Turkey - Earthquake Rehabilitation and Reconstruction Project

Document of The World Bank FOR OFFICLkL USE ONLY Report No: 21855 IMPLEMENTATION COMPLETION REPORT (CPL-35110; SCL-3511A; SCPD-351 IS) ON LOANS IN THE AMOUNT OF US$285 MILLION IN TOTAL TO THE REPUBLIC OF TURKEY FOR AN EARTHQUAKE REHABILITATION AND RECONSTRUCTION PROJECT February 28, 2001 This document has a restricted distribution and may be used by recipients only in the performance of their | official duties. Its contents may not otherwise be disclosed without World Bank authorization.  CURRENCY EQUIVALENTS (Exchange Rate Effective November 2000) Currency Unit = Turkish Lira (TL) TLI.00 million = US$ 1.45 US$ 1.00 = TL690,000 FISCAL YEAR January 1 - December 31 ABBREVIATIONS AND ACRONYMS ECA - Europe and Central Asia FY - Fiscal Year GDP - Gross Domestic Product GOT - Government of the Republic of Turkey GPS - Geographic Positioning System HDA - Housing Development Administration, the Prime Minister's Office IRR - Internal Rate of Return kms - Kilometers m3 - Cubic meters m2 - Square meters m - Square meters MIS - Management Information System MOA Ministry of Agriculture MOP - Memorandum of the President MPWS - Ministry of Public Works and Resettlement PIU - Project Implementation Agency POP - Population QAE - Quality at Entry QAG - Quality Assurance Group RPA - Regional Procurement Adviser SBD - Standard Bidding Document TOR - Terms of Reference UNDP - United Nations Development Program Vice President: Johannes Linn, ECAVP Country Manager/Director: Ajay Chhibber, ECCO6 Sector Manager/Director: Ricardo Halperin, ECSIN Task Team Leader/Task Manager: Henry Boldrick, ECSIN  FOR OFFICLAL USE ONLY CONTENTS Page No. 1. Project Data 1 2. Principal Performance Ratings 1 3. Assessment of Development Objective and Design, and of Quality at Entry 1 4. Achievement of Objective and Outputs 5 5. Major Factors Affecting Implementation and Outcome 8 6. Sustainability 10 7. Bank and Borrower Performance 11 8. Lessons Learned 14 9. Partner Comments 15 10. Additional Information 15 Annex 1. Key Performance Indicators/Log Frame Matrix 16 Annex 2. Project Costs and Financing 20 Annex 3. Economic Costs and Benefits 22 Annex 4. Bank Inputs 25 Annex 5. Ratings for Achievement of Objectives/Outputs of Components 27 Annex 6. Ratings of Bank and Borrower Performance 28 Annex 7. List of Supporting Documents 29 This document has a restricted distribution and may be used by recipients only in the performance of their official duties. Its contents may not be otherwise disclosed without World Bank authorization.   Project ID: P009099 Project Name: EARTHQUAKE REHABILITATION AND RECONSTRUCTION Team Leader: Henry Boldrick TL Unit: ECSIN ICR Type: Core ICR Report Date: March 1, 2001 1. Project Data Name: EARTHQUAKE REHABILITATION AND L/C/ITFNumber: CPL-35 110; RECONSTRUCTION SCL-3511A; SCPD-351 1S Country/Department: TURKEY Region: Europe and Central Asia Region Sector/subsector: UY - Other Urban Development KEY DATES Original Revised/Actual PCD: 03/31/1992 Effective: 09/18/1992 09/18/1992 Appraisal: 04/19/1992 MTR: Approval: 07/23/1992 Closing: 06/30/1996 06/30/2000 Borrower/lImplementing Agency: GOVT. OF TURKEY/HOUSING DEVT. ADM. Other Partners: STAFF Current At Appraisal Vice President: Johannes Linn Wilfried Thalwitz Country Manager: Ajay Chhibber Michael Wiehen Sector Manager: Margret Thalwitz Paula Donovan Team Leader at ICR: Henry Boldrick Mohsin Alikhan ICR Primary Author: Anders Zeijlon; Ibrahim Sirer 2. Principal Performance Ratings (HS=Highly Satisfactory, S=Satisfactory, U=Unsatisfactory, HL=Highly Likely, L=Likely, UN=Unlikely, HUN=Highly Unlikely, HU=Highly Unsatisfactoiy, 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 S Project at Risk at Any Time: No 3. Assessment of Development Objective and Design, and of Quality at Entry 3.1 Original Objective: On March 13, 1992, at 7:42 p.m., the city of Erzincan (pop. 120,000) and its surroundings were hit by an earthquake measuring 6.8 on the Richter Scale. The motion lasted 20 seconds, and the epicenter  was about 3 kms southeast of the city. Two days later, a 6.1 magnitude aftershock ruptured an 80 km segment near the neighboring town of Pulumur. There were about 500 fatalities from collapsed buildings and some 700 persons were seriously injured. The heaviest damage occurred in approximately 60 buildings of four or more stories that collapsed, mainly along the central avenue of Erzincan in an east-west direction. Hotels, office buildings, hospitals, schools, municipal buildings and a vast number of apartment buildings were affected. There was extensive damage in the surrounding rural areas as well. Over the next few months, the Bank helped design a project to support the Government of Turkey's (GOT) Rehabilitation and Reconstruction program for the area. According to the Loan Agreement, the objective of the project was to assist the Government in carrying out its 1992-95 emergency reconstruction program for the Erzincan region through (i) rehabilitation and reconstruction of the areas affected by the earthquake, and (ii) development and implementation of seismic risk prevention and mitigation measures. The Memorandum of the President (MOP) stated that the objective of the project is to reconstruct essential housing, infrastructure and other facilities to promptly restore economic activity in the affected area, a significantly more ambitious statement of the project's objective. In any case, the project was intended to (i) reconstruct/rehabilitate essential housing, infrastructure and other facilities, and procure critically needed equipment and comrnodities to help rebuild the local economy, and (ii) provide technical training and a series of studies to reduce future earthquake vulnerability. For obvious reasons, the project was not anticipated in the Country Assistance Strategy (CAS) for Turkey, which at the time had been shaped by a series of adjustment operations in the late 1 980s, both at the macro level and in key sectors of the economy. It was envisaged that the adjustment dialogue would continue with the overall aim to promote a more outward-looking development strategy, and with sector and project lending focusing on poverty alleviation, social infrastructure and education/health. The selected project objectives cannot be judged central to the country strategy, but rather formed an enclave in the country program, justified by the need to assist Turkey to deal with the aftermath of the earthquake. The project's regional development effort for the poor and remote Erzincan area did provide a link to the overall strategy, but it was limited as the objectives were concerned with reconstruction only. There was no attempt to address whatever regional development issues there had been before the earthquake, or to set forward-looking priorities among the various sectors and development needs. The project objectives were clear and realistic, responsive to the Borrower's efforts to deal with the aftermath of a serious disaster, and clearly focused on the recovery needs for the region. Their broad scope reflected the urgency felt at the time to support the rapid recovery of the local economy, and in particular the speedy rebuilding of the physical infrastructure and provision of recovery-related inputs. They were well coordinated with the Government's overall recovery strategy which included a three-phased program with an initial Rescue and Relief phase, a Temporary Rehabilitation phase, and finally a Reconstruction and Rehabilitation phase of which the project formed a part. The project proved very demanding for the implementing agency, the Housing Development Administration (HDA) was under the Prime Minister's office, located in Ankara. It had little experience of this type of activity and of working with the Bank's procurement guidelines. After an initial period, however, HDA had built up the relevant capacity and experience and became very capable of managing the many contracts carried out under the project. 