Covenant Present Original fulfillment l Agreement Section type Status fulfillment date date Description of covenant Comments DCA Text 3.01(a) 04 CD The Borrower... shall Funds needed provide, promptly as needed, were not flowing the funds... .required for the as expected in the Project. initial years, but improved later. DCA Text 3.03 12 C The Borrower... shall Annual reviews annually review royalty fees completed. and as of July 1 of each year, adjust them by at least the rate of inflation. DCA Text 3.04 03 C The Borrower shall ensure Accounts that MTNRE and MLHUD established as establish independent project required. accounts in the regions and districts... DCA Text 3.05 12 C January 1, The Borrower shall... Completed on 1994 complete an assessment of time. long-term fnancing needs of the forestry sub-sector. DCA Text 4.01 01 C Maintenance of adequate Accountability records and accounts of issues raised by operations, resources and some audits have expenditures. been addressed. DCA Text 4.01 (b) 01 C Furnish to IDA not later than Audits up to date. 9 months after the end of each FY certified copy of audited accounts including (i) those for Special Accounts; and (ii) separate auditors' opinion on SOE. DCA Text 5.01 (a) 10 C Secondment of graduate Completed on and (b) foresters as Regional Forestry time. Extension Officers and appointment of DPC. DCA Para. 1 05 C The Borrower shall establish PCC continued Schedule 4 a Project Coordination operating Schedule Committee (PCC) effectively. DCA Para. 5 09 C The Borrower shall establish Work programs Schedule 4 by December I in each year were prepared furnish to the Association an regularly and l annual report... and a work reports submitted program for the following on time. year.  29 Part 2. Statistical Tables Revised Covenant Present Original fulfillment Agreement Section type Status fulfillment date date Description of covenant Comments DCA Para. 6 10 CD The Borrower shall by end of Completed but Schedule 4 ... year complete ... with some delays. demarcation of ... villages. The program was expanded to more regions. DCA Para. 7 10 CD March 31, The Borrower shall by March Completed but Schedule 4 1993 1993 furnish... (a) extension with some delays. program for Mwanza, and (b) regional forest resource plan for Tabora. DCA Para. 8 10 CD March 31, The Borrower shall by March Pilot revenue Schedule 4 1993 1993 (a) furnish collection system recommendations for revenue in place after reporting; and (b) ... some delay. implement. DCA Para. 9 10 CD Evaluate pilot Review Schedule 4 implementation of revised completed at revenue collection system. project completion. DCA Para. II 10 C May 31, 1994 . . .initiate its activities for Completed on Schedule 4 identification of best time. practices in Forestry furnish a report. DCA Para. 13 10 C May 31, 1995 ... initiate study on land Completed on Schedule 4 policy and furnish to time. Association for review. DCA Para. 14 9 C September 30, ... carry out jointly with the Completed on Schedule 4 1995 Association a mid-term time. review of the Project. Covenant Types Present Status I = Accounts/audits C = Complied with 2 = Financial performance/revenue CD = Complied after delay 3 = Flow and utilization of project funds NC = Not complied with 4 = Counterpart funding CP = Complied with partially 5 = Management aspects of the project of executing agency 6 = Environmental covenants 7 = Involuntary resettlement 8 = Indigenous people 9 = Monitoring, review and reporting 10 = Project implementation not covered by categories I - 9 11 = Sectoral or cross sectoral budgetary or other resources allocation 12 = Sectoral or cross sectoral policy/regulatory/institutional action 13 = Other  Part 2. Statistical Tables 30 Table 11: Compliance with Operational Manual Statements Statement number and title I Describe and comment on lack of compliance No deviation from applicable Operational Manual Statement Table 12: Bank Resources: Staff Inputs Stage of project cycle Planned Revised Actual Weeks US$ Weeks US$ Weeks US$ Preparation to appraisal NA NA NA NA 171.6 324,100 Appraisal NA NA NA NA 169.2 433,600 Negotiations through NA NA NA NA 10.3 31,000 Board approval Supervision NA NA NA NA 156.7 512,200 Completion NA 10 40.9 0.7 2,200 TOTAL 508.5 1,303.10 i/ Including travel cost (Direct Cost Only) NA= Not available-not recorded in the FACT System  31 Part 2. Statistical Tables Table 13: Bank Resources: Missions Performance rating b/ Stage of project Month/ Number of Days in Specialized Implementation Development Types of cycle year persons field staff skills status objectives problemsc represented"' Preparation d/ 03/90 10 230 AG, E, E, F, l F, F, LP, LP, SL, SL Appraisal 09-10/90 9 243 AG, E, E, l Fl, FO, I, 1, NR, SL Post-Appraisal 05/91 5 105 A, E, FO, - - FO, NR Supervision 1 07/92 3 39 E, FO, PR 2 1 CF, PR Supervision 2 10/92 3 42 E, FO, PR 2 1 CF, FP Supervision 3 04-05/93 1 15 FO 2 2 CF, FP, PC Supervision 4 03/94 2 28 FO, SL 2 1 CF, FP, PR Supervision 5 10-11/94 2 42 FO, SL S S - Supervision 6 08-09/95 3 54 Fl, FO, SL S S - Supervision 7 03-04/96 4 91 Fl, FO, PR, S S l SL Supervision 8 03/97 3 43 Fl, FO, PR S S - Supervision 9 01-02/98 3 51 E, FO, PR S S - Supervision 10 09/98 2 26 E, FO S S a/ A= Anthropologist; AG= Agriculturist; E = Economist; EC = Ecologist; EN = Engineer; FI = Financial Analyst; FO = Forester; IN= Institutions Specialist; LP = Land Use Planner; NR= Natural Resource Specialist; PR= Procurement Specialist; SL =Sociologist/Land Policy Specialist. bl S = Satisfactory; U = Unsatisfactory; HlU = Highly Unsatisfactory; HS = Highly Satisfactory; NR = Not Rated c/ CF = Counterpart Funds; FC = Financial Covenants; FP= Financial Performance; PM = Project Management; LC Legal Covenants; PR = Procurement d/ Prepared by the FAO/World Bank Cooperative Program, with the participation of the World Bank, British Overseas Development Agency and the Dutch Development Cooperation.  TANZANIA FOREST RESOURCES MANAGEMENT PROJECT (CR-2335-TA) IMPLEMENTATION COMPLETION REPORT MISSION, MARCH 9 TO 30, 1999 APPENDIX A Aide-Memoire A. INTRODUCTION 1. A mission comprised of Pietros Kidane (Mission Leader, Economist, FAO), Donald Sungusia (Agricultural Services Specialist, World Bank), and Richard Kaguamba (Forestry Specialist, World Bank) visited Tanzania from March 9 to 30, 1999 to initiate preparation of an Implementation Completion Report (ICR) for the World Bank-financed Forest Resources Management Project (FRMP).10 The mission worked mainly with the Ministry of Natural Resources and Tourism (MNRT), the Ministry of Lands and Human Settlements Development (MLHSD) and with the Project coordination Unit (PCU). Field visits were undertaken to Tabora, Shinyanga and Mwanza. Discussions were also held with staff from the Institute of Resource Assessment of the University of Dar es Salaam. 