Tunisia - Fourth Highway Project

Document of The World Bank FOR OFFICIAL USE ONLY Report No. 8648 PROJECT COMPLETION REPORT REPUBLIC OF TUNISIA FOURTH HIGHWAY PROJECT (LOAN 1841-TUN) MAY 18, 1990 Infrastructure Operatinns Division Country Department II Europe, Middle East and North Africa Regional Office This document has a resticted distribution and may be used by recipients only In the perfonmnace of their officiid duties. Its contents may not otherwise be disclosed wtbout World Bank authrizton.  CURRENCY EOUIVALENTS Currency Unit - Tunisian Dinar EXCHANGE RATE Appraisal Year Average US$1 = TD 0.40 Intervening Year Average US$1 TD 0.78 Completion Year Average US$1 - TD 0.82 ABBREVIATIONS AND ACRONYMS CPCT Cellule de Planification et Coordination des Transports (Transport Planning and Coordination Unit) CSM Commission Superieure des Marches (National Tender Board) DGPC Direction Generale des Ponts et Chaussees (Highway Department) ERR Economic Rate of Return HDM Highway Design and Maintenance Standards Model MEH Ministere de 1'Equipement et de l'Habitat (Ministry of Public Works) MT Ministere des Transports (Ministry of Transportation) PERL Public Enterprise Reform Loan VOCs Vehicle Operating Costs FISCAL YEAR January 1 - December 31  voa omius omx THE WORLD BANK Washington. D.C 20433 U.S.A. Olfce of Oct4*-Cwwal tOpinwtns EVAltian May 18, 1990 HEMORANDUM TO THE EXECUTIVE DIRECTORS AND THE PRESIDENT SUBJECT: Project Completion Report on Tunisia Fourth Highway Project (Loan 1841-TUN) Attached, for information, is a copy of a report entitled "Project Completion Report on Tunisia - Fourth Highway Project (Loan 1841-TUN)" prepared by the Europe, Middle East and North Africa Regional Office. No audit of this project has been made by the Operations Evaluation Department at this time. Attachment 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 authovition.  FOR OMCIL USE ONLY PROJECT COMPLETION REPORT REPUBLIC OF TUNISIA FOURTH HIGHWAY PROJECT (LOAM 1841-TUN) TABLE OF CONTENTS Pa8e No. Preface . . . . . . . . . . . . . . Basic Data Sheet .... . . . . . . . . . . . . ii Evaluation Summary . . . . . . . . . . . . . . . . . iv I. Background: The Transport Sector . . . . . . . 1 II. Project Preparation and Appraisal . . . . . . . 2 III. Project Implementation . . . . . . . . . . . . 4 IV. Economic Revaluation . . . . . . . . . . . . . 6 V. Institutional Development . . . . . . . . . . . 7 VI. Impact on the External Environment . . . . . . 8 VII. Bank, Borrower, and Consultant Performance . . 8 VIII. Findings and Lessons . . . . . . . . . . . . . 9 APPENDI 1. Introduction . . 1 2. The Road Network . . . . . . . . . . . . . . . 2 3. The Fourth Highway Project . . . . . . . . . . 3 4. Project Execution . . . . . . . . . . . . . . . 5 5. Procurement. 6. Project Cost and Disbursements . . . . . . . . 21 7. Evaluation of the Fourth Highway Project . . . 23 8. Economic Evaluation of Improved Maintenance . 26 9. Institution Building and Strengthening of Operations . .28 10. Conclusion . .29 ANNEXES 1. Payments and Commitments 2 Project Cost 3. Execution Dates: Civil WorKs 4. Execution Dates: Equipment 5. Execution Dates: Studies 6. Costs: Works 7. Costs: Equipment 8. Costs: Studies and Training 9. Highway Maintenance Trends This document has a restricted distribution and may be used by recipients only in the performance of their oMcial dut;cs. Its contents may not otherwise be disclosed without World Bank suthoriuation.  Table of Contents (continued) 10. Budget: Article 60 11. Traffic 12. Vehicle Operating Costs 13. Costs of Maintenance Operations 14. Vehicle Operating Cost Differences 15. Costs: Rehabilitation Project 16. Investment Increases Resulting from the Fourth Highway Project 17. Rate of Return, by Section 18. Benefits: Rehabilitation 19. Benefits: Improvements in Highway Maintenance 20. Sensitivity Analysis: Improved Maintenance 21. Network Condition 22. Disbursement Charts ATTACHMENT 1. Coments from the Cofinancier MAP IBRD 14732Rl - Tunisia: The Rehabilitation Program  1 i - PROJECT COMPLETION REPORT REPUBLIC OF TUNISIA FOURTH HIGHWAY PROJECT _ (LOAN 1841-TUN) PREFACE This is the Project Completion Report of the Fourth Highway Project in Tunisia, for which Loan 1841-TUN in the amount of US$36.5 million was signed on May 22, 1980. The loan was closed on September 30, 1987, three years behind schedule. The last disbursement took place on July 21, 1988, when the outstanding balance of US$3.6 million was cancelled. The main text of the report was prepared by the Europe, Middle East and North Africa Country Department II, Infrastructure Operations Division. It is based on information derived from the files of the EMENA Information Center, and supervision and completion missions. The Appendix and Annexes were prepared by the Tunisian Highway Department of the Ministry of Public Works. The main text is intended to complement the Borrower's comprehensive ex-post evaluation. It, therefore, focuses on broader issues since specifics are dealt with in the Appendix and Annexes. This PCR was read by the Operations Evaluation Department (OED). The draft PCR was sent to the Borrower and Cofinanciers for comments and they are attached to the Report as Attachment I.  - ii - PROJECT COMPLETION REPORT REPUBLIC OF TUNISIA FOURTH HIGHWAY PROJECT (LOAN 1841-TUN) BASIC DATA SHEET Key Project Data Appraisal Expectation Actual Total Project Cost (US$ million) 92.1 68.4 Underrun -- 351 Loan (US$ million) Disbursed 36.5 32.9 Cancelled -- 3.6 Repaid to June 1989 -- 13.6 Date Physical Components Completed 09/30/83 08/30/88 Proportion Completion by Above Date 50X 100l Proportion of Time Overrun -- 130X Economic Rate of Return 80X 70Xt Cumulative Estimated and Actual Disbursements (US$ million) ma8Z F8 F F FY86 F8 mY88 FY89 Estimated 8.2 22.5 34.0 36.5 36.5 36.5 34.0 32.9 Actual 0.0 4.3 11.0 12.9 21.1 26.3 31.3 32.9 Actual/Estimated (1) 01 191 32% 351 581 721 921 1001 Other Project Data Qriginal Revision Actual First Mention in Files -- 08/23/77 Negotiations 03/31/80 03/31/80 Board Approval 04/22/80 05/13/80 Loan Agreement Date 05/22/80 05/22/80 Effectiveness Date 08/21/80 11/21/80 Closing Date 09/30/84 09/30/85 09/30/87 09/30/86 Completion Date 05/84 09/86 07/88 09/87 Borrower Republic of Tunisia Executing Agencies Highway Department (Ministry of Public Works) and Ministry of Transportation Follow-on Project Name Highway Maintenance and Rehabilitation Project  - i - WISSION DATA Stage of Month No. of go. of DaYS Spetializetion Performace mature of Project Cve e Lewr f s n fietd Reresnted 1J RationL 2 Probt er 31 Identification 10/78 2 4 EC, HE Preparation 06/79 3 4 EC, HE Appraisal 09/79 3 24 EC, HE, HE Supervision 06/80 4 7 EC, HE, NE, TR 1 Supervision 02/81 3 6 EC, HE I Supervision 05/81 2 3 EC, HE 1 Supervision 11/81 3 4 EC, HE, 0A 1 Supervision 01/82 1 1 HE I Supervision 05/82 2 2 EC, HE 1 Supervision 09/82 3 2 EC, HE I Supervision 09/82 3 2 HE, TR I Supervision 02/83 2 3 EC, HE 2 F Supervision 05/83 3 3 EC, HE 2 F Supervision 04/84 3 3 EC, HE 2 F, M Supervision 09/84 2 3 EC, HE 2 F,M Supervision 03/85 2 3 EC, HE 2 F, M Supervision 03/85 2 3 EC, HE 2 F, 1 Supervision 10/85 3 3 EC, HE, DOC 2 F, M Supervision 05/86 1 1 TR 2 F, M Supervision 07/86 2 3 EC, HF 2 F, M Supervision 02/87 1 1 DDC 2 F, ' Supervision 09/87 1 3 EC 2 F,M Supervision 09/88 1 1 DC 2 F,M Coppletlon 01/88 1 I OA 2 Completfon 04/88 1 1 HE Completion 10/88 3 1 EC, HE Total Number of days in the field a 93 1 DC Division Chief; DDC - Deputy Division Chief; EC - Economist; HE - Highway Engineer; 4E - Mechanical Engineer; OA - Operations Analyst; TR - Traitiing Specialist 2 1 - Problem - free or moderate problems; 2 - Moderate problems 3 F - Financial; M - Managerial STAFF INPUT (in stoff Weeks) FY79 FY80 FY81 FY82 FY83 FY84 FY85 FY86 FY87 FY88 FY89 FY90 TOTAL Preapprafsat 36.2 8.7 44.9 Appraisal 35.6 35.6 Negociatfon 8.7 8.7 Loan Processing 0.1 8.0 8.1 Supervision 3.3 11.8 5.9 8.4 7.3 17.7 6.4 6.8 2.5 0.5 70.6 PCR 2.7 0.9 3.6 Other 0.2 0.3 5.1 0.4 0.7 6.7 TOTAL 71.9 28.7 12.0 5.9 8.4 7.6 22.8 6.8 7.5 2.5 3.2 0.9 178.2  - iv - PROJECT COMPLETION REPORT REPUBLIC OF TUNISIA FOURTH HIGHWAY PROJECT (LOAN 1841-TUN) EVALUATION SUMMARY Introduction and Backaround 1. The socio-economic setting against which the project was identified and prepared is discussed in pares. 3-16 of the April 18, 1980 President's Report (PR, Report No. P-2778-TUN). Paras. 2.01-2.33 of the April 18, 1980 Staff Appraisal Report (SAR, Report No. 2810-TUN) describe the status and prospects of the highway subsector in the early 1980s. The salient feature was that whereas the density of the network was generally adequate, rapid traffic growth over the previous decade and a decline of total expenditures in real terms reduced many roads to below standard conditions for the traffic they carried. Narrow pavement widths resulted in high transport and road maintenance costs and contributed to a poor safety record of many roads. Hence, road rehabilitation and upgrading of the safety program emerged as priority activities. 2. When the project was appraised in September 1979, the Bank was familiar with constraints on project implementation in Tunisia. Project Performance Audit Reports (PPAMs) on the Second Port Project (OED Report No. 1049 of January 8, 1976), on the El Borma Gab6s Gas Pipeiine Project (OED Report No. 1078 of February 26, 1976), on the First Railway Project (OED Report No. 2109 of February 13, 1978), and on the First Highway Project (OED Report No. 2772 of May 24, 1979) discuss in detail reasons for cost overruns, time overruns, and partial achievement of project objectives. Objectives 3. The overall objectives of the project were to improve highway network management and to achieve a balanced development of highway infrastructure (SAR, para. 3.02). Specific instruments for the achievement of these objectives were to: (a) meet the urgent backlog of road rehabilitation, estimated to be around 1,700 km (SAR, para. 3.09), of which the project would finance about 50%; (b) exRand road maintenance activities and improve the efficiency of the maintenance organiz-tion; physical targets were established (SAR, paras. 2.24, 3.04 and Annex V), and improvements in performance are discussed (Appendix, para. 4.6);  (c) strenathen the institutional framework through the introduction of imDroved nlanning techniques. information systems, and staff training; this was accomplished to a satisfactory extent (PCR, para 5.02); (d) raise grogressively the recurrent budget for maintenance to an adeguate level; the SAR (para. 3.04) recognized that this was a long-term effort, partially met (Appendix para. 4.6), but adjustments are still needed in financing mechanisms (PCR, para 1.04); and (e) support a program of traffic management and road safety; serious delays have led to the program's being partially implemented under the Fifth and Sixth Highway Projects (para. 3.09). Implementation Experience 4. Experience with the implementation of the different components constituting the appraised project (SAR, para. 3.03) was as follows: (a) a three-year time slice (1981-83) of a six-year program for improving highway maintenance. including the purchase of hi&hway maintenance equipment. The highway maintenance program began on schedule in September 1981 but by 1984, implementation of civil works had slowed down because of: (i) budgetary constraints; (ii) decentralization of highway maintenance, which had begun in 1982; and (iii) an ongoing analysis of needs, benefits, and budgetary requirements. Procurement of highway maintenance equipment experienced relatively minor delays and difficulties. All in all, physical targets were attained (Appendix, para. 4.6); (b) a three-year slice of a six-year grogram of road rehabi'jitation. Some 780 km of road sections out of the 800 km originally planned were rehabilitated. Works consisted mainly of road strengthening or widening, and the rehabilitation of shoulders and drainage on the country's oldest and most highly trafficked links. Civil works were completed more than four years behind schedule mainly due to the inadequate provision of counterpart funds; (c) training for highwav deRartment staff. including refresher courses in Tunisia and overseas. Human resources development was hampered by an ill-defined organizational structure for the training function and inadequate budgetary allocations. It focussed on specific areas of technical competence to the exclusion of overall staffing and