Nepal - Earthquake Schools Rehabilitation Project

Document of The World Bank FOR OFFICIAL USE ONLY Report No. 16480 IMPLEMENTATION COMPLETION REPORT NEPAL EARTHQUAKE SCHOOLS REHABILITATION PROJECT (CREDIT 2047-NEP) April 9, 1997 Population and Human Resources Operations Division Country Department II South Asia Region 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 Name of Currency = Nepalese Rupee (NR) At Appraisal: US$1.00 =26.5 At Completion: US$1.00 = 56.5 GOVERNMENT FISCAL YEAR July 16 - July 15 ABBREVIATIONS BPEP Basic and Primary Education Project ESRP Earthquake Schools Rehabilitationl Project HMG His Majesty's Government ICB International Competitive Bidding LCB Local Competitive Bidding MHPP Ministry of Housing and Physical Planning MOE Ministry of Education MOF Ministry of Finance MTR Mid-term Review NR Nepalese Rupee SOE Statement of Expense SRU School Rehabilitationi Unit UNDP United Nations Development Programme Vice President: : Mieko Nishimizu Director : Robert S. Drysdale Division Chief : Richard Skolnik Staff Member : Brajesh Panth  FOR OFFICIAL USE ONLY IMPLEMENTATION COMPLETION REPORT NEPAL EARTHQUAKE SCHOOLS REHABILITATION PROJECT (CREDIT 2047-NEP) Table of Contents Page Preface .............................................. i Evaluation Summary .............................................. ii PART I: Project hnplementation Assessment A. Statement of Project Objectives .............................................. 1 B. Achievement of Project Objectives .............................................. 2 C. Major Factors Affecting Project Implementation .............................................. 4 D. Project Sustainability .............................................. 7 E. Bank Performance .............................................. 7 F. Borrower Performance .............................................. 8 G. Assessment of Outcome .............................................. 9 H. Lessons Learned ...................................................... 11 PART II: Statistical Tables Table 1. Summary of Assessments .13 Table 2. Related Bank Loans/Credits .15 Table 3. Project Timetable .15 Table 4. Credit Disbursements: Cumulative, Estimated and Actual .15 Table 5. Studies Included in the Project .16 Table 6A. Project Costs .............................................. 16 Table 6B. Project Financing .............................................. 16 Table 7. Status of Legal Covenants .............................................. 17 Table 8. Compliance with Operational Manual Statements .......................................... 18 Table 9. Bank Resources: Staff Inputs .............................................. 18 Table 10. Bank Resources: Missions .............................................. 18 Appendices Appendix A. The ICR Mission's Aide-Memoire Appendix B. Borrower's Contribution to the ICR Appendix C. Organizational Chart 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.   IMPLEMENTATION COMPLETION REPORT NEPAL EARTHQUAKE SCHOOLS REHABILITATION PROJECT (CREDIT 2047-NEP) Preface This is the Implementation Completion Report (ICR) for the Earthquake Schools Rehabilitation Project in Nepal, for which Credit No. 2047-NEP in the amount of SDR 17.6 million (US$22.8 million equivalent) was approved on June 27, 1989 and made effective on December 11, 1989. The original closing date was February 28, 1993. The project was, however, granted three one-year extensions upon the request of His Majesty's Government (HMG), in view of the initial delays incurred, marked improvement in project performance, high level of local demand, magnitude of ongoing procurement, and benefits accruing to many school-age children. Final disbursement took place on June 5, 1996, at which time SDR 1.52 million was canceled. SDR 16.1 million, or 91.4% of the Credit, was disbursed. The ICR was prepared by Brajesh Panth of the Bank's Field Office in Nepal, with assistance from Mark Schlagel of the Population and Human Resources Operations Division, South Asia Country Department II. The ICR was reviewed by Richard Skolnik, Chief, Population and Human Resources Operations Division, South Asia Country Department II and Kazuko Uchimura, Project Advisor, South Asia Country Department II. The preparation of this ICR began during the November 1995 supervision mission. It is based on material in the project files, field visits, and interviews/discussions with beneficiaries, project staff, Government officials and Bank staff. The Borrower contributed to the preparation of the ICR by preparing its own evaluation of the project's preparation and execution. Comments were received from the Borrower on drafts of this report and taken into account in its final version. The cooperation and assistance of His Majesty's Government is gratefully acknowledged.  I  - 11 - IMPLEMENTATION COMPLETION REPORT NEPAL EARTHQUAKE SCHOOLS REHABILITATION PROJECT (CREDIT 2047-NEP) Evaluation Summary Introduction 1. This project was developed to assist His Majesty's Government (HMG) with the reconstruction and rehabilitation of more than 2,300 schools that were either destroyed or seriously damaged, following a serious earthquake which struck parts of eastern and central Nepal on August 21, 1988. The large-scale reconstruction effort, never before tried in Nepal, required the development of a very basic, low-cost school building design that would be quick to construct, and able to withstand seismic activity of similar force (or higher as necessary in this zone) in the future. Project Objectives 2. The project had two main objectives: (a) to provide immediate assistance to HMG in support of its program to reconstruct/rehabilitate about 2,350 schools (over 13,000 classrooms) in the earthquake-affected areas of the Eastern and Central Development Regions of Nepal; and (b) to introduce earthquake-resistant design features into reconstructed school buildings and significantly improve the durability of these facilities. Implementation Experience and Results 3. The project developed its capacity early on in planning, prototype testing and other start-up activities. Much careful thought went into staffing, staff training, logistics, communications, financial control and building design; and a set of standard procedures under a decentralized management structure was tested, piloted and perfected that allowed the project to undertake large-scale construction over a large, often difficult geographic area. The early momentum of the project could not, however, translate into speedy implementation for several important reasons, necessitating three one-year closing date extensions. The project was also restructured in FY95, to limit activities to 13 difficult hill and remote mountain districts that were lagging behind considerably in implementation. 