2. Upon arrival, the PCU provided the mission with a Project Implementation Overview Report, and this has been used in the preparation of this Aide Memoire. The mission would like to express its appreciation for the support extended by officials at headquarters in Dar es Salaam and in the regions. Preliminary mission findings and conclusions are presented below. B. BACKGROUND 3. Before the formulation of the Forest Resources Management Project (FRMP), IDA had financed two forestry projects to establish industrial plantations, from the mid-1970s to the end of the 1980s. During this period, support was also provided to the forestry sector through different integrated rural development projects. A study on wood energy (which also identified potential areas for project investment) was also launched by the Joint World BanklUNDP Energy Sector Management Assistance Program (ESMAP). Forestry development efforts in those days followed traditional lines, and were mainly focused on tackling specific forestry topics and activities. However, this approach was inadequate for dealing with the complex problems that were emerging as a result of population pressure, and the consequent high and conflicting demands for natural resources. These circumstances highlighted the need for a change in the development approach to address broad sectoral issues, including policy and institutional matters, and issues related to the participation of the private sector and communities in tree planting and forest resources management. Such a broader approach to forestry development obtained a general consensus within the main development institutions, including the World Bank. Inspired by this new approach, the Government of Tanzania (GOT) carried out a major 'I P. Dewees, World Bank Task Team Leader for the project participated in discussions during the latter part of the mission and also contributed to preparing this Aide Memoire.  33 Appendix A study of the forestry subsector and, with donor assistance, prepared the Tanzania Forestry Action Plan (TFAP) in the late 1980s. Further, as a follow up to the TFAP, Government, with the assistance of the FAO/World Bank Cooperative Program, drew up a comprehensive investment proposal in March 1990 for World Bank financing. The project proposal, commonly referred to as FRMP, was appraised by IDA in May 1991. 4. The project had three specific objectives: (a) to strengthen institutions in charge of the development of forest policies and natural resources information; (b) to support the development of rural land policies, strengthen crucial governmental and non-governmental organizations in charge of the implementation of these policies, and support the implementation and monitoring of government's policy of village demarcation and titling; and (c) to strengthen regional and district forestry institutions in Mwanza and Tabora Regions. Project components comprised: * Strengthening Natural Resources Monitoring and Forest Policies, through: (i) production of national land use and natural resource maps; (ii) development of a natural resource information center; (iii) improvement and monitoring of the royalty collection system; (iv) identification of best practices for forest activities; (v) preparation of an action plan for the long-term financing of the forestry subsector; and (vi) improvement of accommodations for FBD's main offices. * Strengthening Land Policies and Institutions, by supporting: (i) the development of land policies and (ii) village demarcation, titling and registration. * Strengthening Regional and District Forest Services in Mwanza and Tabora. In Mwanza, activities were to include: (i) survey and planning of the extension program; (ii) district forest service reorganization and training; (iii) development and distribution of extension materials; (iv) community and farm forestry extension; (v) local research and development; (vi) revenue collection and sharing; and (vii) village woodland and forest reserve management. In Tabora, a one-year preparatory phase (Phase I) was to include: (i) vegetation mapping; (ii) reconnaissance surveys/inventories; and (iii) regional forest resource planning. Phase II which was to last five years, subject to the findings of Phase I, was to include: (i) forest reserve management; (ii) village woodland management; (iii) applied research; (iv) community and farm forestry extension; (v) revenue collecting and sharing; and (vi) training. 5. Many of the project's activities were envisaged and designed as pilot operations, intended to test out a range of approaches for improving woodland management. 6. Total project costs were estimated at TSh 6.1 billion, or US $ 20.9 million, with a foreign exchange component of 68 percent. The project was expected to be implemented over a period of six years. About 88 percent of the project costs were to be financed by IDA, 9 percent by GOT and about 3 percent by other donors. l As indicated in the Staff Appraisal Report, it was expected that NORAD would provide US $0.7 million in co- financing to support the Natural Resources Information Centre. This co-financing did not materialize and was not considered during the approval and signing of the credit.  Appendix A 34 7. The management of the project was to be vested with the two ministries and two regions, i.e. MNRT, through its Forestry and Beekeeping Division (FBD) for strengthening forest services; MLHSD for the land policy, village demarcation and registration activities; and Tabora and Mwanza Regions for related forestry and land activities in the field. Coordination of project activities was to be carried out by a high level Project Coordination Committee (PCC) whose members would be largely from the two ministries and two regions, and from the Prime Minister's Office (PMO). A Project Coordinator (PC), to be appointed by MNRT, and a Deputy Project Coordinator (DPC), to be appointed by MLHSD, were to form the secretariat of the PCC. Three tiers of responsibility for implementation were envisaged for the forestry component: national level under FBD, regional level under the Regional Forest Officer (RFO); and district level under the District Forest Officer (DFO). Overall implementation responsibility for the lands component was to be vested with the Survey and Mapping Division (SMD), but the Regional Surveyor (RS), who reports to SMD, was to be responsible for the lands component in each regions. 8. The project was approved by IDA on February 11, 1992, and a Development Credit Agreement (DCA), which provided funds totaling SDR 13.2 million, or US $18.3 million, was signed on April 24, 1992. The Credit was declared effective on July 16, 1992. C. PROJECT IMPLEMENTATION AND RESULTS General Overview 9. IDA approved a Project Preparation Facility (PPF) advance totaling US $650,000 to finance preparatory tasks for project implementation, including the recruitment of the PC, procurement of equipment, material and civil works, as well as for planning of activities and preparation of bid tendering documents. However, the PPF was not utilized, mainly due to communication difficulties between various GOT departments. This meant that preparatory tasks had to be carried out during the time envisaged for implementation, hence delaying the scheduled project launch. The PCC was established and the PC was appointed, consistent with the project's legal covenants. However, the post of PC was vacated soon after the project became operational, since the PC was promoted to the position of Director of FBD, although he continued to act as part-time PC until his successor was appointed in August 1993. The delay in appointing a full time PC hindered the rapid development of the planned organization and management of the project, and significantly delayed the discharging of the PC's responsibilities in preparing work plans and in reporting and monitoring project activities. 