Tunisia - Highways Maintenance and Rehabilitation Project

Document of The World Bank FOR OFFICIAL USE ONLY Report No. 15673 IMPLEMENTATION COMPLETION REPORT REPUBLIC OF TUNISIA HIGHWAYS MAINTENANCE AND REHABILITATION PROJECT (LOAN 2896-TUN) June 3, 1996 Private Sector Development, Finance, and Infrastructure Division Maghreb and Iran Department Middle East and North Africa Region This document has a restricted distribution and may be used by recipients only in the perfortnance of their official duties. Its contents may not otherwise be disclosed without World Bank authorization.  CURRENCY EQUIVALENTS Equivalents in Dinars of US$1 At appraisal 0.83 Project average 0.93 Period average 0.92 Annual averages: 1987 0.83 1988 0.86 1989 0.95 1990 (.88 1991 0.92 1992 0.88 1993 1.00 1994 1.01 1995 0.95 FISCAL YEAR January I - December 31 ABBREVIATIONS AND ACRONYMS DGPC - Direction G6nerale des Ponts et Chaussees ERR - Economic Rate of Return SAR - Staff Appraisal Report  FOR OFFICIAL USE ONLY REPUBLIC OF TUNISIA HIGHWAYS MAINTENANCE AND REHABILITATION PROJECT (2896-TUN) WPLEMENFUMON COMPLDEON REPORT TABLE OF CONTENTS PREFACE ................................... EVALUATION SUMMARY ............................... iv INTRODUCTION ............................... iv PROJECT OBJECTIVES ............................... ; iv Objectives and Covenants ............................... iv Evaluation ............................... iv IMPLEMENTATION EXPERIENCE AND RESULTS ............................... iv Achievement of Objectives ............................... iv Sustainability ...............................v Costs and Delays ................................v Key Factors ................................v Bank and Borrower Performance ................................v Outcome ................................vi SUMMARY OF FINDINGS, FUTURE OPERATION, AND KEY LESSONS LEARNED ............................................... Vi Findings ...................................................... vi Project Operation ...................................................... vi Lessons for Future Projects ...................................................... vi IMPLEMENTATION ASSESSMENT .....................................................I1 STATEMENT / EVALUATION OF OBJ ECTIVES ......................................................1 Objectives at Appraisal .......................................................I Evaluation of Objectives and Risks ......................................................I ACHIEVEMENT OF OBJECTIVES .......................................................2 Sector Policies and Financial Objectives ......................................................2 Institutional Development .......................................................2 Physical Objectives .......................................................3 Economic Evaluation ......................................................3 MAJOR FACTORS AFFECTING THE PROJECT .................................. ...4 Not Generally Subject to Government Control ...................................4 Generally Subject to Government Control .....................................4 Generally Subject to Implementing Agency Control ......................................4 Cost Changes ...................................6 Implementation Delays .....................................7 PROJECT SUSTAINABILITY ....................7 BANK PERFORMANCE ..................7.............7 Identifi cat ion 7 Preparation .8 Appraisal .8 Implementation .8 BORROWER PERFORMANCE .8 Preparation .8 Implementation .8 Tds 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 wiLhout World Bank authorization.  - Hl - ASSESSMENT OF OUTCOME ............................................................ 8 PROJECT'S FUTURE OPERATION ............................................................ 9 KEY LESSONS LEARNED ............................................................ 9 Development Objectives .......................................................... 9 Implementation .......................................................... 9 STATISTICAL TABLES ............................................................ 10 TABLE 1: SUMMARY OF ASSESSMENT ............................................................ 10 TABLE 2: RELATED BANK LOANS/CREDITS ............................................................ I I TABLE 3: PROJECT TMETABLE ............................................................ 1 2 TABLE 4: LOAN/CREDIT DISBURSEMENTS: CUMULATIVE ESTIMATED AND ACTUAL .................................... 13 TABLE 5: KEY INDICATORS FOR PROJECT IMPLEMENTATION ............................................................ 14 TABLE 6: STUDIES INCLUDED IN PROJECT ............................................................ 15 TABLE 7: PROJECT COSTS ............................................................ 16 TABLE 8: PROJECT FINANCING ............................................................ 16 TABLE 9: ECONOMIC COSTS AND BENEFITS ............................................................ 17 TABLE 10: STATUS OF LEGAL COVENANTS ............................................................1 8 TABLE I 1: COMPLIANCE WITH OPERATIONAL MANUAL STATEMENTS .......................................................... 18 TABLE 12: BANK RESOURCES: STAFF INPUTS ............................................................ 19 TABLE 13: BANK RESOURCES: MISSIONS ............................................................1 9 APPENDIXES ............................................................ 21 MISSION'S AIDE-MtMOIRE (EXTRACT) ............................................................ 21 Rapport d 'ach&vement .......................................................... 