4. The extension period experienced markedly improved implementation progress for several reasons. First, the project relied, after some refinement, on the use of a basic shell steel frame structure made from standard, easily identifiable and interchangeable  - iii - materials, consisting of structural members and corrugated galvanized iron sheets for roofing manufactured at the factory. This greatly facilitated storage, porterage and construction time (frame and roof erection and masonry infill to sill level in about 21 days), simplified maintenance and greatly reduced the need for intensive on-site supervision, without compromising on quality. The substantial design improvements made over time also significantly reduced the weight of the structural members, making transport possible even to the remote mountain districts. Second, as part of the project implementation strategy, most staff were hired on a contract basis and deployed to districts under a decentralized management structure which evolved over time. This provided adequate flexibility and contributed generally to high field staff morale and low turnover. Finally, political and local demand for classrooms increased sharply in latter years, as people became more aware of the modality of the project and as communities began to focus attention and resources away from the completed task of rebuilding homes to the next priority of reconstructing schools. 5. The reconstruction objective of the project was achieved, with considerable progress made during the last three years of project extension. The project was able to construct over 15,300 classrooms -- 15% more than the original numerical target. Although communities were expected to complete construction beyond the initial basic shell structure (i.e., a design which included steel roofing, in-built window/door frames, and 'filler' walls to window sill level) with their own resources, primary schools, particularly in poorer communities, had difficulty raising resources immediately, taking up to three or four years to finalize construction. Project Unit estimates indicate that by project completion one-third of construction was fully completed, including doors and windows, one-third was completed up to the lintel level, but lacking doors, windows and glass/shutters, and one-third was completed up to the initial basic shell sill level. The percentage of fully constructed classrooms was much higher in the hill and mountain districts, where full enclosure is critical due to adverse weather conditions. In the low- lying Terai districts, however, classrooms have been able to be utilized after completion of the basic shell structure. While final construction by communities has been slower than anticipated, it is estimated that about 85% of classrooms had classes running by project completion. Since the cost involved in the final stage of construction is modest, it is expected that communities will gradually achieve full completion as they are able to mobilize additional resources. Summary of Findings, Operations and Key Lessons Learned 6. The project was successful in terms of engineering, managerial and transport logistics design. However, for the sake of speed and expediency in the emergency situation that prevailed, a compromise was considered necessary and a uniform architectural design was adopted across all ecological zones. The design was not optimal for either very cold or hot climate zones, or areas where population density was either very high (Terai region) or very low (high mountain regions). Effective collaboration between the Ministry of Housing and Physical Planning (MHPP, the implementing ministry) and the Ministry of Education (MOE, the ministry responsible for pedagogical  - iv - innovations and for creating a positive learning environment) would probably have resulted in greater emphasis on a more flexible design approach and a higher degree of community mobilization for construction completion beyond the basic shell structure provided by the project. 7. During project preparation, a survey was taken and a "master list" developed to identify those schools affected by the Earthquake which would be eligible for project assistance. The bulk of assistance -- about 75 percent -- eventually went to these schools. As might be expected, the master list had less relevance to many local authorities the further the project extended beyond the actual date of the earthquake. There was also sentiment -- likely justifiable -- that master list omissions had occurred during the initial survey and that many non-listed schools were equally deserving, as physical conditions were generally no better than those of listed schools. Nonetheless, project restructuring in FY95 improved overall list compliance. 8. The Bank's involvement with this project was appropriate and expeditious in initial response (4.5 months from start to approval), as the need was clear and the borrower's commitment well-articulated. The gross overall achievement demonstrates that the high degree of replicable standard elements in project design allowed the desired comprehensive coverage, especially in difficult areas. Greater emphasis on social communication, to improve community knowledge of the project, and stronger emphasis on inter-ministerial collaboration at the project design stage might have produced better results in terms of community preparation and local funding for the final stage of enclosure of the basic shell provision. 9. Borrower performance was generally satisfactory. However, the emergency approach adopted created powerful momentum that resulted in an emphasis on technical expedience that led technical staff to complete basic shell structures and move on to new construction, rather than risk the disruption that the introduction of design variants optimally suited for local climatic conditions and classroom size would have entailed. Effective coordination with MOE and the Basic and Primary Education Project (BPEP) proved very difficult to establish despite various attempts, although MOE was represented on the ESRP Project Board. 