Turkey - Highway Rehabilitation Project

Document of The World Bank FOR OFFlCIAL USE ONLY Report No. 8433 PROJECT COMPLETION REPORT TURKEY HIGHWAY REHABILITATION PROJECT (LOAN 2137-TU) MARCH 16, 1990 Infrastructure Operations Division Country Department I Europe, Middle East and North Africa Region This document has a resticted distribution and may be used by recipients only in the performance of their offical duties. Its contents way not otherwise be disclosed without World Bank authorization.  CURRENCY EQUIVALENTS (Appraisal) 6/77 US$ 1 - TL 18 6/83. USS 1 TL 209 6/79 US$ 1 - TL 31 6/85 US$ 1 - TL 576 6/81 US$ 1 m TL 111 6/87 US$ 1 TL 852 ABBREVIATIONS GAP - Southeastern Anatolia Regional Development Project HRP - Highway Rehabilitation Project KGM - Directorate General of Highways MOF - Ministry of Finance MPW - Ministry of Public Works OED - Operation Evaluation Department PMS - Pavement Management System PT - Project Control and Coordinating Team PIR = Public Investment Review PCR - Project Completion Report SPO State Planning Office SHP - Second Highway Project SAR Staff Appraisal Reiort SAL - Structural Adjustment Loan TRRL - Transportation and Road Research Laboratory TTH - Trans Turkey Highway UNDP - United Nations Development Program FISCAL YEAR OF BORROWER January 31 - December 31, 1989  ]FOR OFV1MAL US ONLy THE WORLD SANK Wahwnton. D.C. 20433 U.SA. Ofa. of OWlct4-Gemuri Opa Ivabstm March 16, 1990 MFMORANDUM TO THE EXECUTIVE DIRECTORS AND THE PRESIDENT SUBJECT: Project Completion Report on Turkey Highway Rehabilitation Project (Loan 21 7-TU) Attached, for information, is a copy of a report entitled "Project Completion Report on Turkey Highway Rehabilitation Project (Loan 2137-TU)" 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 7 This document ha a mctd ditrdbution and may be used by recipints only In the performwae of their offcial dutles Its conlents may not otherwis be disclod witbout World kank authoization.  FOR OMFCIL USE ONLY PROJECT CQO ETION REPORT TURKEY HIGHWAY REHABILITATION PROJECT (LOAN 2137-TU) TABLE OF CONTENTS PBe PREFACE . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . EVALUATION SUMY .... . . . . . . . . . . . . . . . . . . . . ii INTRODUCTION AND BACKGROUND ... . . . . . . . . . . . . . ii. OBJECTIVES .... . . . . . . . . . . . . . . . . . . . . ii IM4PLEMENTATION EXPERIENCE .i. . . . . . . . . . . . . . . . i RESULTS . . . . . . . . . . . . . . . . . . . . . . . . . . .iii SUSTAINABILITY .... . . . . . . . . . . . . . . . . . . . iv FINDINGS AND LESSONS . . . . . . . . . . . . . . . . . . . . iv PROJECT COMPLETION REPORT . . . . . . . . . . . . . . . . . . . . . . 1 PART I: PROJECT REVIEW FROM THE BANK'S PERSPECTIVE . . . . . . . . . 1 Sector Development Objective, Policy Context, and Macroeconomic Linkages .1... . . . . . . . . . . . . . I Project Objectives and Description . . . . . . . . . . . . . 2 Project Design and Organization . . . . . . . . . . . . . . . 3 Project Implementation .... . . . . . . . . . . . . . . . 5 Project Results .... . . . . . . . . . . . . . . . . . . . 9 Economic Re-Evaluation . . . . . . . . . . . . . . . . . . . 9 Economic Rates of Return .... . . . . . . . . . . . . . . 9 Project Sustainability .... . . . . . . . . . . . . . . . 10 I This document has a restricted distribution and may be used by recipients only in the performa of ther offlchl duties. Its contents may not otherwise be disclosed without World nk auhorzion.  PlaeNo. Bank Performance . . . . . . . . . . . . . . . . . . . . . . 10 Borrower Performance . . . . . . . . . . . . . . . . . . . . 11 Project Relationships . . . . . . . . . . . . . . . . . . . 12 Compliance with Covenants . . . . . . . . . . . . . . . . . . 12 Consulting Services . . . . . . . . . . . . . . . . . . . . . 12 Project Documentation and Data . . . . . . . . . . . . . . . 12 Findings and Lessons . . . . . . . . . . . . . . . . . . . . 13 PART II PROJECT REVIEW FROM BORROWER'S PERSPECTIVE . . . . . . . . 14 PART III: STATISTICAL INFORMATION . . . . . . . . . . . . . . . . . 15 Table 1 Related Bank Loans ..15 Table 2 Project Timetable . . . . . . . . . . . . . . . . 15 Table 3 Cumulative Estimated and Actual Disbursements . . 16 Table 4 Traffic 1981-1987 . . . . . . . . . . . . . . . . 16 Table 5a Project Costs ..17 Table Sb Project Financing . . . . . . . . . . . . . . ... 18 Table 6 Studies . . . . . . . . . . . . . . . . . . . . . 19 Table 7 Missions . . . . . . . . . . . . . . . . . . . . . 20 Table 8 Status of Covenants . . . . . . . . . . . . . . . 21  PROJECT COMPLETION REPORT TURKEY HIGHWAY REHABILITATION PROJECT (LOAN 2137-TU) PREFACE This is the Project Completion Report (PCR) for the Highway Rehabilitation Project in Turkey for which Loan 2137-TU was approved on May 11, 1982 in the amount of US$71.1 million. The loan was closed on June 30, 1987, as scheduled. The loan was fully disbursed and the last disbursement was made on June 29, 1982. The PCR was prepared by the Infrastructure Operations Division of the Turkey and Pakistan Department of the Bank's Europe, Middle East and North Africa Region. The Borrower assisted with the economic re-evaluation of the civil works projects, but declined to the Bank's invitation to contribute a Part II to the report. Preparation of this PCR was undertaken during supervision missions of the follow-on project, the Second Highway Project (Ln. 2439-TU), and is based inter alia on the Staff Appraisal Report, the Loan Agreement and supervision reports.  - ii - PROJLCT COMPLETION REPORT TURKEY HIGHWAY REHABILITATION PROJECT tLOAN2137 -TU) EVALUATION SoUM4R INTRQOTION AND BACKGROUND I. The socio-economic setting against which the project was identified and prepared is discussed in paras 3-25 of the April 20, 1982 President's Reot (PR). Paras. 2.01-2.27 of the April 22, 1982 Staff Aporaisal feoert (SAR) describe the status and prospects of the highway subsector in the early 1980s. The salient feature was that most of the roads had not been designed to meet the growing traffic demand ad that despite extensive maintenance programs, were deteriorating rapidly. 