10. In addition to the delays recorded on account of not undertaking preparatory tasks, progress in implementation was impaired in the first year by lack of funds, staff, office facilities and delays in opening project accounts in the regions. Only in the second year of the project was some progress in the field observed, following the recruitment of staff and the procurement of a few vehicles. A major leap in project implementation was recorded only in the third year, when the recruitment of consultants was virtually completed, and procurement of considerable numbers of vehicles and other equipment and material was undertaken. By this time, most of the serious implementation problems were resolved, and project management was well positioned to provide the support required by the implementing agencies. Thereafter, implementation progress continued satisfactorily.  35 Appendix A 11. One problem that continued to persist throughout project implementation was a shortage of GOT counterpart funds, particularly with regard to the forestry component. Major crises were avoided mainly due to increased use of FBD funds retained from revenue collection. The project was able to take advantage of a more general agreement between GOT and the Bank that revenue-generating public institutions could retain revenues to meet counterpart funds requirements, rather than to be fully reliant on budgetary commitments from Treasury. The issue of the shortage of GOT funds and the difficulties encountered in financing the project's operating costs were raised by all IDA supervision missions. The final supervision mission expressed its concern about the likelihood of sustaining the project's impacts if funds are not available to the regions to continue carrying out project-related activities such as boundary clearing and maintenance, revenue collection, and the implementation of woodland management plans. It recommended that MNRT should consider preparing a request to Treasury to increase its share of retained revenues to finance on-going operations, and should consider also reducing its own share of retained revenues at HQ in the interests of ensuring that regional budgets are adequate for sustaining the project's investments. This recommendation was not acted upon, but the issue is more generally being addressed by a Government Task Force. 12. A mid-term review was carried out for the project in March/April 1996 by a joint GOT and IDA team. This review confirmed the validity of the original project objectives, and expressed its satisfaction about the project's progress. The original project closing date of June 30, 1998 was extended twice: first to March 31, 1999, and finally to June 30, 1999 (to complete the required audits of project accounts). As of end of March 1999, disbursements amounted to US$ 18.53 million, or slightly over 98 percent of total credit funds. Strengthening Natural Resources Monitoring and Forest Policies 13. Land Use and Natural Resources Mapping. In mid 1997, technical assistance personnel had successfully produced the countrywide maps envisaged under the project, albeit with about two and half years delay from the original plans. The 64 map sheets produced by the project were based on the most recent satellite imagery, and, among other things, they show general classes of forest cover and agricultural and settlement areas. They also provide baseline information on the different uses of land, and this has facilitated the tasks of assessing natural resources and preparing development plans. Another use of the maps has been for regular monitoring of forest resources and changes in land use. The maps have become popular and are widely used by different governmental and non-governmental organizations. The PCC had ruled that the maps should be produced in the facilities of SMD instead of accomplishing this abroad. Hence the project improved the poor state of SMD facilities, contributing significantly to the domestic map production capability. However, in spite of the improvements, the maps were produced abroad, because of doubts about obtaining stable pH level of the water and stable electricity supply. 14. Natural Resource Information Center. In line with project objectives, the Tanzanian Natural Resources Information Center (TANRIC) was established by the Institute for Resource Assessment (IRA) of the University of Dar es Salaam. The project provided technical assistance, computers and other office equipment, and funds for purchasing satellite imagery, maps and for ground survey staff. Training in geographic information system (GIS) was also provided under the project. Presently, TANRIC is an important entity with a considerable information base on  Appendix A 36 natural resources and with capabilities for collecting and analyzing data and delivering the latest information to customers. The demand for its services is increasing, and its business has enabled it to become almost a self-supporting institution. GOT has decided to attach TANRIC to IRA, hence resolving conclusively the uncertainty that prevailed about the center's institutional arrangements. 15. Improvement and Monitoring Royalty Collection Systems. The project aimed at improving the cost effectiveness of GOT's royalty collection. To this end, the project proposed to develop a revenue reporting system, introduce a pilot phase of revised revenue collection method, undertake the feasibility of assessing fee collection from woodfuel used for tobacco curing, and evaluate during mid-term the success of the new system. These activities were carried out with much delay and have frustrated the original project plans. Because of the delay in recruiting consultants, and later, due to a poor dialogue between the consultants and FBD, a new revenue system was introduced only in 1996, initially in the four pilot regions, i.e. Tabora, Shinyanga, Singida and Mwanza, and later extended to all the regions. The system consisted of registering of business ventures in harvesting, processing and trading of forest products, as well as establishing checkpoints, including near train stations, and stepping up forest patrolling activities. During 1998 about 400 people were registered in the four regions, of which about 210 were in Mwanza. 16. Collection of revenue levels improved soon after the introduction of the new system in the four regions from about TSh 90.72 million in 1996, to TSh 277.86 million and TSh 299.17 million for 1997 and 1998, respectively. The increase of revenue in Mwanza was a major jump from TSh 8.70 million in 1996 to TSh 103.0 million in 1998. The contrast between the high collection of revenues of Mwanza, which has little forest resources, and the low collection of Tabora, which has high forest cover, is explained by failure of the collection efforts in Tabora. Mwanza was collecting revenues mainly from wood that was coming from Tabora. In spite of the overall 3-fold increase of revenue (and the nearly 12-fold increase in Mwanza), the total collected from the regions is likely to represent less than 10 percent of the potential revenue which could be collected, indicating yet again the existence of inefficiency in collection. An evaluation of the revenue collection system was prepared in advance of the ICR, and its findings will be incorporated into the ICR. Many reasons have been given for the poor revenue collection, including poor data collection and monitoring, poor patrolling, weak institutional capacity, poor management and dishonesty of some individuals in the field. An important explanation of the problem has also been of an institutional nature i.e. organizational fragmentation. Although the forest reserves owned by GOT are under the jurisdiction of the FBD, the RFOs and DFOs are not employed by FBD, but by the Regional and District Administrations, respectively. Therefore, FBD directives may not be followed totally in the field. DFOs, who are materially collecting revenues from the forest reserves owned by FBD and by the districts, may give priority to district matters. It should be recalled that FBD shares part of the revenues it collects with the RFOs, but not directly with DFOs. In order to obviate these difficulties, MNRT is considering establishment of a forestry 'executive agency,' with one line of command, to be responsible for the overall management of the forest reserves and for royalty collection. 