Sri Lanka - General Education Project

Document of The World Bank FOR OFFICIAL USE ONLY Report No: 18010 PERFORMANCE AUDIT REPORT SRI LANKA GENERAL EDUCATION PROJECT (Credit 2072-CE) June 11, 1998 Operations Evaluation Department This document has a restricted distribution and may be used by recipients only in the perfornance of their official duties. Its contents may not otherwise be disclosed without World Bank authorization.  Currency Equivalents (annual averages) Currency Unit = SLRupees (Rs) Rs 1.00 = US$0.0172 US$ 1.00 = Rs 58 (January 1996) Abbreviations and Acronyms ADB Asian Development Bank GEP General Education Project IDA International Development Association MEHE Ministry of Education and Higher Education OED Operations Evaluation Department PAR Performance Audit Report SAR Staff Appraisal Report Fiscal Year Government of Sri Lanka: January 1 -December 31 Director-General, Operations Evaluation . Mr. Robert Picciotto Director, Operations Evaluation Dept. : Ms. Elizabeth McAllister Manager, Sector and Thematic Evaluations Group : Mr. Roger Slade Task Manager * Ms. Helen Abadzi  FOR OFFICIAL USE ONLY The World Bank Washington, D.C. 20433 U.SA. Office of the Director-General Operations Evaluation June 11, 1998 MEMORANDUM TO THIE EXECUTIVE DIRECTORS AND THE PRESIDENT SUBJECT: Performance Audit Report on Sri Lanka General Education Project (Credit 2072-CE) The Sri Lanka General Education project supported by Credit 2072-CE for US$49 million equivalent, was approved in FY90. After an extension of one year, the credit was fully disbursed and closed on December 31, 1996. This was the first education project IDA financed in Sri Lanka. Its objectives were to: (a) increase the effectiveness of the education system by upgrading facilities and increasing the supply of educational materials; (b) improve efficiency through better management and supervision; (c) reduce regional and local disparities in the quality of schools; (d) support decentralization; (e) strengthen planning; and (f) better calibrate education and training to labor market requirements. The project financed construction and rehabilitation of schools, maintenance, equipment, books and materials, teacher and staff training, technical assistance, and labor-market studies. The project disbursed very little during its first three years owing to the civil war, which also exacerbated institutional weaknesses. However, implementation accelerated after a mid-term review, and by the end of the project, planned activities had been carried out. Targets for civil works and supply of educational equipment were met or substantially exceeded. In-service training of staff principals was successful. Studies of educational finance and management, physical facilities, teacher deployment and absenteeism were successfully completed despite delays in the provision of technical assistance. During the last three years, the project also financed the study of important policy issues, in areas of teacher training and educational finance. The studies and related seminars developed into two follow-on projects that supported long-ranging reforms. Thus a hardware project was turned into a policy-oriented project. An important vehicle for coordination and communication was an education development committee of provincial officers, which met every month during the second half of the project. The Operations Evaluation Department (OED) rates project outcome as satisfactory, institutional development as modest, sustainability as likely, and Bank performance as satisfactory. Both Bank and borrower performance were unsatisfactory during the first half of the project. Lack of counterpart funds prevented staff appointments, and management of the project implementation unit seemed unable to coordinate work with provincial offices. In the Bank, task managers changed repeatedly, and until the mid-term review, attention to the project was desultory. Following resolutions during the mid-term review, however, implementation and disbursements accelerated, and the task manager paid minute attention to policy and implementation details. Thus, Bank and borrower performance during the last half of the project were highly satisfactory. 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.  2 Important lessons stemming from this project are: * Task manager diligence, creativity, expertise, and implementation orientation can significantly influence project outcomes. The Bank should use educators who have detailed knowledge of the field and who are willing to innovate and to take risks. * Organized, long-tern communication between ministries and regional offices clearly improves implementation outcome. The challenge is to provide the circumstances to make this possible. Incentives and easy means for commuting are important in allowing committees to meet repeatedly over time. * Insufficient attention may be paid to the instructional consequences of building designs. Projects should study noise and heat levels in schoolrooms of different types relative to instructional outcomes. * Remote schools are more likely to have cumulative problems, including low-quality construction. More support in time and effort should be made to supervise those sites. * Instructional materials, particularly for science, may stay unused until expertise and a real motive for using them exists. In addition to a clear rationale for choosing each item, training must be provided in their use, and teacher supervision must include use of instructional aids. Attachment  l Contents Ratings and Responsibilities .....................3 Preface .............5 1. Introduction ...........7 2. Project Objectives and Components ...........................................................8 3. Implementation Experience ...........................................................9 Component 1: Increase of Equitable Supply of Educational Facilities and Materials .......................................................... 