Turkey - Road Improvement & Traffic Safety Project

Document of The World Bank Report No: 25541 IMPLEMENTATION COMPLETION REPORT (CPL-40480; SCL-4048A; SCL-40490) ON A LOAN IN THE AMOUNT OF US$ 250 MILLION EQUIVALENT TO THE REPUBLIC OF TURKEY FOR A ROAD IMPROVEMENT & TRAFFIC SAFETY PROJECT November 6, 2003 Infrastructure and Energy Services Department Europe and Central Asia Region  CURRENCY EQUIVALENTS (Exchange Rate Effective August 2003) Currency Unit = Turkish Lira (TL) 1,000,000 TL = US$ US$ 1 = TL Jan 1992 Jan 1993 Jan 1994 Jan 1995 Nov 1995 May 1996 5,113 TL 8,628 TL 14,000 TL 38,801 TL 50,000 TL 75,000 TL Jan 1997 Jan 1998 Jan 1999 Jan 2000 Jan 2001 Jan 2002 Jan 2003 108,000 TL 206,000 TL 315,000 TL 540,000 TL 670,000 TL 1,445,000 TL 1,650,000 TL FISCAL YEAR January 1 December 31 ABBREVIATIONS AND ACRONYMS - AADT Average Annual Daily Traffic - CAS Country Assistance Strategy - EA Environmental Analysis - ERR Economic Rate of Return - GU Gazi University - HDM-3 Highway Design and Maintenance Model, Version III - HIMP Highway Investment and Maintenance Program - ICB International Competitive Bidding - ICR Implementation Completion Report - KGM General Directorate of Highways - MOE Ministry of Education - MOEn Ministry of Environment - MOH Ministry of Health - MOPWS Ministry of Public Works and Settlements - MOT Ministry of Transport - NP National Road Safety Program - PMS Pavement Management System - PP Pilot Road Safety Program - PFPSAL Programmatic Financial and Public Sector Adjustment Loan - QAG Quality Assurance Group - RUC Road User Charges - SAR Staff Appraisal Report - SEE State Economic Enterprises - SPO State Planning Organization - TL Turkish Lira - TTP Turkish Traffic Police Vice President: Shigeo Katsu, ECAVP Country Director: Andrew N. Vorkink, ECCU6 Sector Director: Hossein Razavi, ECSIE Sector Manager: Eva Molnar, ECSIE Task Team Leader: Mirtha Pokorny, LCSFT  TURKEY ROAD IMPROVEMENT AND TRAFFIC SAFETY CONTENTS Page No. 1. Project Data 1 2. Principal Performance Ratings 1 3. Assessment of Development Objective and Design, and of Quality at Entry 2 4. Achievement of Objective and Outputs 4 5. Major Factors Affecting Implementation and Outcome 8 6. Sustainability 10 7. Bank and Borrower Performance 10 8. Lessons Learned 13 9. Partner Comments 13 10. Additional Information 14 Annex 1. Key Performance Indicators/Log Frame Matrix 15 Annex 2. Project Costs and Financing 16 Annex 3. Economic Costs and Benefits 18 Annex 4. Bank Inputs 19 Annex 5. Ratings for Achievement of Objectives/Outputs of Components 20 Annex 6. Ratings of Bank and Borrower Performance 21 Annex 7. List of Supporting Documents 22 Annex 8. Implementing Agencies Contribution and Comments 23 Map IBRD 27396R  Project ID: P038091 Project Name: Road Improvement and Traffic Safety Project Team Leader: Olivier Le Ber TL Unit: ECSIE ICR Type: Core ICR Report Date: November 6, 2003 1. Project Data Name: Road Improvement and Traffic Safety Project L/C/TF Number: CPL-40480; SCL-4048A; SCL-40490 Country/Department: TURKEY Region: Europe and Central Asia Region Sector/subsector: Roads and highways (94%); Central government administration (5%); Health (1%) Theme: Infrastructure services for private sector development (P); Rural services and infrastructure (S) KEY DATES Original Revised/Actual PCD: 09/12/1994 Effective: 09/23/1996 10/10/1996 Appraisal: 06/21/1995 MTR: 11/15/1998 11/15/1998 Approval: 06/20/1996 Closing: 03/31/2003 03/31/2003 Borrower/Implementing Agency: GOVERNMENT OF TURKEY/KGM Other Partners: EGM (Traffic Police), Min. of Education, Min. of Health, Gazi University STAFF Current At Appraisal Vice President: Shigeo Katsu Wilfried Talwitz Country Director: Andrew N. Vorkink Kenneth G. Lay Sector Manager: Eva Molnar Ricardo Halperin Team Leader at ICR: Mirtha Pokorny Mirtha Pokorny ICR Primary Author: Mirtha Pokorny Robert H. Nooter Olivier Le Ber 2. Principal Performance Ratings (HS=Highly Satisfactory, S=Satisfactory, U=Unsatisfactory, HL=Highly Likely, L=Likely, UN=Unlikely, HUN=Highly Unlikely, HU=Highly Unsatisfactory, H=High, SU=Substantial, M=Modest, N=Negligible) Outcome:S Sustainability:L Institutional Development Impact:M Bank Performance:S Borrower Performance:S QAG (if available) ICR Quality at Entry: S Project at Risk at Any Time: No There was no Quality at Entry (QAE) assessment of the project (the QAG did not exist at the time).  