training needs of the Highway Department;  - vi - (d) grovision of laboratory test equipment. traffic counters. and traffic aiPnS. Laboratory test equipment and traffic counters were delivered as planned. Procurement of traffic signs experienced delays and partial cancellation of contracts due to the depreciation of the dinar and import license procedures; and (e) technical assistance for groject monitorinf and evaluation. for overall highway investment planning. and for preparation of annual work programs and a future rural roads proiect. The Highway Master Plan was completed in June 1981. It was based on an examination of the existing network structure in light of the needs arising from long-term transport demand and regional planning programs. The Transport Planning and Coordination Unit was established in December 1983, about two years behind schedule. It became fully operational by 1985 and has since then provided a valuable contribution to the strengthening of overall planning, coordination, and monitoring functions within the Ministry of Transportation. It was instrumental in initiating sector reform and the preparation of a deta base on road and rail transport. The Highway Maintenance Study, completed in March 1985, provided a good assessment of the condition of the paved network. esults 5. The estimated project cost was US$92.1 million, with about US$45.5.million representing foreign costs (SAR, para. 3.16 and Table 3.1). The actual project cost was US$68.4 million (Appendix, para. 6.1). Comparison of the two figures has little meaning because of the massive changes in the US dollar exchange rate, which was between DT 0.499 at appraisal, reached DT 0.890 in March 1985, and stabilized at about DT 0.820 in 1988. Project completion, originally scheduled for May 1984, took place in July 1988. Time overruns were due to lack of counterpart funds resulting from cost increases and the large unspent dollar balance created by the increased exchange rate. The estimated economic rate of return (ERR) was 80X (SAR, para. 4.09), the re- estimated ERR 701 (Appendix, para. 8.6; and Annex, Table 17). Institutional strengthening showed encouraging results (Appendix, para. 9). Sus,tainabilim t 6. The increasing maturity of the principal Executing Agency, the Highway Department, demonstrated by its thorough evaluation of project results in the Appendix, is the best indication that, subject to timely and sufficient allocation of funds in the future, the sustainability of benefits generated by the project is assured. PLndings And Lessons 7. Project experience illustrates the benefits from Bank-Borrower collaboration over a series of lending operations: institut'ons have been strengthened, planning and information techniques improved, and future work prepared taking into account stumbling blocks encountered in the past.  - vii - 8. The project shows that comparisons of estimated and actual cost figures have to be thoughtfully interpreted. For the rehabilitation component, actual costs did exceed those at appraisal but the proportional increase was comparable to the proportional increase in project benefits. 9. The principal difficulty during project execution arose from shortages in counterpart funds. Increases in the US dollar exchange rate demanded corresponding increases in local budget appropriations, which were difficult to obtain. Clearly, this issue deserves careful attention not only in the case of Tunisia but in many other Borrowing countries. 10. A particularly rewarding aspect has been the steady strengthening of institutions, already noted in the PPAM of the Second Highway Project. Tunisian officials have consistently demonstrated their ability to adopt planning and management techniques imported through technical assistance and to make the necessary adjustments to ensure that foreign methods and solutions are appropriate for Tunisian requirements.  - 1 - PROJECT COMPLETION REPORT REPUBLIC OF TUNISIA FOURTH HIGHWAY PROJECT (LOAN 1841-TUN) I. BACKGROUND: THE TRANSPORT SECTOR 1.01 In 1979, when the project was appraised, the transport system comprised some 17,000 km of classified roads, 1,700 km of standard- and narrow-gauge railways, five major commercial ports, one national shipping company, four airports handling international traffic, and two airlines. Road transport was the principal mode used for both domestic and passenger transport and its share of total demand had been steadily increasing over the years. Roads carried about 90% of passenger traffic and 85% of freight, exclud4.ng minerals. 1.02 Since 1979, combined road and rail traffic has increased by about 8% p.a. for both freight and passengers with a continued shift from railways to road transport. About two-thirds of rail freight traffic consisted of phosphates. The share of main-line rail passenger traffic has declined from 37% to 13%. The relatively small size of the country and the concentration of economic activity in the Northeast contributed to declines in both freight and passenger traffic; the fact that there were few long hauls curtailed opportunities for profitable railway traffic. Nevertheless, the Sixth Plan (1982-1986) allocated he.vy investments of dubious economic justification for the expansion and upgrading of the railway network. Total investments and operational subsidies, which compensated for artificially low pasbcnger tariffs as well as administrative and maintenance inefficiencies, grew to about US$100 million equivalent by 1986. Subsidies were needed to enable railway passenger services to compete with road passenger services. Except for the transport of minerals and other long-haul freight, road transport clearly had a comparative advantage. The discrepancy between the decreasing role of the railways in the transport sector and the increasing financial resources needed to run it, at a time when public investments needed to be reduced, was brought to the Governmeavts attention in a Transport Sector Memorandum (Green Cover, July 1985). The Government concurred with the Bank's position, but did not change its policy. The Bank together with the Transport Planning and Coordination Unit (CPCT) of the Ministry of Transport (MT) undertook a brief study analyzing the railway investment program, which the Government followed in preparing the Seventh Plan (1987-91). This analysis was expanded into more comprehensive review of the railways, the results of which were used and updated in a comprehensive Bank public expenditure review1/ and the Transport Sector Strategy (Green Cover, August 1988). 1.03 The transport sector is presently faced with rising budgetary demands as well as evidence of inefficiencies and incompatibility of objectives. The railway component of the Public Enterprise Reform Loan (PERL), approved in July 1989, is expected to reduce the budgetary demands of the national railway company (SNCFT) for investment purposes as well as improve its cash generating t Tunisia: AdaRting Public Exvenditures to Changing Resource Availability, January 1987.  - 2 - capacity. As a result of Bank initiatives and discussion, the Government increased its awareness of the short-, medium-, and long-term costs of uneconomic transport infrastructure investments. Its new transport sector policy, aimed at rationalizing domestic transport, contains the following key elements: (a) transport pricing that reflects long-term marginal costs in an increasingly competitive environment; (b) improvement of cost-accounting practices and introduction of economically sound principles to calculate compensation for (social) services provided below cost; and (c) gradual Government withdrawal from the provision of road freight services and the promotion of road/rail competition. The national freight transport company has been restructured as a competitive operation and offered for sale to the private sector. 1.04 MT is responsible for the coordination of all modes of transport, the improvement of transport regulations and tariffs, the overall development of the transport sector, and the planning of rail, port, and airport infrastructure. The Highway Department (DGPC) in the Ministry of Public Works (MEH) underwent decentralization in 1982. As a result, each of the country's 23 governorates has a regional MEH directorate with a highway department responsible for maintenance, own-account construction, and equipment management and maintenance. Highway subsector financing is included under separate capital and recurrent expenditure budgets. Road users contribute to Government revenues through various taxes and fees such as annual vehicle registration fees, taxes on fuels, lubricants, and tires, and import duties on vehicles and spare parts. The total revenue from road users generally covers total capital and recurrent expenditures. However, there are indications that heavy vehicles may be undercharged. This s'-.ation has been analyzed in the context of a broader study2/, which took Tunisia as its case study. MT through CPCT is in the process of updating the findings of the study and expects to implement recommended schemes for road-user charges and related taxation measures with Bank financing under a forthcoming Transport Sector Loan. II. PROJECT PREPARATION AND APPRAISAL Sector Obiectives and Bank Stratefy 2.01 The main objectives for the Fifth Development Plan (1977-81) for the transport sector were to: (a) improve the existing infrastructure, especially through modernization and rehabilitation works; (b) promote rural development by improving feeder roads; (c) increase Tunisia's share of international traffic; and (d) reorganize regional transport companies. The Bank supported national objectives in the transport sector with particular emphasis on: (a) assistance in the modernization, rehabilitation, and maintenance of existing infrastructure; (b) further improvement of rural roads; and (c) continued strengthening of transport institutions, including improved investment planning. A Fourth Highway Project would address the road maintenance problem while a Fifth Highway Project would provide further assistance in improving 2 Road Transport Taxation in Developing Countries. David M. Newbery Gordon A. Hughes, William D.O. Patterson, and Esra Bennathan; the World Bank, 1988.  rural roads. A Third Port Project would assist the Government in the modernization of berth facilities at the main ports of La Goulette and Sfax and in providing support facilities for improved port operations. Project Obiectives 2.02 The overall objectives of the project were to improve the management of the Tunisian highway network and achieve a balanced development of the highway infrastructure. Specifically, the project was designed to: (a) meet the urgent backlog of road rehabilitation; (b) expand road maintenance activities and improve the efficiency of the maintenance organization; (c) strengthen the institutional framework through the introduction of improved planning techniques, information systems, and staff training; (d) progressively raise the recurrent budget for maintenance to an adequate level; and (e) support a program of traffic management and road safety. Proiect DescriDtion 2.03 The project initially consisted of: (a) a three-year slice (1981-84) of a longer-term program for improving highway maintenance, including the purchase of highway maintenance equipment; (b) three-year slice (1981-84) of a longer-term program of road rehabilitation; (c) training of DGPC staff through in-country and overseas refresher courses; (d) provision of laboratory test equipment, traffic counters, and traffic signs; and (e) technical assistance to DGPC for: (i) the establishment of an Organization and Methods Unit for project monitoring and evaluation, preparation of annual work programs, and traffic, road inventory, and accident data processing; (ii) preparation of studies for highway investment and maintenance planning, including a Highway Kaster Plan; (iii) preparation of economic studies and detailed engineering for the second- and third-year rehabilitation programs; and  (iv) preparation of detailed engineering for about 1,000 km of rural roads in anticipation of the Fifth Highway Project (Loan 2108-TUN). Project Costs and Financing 2.04 The project was appraised at a total cost of US$145.5 million, including a foreign exchange component of US$71.5 million. At that time, the Government was intent on implementing the targeted 1,300 km of road rehabilitation and was confident that