'te project was to support Government efforts to rehabilitate and strengthen the network, with emphasis on highways which played an important part in foreign exchange earning. II. When the project was appraised in November 1981, the Bank was familiar with constraints on project implementation in Turkey. The First Railway Project of 1973 and the Ports Rehabilitation Project of 1979 had revealed important stumbling blocks that had to be removed. The OED report on the First Railway Project (Report No. 5768 of June 30, 1985) and the PCR on the Ports Rehabilitation Project (OED Report No. 6213 of May 21, 1986) discuss in detail reasons for cost overruns, for time overruns, and for partial achievement of project objectives. Q8IECIES III. Intended: Improve through rehabilitation and strengthening priority sections of the network; promote improved planning and control systems; support the Government's efforts to improve highway safety and develop a suitable vehicle loading control system; provide training for highway staff (SAR, para. 3.02; PCR, para. 5). Achieved: The objectives-of the project were substantially achieved. IMPLEMENTATION EXPERIENCE IV. Experience with the implementation of the different components constituting the appraised project (SAR, para. 3.03) was as follows: a A three-vear time slice of the highwav rehabilitation and strengthening program (about 775 km.) to be comoleted by 1986. A total of 965 km. of TETEK road sections were rehabilitated,  - ii. - of which 337 km. were financed by Kuwait Fund, and 628 km. by the IBRD loan. The average cost was about US$158,000 per km. b Procurement of specializged egUiMene for highway rehabilitatign works. This was handled with speed and efficiency. The items procured went into immediate service on KGM's force account maintenance work (para. 14). c EauiDment and materials for {i' the nregaration of future Rrograms of highway rehabilitation: (iL) a vehicle axle loading and control system. and a program to evaluate pavement nerformance and related road user charg2es: and (LiiQ a highway safety proeram. The equipment was procured. A working pavement management system is not yet in place, and the road user charge study was not completed until two years after loan closure. The highway safety program is ongoing and successful (para. 17). d Consultancv services for supRlementarX studies (unommitted at Board airnroval but which could include studies such as imRrovement to the higha planninf process. on asgects of-ro maintenance, and a follo_w-u of the National Tra=rt Master Plan like bulk transRort need. The PCR discusses consultancy services in general under Xpara. 31/ and says that, on the whole, and despite the earlier reluctance of the Executing Agency to use consultants, these were satisfactory. Study results took a long time to come (PCR, para. 17) but have been of moderdte use. The Highway Safety Study has contributed to the implementation of an effective and well-monitored safety program. The completion of the road user charge study was delayed until September 1989, because data on user revenues was slow in coming from other Government departments (PCR, para. 18). e Fellowshigs for training of hiehway staff. Satisfactorily completed (PCR, para. 15). V. The borrower complied with all covenants except the requirement to prepare a Project Completion Report (para. 29). RESULTS VI. The estimated prolect cost was US$264 million (including about US$35 million of taxes and duties) with about US$140 million representing foreign costs (SAR, para. 3.17 and Table 3.1). The IBRD-financed parts of the project had a total appraised cost of US$127.9 million. Actual proiect cost of thi IBRD-financed parts of the project was US$123.5 million.  - iv - Planned grojeot comoletion war in December 1986 (SAR, para. 3.21). Actual nrolect completion was on June 30, 1989, though two civil works contracts were not complete at that time and were finished using funds from the follow-on project. The Loan was closed on June 30, 1987. Calculated under a variety of assumptions, the estimated Economic Rate of Return (ERR) for three road sections ranged from 47X to 349X (SAR, paras. 4.01-4.09). Re-estimated ERRs for three completed sections representative of the three major areas of influence yield from 43 to 681 (PCR, para. 23). SUSTAINABILITY VII. The PCR states (paras. 25-26) that project-generated benefits regarding both vehicle operating costs and institutional improvements are sustainable. FINDINGS AND LESSONS VIII. The PCR shows that the project was successful in achieving its physical targets. There was also considerable progress made in introducing improved methods of project preparation which led to smooth preparation of a follow-on project. The Turkish Government is a good partner in implementing the "hardwarel elements of a project. It was more difficult to get the Borrower's cooperation for the studies and policy elements of the project. The lesson to be learned is that preparation should have been more advanced before Board presentation. In particular, terms of reference (TORs), shortlists, and consultant proposals for studies should have been cleared during preparation.  PROJECT COMPLETION REPORT TURKEY HIGHWAY REHIABILITATION PROJECT (WAN 2137-TU) PART I; PROJECT REVIEW FROH THE BANK'S PERSPECTIVE Project Identity -Project Name: Highway Rehabilitation Project -Loan Number: LN-2137-TU -RVP Unit: EMlIN -Country: TURKEY -Sector: Transuort -Sub-sector: Highways Sector Develo2nent ObIetives. Policy. Context and Macroeconomic Linka2es 1. The Bank's involvement in the transport sector has been small and sporadic. To date, the Bank has made only 7 loans to the sector versus over 30 in industry and over 20 each in agriculture and energy. Of these 7 projects, one was made in the 50s (First Port Project, Ln. 28-TU) and one in 1973 (First Railway Project Ln. 893-TU). Five loans were then made between 1979 and 1986, two each for ports (Ln. 1741-TU and Ln. 2535-TU) and highways (Ln. 2137-TU and Ln. 2739-TU) and one for railways (Ln. 2739-TU). The total amount lent in the sector amounts to US$743 million. Operations have focussed on increasing the productivity of the existing system and the effictency of the transport sector agencies and enterprises; preserving the existing infrastructure and equipment through adequate maintenance; and expanding the system to avoid bottlenecks affecting the productive sectors of the economy. 