Tanzania - Integrated Roads Project

Document of The World Bank FOR OFFICIAL USE ONLY Report No.: 20332 IMPLEMENTATION COMPLETION REPORT THE UNITED REPUBLIC OF TANZANIA INTEGRATED ROADS PROJECT (CREDIT 2149-TA) April 20, 2000 AFT: Transport 1 Africa Region This document has a restricted distribution and may be used by recipients only in the performance of their official duties. Its contents may not otherwise be disclosed without World Bank authorization.  CURRENCY EQUIVALENTS (As of March 31, 1990) SDR 1.0 = US$1.29; US$ 1.0= SDR 0.78 Tsh 1.0 = US$ 0.005; US$ 1.0 = Tsh 200 (As of June 30,1999) SDR 1.0 = US$1.20; US$ 1.0= SDR 0.83 Tsh 1.0 = US$0.001; US$ 1.0 = Tsh 716 Currency Unit Tanzanian Shilling Fiscal Year July 1 to June 30 ABBREVIATIONS AND ACRONYMS AfDB African Development Bank ATAP Agricultural Transport Assistance Program CODAP Coordination Office for Donor Assisted Projects CRRP Core Rural Roads Program DANIDA Danish International Development Agency EIRR Economic Rate of Return ERP Economic Recovery Program EU European Union FINNIDA Finish International Development Agency GOT Government of Tanzania GTZ Deuche Gesellchaft fur Technishe Zusammenarbeit IDA International Development Association IRP 1, 11 Integrated Roads Project 1, II MCT Ministry of Communications and Works MCWT Ministry of Communications Works and Transport MOW Ministry of Works NORAD Norwegian Agency for Development Co-operation NTC National Transport Agency PEHCOL Plant and Equipment Hire Company Limited RETCO Regional Transport Company Ltd. SAR Staff Appraisal Report TANRROADS National Roads Agency TANZAM Tanzania-Zambia UDR Shirika la Usafari Dar es Salaam Ltd. UNCDF United Nations Capital Development Fund UNDP United Nations Development Program USAID United States Agency for International Development Vice President: Callisto E. Madavo, AFRVP Country Director: James W. Adams, AFMTZ Sector Manager: Yusupha Crookes, AFTT1 Task Manager: Yitzhak Kamhi, AFTTI  Tanzania: Integrated Roads Project Preface Implementation Completion Report THE UNITED REPUBLIC OF TANZANIA INTEGRATED ROADS PROJECT (Credit 2149-TA) IMPLEMENTATION COMPLETION REPORT PREFACE This is the Implementation Completion Report (ICR) for the Integrated Roads Project (Credit - 2194-TA) to The United Republic of Tanzania for which SDR 139.9 million (US$180.4 million equivalent) was approved on May 31,1990 and made effective on March 15,1991. The Credit 2149-TA was partially closed on June 30, 1998, the original closing date. It was fully closed on June 30, 1999 after a year extension from the original closing date of June 30, 1998. Sixteen donors provided parallel financing for the project. Final disbursement of IDA credit was made on December 13, 1999. A total of USD 31,165,978.54 equivalent (SDR 22,695,381.36) was cancelled from the credit on December 21, 1999. Yitzhak Kamhi, Task Team Leader (AFTT1) and Ephrem Asebe (consultant) prepared the ICR. The ICR was reviewed by Messrs. James W. Adams, Country Director for Tanzania, and Yusupha Crookes, Sector Manager, Africa Region. Preparation of this ICR began during the Bank's ICR mission in November 1999. It is based on material in the project file. The Borrower contributed to the preparation of the ICR by stating its views as reflected in the mission's Aide-Memoire (Annex A), and by preparing its own evaluation of the project's execution (Annex C). Annex B also includes the views of some co-financiers.   FOR OFFICLAL USE ONLY CONTENTS PAGE No. PREFACE EVALUATION SUMMARY ........................................................... i-V PART l: PROJECT IMPLEMENTATION ASSESSMENT ........................................................... 1 A. PROJECT OBJECTIVES AND DESCRIPTION ........................... ................................ I B. ACHIEVEMENT OF PROJECT OBJECTIVES ........................................................... 3 C. MAJOR FACTORS AFFECTING THE PROJECT ............................................ ............... 9 D. PROJECT SUSTAINABILITY ........................................................... 11 E. BANK PERFORMANCE ........................................................... 11 F. BORROWER PERFORMANCE ........................................................... 13 G. ASSESSMENT OF OUTCOME ........................................................... 14 H. FUTURE OPERATIONS ........................................................... 14 I. KEY LESSONS LEARNED . - ............................................................ 15 PART II: STATISTICAL TABLES ........................................................... 17 TABLE 1: SUMMARYOFASSESSMENTS .................... ....................................... 17 TABLE 2: RELATED BANK LOANSICREDITS .............................. ............................. 18 TABLE 3: PROJECT TIMETABLE ........................................................... 19 TABLE 4: LOAN/CREDIT DISBURSEMENT: CUMULATIVE ESTIMATE AND ACTUAL .......... .......... 19 TABLE 5A: KEY INDICATORS FOR PROJECT IMPLEMENTATION ................................ ................ 20 TRAINING UNDER IDA CREDIT TABLE 5B: KEY INDICATORS FOR PROJECT IMPLEMENTATION ................................................ 21 STATUS OF TRAINED STAFF UNDER IDA CREDIT TABLE 5C: KEY INDICATORS FOR PROJECT IMPLEMENTATION: PEHCOL ................................ 21 TABLE 5D: KEY INDICATORS FOR PROJECT IMPLEMENTATION ................................................ 21 ROAD FUND EXPENDITURE ON ROADS TABLE 6A: KEY INDICATORS FOR PROJECT OPERATION ....................................................... 22 ROADS CONDITION & TARGET ACHIEVEMENTS TABLE 6B: KEY INDICATORS FOR PROJECT OPERATION ....................................................... 23 TRUNK ROAD OPERATIONAL PLAN AND BUDGET TABLE 6C: KEY INDICATORS FOR PROJECT OPERATION ....................................................... 23 RURAL ROADS OPERATIONAL BUDGET AND PLAN TABLE 7: STUDIES INCLUDED IN PROJECT: IDA COMPONENTS ............................. ............... 24 TABLE 8A: PROJECT COSTS (US$ MILLION) ...................... ................................. 24 TABLE 8B: PROJECT FINANCING (US$ MILLION) .................................. ..................... 25 TABLE 9: ECONOMIC COSTS AND BENEFITS ....................................................... 25 TABLE 10: STATUS OF LEGAL COVENANTS ....................................................... 26 TABLE 11: COMPLIANCE WITH OPERATIONAL MANUAL STATEMENTS ................... ................... 28 TABLE 12: BANK RESOURCES: STAFF INPUTS ......................... .............................. 28 TABLE 13: BANK RESOURCES: MISSIONS ................. ...................................... 29 This document (Las a resticted distribution and may be used by recipiens only in the pearfogrmnce of their official duties. Its contents may not otherwise be disclosed without World Bank authorizadion.  Tanzania: Integrated Roads Project Page 2 Implementation Completion Report ANNEX A: SUPPLEMENTARY STATISTICS ............................................. 34 TABLE 1: REALLOCATION OF IDA CREDIT .............. ......................................... 34 TABLE 2: STATUS OF IDA FINANCED COMPONENT AS OF PROJECT CLOSING ............. ................... 35 TABLE 3: PROJECT COSTS OF IDA AND OTHER DONORS ...................... ....................... 37 ANNEX B: ICR MISSION'S AIDE MEMOIRE ............................................. 38 ANNEX C: BORROWER'S CONTRIBUTION AND COMMENTS TO ICR ............................................. 43 MAPS: IBRD No. 22067  Tanzania: Integrated Roads Project Page i Implementation Completion Report Evaluation Summary EVALUATION SUMMARY THE UNITED REPUBLIC OF TANZANIA Integrated Roads Project (Credit 2149-TA) Introduction i. Background. With an estimated project cost of US$ 871 million, the Integrated Roads Project I (IRP I) was the largest single road program in the sector in Sub-Sahara Africa. It was a bold experiment in the sector and represented major breaks in the ways road sector projects were designed in the Bank. It pooled resources with parallel financing from IDA, and 15 other donor agencies:-AfDF, UNDP, EEC, DANIDA, Republic of Germany, Italy, FINNDA, ODA (UK), NORAD, USAID, The Netherlands, Ireland, Switzerland, Saudi Fund and Kuwait Funds. ii. The project was launched in 1991 after three years of consultation, to overcome the impact of the deteriorating road network in a context of the economic recovery program the country was then undertaking. During the period, IDA was instrumental in assisting the Government of Tanzania (GOT) to develop IRP I and had assumed responsibility of coordinating donors' assistance. This Implementation Completion Report (ICR) of Integrated Road Project I attempts to sum up the major lessons leamed from implementing the Project. The scope of this ICR is limited to an assessment of IDA financed component of IRP I. Reference to the total project is made where it is essential to the overall understanding of the IDA component. Bank Project iii. Project Objective. As outlined in the Staff Appraisal Report No. 8367-TA, the main objectives of IRP were to: (i) develop MCWT's institutional capacity to manage the networks'; and (ii) restore Tanzania's trunk and regional roads networks, which were obstacles to the sustainability of the economic recovery program. iv. Project Components. The project was to provide financial assistance for (i) institutional support to strengthen road management capacity of the MCWT; (ii) rehabilitation and improvement program of about 4,600 km of trunk roads and 3000 km of regional roads in agriculturally productive areas including related structures and bridges; (iii) road maintenance support to MCWT including establishment of privately operated plant pools, and development of local constructing capacity to undertake road works; and (iv) management assistance to Air Tanzania Corporation and the National Transport Corporation. v. Project Covenants. The critical agreements reached at negotiation were to (i) undertake the main institutional and policy changes prior to project effectiveness, including (a) changes in the procurement regulations, (b) appropriating the agreed road budget for the first year of project implementation, and (c) reorganize the road administration; (ii) appropriating GOT agreed annual road budget as a condition of IDA funding; (iii) complete engineering designs and technical documentation for the first two years of the program prior to board presentation. vi. Project Evaluation. The Integrated Roads Project objectives were very important to the people and Government of Tanzania, and were in line with IDA's sector priority for the country. The design of the project was conceptually clear. The institutional objective was to address the long-term sustainability of the project, while the physical objective was focused on the urgent task of maintenance and rehabilitating the road network to enhance the economic recovery program. Major challenges facing project design were defining an appropriate program size and finding the appropriate tradeoffs between the two major objectives. A key strategy adopted, to assure sustainable of the outcomes of the project, was to address in parallel several development objectives, including transformation of the ministry from blue collar to white collar organization, 1 That is, transform from blue collar to white collar organization.  