10 Component 2: Improvement of Management and Supervision ................................. 11 Component 3: Strengthening the National Institute of Education and Planning Offices: Education Policy Development ....................................... 12 Studies .......................................................... 13 Results .......................................................... 15 Monitoring and Evaluation .......................................................... 17 Institutional Development ........................................................... 17 Sustainability .......................................................... 17 IDA Performance .......................................................... 17 Borrower Performance .......................................................... 18 4. Issues and Recommendations .......................................................... 19 Teacher Absenteeism .......................................................... 19 Instructional Deficiencies in Teacher Training Institutes .......................................... 19 Extra Burdens of the Poor Students' Schools .......................................................... 20 5. Lessons .......................................................... 21 Annex Basic Data Sheet .......................................................... 23 Box. The Meetings of the Education Development Committee ................................................ 16 Table 3.1 Project Activities .......................................................... 13   3 Principal Ratings Principal Ratings ICR Audit Outcome Satisfactory Satisfactory Sustainability Likely Likely Institutional Development Partial Modest Borrower Performance Satisfactory Satisfactory Bank Performance Satisfactory Satisfactory Key Staff Responsible Task Manager Division Chief Country Director Appraisal James French Smith Martin Karcher Jochen Kreiske Midterm Albert Aime Barbara Herz Paul Isenman Completion Albert Aime Barbara Herz Mieko Nishimizu   5 Preface This is a Performance Audit Report (PAR) on the General Education Project (Credit 2072-CE), which was approved on December 12, 1989, for US$51.4 million equivalent. The project was completed, with the credit fully disbursed, on December 20, 1996, after a one-year extension. The purpose of the audit was to study the effectiveness of the International Development Association's (IDA's) first primary education project in Sri Lanka and the. significant policy changes that it engendered in the country's educational system. The Operations Evaluation Department (OED) found it important to study the effects of IDA advice under the circumstances of a civil war but also considerable commitment to quality education. The PAR is based on the following sources: the Implementation Completion Report (ICR), issued as Report 16654, dated June 9, 1997; the Staff Appraisal Report (SAR); the Credit Agreement for the project; and the project files, particularly the supervision reports. An OED mission visited Sri Lanka in January 1997 to collect other pertinent information. Many thanks are owed to government officials and researchers for their cooperation. Following customary OED procedures, copies of the draft PAR were sent to the relevant government officials and agencies for their review and comments but none were received.   7 1. Introduction 1.1 Sri Lanka achieved remarkable educational progress in the 1980s, despite a civil war that started in 1983. Primary school enrollment rates rose to over 90 percent, and girls achieved educational parity with boys. However, school management remained weak, unexcused teacher absenteeism remained high, and the quality of instructional delivery remained low. The system was highly inequitable: urban "national" schools for high-performing students received about 3000 percent more funding per capita than Type 2 and Type 3 rural and semirural schools.' The schools of the rural poor had inadequate facilities and provided particularly low quality of education. The General Education Project (GEP) was conceived to remedy the problems caused by inadequately distributed resources. It also aimed to support devolution of authority to the eight provinces, a process that the government decreed in 1989 in an effort to end the civil war. 1. Schools in Sri Lanka are classified as Al or "national," for which children are selected through competitive examinations; "Type IA, B, or C" secondary schools for average students with or without science; "Type 2," large primary and secondary schools often in semirural areas; and "Type 3," primary and lower secondary schools in rural areas. Urban students of higher socioeconomic status are much more likely to be accepted into national schools than poorer, rural students.  8 2. Project Objectives and Components 2.1 The project objectives were to (a) increase the effectiveness of the system by upgrading facilities and increasing the supply of educational materials; (b) improve efficiency through better management and supervision, (c) reduce regional and local disparities in the quality of schools; (d) support decentralization; (e) strengthen planning; and (f) begin adjusting education and training to labor market requirements. 