3. Assessment of Development Objective and Design, and of Quality at Entry 3.1 Original Objective: The original project objectives were: (a) the reduction of road transport costs through infrastructure improvements and the protection of past investments in the highway sector through rehabilitation and strengthening of paved highways; (b) the improvement of traffic safety in state and provincial roads; (c) the improvement of the operational efficiency of KGM through the implementation of management systems and computerization; and (d) improvement in the consideration of environmental factors in project selection and design. The objectives were consistent with the 1993 CAS priority of providing support for infrastructure development. The 1993 CAS was under review and revision at the time of project preparation. In any event, improving transport by easing transport bottlenecks was an integral part of the Government's development agenda to promote the growth of the private sector, and the project was a follow-up to a series of projects supporting these objectives. The previous projects, however, while successful in funding civil works and equipment, did not succeed in overcoming KGM's reluctance to implement the technical assistance components for institutional development. This state of affairs was best summarized in OED's comments on a previous ICR stating: "KGM perceives the Bank as a source of funding and not as a partner in development". Based on the previous experience, the Bank decided that the project would only go ahead if it would make inroads progress in achieving institutional improvements in the road sector. Because the high level of accidents on Turkish roads was well known, both internally and abroad, the lack of traffic safety was a potential threat to the increasingly important Turkish tourist industry. As public opinion was starting to exert strong pressure for government action, it was decided to focus the institutional efforts on traffic safety issues. Two other major concerns were key in shaping the project: (i) because in past projects an important share of the loan funds were dedicated to finalize works that, because of costs overruns, were left unfinished under the previous project, the Bank decided that the proposed new operation under no circumstances would finance cost overruns arising from continuous changes in project design; and (ii) past waivers to KGM's full compliance with the Bank's procurement guidelines for ICB contracts, an outstanding issue and the source of continuous friction between the Bank and KGM, would not be acceptable for this operation. The project's components were on the whole consistent with the project objectives. Seventy-eight percent of the estimated project costs were devoted to road rehabilitation, supportive of the first objective. While this was a small proportion of the total Turkish budget for State and rural roads, amounting to less than four percent of the budget spent for this purpose in 1993, budget cutbacks in 1994 had lowered spending for roads to little more than half of the 1993 figure. The traffic safety component, totaling about US$78.9 million, was sufficient to make a substantial impact on spending for traffic safety improvements. The component for institutional strengthening was limited to upgrading of KGM's computer and management systems and finalizing the design and installation of a Pavement and Bridge Management System started under the previous project. 3.2 Revised Objective: On August 17, 1999, an earthquake in Marmara seriously damaged the Istanbul-Ankara expressway, a vital link in Turkey's road network. The Government requested that US$24 million of project funds be diverted for the repair of the damage to this road. An amendment was approved on September 28, 1999, adding the objective of repairing the earthquake-related damages suffered by the Federal road network to the original objectives. Thanks to the extraordinary measures of KGM, roads damaged during the - 2 -  earthquake became fully passable, albeit sometimes precariously, within a week from the event. Subsequent to the reallocation of road funds, which were mainly directed to restoring the Istanbul-Ankara motorway to its pre-earthquake condition, KGM decided to go beyond the original design and advance plans to upgrade and modernize the link. Because of a sense of urgency to restore former levels of road serviceability, KGM decided to finance the works with domestic resources following more expeditious local procurement methods that could not be accepted under the Bank's procurement guidelines. Therefore, the Government requested on September 27, 2000, that the Loan Agreement be further amended to restore the original objectives of the project. The Bank approved the request and thus in the end no changes were made in the original project objectives. 3.3 Original Components: The original components of the project were: (a) The Road Improvement Program, comprising strengthening or upgrading of about 600 km of high priority state roads, and about 300 km of rural (provincial) roads and town passages at a cost of US$304.6 million; (b) The Road Traffic Safety Program that included: (i) Civil works for the improvements of road sections identified as accident black spots. (ii) Provision of road safety materials for improved traffic management for installation by KGM. (iii) Provision of equipment for the Turkish Traffic Police (TTP) for enforcement of traffic regulations in a pilot network. (iv) Medical equipment to improve the emergency response of the University Hospital at Gazi University (GU). (v) Education equipment and literature for carrying out road safety campaigns in selected schools. (vi) Extension of TTP's accident data base to other users through the creation of an Accident Data Bank. (vii) Technical assistance and training for the agencies involved in traffic safety covering the design, implementation, monitoring and evaluation of the pilot and national programs at a cost of US$78.9 million; and, (c) The Institutional Development Program, consisting of the introduction of various management systems and computerization throughout the General Directorate of Highways (KGM) and training of the staff of KGM in the areas of environmental analysis, road planning, construction and maintenance at a cost of US$5.8 million. 