Tanzania: Integrated Roads Project Page ii Implementation Completion Report Evaluation Summary development of a private construction industry, decentralization of the maintenance and rural roads administration, training of manpower, establishment of road fund. But these added to the complexity of the program and the operational strategy had to deal with tradeoffs. Implementation vii. Achievement of the Program Objectives. There is no question that there have been improvement in terms of reductions of transport cost and travel time mainly as result of improvement of certain major links of the country. According to govemment statistics, in 1999 the actual proportion in length of the trunk road network in good, fair and poor conditions were 30%, 39% and 30% of the total trunk roads respectively. This favorably compares to the SAR estimates for 1990/91, the start of the project, which was 26%, 26% and 47% for good, fair and poor roads of the trunk road network length respectively. Between 1990/91 and 1998199, the trunk road network had also increased from 9,629 km to 10,203 km. However, the status of roads in good condition in 1999 which stood at 30% for the trunk roads was significantly below when compared to the appraisal targets of 60% by mid-1996 and 80% by mid-2000. The current condition of rural roads is not know as El Nino has damaged the road constructed. About 2000 km of rural roads were known to have been rehabilitated by the donor community under the IRP I. viii. With respect to the physical achievement under IDA funding, initially a total of 910 km of gravel roads were contracted out. About 683 km of gravel roads were constructed with cost overuns ranging between 65% and 244%. Of the 683 km of constructed roads some 300 km was total loss and some 266 was deteriorated within the first 13 months of construction. The Tanga- Horohoro/Marangu-Tarakea road (118 km) was substantially completed and is now in reasonable good condition; its EIRR is 29%. Regretfully, for the uncompleted and washed out components, the investment value is practically lost, making re-estimation of ERR not worthwhile. Mwanza-Nzega Road (54 km) is yet to be fully completed and no EIRR is estimated. Its designs were not updated before signature of contracts. Field survey was poor and it suffered major claims related to payments delays of GOT portion. The Kayaka bridge and approaches were completed. Implementation experienced many failures of structures, approaches, temporary approaches and bridges. And substantial additional works were carried out after commencement. Six of the seven packages in Emergency Works for Dar es Salaam Roads were implemented. Poor surveys and designs, frequent changes in scope of works resulted in poor quality of works and significant cost increases. Some of the roads also disintegrated soon after completion. Under IDA funding periodic maintenance and emergency works were carried out. The project also funded a portion of TANZAM Highways sections 2, 4 & 5, which were rehabilitated under the Six Highway Rehabilitation Project (Cr,1688-TA). ix. On the institutional front, modest progress has been made towards achieving some of the institutional development objectives. A modest private construction industry has been created. Road maintenance administration has been decentralized in 20 regions. The Road Fund established under the project now can potentially provide some 70% of the required road maintenance funding. Road maintenance expenditure made from the Road Fund has increased from US$7.24 million in 1991/92 to US$22.48 million in 1998/99. The National Road Agency was established in July 1999 and its chief executive took position in February 2000. Significant number of high level manpower has been trained at home and abroad. WVith respect to liberalization, the Plant and Equipment Hire Company is now in the process of privatization. Nevertheless, it is generally agreed by both MOW and donors, that institutional issues, particularly contract administration and project management are still major problems in efficiently managing the road network. Overall, the project achievement is assessed as unsatisfactory. x. Cost and Implementation Time Schedule. At completion, the total funds expended on IRP I was estimated at US$868.0 million including GOT contribution estimated at US$60.0 million. US$166.7 million equivalent (SDR117.2 million) of IDA funds were disbursed and US$ 12.17 million of the original estimates was cancelled along with US$ 19.0 million equivalent from exchange rate savings. However, In spite of its limited impact the cost of institutional support was more than double  Tanzania: Integrated Roads Project Page iii Implementation Completion Report Evaluation Summary the appraisal estimates. This applied for both IDA and for the other donors. After allowing for inflation, cost escalation for civil works ranged from 40% to over 125%. xi. Main Factors Affecting Implementation. The factors which were partly outside the government control and which affected the implementation of the project included: difficult environment for implementation; inadequate design of the project; problems in donor coordination and poor performance of consultants and contractors. The factors within the government's control affecting implementation were: lack of good governance including instances of corrupt practices; failure to apply lessons learned from preceding project implementation; lack of proactive response; frequent reorganization of the implementing agencies; inadequate counterpart funding; and poor staff incentives. Factors within the control of the implementing agencies included: substantive changes in the scope of works without proper technical documentation and authorization by funding agencies; prolonged procurement processing; deficient contract administration; and poor project and financial management. Results and Performance xii. Project Sustainability. Overall, the sustainability of the project is assessed as uncertain. It appears that much longer time than originally planned is needed to assure sustainability. Sustainability of the progress made with respect to institutional objective depends on the prevailing environment, government commitment to maintenance-first policy, proactive leadership, the incentive system to skilled manpower and on follow-up measures to consolidate the progress achieved to date - road fund, road agency, etc. The sustainability of the commercialized activities under the liberalization regime depends on the entrepreneurial abilities of the new contractors and their financial capacities. The sustainability of the maintenance system introduced depends on the performance of the economy to generate road fund revenue and eradication of corrupt practices. xiii. Bank Performance. Bank performance in identification and preparation assistance is assessed as satisfactory. IDA staff rightly identified that transport is the major problem facing the Tanzanian economic recovery during its transition to market-oriented economy. During preparation phase, IDA staff provided critical assistance to GOT in coordinating donor assistance and packaging the program. Appraisal was insufficient with respect to considering the appropriate design options for Tanzania, particularly on the scope of institutional reforms to be undertaken. Second, it was known from earlier IDA financed road sector projects that there were major problems in performing basic project activities. Appraisal, thus, did not fully address the major flaws experienced under the Sixth Highway Credit; particularly in respect to contract administration and project and financial management. Those errors were, therefore, repeated on a larger scale. Third, the risks due to lack of incentives to staff to steer the project implementation was identified but an adequate sustainable solution in the context of the civil services was not found. xiv. The overall assessment of supervision must be seen in the context of unduly ambitious and optimistic project design; early successes following initial contract signatures and initial improvements in road conditions; the unclear and conflicting role of the Principal Transport Engineer (PTE) who was advising the government and the Bank management and staff at the same time, and poor governance conditions in the country. Therefore, the implementation of the program basically failed, particularly in the gravel road component, notwithstanding of the efforts of the supervision. Hence, the Bank's supervision performance is rated as marginally unsatisfactory. xv. Borrower Performance. Borrower preparation is assessed as satisfactory. The government's bold initiative in its Transport Sector Policy Paper and its Transport Recovery Program and its ability to bring some 16 donors to agree in support of its program over a period of three years of consultation, was a major undertaking. It showed leadership and built on a reservoir of good will that the government had among the donor community at the time. Coming to closure was a major accomplishment. The govemment, however, was facing serious difficulties in translating its ambitious program on schedule and within planned budget. It has not been able to make best use of the resources at its disposal. Audit reports of project accounts were continuously and seriously  Tanzania: Integrated Roads Project Page iv Implementation Completion Report Evaluation Summary qualified. GOT's responses were often not timely and sometimes had required the prodding of the donor community, as in the case of the Road Fund and Road Agency. As evidenced in the 1996 Report on Presidential Commission Inquiry Against Corrupt Practices, the operating environment in the implementing agencies was not conducive to good project management practices. In the area of procurement, project management and contract administration, a decade after the launching of IRP I, despite substantial expenditure of resources2, including manpower training, the evidence is that implementing agencies are only marginally better than they were at the start of the project. xvi. Project Outcome. Overall, the outcome of the IDA financed component is assessed as unsatisfactory. However, as a result of the project, there is today a more organized sector program. The government has learned valuable lessons from its implementation experience. A private sector led road construction sector industry has emerged. Road maintenance is now carried by private contractors who have largely come into existence with the support of the IRP I. Despite the temporary relapse in the administration of the Road Fund in 1995/96, the road fund now contributes a significant share of the cost of maintenance. The road fund is administered by the Road Fund Board supported by the Road Fund Act. The TANROADS is finally going to start fully operating by July 2003. However, impact of transport cost reduction on the macroeconomic situation was not as high as expected at the start of this ambitious program. Also, the basics of project and financial management have yet to be mastered. Key Lessons Learned and Future Actions. xvii. This Integrated Roads Project has been the largest and most ambitious project in the sector. As such it has been the subject of multiple reviews including a recent Bank Quality Assurance Group review. This ICR identifies a number of key lessons for future project design and implementation. Complexity of the Project and Country Absorptive Capacity. Project composition and implementation schedule should match the absorptive capacity of the implementing agency. In design of such large investment, involving several donors, the approach should aim at: (i) a phased investment with a better defined triggers; and (ii) long-term donors support. * Project Design. While the concept of sector investment project on which IRP I is based appears workable, its application