10. Sustainability of this project is considered likely once the schools are fully integrated into the MOE/BPEP school mapping framework, which, in addition to maximizing optimal use of classrooms will help rationalize school construction, rehabilitation and maintenance issues. During the preparation of the second phase of BPEP, MOE has indicated that it will prioritize its construction program by emphasizing rehabilitation before undertaking new construction. This approach is likely to ensure optimal utilization of existing classrooms, including ESRP- assisted classrooms. 11. The fact that the project was able to construct over 65 percent of the classrooms during the extension period indicates that this high momentum was possible because the  project had been able to substantially refine its modality during the first three years of project implementation. Key Lessons Learned 12. The project was successful at designing and disseminating a standardized basic shell design over time. The use during construction of standard, light-weight portable structural materials, under the guidance of competent technical staff, obviated the demand for prolonged on-site technical supervision and the major management burden this would entail in Nepal, permitted construction in remote areas requiring porterage of all materials, reduced costs and construction time, and ensured a certain level of quality. However, projects like this one which are implemented principally on a technical emergency footing carry the risk that certain worthwhile innovations will be overlooked. A case in point was the reluctance of the project to incorporate variant architectural designs to suit local climates. A longer initial extension period and acceptance by IDA and HMG of the risk of possible delays in implementation might have yielded higher overall benefits in this regard. Nonetheless, given the short-term nature of the extensions granted and the largely unmet construction targets which still existed at the end of the original project period, the incorporation of design variants may well have introduced a level of complexity too unpredictable for an emergency operation. 13. The ESRP model of staff hiring on a contract basis is particularly well suited when regular government employees are not motivated or capable of taking large tasks of a temporary nature involving protracted field work. Staff deployment followed by intensive training and logistics support, supported by a decentralized management structure were key factors. However, this needs to be complemented by realistic and consistent staff allowances, in order to keep morale high and turnover low. 14. While community participation is critical in an activity such as school construction, it is important to avoid applying uniform expectations across all groups. Of equal importance is the need to assign realistic values to different types of community contribution and to recognize that there are likely to be wide variations in any program of this nature. 15. The project involved construction and rehabilitation of a large magnitude and geographic coverage hitherto untried in Nepal. This required community mobilization and a relatively complex organizational structure to manage procurement and distribution of construction materials and construction. Of critical importance is understanding the time it takes for communities to comprehend the project benefits that accrue to them. Once communities began to fully realize these benefits, the pressure and demand for such assistance rose sharply and implementation accelerated.  IMPLEMENTATION COMPLETION REPORT NEPAL EARTHQUAKE SCHOOLS REHABILITATION PROJECT (CREDIT 2047-NEP) PART I: PROJECT IMPLEMENTATION ASSESSMENT A. Statement of Project Objectives 1. Project Background. On August 21, 1988 an earthquake measuring 6.7 on the Richter scale struck parts of eastern and central Nepal, killing over 720 persons and causing widespread damage, estimated at that time at NRs. 4.2 billion (US$170 million), to housing, schools, hospitals, public buildings roads and bridges. More than 2,300 schools were either destroyed or seriously damaged, and some districts reportedly lost almost all of their schools. The large scale of destruction deprived some 400,000 school children of access to regular classroom accommodations. 2. After providing immediate relief in the form of medical treatment, financial compensation and materials for shelter, His Majesty's Government (HMG) of Nepal turned its attention to rehabilitating housing and school structures through assistance from IDA. IDA was approached by HMG and school rehabilitation was initiated under the Earthquake Emergency Schools Rehabilitation Project (IDA Credit 2047-NEP) after an initial effort to quickly mobilize funds from other multilateral and bilateral agencies showed limited success. 3. The likelihood of severe structural damage following an earthquake has always been high in Nepal, given that traditional construction methods are largely of low quality. Following the August 1988 earthquake, it was necessary to construct quickly several thousand classrooms, to provide children with access to basic education facilities. This required the development of a building design that could be constructed quickly and economically on a scale never before tried in Nepal, and that could withstand similar seismic forces (and higher as necessary in this zone) in the future. The project was launched under the leadership of the Ministry of Housing and Physical Planning (MHPP), with high spirit and commitment from HMG and IDA. 4. Main Objectives: The main objectives of the project were to: (a) provide immediate assistance to HMG in support of its program to reconstruct/rehabilitate about 2,350 schools (over 13,000 classrooms) in the earthquake-affected areas of the Eastern and Central Development Regions (29 districts including 13 difficult hill and remote mountain districts) of Nepal; and (b) introduce earthquake-resistant design features into reconstructed school buildings and significantly improve the durability of these facilities.  - 2 - These objectives were based on a survey conducted during the project preparation phase to identify schools that were affected by the Earthquake. The schools on this list were eligible for project assistance and were called the master list schools. 