17. Identification of Best Practices for Forest Activities and Long-Term Financing of Forestry. FBD recruited consultants to undertake these two studies. The study on best practices  37 Appendix A provided useful and innovative ideas on many issues, including the involvement of local communities in the management of public lands and forest reserves, and the private sector in the management of plantations. The findings and recommendations of this study have been reflected in the new Forest Policy. Preparation of the latter has benefited significantly from FRMP, as the project provided finance for many studies and supported workshops and seminars for the realization of the policy. The study on long-term financing has also provided useful information on the present valuation of forest products and recommended a change of the current valuation system. The study indicated that self-financing of future forestry operations would be guaranteed, if effective collection of the revenues could be undertaken. 18. Improvement of Infrastructure. The project proposed to resolve FBD's problem of limited office space by constructing and rehabilitating offices on a government site in Dar es Salaam. Although an initial design for the structure was already developed before project appraisal, and architects were appointed at the end of 1993, FBD has not succeeded in having its new headquarters built. Last minute changes in site plans and excessive delay in advertising tenders did not allow tenders to be finalized or awarded by the end of July 1997, the latest dead- line agreed with IDA supervision mission. Consequently, the construction of the building was excluded from the project, and the funds were reallocated to other project components. Strengthening Land Policy and Institutions 19. Land Policy Development. The land policy studies envisaged under the project were modified in mid 1993 to include support for the process of land policy formulation. Accordingly, consultants were recruited to assist this important GOT policy formulation exercise, which was completed with the adoption of a National Land Policy by Cabinet in April 1994, and by the Parliament in July 1995. The Land Policy eventually led to the promulgation by Parliament of new legislation in February 1999 (the Land Act and the Village Land Act). 20. Lands Records Storage and Management. The project also assisted MLHSD in improving the organization of land records and cadastral information system. Following a review of the poor state of land records storage and management, a plan was drawn up to initiate an improvement whose main thrust was to save critical information contained in the deteriorating records by storing them in a computerized data base. To this end, the project developed a comprehensive system in computerized data management, and provided staff training, technical assistance, computers and other office equipment. The project succeeded in introducing a system for the safe storage and easy use of land records, and about 45,000 records or some 18 percent of the total have been already computerized so far. 21. Village Demarcation, Titling and Registration. Demarcation, survey and titling of village lands was envisaged to take place in five regions, but this was later expanded to cover eleven regions. Project activities commenced with demarcation, and progress accelerated after the mid-term review of the project, when staff mobility improved. Demarcation activities were undertaken with the involvement of the village people, after project staff fully appraised them about the purpose of the demarcation. Initially, the involvement of villagers in the demarcation process was less satisfactory, but this was later corrected, following a conscious effort by project management to improve the participation of the people. Demarcation of village lands was completed in Dodoma, Morogoro, Singida, Tabora, Shinyanga, Mwanza, Mara, Ruvuma and  Appendix A 38 Kagera regions. Small areas of Arusha and Kilimanjaro regions remain undemarcated, pending settlement of disputed boundaries. Final surveying commenced only in August 1998 due to the delayed arrival of the necessary equipment, particularly the global positioning system (GPS) units. Once the equipment was in hand, however, the process of surveying progressed rapidly. At completion, the project has demarcated about 3,560 villages and surveyed some 1,820 villages. With the new Land Policy, the original plan to issue village land title to the village administrations was changed, as titles are now to be issued directly to land owners, whether they are individuals, clans, or the village administration. Instead, the village councils will be provided with an administrative certificate, together with the boundary maps, of their respective villages. Strengthening Regional and District Forest Services 22. Forestry Extension. Given the limited vegetative cover in Mwanza, the project's forestry extension efforts promoted tree planting in this region. The involvement of people in tree planting and in the management of woodlands, was also considered important in Tabora as a means of checking deforestation caused by the use of wood for tobacco curing. It was originally an objective of the project to rely on technical assistance inputs to develop innovative and pilot approaches to forestry extension, to develop good baseline information and to put in place an effective monitoring and evaluation system, and to draw heavily in this regard on the experience of NGOs which had been working in this area. The consultants who were contracted to provide these inputs had insufficient experience to explore or promote alternative extension approaches, and the result was that forestry extension was developed along fairly traditional lines, i.e. the project simply promoted tree planting, independently of any understanding of the social and economic factors which influence household decision-making in this regard, or for otherwise understanding how tree and woodland management could be incorporated into land-use practices. 23. The project did however build on earlier experiences in Mwanza by focusing on providing support for Village Extension Agents (VEAs) who were designated by the villagers to become the focal point for receiving technical messages and training in nursery and tree planting techniques. The main topics covered by forestry extension were tree planting, individual or community woodland management (ngitiri), improved stoves and Joint Forest Management (JFM), i.e. management of government forest reserves with communities. 