to a specific country requires making tradeoffs between major development objectives based on detailed analysis of major risks in achieving those objectives. In particular, in the civil works, the absence of detailed engineering for all major roads immediately prior to start of construction, made variation orders almost inevitable, leading to delays and cost increases. In institutional capacity building too, the many demands for reforms, taxed the capacity of the govemment to deliver on budget and on schedule. Therefore, in projects of IRP I type, it is important to introduce "checks and balances" with early warning signals to be able to adapt the program when required. * Corrupt practices. The poor implementation performance of IRP I testifies that the potential for corrupt practices can significantly impact project implementation performance. In this case, roads that would have been constructed under the project to good standard remained uncompleted partly because of corrupt practices and the absence of systems and procedures that could check against their occurrence. The Government should institute measures to safeguard against such practices. Physical progress should be verified against design specifications. Stringent measures should be taken under situation where audit reports are constantly delayed and qualified in order to fight potential corrupt practices from blossoming. * Learning from Past Mistakes. The lessons of experience of past project implementation were to safeguard against future mistakes both for IDA and the borrower. As a matter of fact, major 2 A total of us $105 million was expended by all donors on capacity building as estimated by Carl Bro International a/s, Comprehensive Review of the IRP, Third Interim Report, Revised June 1997.  Tanzania: Integrated Roads Project Page v Implementation Completion Report Evaluation Summary flaws experienced in the Sixth Highway Project Credit were also experienced under IRP 1. For example, appropriate technical documentation and provision of adequate resources should be fully completed before commencement of works. In short, more attention needs to be paid to past lessons of experience during project design and implementation. * Composition of Project Team. Supervision team should have professional mix with key persons being sufficiently experienced in the major fields of a project activities such as engineering, institutional reforms, etc. In addition, there should not be frequent changes of supervision management and teams, in order to ensure continuity and consistency of efforts in achieving quality and quantity of the planned targets. * Technical Documentation. Technical documentation necessary for implementation of civil works should be updated before actual commencement of construction works. * Donor Coordination. The spirit of cooperation of the donor community in support of the Integrated Roads Project I had no parallel. It was the first time so many donors have committed themselves to such a large common program. Unfortunately, in IRP 1, it has not been easy to maintain the interests of all donors' cooperation during implementing phase as each donor focus on execution of its own mini program under the parallel financing arrangement. In future, in large investment supported by several donors, a professional coordination managed by the government should be established, with agreed terms of references. * Technical and Financial Audits. The Government should among other things, carry performance technical and financial audits of the works carried out. Such audits should, besides frequently revisiting the overall status of implementation, also include verification of the physical and financial targets. * Initial Successes in Long-term Program. Usually during the initial phases of project cycle of large programs, there is high optimism of the team involved and there is a tendency to underestimate the operation environment of the specific country for which the project is designed, even when there are potentially implementation problems. Initial successes can mask deficiencies related to preparation, selection of contracts and judgements of capacities. It is therefore important that project implementor should not be taken in by the initial euphoria about early successes, and should continue to ensure that measurable output targets are met. The project should have well defined qualitative and quantitative indicators, which should be adequately designed so that any deviation from initial targets can be detected and corrected in real time. xvii. Key Elements of Future Planned Operation. To lay the foundation for excellence in institution building, there is need to assure that the modest capacity achieved to date under the project, especially the trained manpower and the creation of a road agency, is sustained. Assuring adequate resources for road maintenance will remain essentially an ongoing issue. Given past performance of the economy, efficient and effective use of the available road fund will be another critical area of focus.   Tanzania: Integrated Roads Project Page 1 Implementation Completion Report Part 1: Project Implementation Assessment PART 1: PROJECT IMPLEMENTATION ASSESSMENT A. PROJECT OBJECTIVES AND DESCRIPTION Background 1. With an estimated project cost of US$ 871 million, the Integrated Roads Project I (IRP 1) I was the largest single road sector program in the sector in Sub-Sahara Africa. It was a bold experiment in the sector and represented major breaks in ways road sector projects were designed in the Bank. It pooled resources on the basis of parallel financing from IDA, and 15 other donor agencies: -AfDF, UNDP, EEC, DANIDA, Republic of Germany, Italy, FINNDA, ODA (UK), NORAD, USAID, The Netherlands, Ireland, Switzerland, Saudi Fund and Kuwait Funds. 2. The International Development Association (IDA) had been instrumental in assisting the Government of Tanzania (GOT) develop IRP I and had assumed responsibility of coordinating donors' assistance. Prior to the IRP I, the Bank group has extended credits and loans to help finance six highway projects, one trucking project, two railway projects and five port projects. IRP I was designed in support of GOT's economic recovery program. During the 1970s and early 1980s, declining funding for road maintenance resulted in the deteriorating road condition in Tanzania. However, when the Government of Tanzania undertook an Economy Recovery Program starting 1986 designed to liberalize and adjust the economy, it became apparent that the functioning of the transport sector was critical to sustain the momentum of Tanzania's economic recovery. In December 1987, GOT called a Transport Sector Conference. At the Conference, GOT presented a draft National Transport Policy along with a conference document entitled "Programme for Transport Sector Recover'. The GOT initiative was based on a study to improve conditions of rural roads to essential agricultural production area. 3. Following three years of coordination and consultation on basic policy, institutional changes and rehabilitation requirements of the sector needed to adequately respond to the economy, the donor community agreed to support the govemment initiative. GOT committed itself to undertake strategic changes regarding the organization of road maintenance and also pledged to make major shifts in its public expenditure, committing itself to provide 20% of the development budget expenditure for transport infrastructure, and to finance road maintenance from its recurrent budget. Finally, agreement was reached between GOT, IDA and 15 other donor agencies - AfDF, UNDP, EEC, DANIDA, Republic of Germany, Italy, FINNDA, ODA (UK), NORAD, USAID, The Netherlands, Ireland, Switzerland, Saudi and Kuwait Funds - to finance the IRP I. A total of US$756.8 was pledged or secured from the donor community at the launching of IRP for which IDA's contribution was US$180.4 million equivalent. GOT was to cover US$80.5 million and only US$33.8 million or 4% of the total project cost was expected to be filled through future donor coordination conferences. The project would be parallel financed following different implementing procedures. Bilateral donors were to use their own staff to undertake implementation and supervision of components financed by them. Others like IDA considered project implementation and field supervision to be primarily the tasks of GOT and it's implementing agencies. 4. Initially, the ICR was to provide a comprehensive assessment of the IRP I including the outcomes and results of the works financed by all the donors. During the ICR mission, it became apparent that relevant data was not readily available as the government had not kept adequate records and the donor community has almost closed their activities related to the project. It became apparent that undertaking such a task was beyond the budget allocated for the ICR. Thus, the focus of this ICR is mainly on IDA financed component of IRP I. 1 IRP was essentially a program covering not only the road infrastructure but also covering the development of road and air transport sectors  Tanzania: Integrated Roads Project Page 2 Implementation Completion Report Part 1: Project Implementation Assessment 5. Project Objective. As outlined in the Staff Appraisal Report No. 8367-TA, the main objectives of IRP were to: (i) develop MCWs institutional capacity to manage the networks; and (ii) restore Tanzanian's trunk and regional roads networks, which have become an obstacle to the sustainability of the economic recovery program. The institutional objective was to focus on: assisting the government in developing strong management and technical capacity to maintain the trunk and regional road networks, and transformation MCWT from a construction-oriented ministry to an administrative/ contract management-oriented ministry. The physical objective was to increase the trunk roads in good condition from 15% to 60%; and to selectively rehabilitate the regional road network initially in eleven of the most agriculturally productive regions thereby increasing the road network in good condition from 10% to 50%; and institute regular maintenance on 80% of the trunk and 60% of the regional road networks. 6. Project Components. The project was to provide financial assistance for: (i) institutional support to strengthen road management capacity of the MCWT; (ii) rehabilitation and improvement program of about 4,600 km of trunk roads; 3000 km of regional roads in agriculturally productive areas, and other related structures and bridges; (iii) road maintenance support to MCWT including establishment of privately operated plant pools and development of local construction capacity to undertake road works; and (iv) management assistance to Air Tanzania Corporation and the National Transport Corporation. 7. Changes in Project Scope. Three amendments to the credit agreement have been undertaken. The first was on August 21, 1991, to include completion of the rehabilitation of three sections of TANZAM highway. The second was on June 19, 1992, to reflect the split in the Ministry of Communications and Works, and establishment of the Plant and Equipment Hire Company. The third amendment October 4, 1992 was to include the establishment of computer course offerings for highway technicians at Dar es Salaam Technical College. In 1995, attempts were made to restructure the project, following failures of gravel roads. However, as CODAP was not prepared to produce documentation for making appropriate decisions, nearly two years would pass before a basis for assessing the status of the utilization of the IDA credits was established. In 1997 agreements were finally reached based on the findings of an independent consultant not to further restructure the project. Instead it was agreed to: (i) give priority to the ongoing contracts; (ii) update the scope of works and quantities before signing any contracts; and (iii) re-orient the training programs to focus more on building capacity particularly contract administration, procurement project management. 