5. The project became effective on December 11, 1989. A comprehensive planning, organization and staffing phase was successfully carried out that included the pilot testing of school designs, construction procedures and manpower skills development. However, the expected implementation targets could not be sustained, as the project suffered considerable lags for about two and a half years before picking up again. The major problems included supply problems caused by the Nepal-India trade and transit dispute of 1988/89, severe and arbitrary budget cutbacks by the Ministry of Finance (MOF) and a contracts "freeze" imposed by the Government (following the political upheavals of April/May 1990) and a Government policy decision of late 1990 requiring foreign bidders to have local agents that ran counter to IDA procurement guidelines. In view of the initial delays incurred, the marked improvement in project performance, the growing high level of local demand, the magnitude of ongoing procurement and the benefits accruing to many school-age children of Nepal, the project was granted three one-year extensions upon HMG requests. With these extensions, the project was able to achieve construction of just over 15,300 classrooms, representing 2,200 (15%) more than the original numerical target. B. Achievement of Project Objectives 6. By the original closing date of February 28, 1993, the project had achieved less than 35% of the original numerical target for classroom construction (Table 1), leaving a substantial amount of work still incomplete. However, despite several implementation constraints as noted above, the project staff by this stage had demonstrated a capacity to carry out the planned construction. It was apparent that by extending the credit closing date, it would be possible to sustain the momentum, as many of the earlier hurdles had been overcome and the project had begun to demonstrate marked improvement. The extensions did allow the project to make a substantial achievement in all project districts in general and in hill and mountain districts in particular.  - 3 - Table 1: Summary of Achievement /L (Number of Classrooms) Original Actual Achievement Target I Original Period Extension Period Grand I l (12/89 - 2/93) (3/93-2/96) Total |_______ |Primary Second. Total Primary Second. Total Difficult 6.960 934 740 1,674 3,854 2,582 6,436 8,110 Remote r Mountain Districts Low-Lying 6.394 1,316 1,658 2,974 2,084 2,158 4,242 7,216 (Terai) I Districts l Total /2 3-354 2,250 2,398 4,648 5,938 4,740 10,678 15,326 /1 All figures based on project records. As part of project restructuring in FY95, it was agreed with HMG that the project would focus during the last 18 months of implementation only on the 13 difficult hill and remote mountain districts, in order to avoid overlaps with construction initiated under the Basic and Primary Education Project (CR 2357-NEP) and recognizing that completion targets had already been substantially achieved in the lower-lying Terai districts. /2 Following project restructuring in FY95, the original target was revised upward, to 15,578 classrooms (7,789 blocks), to account for some deserving but unlisted schools and to adjust, as needed, the number of blocks originally allocated to listed schools, which in some cases had been set unrealistically low. Overall list compliance improved considerably following restructuring, with continued emphasis directed at assisting the listed schools. 7. The reconstruction objective of the project has been largely achieved, as reflected in Table I. Much of this was done during the extension period. The project was able to construct over 15,300 classrooms -- 15% more than the original numerical target. Although communities were expected to complete construction beyond the initial "sill level" of construction (i.e., basic shell structure with steel skeleton and roofing, in-built window/door frames, and 'filler' walls to window sill level) with their own resources, many schools, particularly primary schools in poorer communities, had difficulty raising resources immediately, taking on average three to four years to finalize construction. 8. Project Unit estimates indicate that by project completion one-third of construction was fully completed, including doors and windows, one-third was completed up to the lintel level, but lacking doors, windows and glass/shutters, and one-third was completed up to the initial basic shell sill level. The percentage of fully constructed classrooms was much higher in the 13 difficult hill and remote mountain districts, where full enclosure was critical due to adverse weather conditions. In the low-lying Terai districts, however, classrooms have been able to be utilized soon after completion of the basic shell structure. While final construction by communities has been slower than anticipated, it is estimated that about 85% of classrooms had classes running by project completion. Since the cost involved in the final stage of construction is modest (approximately US$100 per classroom), it is expected that communities will gradually achieve full completion as they are able to mobilize additional resources.  -4 - 9. The two-classroom basic shell design, conforming to the Indian code of practice for Zone V seismic prone areas (considering a seismic intensity of 7 to 8 on the Richter Scale) appears to have successfully met the Project's second objective, as attested by the fact that ESRP-assisted schools were unaffected by floods in the Sarlahi district in 1992 and a tornado in the Morang and Jhapa districts in 1996, even though many non-ESRP schools were severely damaged. In both instances, the quality of construction was widely acknowledged by concerned authorities and communities. The project also won a United Nations Habitat Award for its contribution to the rehabilitation effort. 