3.2 Revised Objective: An expansion of the geographical scope of the project objectives was approved by the Board of Executive Directors on two occasions. On December 19, 1995, the Loan Agreement was amended to - 2 -  include reconstruction of municipal infrastructure and flood and erosion control measures in the city of Senirkent following mud slides in the city, and the closing date was extended to December 31, 1999. On July 28, 1997, a similar amendment was made to include the reconstruction of urban infrastructure and flood control measures in the town of Sutculer following a flood in late 1995, as well as the design of seismic strengthening measures for hospitals in Istanbul and Izmir. The geographically expanded project objectives remained broadly consistent with the original objectives: disaster-related reconstruction and institutional development measures to reduce the impact of any future disasters. 3.3 Original Components: The original project components were well selected and designed to achieve the project objectives. The June 1999 QAG Rapid Supervision Assessment also concluded that the project design was sound. An initial damage assessment provided an inventory of all reconstruction needs and formed the core of the GOT's overall reconstruction program estimated at US$650 million. Following agreements on who should do what, the Bank-supported project components were designed to avoid duplication of efforts and to complement the other reconstruction activities of the Government of Turkey, and amounted to US$285 million, or 44 percent of the total program, and the closing date was set at June 30, 1996. The project components included: (a) a comprehensive urban reconstruction/rehabilitation program for cooperative and government housing, hospitals, schools, public buildings, infrastructure, shops and businesses, building demolition and debris removal, various lots of equipment, approximately 76 percent of total costs; (b) a similar but smaller program for the affected rural areas including bams, irrigation/water supply and animals/equipment, approximately 9 percent of total cost; (c) training and studies including a loss-reduction study, training for the construction industry, a study developing an earthquake Insurance Industry, a study on Urban Vulnerability, and a study on Emergency Preparedness, approximately 1 percent of total costs; and (d) consultant services for design, supervision and project management, institutional strengthening of the concerned agencies, and some supporting equipment, approximately 14 percent of total cost. Although the project components were developed in coordination with the overall reconstruction program, they came to include activities in many sectors. This made the project cumbersome to implement and difficult to supervise, with a large number of sometimes very small contracts and relatively small amounts of financing for many of the included components. The dominance of the construction activities, at 85 percent of total cost, reflected the priorities at the time to quickly carry out physical reconstruction works, but covered several sectors, involved many different Government agencies and demanded substantial coordination and supervision efforts. The administrative capacity of the implementing agency was initially not sufficient to handle the project work load, and substantial time and support was needed before implementation took off. 3.4 Revised Components. There were three significant revisions of the project components, two of which involved amendments to the Loan Agreement and introduced new physical components, and one which revised the institutional development component, i.e., the efforts to enhance Turkey's preparedness for future disasters. (1) Following the first amendment in the Loan Agreement in late 1995 the closing date was extended to December 31, 1999 and the US$20 million Senirkent Rehabilitation program was included in the project. It was designed to support recovery efforts after mud slides in the city. It was, although much smaller than the Erzincan program, modeled after the original project concept and included reconstruction of essential infrastructure, measures for flood and erosion control, disaster mitigation, and supporting equipment. -3 -  (2) After the second amendment in 1997, the closing date was further extended to June 30, 2000, US$42 million of unused funds were canceled, and a similar US$7 million program for the town of Sutculer was included following floods in the area in 1995. This also included a US3.5 million provision for the Design of Seismic Retrofitting of 26 hospitals in Istanbul and 31 in Izmir. In addition to these expansions, one sub-component originally in the project, the rehabilitation of the 60 km Refahiye-Erzincan-Tercan Road, was later dropped. The decision to drop this component followed disagreements between the Bank and the Highways Directorate on the need to complete detailed design before bidding the civil works contract, and on whether an earlier list of bidders from another project could be used as short list. (3) Originally, the following components were included in the institutional development component: (a) a Loss reduction study including: (i) revisions of the Disaster Law; (ii) revisions of the Reconstruction Law, Building Code and other laws related to construction; (iii) microzoning plan for Erzincan, and (iv) a program of assessing and improving public buildings in Turkey; (b) training of construction industry; (c) an earthquake insurance industry study; (d) an urban vulnerability study, and (e) an emergency preparedness study. None of these activities had started by 1995, and a revised component was developed and agreed on, including: (i) a study to revise the Reconstruction Law and include aspects on disasters and (ii) preparation of a public infonnation campaign and development of training materials for the construction industry. The originally included MIS system for Ministry of Public Works and Resettlement (MPWS) was also included in the revised component. During discussions of the second amendment, installation of a Geographic Positioning System (GPS) system in the Marmara Region to identify the areas of the highest seismic risk was added to the institutional development component. The agreement to expand the project to include the Senirkent program came after the original reconstruction components were almost complete. About US$170 million of the loan funds had been used up by end 1995 and it was obvious that there would be significant unused funds. The Bank had made several inquiries to the Government about canceling the remaining funds. GOT was reluctant to cancel any loan funds and proposed other uses for the funds. The Bank replied that it would be more appropriate to consider a new project on its own merits rather than add new activities to make use of the surplus funds. GOT's unwillingness to cancel the remaining funds was the dominant factor behind the request for amendment and extension, and the Bank finally agreed since the proposed activities in Senirkent (and later in Sutculer) were similar in nature to the original project activities. 3.5 Quality at Entry: The project was very quickly prepared in response to the Borrower's call for quick action to respond to the disaster. A mission to prepare the operation was sent to Turkey from April 9-24, 1992 (this mission was later upgraded to appraisal) and produced a damage assessment, a list of reconstruction needs, a proposed division of responsibility for the reconstruction activities, and proposed implementation arrangements. The Board of Executive Directors approved the operation on June 23, the loan agreement was signed on July 27, and declared effective on September 18, 1992. As a result of the rapid appraisal and approval of the project, preparation work was not as rigorous as in more typical Bank supported projects. By Board approval, there were no designs or draft tender documents prepared for the civil works, no agreed TORs for the studies and training activities, and no capacity in the implementing agency for managing the preparation of the tender documents and subsequent contracts. Nevertheless, the QAE is rated S given the unique circumstances for the project start. The QAG Rapid Supervision Assessment (July 1999) similarly concluded that the project was ready for implementation at approval as the project was a