3.4 Revised Components: The components remained unchanged in their general outline during the life of the project except for the amendment relating to the Marmara earthquake, as explained above. 3.5 Quality at Entry: In general, the preparation of the project appears to have been adequate. The objectives of the project were consistent with the objectives of the CAS and Government priorities. The project design, focusing mostly on traffic safety and with a relatively limited institutional strengthening objectives for KGM, was reasonable, considering that this was the fourth road project that the Bank had undertaken in Turkey and there was a clear intention of making inroads on well selected institutional areas, thus reversing the trend of previous projects which ended up financing almost exclusively civil works and equipment. The preparation - 3 -  took the Bank's safeguard policies into account. While there was no Environmental Management Plan as such, the environmental objective of the project gave added emphasis to this aspect of project implementation. Efforts were made to limit the cost overruns that had been endemic in the previous project by limiting the Bank's financing to justified cost increases not exceeding 15 percent over the original contract cost. The Traffic Safety Program was very ambitious, but consistent with Turkey's poor traffic safety record, and in retrospect was especially well conceived. Shortfalls in project preparation relate to the degree of planning and preparation for the upgrading of state roads, where initial plans for the improvement of two lane roads were changed to major works for their upgrading to four lanes in some cases. However, on the whole the ICR finds the quality at entry to be Satisfactory. There was no Quality at Entry (QAE) assessment of the project (the Quality Assurance Group did not exist at the time). 4. Achievement of Objective and Outputs 4.1 Outcome/achievement of objective: The degree of achievement of the four objectives varied, as follows: (a) Approximately 356 km of two lane equivalent roads were improved out of a total main road network (motorways, state and provincial roads) of 61,000 km in 1994. Thus, while the project produced a significant improvement in the roads that were included in the project, it represents less than 1 percent of the main road network. It might be claimed, however, that the project helped to divert already excessively atomized funds to high priority projects. During negotiations, the Bank accepted to finance under the project works the Ankara-Polatli-Srivihisar road, which had already been procured. This "retroactive inclusion" was agreed because the long sought acceptance of complete adherence to the Bank's ICB procurement guidelines by the Turkish Government was considered as a major breakthrough. These works had been identified during appraisal as one of the top priorities in KGM's investment program, and out of 580 km of road sections so identified, 356 km were completed under the project. (b) The traffic safety improvements, which were sufficiently concentrated, and the consultant support, which was sufficiently effective, have made a measurable improvement in the safety of Turkey's roads. While statistics regarding traffic accidents are uncertain (the official statistics are significantly lower than the figures shown in the final consultant report), the frequency of accidents per 10,000 vehicles has at least leveled off over the last ten years and appears to be declining. While this improvement started before the project was undertaken and other factors played a role in this result, the project no doubt made a contribution to sustaining this trend by raising awareness among officials and helping to create a multi-disciplinary body at the technical level with the potential to act as the Secretariat of the political body responsible for traffic safety (i.e. the Traffic Safety Council). Even more impressive is the reduction in accidents and fatalities occurring in the 317 areas identified as accident black spots, where physical improvements were made. In these locations, accidents were reduced from a measured level of 7,389 accidents and 909 deaths in the year preceding the improvements to an average of 1,934 accidents and 54 deaths per year in the years