2. This project marks the first involvement of the Bank in the highway sub-sector in Turkey, mainly because the road network had been developed to fairly good standards during the postwar years with the assistance of USAID. With the rapid increase in domestic traffic and international transit between Europe and the Middle East, roads deteriorated rapidly in the late 70s. In 1977, the Government asked for Bank assistance for the rehabilitation of the Trans-Turkey Highway (TTH) which not only serves as a transit route, but also links the major 4conomic centers of the country from the west to the eastern borders with Iran and Iraq. After promising initial contacts, project preparation slowed considerably, as the Borrower tried to decide the extent of  cooperation it wanted with the Bank in the highway subsector. This was compounded by the serious economic and political problems affecting Turkey in 1979/80. 3. It was only after a productive macroeconomic dialogue had been reestablish, through Structural Adjustment Loans (SALS) which began in 1980, that Bank dealings with tke transport sector also improved. Preparation of this project was resumed in 1981 and was well linked with the macroeconomic objectives prevailing at the time. tor instance, the 1981 Public Investment Review (PIR)V called for a reorientation of public investment toward critical infrastructure needs and an improvement in economic evaluation of projects by implementing agencies. The third SAL I, approved at the same time as this project, argued that immediate priority should be given to rehabilitation and expansion of investments in infrastructure and energy since such projects were crucial to remove production bottlenecks, support private sector growth and facilitate export-led growth. Within this framework, preparation of this project resumed and was completed quickly. In retrospect, the decision to resume preparation in 1981 within the adjustment framework was appropriate and timely. Proiect Objectives and Descriotion 4. The experience with the railway project (Ln. 893-TU) of 1973 suggested the need for realistic objectives; originally scheduled for completion by 1976, the project was only completed in 1981. Also, since this project was the first operation in the highway sector, it was decided to keep it reasonably simple. This decision turned out to be wise. 5. The objectives of the Highway Rehabilitation Project (HRP), which was signed in May 1982, were firs, to improve the highway network by rehabilitation and strengthening of certain priority sections; V sec to improve planning and control systems in the General Directorate of Highways (KGH); third, to support Government efforts to improve highway safety and develop a suitable vehicle load control program; and fourth, to provide training for KGM staff. The project consisted of the following components: !J Report No. 3472-TU, December 1981. Board Approval May 1982. 3V Priority was given to sections yielding hish short tozr benefits; particularly those oriented towards exports.  (i) a three-year slice of the national highway rehabilitation and strengthening program, covering about 775 km., to be initiated 4.ring 1982-84 and completed by the end of 1986; ' (ii) provision of specialized highway paving and quarrying equipment to complement existing equipment at about five quarry and rehabilitation sites; (iii) development of a planning system for highway rehabilitation and strengthening, including necessary equipment; (iv) development of a vehicle load control program and a highway safety program, including necessary equipment; (v) provision of 60 man-months of fellowships for highway related staff: and (vi) supplementary studies on highway or transport operations as agreed. Proiect Design and Organization 6. The government's expressed desire to upgrade priority sections of the Trans Turkey Highway was initially a response to severe criticism in Europe and the Middle East regarding the poor condition of Turkish highways in the mid-70s. It was recognized that highway expenditures had not kept pace with the growth in traffic, and therefore road standards were no longer adequate to satisfy rising volumes of domestic and international traffic. A project brief was prepared in December 1977, but languished until November 1980, when a formal identification mission took place. During this period, a consensus developed among Bank staff and officials at the Ministry of Public Works (MPW), KGM, and the State Planning Office (SPO) that alleviating transport constraints was essential to the realization of Turkey's macroeconomic objectives in the context of high inflation, severe balance of payments problems, and a large foreign debt. The urgency of this task became more apparent after the Project Brief highlighted key transportation bottlenecks w'nich could thwart achievements under the Bank's SAL program which supported Turkey's shift to export-led growth. The renewed emphasis on alleviating transport constraints was thus heavily influenced by the deteriorating macro-economic balances and the shift in national development strategy.v 4So be financed by IBRD and rZuwait Fund.each providing about US$70 million. The Bank 's first SAL for Turkey was approved in 1980. The third SAL was approved by the Board during the week following approval of the Highway Rehabilitation project, thus emphasizing the complementary nature of adjustment operations and infrastructure investments. Specifically, the highway rehabilitation project was expected to facilitate higher exports and reduce in maintenance expenditures, measures designed to support external and domestic adjustmewts respectively.  - 4 - 7. The physical components of the project were well prepared, an important contributing factor to later success in implementing physical works. It was simple in design and appropriate for the Bank's first investment in the sub-sector. Government's administrative and coordinating arrangements had improved by 1980 and given the simple nature of the physical works, KGM was expected to be able to carry out the project efficiently. Consultants would undertake studies when necessary to supplement KGM staff expertise. 8. KGM was responsible for the project preparation effort, but received strong support from several Bank missions (Table 8b). Most economic and technical feasibility studies for proposed sub-projects had been completed and approved by the Bank within one month of loan signature (May 82) and local funds were committed by the government. Bid documents for 70X of equipment to be procured during the first tranche of the operation had already been prepared by the time of loan signature and US$60 million in ICB contracts had been let for rehabilitation works within the first six months. The strong preparation effort also enabled the government to secure assurances of parallel financing from the Kuwait Fund prior to Board presentation in May 1982. Loan signature was two days after Board presentation. (Tabie 2). 9. The fact that a single agency (KGM) was responsible for all aspects of the project contributed to its success. The project also gave appropriate emphasis to institutional objectives. The need for institutional strengthening was recognized by KGM itself during the preparation process. Solutions were thus not imposed, but worked out in a collaborative manner during preparation and taken up by the executing agency. The long period of preparation led to a high degree of political commitment from the government and counterpart agency. 10. The Project Control and Coordinating Team (PT) was established within KGM prior to negotiations. The unit proved to be an effective and efficient organization with requisite leadership and backing from senior management in KGM. The PT was involved in project design from an early stage with the Coordinator attending negotiations. This helped to prevent future misunderstanding regarding their role and responsibilities. The necessary support of other Ministries for project financing and processing was also assured at the outset thus allowing KGM to proceed with a clear mandate and resources. Establishing monthly cash flow requirements and backing from Treasury in particular, helped to ensure that the major project risk identified during appraisal, i.e. shortages in local cost financing, did not occur. 11. The time frame established at appraisal for implementing rehabilitation works and costs was generally realistic. However, works as initially designed underestimated actual traffic growth and did not anticipate legislation increasing axle load limits (para. 16). Modification of original pavement designs to increase structural capacity was necessary in 1985 resulting in cost e_calation. As a result, completion of two paving contracts were completed using Second Highway Project funds, once loan 2137-TU funds had  -5- been exhausted. The time frame and terms of reference established for the axle load, pavement research, and traffic safety studies were also overlv optimistic given the technical issues involved and KGM's reluctance to recruit necessary staff and employ consultants. (para. 17) Difficulty in ottaining information from other governmental entities also hampered progress on the user charge study. Project Implementation 12. The implementation phase of this project was characterized by relatively fe .jor problems in construction and supervision of physical works. Although there had been frequent changes of director-general in KGM during preparation, there was only one change during the implementation period. There was no major re-organization of KG?M during this period, and the committees set up to span the otherwise rather rigid inter-departmental boundaries worked well. The PT exercised its monitoring, control, evaluation, and re.porting functions efficiently. The need for additional staff was flagged early in supervision (October 1982) when reporting to the Bank was in arrears. This was quickly remedied when KGM agreed to strengthen the team with two new positions, finance additional staff training in project management, and computerized monitoring systems were set up with assistance from Bank specialists. Withdrawal applications, quarterly progress and audit reports were provided to the Bank on a timely basis. 13. The planning, design and preparation of contract documents for the sixteen road contracts was well handled. With Bank advice on methodology, the Planning Directorate produced gor. economic evaluations of each road section, and the design and contract documents basically followed KGM's existing practice, with modernization of pavement design methods, and document changes to comply with Bank guidelines on procurement. The departments concerned were strengthened by exposure to the methods of the Bank and the requirements of FIDIC and international competitive bidding. This benefit was retained and made the preparation of the second project much smoother. However, although KGM's contract document was extensively modified (and translated into English), one established Turkish practice --that of calling for bids expressed as a percentage reduction in the Government's published estimate-- was retained. KGM argued that there was no difference in the competitive effect of the two methods, but their contractors were familiar with the system in force, which was in fact mandated by law. The Bank accepted this argument for the project, but this led to serious problems during start-up of the second project, when following a change of management, the Bank would no longer accept this practice. Turkey claimed that the existing practice already accepted by the Bank should be retained, and bidding was delayed. It should be noted that the Turkish contracting market is very competitive (some 60 firms applied for prequalification) and this practice nevertheless achieved fairness and competitior.  