disaster relief operation. -4 -  4. Achievement of Objective and Outputs 4.1 Outcome/achievement ofobjective: The outcome/achievement of project objectives is rated satisfactory. Taking as a project objective the MOP's more expansive goal of rebuilding the Erzincan local economy, analysis of available regional GDP data for Erzincan suggest that the Erzincan economy was rebuilt in the 3-4 years following the disaster. The earthquake caused a 10-14 percent reduction in economic activity in 1992 and 1993. This was followed by a local "boom" in 1994, probably fueled by the reconstruction activities. Then came a period of higher-than-1991 GDP levels from 1995 onwards with continued growth in 1996 and 1997. The "economic loss" suffered can be illustrated by the fact that the econornic growth in Erzincan was limited to 9.4 percent between 1991 and 1997, while the corresponding figure was 14.4 percent for eastern Turkey, and 33.5 percent for Turkey as a whole. It is difficult to say how much of the recovery was due to the reconstruction program. The Erzincan region's GDP was about US$260 million in 1992. There is therefore little doubt that the Bank's funding of $180 million for the region, as well as the originally planned overall US$650 million reconstruction package, spent over 3-4 years were very large interventions. It would appear safe to conclude that the Bank-funded construction activities contributed to rebuilding the local economy. In addition to the direct economic benefits, the construction activities are likely to have triggered and accelerated the private sector-driven recovery and expansion of the local economy in the area. There is no data available describing this process, but there is anecdotal evidence that many of the entrepreneurs and richer parts of the population left Erzincan after the earthquake. Many of those who could leave did so, and settled in other parts of Turkey, mainly in the economically more active western parts. Those who left were in many cases owners and operators of factories and businesses. They often appointed caretakers or managers to look after the businesses. Demographically, therefore, the earthquake caused out-migration of the richer and most enterprising people which in turn was matched by some in-migration from the surrounding areas. The forward-looking institutional objectives seeking to reduce Turkey's vulnerability to future earthquakes were only partly achieved. Some of the activities were completed, others were partly completed and will be completed on their own or as part of other Bank-supported projects, and some were not undertaken. Please see 4.2 (ii) below for details. The objectives for the smaller parts of the project that were added in 1995 and 1997 were substantially met. The rehabilitation programs for both Senirrkent and Sutculer have been completed and both their municipal infrastructure and actions to reduce risk for and vulnerability to future floods and mud slides have been completed. Similarly the smaller added components of retrofitting hospitals in Izmir and Istanbul, and the establishment of a GPS system in the Marmara area were completed. 4.2 Outputs by components: The components (original and later added) designed to meet the physical objectives were generally successfully implemented while performance of the components supporting institutional strengthening was mixed. The summary provided here is illustrated in more detail in Annex 1. (i) Physical objectives under the Erzincan program: this component is rated Satisfactory as almost all components were implemented as planned. It would have earned a Highly Satisfactory rating had it not suffered from initial delays. Most of the construction activities did not start until - 5 -  Spring/Summer 1994, i.e. two years after the earthquake occurred, and most were completed during late 1995, almost 4 years after the earthquake. This delay did not stop the project from reaching its objectives, but it certainly delayed it. Under the housing component, 252 Government apartments were reconstructed compared to the 250 originally planned. However, the 1,131 units of cooperative housing actually repaired or reconstructed fell short of the 3,640 units planned. This was due to inaccuracies in the original damage assessment which formed the basis for the Bank's appraisal. The actual number of eligible damaged units turned out to be a much smaller number. Under the community facilities component, three hospitals were reconstructed with 475 beds planned as compared to the 350 beds for patients and 150 units of nurses' lodgings that were actually built. The hospitals are fully operational and well maintained, although there is a shortage of specialized doctors (as is the case in all of Eastern Turkey). Another 500 units of housing for medical staff were completed as planned, as was the delivery of equipment to the hospitals. The planned rehabilitation of a health post was dropped. Equipment for the area schools were delivered as planned, as was the reconstruction of 300 units of teachers' housing. Under the shops and business component, 387 shops and 275 offices were reconstructed in six shopping centers, another 512 shops and 143 apartments were repaired in 39 structurally rehabilitated buildings. About 95 percent of the units are used, and most of them are rented out by the original beneficiaries. Under the public buildings component, municipal offices, the Adult Education Center, the Horticulture Research Center, the National Education Center, four municipal workshops, the Rural Affairs Building and a bus terminal were reconstructed, mainly as planned although the rebuilding of a sports stadium wall was dropped. Civil works and equipment under the infrastructure components were carried out and delivered much as planned and included inputs and investments for many sectors of the municipal economy. The main difference from the original plans was that the approximately 100 kms. of road repairs originally planned never took place. The building removal and debris component took place as planned. The rural component was implemented very quickly even if the originally planned number of barns to be constructed (4,100) was larger than what was achieved (2,885). Again, this was due to inaccuracies in the original damage assessment used for appraisal. The barns were built very quickly after loan effectiveness, by the beneficiaries themselves with lump sum payments made for construction materials. The repair of the sugar factory chimney and the delivery of equipment to the Civil Defense Directorate took place as planned. (ii) Institutional development component. As mentioned in 3.4 above, this component was scaled-back and revised in 1996. It is rated Unsatisfactory since many of the originally intended activities never took place, and as only part of the work actually done has led to the intended actions or follow-up. The study to revise the Reconstruction Law and include aspects on disasters in it was initiated in October 1997, and completed in draft in September 1999. There have been some modifications in the MWPS bylaws governing construction but no legal changes have been presented to parliament. The preparation of the public information campaign and development of training materials for the construction industry started in June 1998 and were completed in June 2000, but the campaign and training activities could not be initiated in time to take advantage of Bank agreement to fund them under the Erzincan loan and it was decided to fold it into similar activities under another Bank-financed project. The Installation of the GPS system in the Marmara Region was completed by November 1999 (most parts of the system were in place when the August 1999 earthquake hit the region, and provided valuable inputs to assess the situation). In addition, a component to automate and replace the MWPS management information system -6-  was included in the original project and was largely implemented as planned. Some of the specially designed software was rejected by MWPS, but off-the-shelf software was procured and is working well with the installed hardware. (iii) Components in the first amendment of the Loan Agreement, January 1996 - Senirkent Reconstruction Program. This project component is rated Highly Satisfactory as all parts of it were implemented on a timely basis and according to plans. The infrastructure works and equipment were completed as planned in October 1999, and the preventive flood and erosion control components were completed as planned in June 1999 and included 3.7 million trees planted, 24 thousand meters of fencing, and 2,603 hectares planted. (iv) Components in the second amendment of the Loan Agreement, August, 1997 - Sutculer Reconstruction Program and the Assessment and Design of Hospitals in Istanbul and Izmir. These components are rated Highly Satisfactory. All parts of the reconstruction program were completed by June 2000 (except for the water treatment plant component which was dropped because it could not be technically justified), including the preventive flood & erosion control activities to protect a 860-hectare erosion control zone. The design work for the Istanbul and Izmir hospitals was completed by September 2000. 