following the improvements. (c) The improvement of the operational efficiency of KGM, which was from the outset expected to be only marginal and limited to the completion of the design and implementation of a pavement and bridge management system (started by an individual consultant during the previous project) did not take place as expected. The main reason is that the scope of the component changed substantially with the decision of KGM to design and implement a much more sophisticated Highway Information System including data gathering from a Global Positioning System (GPS) also to be procured under the project. For reasons discussed in the following section, this more ambitious component could not be completed within the life of the project. KGM did continue its planned reduction of direct hire staff, which declined from 35,100 in - 4 -  1994 (5,600 technical and administrative staff and 29,500 laborer) to 24,000 today (5,300 technical and administrative staff and 18,700 laborers). But KGM's method of contracting remains less than optimal despite four World Bank projects, as will be explained below, even though the physical results of the road improvements remain of high quality. (d) Achievement of an improvement in the consideration of environmental factors in project selection and design was limited to assuring that these factors were taken into account during the design of the components financed by the Bank. This was done appropriately. It should be noted that project preparation coincided with the consolidation of the normative and enforcement role of the Ministry of Environment in the screening and approval of road investments. By loan effectiveness KGM's Environmental Directorate was well established, fully staffed, and carrying out its duties in a satisfactory manner, thus reducing the need for any further Bank involvement. In summary, the reduction in road transport costs was satisfactory, as was the achievement of consideration of environmental factors. The improvement of traffic safety conditions was extremely significant, while the improvement in KGM's operational efficiency was minimal at best, and is considered as unsatisfactory. Taken together, the overall achievement of the project objectives can be considered as satisfactory. 4.2 Outputs by components: (a) The Road Improvement Program: State Roads: Nine major road segments were upgraded under the project for a total length of 356 km in two lane equivalents at a cost of US$221.6 million. These works, which were mainly upgrading of two lane roads to four lanes, cost an average of US$622,472 per two lane km. Furthermore, three of the sections were not completed during the life of the project (while six sections completed were started before the beginning of the project), and will require an estimated US$32.8 million more to complete with KGM's own resources, which would bring the average cost per two lane km to US$714,607 per km. This compares with the estimated 600 km of two lane equivalent upgrading planned during project preparation at a cost of US$421,300 per km. Efforts were made to limit the cost overruns that had been endemic in the previous project by limiting the Bank's financing to justified cost increases not exceeding 15 percent over the original contract cost. Another unusual and generally unsatisfactory standard procedure followed by KMG is the stretching out of the implementation time for each contract over an excessive number of years by limiting the amount of counterpart financing that was made available. In most cases, the major contracted works, which could have been completed in two years, were implemented over a four or five year period and even eight years for the section Usak-Kula. The Bank, to avoid participating in the financing of too many contracts with limited counterpart funding, refused to include new works under the loan and kept the issue of poor allocation efficiency high in the dialogue with the government. The allocation for the Rural Roads component (100 percent financed by KGM) was reduced under the project as KGM gave the first priority to the improvement of accident black spots and provision of traffic safety measures on the highways. One hundred and forty one small works, mostly town passages, were carried out at a cost of US$10.2 million, compared to an estimated cost of US$21.6 million. Road design and supervision costs were provided by KGM and were not financed from the project funds as originally planned. (b) The Road Traffic Safety Program: Preparation of this component was the result of extensive Bank discussions with KGM centering on difficulties to reconcile assumptions. While KGM officials felt that foreign consultants would not add tangible value to their road safety programs, the Bank was of the idea that only the presence of experts with an integrated multi-disciplinary view of the problem would curb a potential bias towards solutions involving expensive civil works. This, coupled with the failure in previous projects to