it could obtain Saudi Fund cofinancing to cover a sizeable portion of the foreign and local costs. In the end, however, the Government was unable to secure Saudi cofinancing. This financing gap resulted in a reduction of the rehabilitation program to 800 km. It also lowered the total project cost, substantially, to US$92.1 million, with a foreign exchange component of US$45.5 million, which would be financed through a US$36.5 million Bank loan and US$9 million in suppliers' credits. Soon after the loan became effective, the OPEC Fund agreed to finance the equipment component, obviating the need for most of the suppliers' credits. Subsequent Changes in the Proiect Scope 2.05 In 1981, the Loan Agreement was amended to include financing of technical assistance for the establishment of CPCT within MT. Although this component had been discussed during appraisal in October 1980, it was not until the following year that overall agreement was reached on the need for assistance in transport planning. The Loan Agreement was amended again in June 1985 to include the financing of a Traffic Management and Pilot Safety Program and a tranche of road works comprising the rehabilitation of three highways (MC33, MC36, and GP6). III. PROJECT IMPLEMENTATION Overvi 3.01 Project implementation took over four years longer than originally planned. The primary factor was the country's macro-economic performance, which had begun to deteriorate during the early 1980s. Consequently, the Sixth Plan (1982-86) proposed slower growth in new investments and recurrent expenditures, which had a negative impact on the availability of counterpart funding. Other main causes of implementation delays were inadequate provision for staffing and cumbersome administrative procedures for obtaining import licenses. The former is now being addressed in a comprehensive manner under the Sixth Highway Project (para. 5.03) and the latter under the Structural Adjustment Loan (2962-TUN), which supports a medium-term adjustment program with measures to liberalize trade and reduce import restrictions. By 1992, most restrictions will have been removed. These measures should likewise facilitate the issuance of import licenses and are therefore a major step in the right direction.  Hiahwav Naintenance 3.02 The highway maintenance program began on schedule in September 1981 but by 1984, implementation of civil works had slowed down because of: (i) budgetary constraints; (ii) decentralization of highway maintenance, which had begun in 1982 (para. 5.02); and (iii) an ongoing maintenance study, which provided an updated analysis of needs, benefits, and the necessary budget. This component experienced relatively minor procurement problems. For example, the agent for the selected four-wheel drive vehicles could not assure after-sales parts or service with the result that the contract was cancelled. Another example is the procurement of service vehicles for project supervision and control. This was blocked by the Commission Superieure des Marches (CSM), which wanted smaller vehicles than those offered in the evaluated bid. Further discussion of the procurement of road maintenance equipment and implementation of the maintenance program is provided in the Appendix, paras. 4.3 and 4.6. Civil Works: Road Rehabilitation 3.03 Civil works procurement posed no major problems, since general prequalification had been carried out in 1979 during project preparation. Start-up of the first slice, nevertheless, took place five months behind schedule in April 1981 because of the slow provision of counterpart funds. The second and third slices experienced similar start-up delays. By late 1983, when rehabilitation works were to have been completed, severe budgetary constraints caused a delay of about one year. By the end of 1984, overall progress was about 58% behind the appraisal estimates. Time-consuming administrative procedures also contributed to delays in civil works implementation. For example, the contract for the (MC82 road) rehabilitation was awarded in March 1983 to a contractor who was later declared bankrupt. It took about one year to retender and award the contract. Another contract (GP6 rehabilitation) was amended to take into account higher design standards resulting in a cost increase of 25% over the original contract amount. Koreover, the MC33 road, which runs through the industrial suburbs of Tunis, was delayed considerably because problems with utilities in the right-of-way had been underestimated at the design stage. During implementation, the Tunisian dinar depreciated against the US dollar at a faster pace than the differential inflation rate between local and foreign costs. It was, therefore, necessary to amend the Loan Agreement in 1984 and again in 1985 to raise disbursement percentages for civil works from the original 45% to 50% and 62X, respectively, to cover the foreign exchange cost of the component. Further discussion of road rehabilitation implementation difficulties is provided in the Appendix, para. 4.2. Training 3.04 In early 1981, DGPC appointed, as planned, a permanent training officer. She coordinated and oversaw a program of overseas and local training activities targeted to engineers, technicians, and works and site supervisors. Her work was hampered by an ill-defined organizational structure for the training function within DGPC and an inadequate budget. Further discussion of project-related training is provided in the Appendix, para. 4.5.  - 6 - aboratory Test Equipment. Traffic Counters. ant Traffic Si8ns 3.05 Laboratory test equipment and traffic counters were delivered as planned. But the original order of traffic signs had been split among three suppliers because the order was too large to be met by a single firm. Following the contract award decision, a rise in the value of the US dollar significantly increased the cost of the imported reflective sheet used in sign manufacture with the result that only one of the three firms delivered signs. Delivery was delayed due to the lengthy import license procedures followed by the CSM. The Bank also agreed to finance a lot of guardrail under this component. Further discussion of the procurement of laboratory test equipment, traffic counters, and traffic signs is provided in the Appendix, para. 4.3.). Technical Assistance 3.06 Highwav Master Plan. Completed in June 1981, the Plan examined the existing network structure in light of the needs arising from long-term transport demand, regional transport demand, and highway investment planning. 3.07 Transnort Planning and Coordination Unit. CPCT was established within MT in December 1983, about two years later than foreseen. It was behind schedule mainly because of delays in recruiting foreign consultants. A Unit Chief was named in May 1984. Allocation of offices and office equipment resulted in further delays. However within a year, it had a staffing complement of five Tunisians in addition to the consultants and became fully operational. 3.08 Highway Maintenance Study. The study provided a good assessment of :he condition of the paved network. It was followed up by further analysis of paved roads over a longer period and to cover gravel roads, and by regularly updated, axle-load surveys. Technical assistance and specialized equipment for these activities are being financed under the Sixth Highway Project. From an organizational standpoint, the study confirmed earlier interest in establishing a Road Maintenance Organization and Methods Unit in DGPC. This is also being financed under the Sixth Highway Project. 3.09 Traffic Management and Pilot Safety Progrm. This program was approved in mid-1985 but was held in abeyance until the new Ministry of Transport team took office. It was then decided that it would not be implemented under this loan, as insufficient time was left to establish and staff a Steering Committee and Working Group and to reach agreement on bidding documents and procurement procedures. The program is now being partially implemented ander the ongoing Fifth and Sixth Highway Projects. 3.10 Further discussion of the technical assistance program is provided in the Appendix, para. 4.4. IV. ECONOMIC REVALUATION 4.01 An economic revaluation was undertaken based on actual project costs and traffic volumes as well as updated vehicle operating costs (VOCs). Generally, traffic volumes have increased at a somewhat lower rate than  - 7 - forecasts made at appraisal, namely in the range of about 71 p.a. from 1980 to 1987 compared with forecasts of 10 p.a. VOCs have increased at a rate far above the 91 annual inflation rate during the period 1981-87, mainly as a result of the devaluation of the dinar. The increases range from as much as 301 p.a. for the price of passengers cars and fuel costs to around 121 p.a. for tires (Appendix, para. 7.3). In contrast, construction and maintenance costs have increased at rates much closer to the overall rate of inflation, namely at about 101 p.a. 4.02 The economic revaluation was conducted for about 50 individual sections rehabilitated under the project. As at appraisal, use was made of the Bank's Highway Design and Maintenance Model (HDM - Version 2). Most sections show high ERRs with the average ERR exceeding 1001. Only one section, the GP7, showed an ERR of just under 10, this being due to higher than expected construction costs and lower traffic volumes. Generally, the ERRs are slightly higher than at appraisal, mainly as a result of the major increases in VOCs since 1980. An economic revaluation was also undertaken for the overall improved road maintenance program. This involved assessing the surface condition of the road network for each category of road, classified into five groups by range of traffic level. This assessment was based on a detailed road inventory conducted in 1983 as part of the Highway Maintenance Study financed under the project (para. 3.06). The HDM model was used to provide VOC estimates over the network over a ten-year period (1981-90) based on traffic volumes, road surface conditions, and vehicle characteristics and costs. These estimates were made for the network as it is under the improved maintenance program and for the network as it would have been in the absence of such a program. Taking into account the investment costs of the improved road maintenance program, the ERR is estimated at about 701. V. INSTITUTIONAL DEVELOPMENT Sectorvide Coordination 5.01 A major project achievement was the establishment of CPCT. In spite of these initial difficulties, the Unit provided a valuable contribution to the strengthening of overall planning, coordination, and monitoring functions within MT. Among other items, the Unit was instrumental in initiating reform in the sector and the preparation of a data base on road and rail transport, and in preparing the sector analytical review for the Seventh Development Plan (1987-91). The Unit continues to operate within MT's newly restructured Planning Directorate and further assistance is now planned with funds remaining under the Third Urban Development Project (Loan 2223-TUN). Maintenance Oxganization 5.02 In 1982, highway maintenance operations were decentralized to reduce the control of DGPC headquarters over routine maintenance tasks. It took several years for DGPC to adjust, resulting in lower output of road maintenance activities. The Highway Maintenance Study (para. 3.08) provided an evaluation of the new system as well as recommendations, some of which were implemented under the Fourth Highway Project (e.g., organization of maintenance crews, definition of equipment needs, and acquisition of a first tranche of maintenance equipment). By project completion, global highway  - 8 - maintenance output matched the physical targets agreed under the project. However, the Highway Maintenance Study identified further increases in productivity that could be achieved. This is now being pursued under the ongoing Sixth Highway Project (Loan 2896-TUN). Human Resources Development 5.03 Project-related training activities improved the technical competence of professional staff but fell short of actual needs, particularly in the fields of mechanical equipment ant network and site management. Overall results were moderate because of limited commitment to training, insufficient budgetary allocations, inadequate job definition, and cursory training needs assessment. Staff trainirg improvements continue under the Sixth Highway Project. In fact, the establishment