4.3 Net Present Value/Economic rate of return: There was no NPV or IRR calculated for the project during preparation/appraisal. During the ICR preparation, a limited ex-post cost-benefit analysis was carried out for the housing and shops/business components, and the per-unit costs for the hospitals component were assessed. Contracts and other materials available for the other components were analyzed to see if least-cost solutions were sought. Details of the calculations are found in Annex 3. A: The project's housing component amounted to US$44.2 million or 22 percent of total project cost. For these funds, a total of 2,183 apartments of on average 90m were reconstructed or repaired at a cost of $20,200 each. The market value of each apartment is assessed at $200 per month. Comparing the assessed market rental value with the costs for the three reconstruction/repair contracts yields a rate of return of 11 percent. B: The project's shops and businesses component amounted to US$22.5 million or 11 percent of total project cost. A total of 387 shops were reconstructed in six shopping centers, 512 shops rehabilitated in 39 different buildings, 275 offices reconstructed in six shopping centers, and 143 apartments reconstructed in 39 different buildings, implying a renovation cost of US$16,500 for each unit. The rental market value is assessed at $130 per month for each shop, $50 per month for each office, and $200 per month for each apartment. Average rental value for each unit in this component is $155 per month. Comparing the assessed market rental value with the costs for the three reconstruction/repair contracts yields a rate of return of 7 percent. C: The project's hospitals component amounted to a total cost of US$35 million or 17 percent of total project cost. Three hospitals with a total of 350 beds were reconstructed, furnished and equipped. This included 150 associated nurses lodgings. This corresponds to 85m' per bed, $59,000 per bed in rehabilitation/reconstruction costs, $41,000 in cost per bed for fumiture and equipment, or a total cost per bed of $100,000. Per square meter, the costs are $683 for the reconstruction, and $476 for furniture and equipment, in total $1,059 per square meter. The costs for this component are very low and the standards - 7 -  appropriate. The 85m2 per bed compares to a norm of lOOm' in many industrialized countries. The $1,059 total cost per square meter compares to a figure of $1,700 for a recently completed Bank-supported medical facility in Estonia (excluding beds). The total cost per bed, at $100,000, compares to an industry average of $250,000-350,000 for first rate hospitals in many industrialized countries, with $120,000 being the benchmark for less modem facilities in poorer countries. A review of available data for components D, E, and F suggest that the activities pursued by the project represented the least cost solutions for the outputs and associated benefits: D: Equipment, total cost US$22 million = 11 percent of total project cost E: Other civil works, infrastructure, total cost US$65 million = 33 percent of total cost F: Consultants services, total cost US$12 million = 6 percent of total project cost 4.4 Financial rate of return: There was no financial rate of return estimated for the project. 4.5 Institutional development impact: Overall, the project did not make significant contributions to the ultimate institutional development objective: to improve Turkey's preparedness for future natural disasters. A more thorough implementation of the activities designed to limit the damage of future earthquakes would have been desirable. The smaller preventive components for the Senirkent and Sutculer programs, the Marmara GPS system, and the planned retrofitting of the hospitals in Istanbul and Izmnir do compensate for some of the earlier neglect in the project. In summary, the institutional development aspects were neglected as the original construction activities were carried out, but pursued through a scaled-down program later, and through the amendments of the project. HDA/PIU grew into a skilled organization and became a good training ground for project management including procurement and contract management. 5. Major Factors Affecting Implementation and Outcome 5.1 Factors outside the control of government or implementing agency: There were several problems during the first three years of the project with the selection and work of the consultants who assisted HDA with design and construction supervision for the four main construction components. These difficulties significantly contributed to the initial implementation delays. By mid 1993 all four consultants were contractually late in the preparation of bidding documents, and frequently submitted the wrong or incomplete versions of the documents to HDA. Most of these problems were eventually solved, but HDA in 1996 terminated the contract for one of the consultants and converted the security bond (bank letter of guarantee) to cash. This led to a lawsuit against HDA which is still underway and to which HDA has issued a counter claim. In addition, three construction companies have brought forward six other lawsuits against HDA, mainly related to the applicability of the correct price index in payment certificates and the claims for time extensions. HDA has won four of the six cases, lost and appealed one, and is awaiting the outcome of one. It is not expected that the lawsuits will have any impact on the project outcome. - 8 -  5.2 Factors generally subject to government control: There was poor Govermment commitment to the institutional components attempting to strengthen Turkey's preparedness for future earthquakes. The project was clearly viewed as a reconstruction program, and there was little interest in the broad studies and training program included in the original project design. None of the planned activities had started by the end of 1995. The modified program agreed on in 1996 has been implemented, but interest in the findings of the studies and in the information campaign and training materials prepared has been limited. However, GOT interest in the Marmara GPS system and in the envisaged renovation of hospitals in Istanbul and Izmir has been more positive. The selection of HDA as implementing agency was a compromise after the first proposals to use MPWS failed. The Bank's identification mission in April 1992 recommended that the directorate in MPWS responsible for disaster management be made the implementing agency, as this directorate was to play a central role in the overall reconstruction program. Within weeks, however, it became clear that MWPS did not agree to the envisaged modalities of the Bank supported activities (see section 7.4 below) and as a result the government suggested that the Bank supported part of the reconstruction program be implemented by the smaller and more flexible HDA. Although HDA became a very competent implementing agency, it is still regrettable that the Bank and MWPS did not manage to establish a better basis for cooperation. The ministry is the key agency dealing with many aspects of disaster management (including construction standards), and