21 Conclusions provisoires ........................................................... 2 1 BORROWER CONTRIBUTION TO THE ICR ............................................................ 36 . .  REPUBLIC OF TUNISIA HIGHWAYS MAINTENANCE AND REHABILITATION PROJECT (2896-TUN) PREFACE This is the Implementation Completion Report for the Highways Maintenance and Re- habilitation Project in the Republic of Tunisia, for which Loan 2896-TUN in the amount of $ 63 million was approved on December 17, 1987, made effective on October 17, 1988 and closed on June 30, 1995. The remaining undisbursed balance of $ 1,212,525.95 was canceled on November 6, 1995. Final disbursement took place on March 27, 1996 when the Government reimbursed the Bank $ 589.25, the outstanding balance of the Special Account, which was canceled from the loan the same day. The report was prepared by Jaffar Bentchikou (Sr. Highway Engineer), MN1PI and re- viewed by Amir Al-Khafaji, Division Chief, MN1PI and Rene Costa, Projects Advisor, MNI. Preparation of this report began during the November 29 to December 14, 1995 mission. Findings were discussed with the Borrower during the March 11 to 20, 1996 mission. Comments on the report have been received from the staff that appraised the op- eration. The Borrower did not comment in writing on the draft report but has nevertheless issued its own evaluation (second appendix).   REPUBLIC OF TUNISIA HIGHWAYS MAINTENANCE AND REHABILITATION PROJECT (2896-TUN) EVALUATION SUMMARY INTRODUCTION 1. The project was the sixth Bank-financed highway project and the eleventh in the transport sector in Tunisia. A subsequent Rural Roads Project focuses not only on rural transport needs but also on preparing a major shift of road maintenance activities from force account to contracting them out. PROJECT OBJECTIVES Objectives and Covenants 2. The overall project objective was to maximize the impact of the resources avail- able in the highway subsector. It was translated into implementation of a maintenance program, including 800 km of road rehabilitation and 225 km of bituminous concrete overlays; improvements to personnel management and training; establishment of road management and cost accounting systems; increases in the density of traffic counts; and improvements in road equipment maintenance. The Bank identified key organizational actions that were used as conditions for loan processing. Loan covenants covered organ- izational actions and a detailed action plan and were designed to promote achievement of project objectives. Evaluation 3. The objectives were appropriate. The Bank required that formal organizational steps be taken and this delayed negotiations until the Bank, realizing the complexity of the legal steps involved, changed this condition from negotiations to effectiveness. Dur- ing implementation, DGPC gave uneven attention and support to the project objectives and resisted some institutional changes. This weak Borrower's ownership was a risk un- foreseen at appraisal. IMPLEMENTATION EXPERIENCE AND RESULTS Achievement of Objectives 4. Sector policies and financial objectives were only partially achieved. The tight budget for recurrent maintenance supplies and operations increased satisfactorily during the five-year period covered by the SAR and beyond, and many physical road mainte- nance targets were reached. Some of the more important targets set for 1992, however, have yet to be reached in Tunisia's 1996 budget.  - v - 5. Institutional development objectives, as well, were only partially achieved. Im- provements to personnel management and training and to the road laboratory, and the es- tablishment of a cost accounting system, were successful. The road management system, however, is not yet operational, and traffic data collection and the maintenance of road equipment have not improved. 6. Physical objectives were also only partially achieved, (107% of bituminous con- crete resurfacings but only 75% of road rehabilitations), mostly due to cost changes and rigidities in budgeting and programming. Sustainability 7. Project sustainability is likely. It will depend, however, on the Borrower's com- mitment to making in-depth reforms to road network management and financing, and to contracting out maintenance. Such reforms are being prepared under a follow-up project. Costs and Delays 8. Cost changes, due to the long delay between engineering studies for the first year of works and implementation, and higher than expected local inflation, were important for the road rehabilitation element. The availability of counterpart financing was a limit- ing factor and led to the cancellation of about $ 1.2 million of the loan amount at closing. There were some implementation delays due to slow decisionmaking (in particular for procurement of consulting services), restricted flow of information, the slow compliance with formal institutional covenants at effectiveness, and the Gulf War. Key Factors 9. Consulting services had limited impacts in some areas due to timid DGPC owner- ship of institutional objectives, counterproductive requirements by the highest procure- ment body, and the lack of on-site construction experience of local consultants. 10. The need for pavement overlays was underestimated, in part due to the delay be- tween engineering studies and implementation, and this led to not optimizing impact on road users for the road rehabilitation works. Furthermore, bridge works were programmed separately from pavement works, resulting in either road safety hazards, or in leaving project roads subject to flood disruptions. Bank and Borrower Performance 11. The Bank's performance was satisfactory overall. However, during the appraisal period, procedural difficulties in meeting the formal institutional conditions the Bank had put for negotiations led to unusual delays from appraisal to negotiations to effectiveness. The conditions led to a successful human resource element but also led to reduced quality of the road rehabilitation works by making the technical studies outdated. 