and staffing of a Training Directorate were conditions of loan effectiveness. Project Sustainability 5.04 The project rehabilitated roads, improved maintenance, and strengthened planning and monitoring capacity. Project successes, as well as shortcomings, of this project ha.c been systematically followed up by the Fifth and Sixth Highway Projects. In fact, methodology, criteria, and procedures for technical and economic appraisal of rehabilitation subprojects to be financed under the Sixth Highway Project were developed urnder the Fourth. Subject to .he availability of funds, the operational and institutional benefits generated by the project will be sustainable. VI. IMPACT ON THE EXTERNAL ENVIRONMENT 6.01 Civil works components were concerned mainly with the rehabilitation of existing road structures and pavement, including shoulders and associated drainage. Road sections rehabilitated under the project were key sections of Tunisia's main road network and supported regional development. There was no adverse environmental impact in terms of land use, expropriation, resettlement, or any other ecological or sociological aspect. VII. BANK. BORROWER. AND CONSULTANT PERFORMANCE Bank Performance 7.01 Close project supervision by the Bank, both during missions and through desk reviews, helped the project survive a protracted period of austerity. The Bank showed flexibility in adjusting to differential inflation rate evolutions and in approving Government requests to include priority subcomponents in the project (e.g., fourth tranche of road works, extension of the traffic sign component). The Bank was also instrumental in having MT participate in the project, through the establishment and operation of CPCT. However, in spite of strong Bank assistance in preparing the road safety component, the Pilot Traffic Management and Safety Program did not materialize. The Bank has included this item in its current sector strategy dialogue with the Government.  -9- B21orFr POXtoKmunce 7.02 MEH/DGPC, the main executing agency, performed generally well. Some delays occurred because of lengthy procurement procedures that led, in the case of the acquisition of traffic signs to cancellation of part of that program. However, the main cause of delays was the lack of counterpart funds. * The Ministry of Finance provided investment funds, albeit generally late, while funding for the recurrent budget remained scarce. To remedy this situation, the Sixth Highway Project included recurrent maintenance operations. Compliance with loan covenants was generally good. In 1985, MEH was about one year behind in the submission of quarterly progress reports because of staff turnover. DGPC prepared a thorough, in-depth ex-post evaluation of the project and produced an excellent PCR. MT was not involved in the project to the extent desirable. Whereas the work carried out by CPCT was impressive, MT did not maintain the momentum needed to carry out the Pilot Traffic Management and Safety Program. Relationship between the Bank and the Borrower 7.03 Good relations between the Bank and MEH/DGPC contributed to the clarification and resolution of many issues that arose during project implementation. Relations with MT were less effective since three different teams succeeded at the head of the Ministry during the project period and were less familiar with Bank procedures. Consulting Services 7.04 The Highway Maintenance Study and other technical assistance activities were satisfactorily completed. Special mention has to be made about the consultants for CPCT. Their performance was outstanding in view of the transfer of knowledge and enthusiasm to the three original counterparts, who remain at their jobs. Proiect Documentation and Data 7.05 Preparation documents (project briefs and the decision memorandum) outlined the objectives and major issues confronting the project. The Staff Appraisal Report provided in-depth analyses of the project's economic and institutional viability. Supervision reports, as well as legal addenda, clearly documented the circumstances surrounding various changes to the project scope and to the loan, such as new project items, extended closing dates, and revised disbursement percentages by category. The Borrower's PCR provided comprehensive statistical data. VIII. FINDINGS AND LESSONS 8.01 Project experience illustrates the benefits from Bank-Borrower collaboration over a series of lending operations: institutions have been strengthened, planning and information techniques improved, and future work prepared taking into account stumbling blocks encountered in the past. The establishment of CPCT was a major project achievement even though it took place two years behind schedule causing significant delays in project execution. The lesson learned is that whenever important institutional  - 10 - improvements have been identified at appraisal, the early implementation of corresponding measures should generally permit a more timely execution of the project. 8.02 Some significant changes were made to the project during the first year of implementation. One change was due to confirmation of cofinancing arrangements, which had not been clearly defined at appraisal. Another change involved the provision of technical assistance for transport planning, which was added at the request of MT. The Bank showed flexibility in respect of these changes, which as a whole served to improve project impact. 8.03 The principal difficulty during project execution arose from shortages in counterpart funding. Inc-eases in the US dollar exchange rate demanded corresponding increases in budget appropriations, which were difficult to obtain. Clearly, this issue deserves careful attention not only in the case of Tunisia but in many other Borrowing countries. 8.04 ProJect implementation delays were due, in part, to procurement bottlenecks. This demonstrates that the Bank and the Borrower ought to seek specific measures to alleviate cumbersome administrative procedures, particularly those involving the issuance of import licenses. The recently approved SAL is a good start in that direction. 8.05 The project shows that comparisons of estimated and actual cost figures have to be thoughtfully interpreted. For the rehabilitation component, actual costs did exceed those at appraisal but the proportional increase was almost the same as the proportional increase in project benefits. Thus, despite the cost increase, the ERR for the rehabilitation component was over 1001. 8.06 Human resources development focussed on specific areas of technical competence to the exclusion of the broader training function. More careful attention should be given to all the Borrower's staffing and training needs. This is being followed up by the Sixth Highway Project. 8.07 During preparation of this project, the Bank tested the HDM Model for the first time and, with the help of the Borrower, was able to identify its more sensitive parameters. The Model, in its subsequent versions, is recognized as a useful tool for the preparation of highway maintenance programs. 8.08 A particularly rewarding aspect has been the steady strengthening of institutions, already noted in the PPAM on the Second Highway Project. Tunisian officials have consistently demonstrated their ability to adopt planning and management techniques imported through technical assistance and to make all necessary adjustments so that foreign m_thods and solutions become appropriate for Tunisian requirements.  I~~~~~~~~PNI  1. INTRODUCTION 1.1 Sector Content Strategic Obiectives and Transport Sector Issues When it became independent in 1956, Tunisia took possession of a diversified transport system that was quite well adapted to the requirements of the time and was adequate to serve the country's economic development needs until the end of the 1960s. However, since 1970 the transport sector has become more important because of rapid economic growth, which has produced a very considerable increase in traffic. During execution of the Fourth Development Plan (1973-77), Tunisia succeeded to some small extent in increasing the capacity of its transport sector, through investment in infrastructure and vehicles. Although the road network adequately serves the various areas and supports rural and regional development, it is insufficiently maintained, and is becoming inadequate for the rapidly increasing volume of traffic, being very old and designed for a different type of tr4ffic, with low volume (in 1956, the average was 192 vehicles per day, heavy vehicles constituting 16.22). The increase to 623 vehicles per day since 1970, and the demands of economic and social development, necessitate the modernization and strengthening of the network, which has become inadequate because of its wioding curves, narrow roads, hogbacks, frequent blockages and poor pavement structures. In order to prevent the rapid deterioration of the network and enable it to play its full role, the government has decided to concentrate its efforts on modernizing and rehabilitating tne existing network, and on improving maintenance. 1.1.2 Consequently, the Fifth and Sixth Development Plans (1977-81 and 1982-86) identify the following major objectives for the transport sector: (i) more efficient utilization of the existing infrastructure, together with its maintenance and upgrading, if need be; (ii) promotion of rural development by upgrading feeder roads; (iii) the reorganization of SRTs (Regional Transport Companies). 1.2 The Role of the Bank and the Sector Loan Strategy Since 1964, the Bank has assisted the Tunisian government with eight transport projects. The First Highway Project (Loan 746-TUN, 1971, US$24 million), the Second Highway Project ('>an 1188-TUN, 1976, US$28 million) and the First  -2- Railway Project (Loan 606/Credit 150-TUN, 1964, US$17 million) focused on modernization and rehabilitation of primary and secondary roads, and renewal of track and purchase of railway equipment. Two smaller loans in the port subsector (Loan 380-TUN, 1964, US$7 million and Loan 573-TUN, 1969, US$8.5 million) provided financing for support facilities for port operation, mainly dredging, maintenance operations and cargo handling equipment. A Third Highway Project (Loan 1601-TUN, 1978, US$32 million) initiated an integrated program for improving about 1,100 km of rutal roads. The Fourth Highway Project (Loan 1841-TUN, 1979, US$36.5 million) is for the rehabilitation of about 800 km of paved roads. The Fifth Highway Project (Loan 3722-TUN, US$35.5 million) continues the Third Project, and is concerned with the upgrading of 1,200 km of rural roads. The Sixth Highway Project has been appraised, its objectives being to upgrade the maintenance management and rehabilitate 800 km of priority roads (Loan 2896-TUN, November 2, 1987, US$6.3 million). The Third Port Project (Loan 1797-TUN, of February 8, 1980, US$42.5 million) will assist the government in modernizing the ports of La Goulette and Sfax. The purpose of the Second Urban Transport Project (Loan 2429-TUN of August 28, 1984; US$33 million) is to upgrade transport in Greater Tunis. 2. THE ROAD NETWORK 2.1 The Tunisian road network consists of about 18,600 km of roads officially maintained by DPC, and more than 22,000 km of unclassified tracks of economic and social imporLance. The classified roads, for which DPC is responsible, are divided into three groups: (a) primary highways (routes de grand parcours, GP), defined as highways linking major centers in Tunisia with neighboring countries; (b) secondary highways (routes de moyenne communication, MC) for inter-regional service; (c) tertiary or local roads (routes vicinales d'etat, RVE). 2.2 Road standards are based on traffic volumes. Thus, in the more densely trafficked northeast of the country and on the coast, many secondary and tertiary roads have a bituminous surface, whereas primary roads in the south or northwest can have gravel or earth surfaces.  -3. The table below shows the distribution of types of surfacing by road classification: NETWORK OF ROADS MAINTAINED BY DPC (as of January 1986, in km) Surface TyPe Length Primary Secondary Tertiary Unclassified Paved 10,382 3,670 4,023 1,622 1,067 Gravel and improved earth 2,703 193 780 1,274 456 Unimproved earth 5,519 225 1,858 3,228 208 3. THE FOURTH HIGHWAY PROJECT 3.1 General Data 3.1.1 Road maintenance and rehabilitation lost momentum after comple-ion of the First Highway Project. As a result, by 1978 the network's service standard had fallen below the level required for satisfactory economic development, and the need to regain this lost momentum became urgent. The Fourth Highway Project was designed to achieve this objective. It was Identified in November 1978, and was prepared by DPC, with the assistance of Bank preparation missions in March and June 1979, and the appraisal mission in October 1979. 