2.2 Total project cost was US$75 million, of which IDA financed US$51.4 million equivalent (SDR 35.4 million). Project design was based on a sector study funded by a Project Preparation Advance. The project focused on hardware, and about 90 percent of proceeds were to be spent on civil works. Policy content was limited to agreements that solutions to teacher absenteeism would be studied. Though the project supported decentralization, it did not initially promote specific policies about it.  9 3. Implementation Experience 3.1 The SAR identified several risks: (a) lack of counterpart funds, (b) concerns about military security, and (c) piecemeal implementation of the management component, because it relied on the informal cooperation of central and provincial staff. These risks and related problems materialized soon after the project started implementation. 3.2 Civil strife caused communication difficulties and security problems. Due to financial difficulties, the Central Bank created lengthy delay in the release of IDA funds. The consequent lack of funds resulted in fitful starts and stops of activities and prevented the hiring of staff, such as accountants and a procurement officer, during the first two years. There was some confusion about when the project really started, and the director of the project implementation unit (PIU) proved unable to respond to the many tasks at hand. The national devolution policy required that provincial offices carry out civil works, but the policy was fraught with uncertainties, and the provincial staff initially lacked the capacity to implement the necessary activities. The Bank's project management was also less than optimal; task managers changed repeatedly, and Sri Lanka changed departments within the Bank. These events caused some neglect of the project. Although disbursements were low, the supervision reports indicated that progress was satisfactory. However, due to all these factors, only SDR 4 million had been disbursed by mid- term review. 3.3 During the third year of the project, prospects changed dramatically. A new Minister and Secretary of Education became keenly interested in the project and took important steps to revive it. About the same time, the Bank appointed to the project a task manager who had long experience in managing educational systems before entering the Bank and who focused on results. The govermment upgraded the project coordination unit (PCU) staff, established a constructive dialogue with the provinces, and increased implementation monitoring. The new government that took over in 1994 had a very clear commitment to education and further facilitated implementation. With a one-year extension, credit proceeds were fully disbursed and most numerical targets were met or exceeded (Table 3.1). 3.4 In addition to civil works and fellowships, broad systemic reforms were needed to reach the project goals. These related to per capita student expenditures, teacher careers and needs, curricular and textbook changes, classroom supervision, and school mapping. The initial project design had not emphasized these, but the task manager who took over after the mid-term review focused on these issues. He initiated detailed analyses of the issues and convinced the government that the reforms would be desirable. To carry them out, two separate follow-on projects were developed: Teacher Training and Teacher Development project (Credit 2881-CE), which became effective in August 1996, and the General Education II project, which became effective in January 1998. Thus, through the desire of the govermment and the efforts of the task manager, a number of reforms became effective during the GEP or shortly thereafter. Among those reforms were teacher salary increases,2 increasing the education budget to sustainable levels, decisions on school rationalization through school mapping, and equitable distribution of 2. Teacher pay was low, but increased 48 percent in 1995, an amount that was fiscally unsustainable. Also, increases did not differentiate between trained and untrained teachers.  10 financial allocations to schools, so that national schools would no longer get the lion's share. As a result, a hardware-oriented project became the basis of far-reaching reforms. 3.5 Despite efforts to distribute project benefits equitably throughout the country, the effects of the civil war confined project activities to predominantly Sinhalese areas. No civil works were carried out in the predominantly Tamil areas of the north and east. Training of Tamil-speaking principals also became problematic because they were hesitant to leave their areas, and trainers were hesitant to go to them. However, equipment and materials were sent to the Jaffna schools in 1996 through the military commander. Component 1: Increase of Equitable Supply of Educational Facilities and Materials 3.6 The project supported a construction program in rehabilitating and enlarging about 3,000 Type 2 and Type 3 schools. Schools received classrooms, toilets, wells, furniture, and equipment. Construction plans were designed by the Ministry of Education, and the provincial offices implemented appropriate choices. At their discretion, they identified the schools where construction was needed. They costed the civil works, prepared tenders, evaluated bids, selected contractors, and the provincial engineering offices supervised the works. The ministry supervised and reviewed the execution of civil works but had no on-site control over construction. 