10. Besides achieving its two stated objectives, the project was able during execution to improve the overall design of construction materials and to reduce overall weight, from 2,920 kg. at the initial pilot stage to 1,100 kg. (Table II). This greatly facilitated the transportation of materials to the hill and mountain districts. Local manufacturers were very competitive and were successful in winning the bulk (about 75%) of contracts awarded (Table III). Table II: Improvement in Construction Design Design/Year (NFY) Type of Structural Members Weight Excl. Roofing Sheets Original Design (89/90) Twin-column/ pipe section 2920 kg First Revision (90/91) Single post/ pipe section 2235 kg Second Revision (91/92) Half post/ pipe section 2081 kg Third Revision (92/93) Channel section/ Red Oxide 1100 kg Fourth Revision (93/94) Channel section/ Zinc Oxide 1100 kg Table III: Procurement: Quantity, Cost and Type FY No. of Two- Tender mode Average Cost Per Unit Award by Domestic/ Classroom in NRsg International Blocks Procured 1988/89 41 Local shopping n.a. all domestic 1989/90 363 LCB n.a. all domestic 1990/91 430 LCB 75,506 all domestic 1991/92 1,650 LCB 124,307 all domestic 1992/93 275 LCB 107,067 all domestic 1992/93 1,600 ICB 93,825 D = 900; I = 700 1993/94 2,000 ICB 130,089 D = 1,200; I = 800 1994/95 1,430 ICB 103,441 D = 1,035; I = 395 Totals 7,789 105,706 /I D = domestic manufacturer and I = manufacturer outside of Nepal /1 About US$940 per classroom. C. Major Factors Affecting Project Implementation 11. The project developed its capacity early on as planning, prototype testing and other start-up activities were initiated with prefinancing from the Primary Education Project (Cr. 1463-NEP). The School Rehabilitation Unit (SRU), established immediately  - 5 - after appraisal, gave much careful thought to the administration of the project in terms of staffing, staff training, logistics, communications, financial control and building design, which were all essential pre-requisites for this complex construction undertaking. This was followed by strengthening monitoring capacity and staffing core PIU personnel. Procurement delivery/stockpiling of materials for 750 classrooms was efficiently undertaken and construction involving up to 150 communities was carried out relatively smoothly, with acceptable levels of community involvement. 12. A construction effort of this magnitude was untried before in Nepal. It required creating an appropriate structure (see Appendix C), as the implementing ministry did not have the capacity to undertake such a large-scale operation over such a wide geographic area. The project contracted one engineer and an average of six overseers/suboverseers per district, set up depots and sub-depots at strategic locations to ensure speedy distribution of construction materials, created transportation mechanisms from the manufacturers to depots through trucks and from depots to schools through porters, and managed the procurement of construction materials through LCB and ICB. The project contracted a national consultant (structural engineer) for the design work, and undertook several in-service training workshops in project management/administration, quality construction and community participation for technical staff, local masons and communities. Much of the staff training and considerable assistance to the management and monitoring effort were made possible through technical assistance provided by UNDP and managed by a chief technical advisor. The size and complexity of the project required establishing a set of standard procedures under a decentralized management structure which was tested, piloted and perfected to construct the basic shell structure. 13. The initial momentum of the project could not translate into speedy implementation due to the problems stated earlier, and the project ultimately required three one-year extensions, as it was not possible to make up the cumulative delays from the early period. During the third year, the implementation effort improved markedly, resulting in the speedy delivery of classrooms and excellent progress in construction work. 14. Whereas implementation priority had initially been given to the populous Terai districts (where the logistics and deployment process was less complex), the project was restructured in FY95 to limit its activities exclusively to the 13 hill and mountain districts, where construction progress was still lagging behind considerably. This geographical focus resulted in considerable progress by project end, aided in large part by several factors. First, the project relied after some refinement on the use of a basic shell steel frame structure made from standard, easily identifiable and interchangeable materials, consisting of structural members and corrugated galvanized iron sheets for roofing manufactured at the factory. This greatly facilitated storage, porterage and construction time (frame and roof erection and masonry infill to sill level in about 21 days), simplified maintenance and greatly reduced the need for intensive on-site supervision, without compromising on quality. In addition, the substantial design improvements made over time reduced the weight of the structural members, making  - 6 - transport by porters possible even to the remote mountain districts. Second, as part of the project implementation strategy, most staff were hired on a contract basis with field allowances and deployed to districts under a decentralized management structure which evolved over time. The project also trained a cadre of artisans from project districts in assembly of the basic shell structure who were contracted as required on a per job basis. This provided adequate flexibility and contributed generally to high field staff morale and low turnover. Each district had an engineer responsible for all the construction work in that district. The contract approach provided the SRU with great flexibility in the use and mix of staff. Finally, the political and local demand for classrooms increased sharply in latter years, as people became more aware of the modality of the project and as communities began to focus attention and resources away from the completed task of rebuilding homes to the next priority of reconstructing schools. 15. Community Participation. One of the positive features in ESRP design was the involvement of communities in school construction and management under the technical guidance of the project. Although this was not a new concept in Nepal, this approach further strengthened the roles of communities that have traditionally assumed the responsibility of school construction from their own resources. A school reconstruction committee at each school was constituted by the communities with whom the project entered into a contract. This committee not only managed construction, it also arranged for labor for porterage of construction materials and construction; and local materials for foundation, walls, floors, door and window shutters, plastering, painting and furniture. The project provided steel frames, roof sheets, window and door frames, reinforcement and cement. Community participation averaged around 30% of the total value of construction in a completed two classroom block. 