8. Project Covenants. The critical agreements reached at negotiation were to: (i) undertake the main institutional and policy changes prior to project effectiveness, including (a) changes in the procurement regulations, (b) appropriating the agreed road budget for the first year of project implementation, and (c) reorganize the road administration; (ii) GOT appropriates agreed annual road budget as a condition of IDA funding; (iii) complete engineering design works for the first two years of the program prior to board presentation; (iv) develop closer coordination among the government, donors and IDA; (v) station in Tanzania an experienced highway engineer from the Bank to assist the government in coordinating implementation of the project. Assessment of Project Objectives. 9. The Integrated Roads Project objectives were very important to the people and Government of Tanzania as the transport sector was the major constraint to the success of the country's Economic Recovery Program. It was also IDA's priority to assist the Government in further development and implementation of specific policy and institutional reforms in the sector, in the context of the continuing ERP and as supported by Bank policy. The project objectives were thus in line with IDA's overall and sector policy for Tanzania. 10. The design of the project was conceptually clear. The institutional objective was to address the long-term sustainability of the project, while the physical objective was focused on the urgent task of maintenance and rehabilitating the road network to enhance the economic recovery program.  Tanzania: Integrated Roads Project Page 3 Implementation Completion Report Part 1: Project Implementation Assessment A major challenge facing project design was defining an appropriate program size and finding the appropriate tradeoffs between these two major objectives. A key strategy adopted, to assure sustainable outcomes of the project, was to address in parallel several development objectives, including transformation of the ministry from blue collar to white collar organization, development of a private construction industry, decentralization of the maintenance and rural roads administration, training of manpower, establishment of road fund. But this added to the complexity of the program and the operational strategy had to deal with tradeoffs. For example, the project assumed proper designs and technical documentation would be available prior to commencement of works, but on the grounds of emergency needs, poor and insufficient designs, and weak procurement documentation were actually used. 11. The project was indeed complex given the borrowers implementation capacity - with many development objectives, coordination of several implementing agencies and some sixteen donors, five types of disbursements and an environment of frequent organization changes. Building consensus among the donors was difficult because of different strategies and different aspirations across the donor community. Given the prevailing environment, the motivation of the implementing staff, the readiness of the country for such a huge program, the design of project was also too optimistic while the risks in this respect not properly addressed. The country was already having difficulties in effectively implementing the preceding Sixth Highway Project, a substantially smaller project. B. ACHIEVEMENT OF PROJECT OBJECTIVES 12. Overall Assessment. . There is no question that there have been improvement in terms of reductions of transport cost and travel time mainly as result of improvement of certain major links of the country. According to government statistics, in 1999 the actual proportion in length of the trunk road network in good, fair and poor conditions were 30%, 39% and 30% of the total trunk roads respectively. This favorably compares to the SAR estimates for 1990/91, the start of the project, which was 26%, 26% and 47% for good, fair and poor roads of the trunk road network length respectively. Between 1990/91 and 1998/99, the trunk road network had also increased from 9,629 km to 10,203 km. However, the status of roads in good condition in 1999 which stood at 30% for the trunk roads was significantly below when compared to the appraisal targets of 60% by mid-1996 and 80% by mid-2000. The current condition of rural roads is not know as El Nino has damaged the road constructed. About 2000 km of rural roads were known to have been rehabilitated by the donor community under the IRP I. 13. With respect to the physical achievement under IDA finding, initially a total of 910 km of gravel roads were contracted out. About 683 km of gravel roads were constructed with cost overuns ranging between 65% and 244%. Of the 683 constructed road some 300 km was total loss and some 266 was deteriorated within the first 13 months of construction. The Tanga-Horohoro/Marangu- Tarakea road (118 km) was substantially completed and is now in reasonable good condition; its EIRR is 29%. Regretfully, for these uncompleted and washed out components, the investment value is practically lost, making re-estimation of ERR not worthwhile. Mwanza-Nzega Road (54 km) is yet to be fully completed and no EIRR is estimated.. Its designs were not updated before signature of contracts. Field survey was poor and it suffered major claims related to payments delays of GOT portion. The Kayaka bridge and approaches were completed. Implementation experienced many failures of structures, approaches, temporary approaches and bridges. And substantial additional works were carried out after commencement. Six of the seven packages in Emergency Works for Dar es Salaam Roads were implemented. Poor surveys and designs, frequent changes in scope of works resulted in poor quality of works and significant cost increases. Some of the roads also disintegrated soon after completion. Under IDA funding periodic maintenance and emergency works were carried out. The project also funded a portion of TANZAM Highways sections 2, 4 & 5, which were rehabilitated under the Six Highway Rehabilitation Project (Cr,1688-TA).  Tanzania: Integrated Roads Project Page 4 Implementation Completion Report Part 1: Project Implementation Assessment Status of Major IDA funded Civil Works Road Section Expenditure Results (US$) 1. Lusahunga-Usagara (266 kmi) 6,069,740 Works completed, but deteriorated within the first 13 months after rehabilitation. Cost increases of 65% 2.Tundoma/Lichet 8,702,892 Some 50% of 226 km were done. scope of work (Sumbawanga) (226) substantially changed. Work abandoned. Almost total loss of investment. Cost increases of 244% 3. Shelui-SingidalBabati/Bereku 8,178,078 Some 60 % of civil works was completed. (310km) Scope work substantially changed. Work abandoned. Almost total loss of investment. Cost increases 4. Tanga/Horohoro & Marangu 9,099,438 Gravel road completed excluding bridges. Tarakea (118km) Sub-total 32,050,148 6. Rehab. - Tanzam 2 3,999,922 Substantially completed under Cr.1688: Six 7. Rehab. - Tanzam 4 5,368,290 Highway Rehabilitation Project; but payment; 8. Rehab. - Tanzam 5 96,319 represent payments made under IRP i. Sub-total 9,464,531 9. Kayaka Bridges and 7,741,022 Implementation experienced many failures of approaches structures, approaches, and temporary bridge. Substantial additional works included after commencement. 10. Emergency Works for Dar es 35,343,683 The program had seven contract packages; but Salaam Roads only six packages were implemented. Poor surveys and designs and frequent changes in scope of works resulted in poor quality and cost increases. Some of the roads disintegrated. Total 84,599,384 14. On the institutional front, modest progress has been made towards achieving some of the institutional development objectives. A modest private construction industry has been created. Road maintenance administration has been decentralized in 20 regions. The Road Fund established under the project now can potentially provide some 70% of the required road maintenance funding. Road maintenance expenditure made from the Road Fund has increased from US$7.24 million in 1991/92 to US$22.48 million in 1998/99. The National Road Agency expected to be established in July 2000, and fully operational by June 2002. Significant number of high level manpower has been trained at home and abroad. With respect to liberalization, the Plant and Equipment Hire Company is now in the process of privatization. Nevertheless, it is generally agreed by both MOW and donors, that institutional issues, particularly contract administration and project management are still major  Tanzania: Integrated Roads Project Page 5 Implementation Completion Report Part 1: Project Implementation Assessment problems in efficiently managing the road network. Overall, the project achievement is assessed as unsatisfactory. 15. At completion, the total fund expended on IRP I was estimated at US$868.0 million including GOT contribution estimated at US$60.0 million. US$166.7 million equivalent (SDR1 17.2 million) of IDA fund was disbursed and US$ 12.17 million of the original estimates was cancelled along with US$ 19.0 million equivalent from exchange rate savings. However, the costs of institutional support were more than double the appraisal estimates (para.). This applied for both IDA and for the other donors. Cost escalation for civil works ranged from 40% to over 125%. Delays in procurement totaling 30 to 36 months have been experienced, although major improvement has been observed lately. I. Institutional Reforms. 16. Reorganization. Decentralized decision making to the regions for planning and implementation of road programs were introduced and staff was assigned to all 20 regions. Following the Mramba Commission on restructuring of the Tanzanian Government, all transport modes in different ministries were brought under one oversight agency, the Ministry of Works, Communications and Transport (MWCT). REOs reverted-to MWCT from the Regional Authorities in July in 1990, incorporating the MOW trunk roads maintenance organization in the 20 regions. Subsequently, following three years of discussions to create a more streamlined organization for overall management of the transport sectors, the ministry was further split into MOW and MCT. MOW, Department of Roads (which was split from Buildings in November 1991) was to provide better focus on road infrastructure needs. In line with the Parastatal Sector Reform Commission, one objective of the later reorganization was directed at preparing for private partnership or ownership of the transport functions (road, air, and maritime transport) and creation of public work parastatal for road equipment (PEHOL). The establishment of an agency specifically related to road construction, TANROADS, remained an important agenda of reform supported by donors throughout the project life, but the government was reluctant to pursue this in the initial years of the project. It was only in the last two years that the GOT changed its stance and today TANROADS is being established. 