14. Disbursements on the HRP were slower than expected through FY84. ! This was due to an exceptionally harsh winter in 1982/83 during which all construction was halted. Although this could not be anticipated, KGM and contractors accelerated works, and disbursements improved rapidly to converge on appraisal estimates by the end of FY 83 (see Table 3). The Bank allowed an increase in the disbursement percentage for paving from 51% to 65X and for earthworks from 35% to 49X in 1984, and remaining funds were disbursed on schedule. There were revisions of schedule 1 of the Loan Agreement in December 1984, February 1985, and June 1986 to adjust for the changes in disoursement percentage (see table 5b). Throughout implementation there were no major delays due to counterpart funds not being available. Procurement was also handled efficiently. 80% of equipment for phase 1 of the project had been delivered by October 1983, only 14 months after the loan became effective. All bids had been awarded by March 1984 and equipment t-as delivered according to schedule. A comparison of estimated and actual project cost3 is shown in Table 5a. Table Sb shows the distribution of the loan between disbursement categories, and how it evolved. Total cost was US$4.43 million less than the appraisal estimate, but this result masks the heavy devaluation of the lira that took place over the implementation period, and does not account for the two contracts that were completed under the follow-on project at a cost of $2.07 million. This also masks the cost of thickening the asphali layers to absorb higher axle loads (see para. 16), so cost comparisons are not very meaningful. 15. A component not sufficiently detailed during preparation was overseas training for KGM staff. This was recognized during the first supervision mission. A training consultant from UNDP was immediately engaged by KGM as suggested. He arrived 3 months after loan effectiveness, and recommended a three year program of courses for engineers and supervisory staff. To minimize disruption to normal duties, KGM settled for sherter courses (generally 2-3 weeks) for groups of 10-14 where interpreter use could be maximized. Training began in June 1983. The project budget was adjusted from US$200,000 to US$300,000 to cover the extra cost of the new program. Training took place mostly in the United Kingdom where the logistics and placement of trainees was sub-contracted to the British Council, an arrangement which worked well. A total of 89 KGM engineers received training under this arrangement. 16. Construction of physical works, drainage, earthworks, and sub-base, progressed as scheduled. These works were executed by "tacherons" (i.e. small contractors) selected by local competitive bidding from the lists of firms registered with KGM for its own works. Paving contracts, however, progressed at slower-than-anticipated rates, due to weather and supply constraints. Bitumen shortages were experienced by some contractors during 1985. In the follow-on project it was finally realized that domestic production could not keep pace with the increasing volume of works, and the loan agreement was amended to allow for the import of bitumen from the international market, v Actual disbursement were 231 of appraisal estimates in FY83, 711 in FY84.  -7- which incidentally provided a higher quality product. The major variance in project estimates, however, resulted from legislation in June 1985 which raised the maximum load for a single axle from 10 tons to 13 tons. The increase in axle loads compounded with heavier than expected heavy vehicle traffic on trunk routes to Iraq U required design thickness increases on all IBRD-financed paving contracts, which raised the cost of paving contracts by about 25X. Most of this overrun was covered by continuing devaluation of the lira. However, two of sixteen paving contracts V were still incomplete by the time the loan had been fully disbursed and were continued under the follow-on Second Highway Project (Ln. 2439-TU, Board approval - June '84). 17. The other major variance in implementation occurred in studies to examine highway safety and axle load control. Data collection for the safety study was initiated and an interim draft submitted according to schedule in March 1983. Deficiencies in data regarding accident location, travel direction etc. caused the Bank to reiterate its earlier recommendation that a specialist in highway safety be engaged to assist KGM. Delays were anticipated by the Bank and the original due date for the final study was extended from July 84 to December 1985, and later to December 1986. During this period, Bank missions continued to press KGM to hire consultants and even drafted Terms of Reference. KGM initially agreed to this proposal, but later changed its position. The government did finally engage Middle East Technical University and Istanbul Technical University to assist in "black spot" analysis. Vehicle inspection was improved through German Technical assistance which has eventually resulted in the launching in 1988 of a large program of vehicle testing based on 90 fully instrumented test centers operated by the private sector, for which bids had been received. The final safety study was completed in 1987 under covenants in the Second Highway Project Loan Agreement. Delays were the result of KGM's lack of experienced staff in this area, reluctance to recruit outside consultants on a timely basis, and deficiencies in data available from accident reports. There was clear recognition among all parties that improving highway safety was an important priority and KGM constructed crash barriers, posted highway signs, and trained traffic officers in other ministries during the project implementation period. There is now an effective and well monitored safety program in KGM with engineering measures having reduced the aggregate length of road considered "critical" by 35 percent, between 1984 and 1987. 