the project components dealing with Turkey's preparedness for future earthquakes would have had a greater chance of success had they been part of the ministry's own work. The project came to cover many sectors of the region's economy, and many government agencies had to be involved in the work. This meant that HDA initially had to spend significant time and effort to establish a series. of protocols with the agencies it had to work with. In the end, the Government's intra-agency collaboration worked quite well as far as the construction activities and procurement of goods were concerned. The Ministry of Agriculture quickly administered the construction of barns and procurement of animals, the Ministry of Education procured and distributed the school materials, etc. But it also meant that HDA became overloaded with a lot of detailed questions about Bank procurement and procedures from a vast number of agencies each struggling to implement their part of the project. One Bank report in 1994 noted that HDA was swamped with these kind of questions and that all resources went into the day-to-day handling of them. The staff had little time or energy to focus on the overall implementation of the project, and this is possibly yet another explanation to why the institutional strengthening components were neglected for so long. 5.3 Factors generally subject to implementing agency control: HDA became a very efficient implementing agency. The successful completion of the reconstruction efforts in Erzincan, and later in Senirkent and Sutculer, is a result of HDA's competent project management and handling of contracts and procurement. This is particularly true for the time from 1995 onwards. The early years of project implementation were difficult for HDA. Without the necessary staff, it attempted to get a large number of complex construction and infrastructure operations in motion, employing Bank rules and procurement guidelines, while also negotiating protocols with the larger ministries dealing with specific components of the operation. It took a long time for HDA to get the right key staff, i.e. with knowledge of procurement and contract management, and also to secure the services of -9-  consultants to help as needed with the preparation of bidding documents and supervision of the construction activities. 5.4 Costs andfinancing: Both the funding needed to complete the original project components and the speed by which disbursements would take place were overestimated during project preparation. By 1996, the original physical components of the project had been completed at a cost of about US$180 million, compared to a loan amount of US$285 million, implying that the original costs were overestimated by about 58 percent. Most of the over budgeting (about 80 percent of the US$105 million of excess funds) originated from four subcomponents; (i) Infrastructure - US$26 million of the US$55 million budgeted was spent, (ii) Housing - US$47 million of US$69 million was spent, (iii) Rural reconstruction - only US$7 million out of US$25 million was spent, and (iv) PMU/project design and supervision - US$20 million out of US$38 million was used. The speed with which the funds would be disbursed was also overestimated. For the $285 million made available, the MPWS envisaged that 49 percent of the funds would be spent during FY93, with the remainder at a declining rate during the next three FYs at 33 percent, 10 percent and 7 percent respectively. In fact, the disbursement-profile of the spending of the $180 million actually spent before the project restructuring was much more back-loaded, with only 9 percent and 7 percent spent in 1992 and 1993, respectively, and 49 percent, 30 percent and 5 percent during the following three years. There are several reasons why the funding needs and speed of their use were overestimated. First, the rapid preparation did not allow for realistic plans and cost estimates. Second, the pressure to get the recovery program started fast led to a situation where generous amounts of funding were made available to make sure that shortage of funds would not hamper the recovery efforts. Third, the enthusiasm and urgency to implement the recovery efforts were mistaken for actual implementation and absorption capacity. While such situations should be avoided in the future, the need to respond to urgent requests after natural disasters makes it difficult to get cost estimates as accurate as are available in the normal Bank investment operation. The most obvious implication of the over-budgeting was that a large amount of funds was underutilized. This became clear in 1995 whereupon the Bank advised GOT to cancel the extra funds. However, GOT did not agree to this, and instead raised various ideas about how the funds could be used. A situation emerged where the idle funds were seen as available credit to be drawn on for any disaster related activities. This resulted in the two amendments: to add the Senirkent, Sutculer and Hospital Retrofitting components. The costs for the added components (and the revised institution strengthening) were more realistically assessed, and parts of the loan were eventually canceled on three occasions. The first cancellation amounted to US$42 million and was dated May 7, 1997, the second of US$19.5 million took place on August 9, 1998, and the final one of US$17 on April 5 1999, in all bringing the total amount of the loan down from US$285 to $206.5 million. 6. Sustainability 6.1 Rationale for sustainability rating. The sustainability for the project as a whole is rated Likely, and includes three parts; (i) the project's physical components rated highly likely, (ii) the reduced institutional component rated likely, and (iii) the abandoned parts of the institutional component that were never carried out. The physical components (construction/rehabilitation of buildings and infrastructure, equipment) sustainability is highly likely. The infrastructure/investments provided to the three municipalities are integral parts of the cities' - 1 0 -  provision of services to its inhabitants, and are being used and maintained. Virtually all housing/apartments supported by the project are lived in with market-based rents used for the cooperative houses and subsidized rents for the Government owned buildings where the civil servants are typically charged US$75 per month (as compared to a market rent of US$200 per month). About 95 percent of the shops and offices provided are used, with the majority being rented out by the original beneficiaries. The three hospitals in Erzincan are fully functional and well maintained, the main problem being to attract suitable medical specialists (an issue common to all of Eastern Turkey). The seismic-resistant design developed for hospitals in Istanbul and Izmir is likely to be used in the future. The sustainablity of the project's reduced institutional parts is rated likely since the GPS system for the Marmara region as well as the management information system for MPWS are operational. Similarly, the training and information materials developed for the public and the construction industry are likely to be used, and the proposed revisions of the reconstruction law are likely to be considered. Some of the abandoned parts of the institutional component are being continued under the MEER project, i.e. revision of key legislation and the development of a disaster insurance industry. 6.2 Transition arrangement to regular operations: The transition for the physical parts of the project is complete. The respective agencies and municipalities manage, operate and maintain the equipment, buildings and investments made. Similarly, the Marmara GPS system is handled by TUBITAK, a scientific and technical research council, and the MPWS' MIS system is fully handled by the ministry. The training and information campaign materials have been handed over to MPWS, as has the report listing desireable revision to the Construction law. It is recommended that the Bank, possibly as part of the MEER project, support the future use of these materials and recommendations for law amendments. Similarly, the Bank should support the continued revision of laws and regulations, relevant studies and the development of an insurance industry for disasters. Finally, discussions are underway with the Government about support for reducing seismic risk in Istanbul, an effort which could include retrofitting of some hospitals in Istanbul and Izmir. 