12. The Borrower's performance was satisfactory at preparation but deficient at im- plementation, due to slow decisionmaking, rigidities in procurement and budgeting, and timid support given to institutional elements. v  - vi - Outcome 13. The project outcome is rated satisfactory. Economic returns are high, road main- tenance budgets have increased dramatically, the backlog of road rehabilitation has been reduced, and the training of personnel has been improved. However, the management of force account road maintenance activities could not be improved. This is not surprising, as it has now been established that the best road maintenance practice is to rely mostly on the private sector for both a large part of road maintenance works and the maintenance of the residual equipment fleet. Thus, road maintenance activities will now need to be reor- ganized around a drastic reduction in force account capacity (reduced equipment fleet and work force), the contracting out of most maintenance activities to local contractors, the transfer of rural and regional roads ownership to the regional councils, and corresponding budgetary reforms. SUMMARY OF FINDINGS, FUTURE OPERATION, AND KEY LESSONS LEARNED Findings 14. The Borrower's ownership of the institutional objectives was overestimated dur- ing the appraisal period. 15. The lack of timely availability of counterpart funds played a negative role in proj- ect implementation. 16. The highest procurement body (Commission Supdrieure des Marches) made counterproductive requirements when it required DGPC to provide vehicles, equipment, office space, and counterpart personnel to consulting assignments. Since these elements were already in short supply in DGPC, this made the assignments difficult to start and to implement and contributed to ineffective use of costly foreign expertise. Project Operation 17. The Government did not provide a plan for the project's future operation. DGPC, however, indicated that it plans to overlay project roads with bituminous concrete to im- prove pavement roughness; widen their narrow bridges, and replace most of their fords with bridges. DGPC is also preparing reforms to contract out road maintenance; the ex- tent and implementation speed of these reforms could be used as indicators in monitoring progress of road maintenance management. Lessons for Future Projects Development Objectives 18. The Bank should assess the complexity of the legal steps involved and the related risks of delays when attaching conditions to loan processing steps. 19. Appraisal should estimate the balance of financing between dinars and dollars and related foreign exchange risks, by using variable annual exchange rates based on the vi  - vii - constant purchasing power rule. Similarly, a project account system should be set up al- lowing to reconcile payments in US dollars and in dinars. 20. Road rehabilitation should extend to solving all identified problems at the same time for a given section; the needs usually include widening / construction of bridges and culverts, raising the road surface above flood levels, and treatment of spots with an ab- normal concentration of accidents. 21. Consultants should be more involved in supervision of works to build experience on the site and feed it back in the design of subsequent similar projects. Implementation 22. Due to the need to ensure continuity in project management from preparation to completion, organizational structures where different units are responsible for different phases of the project cycle (preparation, appraisal and implementation) should be avoided. vii  REPUBLIC OF TUNISIA HIGHWAYS MAINTENANCE AND REHABILITATION PROJECT (2896-TUN) IMPLEMENTATION ASSESSMENT STATEMENT / EVALUATION OF OBJECTIVES Objectives at Appraisal 1. The principal objective was to improve the use of resources by improving the management of the road network. Specifically, the project aimed at: * improving the efficiency of road maintenance through organizational changes and strengthening; * achieving a proper balance in highway expenditures among maintenance, re- habilitation, and new construction; * reducing the backlog of road rehabilitation; and * improving personnel management and training. 2. To achieve these objectives, the project comprised: * a five-year maintenance program, of which the Bank financed all periodic bituminous concrete resurfacing carried out by contract (225 km), renewal of road equipment and facilities, and equipment for repair workshops; * 800 km of road rehabilitation by contract; * improvement of the human resources management system, of which the Bank financed consulting services and overseas training; and * institutional development and studies, including the establishment of a new road maintenance unit, the improvement of equipment maintenance, the rein- forcing of laboratory services, and economic and detailed engineering studies. Evaluation of Objectives and Risks 3. The objectives were clear, realistic, and important for Tunisia's development strategy. They were not complex or demanding for DGPC. 4. The SAR mentioned that the only foreseeable risk was related to the timely avail- ability of counterpart funds and that speciai supervision and monitoring measures were included in the project to minimize this risk. 5. During implementation, the Borrower's priorities shifted somewhat. Some project elements (mostly the "hardware" part of the project, i.e. road works and supply of road equipment) received much more management attention than others (the "software" part, i.e. the institutional objectives). The overall objective of better road network management  ICR - 2 - 2896-TUN was also a casualty of DGPC's emphasis on physical results and resistance to institutional changes. The lack of timely availability of counterpart funds played a negative role in project implementation (the project's road rehabilitation length was cut, in part for lack of counterpart funds, while about $1.2 million of World Bank financing was canceled at project closing), but the most important limiting factor was DGPC's rather timid com- mitment to the institutional objectives and its unwillingness to jostle bureaucratic aspects in order to fulfill them. This weak Borrower's ownership of these objectives was a risk unforeseen at appraisal. ACHIEVEMENT OF OBJECTIVES Sector Policies and Financial Objectives 6. The achievement was partial. The stated objective of achieving a proper balance in road expenditures among maintenance, rehabilitation and strengthening was only par- tially translated into one specific project goal and one procedure spelled out in the project action plan. The goal was to increase the quantities of road rehabilitation and mainte- nance works over the 1988-1992 period in accordance with agreed targets, which rebal- anced the activities with major backlog and ensured an overall growth of 6% per year in real terms. The procedure was a mechanism by which the Bank was consulted before the annual road budgets were finalized. Project design was weak on financial objectives, which were not explicitly spelled out in the SAR. Furthermore, the figures and criteria for establishing adequate recurrent and capital budgets were inconsistent and scattered throughout the SAR. 7. Results were globally positive, the awareness increased. major imbalances were corrected, some of the more important physical targets were reached or surpassed over the period while others were not. The budget for recurrent maintenance supplies and op- erations, which is where the crunch was, was very close (-4%) to the SAR targets ad- justed for actual local inflation over the 1988-1992 period and again in 1996 is very close (-3%) to adjusted SAR targets. Many physical road maintenance targets were reached or surpassed, however, some of the more important ones set for 1992 (patching of paved roads and grading and regravelling of unpaved roads) have yet to be reached in the 1996 budget. Institutional Development 8. The achievement was partial. Improving the efficiency of road maintenance through organizational changes and strengthening and improving personnel management and training was translated into (a) establishing an organization and methods unit (for establishing road management and cost accounting systems and improving the density of traffic counts); (b) improving mechanical equipment maintenance; (c) reinforcing labora- tory services; (d) establishing a training directorate; and (e) preparing a five-year training action plan. 9. The organization and methods unit was established, but the initial objective of countrywide use of modem road network management tools had to be scaled down dur-  ICR - 3 - 2896-TUN ing implementation to a pilot introduction in six governorates. As it turned out, even this reduced objective was not reached, since only the analytical accounting package is being used while neither the road management system nor the new approach on traffic counting are operational. 10. Four new regional workshops were to be built and equipped, and technical assis- tance was to help the equipment directorate of DGPC establish a five-level set up for equipment maintenance; develop equipment inspection functions; improve equipment maintenance tracking and monitoring and inventories management. Also, a few existing provincial workshops were to be rehabilitated. The regional workshop and the technical assistance programs, however, were canceled for lack of funds and lack of DGPC owner- ship, while only limited reforms were introduced (equipment inspection) and the provin- cial workshops were rehabilitated. The funds for technical assistance were not used for this element. Currently, the equipment fleet remains oversized, heterogeneous, underused and under-maintained and equipment management and supervision is still weak. Why did the Bank agree to the cancellation of the regional workshop program? At the time of project appraisal, it was still conceivable to carry out most road maintenance by force ac- count and thus to develop a large program to address its traditional weakest point, the maintenance of the equipment. Now it has been established that the best practice is to rely mostly on the private sector for not only a large part of road maintenance works but also the maintenance of the residual equipment fleet required by a reduced force account ca- pability. Thus the program would have been a step in the wrong direction. 