18. Delays were experienced in the axle load study and related analysis of user charges. The study was originally due on December 31, 1983, however, government was already enforcing maximum axle loads using 21 fixed weigh stations and 48 mobile stations by November 1982. The companion analysis of user charges and of the damaging effects of heavy vehicles had to be postponed An important contributing factor to the increase in traffic was the continuing war between Iran/Iraq which spurred demand for Turkish exports. While the appraisal did try to anticipate such occurrences, the level and duration of this conflict and consequent demands on Turkish infrastructure were of course impossibLe to predict. V Komurler-Gaziantep, Saklan-Habur.  - 8 - until December 1985, by which time legislation raising axle loads had been passed. KGM's Research and Planning departments had completed an excellent study on incremental highway costs resulting from heavy vehicle use by April 1984, but additional analysis was required to compare the optimal solutions presented with the actual user charges in effect. This comparison was delayed until September 1989, due to the inability of KGM to obtain necessary data on road user revenues from SPO and MOF. Inability to secure cooperation from other departments not directly involved in implementation thus thwarted project objectives. 19. The establishment of experimental road sections for pavement research was also delayed. Its purpose was to test different pavement design and materials under varying local conditions and to develop appropriate designs. Four test sites were originally planned in June 1982. This was quickly scaled back in November to one section that would be constructed on the Ankara-Eskisehir highway. The Bank recommended that specialists from the Transportation and Road Research Laboratory (TRRL-U.K.) be engaged to assist. The Bank and TRRL prepared an options paper on pavement research in October 1983, but KGM was not receptive however, and preferred to continue with its plan to construct 36 test sections of 50m each at one site on the Ankara- Eskisehir highway. It planned to complete this during 1984 by force account. The Bank agreed to finance these works and provided technical assistance through TRRL. The Eskisehir test sections were finally completed as part of the Second Highway Project in the summer of 1987, and are currently being monitored by KGM's technical research department for performance under traffic. There was little formal progress in the implementation of a new planning system. Although everything required was done for implementing specific project-financed works, and though new techniques were learned in KGM (para. 13), it would not be correct to say that any major change took place in KGM's planning procedures generally, as a result of the project. The eventual aim of this work is to provide input to a KGM-wide pavement management system (PMS). KGM's Planning Division is gathering inventory data, traffic and loading data, and its staff are becoming conversant with the operation and data requirements of the Highway Design and Maintenance Standards Model III software program, recently supplied by the Bank, and they are currently discussing with supervision missions, the purchase under the second project of PMS data gathering equipment-- deflectometers, bump integrator and the like. Thus, though there has been no formal commitment by KGM senior management to set up a PMS in any formal sense as a response to Bank reques.s, informal contacts are beginning to produce the same result. 20. During bilateral talks prior to Board Presentation it was understood that the Kuwait Fund would finance US$70 million of the cost of the joint project, with their financing going for clearly defined road rehabilitation contract lengths. However Kuwait Fund finally made two loans to Turkey, Ln. 240 for Kuwaiti Dinars (KD) 5.5 million,(US$19.0) and in 1983 Ln. 268 for KD 5.0 million, (US$17.2 million). Disbursement details of Kuwait loans were not available to Bank in preparing Part I of this report.  -9- Project Results 21. The project did achieve most of its specific objectives on physical works. Priority sections of the Trans Turkey highway amounting to 967 km (of which 297km were financed by Kuwait Fund and 667km by IBRD) were rehabilitated and strengthened. Computerization of project monitoring and introduction of methodologies for evaluating economic and technical viability of all sub- projects proposed in the 5-Year Plan indeed also went beyond the planning and control objectives foreseen at the time of appraisal. Studies completed on safety and axle load monitoring were delayed, but ultimately have also enabled the government to establish suitable programs in these areas. The long term impact of these studies on highway safety and optimizing use of transport infrastructure should be high. KGM has been constructing crash barriers, posting signs, and carrying out training activities to improve highway safety throughout the implementation period. These engineering measures have shown a major reduction in the aggregate lengths of road considered "critical", but nevertheless accidents have continued to increase nationally. Road Safety requires in addition to engineering, a matching effort in education and enforcement currently being pursued as part of the Bank's transport dialogue. Axle load monitoring and control is now institutionalized in Turkey--the next stage will be to ensure that the incidence of user charges fully reflect the costs imposed by various classes of vehicles. Completion of the study comparing present and optimal user charges is a necessary step in this regard, and is expected by late 1989. The benefits of the pavement research program on improved highway design and maintenance costs can only be expected after 5- 10 years. 