7. Bank and Borrower Performance Bank 7.1 Lending: The Bank quickly responded to the disaster. The earthquake took place on March 13. On March 15, a meeting was held which concluded that the Bank should offer assistance to prepare an operation with limited objectives and with a realistic time schedule to (i) do a comprehensive damage assessment, (ii) help fill the financial and technical requirements to restore the productivity of the area, and (iii) take actions to mitigate against damage in case of any future earthquakes. Letters outlining the Bank's possible support were sent to the President and the Prime Minister on March 16. The offer was gratefully accepted, and the Bank launched an identification/pre-appraisal mission in April (9 - 24) which was later upgraded to appraisal. This seven-person mission produced a comprehensive assessment of the damage, and identified the scope and content of the project. Disagreements about the implementation arrangements, however, made it necessary to launch a post-appraisal mission (May 31 - June 9). During this mission it was agreed that the Bank-supported activities would be implemented by the HDA, while MWPS would be in charge of the part supported by an ECU150 million credit from the Council of Europe. After Board approval, the Bank sent a third mission which worked with HDA to strengthen its implementation capacity, comprising one HQ staff and one Procurement Expert from the Bank's Ankara Office. The project was for obvious reasons not envisaged in the Bank's Country Strategy for Turkey, but supported the Government's reconstruction efforts and was well coordinated with GOT's overall -11 -  post-earthquake reconstruction program. The work carried out by the Bank staff was of high quality and provided a comprehensive assessment of the reconstruction needs. Appraisal of the Government's commitment was correct for the reconstruction activities, but not realistic for the institutional components for which the commitment was overestimated. The assessment of HDA's ability to implement the project underestimated the problems this agency came to face in implementing a complex project in a remote part of Turkey. Similarly, the plans for establishing a local PIU in Erzincan proved very hard to carry out. The project design was straightforward and appropriate even if it spanned several sectors of the local economy and came to involve many agencies. The funding amount provided proved overly generous and the disbursement profile too optimistic. 7.2 Supervision: Overall, the Bank's supervision efforts were satisfactory, and can be divided into three phases. During the initial two years of the project, the Bank's supervision efforts focused on assisting HDA with the hiring of consultants for the design and supervision of reconstruction work and for the preparation of bidding documents. The Bank also put significant efforts into familiarizing HDA staff with Bank procurement and disbursement procedures. During this phase, the emphasis was on dealing with the delayed construction efforts while the institutional objective of the project was largely neglected. In the project's second phase, 1995-96, the originally planned reconstruction components were largely completed, and attention was again turned towards the second project objective, and towards reallocating the un-used part of the loan proceeds. Finally, during the third phase in 1997 and 1998, the scaled down institutional component and the Senirkent and Sutculer reconstruction programs were implemented, and the Bank's supervision focused on these activities. Throughout implementation, a large part of the supervision and day-to-day task management were carried out by Ankara office staff, supported by periodic missions from HQ. This arrangement proved very helpful to address issues related to procurement and contract management, and was much appreciated by the borrower. One shortcoming in the supervision (as noted in QAG's RSA report) was that reporting focused on the output level rather than on outcomes, i.e. on the construction activities rather than on the re-established economic activity in the affected areas. Implementation progress was adequately reported and implementation problems identified and addressed. The performance ratings given during supervision (S throughout) were realistic. More attention should have been paid to the lagging behind of the project's institutional strengthening components during 1993-1995. The advice given to the implementing agency was of high quality and appreciated. All loan covenants were enforced and complied with. The skills mix and continuity of Bank staff during supervision was good. 7.3 Overall Bank performance: Overall Bank performance was satisfactory. During preparation the Bank acted quickly, but also underestimated the problems with having an implementing agency not used to Bank projects dealing with a project covering many sectors and many government agencies. During supervision, the Bank provided good support to the reconstruction efforts but initially neglected the institutional development objective (preparedness for future earthquakes). Once the institutional component started, supervision of it was very good as was the supervision of the other added project components. Borrower 7.4 Preparation: The borrower's performance during preparation was satisfactory. The government provided - 12-  adequate inputs for the preparatory work, and the Bank supported activities were shaped into a consistent part of the overall reconstruction efforts. The only complication concerned the selection of the implementing agency. It was initially assumed that MPWS would be implementing the project as its directorate for disaster management played a key role in handling the overall reconstruction program. However, disagreements between the ministry's practices and the Bank's guidelines proved hard to overcome. The ministry wished to use its own procedures for the design work (to be done by the ministry) and construction (using MPWS procurement rules), which the Bank did not judge satisfactory. The Bank further argued that MPWS's organizational structure did not permit accountability and responsibility to be clearly assigned. Some modifications in MPWS' procedures were agreed on, but it was judged to be in the best interest of implementation if the Bank supported part of the reconstruction program was implemented by HDA rather than MPWS. 7.5 Government implementation performance: Government performance during implementation is rated satisfactory. The construction and infrastructure components were implemented with full Government support. The cross-sectoral and comprehensive nature of the project made it difficult at times to achieve focused Government implementation. Each respective agency was supposed to carry out the implementation and procurement of its own component (school materials by Ministry of Education, etc.) with HDA providing overall guidance as needed. In fact, HDA ended up handling most of the detailed implementation. There was little Government interest in the project components aiming to strengthen Turkey's preparedness for future earthquakes. The broad and ambitious program designed for the original project was not pursued during the first three years of the project. Implementation only started after the program had been severely scaled back and after Bank staff invested large amounts of time in developing TORs and advancing the program. 