3.1.2 The overall objectives of the project are to improve the management of the Tunisian highway network and to achieve balanced development of the highway infrastructure. Specifically, the project is designed to: - meet the urgent backlog of road rehabilitation; - expand road maintenance activities and improve the efficiency of the maintenance organization; - strengthen the institutional framework through the introduction of improved planning techniques, information systems, and staff training; - progressively raise the recurrent budget for maintenance to an adequate level, and support a program of traffic management and road safety. 3.2 Appraisal Project appraisal followed an overall study of the road network. The following parameters were used in the breakdown: climatic regions, roughness, structure, traffic and surface condition.  -4- The Highway Design and Maintenance Standards Model (HDM), was adjusted tc match Tunisian experience and data and used to define optimum maintenance policy for the network and to evaluate the economic benefits of the rehabilitation program for the first year. The latter showed very high rates of return, indicating (if such indication were necessary) that the network is unsuitable for current traffic and that rehabilitation would considerably reduce vehicle operating costs, thus resulting in very large savings for the community. Implementation of an optimum policy requires improved utilization of available equipment and resources, together with the purchase of maintenance and inspection equipment. Consequently, the project consists of the following items: - rehabilitation of 800 km of the 1,700 km of roads in need of attention; - purchase of maintenance equipment and spare parts for increasing the efficiency of road maintenance, thus increasing the surface dressing program to 800 km per year; - development of a training unit responsible for planning and coordinating training; 3 financing of about 120 man-months of consulting services for establishing an organization-and-methods unit and preparing studies on the planning of road maintenance and the Highway Master Plan; - purchase of laboratory and workshop equipment for developing site tests and providing DPC with test equipment for improving maintenance planning; - provision of 25 (subsequently increased to 50) traffic counters for the statistics section, and 6,000 road signs for the road safety program. The project cost estimates and allocations were as follows: Financing in millions of US$ Millions Millions of Foreign Supplier of D US$ 1/ Exchange Bank Credit Rehabilitation 27.5 68.7 44.6S 30.6 0 Maintenance equipment 7.2 18 62.5b 2.1 9 Workshop, laboratory,signs, traffic counters 1.3 3.2 62.5% 2 0 Technical assistance and training 0.9 2.2 80% 1.8 0 TOTAL 36.9 92.1 50 36.5 9 _/ US$1 = D0.400.  -5- In the negotiations, the government made the following commitments: (i) Before issuing invitations to bid, DPC would submit for Bank approval the technical and economic studies on each of the road sections, in accordance with an agreed methodology, criteria and procedures. (ii) DPC would execute the maintenance program in accordance with a jointly agreed three-year action plan. (iii) DPC would submit its annual program of activities to the Bank on September 30 each year. (iv) Agreements on export credits relating to the maintenance equipment should be implemented on September 30, 1981. The Bank's Board approved a US$36.5 million loan at an 8.25S interest rate, for 17 years, including a four-year grace period. The Loan Agreement was signed on May 22, 1980, and the effective date was November 28, 1980. The closing date was to be September 30, 1984. Project implementation was the responsibility of DPC, with the participation of all the headquarters divisions. Supervision was carried out by the regional subdivisions, with the assistance of headquarters staff. 4. PROJECT EXECUTION 4.1 Start-Up and Changes 4.1.1 Initial schedule Construction works were initially scheduled to begin on April 1, 1921, and to be completed on September 30, 1983. The main dates in the original schedule were as follows: - Civil Works: Beginning End First Slice April 1, 1981 April 1, 1982 Second Slice January 1, 1982 January 1, 1983 Third Slice July 1, 1982 September 30, 1983 - Delivery of Equipment: From September 1, 1981 to October 1, 1982. - Technical Assistance: From September 1, 1980 to October 1, 1982.  -6- 4.1.2 Execution dates In the first year, works began in accordance with the agreed program (see Annexes 3, 4 and 5). However, time overruns affected execution. The main dates in the execution of program were as follows: - Civil Works (see Annex 3): Beginning End First Slice September 1981 March 1983 Second Slice May 1982 June 1987 Third Slice November 19813 Lot 11 December 1984 August 1986 GP6:MC33:MC36 From December 1984 Aigust 1988 to August 1985 - Delivery of Equipment (see Annex 2): From September 1981 to July 1984. - Technical Assistance (see Annex 3): From November 1980 to December 1987. The project began on schedule at the beginning of September 1981 with the works relating to the first four lots. Lot No. 2 was subsequently cofinanced under the Second Highway Project. The maintenance and laboratory equipment lots were initiated in June 1981. As regards operating equipment, changes were made consisting of the cancellation of two road sign lots, and the introduction of a lot comprising signalling equipment for 17 intersections. Equipment purchases were not as planned at appraisal, because of the change in financing. The Tunisian Government undertook to obtain a supplier credit for US$9 million by September 30, 1981. This credit was replaced by a US$6 million OPEC loan. Problems due to the lack of resources, together with the rise in the dollar (see Annex for chart), made it necessary for the closing date to be postponed on several occasions up to September 30, 1987, and the disbursement date to be postponed to June 30, 1988 so as to limit the unspent dollar balance.  -7f In addition, the repayment percentage for the rehabilitation component increased from 45% to 55S and then 63X, because of the increase in the foreign exchange share of the works. 4.2 The Rehabilitation Program 4.2.1 Content Rehabilitation of the paved network, originally consisting of three annual slices, affected 780 km of the 800 km originally planned, and consisted of four lots. The project applied to priority sections of the paved network, and consisted of coordinated work on the most trafficked main links, so as to adapt them to increasing volumes. Essentially, they consisted of road strengthening or widening, and the rehabilitation of shoulders and drainage. The benefits of these works result from a reduction in vehicle operating costs (due to improved surface conditions), road operating costs and supplementary maintenance expenditures. The HDM Model (pavement strengthening) was used to calculate these benefits. As regards widening, DPC prepared a methodology that enables it to use the same Model to assess the economic rate of return from widening. This employs a system of equivalences between widened and strengthened pavements. The internal rate of return based on original cost and traffic projections exceeded 15X. 4.2.2 Description The rehabilitation program consists of 17 lots distributed over 17 governorates. In addition to the rehabilitation of various highly trafficked urban or suburban sections, the following were the main items in the program: - GP 1: El Jem - Sfax: strengthening and widening to 7.40 mi; - GP 3: Cheylus - El Fahs: widening to 7 m; - GP 4: El Fahs - Siliana: in the first phase, widening to 7 m of the El Fahs - Robaa section; - GP 5: Medjez El Bab - Le Kef: widening to 7.40 m, and strengthening of the whole section; - GP 6: Medjez El Bab - Beja: widening to 7.40 m, and pavement strengthening, with modernization works; - GP 7: Tabarka - Algerian frontier: strengthening of entire section, which has considerably deteriorated;  -8- - GP 13: Menzel Chaker - Lassouada: widening to 7 m; - GP 19: Tataouine - Ramada: widening to 7 m; - GP 3: Hajeb El Ayoun-GP 13: widening over 30 km; - GP 3: Tozeur - Nefta: widening to 7 m over 23 km; - MC 116: Houmet Souk - Ajim: strengthening and widening to 7 m over 19.6 km; - MC 117: El Kantara - Houmet Souk: strengthening and widening to 6 m over 25.20 Kcm; - MC 52: B6ja - Nefza: Although the original project provided only for strengthening, the road was also widened to 6 n in view of the traffic that will be generated by the development of Mentazah Tabarka; - MC 82: Chebba - Sfax: widening to 6 m or 7.40 m over 58 km; MC 33: Port of La Goulette: modernization over 9 km, together with construction of two engineering structures and doubling of width over 7 km. It should be noted that several sections included in the Fourth Highway Project are continuations of sections improved under the Second Road Project. The above indicates the importance attributed in the Fourth Highway Project to works on the important links among the various major economic centers. These works are particularly necessary because these links are the oldest in existence and carry a large volume of passenger car and heavy vehicle traffic. To be noted among these sections are GP 1, GP 3, GP 4, GP 5 and GP 6. 4.2.3 Execution schedule and time overruns (see Annex 1) According to the original schedule, the Fourth Highway Project was scheduled to be completed on September 30, 1983 and the closing date of the Loan was September 30, 1984. The closing date was finally postponed until September 30, 1987. As of March 30, 1988, certain items had not yet been provisionally accepted. If all lots were assumed to be accepted by August 1988, the time overrun in the execution of the works component would be more than four years. These overruns are essentially due to a lack of ministry resources resulting from the increase in the cost of works compared with the original appraisal and the large unspent dollar credit balance resulting from changes  -9- in the exchange rate. Thus, in 1983, the allocations in dinars earmarked at the beginning of the project for financing the rehabilitation works had been totally committed, whereas much of the loan had yet to be committed. In addition to the above two reasons, the 1984-85 economic crisis in the country-characterized by a lack of liquidity and the problems faced by enterprises in making payments-aggravated these difficulties and hindered progress, even leading to the bankruptcy of several road-construction enterprises (STC and SABAG). In order to reduce the unspent balance, the World Bank agreed to the financing of a fourth set of lots (MC 36, MC 33 and GP 6) for which Tunisian credit was already available. The works comprising this fourth lot did not begin until December 1984. The main overruns in works execution resulting from problems with enterprises or arising on site are as follows: - Lot 5: The overrun was due to the late start of work. - Lots 8 and 9: Both were delayed by more than two years, thus doubling the execution period. It should be noted that the enterprise hired for these two lots was originally prequalified for a single lot. In order to enable the enterprise to compensate for a loss of income due to an error in the bid (amounting to more than D 200,000), the government decided exceptionally to grant the enterprise the two lots (8 and 9) for which it was the lowest bidder. - Lot 13: The works were considerably delayed because of the problews of STC, which eventually went bankrupt. - Lot 16: The overrun was due to the late start-up of works. - GP 6: Because the highway rapidly deteriorated between the presentation of the study and execution, the strengthening structure had to be reexamined and the contracts committee, had to approve the changes. Consequently, the completion of execution was delayed. - MC 33: The overrun resulted from the contractors' works, and from construction problems arising in the structure over the railway. Certain lots, such as Nos. 3, 4, 6, 7 and 12 were executed on schedule. 