3.7 The mission visited 14 Type 2 and 3 schools in three provinces. In most areas, users reported that civil works were satisfactory. However, there was some concern that contractors carried out prearranged designs without taking local preferences into account. Also, two remote schools were found to have significant problems, apparently caused by contractor neglect. According to some provincial engineers, it was difficult to find qualified contractors for more remote sites and to get a satisfactory product. One case of a defective building had dragged on for years. Satisfactory constructions are likely to suffer problems in the future because there is no preventive maintenance.3 3.8 Although construction quality was satisfactory, location was not always optimal. There was no updated school mapping, which took into account population growth and changes. Without it, decisions regarding site selection were open to political influence. Some new buildings were built where they were not needed, while in other sites students remained crowded. After a school mapping exercise carried out toward the end of the project, it became apparent to the government that some schools where IDA financed construction should close. 3.9 The building design paid no attention to noise levels. The buildings were mainly large, open halls. Very few Type 2 and 3 schools, old or new, had any partitions between classrooms. ("National" schools, however, typically have closed classrooms.) The open design has been used traditionally and is still preferred because (a) it allows air to circulate, (b) assemblies can be held inside the buildings, (c) classroom spaces can adjust to changing class sizes, and (d) teachers can keep an eye on neighboring classes when teachers are absent. With classes sitting next to one another, teachers compete for students' attention. The noise levels observed by the mission were high, particularly when teachers were absent and classes were unsupervised. Principals and 3. A 1997 audit of the two Bank-financed construction industry projects in Sri Lanka (Credits 11 30-CE and 1698-CE) has highlighted difficulties with contractor performance.  11 teachers reported that they do not like this situation and teachers often complain of the stress that the open space creates. 3.10 To lower the noise and distraction somewhat, some schools use low partitions. However, those clearly do not block the noise. To escape the noise teachers often hold class in the open air, but then they do not have access to the blackboard and teaching aids. Despite attention to quality, this issue was not given any attention during the project. 3.11 The mission found that the project had built some closed classrooms, but the air in the rooms did not sufficiently circulate. Glass windows with metal frames permanently closed off half the window opening, creating a stifling environment. 3.12 The mission heard several complaints regardingfurniture. In some schools, the wood used to build the furniture was untreated and was being eaten by ants. The parent committees had the option of painting the furniture, and some did, but the paint easily wore off, making the furniture look much older than it was. Screws holding stools together easily came out. Efforts were made in the last two years to obtain better and larger furniture, but overall, funiture quality was unsatisfactory. 3.13 Educational equipment was supplied to schools, and the mission found it in all school cabinets. However, in most cases the materials were not being used. The principals explained this by citing lack of training or the recent arrival of the material. Some said that if teachers feel the need, they come to use them. This raises questions about how well teachers are trained to use the materials, particularly the chemistry and physics equipment. 3.14 Overall, some progress was made in breaching the gap between the rural and the urban schools as a result of civil works, furniture, and equipment. National schools continue to have better infrastructure (such as classrooms with walls), maintenance, and teacher distributions. However, national schools are also very crowded, and this compromises their capacity to deliver quality instruction. True equity, and a relief of overcrowding for the better schools, will be achieved only when resources are equitably distributed. Component 2: Improvement of Management and Supervision 3.15 Absenteeism. Teacher absenteeism had been a major problem at the time of project appraisal. Many efforts were made during the project to reduce it, and reportedly some progress was made. However, the evaluation mission visited many classes that were unsupervised. It was not unusual to find 20 to 25 percent of the teachers absent on a given day. Principals explained that teachers, who have 42 days of leave during the school year, take leave at their discretion. Moreover, while they are supposed to give advance notification of an absence, they often do not do so. As a result, their classes remain unsupervised. Children may roam the corridors, and the noise of unsupervised classes in the open school buildings makes learning more difficult for other students. 3.16 Principals expressed frustration with the situation and helplessness in solving the problem. Many felt that they could not take action against noncompliant teachers. In principle they could refuse to approve the unannounced leave, but then teachers could go to the provincial offices and get it approved. Principals also felt that politically connected teachers could create problems for them if leave was disapproved.  