7.6 Implementing Agency: HDA's performance is rated satisfactory. This assessment is based on (i) consistently highly satisfactory performance since 1995-96, (ii) initial delays in project implementation as well as in solving staffing issues during 1992-1994, (iii) disregard for the project's institutional strengthening components until their restructuring in 1996. HDA grew into a very effective implementing agency after the first few years of work. The management of HDA was generally efficient, including the financial management. Following a period of inadequate staffing during the first years of implementation, the agency developed the right staffing and skills mix. There were no remarks related to the execution of a US$60,000 Trust Fund from UNDP. The restructuring of the project to include Senirkent, Sutculer and the design for renovation of hospitals in Izmir and Istanbul, and the implementation of these new activities was handled very well. During the first few years of implementation (1992-94) HDA found it hard to carry out the assigned tasks. One year after effectiveness, the Bank advised the PIU to do the following urgently: (i) complete the staffing of the PIU, and strengthen the procurement function of the PIU with a TA contract, (ii) complete the staffing of the Erzincan PIU, (iii) push the consultants to get the bidding documents ready for the four major construction components, (iv) to finalize the Turkish translation of the SBDs, and (iv) complete the audit of the 1992 accounts which was overdue. The absence of experienced staff in the areas of procurement and contract management proved particularly harmful. At this time a lot of effort went into establishing the official protocols governing the PIU's relations with the other parts of the Government - 13 -  responsible for their respective parts of the project. As the volume of activities increased, the PIU found itself flooded with many smaller questions and issues, often of a very technical nature, from other agencies, contractors, consultants, etc. It was also difficult for a long time to get staff for the PIU in Erzincan due to unrest in the area. HDA showed little interest in the implementation of the studies and training components to enhance Turkey's preparedness for future earthquakes. Once these components were restructured, HDA effectively implemented the studies agreed upon. A review of the project's procurement by the ECA RPA's office in mid-2000 did not reveal any major issues but stated the following: (i) Legal proceedings are under way with 4-5 contractors on price adjustment; arbitration process as per contract was not followed; (ii) Bank agreed to restricted advertisement (not through Development Business) but information was forwarded to Embassies only; (iii) Construction of barns was implemented under the project, but was not provided for in the procurement schedule of Loan Agreement, (iv) no information on aggregate value for different method to establish compliance with legal agreement, (v) consultant selection for Project Management Automation (value>$200,000) included only four local institutions, (vi) procurement of stabilization equipment was not clear; repeated sole sourcing was awarded, (vii) role of TUBITAK, a research institute, as beneficiary and provider of services for the Marrnara GPS system was not clear, (viii) extension of contract with one company was approved without clearance of the RPA, (ix) no post reviews conducted. 7.7 Overall Borrowerperformiance: The Borrower's overall performance is rated satisfactory based on the assessments above. 8. Lessons Learned * Give preventive and institutional development measures a more prominent role: The project's objective to strengthen Turkey's preparedness for future earthquakes was not fulfilled as almost all attention in the Erzincan project was focused on the reconstruction efforts to repair damage done. In August 1999, seven years after the Erzincan earthquake, the Marmara Region was struck by a serious earthquake with extensive damage and many casualties. Obviously, forceful implementation of the project's comprehensive training and studies program that was designed to mitigate against damage in case of future earthquakes would have served Turkey well, and a vigorously pursued loss-minimizing program during the 1990s would have yielded substantial benefits. In the case of any future operations seeking to meet institutional or preparedness objectives, the following measures could be considered: (1) Make sure that the preventive/forward looking project components have their own implementing agencies and interested audiences, i.e., avoid putting these components as add-on's to a larger and implementation intensive reconstruction program. (2) Consider putting progress on the preventive/forward looking activities as conditions for disbursement on the reconstruction components. (3) Develop self-standing operations with preventive/forward looking objectives, i.e., not linked to any reconstruction activities. (4) Prepare the preventive/forward looking components in more detail (at least agreed-on TORs) before Board approval. * Maintain reasonable Quality at Entry even for urgent projects: The Bank made a significant amount of funding available within six months after the earthquake. However, funding in place proved not to be the key constraint. Instead, poor implementation capacity as well as limited absorptive capacity locally meant that it took two years for the reconstruction activities to start, - 14 -  and three years to finish. More attention to preparation, development of bidding documents and the creation of a better equipped PIU would have contributed to speedier implementation. The urgency and pressure to move fast after disaster strikes make it tempting to focus on making money available fast, and to push the more difficult problems aside. Would an extra six months of preparation of the original project have led to the completion of reconstruction after two years instead of three? * Local involvement is key to success. The rural component rebuilding barns was a big success and quickly implemented as it closely involved the beneficiaries. It would have been interesting to know if closer community involvement in the identification and construction activities would have resulted in speedier implementation of other activities. * Seek to limit the number of sectors included: The reconstruction activities covered almost every sector of the local economy, leading to a large number of contracts worth often small amounts of funding available for each activity. It would have been beneficial if the division of responsibilities between the Bank and other actors had left the project with 2-3 sectors to focus on. The initial delays could have been reduced if focus had been on, say, housing and hospitals only. * Avoid extensions: Rather than extending the project and adding on two major components, it might have been better to allow the project to close after the original construction activities were completed. The extensions and added components (which were well designed and implemented) were really the result of excess funds being available. These were also activities of very small size and scope which would not have justified Bank intervention. The result was that a relatively small amount of loan proceeds (about US$20 million) was used during the last four years, while the budget costs for this to happen from the Bank's side included supervision for four years and the drafting of two extensions. All in all, it would have been better to close the project in 1996 and identify new operations in response to future disasters on the basis of the merits of those cases. 