11. Consulting services for the road laboratory and for training had satisfactory achievements, as did the establishment of a human resource system for personnel man- agement, and training. A comprehensive training action plan was prepared and imple- mented. Training actions, still would be more effective were they integrated into a com- prehensive career and professional growth system. This success follows previous at- tempts to improve personnel management and training, which had failed because of the lack of institutional framework. A study to prepare a follow-up (rural roads) project was added and had satisfactory achievements. Physical Objectives 12. The achievement was partial. The physical objectives were (a) 225 km of bitumi- nous concrete resurfacings, and (b) 800 km of road rehabilitations. The actual achieve- ments were: (a) 240 km of bituminous concrete resurfacings (107% of objectives), and (b) 596 km of road rehabilitations (75% of objectives). The reduction in the length of road rehabilitations is due to a higher per km cost and to rigidities in budgeting and pro- gramming, which did not permit increasing annual budgets for the element. In terms of value, the physical objectives were reached (para 23). Economic Evaluation 13. The SAR does not mention an overall ERR. The road rehabilitation element had a minimum ERR of 30%. The road maintenance program had ERRs ranging from 28% to 48% at appraisal, and no ERR was reestimated at completion because, despite moderate  ICR -4 - 2896-TUN cost increases, the traffic levels and the improvements in pavement roughness are in the ranges that usually bring high ERRs. 14. ERRs for 584 km of road rehabilitation ranged from 30% to 100% at appraisal. An economic reevaluation was carried out at completion on all rehabilitated sections on the basis of actual cost of works, traffic counts in 1987 and 1992, updated estimates for 1994 traffic and beyond, updated VOCs, and estimated maintenance costs based on opti- mum strategies. The revised ERRs range from 26% to more than 100%, with a median of 70%. MAJOR FACTORS AFFECTING THE PROJECT Not Generally Subject to Government Control 15. Several project consulting assignments had a two-and-half year period between the start of the procurement process and the start of the assignment, due in part to the Gulf Crisis. This delay impacted negatively on institutional development. Generally Subject to Government Control 16. Consultant services were procured with some delays and used with uneven results. This was mainly because, in order to reduce foreign costs, the highest procurement body (Commission Superieure des Marches) made counterproductive requirements when it re- quired DGPC to supply vehicles, equipment, office space and counterpart personnel. Since these elements were already in short supply in DGPC, this made the assignments difficult to launch and to implement. 17. In addition, local laws and regulations do not differentiate procurement of con- sultant services from procurement of civil works. Thus one could find unexpected pro- curement procedures or clauses in consultant contracts such as late penalty clauses and bid and performance bonds. Also, tax laws applicable to consultant contracts are com- plex. There is, for example, a registration fee for all contracts; its amount is nominal for studies but reaches a high 3% of the contract value for technical assistance. 18. Furthermore, local consultants in Tunisia are used almost exclusively to prepare feasibility and detailed engineering studies and are rarely involved in works supervision. This results in studies that may be unrealistic or difficult to implement, since consultants' staff are rarely given the opportunity to accumulate field experience. On-site construction experience and feedback should be used to improve the design of subsequent similar projects. Generally Subject to Implementing Agency Control Impact on road users 19. The sections of the road rehabilitation element for the first year were selected on the basis of their widening needs, to improve capacity and safety, with the strengthening  ICR -5 - 2896-TUN needs (on about one-third of the length) secondary. Usually the order of priority is re- versed for technical reasons' and to better answer users' needs (because strengthening can drastically improve pavement roughness, the main parameter of user comfort), while widening-only rehabilitations are kept to a minimum, when the existing congestion calls for urgent action. During implementation, the order of priority was rightly reversed at the light of increased pavement failures which led DGPC to increase the proportion of strengthening to 62% of project roads. The remaining sections will be programmed to rapidly receive a bituminous concrete overlay, which should bring their roughness to ac- ceptable levels. 20. DGPC could have paid more attention to optimizing the positive impact of the road works on the road users by aiming to make itineraries more homogeneous, and at the same time focusing on solving all the needs of the project sections instead of program- ming works by type. Some road sections are still using fords or submersible bridges (Irish bridges) instead of (high) bridges and culverts or have sudden variations of road width because bridge and culvert construction or widening was deemed of lower priority and would be carried out separately. Some other sections are still regularly flooded, and rais- ing them could have been included in the project with substantial cost savings compared to raising them later. Finally, spots with abnormal accident concentration were identified and treated not systematically but on an ad-hoc basis. It should be said, however, that an- swering all of these needs would have led to more costly rehabilitation works for which a larger budget would have been needed at the expense of other road investments. A fully operational road management system would assist considerably the decision makers on that matter. 