24. By far the most important extension activity, in terms of project resource use, has been the promotion of tree planting through individuals, groups and institutions. Extension messages and training in nursery techniques, species selection, tree planting and tending were, therefore, given to a large number of individuals and to appreciable number of institutions, particularly schools. Over 10,000 nurseries of different sizes have been established in both regions, and about 9.5 million seedlings have been produced. The project contacted over 34,600 households and provided training to about 13,700 farmers and school teachers. In addition to delivery of extension messages, tree planting was also promoted through the provision of tree seeds, polythene tubes and pesticides. The provision of inputs in this manner, however, is not clearly sustainable. 25. The tree planting program has largely met the physical targets set at appraisal, but the impact of this cannot be assessed due to the failure of the consultants to assist in establishing a  39 Appendix A monitoring and evaluation system for the extension activities. About 80 percent of tree planting activities were carried out in Mwanza. In Tabora tree planting efforts were geared mainly towards tobacco growers in order to induce them to produce their own woodfuel for curing tobacco. However, since woodfuel is still abundantly available in the natural woodlands of Tabora, the prospects for convincing farmers to grow trees for woodfuel purposes are limited. Hence, exploring different extension approaches suitable for this region would have been in order. An evaluation of the project's extension efforts was carried out in advance of the ICR mission, and the findings of this evaluation will be drawn upon when preparing the ICR. 26. Despite the weak support provided for the development of new approaches toward forestry extension, the project did support two particular sets of initiatives which were relatively unconventional. The first involved building on traditional practices of setting aside tracts of land for later or emergency use. Originally a concept which was applied to grazing areas, ngitiri, the project extended the idea to residual pockets of woodlands. The second approach was an effort to share the benefits and responsibilities for the management of Government Forest Reserves with communities through what was called Joint Forest Management (JFM). An independent evaluation of the project's ngitiri and JFM activities was carried out, and the findings will be reported upon more fully in the ICR. 27. One of the main observations of the ngitiri program (which began on a modest scale in Mwanza, and then was adopted in Tabora) was that it did not go far enough in reconstructing and modernizing the concept. This is partly because the approach only received halting support from the technical advisers who were engaged to provide assistance in this area. In some areas, ngitiri were simply registered, and there was inattention given to: where and how authority over common land ngitiri is lodged; to guiding villagers about protection strategies; and to encouraging villagers to exploit the considerable powers they possess to regulate access, use and management actions in general. The evaluation also noted the large discrepancies in the areas reported to be under ngitiri management, and suggested that hectarage was, in any event, not an appropriate measure of success. Rather, the reviewers noted, the value of ngitiri is in the support it gives to ordinary people to act to halt the disappearance of what little woodland is left, and to bring as many pockets of woodland as possible under rehabilitation. Dismissing ngitiri as insignificant because they are small (as the technical advisers seemed to have done) is unwarranted, simply because they are highly significant to their owners. The evaluation concluded that the ngitiri concept is fundamentally sound, and has good replication potential throughout the country. The most positive aspects of the ngitiri program are those which targeted villagers as the logical, long-term guardians of forest resources, and which work with them to protect and conserve these resources themselves, at little cost to government, and through socio-legal mechanisms which they themselves control and maintain according to their own interests. 28. Joint Forest Management was an approach adopted on a pilot basis in the Urumwa Forest Reserve, a 13,000 ha reserve 15 km or so from Tabora. In practice, 6 user groups have entered into contractual arrangements with Government which provide them with exclusive harvesting rights of specified small coupe areas in return for protecting these and other areas from illegal use. In addition, each household within each user group is supposed to pay an annual fee of TSh 500. The evaluation concluded on the one hand, that community involvement in Urumwa had led to an improvement in the condition of the Reserve, and to a sharp reduction in the threats  Appendix A 40 it faces, but on the other, that (like the ngitiri effort), the approach taken didn't go far enough, particularly in regard to more fully transferring decision making responsibilities to the communities involved. The evaluation also noted that there had been a number of very positive shifts during implementation in the narrow approach originally proposed, and that this suggested some interesting opportunities for expanding upon a significantly modified approach in future. 29. Under the project, applied research in dry woodland management is being carried out by the Tanzania Forestry Research Institute (TAFORI), and it is hoped that some important technical advise in managing the woodlands will emerge in the future. Although the research started late i.e. in 1996 some experiments such as the regeneration characteristics of clear felled woodlands, and in understory burning as a management tool appear to give promising results. 30. An important project achievement, not foreseen at appraisal, was the introduction of improved stoves. The wood stoves being introduced particularly in the rural areas are simple to construct and made from local material i.e. from mixing different types of clays. The extension method is based on showing women how to install stoves by actually constructing them, and letting each person to replicate the process separately or in a team. This approach has had a domino effect, as the trained women are showing others how to construct the stoves wherever they go. Until now, the main emphasis was on promoting wood stoves, but improved charcoal stoves, manufactured by women, are also being introduced by the project. Studies carried out by project staff indicate that the improved stoves are more woodfuel efficient and a saving of about 50 percent of woodfuel is obtained compared to the traditional stoves. 