implement those components involving international consultants, contributed to shaping a - 5 -  component with strong safeguards to ensure a balance between both views. Eventually, one of the preparation missions was unexpectedly presented with a National Road Safety Program (NP) involving a list of civil works for black spot improvements, police enforcement equipment, and emergency response services from Gazi University Hospital. The Bank, taking into account that the NP had an unusual strong ownership, agreed to finance its early stages of the black spot improvement program (up to a certain limit) and conditioned the rest of the program on the provision of the multi-disciplinary international experts to be financed under the project. In addition, it was agreed that a pilot project would be developed with the help of the consultants to incorporate the state of the art knowledge and methods of road safety in the world. The consultants (SweRoad) were engaged to work with the Government for the implementation of this program. It was agreed that a Pilot Road Safety Project (PP) would be developed for a selected section of the road network as a basis for testing what methodologies worked best, which could then be incorporated into the larger National Road Safety Program (NP). All of the phases of the program were implemented slowly, however, including the selection of the consultant, agreement on the PP, and the procurement actions needed for its implementation. The alternative of preparing the PP and advancing in the process of procuring consultant services prior to Board presentation was not considered, since there was strong pressure to process the loan as soon as possible to start addressing the most dangerous black spots in the country. The pressure was not only from KGM but from the Government generally, since the loan ended up being the only Bank operation in Turkey in FY96. This led to a strong interest from the Government for the Bank to process the loan at a rapid pace. As it became apparent that the implementation of the pilot phase would take most of the time up to the Closing Date of the Loan, the Government proceeded to implement the NP along the lines of the original agreements, but with screening by the consultants to ensure the cost effectiveness of the proposed solutions. Implementation of a PP as originally envisaged turned out to be extremely difficult to achieve. The rigidities of the Turkish system, both in terms of budgetary allocations and in terms of making decisions, are not conducive to a trial and error exercise involving a high level of discretion by KGM's technical staff. Therefore consultants assisted in the implementation of the entire safety program, including preparing designs for the improvement of accident black spots, training the KGM staff in design techniques, setting up and carrying out the training programs, and generally sharing their expertise about traffic safety issues. In summary, all aspects of the Traffic Safety Program were carried out effectively. Three hundred and seventeen accident black spots were improved with the remarkable results reported above. Safety materials were procured and installed on the highways; Gazi University Hospital's emergency response capability was upgraded with the addition of equipment and training; the Traffic Police were trained in traffic enforcement introducing a target approach; and, the Ministry of National Education prepared and installed a safety training program for the national elementary school system, including the training of large numbers of teachers and introducing traffic safety lectures in the elementary curriculum. A traffic Safety workshop was organized in cooperation with the Turkish Parliament. A total of US$75.6 million was spent for this program compared to the originally planned US$78.9 million. The consultant also prepared a National Road Traffic Safety Strategy for extending Turkey's efforts to reduce traffic accidents and fatalities. The proposed strategy was submitted for comments to government and non-government institutions before finalizing it, and the Government has now approved it as the formal guideline for road traffic safety in Turkey. The strategy proposes, inter alia, measures to make operational what is now a theoretically sound institutional setting, i.e. the existence of a High Traffic Safety Council under the Prime Minister's office. (c) The Institutional Development Component, as appraised, was of a modest scope and centered, as explained above, in finalizing work on the design and implementation of a simple Pavement Management - 6 -  System that was started by an individual consultant in the previous project. During project implementation, KGM decided to expand the scope of the component and include a rather comprehensive Highway Information System. The KGM staff made substantial efforts to design the component, but eventually the task, which required coordination with the Turkish Mapping Agency under the Turkish Army and with