4.2.4 Cost of works (see Annex 6) Cost overruns in the works component resulted from the following two causes: - the difference between actual cost and the contract amount;  -10- the difference between actual cost and the cost calculated at appraisal. Annex 6 shows-by individual lot-initial costs, price changes and additional works, and also shows these as a percentage of the final total. Cost overruns expressed in dinars vary considerably from one lot to another, ranging from 0% for lot 6 to more than 501 for lots 8, 9 and 11, the average being 271, with the following breakdown: - 132, resulting from orice changes; - 141, resulting from increases in the amount of works. Overall, cost overruns are as follows: - Total procurement costs D 34.300 million - Total cost of additional works D 4.816 million - Total price changes D 4.403 million - Actual total cost D 43.519 million The reasons for the cost overruns, as compared with the original contracts, are as follows: - Contract changes in prices, which were aggravated by the following factors: - Contract periods were quite long (because of the need to maintain traffic flows during the works; i.e. lots 5, 8 and 9). - All the lots were awarded to Tunisian enterprises, so that all costs were subject to fluctuations. - Amendments to the original contracts, such as the widening of MC 52, which had not originally been planned. - Rehabilitation or reconstruction of structures on GP 5 (lot 9), which were too narrow for the road. - Development of the Medjez El Bab and Le Kef crossing, which was not originally included in lots 8 and 9. - The reexamination of the pavement structure of GP 6, which was justified by the rapid and unexpected aging of the road, deflection having increased by more than 600S1 - Specific constraints relating to the proximity of housing to the pavement, making it necessary to reduce the preexisting surfacing (lot 4).  -11- Total project cost-expressed in dinars-exceeded the appraisal projections by 58Z (see Table 2). This overrun is justified by the age of the roads that were rehabilitated and their advanced state of disrepair, which made it necessary to increase strengthening beyond the degree deemed necessary in the appraisal, and make greater use of bituminous and slag aggregate. Price increases resulting from imports (i.e. equipment purchases and operating costs) produced successive increases in the costs per km of works. so that the cost of comparable activities was 46S higher in 1982 than in 1981. Other factors were the modernization works for GP 6, MC 33 and MC 36, which involved (in the case of GP 6) the construction of new roads with large-scale earthworks, together with the construction of engineering structures for MC 33. These works were not included in the appraisal. 4.3 Execution of Part II of the Loan 4.3.1 Purchase of equipment The loan amounts originally intended for purchasing maintenance equipment, workshop and laboratory equipment, road signs and traffic counters total US$13.1 million, with the following breakdown: Supplier World Bank Credits Total (in millions of US$) Road maintenance equipment 2.1 9.0 11.1 Workshop and laboratory equipment, road signs and traffic counters 2.0 2.0 4.1 9.0 13.1 The US$9 million in supplier credits were subsequently replaced by US$6 million from the OPEC Fund. Available financing totaled US$10.1 million: Supplier World Bank Credits Total (in millions of US$) Road maintenance equipment 2.1 6.0 8.1 Workshop and laboratory equipment, road signs and traffic counters .0 2.0 4.1 6.0 10.1  -12- Consequently, the modifications in the equipment procurement program were the result of financial changes. 4.3.2 Road maintenance equipment (see Annexes 4 and 7) (a) Procurement program: After these changes, DPC planned to purchase the following items of road maintenance equipment (see 2A): six tankers (20 t), 29 hand-operated sprayers, 8 brooms, 15 chip spreaders, 1 tow truck, 1 forklift truck, 70 graders, 1 truck equipped with a boom and aerial platform, 177 trucks, 6 compressors, 1 "Junior" marker (traceuse), 12 markers, 108 tanks (3,000l1), 2 road tractors, 35 light trucks + 32 light trucks, 20 all-terrain vehicles, 10 transport vehicles, 1 crushing unit, 1 crane (16 t). (b) Changes: The following changes were subsequently made to these equipment purchases: - cancellation of the crushing unit and the 20 all-terrain vehicles; - purchase of two additional road tractors and six trailers; - a reduction in the number of tanks from 108 to 40. (c) Contract deadlines (see Annex 2): The main contracts for these components were made between the end of 1981 and the beginning of 1982, i.e. approximately within the periods planned at appraisal. The tankers, semitrailers and trucks were delivered more than one year later than the dates stipulated in the contracts and planned at appraisal. (d) Cost of maintenance equipment (see Annex 7): Initially, the contracts for this component totaled D 6,042,736.340. The actual cost was D 6,352,063.309. The reason for the discrepancy was the purchase of three semitrailers and two additional tractors, together with changes in customs duties (+ D 48,804.868) and in the exchange rate (+ D 71,879.269). In addition, nine contracts for the purchase of spare parts were made between 1983 and 1985, totaling D 881,150.  -13- 4.3.3 Laboratory equipment, traffic counters and road signs (see Annexes 4 and 5) (a) Procurement program and changes: - Laboratory equipment: Financing was provided for the purchase of testing equipment for the regional laboratories at Sousse, Sfax, Bizerte, Le Kef and Gabes and for the two new laboratories at Gafsa and Kairouan. The equipment covered require- ments for carrying out geotechnical, bitumen, concrete and soils testing. The Fourth Highway Project provided for the strengthening of the main laboratory in Tunis and the establishment of a pavement testing unit with modern equipment for studying and monitoring pavement behavior. Four contracts for the supply of this equipment were made at the end of 1981. They totaled D 713,000. The equipment was delivered in July 1982, i.e. within the period planned at appraisal. - Traffic counters: To enhance the capability for analysis, design and refinement of the maintenance programs, the statistics section was provided with 50 traffic counters, together with spare parts. For the sake of standardization, a negotiated contract was made with Socikt6 Tunisienne d'Etudes et d'Assistance, with the approval of the World Bank. - Traffic Lights: In March 1981, the World Bank agreed to a request from DPC to finance installations on 12 intersections, this number subsequently being increased to 17. The railway safety program was to have received the following three sets of signs: - Lot 1: 4,200 police signs and 4,000 danger signs; - Lot 2: 2,200 advance signs and 800 m of place names; - Lot 3: 2,000 m of direction signs (6,000 signs). A contract was issued only in the case of lot 3, and no further action was taken on the other two. (b) Contract Dates and Changes (see Annex 4): Except as regards the signs, contracts were issued at the end of 1981 and the beginning of 1982, i.e. within the periods planned at appraisal.  -14- The final delivery dates were also on schedule. The contract for signs was not awarded until one year after bid opening. Similarly, import problems delayed the delivery of this equipment, and even impacted the quantity delivered (i.e. 2,073 units instead of 6,000). (c) Costs (see Annex 7): The total cost of equipment, statistical and operating contracts was as follows: Laboratory D 712,942.357 Operations and Statistics D 499,852.979 Total D 1,212,795.336 Actual costs were4 as follows: Laboratory D 713,812.707 + Customs D 810,363.707 Operations and Statistics ID 480,426.676 TOTAL D 1,290,790.383 The reasons for the difference between contract costs and actual costs is the change in the volume of works (as in the case of the traffic lights and the contract for signs), together with the increase in customs duties. 4.4 Technical Assistance (see Annexes 5 and 8) 4.4.1 Original program and changes According to the appraisal, this component was to provide for 120 rn-mnths of consulting services to assist in: - establishing an organization and methods unit for project monitoring and evaluation, including road inventory and accident data processing; - preparing studies necessary for the planning of road maintenance, including a highway master plan; - preparing economic studies and engineering for the second and third years' rehabilitation programs; - carrying out a training program for DPC staff in Tunisia and abroad; - preparing engineering for about 1,000 km of rural roads.  -15- Subsequently, this component was considerably amended to consist of the following studies: - the highway master plan; - the road maintenance study; - the study of the transport coordinating unit. Establishment of an organization and methods unit was postponed when the maintenance study began, because the latter was to result in possible changes in maintenance and organization. The engineering studies for the second and third year's rehabilitation programs were prepared on force account, and not included in the Fourth Highway Project. 4.4.2 Preparation of studies (a) Study of the Highway Master Plan The Highway Master Plan was prepared in light of the existing network structure, the needs arising from long-term transport demand, regional planning programs, and the planning of highway investment. The study began on October 17, 1980 and was completed on June 2, 1981, being executed by SETEC under a negotiated contract. (b) The Road Maintenance Study This began on November 15, 1982 and was completed on March 15, 1985. It was prepared by a group of French and Tunisian consulting firms (SETEC, SOTUETEC and SOTINFOR), and had the following objectives: - examination and diagnostic study of the existing road network; - examination of road maintenance methods and methods of new pavement design and strengthening; - analysis of road maintenance resources and organization; - examination of the proportion of the budget allocated to maintenance. Changes became necessary during the course of the work, particularly In the case of the study of pavement behavior and the modifications in the HDM surface deterioration model. This latter aspect of the study was disappointing because of the lack of information and the relatively short duration of the study. Nevertheless, a pavement monitoring methodology was prepared. In view of the serious problem of equipment, this study examined that aspect in greater depth, emphasizing the need for:  -16- - training road maintenance personnel; - improving the management of spare parts, which is initially manual, but will later be computerized; - improving spare parts procurement by selecting suppliers more judiciously and paying them within a reasonable period; - establishing a system of five levels for equipment repair workshops. (c) The Transport Planning Unit In April 1982, the World Bank approved an amendment to the Loan to include the financing of consulting services to assist the newly established Transport Planning Unit, which has the following responsibilities: - improving transport management; - systematically examining and planning inv%estment in the sector. At the end of 1983, a consulting firm was appointed, although the unit was not provided with a sufficiently large staff. In July 1986, when changes were made in the upper levels of the Ministry of Transport, the consultants completed their work and the Unit (which, since its establishment in January 1985, had consisted of three engineers) was attached to the Directorate of Planning. 4.5 Training At the beginning of 1981, DPC appointed a person to be responsible for planning and coordinating training, particularly for engineers, technicians, works supervisors (conducteurs de travaux), and site supervisors. Fellowships were provided to enable engineers to take courses lasting one or two months, and seminars were organized in Tunisia. The personnel consists of 5,000 persons, and the training provided (in man-months) can be summarized as follows: - Engineers: 50 - Technicians: 3 - Foremen: 12.5 - Brigade Chiefs: 50 - Equipment Chiefs: 10.5 - Mechanics: 19  -17- - Drivers and Workers (refresher or initial courses in the operation of graders): 106 Consequently, training has been intermittent, falling short of needs, particularly in the field of mechanical equipment and network and site management, which is very largely the responsibility of the regional services. Training has suffered from the following shortcomings, among others: - a lack of annual funding allocations; - a failure to define the duties of the personnel; - a lack of supervisory and monitoring bodies representing the government agencies, for assessing both needs and the effects of training. Expenditures under the Fourth Highway Project total about D 35,000, as follows: - ESTP Agreement: D 29,868.75 - Road Safety Seminar: D 4,900 4.6 Inprovements in Maintenance (Annexes 9 and 10) DPC has prepared a road maintenance improvement program consisting mainly of an increase in the volume of routine and periodic maintenance and the application of a new loi cadre to strengthen maintenance teams. 