31. Forest Reserve Management. Expectations at Appraisal, with regard to the preparation of detailed forest management plans for priority forest reserves were never met. A considerable amount of work was carried out in Geita (in Mwanza), but this has been undermined because of mining interests -- a concern which had been identified at Appraisal. The proposed inventory for priority forest reserves in Tabora Region has not been undertaken as the necessary maps have not been delivered by IRA, and the consultant Forest Management Specialist (who was to provide assistance in carrying out the inventory) left in late June, 1998. Even with the type of information which would be included in a forest management plan, there is wide agreement that implementation would be problematic in any event, given the difficulties of forest protection because of encroachment and illegal timber harvesting and charcoal production. The dearth of forestry staff in the districts and the lack of cooperation from many districts has exacerbated the problem. Joint Forest Management is seen as a partial solution. It was expected that the management plans were to describe reserve demarcation, forest inventory, timber regulation, community participation and pilot management studies. The failure in preparing forest reserve management plans, is mainly due to the poor performance of the same consultants who failed to provide adequate support for the development of extension forestry. The project had carried out aerial photography of some forest reserves of Tabora as an input to the management plans, and these are currently with TANRIC. It is hoped that these will be useful for preparing management plans for some forest reserves in the future. 32. More general planning activities were, however, undertaken, including the preparation of regional forest development plans for both Mwanza and Tabora. All of the districts in Mwanza region prepared district forest development plans, while around 30 percent were completed in Tabora.  41 Appendix A 33. Even in the absence of detailed priority reserve management plans, the project did achieve some level of success in re-establishing forest reserve boundaries to provide a basis for management and protection. In Tabora region, where these activities were undertaken, around 940 km of boundaries were cleared (having been allowed to deteriorate since the early 1970s), and these have served to separate the reserves from other types of land use and ownership, and from JFM and applied research mentioned above. Boundary patrolling by bicycle and motorcycle was also reintroduced during the project period. 34. Forest reserve management in Tabora Region has also been hindered by illegal encroachment. As many as 65,000 people have been in the past reported to have encroached in reserved areas, and this has of course hindered the introduction of better management. While the problem was not evident during project preparation or at appraisal, and only developed during the project period (coincident with the many political and economic reforms which took place during this period), the situation has been exacerbated by a shortage of forest guards and by the resulting poor state of protection afforded to the reserves. The problem is also an institutional one because of an almost complete lack of cooperation from districts, particularly district forestry staff, who are the principal front line management and protection officers. Despite Government's efforts to deal with encroachment through other non-project mechanisms,12 problems persist. Illegal logging and charcoal production continue in many regions. Organization and Management 35. It took almost two years before project management was put on a strong operational footing, mainly due to the failure in using the PPF, and delays in replacing the PC and in the recruitment of several core staff. The time passed without a full time PC and other core staff has meant that several project tasks could not be carried out as planned. These included preparation of adequate annual plans and budgeting, procurement of goods and services, reporting and monitoring of project activities, and the maintenance of strong linkage with staff in the field. Fortunately, these difficulties were largely resolved towards the end of the second year of the project, as project management went into its full strength and assumed its expected role in facilitating project implementation. The PCC met regularly to review and approve project work plans and to resolve eventual issues. The PC provided continuous support to regional project staff, and facilitated linkages with other concerned bodies. The DPC has also provided similar support to project staff implementing the land component, and facilitated improvement in their performance. 36. Implementation plans were prepared based on the existing GOT organizational structure, which was characterized by a fragmented line of institutional responsibility. It was stated that overall responsibility for the forest component, as well as implementation of the components regarding natural resources monitoring and forest policies were vested with FBD/PCU. However, in the field, the RFOs were given the responsibility for implementing the project at the regional level and for coordinating and monitoring district level activities, while DFOs were in charge of the actual implementation of project activities in the districts. Even the flow of funds 12 During the project period, around 1,200 illegally encroaching people were evicted from forest reserves in Tabora Region, but this was undertaken without the involvement of project staff or funds. Large numbers are also believed to have moved from reserved areas voluntarily because of a public campaign against illegal encroachment, and following clarification of boundaries.  Appendix A 42 to the field offices were different i.e. through FBD/PCU to the regional accounts, and through PMO/PCU to the district accounts. Since the regions and the districts have their respective programs in their areas of jurisdiction, and since they are answerable to the commissioners in the regions and districts, the RFOs and DFOs tended to implement the project along their habitual independent manners. In this process, FBD felt neglected and with little ownership for the project, although it was still expected to have overall responsibility for implementation as the technical Division of MNRT. Early during implementation, FBD expressed its misgivings about the independent mindedness of the RFOs and about its own limited involvement in implementation, because even the staff it seconded to the regions, i.e. the Regional Forest Extension Officers (RFEO) were answerable to the RFOs. Project management tried to resolve the issue by requesting the RFOs to inform the FBD on all project related matters. The RFOs, in turn, complained about the independent-mindedness of the DFOs, although they were meant to coordinate and monitor district level activities. These problems did not occur with regard to the land component, because this was being implemented by the Director of Survey and Mapping Division through the RS. 37. The diffused line of command has not facilitated the emergence of a clear organizational setup for implementing the forestry component. Despite these difficulties, the ad hoc organization and management system put in place by the project, i.e. the PCC and PCU, has been successful in implementing the complex development program under FRMP. However, the wisdom of establishing a management system that artificially tended to centralize activities that have been long decentralized, is questionable, although this is being justified as a necessary mechanism to coordinate the activities of two ministries. The fact is this coordination