21. The overall quality of the road rehabilitation element was barely satisfactory, due to several factors: the low quality of the technical studies, the time gap between these studies and the actual works (with the pavements deteriorating further during the period), the quality of contractors' work, and the quality of supervision. Implemnentation of studies and technical assistance 22. DGPC was late in supplying foreign consultants with office space, telephone lines, vehicles, specialized equipment, and counterpart staff. When supplied, it was often short of contractual requirements, and contributed to ineffective use of costly foreign ex- pertise. DGPC did not give full support to the technical assistance teams, particularly the one working on the network management tools. As required by the legal documents, DGPC established a training directorate and a training steering committee, and appointed a training director; however, when the appointed official was transferred to another posi- tion outside DGPC early in the project period, no replacement was ever made. 1 Widening a road without concurrent pavement strengthening is technically challenging to do correctly due to the need to reach high pavement densities on the two narrow pavement strips on each side to avoid differential settlements; the presence of an overlay allows one-sided widenings, which are eas- ier to compact correctly, and reduces the effects of differential settlements.  ICR -6 - 2896-TUN Cost Changes 23. The main project component, the 800 km of road rehabilitations, were estimated at appraisal to cost DT 50.440 million or $60.770 million, including contingencies. This translates into about 63,000 DT / km or $76,000 / km. At completion, 596 km of rehabili- tation had been carried out at a cost of DT 55.425 million (a 9.9% increase) or $58.089 million (a 4.4% decrease). This translates into about 93,000 DT / km (a 47.6% increase) or $97,500 / km (a 28.3% increase). 24. These wide variations, justified in part by changes in exchange rate and by the need to increase the proportion of pavement strengthening, show that: * the engineering studies for the first year works on which the appraisal was based had become too old at implementation time to be accurate; * SARs should be conservative in estimating: (i) the physical contingencies (changing at negotiations to a higher ratio than the standard 10% could have been justified to compensate for the ripening quality of the engineering stud- ies); (ii) the price contingencies 2; and (iii) the exchange rate risk (at appraisal cost estimates were computed in local currency and converted to US Dollars using a fixed exchange rate of DT 0.83 / US $1.00, which was prevailing at appraisal time3; the actual exchange rate varied moderately during the 1988- 95 period, between DT 0.83 and 1.01 per US $, with a period and project av- erage of DT 0.93 / US $); and * a minimum of flexibility is necessary in the Government's annual budgeting and programming process to match the World Bank financing without large cancellations at the project completion. 25. Thus the SAR had an inaccurate estimate of the balance of financing between di- nars and dollars. The problem was compounded by the Government's reluctance to pro- vide additional local financing during the implementation period to match the World Bank financing. At the end, $1,213,185.20 was canceled from the loan. This cancellation would have been about $3.6 million larger, had the Bank not agreed to finance, late in the project, the procurement of 40 additional graders. Also, DGPC had the handicap of not having set up a project accounting system capable of reconciling payments in dinars and in US dollars. 2They were computed on the 5-year expenditures program ending in 1992 while the disbursement sched- ule was in line with the 7.5-year disbursement profile ending in June 1995 and which was proven ac- curate; a 13.7% price contingency was used while the actual international price inflation was 19.8% for the period 1988-92 or 29.4% for the period 1988-95 and local price inflation was 40.8% for 1987- 92 or about 490/o for 1987-95. 3The World Bank software Costab has the option of using Constant Purchasing Power Parity exchange rates, where annual exchange rates are computed on the basis of the inflation differential between the local and foreign elements. This method would have minimized the exchange rate risk.  ICR - 7 - 2896-TUN Implementation Delays 26. The project was declared effective ten months after Board presentation, due to delays in setting up and staffing the training directorate and the training steering commit- tee and appointing the project coordinator. Thereafter, lengthy procurement procedures, followed with zeal but with no sense of urgency, delayed (in combination with the Gulf war) the start of the first road works by another seven months and the start of the studies and technical assistance by about two years. Shortages of counterpart funds also contrib- uted to delays by forcing implementation in phases. 27. The main reason for project delays was, however, organizational. The slow deci- sionmaking process was compounded by the restricted flow of information and docu- ments within DGPC and to the regional directorates. The project was prepared by the Study Directorate of DGPC and the road works were implemented by the regional direc- torates. A project coordination unit was created but was not involved in preparation, and during implementation was not directly involved in contract management and financial aspects. At the end, however, the loan closed on time because, even if the initial project implementation schedule and provisions for contingencies were too optimistic, the loan closing date was rightly set to match the standard disbursement profile for highway proj- ects. PROJECT SUSTAINABILITY 28. Project sustainability is likely. There is a caveat, however, as it has now been es- tablished that the best road maintenance practice is to rely mostly on the private sector for both a large part of road maintenance works and the maintenance of