17 Training. Significant manpower has been trained under IRP I. In MOW, under IDA funding alone, a total of 63 staff obtained Master of Science Degrees: 54 in engineering; 3 in construction management; 4 in human resources and 2 in transport economics. However, of the 23 foreign educated professionals, only 11 are currently with the ministry. Out of the remaining 40 who graduated from the University of Dar es Salaam, all are with the ministry except the two who died. Regarding the short-term trainees, 14 out of 32 of the foreign trained, and 336 of 379 locally trained are still with the ministry. Also, in the National Construction Council, a total of 15 staff had received advanced training and 12 of those are currently with NCC. Similarly, other donors have provided training. While this is a significant capacity and it should have translated into effective performance, the consensus of donors and the ministry itself is that the ministry has yet to significantly overcome its capacity problems. One reason, in the words of one of the staff who received post-graduate degree under IRP I financing, "with wages which hardly cover two weeks living expense, you are expected to devote yourself to your work for a full month". Although a staggering US$105 million3 (from all donors) was spent on the capacity building effort over a nine year-period, the absence of an adequate incentive structure undermined these efforts. 18. Contract Management. MOF undertook a study and implementation of a revised procurement and supplies management after much delay but its findings and recommendations were not made effective and had little effect on the implementation of IRP I. However, improvements are 3 A total of us $105 million was expended by all donors on capacity building as estimated by Carl Bro International a/s, Comprehensive Review of the IRP, Third Interim Report, Revised June 1997.  Tanzania: Integrated Roads Project Page 6 Implementation Completion Report Part 1: Project Implementation Assessment evident since 1998. Currently, procurement time on civil works have tended to be within the tender validity period, while processing consultancy service contracts still exceed the tender validity period, partly due to two stage evaluation and each stage undergoing internal clearances before obtaining "no objection" from IDA. 19. Technical Capacity for Road Maintenance. Road Maintenance Management System was introduced. Socio-economic criteria were used for selecting priority road improvements. Labor- based contract training was undertaken; and a management information system was introduced. 20. Road Fund. The road fund was established in July 1992. The Sub-Sahara Africa Road Maintenance Initiative (RMI) and the donor community supported the program. The revenue is derived from fuel toll levy. In principle, significantly more resources should be available to pay for road maintenance with road users shouldering an increasing share of the cost of maintenance. Currently, 70% of the total collected revenue from fuel levy is allocated to road maintenance. The amount available in the future depends on the performance of the economy and on the levy rate applied. 21. In the first three years of its establishment, the road fund was covering significant share of the resource requirement for the maintainable road network. Over time, delays in transfer of road funds to the Ministry of Works became frequent. This affected the effective execution of maintenance activities. There was also misapplication of the road fund. In 1995/96, there was a crisis with the road fund. Following short falls in the government revenue, GOT diverted the road fund to the general account. The government saw the issue as a balancing act between important national priorities due to tight budget to meet resource needs for counterpart funding; rising debt payments, both external and internal; food deficit due to drought etc. Among donors, the 1995/96 'road fund crisis" rekindled the donors' concern that maintenance was not being given adequate priority. 22. Audit reports on the road funds were highly qualified. For example, the audit reports of the 1994/95, 1995/96, and 1996/97 showed that the fund was used for ineligible expenditure which included: terminal benefits, night out allowances to casual laborers, subsistence allowances, severance allowances and transportation of personal effects. Contractors with poor performance continue to be awarded contracts paid from road fund. These led several donors to question the transparency and accountability of the road fund. In the last two years. In response the GOT has undertaken confidence-building measures, including establishing Road Fund Board, and appointing a chief executive for TANROADS scheduled to commence work in July 2000. II. Commercialization. 23. Construction Industry. In line with its commitment, GOT adopted in 1992 the National Construction Industry Development Strategy. This led to the establishment of a private sector led construction industry. Local contractors and consultants received training. By the end of 1996, the number of private civil works contractors had significantly increased from about 43 in 1989 to 610, of which some 200 were in road sector. Today, there are more than 50 contractors with capabilities to undertake basic unpaved road rehabilitation and periodic maintenance works. Maintenance works are now carried primarily using local contractors instead of force account. In the 1995/96 "road fund crisis" the contractors played active role by lobbying for the timeiy disbursement of the road fund. Yet despite significant increase in the 1990s in the number of qualified domestic contractors, their share of the market has remained low, around 10% of total value of contracts. This is the major concern to the nascent private contracting industry in Tanzania. 24. Plan and Equipment Hire Co. Ltd (PEHCOL). Under the credit GOT was committed to manage all its road equipment along commercial lines, with users paying full hire rates. However, because of lack of reliability of the PEHCOL's equipment, commercialization objective was not as successful in the regions where PEHCOL had workshops as elsewhere. Much effort and resources were expended to bring the equipment in good service condition. A consultant was hired to advise on rehabilitation strategy. A total of US$ 11.45 million was utilized for the purpose from the IDA Credit of  Tanzania: Integrated Roads Project Page 7 Implementation Completion Report Part 1: Project Implementation Assessment which about US$5.0 million directly on repairs and rehabilitation and US$ 6.0 million indirectly in the forms of supervision vehicles, studies, technical assistance. 15 plants and 7 vehicles were rehabilitated; 35 plants and 12 vehicles had overhaul of major components; and 19 plants and 47 vehicles received upgrades. But the impact of the above expenditure on improved equipment availability was marginal as most of the equipment in PEHCOL fleet was old. In retrospect, the economic merit of rehabilitation of the old equipment is questionable and there is need to look for new approaches to establish equipment-leasing companies in remote rural areas. GOT has now decided to sell 51 % to 75% shares of PEHCOL to private investors. Bids were opened in May 1999. Valuation of bids is ongoing under the Presidential Parastatal Sector Perform Commission. 25. National Transport Corporation (NTC). Following restructuring exercises, measures were taken to bring about reduction of fleet sizes and manning levels of the companies with a view to improve efficiency, reduction in costs and therefore improved profitability. Two of the RETCOs showed profit in 1997/98. The previous years, RETCOs showed losses due failure to adjust to shortage of consolidated cargoes arising from the liberalization of agricultural crop marketing. Starting 22nd of August 1997, NTC and the associated companies became a Specified Public Corporations and are earmarked for divestiture during FY 1999/2000. Government has decided to sell the RETCOs to the workers, the public and institutions in the respective regions. As regard to Shirika la Usafari Dar es Salaam Ltd. (UDA) it has been decided that 75% of the shares will be sold to investors in 2000. 26. Tanzania Air Corporation (ATC). Management and administration study funded under the Credit had some impact on the profitability of TAC. Of the seven items TAC planned to procure under the project five were implemented. Two items, the airline revenue accounting package, and purchase of 18 reservation terminals totaling US$132,302 will be financed under IRP II. Ill. Physical 27. Overview. Overall, achievement of the IDA civil works component is now assessed as partial. Assessment of project progress was initially more positive. In mid-1993, based on initial progress, but not based on actual physical achievements, the project was considered a success and therefore a follow-on project, IRP II, was appraised. The basis for the partial rating is for the physical achievements is already highlighted in paras.12-13. However, the gravel roads implementation merits further details as it eventually led to major procurement problems. 28. Gravel Roads. All the gravel roads constructed between late 1992 and early 1994 under IDA financing were substantially changed from the original scope of works from spot improvement, regular periodic maintenance-rehabilitation to works to full-scale reconstruction. Moreover, this was done without full assessment of the financial and budgetary impact of the decisions being made and without an IDA "no objection". While the designs were adequate for the original scope of works, they were not adequate for the new scope of works. Moreover, no detailed designs were made available to the contractors or given to the Bank for prior review. The design changes were made on site through variation orders issued by supervision consultants. For example, in the cases of Lusahunga-Usagara (266 km) and Tunduma-Sumbawanga (226 km), construction were undertaken without proper design, drainage and widening of roads from 6.0 m to 7.5 m were undertaken. In the cases of Babati area roads (310 km), there was a widening of roads from 6.5 m to 9.5 and finally to 10.2m and reducing the road length to 107 km by the omission of sections (i) Babati-Bereku (35 km) and (ii) Babati-Singida (168 km); and introduction of a major realignment at Sekenke escarpment which involved major earth works and drainage structures. In the case of Tanga-Horohoro (68 km) and Marangu -Tarakea (52 km) roads, two projects about 500 km apart were lumped together in one contract. Partially as a result of these changes many sections of these roads failed during the first major rains. Similar conceptual and implementation mistakes occurred with the emergency rehabilitation of Dar es Salaam Roads. 