system will be terminated with the closing of the project, and the management experience gained may be lost to the implementing agencies. Technical Assistance and Training 38. In order to improve institutional capabilities and to undertake studies required for land and forestry policies formulation, the project has provided a strong technical assistance input. In addition to short-term individual consultancies, five firms have provided medium to long-term technical assistance in the fields of natural resources mapping, resource information, improvement of royalty collection, extension forestry, natural resources management, land policy development, improvement of land records and registration and design of FBD headquarters. Excluding the architectural consultancy, a total of 612 person month (385 expatriate and 277 local) of technical assistance was provided by these firms. In general, most of the consultancies provided by the firms were satisfactory except one firn whose performance was deficient in the areas of forest extension and management. A large training program was also provided under the project, and this has helped implementation in the field. Most of the training focused on improving the operational capabilities of staff in the fields of forestry extension and management and on demarcation and survey of land. The training program consisted primarily of short-term courses and study tours. Procurement 39. Procurement was carried out by the PCU and by the Central Tender Board (CTB). The procedure followed was to deliver the tendering documents prepared by PCU to CTB, after a "no  43 Appendix A objection" was given by IDA. After being reviewed, the tenders were advertised by CTB and bids were directly received and opened by it. The bids were then sent to the PCU for evaluation. Following the reviewing of the evaluation made by the PCU, and having satisfied itself of the result, an approval for awarding the contract was given by CTB to the PCU. This system has worked out quite well in most cases, although some delays were encountered in some cases due to CTB's numerous queries. The only major procurement problem faced by the project was the failure to construct the FBD headquarters as discussed earlier. FBD management submitted the evaluations only in November 1997, about three months after the last deadline set by IDA for forwarding the contract. CTB's approval had been obtained in April 1998, which was too late to commence and finish construction by the project's closing date. Delays in the procurement of GPS units were also experienced due to several misunderstandings resulting from a lack of clarity about IDA procurement procedures for reviewing contract modifications. Borrower and IDA Performance 40. GOT has given strong political support through the PCC and its high level members, and this was an important asset to the project, as it was clearly perceived by staff at all levels as a reflection of firm government commitment to the project in particular, and to the development and management of natural resources in general. However the failure to use the PPF, the delayed replacement of the PC and the inability of GOT to provide adequate counterpart funding, can be considered as deficiencies in performnance. The failure to finalize the necessary preparations and studies for the FBD headquarters was a major shortcoming and a missed opportunity for the Division to improve its working conditions and operational efficiency. IDA's performance has been generally satisfactory, as project design was in tune with the set objectives, and supervision missions were apt at resolving problems faced by the project. However, increased efforts in reminding the concerned institutions about the availability of the PPF would have been warranted, as this could have perhaps increased the chance of its utilization. In retrospect, it can also be said that IDA should have endeavored to assist in bringing about less fragmented organizational arrangements in forestry management. Project Impact 41. The project has positively contributed towards improved the management of natural resources. First of all, it has assisted in a significant manner in the formulation of both the Forest and Land Policies. The Land Policy has led to the enactment of new land legislation. An important achievement of the lands component was also the strengthening of the departments concerned in surveying and in land recording and registration by improving their working capabilities through the provision of equipment, training and the introduction of improved working methods. The survey of village lands will facilitate development planning and land administration at the village level. Within the forestry component, the promotion of private nurseries has enabled the increased production of tree seedlings. The tree seedlings produced and planted under the project are estimated to cover at least 3,000 ha of land, and this will positively contribute to the future supply of construction poles and woodfuel. The stoves are already helping families to reduce their woodfuel consumption due to their improved efficiency. One of the most tangible project results has been the marked increase in royalty collection, and the demonstration that this could be further systematically improved.  Appendix A 44 42. The project has also strengthened the forestry services in the two regions through training and improved mobility, and are now capable of undertaking much larger programs with more confidence, given their experience under the project. The introduction of the concept of joint forestry management and improved management of individual or community woodlands has stimulated people to think about new way of handling the country's forest resources. This is expected to direct future development efforts to participatory and cost effective resource management systems. Finally, the project has reduced the rate of encroachment of the forest reserves in Tabora by establishing boundaries and sending strong signals to the people, that violation of the forest reserves will not be tolerated any more. These signals, which come from the regional and district authorities, have demonstrated their resolve by evicting encroachers from the forest reserves. The phenomena of generalized encroachment on forest reserves was not known when the project was being formulated, but it coincided with its implementation. It is believed by practitioners in Tanzania that encroachment in Tabora would have been far more dramatic, if the project was not being implemented during the years of high incidence of encroachment of the forest reserves. Sustainability 43. Sustainability of the privately operated activities such as nurseries, tending of privately established trees, installed stoves, is practically ensured since many people are already undertaking these activities on their own and without project support. What might disrupt nursery activities could be the lack of availability of inputs such as seeds and polythene tubes, but this could be tackled by working with the private sector to provide these services. Likewise, the sustainability of the land component is not in question because follow-up services will be carried out against payments by those who demand services, such as surveying for obtaining land titles. Sustainability of continuing the services provided by the field forestry