the residual equip- ment fleet. Thus project sustainability will depend greatly on the Borrower's commitment to making in-depth reforms in road network management, maintenance, and financing. Such reforms are being considered and prepared under the follow-up project (Loan 3840- TUN, Rural Roads Project). They would include a major push in favor of contracting out most road maintenance, the transfer of rural and regional roads ownership to the regional councils, and corresponding budgetary reforms. Furthermore, DGPC plans to systemati- cally resurface project roads with a bituminous concrete overlay. This, combined with a program to construct, rehabilitate, and widen bridges and culverts, would in a short period bring the level of service of the project roads that were only widened to satisfactory lev- els. BANK PERFORMANCE Identification 29. Performance is rated satisfactory. The project was identified as a follow-up phase of the Fourth Highway Project (Ln. 1841-TUN), under which a major road maintenance study was carried out. It included five-year slices of road and bridges rehabilitation and maintenance programs, and emphasized the need to develop and implement a pavement  ICR - 8 - 2896-TUN monitoring and evaluation system, a training program, and a traffic management unit. The project also included elements to be implemented by the Ministry of Transport. Preparation 30. Performance is rated satisfactory. The Bank emphasized the need to reorganize DGPC for more efficient road maintenance and to appoint a project coordinator. Appraisal 31. Performance is rated marginally satisfactory. The Bank's insistence on the formal establishment of a training directorate and a training steering committee within DGPC, and personnel had positive impact on the satisfactory achievement of the human resource element. However, this insistence, together with organizational changes in both the Gov- ernment and the Bank, was also the main reason for the unusual 19-month period be- tween appraisal and negotiations. When it was realized that a decree from the Prime Minister was necessary for the project to go forward, that became a condition of effec- tiveness, which took another year to be fulfilled. The loss of momentum and the increased risks in project implementation (cost overrun or quality of works) that resulted from the increased time gap between studies and actual road rehabilitation works and from the rushed post-appraisal could have been prevented with better communication. The SAR was weak on cost estimates and had some minor incoherence. Implementation 32. Performance is rated satisfactory. BORROWER PERFORMANCE Preparation 33. Performance is rated satisfactory. Implementation 34. Performance is rated deficient. Decisionmaking was slow, in particular for pro- curement of consulting services. The departments responsible for some of the institution building did not have the commitment nor the capacity to handle technical assistance and there was no strong pressure from the hierarchy to make things change. ASSESSMENT OF OUTCOME 35. The project outcome is rated satisfactory. Economic returns are high, road main- tenance budgets have increased dramatically, the backlog of road rehabilitation has been reduced, and the training of personnel has been improved. However, the management of road maintenance activities could not be improved and will need to be reorganized around a drastically reduced equipment fleet and work force and the contracting out of most maintenance activities to local contractors.  ICR - 9 - 2896-TUN PROJECT'S FUTURE OPERATION 36. The Government did not provide a plan for the project's future operation. DGPC, however, indicated that it plans to overlay project roads with bituminous concrete to im- prove pavement roughness, to widen their narrow bridges, and to replace most of their fords with bridges. DGPC is also preparing reforms to contract out road maintenance; the extend and implementation speed of these reforms could be used as indicators to monitor progress in improving management of road maintenance activities. KEY LESSONS LEARNED Development Objectives 37. The Borrower's ownership of the institutional objectives was overestimated dur- ing the appraisal period. 38. The Bank should assess the complexity of the legal steps involved and the related risks of delays when attaching conditions to loan processing steps. 39. Appraisal should estimate the balance of financing between dinars and dollars and related foreign exchange risks, by using variable annual exchange rates based on the constant purchasing power rule. Similarly, a project account system should be set up al- lowing to reconcile payments in US dollars and in dinars. 40. Road rehabilitation should extend to solving all identified problems at the same time for a given section; the needs usually include widening / construction of bridges and culverts, raising the road surface above flood levels, and treatment of spots with an ab- normnal concentration of accidents. 41. Consultants should be more involved in supervision of works to build experience on the site and feed it back in the design of subsequent similar projects. Implementation 42. The lack of timely availability of counterpart funds played a negative role in proj- ect implementation. 43. The highest procurement body (Commission Supirieure des Marches) made counterproductive requirements when it required DGPC to supply vehicles, equipment, office space, and counterpart personnel. Since these elements were already in short sup- ply in DGPC, this made the assignments difficult to launch and to implement and con- tributed to ineffective use of costly foreign expertise. 