4.6.1 Increases in the volume of maintenance (Annex 9) A comparison of the volumes of maintenance activities for 1978-80 and 1981-86 shows the following annual increases: - 67 km of surface dressing; - 58 km of shoulder regraveling; - 42 km of treatment with asphaltic concrete; - 4,400 km of shoulder reprofiling; - 2,500 km of earth-road reprofiling. There were the following reductions in the regraveling of earth roads and patching: - Patching: 175,000 mz; - Regraveling of earth roads: 243 km.  -18- Although the reduction in patching is logical and economical, in light of the increase in surface dressing, this is much less true of earth road regraveling. Nevertheless, to offset the impact of this reduction, efforts were devoted to reprofiling these roads in order to improve their serviceability. On average, the frequency of surface dressing on paved government roads was once every 19 years in 1979, once every 14 years in 1980, and once every 15 years in 1987; i.e. there is still some way to go before the target of once every 10 years is achieved. This target was confirmed by the Road Maintenance Study, and the rate of return to be expected has been calculated as over 302. 4.6.2 Maintenance budget Annex 10 sunmmarizes the financial components of Article 60 of DPC, in current dinars. The following table shows maintenance costs governed by Article 60, at 1987 prices and in light of price change indicators: Sz 9 I8 m2 12m I S nu A lsa6 1282 I8 Amount 2.37 2.58 2.92 3.34 4.2 4.8 5.6 6.11 4.91 5.05 8.0 (in millions of current dinars) Price Change 278 264 242 224 172 158 155 151 143 100 92 Index Amount 6.59 6.81 7.08 7.48 7.22 7.58 8.68 9.24 7.03 5.06 7.36 (in millions of dinars at 1987 prices) The budget has, therefore, increased from an average of D 6.83 million from 1978 to 1980, to D 7.46 million from 1981 to 1988, an increase of 9b. Adjusted to actual length in km, this increase is only 2S, and this figure does not take into account the complexity of certain structures built during the 1981-88 period and requiring very high standards of maintenance and very large allocations (e.g. the Tunis-Rammamet highway and the Bizerte Bridge). 4.6.3 Objectives and performance: 1981, 1982 and 1983 A comparison of project objectives and performance from 1981 to 1983 shows that the volume of principal maintenance activities actually performed was satisfactory, being 90S or more of the planned total. A very low performance rate (about 501) was evident only in the regraveling of shoulders and earth roads, network rehabilitation and pavement markings.  -19- The following table shows a breakdown by operations of project objectives and performance: COMPARISON OF ACTION PLAN AND WORKS PROGRAM (Averae Per Year) Objective Actual S of Objective Unit 1981-83 1981-83 Routine laintenance Patching m2 435,000 480,000 110 Shoulder leveling lum 9,700 12,000 124 and cleaning out of ditches Rehabilitation of U 2,000 4,227 211 signing Installation of U 1,500 4,100 173 nw signs Pavement marking km 600 260 43 Levling of km 9,700 9,700 100 unsurfaced roads Periodic Maintenance Surface dressing, on km 660 600 91 force account Overlays km 50 50 100 Regraveling of km 290 shoulders 700 Regraveling of km 11.3 43 earth roads Rehabilitation On force account km 80 72 90 By enterprise km 250 780 over 7 years 44 (110 km) 4.6.4 Needs and performance: 1985. 1986 and 1987 From 1985 to 87, the percentage of works executed compared with real needs (determined in 1984 by the Road Maintenance Study) declined. In addition to shoulder and earth road regraveling and pavement marking, surface dressing overlaps and shoulder reprofiling declined. The volume of work actually executed was less than 702 of needs in each of these categories.  -20- For this period, the following percentages of needs were met: Surface dressing 651 Shoulder regraveling 70S Regraveling of earth roads 4S Asphaltic concrete overlays 661 Reprofiling of shoulders 462 Reprofiling of earth roads 94% Pavement markings 36t New and maintained signs 831 Special operations 561 4.6.5 Organization of maintenance In 1982, a law defining the various types of maintenance team was established, as part of the decentralization process. In 1988, the maintenance services (divided among 23 governorates) consisted of the following: - 126 reprofiling teams; - 76 patching teams; - 23 engineering structures teams; - 20 signs teams; - 9 sand-clearing and snow-clearing teams; - 23 transport groups; - 23 earthworks groups; - 23 asphalting groups. Periodic maintenance works are planned annually by the Directorate of Highway Maintenance and Operation, in response to proposals from Regional Services, which are also responsible for supervising and organizing the planned works and drawing up quarterly reports. The principal manuals used for planning rehabilitation and surface dressing, together with calculating the cost of the various maintenance  -21- activities, are held by the districts (arrondissements) (i.e. the atlas and the cost accounting procedures). This organization is very different from that in existence at appraisal. In 1980, routine maintenance, together with the maintenance of engineering structures and signs, was executed by regional teams that were independent of the Subdivisions. Routine maintenance was executed by national groups supervised by the Directorate of Highway Maintenance and Operation. These groups operated in rotation through the various Subdivisions, in accordance with plans and programs prepared jointly each year by the Directorate of Maintenance and the Regional Services. There were the foilowing eleven groups: - 6 national rolling and asphalting groups; - 3 earthworks groups (excluding regraveling); - 2 groups specializing in earthworks and the laying of well graded slag aggregate. 5. PROCUREMENT Civil works procurement posed no major problems. In 1979, a general prequalification was carried out, and has been periodically updated. Contracts were been grouped into lots with totals exceeding US$2 million. The provisions and procedures specified in the Loan Agreement were fulfilled. The equipment component was the subject of international competitive bidding, except for the contract for 50 traffic counters. Standardization was necessary in this case, so a contract was negotiated with Societe Le Mat6riel, once Bank approval had been obtained. Consulting service contracts were issued after prequalification, and on the basis of limited bidding. An exception was the contract for the Highway Master Plan, which was negotiated. Franco-Tunisian groups were hired for all consulting services, except the study on transport coordination. For this, a French consultant was hired that had no connection with any Tunisian firm. 6.0 PROJECT COST AND DISBURSEMENTS (ANNEXES NO. 1 AND NO. 2) 6.1 Costs Project costs changed considerably because of changes in the dollar exchange rate. The value of the dollar increased from D 0.499 at appraisal to D 0.890 in March 1985, and stabilized at about D 0.820 in 1988.  -22- In dinars, actual project cost increased by more than 452 compared with the original calculations, in spite of the reduction in equipment procurement. In relation to the dollar, the estimated reduction in these costs, taking into account the amount committed, the disbursements made, and the eventual exchange rate (US$1 = D 0.820), is more than 262. The reason is the considerable increase in the value of the dollar, which averaged D 0.20 during project execution, i.e. about double the value at appraisal (D 0.400). ESTIMATED AND ACTUAL PROJECT COSTS (In millions) Cost: Actual/Appraisal Actual Cost Appraisal Dinars US$ Dinars USV Dinars US$ Civil Works 43.52 55.93 27.5 68.7 158 81 Equipment: - Maintenance 6.35 8.20 7.2 18 88 46 - Workshop, laboratory 2.17 2.76 1.3 3.2 167 86 Assistance and 1.30 1.44 Training 0.04 0.04 0.9 2.2 148 67 TOTAL 53.4 68.4 36.9 92.1 145 74 6.2 Disbursements The annexed chart shows actual disbursements in millions of US$, and also shows what disbursements would have been if the exchange rate had remained constant at the appraisal level. At the closing date (September 30, 1984), the loan would have totaled US$12.994 million, instead of US$36.5 million. As of February 26, 1988, disbursements totaled US$30.25 million. The balance (US$5.25 million) could not have been completely disbursed by June 30, 1988, because commitments totaled US$68.4 million, corresponding to a financing of US$39.16 million (including US$3.622 million in equipment financing provided by OPEC. The sum theoretically available from the Bank was US$35.54 million, so that, in theory, the unspent balance was US$0.96 million. However, because of the difficulty of utilizing the remaining US$6.25 million before June 30, 1988, DPC decided to cancel US$2.5 million, and applied to do this on April 12, 1988. In addition, if the exchange rate had remained constant, the original loan amount (U$36.5 million) would not have covered total disbursements  -23- (US$48.958), assuming that the repayment rates had remained those originally planned at appraisal. At the same exchange rate, and with the original repayment rates, the actual repayment rate is 652 of that originally planned. The reason for this is the delay in project execution. 7. EVALUATION OF THE FOURTH HIGHWAY PROJECT 7.1 Introduction The purpose of this section is to present the data and results of the evaluation of the programs for strengthening and improving the maintenance of the paved network. 7.2 The Strengthening Program 7.2.1 Content The network strengthening program was prepared between 1981 and 1985. It consisted of four annual slices, execution being distributed over the period from 1982 to the end of 1987, except for a small residual amount which is expected to be completed in August 1988. The original program was to be applied to 782.5 km with the following breakdown, based on type of strengthening operation: Operation km b 1. Strengthening and Widening 385.3 49.2 2. Pavement Widening 259.5 33.2 3. Strengthening 87.5 11.2 4. Overlays 50.2 6.4 TOTAL 782.5 100.0 7.2.2 Economic evaluation Economic appraisal of this program was performed in 1981/82 and 1983, using the HDM Model (version 1), except for the fourth slice (MC37, GP6 and MC33). Most of the data used were calculated at 1981 factor cost. In particular, traffic per trunk road was calculated using the results of the 1977 census and the growth rates from 1972 to 1977. However, because of the wide differences among the results obtained, the annual growth rates assign 'd were 102 for the 1977-86 period, and 8S after 1986.  -24- The results from the EDM Model showed very high internal rates of return for all the planned operations, justifying the adoption and implementation of the program. The HDM Model was not used in appraising the fourth slice. This was studied separately, and included the following items: - MC26, as part of the transfer of heavy transport: doubling of lanes to reduce traffic generated by the new development center; - MC33, in relation to the construction of the new port of Rad&s: widening and strengthening from GP1 to the new port, in order to deal with the traffic diverted and generated by the port; - GP6, as part of a national program for improving the northwestern network (i.e. correction of design, and the widening and strengthening of certain sections). These three projects were the subject of a feasibility study, prepared in association with their related projects. Furthermore, they constitute only 62 of the total length of roads affected by the program as executed. Consequently, they were not included in this evaluation. 