the Turkish Police and its own information system, proved to be too difficult to complete without specialized assistance. Eventually, it was agreed that the Bank-financed training contract with the US Federal Highway Administration would be expanded to finance two experts on information technology. The consultants helped to disaggregate the component into manageable modules, concentrating the first phase in modules related to road safety. From then on, progress on this complicated component continued to be slow, mainly because of: (i) extensive discussions on the best way to procure the needed equipment/services (goods vs. consulting procurement); (ii) scarce availability and excessive demand of qualified Bank support for a very specialized field; and, (iii) complex procedures in KGM to move forward at an acceptable pace. In conclusion, the project helped to define and dimension a state of the art tool for modernizing KGM, which is going to be procured with their own resources. Its future impact will depend on the political will of the Government to implement its plans for public sector reform, addressing problems of efficiency and accountability, which transcends KGM and the transport sector. It should nevertheless be mentioned that 56 members of KGM's staff were trained in ten areas of expertise, including Environment and Design, Bridges, Tolls, Highway Planning, Soil Rehabilitation Techniques, Information Management, Materials Technologies, Pavement Technologies, Contract Administration, and Training Activities. A total of US$1.3 million was spent for these activities compared to the planned level of US$5.7 million. The first two components were implemented satisfactorily, while the third one, which represented only about 1.2 percent of the loan, was not fully implemented and can therefore be considered unsatisfactory. Taken together, the outputs can be considered as satisfactory. 4.3 Net Present Value/Economic rate of return: Road Improvement Program: The Economic Rate of Return (ERR) for the improvement of the state roads was calculated on the basis of the World Bank HDM-3 computer model by KGM. This methodology bases the ERR on a comparison of costs and benefits, with the benefits mainly based on savings in vehicle operating costs. The anticipated ERR in the SAR for the first year's work program (Ankara-Kirrikale road improvement) was 25 percent, later modified to an estimated 18.61 percent on revised designs, while the anticipated ERRs for all of the major works taken together after specific road sections and designs were decided was an average of 25.6 percent. The actual average ERR for all of the road sections taken together was 16.4 percent. KGM explained that the principle reasons for the difference between the estimated and the actual ERRs was the increased cost of the works compared to the original estimate and the decreased volume of traffic compared to the forecast. The reduced traffic was the result of the economic crisis of 2001, which induced a significant decline of traffic over 2001 and 2002 compared to a regular traffic growth the preceding years. The results, while significantly lower than the anticipated results, are still on average greater than the marginally acceptable rate of 12 percent that was given in the SAR as the minimum rate for inclusion of a road section in the project. KGM also evaluated 152 of the completed Accident Black Spot improvements and determined, on the basis of results achieved in the years following the improvement, that 143 of these improvements were considered "effective". Considering the reduction in accidents from 7389 to 1934 per year and of fatalities from 909 to 54 per year at a cost of US$29.4 million, there is little doubt that the program was very cost effective. - 7 -  4.4 Financial rate of return: Not applicable, as KGM's operations are financed from the Government's central budget. 4.5 Institutional development impact: The institutional development impact of the project was disappointing in several respects. While KGM was already a competent functioning organization at the start of the project, there were a number of ways in which its operations could have been improved and its efficiency raised. However, the delays in working out the procurement of modern computer-based information and management system, which may have represented a lack of interest on KGM's part, prevented this from going forward although there certainly would have been time to do so during the six-and-one-half years that the project was under implementation. Also, KGM's method of contracting for major works remains inefficient and defective, as outlined above, and did not change during the life of project in spite of efforts by the World Bank staff to encourage changes. On the other hand, the improved methodologies introduced for improved traffic safety, not only within KGM but also in the Gazi University Hospital, the Traffic Police, and the Ministry of National Education, appeared to be effective and well accepted. 