44. Due to the need to ensure continuity in project management from preparation to completion, organizational structures where different units are responsible for different phases of the project cycle (preparation, appraisal and implementation) should be avoided.  REPUBLIC OF TUNISIA HIGHWAYS MAINTENANCE AND REHABILITATION PROJECT (2896-TUN) STATISTICAL TABLES TABLE 1: SUMMARY OF ASSESSMENT A. Achievement of objectives Substantial Partial Negligible Not applicable Macro policies 0 n n 0 Sector policies I] 3 0 Financial objectives n 0 0 0 Institutional development 0 R3 0 Physical objectives iii iii Poverty reduction E E 0 Gender issues Z o Other social objectives E 0 Environmental objectives F l 0 Public sector management E 0 El Private sector development E E 0 Other El E . B. Project sustainability Likely Unlikely Uncertain 10 Ol O _ C. Bank performance Highly Satisfactory Deficient satisfactory Identification E 0 E Preparation assistance E 0 E Appraisal E 0 E Supervision E 0 E  ICR -11 - 2896-TUN D. Borrower performance Highly Satisfactory Deficient satisfactory .................................................................................................................................................................................................................................. Preparation [ J Implementation E ]J Covenant compliance Q Q Operation (if applicable) E 5 5 E. Assessment of Highly Satisfactory Un- Highly outcome satisfactory satisfactory unsatisfactory E] El . ... TABLE 2: RELATED BANK LOANS/CREDITS Loan Number Purpose Fiscal Status & Title Year ............................................................................................................................................................................................................................................ Preceding operations Loan 746-TUN Improvement and construction of 275 km of primary and secon- FY71 Completed; First Highway dary roads; reconstruction and rehabilitation of 51 bridges and PPAR no Project g a culverts; resurfacing and rehabilitation of 1,920 km of paved 2732 of 12/79 Project roads; reorganization and strengthening of hughway m2 onte- nance operations; improvement of arrangements for transport coordination; study of road user taxation and road tariff regula- tions. .................................................................................................................................................................................................................................... Loan 1188-TUN Improvement of 225 km of primary and secondary roads; updat- FY76 Completed; Second Highway ing of transport survey; preparation of rural roads program. PCR no. 5647 Project of 5/85 ........................................................................................................................................................................................................................................... Loan 1601-TUN Improvement of about 1,000 km of rural roads and complemen- FY78 Completed; tary agricultural investments; purchase of road maintenance PPAR no. Phird Highway equipment, spares and tools, studies of about 2,000 km of addi- 7182 of 4/88 Project tional selected rural roads. ...................................................................................................................................I......................................................................................................... Loan 1841-TUN Rehabilitation of about 800 km of paved roads; maintenance FY80 Completed; Fourth Highway program including acquisition of equipment and materials; PCRno. 8648 Foucth Higway studies and training; materials testing and traffic counting Project equipment; studies and engineering of about 1,000 km of rural 5/90 roads. ........................................................................................................................................................................................................................................... Loan 2108-TUN Improvement of about 1,200 km of rural roads and upgrading FY82 Completed; Fifth Highway the agricultural areas surrounding those roads; construction of PCR no. (Rural Roads) other rural public works such as irrigation schemes, studies of 12468 of 8/93 Pro ect rural road construction and maintenance techniques and of a ProJect follow-up rural roads project, and technical assistance and training. ...........................................................................................................................................................................................................................................  ICR - 12 - 2896-TUN Following operation Loan 3840-TUN Preparation and initial implementation of an institutional reform FY95 Ongoing Seventh High- program aimed at decentralizing ownership and management of rural roads, budgetary reforms and contracting out maintenance; way (Rural periodic rehabilitation by contract of about 300 km of gravel Roads) Project rural roads and upgrading of 715 km of rural roads. TABLE 3: PROJECT TIMETABLE Planned Actual ............................................................................................................................................................................................................................. Identification 4/82 4/82 Preparation 9/82 - 9/85 9/85 Appraisal Mission 6/83 - 2/86 3/86 Loan Negotiations 10/86 10/87 Board Approval 12/86 12/87 Loan Signature 1/88 Loan Effectiveness 4/88 10/88 Loan Closing 6/95 6/95 Project Completion 6/95 6/95  ICR - 13 - 2896-TUN TABLE 4: LOAN/CREDIT DISBURSEMENTS: CUMULATIVE ESTIMATED AND ACTUAL Disbursements 2896-tun 100 90 _ 80 60 c co T0 O 50 D) '4 60S