services, will, however, depend on the level of support that these will receive from GOT. Based on past experience regarding GOT's difficulties in providing adequate budget, it can be deduced that funds to sustain forestry activities may not flow easily. This may put their sustainability in doubt, although FBD's stand in financing forestry development has improved markedly with the increase in revenue collection. Continuing with the system of collection introduced by the project may be difficult under current administrative arrangements. For instance, it will be difficult to continue payment of relatively high levels of allowances to implement project-related activities. Allowances such as these which provided incentives for staff to collect royalties were not sustainable, at least given existing institutional and administrative arrangements. This consideration appears to cast doubt on the sustainability of the quality and level of services. Therefore, the present search for an alternative organizational and management setup that ensures the sustainable development of forest resources deserve considerate attention by the policy makers. Lessons Learned 44. Many lessons can be learned from the implementation experience of FRMP, and these will be further elaborated in the ICR. In the meantime, the following lessons are drawn from the project:  45 Appendix A * While establishing separate organization and management may be a convenient option for implementing a project within the existing fragmented organizational setup, project ownership by the different agencies becomes difficult. In addition, this approach is not sustainable, since each agency will implement the project under the normal government administrative structure after the project is completed, * Although payment of project allowance are given in order to motivate staff and to compensate them for the increased hardship and responsibility, this has its own drawbacks. Firstly, it has created rivalry between project and non-project staff, and secondly it is not sustainable as the practice will be dropped after the completion of project. * Revenue collection from forest reserves can be increased significantly, as demonstrated by the project, if an improved system is introduced, and staff are motivated and well equipped. * As proven by the project, community participation is cost effective, rapid and sustainable in carrying out village demarcation and surveying. * People are responsive to new forest management systems, such as the ngitiri and Joint Forest Management, when they can perceive a long-term and sustainable benefit from the very forests that are subjected to misuse and encroachment. * Simple technology that is based on the use of inexpensive local material, such as the improved stoves, is enthusiastically adopted by farmers. * According to preliminary observations, it appears that clearfelled dry woodland (characterized throughout the region as miombo woodlands) can grow rapidly, contrary to the general belief these woodlands grow very slowly. This may open new venues for miombo woodland management.  Appendix A 46 Forest Resources Management Project Progress in Meeting Actions Previously Agreed during September 1998 Supervision Mission Task | By whom Target date Remarks Regional and District Forestry I . Revenue collection Regional, district December Because of poor for 4 pilot regions to forestry 31, 1998 (for performance during the first reach TSh 350 officers/FBD CY 98) half of the year, targets were million (excluding reduced from TSh 440 plantation revenues) million. Actual collection totaled TSh 299 million against targets of TSh 350 million. Revenues collected against targets set for the pilot regions are below: Tabora: collected: TSh 160 million (target: TSh 225 million); Singida: collected: TSh 8.4 million (target: TSh 40 million); Mwanza: collected: TSh 103 million (target TSh 60 million); Shinyanga: collected: TSh 27.8 million (target TSh 25 million); 2. Prepare evaluation of Consultant February 15, Completed. Recommended forest revenue 1999 streamlining and collection system improvement of system. 3. Convene workshop FBD/PC by January Not convened on forest revenue 31, 1999 collection system 4. Revenue returns RFOs/PC monthly Summary included in from 4 pilot regions evaluation study. to FBD and IDA  47 Appendix A Task By whom Target date Remarks 5. Prepare evaluation Consultant January 31, Completed. and assessment of 1999 forestry extension system in Mwanza and Tabora. Report to FBD and IDA. 6. Complete forest IRA December Completed management maps 15, 1998 for Tabora Region. Maps to FBD and RFO Tabora 7. Fully-costed and RFO Tabora September Completed. updated plan for 30, 1998 accelerated boundary clearing in Tabora Region, to IDA and FBD. 8. Prepare RFEOs in November Completed documentation Mwanza and 15, 1998 outlining the Tabora approaches undertaken to establish and to register ngitiris together with observations and constraints associated with their registration and use. 9. Undertake evaluation Consultant Consultancy Completed of ngitiris and Joint to commence Forest Management. before Report to FBD/IDA. November 15, 1998 Lands Component  Appendix A 48 Task By whom Target date Remarks 10. Complete village MLHSD October 31, Demarcation of village demarcation in 1998 lands was completed in Kagera, Singida, and Kagera and Singida. Small Arusha regions areas of Arusha and Kilimanjaro regions remain undemarcated, pending settlement of disputed boundaries. 11. Complete surveying MLHSD October 15, Completed. of demarcated 1998 village boundaries in Tabora Region 12. Complete surveying MLHSD March 31, Mwanza has been of demarcated 1999 completed. Mara to be village boundaries in completed in April 1999. Shinyanga, Mara, Shinyanga to be completed and Mwanza regions by May 1999. Project Management 13. Reach agreement MNRT/PC October 15, Decision reached that IDA with GTZ about need 1998 funds would not be for project financing required. GTZ-funded of forest legislation workshop to be held in workshop scheduled April, 1999. for February. 1999. Convey conclusion to IDA 14. Prepare Cabinet FBD December Decision made to defer Paper recommending 15, 1998 preparation of Cabinet institutional option Paper. for establishment of Tanzania National Forest Resources Management Service. Seek Cabinet approval.  49 Appendix A Task By whom Target date I Remarks 15. Prepare and submit FBD/MNRT January 15, IDA not in receipt of copy request to Treasury 1999 of request to Treasury to increase MNRT's share of retained revenues to finance on-going regional field operations. I Copy to IDA 16. Identify members of FBD/MNRT November Members identified new project 15, 1998 preparation working group 17. Prepare proposal FBD/MNRT October 5, Japanese Grant proposal which outlines the 1998 submitted. Approval inputs required for anticipated, shortly. PPF project preparation request submitted; for the next IDA additiona' required project I documentation under preparation. 18. Prepare final project PC/DPC February 1, Completed. overview as an input I1999 into the Implementation Completion Report ____ | 19. Prepare Operational PC/DPC February 1, Completed. Plan which outlines 1999 activities which will continue after the project's close, with key performance indicators and an indication of financing gap   1v.Uasaks r ( " i iGAND tToKi4i IBRD 22789 t / v }