29. By May 1994, the poor quality of works and high unit cost of ongoing projects began challenging the initial assessment. In the case of gravel roads financed under IDA, substantial  Tanzania: Integrated Roads Project Page 8 Implementation Completion Report Part 1: Project Implementation Assessment changes in scope of work without appropriate engineering design and the Bank's approval led to significant quality of works failures and major cost overrun over the contracted amounts, making impossible achieving the physical targets set at appraisal. Serious attempts were made to salvage the above components, but the operating environment proved intractable. In March 1995, IDA requested the government to appoint an independent consultant to look into the causes of failure of the rehabilitated gravel roads and to suggest actions to be taken to rectify the situation. In July 1997 when the consultant report came out, IDA was to proceed with preparatory process for a major restructuring of the project but progress halted when the entire senior management of the ministry was criticized in the Warioba Commission Report on corrupt practices. According to the Report, '...Officers of the Ministry of Works receive bribes in order to give favor in awarding tenders; to accept upward variation of contracts; to conceal the weakness of the contracts; and in approving payments. Moreover bribes are offered at roadblocks in order to let undeserving vehicles through." Further, the previous consultants' and contractors' claims for the gravel roads remained outstanding. It was only in March 1998, that the reorganization of the ministry took place and in March 1999, the claims of the contractors and consultants were settled. Moreover, there was not adequate funds under the credit to fully finance the required works. Therefore. it was decided to close the project after one-year extension from the original closing date of June 1998. 30. The major finding of an independent consultants carrying out the project review was that the government supervision consultants did not provide MOW adequate waming signals of emerging problems. The consultants' review also indicated that the gravel road contracts have had procurement and project management problems starting from the pre-qualification of consultants, preparation of inadequate engineering designs to the contracting and construction supervision monitoring, actually since commencement of the Project in 1991. Poor direction and weak project management by GOT influenced the performance of the consultants. The production of the tender documents including the field survey took an abnormally long time (about two years to get to the tender stage). Pre-qualification of contractors was advertised between 31 January and 5th February 1988, after which, out of 32 contractors who applied, 27 were pre-qualified. This list was then used for actual biding for the gravel road contracts more than two years later. 31. Bridges. The SAR budgeted US$8.0 million for some 20 trunk road bridges and ten ferry crossings. But only one bridge, Kyaka Bridge and its associated works in Kagera district was completed at cost of nearly US$7.7 million exhausting the available resources for bridges and ferries. Implementation was followed by many failures of structures, approaches, temporary approaches, and temporary bridges. 32. Emergency Works. The IDA component of the project also financed road maintenance and a portion of outstanding balances for emergency works on Tunduma-Sumbawanga and other gravel roads. 33. Regional (District and Rural) Roads. Several donors financed rural roads under IRP I: including the EEC (200 km), UNDP (52 km) and USAID (1697.0 km). USAID under its Agricultural Assistance Program (ATAP No. 621-0166) was the major donor. In January 1996, USAID issued impact assessment report on its program. The report covered a review of the institutional, financial, road industry and local people-level impacts and the project sustainability. The report concluded that ATAP was making reasonable progress. It assessed the outcome as likely sustainable, provided the program continues the support of the GOT. 34. Studies. Various studies have been carried out in the areas of construction sector industry (Table 8). The extension and update of the Zanzibar's road network study from 1991, successfull completed on time, is one such project that is likely to attract financing.  Tanzania: Integrated Roads Project Page 9 Implementation Completion Report Part 1: Project Implementation Assessment IV. Macro-economic Impact of the Program 35. Significant transport cost reduction has been achieved on roads rehabilitated to good/fair standard (para.27). However, because the targets for roads in good condition was not realized under the Credit, the outcome compared to the high expectation at appraisal has not materialized. C. Major Factors Affecting Project Implementation 36. Overview. The factors that were partly outside the government control and which affected the implementation of the project included: difficult environment for implementation, inadequate design, and problems in donor coordination and poor performance of consultants and contractors. The factors within the government's control were lack of good governance including corrupt practices, failure to apply lessons learned from preceding project implementation, substantive changes in the scope of works without proper documentation, lack of proactive response, frequent reorganization of the implementing agencies, inadequate counterpart funding, and poor staff incentives. Factors within the control of the implementing agencies included procurement problems, deficient contract administration, and poor project and financial management control. 37. Factors Not Generally Subject to Government Control. * Lack of realism in the project design. In principle, project design should prioritize investments and ensure an appropriate balance between the development of the productive capacity (road network) and the institutional and policy capacity to oversee the network. In this project, institutional aspects were stated as primary objective (evidence by the order of presentation of the two objectives in the project), but there was never a clear consensus on this between the government and donors. * Effective Donor Coordination remained an issue throughout, although a senior technical advisor was appointed as part of the project implementation arrangements. Information on the activities of the donors and resources were difficult to come by and to achieve coordination of priorities. There were no adequate documentation to help find out what each donor is doing and how many resources each donor was spending. Lack of coordination among donors resulted in excess local contracting capacity creation. Also, not all donors were active in the implementation issues. Finally, GOT received a number of conflicting signals, for example, on road fund, road agency, reorganization of the transport sector, on the speed at which these changes were to be implemented and on the degree of achievement of the project objectives. * The government did not always get the best advice from its advisor and consultants on status of project management as evidenced in the poor quality civil works, cost over runs, delayed completion of subprojects and unfavorable terms of contracts. 38. Factors Generally Subject to Government Control. * The prevailing implementation environment was difficult. As noted in the Warioba Commission Report (Dec. 1996), the governance system prevailing in the ministry was not conducive to timely completion of project. Audited accounts also showed serious shortfalls. Macroeconomic performance was below expectations, and country performance ratings of inflation, fiscal deficit, savings and expenditure management were poor. OED reviews pointed to difficulties in controlling the budgetary pressures that were fostering inflation. * The lessons of experience of the preceding Highway Rehabilitation Projects appear to have little impact. Poorly documented engineering design were one of the main causes of delays and cost escalations under the Sixth Highway Rehabilitation Project and remained so in IRP I.  Tanzania: Integrated Roads Project Page 10 Implementation Completion Report Part I: Project Implementation Assessment * Counterpart financing was a problem. The prevailing macro-economic conditions made it difficult for the government to meet its commitment to make available the proceeds from road funds on time and in the amount required for road maintenance. Because of the development budget funding problem, MOW used the road fund revenues to fund emergency expenditures. * There were several implementation issues for which adequate and timely attention on the part of the GOT would have radically changed the situation on the ground and the perception of donors towards the government's performance. Lagged responses on the part of the government raised concern as to the government's commitment. Some of these included: decision to change the scope of works from rehabilitation to reconstruction without proper engineering design; delays in commercialization of PEHCOL and delayed establishment of road agencies; prolonged delays on taking action against corruption, use of dedicated road fund for the purpose not intended, etc. The governments delayed responses to the donor community's queries on the 1995/96-road fund allocation and road agency issue led to a request by the donors to meet with the President on these issues. * Frequent reorganization of the implementing ministry in the midst of implementing such a large project had also adverse impact on project implementation. The split of MWCT into two separate ministries diverted senior management attention away from implementation of IRP projects towards reorganization. The reorganization also resulted in lack of coordination between the two ministries dealing with transport and was for a time a source of delays until functions were sorted out. - The weak incentive system in the implementing agency did affect staff proactive performance, contributing to delay completion of the project, leading to substantial changes in scope of work without supporting technical preparation. The result was of significant consequence to project implementation - cost escalation, including changes in-contract prices of over 100% of the original amount. * The weak contracting industry also contributed to the poor quality of works and cost escalation. 39. Factors Generally Subject to Implementing Agency Control. * The project faced several problems with timely decision making processes, delays in execution, cost overruns, implementation problems, technical/financial problems, technical and financial monitoring and quality of executed works. The selection of contractors, their financial capabilities and the contractors' capabilities to import equipment created also problems for the project. Variation orders in magnitude well above than 15% of the project contract without the Bank's approvals were allowed contrary to Appendix I paragraph 3 of the Guidelines for procurement under IDA Credit. Throughout implementation, the coordinating office was not adequately equipped or staffed to effectively handle these issues for such complex project. Generally, it was less proactive and left issues unresolved for too long. * Procurement and contract administration were focal points of IRP I problems through out the project life. In particular, the area of procurement and negotiation of contracts, the implementing agency could have contributed much more than it did. Either due to omission or commission, procurement problems, and deficient contract management contributed to cost escalations and poor quality. Moreover, the contracts provided generous price escalation payments. For example, foreign currency exchanges were fixed at unrealistic rates resulting on variation on the dollar to almost 20% per annum whereas in real terms it should have been only 4% at most on the international market. Moreover, the foreign currency portion should not have been escalated for the first 12-months of the contract. In addition, the Tanzanian portion should not have been subject to price escalation before the first 12 months of the contract expire, and thereafter price escalation should have reflected only the inflation prevalent in the country in the various inputs. Overall project implementation delays were between 30 to 36 months. Procurement delays  Tanzania: Integrated Roads Project Page 11 Implementation Completion Report Part l: Project Implementation Assessment averaged 189 weeks from tender to signing of contracts; poor contract administration contributed to price escalation in majority of cases exceeding 100% of the contract price. Inadequate contractors capacity contributed to poor quality of works resulting on failing road structures and road surface, particularly for gravel roads. D. PROJECT SUSTAINABILITY 40. Project Sustainability. Overall, the sustainability of the project is assessed as uncertain. It appears that much longer time than originally planned is needed to assure sustainability. Sustainability of the progresses made with respect to institutional objective depends on the prevailing environment, government commitment to maintenance-first policy, proactive leadership, the incentive system to skilled manpower and on follow-up measures to consolidate the progress achieved to date - road fund, road agency, etc. The sustainability of the commercialized activities under the liberalization regime depends on the entrepreneurial ability of the new contractors and their financial capacity. The sustainability of the maintenance system introduced depends on the performance of the economy to generate road fund revenue and eradication of corrupt practices. E. Bank Performance 41. Identification and Preparation IDA's performance during identification and preparation phases is assessed as satisfactory. IDA staff rightly identified that transport was a major constraint facing the Tanzanian economic recovery in its transition to market-oriented economy. During preparation IDA staff provided critical assistance to GOT in coordinating donor assistance and packaging the program. 