7.3 Hgyotheses and Basic Data for the Evaluation 7.3.1 Base year 1987 is the base year for all data used in the evaluation of the strengthening project. In fact, the most recent information on traffic and vehicle operating costs is for 1987. This also applies to the unit costs of road network maintenance. 7.3.2 Traffic (see Annex 11) Data from the 1982 census and the first six months of 1987 have enabled calculations to be made of traffic per section and growth rates, by type of vehicle, during the 1977-87 period. As regards future traffic trends, the periods 1977-82 and 1982-87 have been taken separately, and the minimum annual rates for 1977-87, 1977-82 and 1982-87 have been used. In contrast to the 1981/82 assessment, we have adopted rates that differ according to the type of vehicle in question. Annex 1 presents, by section, average daily traffic by type of vehicle in 1987, and the growth rate calculated. We have also indicated traffic rates taken into account for the 1981-82 assessment. Comparatively large discrepancies appear, essentially because of the inadequacy and unreliability of the data available at the time. Consequently, the use of estimates is the reason for these discr;jancies. 7.3.3 Vehicle operatina costs (see Annex 12) Vehicle operating costs have been adjusted to 1987 levels using the results of surveys of franchise holders, insurance companies, and suppliers of spare parts and tires.  r2S- The following five types of vehicles were used, based on knowledge of those actually in service: 1. Passenger car: of the 305 or R9 type (6 HP). 2. Light truck: of the Peugeot 504 or 404 type. 3. Truck (5 ton): OM - Fiat. 4. Truck (15 ton or larger): Berliet or Fiat type. 5. Bus: intercity type. The unit costs collected for use in the HDM Model included all taxes, and were then converted to costs net of tax using the tax rates applied in 1987. Annex 12 presents the costs net of tax, together with those for 1981. The differences in these costs should be noted: +301 per year for passenger cars, and +21' for heavy vehicles; +12% per year for tires, and +11% for rf utenance labor costs; +17f for the wages of driving crews; +26Z/291 per year for fuel. These increases, particularly for vehicle purchase and fuel, are excessively high compared with the overall movement of the price index in Tunisia, which was 9S per year from 1981 to 1987. The 1986 devaluation of the dinar partly explains this large increase in vehicle operating costs. 7.3.4 Maintenance costs (see Annex 13) The costs net of tax of the various maintenance activities were calculated from DPC's cost accounting for 1986, adjusted to a rate of 92 for 1987 (in line with changes in public works costs). Annex 13 shows the unit costs calculated. 7.3.5 Maintenance standards The program was evaluated in light of the following actual maintenance standards: Patching: 30% of the cracked surfaced, not exceeding 100 ma/km/year. Surface dressing: once every 15 years once the area affected by cracks is greater than 40X.  -26- Routine maintenance: once every year (i.e. drainage, reprof liog, etc.). 7.4 Investment (see Annex 15) Works were performed between 1982 and 1987, depending on the individual section. In order to evaluate these projects, we determined a schedule of actual expenditures for each section from 1982 to 1987, and adjusted them to 1987 factor costs (i.e. net of tax). The inflators for the public works sector from 1982 to 1987 were used in making these adjustments. In general, inflation in this period was about 9% per year (Source: National Statistics Institute). Annex 3 shows the costs of each section as calculated in 1981 and adjusted to 1987 dinars, together with actual expenditures at 1987 factor cost. : The differences between the costs calculated in 1981 and actual expenditures are acceptable, because those calculations were based on unit coits and volumes of works which were subsequently considerably changed (as regards project content, nature of works, etc.). 7.5 Results (see Annex 17) The results of the economic evaluation of the program were provided by the EDM Model (version No. 2). The table shows internal rates of return calculated before and after the program, by section and by lot. tn general, the economic benefits (IRR) are only marginally different from the 1981/1982 calculations, except for the GP7 section, which has a comparatively low internal rate of return (9.21), essentially due to its low traffic volume and the increase in costs. The net benefit (adjustment rate a 201) is D 456 million, and investment totals D 34.25 million at 1987 prices (see Annex 17). The overall internsl rate of return is above 1002. 8.0 ECONOMIC EVALUATION OF IMPROVED MAINTENANCE 8.1 Increases in the Volume of Maintenance Activities (see Annex 9) Since 1980, the volume of the main road maintenance activities has increased in comparison with preceding years. The following increases took place from 1981 to 1983: - more than 270 km of surface dressing; - 119 km of overlays; - 16,340 km of shoulder reproflilngg - 413 km of shoulder regraveling. In addition, the increases in surface dressing and overlays allowed expenditure on patching to be reduced.  -27- 8.2 Network Condition (see Annex 21) Assessments of network condition with or without improved routine and periodic maintenance are based on the following factors: - In 1983, the calculations contained in the Road Maintenance Study, adjusted to allow for the fact that the bump integrator was not calibrated, and roughness are overestimated by more than 30X. - In 1987, the hypothesis that, without rehabilitation and without maintenance strengthening, network condition would change from good to average or from average to bad over a period of 15 years. - Because the results of the 1987 traffic count were unavailable, the network condition has been assessed in accordance with 1982 traffic. In the economic calculations, traffic increases are assumed to be 102 per year. Average roughness assumed for the three surface conditions are as follows: - 2,600 mm/km, for good pavement; - 3,300 mm/km, for a fair pavement; - 4,100 mm/km, for bad pavement. 8.3 Vehicle Operating Costs The additional costs for each of these degrees of roughness, compared to a refercice quality of 2,000 mm/km, are presented in Annex 14. The following are the costs per 1000 km, in 1987 dinars: Roughness Roughness Roughness 2,600 mw/km 3,300mm/km 4100 mm/km D 24.800 D 42.400 D 63.000 8.4 Calculation of Benefits A comparison of the two scenarios (with or without improvements in maintenance) shows that the benefits due to reduced vehicle operating costs (at 1987 prices) are D 4.54 million in 1983 and D 16.40 million in 1987.  -28- The following table shows annual benefits: Benefits Due To Reduced Vehicle Operating Costs Year (in millions of 1987 Dinars) 1981 1.25 1982 2.75 1983 4.54 1984 6.82 1985 9.52 1986 12.68 1987 16.40 1988 20.74 1989 25.78 1990 31.62 8.5 Increases in investment Annex 1 shows the additional investment in improved maintenance, in millions of 1987 dinars. The unit costs of surface dressing, overlays, regraveling and reprofiling of shoulders are based on the net-of-tax costs of these activities in 1987, adjusted for additional overheads. Costs of equipment, studies and training have been distributed annually, in accordance with procurement dates, and adjusted by 9% per year. Net-of-tax equipment costs differ by between 622 and 831 from the total amount, and the average level is 78X. Costs of studies and training are 952 of the total. 8.6 Benefits Due to Improved Road Maintenance (see Annex 19) The benefits--discounted by 10% and 20%-are D 49.70 million and D 25.83 million respectively, and the economic rate of return is more than 70%. A sensitivity analysis consisting of a 402 reduction in vehicle' operating costs indicates that benefits nevertheless remain large, being D*20.77 million and D 8.36 million (10% and 20% respectively). The economic rate of return is still over 30% (see Annex 20). 9. INSTITUTION BUILDING AND STRENGTHENING OF OPERATIONS 9.1 Because of weaknesses in the organizational resources of the Ministry of Transport and Communications affecting the application of transport studies, a transport planning unit was established, initially attached to the Minister's Office, and subsequently to the Directorate of Planning. The small number of staff members, although appointed to the unit well after the appointment of the organization agency, nevertheless participated in the studies, thus receiving on-the-job training.  -29- 9.2 During the project, progress was made in institution building through the establishment in 1981 of a training unit. This consisted of a small staff and began to operate informally, planning and organizing short courses in Tunisia and abroad. The persistence and dedication of its chief enabled it to continue operating and finally be officially included in the new organization structure. As stated previously, the lack of an organization structure and annual appropriations considerably hindered the efficient progress of training. 9.3 The Road Maintenance Study, in which many DPC personnel were actively involved, revitalized DPC by updating important working documentation, such as the roads inventory, computerization of the network and equipment, maintenance guides and manuals, etc. Recommendations were submitted in this study, particularly with respect to equipment and its maintenance, and training. 9.4 The organization structure. for DPC and its components was prepared in 1972/73 as a result of recommendations presented as part of the Road Maintenance Study prepared SCET-BCEOM. It has been reorganized over recent years, with new and increased powers being vested in the regions, and without any precise definition of the division of responsibilities between the regional directorates and the central authority. The reorganization has disrupted operations because of the lack of adequate qualified supervisory staff, and the lack of equipment resources. The Fourth Highway Project has only partly solved these problems. 9.5 DPC has a detailed program for the gathering of traffic accident data, and has improved its processing and analysis of this information. In contrast, there has been no progress on the road safety program, which consists of the installation and replacement of road signs. 9.6 Improvements were made in the volume of road maintenance activities to be executed. Unfortunately, budget appropriations did not keep pace with the increase in needs, and road maintenance objectives for the main activities were not entirely achieved. It should also be noted that an item specifically for earth road maintenance was included in the government budget, thus providing an official framework for such maintenance until the new road classification is issued. 10. CONCLUSION 10.1 Overall, the project achieved its objectives, and works were performed on 782 km of the 800 km initially planned. Although costs exceeded those calculated at appraisal, they increased in almost the same proportion as vehicle operating costs, and the economic rate of return for the rehabilitation component is 3452. 10.2 Road maintenance has been strengthened over recent years (in comparison with the 1978-80 period) and this justifies the additional  -30- investment made under the project. The resulting benefits were calculated by standard methods, without using the HDM Model (which assumes that all ancillary works are performed properly). At a rate of 20S, the discounted benefit is D 25.8 million (at 1987 prices), and the rate of return is 72.432. 10.3 Problems in project execution were due essentially to the lack of budget appropriations for financing the local contribution, and to the increase in the dollar exchange rate, which considerably increased the costs of works and procurement. 10.4 The studies, on-the-job training, and installation of the HDM program enhanced the resources available for transport planning and rehabilitation and maintenance works planning.  - 31 - ANNEXES [TN: IN THE FOLLQWING TABLES, COMAS REPRESENT DECIMAL POINTS, AND VICE VERSA, WHILE DATES ARE GIVEN IN THE FOLLOWING ORDER: DAY, MONTH, YEAR]  - 33 - ANNEX 1 PAYMENTS AND CONKITMENTS  - 35 - ANNEX 1 PAYMETS DISBURSED US$ Equivalent Categorv Amount D (thousands)* (tho 8sad)* i ( 1 ) i 3R,469 i 1,t59 27,248i ! ( 2 ) ! 6,319 OPEC 2, 970 3,622 , ; IBRD 1,011 1wSI6 (38) 0,763 0,487 C,658 *~~~~ 4 (SA) i 0,029 0,029 i ,044 aininain $ inininaain ~-- - - - - -______ -- -- SUBTOTAL i 2332 90e & !TOTAL : t 46,556 ; 26,282 33,870 C015I0~T