5. Major Factors Affecting Implementation and Outcome 5.1 Factors outside the control of government or implementing agency: Factors outside of the control of the Government or implementing agency that affected the outcome of the project were the 1999 earthquake and the general depressed economic conditions in Turkey, which reduced Government revenues and hence the availability of Government financing for the project and for KGM's operations. Revision to the project because of the earthquake was indeed minor, as mentioned in para. 3.2. 5.2 Factors generally subject to government control: The project was implemented during a period of political and financial instability, with frequent changes of Government and of KGM's management. While it would be difficult to assess all of the factors affecting the Government during this period, a stable environment could probably have led to a more meaningful dialogue on the possible creation of an autonomous traffic safety agency or any alternative that would be capable of securing the funding and creating an appropriate governance structure for road maintenance. 5.3 Factors generally subject to implementing agency control: Implementation of the major cost element of the project, the civil works for the improvement of the State, provincial and rural roads, was affected by the methodology that seems to be the standard operating procedure for KMG in carrying out contracted works. This procedure starts with bidding for works on the basis of preliminary (and sometimes obsolete) designs, followed after the selection of the bidder by a preparation of the final design that usually significantly changes the cost of the project. The contract is then re-costed on the basis of the new design using the unit costs (where applicable) that the bidder submitted as part of his successful bid. In some cases, however, items may have been added, such as bridges, for which no unit costs were included in the original bid, resulting in a negotiated price. In most cases, the re-costed contract would have a significantly higher price than the bid price. Then during implementation, further price increases may occur, as in all works where unit prices are applied to actual quantities. The Bank, aware of this procedure and the upward effect on project costs from its experience with earlier road projects in Turkey, tried to limit the cost increases by (i) requesting when possible to see final designs before bidding; and, (ii) financing only fully justified changes amounting up to a 15 percent increase for the cost of any contract. However, in cases exceeding the 15 percent limit, KMG simply financed the cost increases above the Bank's limit with its own funds, while the final cost of the works was not affected. Another unusual and generally unsatisfactory standard procedure followed by KMG is extending the implementation time for each contract over an excessive number of years by limiting the amount of - 8 -  counterpart financing that was made available. This extension is the result of poor prioritization of operations leading to approval of an excessive number of projects, and results in distributing limited allocations over too many projects. Under this Bank project, the major contracted works, each of which could have been completed in two years, were implemented over a four or five year period (which is indeed much better than the average completion time for highways projects in Turkey). This resulted in increased contractor costs, since equipment can not be used efficiently, materials must be bought in smaller quantities and sometimes held in storage, and the road deterioration during the implementation period further increases the upward cost adjustments. The Bank, to avoid participating in the financing of too many contracts with limited counterpart funding, refused to include new works under the loan and kept the issue of poor allocation efficiency high in the dialogue with the government. Alteration of both of these poor practices was finally addressed, at least partially, under covenants of the PFPSAL II with a reduction in the size of the portfolio of approved projects (thus reducing the implementation time of projects) and the preparation of a final design with a systematic audit of proposed projects by independent experts for any new foreign funded projects. In 2001 and 2002, a major downsizing of the highway and motorway projects within the Investment Program was carried out. Some 134 highway projects were removed from the Program, along with two motorways (182 km Aydin-Denizli and 78 km Iskenderun-Antakya-Cilvegozu motorways) and the 348 km Ankara-Kemerhisar section of the 432 km Ankara-Pozanti Motorway. The total cost of the projects (actually the cost minus any money spent for them) removed from the Program was roughly $2.5 billion for the motorways and $2.3 billion for the highways, bringing the total to about $4.8 billion. This downsizing reflected itself in reduced average project completion durations (estimated value of approved projects divided by the yearly allocation) from 22 years in 2001 to 13.5 years in 2002. Furthermore, the systematic audit of proposed road projects by independent experts has been incorporated into the Guidelines for Preparation of the 2003 Investment Program (Prime Ministry Circular No. 2002/4), along with the requirement that environmental impact assessments need to be done. These stipulations for foreign funded projects have been extended to all new projects, at least theoretically. 