42. Appraisal. Appraisal did address many issues and was cognizant of the complexity of the project design. In retrospect, however, there were some shortcomings in its assessment: defining the alternative design of the project (e.g. considering a series of several projects based on an agreed program suggested at the time); the inadequate incentive regime as well as the readiness of the implementing agencies to perform basic tasks. Appraisal was also weak in assessing the risks on the physical objectives due to multiplicity of development objectives, which required major institutional structural changes. Moreover, given that the rationale for undertaking the project was to remove the transport constraints on the ongoing economic recovery program, the primary development objective of the project should have been the physical objective (83% of the project cost), and institutional capacity should have focussed more, and first on the basics of procurement, contract administration and project management with the scope of institutional reforms more phased. Certainly, the perception at appraisal that the primary risk of the project would be from "slow liberalization!" is a reflection of a failure to assess the real reasons of the project of this size. Second, the risks due to lack of incentives to staff was identified at appraisal as a risk, but an adequate sustainable solution was not found in the context of the civil service at the time. Third, since it is known from experience that earlier IDA financed highway sector project had experienced basic implementation problems in performing the basics tasks of procurement, focusing on contract administration and project management at appraisal would have led to a more focussed project. In summary, appraisal is assessed as deficient. Supervision 43. As noted in the earlier sections of this ICR, the conceptual design of the program and its large- scale investment was aimed at meeting the enormous requirements of the country for a well- functioning road infrastructure at a time when the economy was recovering. The deteriorated road network posed serious constraints to the movement of both people and goods, and particularly the flow of agricultural inputs to farms and outputs to markets, and the mobility of people to access social services. The donor community considered that a large program encompassing both capital  Tanzania: Integrated Roads Project Page 12 Implementation Completion Report Part l: Project Implementation Assessment construction and basic capacity building was needed to elicit a supply response to the macroeconomic reforms, which the government was undertaking. 44. Against this background, the donor community agreed to finance a complex and high-risk program totaling US$871 million. There was great optimism that the program would work and this sense of optimism was reinforced by the apparent eagerness of the government to reform its public administration. However, the outcome of the program execution shows that, given the scale of the appraised program and its complexity combined with the weak institutional structure, the donor community's optimism was misplaced. 45. The project achieved initial success: the launching ceremonies, the signature of the first large contracts and the completion of one or two links during the first two years, the implementation of the of the reforms in decentralization of road maintenance, maintenance planning, improved reporting, the quality of presentation at mid-term review. However, this early success later proved to be a liability, since it obscured the ability to detect and respond to implementation problems which soon arose. As the number of contracts increased, the existing infrastructure to implement the project did not grow in tandem and as a result enormous delays in procurement processing occurred leading to huge increases in program costs. 46. Although the design of the project anticipated the stationing in Tanzania an experienced highway engineer from the Bank to assist the government in coordination of the project, eventually an ex-Bank Principal Transport Engineer (PTE) was located in the Bank's Resident Office answerable to the Permanent secretary of MOW (who was responsible for implementing the Project). The arrangement appeared to be working. However, as the works progressed, conflicting signals appeared regarding the role of the engineer, causing confusion to the client, Bank management, contractors and consultants. The optimistic views of the engineer on the program performance and his communication with other players actually contradicted the findings of the headquarters-based task team and its early warnings that the project's institutional structure could not manage the planned activities, that the planned milestones could not be achieved within available resources and that the project was moving towards a crisis. The conflicts implied in the engineer dual roles in acting as both an adviser to the government and as a supervisor by the client during the first four years of the program clearly delayed the emergence of a consensus among the task team, Bank management and donors on required measures to turn around the performance of the program. 47. The influx of funds and the complexity of the project combined with the inability of the implementing organization to increase its capacity at a rate commensurate with the growing volume of works, and the overall poor governance environment in the country, led to major mismanagement of the program and eventually to corruption. By early 1994, the Bank supervision reports began to clearly state the problems, but Bank management influenced by the resident engineer, and the optimistic donor perceptions did not immediately grasp the seriousness of the situation. However, by the beginning of 1995, Bank management decided to strengthen the headquarters supervision mission team by including an experienced senior highway engineer and the magnitude of the problems facing the project were clearly identified. 48. In retrospect, questions could be raised such as: (i) why the early concerns raised during initial supervision missions were not responded to more quickly; (ii) why questions were not raised earlier on the growing disconnect after 1992 between the resident engineer and the supervision missions; and (ii) why an independent engineer was not added to the supervision team until after substantial commitments were locked into. Clearly the initial progress led to a misplaced level of optimism but more important the decision to place a senior engineer into the field as both an advisor and implementor led to a conflict which contributed to substantial implementation difficulties later on. 49. By 1995 the government had lost control of the program, arguably a result of mismanagement. The headquarters task team downgraded the rating of the project to unsatisfactory, stopped giving no objections to new contracts, focused on ongoing contracts and on ascertaining the precise physical and financial status of the project with a view to restructuring it. Bank management, now  Tanzania: Integrated Roads Project Page 13 Implementation Completion Report Part l: Project Implementation Assessment convinced about the deficiencies in implementation of the project, aggressively supported the task team and intensively dialogued with the government on the required project management and institutional changes. 50. In 1996, the government concern about mismanagement in all facets of public administration led to a report on the State of Corruption in the Country (Warioba Report) in which the Ministry of Works was highly criticized. The report stated that ".Officers of the Ministry of Works receive bribes in order to give favor in awarding tenders, to accept upward variation of contracts; to conceal the weaknesses of contractors; and in approving payments. Moreover bribes are offered at road blocks in order to let vehicles through." The report recommended a total reorganization of the Ministry and removal of the management team, the same team which had been earlier seen as "champions" of the program by the Bank and the donors. More than two years after the Warioba Report, in February 1998, the government eventually replaced the top managers and engineers who had been implicated by the report. The Ministry was re-organized and five months later, the Bank was able to partially close this project and to restructure the ongoing Second Integrated Roads Project. In June 1999 the government passed legislation for establishment of the Road Act and Road Fund Board and for establishment of TANROADS. 51. With regard to basic capacity building, which was to take place in parallel with the works programs, the results have been poor because the conditions for effective capacity building did not exist. Although large amount of resources was spent on this effort over a nine year-period (para.17), the absence of an adequate incentive structure undermined the efforts. The Ministry's highly qualified professionals received monthly salaries barely adequate to meet living costs for two weeks. At the same time they had to work side by side with counterparts, provided by donors, contractors and consultant's staff, who eamed 100 times their own salaries. This arguably led to an environment that undermined the project implementation. 52. The arrangements and safeguard measures planned in the project did not adequately recognize the risks given the large scale of the project, its complexity and the weak institutional structure. The unrealistic optimism that the government's apparent commitment to reforming public sector administration would be translated, through the capacity building component, into real capacity needed for project implementation proved to have been a miscalculation. The failure to adequately take into account the lessons learned from the Sixth Highway Project, a smaller scale operation that nevertheless showed similar problems, proved fatal in a much larger and ambitious project. Status of Civil Works of IDA financed component, ref. Paragraph 13 above in the text, shows the cost increases of the IDA financed road works and the magnitude of the deviations from original contract values. 