22. The training objective was also met through a well-conceived and implemented series of seminars held in the U.K. The 89 senior engineers and staff benefitted from a total of nine courses in highway design, construction, maintenance, fleet management/spare parts control and other management courses and site visits that took place in the U.K. The results of the program were very satisfactory according to participants and KGM. Economic Re-evaluation 23. The economic analysis of the project roads was based on a estimated general traffic growth rate of 5X p.a. Between 1981 and 1987, traffic growth on various sections ranged from 8 to 10 p.a. (see Table 4). Vehicle operating cost savings were estimated to make-up 98% of the project benefits. Economic .ates of return (ERR) were recalculated by KGM for three typical sections; on the same basis as the original estimates: Economic Rates of Return Road Section Original ERR Recalculated ERR Kulu-Kochisar 59 43 Ssklan-Habur 52 51 Adana-Toprakkale 42 68  - 10 - 24. The project also had a positive impact on the Bank's macroeconomic objective of fostering trade. Exports to Iraq, as measured by the number of crossings at the Habur( Iraq) border, have increased from 2250 commercial vehicles per day in 1982 to 3837 in 1987. Although it is not claimed that all goods moved by road, gross exports to Iraq rose (in value) from $134.8 million in 1980 to $945.3 million in 1987. They have fallen somewhat in 1988, as Turkey limited the issue of letters of credit in favor of Iraq, because of non-payment of arrears. The impact on poverty has not been determined, but this was not an explicit objective of the project. It should be noted, however, that major portions of the highway financed by the Bank loan are in the most economically depressed regions of Turkey. It is reasonable to assume that the investments will contribute to an acceleration of economic activities in these areas and particularly of the massive Southeastern Anatolia Region Development Project, (GAP) which is currently underway. Project Sustainability 25. The project is likely to maintain an acceptable flow of net benefits throughout its economic life. Important factors in this regard are whether the government will continue to provide adequate funds for maintenance, will ensure optimal use of resources through user charges, and whether Turkey will continue its strategy of export-led growth. A related short-term factor will be the impact on export demand and projected traffic growth from the cessation of hostilities between Iran and Iraq. There was a lull in Iraq traffic at the height of the war, but over the period 1985-87 traffic still increased by 15X. This lull is not, however, expected to continue since economic recovery in both countries and the consequent reconstruction needs that will arise should contribute to continued buoyancy in the demand for Turkish exports. 26. An important institutional objective was met when methodologies adopted by the PT for evaluating technical and economic viability of rehabilitation works financed through the IBRD loan were adopted throughout KGM for investments proposed in the 1983-88 Plan. This was a major step forward and should help to maintain an underlying objective of the project- that of ensuring that highway rehabilitation funds are invested to maximize economic benefits. Under the Second Highway Project this process is being taken a step further with the introduction of HDM3 into KGM's Planning Department, for program evaluation. Bank Performance 27. Bank performance was conscientious in the preparation and supervision of the project. During the preparation period there were eleven directors-general of Highways and frequent changes of government. Following appraisal, stability returned and with it good cooperation and speedy implementation. The phasing of supervision was timely (see Table 7). Early supervision appears to have been an important means of ensuring that the project began on a solid foundation. Subsequent supervision during the first 21 months was carried out with greater frequency (4 months) than later missions, which were undertaken at 6 months intervals. The later missions  combined supervision of the first with preparation of the Second Highway Project. Teams were also well balanced, with transportation engineers and other specialists examining technical progress and economists concerned with the efficiency issues. 28. One area where Bank might have acted more forcefully was in urging the government to complete the safety and axle load studies according to the original schedule by employing outside consultants despite the latter's reluctance. Turkey's reluctance to employ consultants is a problem common to many projects in Turkey. Compliance on the safety study was ultimately ensured only through negotiations on the next operation, despite the fact that both studies were included in loan covenants for the Highway Rehabilitation Project. The Bank was also not successful in getting KGM to complete the study comparing optimal user charges with prevailing charges. Eventually under the follow-on project, KGM has completed its work on determining what road use charge should be levied to compensate the highway authority for expense incurred by the passage of a vehicle, but the lack of inter- departmental co-operation prevented KGM until recently from getting data on revenues actually being collected, and thus ultimately determining whether the existing charges are equitable. Borrower Performance 29. Most civil works were implemented according to schedule, despite design changes and the vagaries of weather. When these factors caused slow- downs, government acted quickly to arrange advance payments to contractors to allow greater capital intensity and completion on schedule. Indeed, the design changes, forced as a result of axle load legislation and higher traffic growth, indicate a growing capacity within KGM to monitor and control for such variables and willingness to adjust highway design specifications when appropriate. Technical supervision of the project by the borrower was strong, and contracts and procurement were handled well. The requirement to employ consultants to assist with certain studies was only partially complied with. KGM delayed considerably in employing consultants on safety and axle load studies. Table 6 shows a list of the studies undertaken and the outcome. KGM acted responsibly in deploying necessary internal staff for the project. Reporting was regular, but focussed exclusively on physical achievements or variances,with no analysis or narrative to explain the causes of such variances. The only area of non-compliance with the conditions of the Loan Agreement is that KGM has thus far has not been willing to prepare Part II of the PCR.  - 12 - Project Relationships 30. The cooperation between the Bank and KGM was excellent throughout the project, where construction and equipment procurement were concerned. The only area of outright disagreement related to KGM's reluctance to engage consultants to assist in certain tasks (para. 17). Throughout the project, however, there was resistance to the idea that the Bank had any role to play in influencing purely policy matters. Although KG) gave the impression that its actions in connection with policy matters -road safety, axle load control, pavement research- were .ot done to comply with loan conditions, it is important to note that acceptable action has been taken, and except as explained above, all loan covenants were complied with. Compliance with Covenants 31. . With the exception of the PCR noted in para. 29, all covenants were complied with. Consulting Services 32. KGM is a highly skilled organization capable of handling the bulk of analytical and technical requirements of investment without outside consulting assistance. The reluctance of KGM to utilize consultants is based on confidence in their own ability to carry out most types of analysis, rather than lack of funds. Divisional offices of KGM are responsible for most engineering with the central office exercising overall supervision. Consultants were thus not used for this phase of the project. KGM field offices supervised construction. Most of the studies including axle loading and safety planned at appraisal were also to be undertaken by KGM itself with consultants limited to unidentified supplementary studies. Consultants were therefore expected by KGM's staff to play a very minor role in this project. As the complexity of axle loading and safety studies became more apparent and constraints on necessary staff expertise arose, use of consultants increased though delays occurred when KGM was reluctant to accept thi.i conclusion. After adopting Bank recommendations on engaging outside expertise, consultants ultimately made valuable contributions to the axle load and safety studies begun by KGM and in establishing the pavement research program. A consultant was also brought in at a critical juncture in the development of the overseas training program and performed well. Likewise, consultants were used with satisfactory results in the equipment use/spare parts study, and other supplementary studies including one related to constructior. of the Bosphorous Bridge. Project Documentation and Data 33. The legal agreement and the appraisal report provided a useful and realistic framework for implementation. Data to complete the PCR were readily available in the project files, Management Information System and the Time Recording System. Supervision reports were also satisfactory and provided the most critical information regarding the implementation process. Narrative  - 13 - portions of quarterly and annual reports submitted by the borrower were not very helpful,(see para. 29), but audit reports were satisfactory and timely. In general, handling of disbursements was well done and on time. Findings and Lessons 34. This report shows that the project was mostly successful in achieving its physical targets, though two contracts were completed using funds from the follow-on project. There was also considerable progress made in introducing improved methods of project preparation which have already produced as a further benefit a quick and smooth preparation of the follow on project. It is clear that the Turkish Government is a good partner in implementing the "hardware" elements of a project. It was more difficult to get the Borrower's cooperation for the studies and policy elements of the project, and the lesson to be learned is that preparation should have been more advanced before Board preparation. In particular, TOR's, shortlists, and consultant proposals for studies should have been cleared during the preparation stage.  - 14 - PART II JECT REVIW FROM BORROWER'S PERSPECTIVE The Bank extended the due date for the PCR from December 1987 to March 1989 to enable KGM to complete work on the final two highway sections that were being completed with Second Highway Project funds. Despite this postponement, KGK did not prepare a Part II, and recently advised the Bank that it would not do so. A severe constraint on English-speaking staff was suggested as a reason. The lack of an input from the Borrower is regrettable, particularly given the generally satisfactory record of the project.  - 15 - PART tI I Table 1 Related Sank Loans tou Tit1 h *t~~~~~ear of Loan Title Purpos Apmt sae "icoe lignmy Projet - - - LU 2439-1t Remb. 1 19684 oqui Transport Sector Adjustant Adjustuut prwration I/ Rhbabilitation hihlay netpork, iaprov laming and Msanit coatiu obJetivn nUtbl ishd in HgtWy aifitutimn LoaM (L.12137-TGl Table 2 Project Tizetable Ites t. O Date Coents planod Actul -RICmOuaSSIONe 611 -preparation 9071 9/07 -Projet Brief 12/77 Duri this period Ss't ul 16 wr undergoing maor -liatitiftion ct 1110 u an were Itbe to totto bomlrroi -AWrail ission 9l78 IIIII -Le NlgotiatIons 2123182 -foard Aproval FY60 5/11/6 -Loan lignature 5113/2 .-Lon Effctivennss 6113182 -Lan Clasimq i 30/61 4130/87 -Loa Co,oletlon 12/31/8 6130187  - 16 - G .1.Uw Etiast Mi ktatStwi kV" $W0um 115 P1 i,p 1m