16. The tradition of community participation continues to play a critical role in school construction in Nepal, where contractors are scarce and expensive. This feature may be diminishing in urban settings, where the opportunity cost of voluntary labor is relatively higher than in rural hills or mountains. However, even where participation is likely, it cannot be ensured across all groups. Primary schools were generally found to have difficulty in assuming this responsibility because they were less endowed with resources and had to rely almost exclusively on government grants or contributions from communities. Primary schools that largely serve poorer communities with very limited resources tend to face severe resource constraints. It is important to avoid putting an unnecessary financial burden on such schools and communities in the process of requiring community participation. In the light of initial feedback on rates of completion by communities and after considerable deliberation lasting nearly two years, due primarily to inadequate information on deserving schools and budgetary delays, the project authorities provided a grant of NRs. 10,000-20,000 (US$177-354) per two- classroom block (around 5 - 10% of the total cost of fully complete blocks), during the latter half of FY96 to the listed primary schools that had been unable to complete the final stage of wall construction beyond the basic shell sill level. This marginal support generated additional contributions from communities and allowed many schools to complete construction.  - 7 - 17. In addition to head teachers, school management committees and local politicians, credit for mobilizing community support goes to project field staff, particularly engineers and overseers, who operated not only as technical staff in construction but also as prime motivators and negotiators throughout the entire process. Although the engineers and overseers had to work in extremely difficult terrain and political conditions, they were instrumental in bringing various leaders and opposing groups to a consensus. A low staff turnover allowed the technical staff to use their accumulated knowledge to constantly improve their work in the field. D. Project Sustainability 18. The project was implemented by a technical ministry, but all ESRP schools, being public schools, will have to be integrated into MOE/BPEP's school mapping framework. Although maintenance requirements will initially be minimal since these structures are new and of reasonable quality, it is critical to integrate the ESRP facilities into the maintenance training program of BPEP, for which considerable background work has been done under BPEP. Since construction was done with community participation, it is essential to maintain this support from communities to make the maintenance program sustainable. 19. Another form of sustainability can come by rationalizing the BPEP construction program, where current and future projects are active in ESRP-assisted districts, to support community completion efforts. By the nature of the emergency basic shell design, up to one third of ESRP classrooms may still have had walling incomplete beyond the sill level at project closure. By providing marginal grant support during the last year of the project, most incomplete blocks are now likely to be quickly completed by communities. This will maximize utility of the investment already made under ESRP and will allow a wider coverage of the expanding classroom construction program under BPEP. 20. Finally, it is also important to encourage the use of local materials that are commonly used by communities. Motivation of communities to use simpler finishing inputs such as bamboo in-fill, mud mortar, etc., led in various instances to expeditious construction and probably simplified maintenance requirements, since communities are more familiar with these inputs. Since the basic shell structure is strong enough to last for up to 50 years, the use of local materials for the in-fill work would not necessarily lead to a serious compromise on overall quality and durability. E. Bank Performance 21. The Bank's performance was satisfactory with respect to project assessment, preparation assistance, appraisal and supervision. Bank involvement in this emergency project was appropriate and expeditious in initial response -- 4.5 months from start to approval -- as the need was clear and the Borrower's commitment well articulated.  -8 - 22. The project was well designed and supported by UNDP-funded technical assistance, which covered staff training, management guidance and monitoring. As a result, IDA supervision missions were more focused and devoted to construction quality and implementation arrangements that were critical to successful implementation. 23. Supervision was managed from headquarters throughout the first five years of the project, with ongoing side-support increasing over time from the Nepal field office. During the last two years of the project, primary supervision responsibility was delegated to the field office. Throughout supervision, the Bank provided suggestions in terms of the effectiveness of the Borrower's supervision and delivery mechanisms that helped to increase the overall efficiency of the project. 24. When implementation progress failed to meet original expectations, and it became evident that the project would not be able to fully complete construction as expected within the original period, IDA supported HMG's request to extend the project. Three extensions were possible due to IDA commitment. This support, necessary to sustain the strong capacity developed by the project team, paid off well in terms of overall construction achievement, particularly in the remote mountain districts. Since most classrooms could be utilized soon after the completion up to the basic shell structure, project inputs proved critical in terms of reconstructing many of the damaged schools fairly quickly and providing safer classrooms to many school children. 