53. Thus, it is in the context that both appraisal and supervision are rated at all level of management as marginally unsatisfactory - a set of good ideas undermined by an unduly ambitious and optimistic project design; the early successes following initial contract signatures and improvements in road conditions making it difficult to identify and gain acceptance by subsequent delays and problems; the controversial role of the Principal Transport Engineer in advising both the government and the Bank management and staff; and poor governance conditions in the country all had a role in the project's eventual failure. Notwithstanding the substantial efforts of the headquarters based task team, in identifying the problems during the supervision as the project's midlife and their extensive efforts to get things on track throughout the rest of the project's life, the conditions in-country and in the donor community and the role of the engineer undermined the effectiveness of the supervision efforts during the critical years when the bulk of the investments were committed. F. BORROWER PERFORMANCE 54. Preparation. Overall Borrowers preparation is assessed as satisfactory. The government's initiative in its Transport Sector Policy Paper and its Transport Recovery Program and  Tanzania: Integrated Roads Project Page 14 Implementation Completion Report Part l: Project Implementation Assessment its ability to bring some 16 donors to agree in support of its program over a period of three years, was a major undertaking. It showed leadership and made good on the reservoir of good will the government had among the donor community at the time. Coming to closure was a fit of accomplishment. On the other hand GOT had some shortcomings. It had not made adequate preparation to correct the weaknesses revealed in the course of implementation of a series of Highway Projects IDA financed earlier. It had not piloted or pre-tested the process of procurement and contract management process to be used in administering the project. 55. Implementation. GOT has overestimated its implementation capacity. The government was not able to execute its ambitious program on schedule and within budget. Audit reports of project accounts, as the scale of implementation increased, were highly qualified. In particular, GOT and its implementing agencies failed to make good use of the resources at their deposal as in the case of the gravel roads. Where the govemment has ultimately made good on its commitments, the responses were not often timely and sometimes had required the prodding of the donor community as in the case of the Road Fund, Road Agency. In the area of procurement, project management and contract administration, a decade after the launching of IRP I, and after continuous reminder of the problems in these areas, and substantial expenditure of resources for manpower training, there is no evidence that the implementing agency is significantly better than it was at the start of the project. GOT management of its consultants was not demanding. It must also be stated that the consultants and the advisor to the govemment did not provide always quality professional services. These professionals could have contributed to better quality of works but failed to do so as evidenced from their implementation performance. Most disappointing was the fact that the project advisor failed to steer the project implementation in the right course. 56. In particular, there were problems related to the use of the Special Account. An initial deposit of USD12.0 million was made to the Special account (SA) in March 1991. By January 1996, the balance in the SA had dropped to about US$5.6 m. CODAP had made payments of about US$3.2 million on account of seven addenda for civil work and consultant services. As no formal "no objections" were given to the amendments covering such additional works and services, CODAP failed to follow IDA procurement procedures for supplementary contract increases and made payments without following IDA procurement procedures for supplementary contract increases. Until an independent consultant's report on the addenda was published, IDA limited the SA to the available USD2.4 million, suspended replenishment starting July 1996 and the minimum size application correspondingly reduced from USD1.2 million to USD150, 000. Outstanding invoices amounting USD5.6 million were directly paid from project accounts. G. Assessment of Outcome. 57. Project outcome. Overall, the outcome of the IDA financed component is assessed as unsatisfactory. However, as a result of the project, there is today a more organized sector program. The government has learned valuable lessons from its implementation experience. A private sector led road construction sector industry has emerged. Road maintenance is now carried by private contractors, who have largely come into existence with the support of the IRP I. Despite the temporary relapse in the administration of the road fund in 1995/96, road fund now contributes a significant share of the cost of maintenance. The Road Fund Board supported by the Road Fund Act administers the road fund. The TANROADS is finally going to be fully operating by July 2003. However, the impact of transport cost reduction on the macroeconomic situation was not as high as expected at the start of this ambitious program. Also, the basics of project management and contract administration have yet to be mastered. H. Future Operations 58. Key Elements of Future Planned Operation. To lay the foundation for excellence in institution building, there is need to assure that the modest capacity achieved to date under the  Tanzania: Integrated Roads Project Page 15 Implementation Completion Report Part l: Project Implementation Assessment project, especiaUly the trained manpower and the creation of a road agency, is sustained. Assuring adequate resources for road maintenance will remain essentially an ongoing issue. Given past performance of the economy, efficient and effective use of the available road fund will be another critical area of focus. 1. Key Lessons Learned and Future Actions. 59. Key Lessons Learned and Future Actions. This Integrated Roads Project has been the largest and most ambitious project in the sector. As such it has been the subject of multiple reviews including a recent Bank Quality Assurance Group review. This ICR identifies a number of key lessons for future project design and implementation in Tanzania. * Project Design. While the concept of sector investment project on which IRP I is based appears workable, its application to a specific country requires making tradeoffs between major development objectives based on detailed analysis of major risks in achieving those objectives. In particular, in the civil works, the absence of detailed engineering for all major roads immediately prior to start of construction, made variation orders almost inevitable, leading to delays and cost increases. In institutional capacity building too, the many demands for reforms, taxed the capacity of the government to deliver on budget and on schedule. Therefore, in projects of IRP I type, it is important to introduce "checks and balances" with early warning signals to be able to adapt the program when required. * Corrupt practices. The poor implementation performance of IRP I testifies that the potential for corrupt practices can significantly impact project implementation performance. In this case, roads that would have been constructed under the project to good standard remained uncompleted partly because of corrupt practices and the absence of systems and procedures that could check against their occurrence. The Government should institute measures to safeguard against such practices. Physical progress should be verified against design specifications. Stringent measures should be taken under situation where audit reports are constantly delayed and qualified in order to fight potential corrupt practices from blossoming. Learning from Past Mistakes. The lessons of experience of past project implementation were to safeguard against future mistakes both for IDA and the borrower. As a matter of fact, major flaws experienced in the Sixth Highway Project Credit were also experienced under IRP 1. For example, appropriate technical documentation and provision of adequate resources should be fully completed before commencement of works. In short, more attention needs to be paid to past lessons of experience during project design and implementation. * Composition of Project Team. Supervision team should have professional mix with key persons being sufficiently experienced in the major fields of a project activities such as engineering, institutional reforms, etc. In addition, there should not be frequent changes of supervision management and teams, in order to ensure continuity and consistency of efforts in achieving quality and quantity of the planned targets. * Technical Documentation. Technical documentation necessary for implementation of civil works should be updated before actual commencement of construction works. * Donor Coordination. The spirit of cooperation of the donor community in support of the Integrated Roads Project I had no parallel. It was the first time so many donors have committed themselves to such a large common program. Unfortunately, in IRP 1, it has not been easy to maintain the interests of all donors' cooperation during implementing phase as each donor focus on execution of its own mini program under the parallel financing arrangement. In future, in large investment supported by several donors, a professional coordination managed by the government should be established, with agreed terms of references.  Tanzania: Integrated Roads Project Page 16 Implementation Completion Report Part 1: Project Implementation Assessment * Technical and Financial Audits. The Government should among other things, carry performance technical and financial audits of the works carried out. Such audits should, besides frequently revisiting the overall status of implementation, also include verification of the physical and financial targets. * Initial Successes in Long-term Program. Usually during the initial phases of project cycle of large programs, there is high optimism of the team involved and there is a tendency to underestimate the operation environment of the specific country for which the project is designed, even when there are potentially implementation problems. Initial successes usually mask deficiencies related to preparation, selection of contracts and judgements of capacities. It is therefore important that project implementor should not be taken in by an initial euphoria and ensure that measurable output targets are met. The project should have well defined qualitative and quantitative indicators, which should be adequately designed so that any deviation from initial targets can be detected and corrected in real time.  Tanzania: Integrated Roads Project Page 17 Implementation Completion Report Part II: Statistical Tables TANZANIA INTEGRATED ROADS PROJECT (IDA COMPONENT) (CREDIT 2149-TA) PART II: STATISTICAL TABLES TABLE 1: SUMMARY OF ASSESSMENTS A. Achievement of Objectives Substantial Partial Negligible Not Applicable Macroeconomic policies E El x Sector policies E X E E Financial objectives D X a E Institutional development