5.4 Costs and financing: While the components remained unchanged in their general outline during the life of the project (except for the amendment relating to the Marmara earthquake, as explained above), the level of actual expenditures in individual categories varied as follows: (a) The Road Improvement Program spending, planned for US$304.6 million, actually financed a total of US$232.9 million of works by the time of project completion, US$221.6 million of which was for the improvement of state roads, US$1.1 million for provincial roads, and US$10.2 million for 141 small contracts for improvements in town passages. A total of 356 km (two lane equivalent) of state roads were improved compared to the planned amount of 600 km. Most of the improved roads involved the upgrading of four lane motorways, as well as the inclusion of a major traffic interchange. (b) The Traffic Safety Program financed US$29.4 million for the improvement of accident black spots compared to US$42.1 million planned for this purpose, US$31.1 million for traffic safety equipment and materials compared to a planned amount of US$33.2 million, and US$2.9 million for consultant services related to traffic safety compared to a planned level of US$1.3 million. (c) The Institutional Development Program financed US$0.6 million for technical assistance and training for the institutional strengthening of the road administration compared to a planned level of US$2.6 million and US$11.5 million for equipment and data collection services for KGM compared to US$3.1 million planned for this purpose. - 9 -  The Turkish Government requested that the original Closing Date of March 31, 2003 be extended to December 31, 2003 so that the remaining funds in Loan No. 4048A-TU could be utilized, but the World Bank did not concur with this request as explained in paragraph 7.2. Consequently Euros 21.1 million were cancelled at project closing as the works could not be finished before the project closed. The original amount of the two loans financing the project was the equivalent of US$250 million. US$100 million of this amount was made available under Loan No. 4049-TU, denominated in dollars, and fully disbursed. The Turkish Government requested that the remaining US$150 million be made available in Deutschmarks, later converted to Euros. As of July 31, 2003, total disbursements under the two loans totaled the equivalent of US$202,319,040. US$20,251,025 equivalent of the second loan was cancelled, and the rest of the loan has been fully disbursed. The shortfall from the original US$250 million compared to actual total disbursements plus the cancelled amount resulted from changes in currency values during the implementation period. 6. Sustainability 6.1 Rationale for sustainability rating: Turkey places a high value on the upgrading of its road network, as seen from the relatively large budgets (but poorly allocated) made available over the years for this purpose. Therefore, it is highly likely that the improvements in the major roads, which accounts for most of the project costs, will be sustained. The sustainability of the improvements in road traffic safety is very likely, since the project complemented and enhanced Government efforts to respond to public pressures to reduce traffic accidents. Even within a "second best" framework of little inter-agency coordination, the adoption of road safety audits in KGM and the substantial improvements in police enforcement should have a lasting effect in traffic safety. Moreover, accident black spot improvements are physical and of a permanent nature, and therefore will have a lasting impact. The changes that were intended for the institutional strengthening component were not implemented, and therefore are not sustainable. Summarizing the foregoing, most of the project expenditures are expected to be sustained in the years ahead. 6.2 Transition arrangement to regular operations: KGM's operating procedures are well established and are not likely to change, especially as the Bank-financed project represented a very small percentage of KGM's annual budget. Also, there is already a keen interest on the part of the Government not only to sustain, but to upgrade its road network. Therefore, no special transition arrangements were considered necessary. 7. Bank and Borrower Performance Bank 7.1 Lending: The Bank performance in identifying, preparing and appraising the project was satisfactory. The project that was identified was intended to move the Bank