25. With hindsight, had the project been granted at the outset a longer extension, linked to an MTR and restructuring, the outcome might well have been a more positive Borrower reaction to the inclusion of design variants. The project was restructured in FY95, recognizing the fact that (i) many district construction targets were substantially completed in the lower-lying Terai districts; (ii) a potential for duplication of efforts between BPEP (which started in July 1992) and ESRP existed if both projects were to operate in the same districts; and, (iii) primary schools were experiencing financial difficulties in completing construction beyond the basic project provision. The restructuring allowed substantial construction achievement in the 13 hill and mountain districts that had lagged behind, greater involvement of District Education Committees in obtaining MOE endorsement for some non-listed schools, and a higher rate of community completion of classrooms, through provision of modest grant support to listed primary schools that had been unable after several years to complete construction. F. Borrower Performance 26. Overall, borrower performance was satisfactory, although better physical progress would have been possible with greater inter-ministerial coordination. Although a high level inter-ministerial board chaired by the MHPP Minister was effective in resolving many implementation issues relating to procurement, distribution and project management, there were two shortfalls. Eirst, the project was technically driven, as the staff were more interested in completing basic shell structures and moving on to new construction sites than they were in monitoring the final completion of schools by  - 9 - communities. The remuneration for technical staff in the field was also reduced at project mid-course, after a fairly reasonable field allowance of 60% of basic salary was replaced with a completion incentive, averaging two-thirds the earlier amount. This adversely affected the morale of staff and was a disincentive to site visits. After the UNDP- financed technical assistance program was completed, the project had inadequate funds budgeted for monitoring and almost no vehicles. Despite IDA's urging, the Government did not approve the procurement of vehicles and made only modest adjustment in the field allowance. 27. Second, coordination with MOE and BPEP was not effectively established, despite repeated efforts from IDA and representation of MOE on the project board. In districts where ESRP blocks had already been allocated and constructed, a closer coordination would have allowed BPEP to update its school mapping and enabled a more efficient merging of the two programs. Being a purely technical ministry, the MHPP was less successful in mobilizing the people in education, such as district education officers and teachers. More systematic execution of final community inputs above the sill level might have been possible with a closer coordination with district education offices, as the SRU was already burdened with completing the minimum basic shell structure in 29 districts. 28. A positive feature was the functional decentralized management structure established by the Borrower, which was the key to overall success. ESRP was effective in the deployment of technical staff to the field, but less so in the case of administrative staff, particularly at headquarters in Kathmandu. Despite repeated concerns expressed by several IDA supervision missions regarding administrative overstaffing, little correction was made. Except for some key personnel, the central office was much less effective than the field offices. However, the central office management was generally very supportive of field staff, which allowed the field offices to be effective in their operation. G. Assessment Of Outcome 29. In retrospect, some issues have emerged. While the project was successful in terms of engineering, managerial and transport logistics design, it might have benefited by adapting the basic design module to introduce one or more variants. The low output in the first two years encouraged the emergency psychology to prevail throughout the project life even though the project was eventually extended for three years. 30. For the sake of speed and as an emergency response, a compromise was made from the outset in adoption of a uniform design approach. This resulted in a structure which is probably less than optimal for either very cold or hot climate zones of Nepal. The given size of the classrooms might have been appropriate in areas where population density was medium (mid-hills), whereas it might have been less so in areas where population density was either very high (Terai region) or very low (high mountain regions). However, it would have been necessary to address this issue at an early stage in the project cycle, at a time when project performance was poor and the risk of loss of  - 10- confidence in the project by HMG and IDA was high. A more effective MHPP-MOE collaboration would also have allowed the project better access to MOE officials and teachers for community mobilization towards rapid completion beyond the sill level. 31. Although the bulk of the assistance (75%) eventually went to master list schools, the distribution could have been more tightly controlled. The structure and staffing was adequate in general, but the discipline required in decision making to reach as many listed schools as agreed upon was less optimal. During the first 3.5 years of the project, the blocks mostly went to listed schools. Although around 10 percent deviation to non-listed schools was expected and accepted, the project went beyond this limit. IDA missions repeatedly underscored the need to adhere to the master list, but this proved an elusive task amidst tremendous community and political pressures for inclusion of non-listed schools. However, project restructuring and strong emphasis by IDA in FY95 ensured improved list compliance. 32. Several factors accounted for the inclusion in the project of non-listed schools. First, as the project duration lengthened, the original master list had less relevance, with some listed schools over time either mobilizing other non-project resources to complete construction or, in the wake of growing population pressure, re-prioritizing limited school resources to support secondary level expansion activities (e.g., hiring additional teachers, putting down deposits, etc.) -- a need still not recognized and funded by the government. For some other listed schools, non-participation in the project was more simply the result of school/local resource limitations. Accordingly, by the last stage of the project, funds that might have gone to some of these non-participating listed schools were directed to other deserving non-listed schools. 33. Second, many communities, governrnent officials and project staff, including district engineers, argued that the original survey did not include all the deserving schools, citing examples of geographic pockets that were omitted during the initial survey. This point appears largely justifiable, particularly given that the physical conditions of nearly all of the non-listed schools funded under the project were generally found to be no better than those of listed schools. 