9. Partner Comments (a) Borrower/implementing agency: See attachment. (b) Cofinanciers: No comments were received from the two organizations involved in the Erzincan Reconstruction Program, i.e., the UNDP and the Council of Europe. (c) Other partners (NGOs/private sector): N/A 10. Additional Information N/A - 15-  Annex 1. Key Performance Indicators/Log Frame Matrix Outcome / Impact Indicators: Restoration of production capacity and Regional GDP data at higher-than employment in Erzincan local economy pre-disaster levels from 1996 onwards Restoration of production capacity and No data available. The urban economy employment in the Senirkent and Sutculer appears restored in both cities local economies Strengthening the country's capacity to deal No data available with future disasters Output Indicators: A: Onginal Project Components A: Orginal Project Components 1: Housing 1: Housing (a) Housing reconstruction of 250 (a) 252 Govemment apartments Government apartments (80m2 each) and reconstructed, 354 cooperative apartments 940 cooperative apartments (100m2 each) reconstructed (b) Housing rehabilitationlrepair of 2700 (b) 777 oooperative apartments repaired cooperative apartments (100m2 each) 2: Community facilities 2: Community facilities (a) Hospital reconstructionrehabilitation (a) Hospital reconstruction/rehabilitation Hospital 1: 275 beds Hospital 1: 150 beds and 60 beds in nurse Hospital 2: 100 beds lodgings (10595m2) Hospital 3: 100 beds Hospital 2: 100 beds and 60 beds nurse Staff apartments: 500 x 75m2 lodgings (10134m2) Equipment: for 475 beds Hospital 3: 100 beds and 30 beds nurse Rehabilitation Health Post lodgings and a heating facility building (9752m2) 500 staff apartments x 75m2 Equipment delivered, except two contracts stil pending Rehabilitation of Health Post canceled (b) Schools reconstruction/rehabilitation (b) Schools reconstrucion/rehabilitation Equipment: 1410 classrooms Equipment: Delivered without delays Teachers' housing: 300 units 300 teachers' apartments 3: Shops and Businesses 3: Shops and Businesses Reconstruction: 1300 units x 60m2 39 buildings in 10 groups Rehabilitation: 2400 units x 80m2 3 shopping centers 4: Public buildings 4: Public buildings Municipal offices reconstruction: 100 x Municipal offices, 6,000m2 100m2 Municipal offices rehabilitation: 400 x 100m2 Adult Education Center 2,900m2, Bus terminal: 5000 pass/day Horticulture Research Center 3,675m2, Municipal workshops: 4 National Education Center 3,792m2 Sports arena rehabilitation Municipal Workshops 3,100m2 Rural Affairs Building 3,255m2 Bustermal 1253m2 - 16 -  5. Infrastructure 5. Infrastructure (a) Municipal works (a) Municipal works Water supply: repair 650 leaks, 1200m of Wastewater treatment plant damaged pipes in 60 loations, Solid waste Landfill reconstruction 1300 inspection chambers and house Sewerage, drainage, water suppty and urban connections, equipment for maintenance roads Ditch digging Water meters and leak detection. 265 kms Urban streets: resurface 20 kms inspected, 300m3/hr of leaks detected and Solid waste landfill repaired Storm drainage Traffic lights at 6 junctions Sewage treatment plant Electric network Traffic light system rehabilitation Electrical distribution line repair 5 km overhead and 1.5 kms underground Electrical distribution transformers: 33 Rehabiritation transformation stations: 4 Improve network Buildings: 4 (b) Equipment (b) Equipment 10 dump trucks, 3 compression tnucks, 1 forklift, 2 loaders, 2 excavators, 1 compactor, I bulldozer, 1 grader, I roller, 10 pick-ups, 1 sewerage truck, I crane- mounted truck, 1 landfill compactor. (c) State and provincial roads; (c) State and provincial roads; 100 kms of road repair canceled Equipment 8 dump trucks, 2 truck tractors, 1 tree transplanting machine, 10 camp trailers, 2 sprinkler trucks, 10 forklifts, 2 generators, 1 crawler tractor, 10 hydraulic breakers, 1 asphalt plant, 2 road marking machines, 2 bridge maintenance platforms, 5 mobile cranes, 1 wireless communication system 6. Building demolition 6. Building demolition Demolition: 360,000m2 or about 100 lrge No data available/completd as planned buildings, 400 small buildings and 3,700 dwellings Debris removal: 625,000 m3 7. Rural development 7. Rural development Cattle bams: 2500 x 120m2 2885 bams built Storage bams: 1600 x 100m2 Irrigation: 27 village groups Animals: 10,900 Water supply 22 village groups MOA Heavy Equipment Rural Roads 7 groups Irrigation repair Animals 10348 Water supply repair Equipment: 2 vibrating rollers, 1 bulldozer, 1 grader, 2 wheel loaders 8. Sugar factory, civil defense 8. Sugar factory, civil defense Repair sugar fadory chimney Sugar factory chimney repaired Materials, equipment for the dvil defense Equipment: 5 rescue vehicles, 2 department decontamination trucks 6 contracts for 12 lots of equipment -17 -  9. Project management automation 9. Project management automation Replace MWPS management information Completed, but some software not accepted system and contract with supplier terminated. Off the shelf software is used so the system works. Only some specific software is not in place - first 4 lots of 12. (adm, invest mon/eval, regulations, documents). The 486 PCs have been replaced. 10. Training and studies 10. Training and studies (a) Loss reduction study: (i) revisions of Disaster Law; (ii) revisions of (i) A study to revise the Reconstruction law Reconstruction law, Building Code and other and include aspects on disasters was laws related to construction; (iii) microzoning completed in draft September 1999; plan for Erzincan; (iv) program of assessing (ii) a Public Information Campaign and and improving public buildings in Turkey Development of Training Materials for the (b) Training of construction industry Construction Industry. Materials prepared (c) Earthquake Insurance Industry study and campaign was planned, but no (d) Urban vulnerability study campaign or training activities to date; (e) Emergency preparedness study (ih) Installation of a GPS system in the Marmara Region completed by November 1999 11. Project management 11. Project management B. Components added at first project B. Components added at first project modification modification 1. Senirkent Rehabilitaton Program 1. Senirkent Rehabilitation Program 1.1 Infrastructure works and equipment 1.1. Infrastructure works and equipment (a) water supply and distribution network, (a) Completed as planned 10/99 wastewater network, storm water drainage network, and roads (b) Electrical distribution network (b) Completed as planned 12/99; 6 transformer stations, 489 poles, related network and street lighting (c) Telecoms network (c) Canceled 1.2. Flood & erosion Control 1.2 Flood and Erosion Control Completed as planned 6/99, 3.7 million trees planted, 24 thousand meters of fencing, 2,603 hectares planted. C: Components added at second project C: Components added at second project modification modification - 18 -  1. Sutculer Rehabilitation Program 1. Sutculer Rehabilitation Program 1.1. Infrastructure 1.1. Infrastructure (a) Water supply and distribution network, Completed as planned 6100, with the Water wastewater and storrn water drainage treatment plant component dropped. networks, simple wastewater treatment system, and roads (b) Electrical distribution network (c) Telecommunication network (d) Equipment for op & maintenance 1.2. Flood & erosion Control 1.2. Flood & erosion Control 600 hectares of land prepared 12,100 meters of fencing completed to protect the 860 hectares erosion control area planted wivh 13,160 trees, 6,842 meters of stone masonry walls to control erosion 2. Retrofitting of 26 hospitals in Istanbul and 2. Retrofitting of 26 hospitals in Istanbul and 31 in Izmir, preliminary design and studies 31 In Izmir Completed 9100 for the final design drawings End of project Note: The "Projected in last PSR" column is left blank because the estimates of the last PSR and the "Actual/Latest Estimates" are the same. - 19 -  Annex 2. Project Costs and Financing Proect Cost by Cor onent (in US$ million e uivalent) Erzincan Rehabilitation and Construction 242.50 160.50 66 Training, studies, management automation, project 42.50 16.50 39 supervision and design Senirkent Rehabilitation and Construction 0.00 16.20 Sutculer Rehabilitation and Construction 0.00 4.50 Istanbul and Izmir Hospital Retrofitting, Marmara GPS 0.00 4.00 Total Baseline Cost 285.00 201.70 Total Project Costs 285.00 201.70 Total Financing Required 285.00 201.70 Project Costs by Procurement Arran ements (Appraisal Estimate) (US$ million eqlivalent) 1. Works 0.00 142.90 34.20 0.00 177.10 (0.00) (142.90) (34.20) (0.00) (177.10) 2. Goods 20.70 40.90 3.60 0.00 65.20 (20.70) (40.90) (3.60) (0.00) (65.20) 3. Services 2.50 40.20 0.00 0.00 42.70 (2.50) (40.20) (0.00) (0.00) (42.70) 4. Miscellaneous 0.00 0.00 0.00 0.00 0.00 (0.00) (0.00) (0.00) (0.00) (0.00) 5. Miscellaneous 0.00 0.00 0.00 0.00 0.00 (0.00) (0.00) (0.00) (0.00) (0.00) 6. Miscellaneous 0.00 0.00 0.00 0.00 0.00 (0.00) (0.00) (0.00) (0.00) (0.00) Total 23.20 224.00 37.80 0.00 285.00 (23.20) (224.00) (37.80) (0.00) (285.00) - 20 -  Project Costs by Procurement Arrangements (Actual/Latest Estimate) (US$ million equivalent) ,;Pv , k -/fw f Si rt M