34. Finally, given the strong implementation capacity developed by the project, the Government discussed with the IDA mission during 1994 the strategy to fully complete the original project and to utilize an uncommitted balance of about US$6 million to support the major ongoing thrust in basic education financed by IDA, ADB and other donors. Since the project was able to construct 15% more than the original numerical target, it was natural that a higher percentage of schools not on the master list received project assistance. 35. Overall, some important aspects may have been less optimal, but given that closing date extensions were granted one year at a time as per the Bank policy, the ESRP was most reluctant to accept the burden of additional delays required to make substantial  - 11 - changes in terms of facility designs and in the modus operandi of a structure which was belatedly becoming highly successful in achieving its original numerical targets. H. Lessons Learned 36. Technical Design/Simplicity. The project was successful at designing and disseminating a basic, highly repetitive emergency formula for construction that was later modified through innovations in engineering design, without adding complexity. Based on experience gained, the project was able to improve the quality of pre-fabricated steel materials and reduce weight substantially, under the guidance of competent technical staff. This obviated the demand for prolonged on-site technical supervision and the major management burden this would entail in Nepal, permitted construction in remote areas requiring porterage of all materials, and reduced costs and construction time. 37. While the project achieved its physical targets due in large measure to its basic simple formulation, the introduction of design variants to deal more effectively with geographic differences in climate and class size might have improved the final product. However, this may well have introduced a level of complexity too unpredictable for an emergency operation. Given that IDA treated extensions on a single year basis, the reluctance of project officials to vary a productive implementation formula is understandable. Willingness on the part of IDA to accept a longer initial extension may have set the stage for the effective introduction for one or more design variants by HMG. 38. Decentralized Management and Staffing. The project's development of a lean, but fragile decentralized management structure, able to handle the technical and administrative interface between a single basic physical structure and execution at the community level, was a key factor for success. Intensive training and logistics support to staff was an integral part of this process. Most district-based engineers, hired on a contract basis, were successful not only because they were technically competent, but because they stayed with the project, were motivated due to the authority and flexibility given to them, had minimal interference from the center and were able to relate to a wide range of people, from district-based government officials to people in different communities. This indicates that in Nepal, if regular government employees are not motivated or capable of undertaking complex temporary assignments involving protracted field work, it is possible to successfully utilize competent, contracted staff, supported by a decentralized management structure. Contracting must, however, be complemented by rewarding and consistent staff allowances, to ensure high morale and low turnover. 39. Community Participation. Once communities began to fully realize the benefits accruing to them, the pressure and demand for assistance from the project rose sharply and implementation accelerated. The project relied on the use of project staff and radio announcements to raise community awareness, however, a targeted social communication component, developed during the initial phase of implementation, could have enhanced  - 12 - the extent of participation and helped communities understand at an earlier stage their role in and features of project implementation. 40. While community participation is critical in an activity such as school construction in Nepal, it is important to avoid applying uniform expectations across all groups, since not all communities may be able or sufficiently interested in participation. This requires a more flexible approach. When strategies, such as the reimbursement policy for transportation costs, are applied to enhance community participation, it is essential to study the implications and make necessary adjustments for optimal results. Of equal importance is the need to assign realistic values to different types of community contribution and to recognize that there are likely to be wide variations in any program of this nature. It may be equally important and more sustainable to encourage communities to accept easily maintained local materials, such as bamboo in-fill, mud mortar, etc. as viable alternatives to more durable modern materials which may be difficult to obtain and maintain in some areas. 41. Technical Assistance. When the Government is reluctant to use credit resources for technical assistance, an IDA project is likely to perform better when complemented by a grant component. This project was primarily funded by IDA, but supporting technical assistance from UNDP, in the form of a chief technical advisor for the first four years of the project, two vehicles and supporting funds provided valuable flexibility to the project in key areas such as staff training, management guidance and monitoring.  - 13 - PART II: STATISTICAL TABLES Table 1: Summary of Assessments A. Achievement of Objectives Substantial Partial Negligible Not Applicable Macro policies Sector Policies W E Financial objectives W Institutional development | | Physical objectives W x Poverty reduction | [Z Gender issues W Other social objectives Ex Environmental objectives W 7 Public sector management W E Private sector management W  - 14 - Table 1: Summary of Assessments (continued) B. Project sustainability Likely Unlikely Uncertain F O W C. Bank performance Highly Satisfactory Deficient satisfactory Assessment E L Preparation assistance I W I Appraisal EII Supervision F xE