12 3.17 Management training. Although delayed by two years, most principals' training activities were completed satisfactorily. Thirty-two young educators received one-year foreign fellowships to become trainers. During their fellowships and upon their return, they developed training materials at the National Institute of Education and were then assigned to provinces, where they conducted 20-day training courses for principals. The courses used participatory methods and focused on relations with teachers and parents, administrative and financial regulations, and classroom observations. The trainers later visited the new trainees and held shorter follow-up courses on problem they had identified. Principals were eager to participate in the courses, which were held during vacations. 3.18 The 12 principals interviewed by the mission expressed satisfaction with the training. The main benefit they cited was a greater understanding of financial and administrative regulations. Another was teacher supervision techniques. During the courses, principals designed teacher observation forms, which the mission found in use in several schools. The training therefore may have helped increase classroom control. Unfortunately, lack of funds prevented its repetition after the project ended. More training is expected to take place under the follow-on project, GEP II. 3.19 Possibly as a result of training and other project inputs, many of the schools visited by the mission had some of the characteristics that make schools effective. Buildings were kept clean, the grounds were neat, and students took care of them. Textbooks were available, and staff was willing to receive training during vacations. On the other hand, the many unsupervised classrooms resulted in noise, and inefficient use of class time. Component 3: Strengthening the National Institute of Education and Planning Offices: Education Policy Development 3.20 The National Institute of Education trains various categories of educational staff and contributes to the development of some aspects of educational policy. Thirteen of its professors received foreign fellowships for masters degrees and doctorates. All but two trainees returned to Sri Lanka and resumed their work. 3.21 The National Institute of Education selected fellowship recipients for principals' training and helped develop training materials. It also monitored training. Ultimately its role in developing policy, as envisaged in the SAR, seems to have been limited. 3.22 The management information system established by the project has worked satisfactorily in most areas. To develop expertise for it, training was offered nationally and abroad. Unfortunately, many trainees were selected through political criteria and were not suitable for the tasks. In general staff judged the national training they received to be insufficient for the for the task. However, the mission observed in two districts computer terminals in use for the of data. The inappropriate staff selection and inadequate training, however, meant that some provincial offices could only create simple statistics. They did not have the capacity to create and update data bases for individual teachers, students, and schools.  13 Studies 3.23 The SAR had envisioned two studies: one of teacher attendance and the other of linkages between labor and education. The studies required much effort and supervision to complete. The labor linkages study did not provide particularly useful recommendations. Nevertheless, it served as an input in an Asian Development Bank (ADB) vocational education project. The teacher attendance study served as a major vehicle for the follow-on project. 3.24 When the project was revived at mid-term review, IDA recommended several more studies. (Table 3.1). These included mapping of school facilities, management structures of the Ministry of Education and Higher Education (MEHE), financing of education, teacher absenteeism and education, textbook quality, and a maintenance handbook. Several seminars were carried out on study results and implications during project supervisions. The studies were completed and provided the government with much-needed analyses and became the basis for the reform package being carried out in the two follow-on projects. Table 3.1: Project Activities Components/ Target Subcomponents Activities Achievement Outputs Outcomes A. Equitable supply of educational facilities and materials Upgrade substandard 166% of target 100% of students in Classes seat more schools achieved Type 2 & Type 3 students schools seated, compared to 40% Enrollments did not before the project increase as a result Supply teaching aids and 140% of target Teaching aids in Teaching aids not science equipment achieved schools sufficiently used Provide water in schools 118% of target Water available in Possibly fewer where unavailable achieved almost all schools intestinal problems Provide latrines in 123% of target Latrines available Students able to use schools that lacked them achieved in schools them; some do not function Provide school fumiture 330% of target Despite targets, Students use them, but and equipment achieved fumiture is of poor they may not last quality Provide provincial 82 vehicles given All were reportedly Increased mobility officials with transport in working order makes inspections and for supervision and used for work- meeting attendance related purposes easier, improved teacher attendance B. Improvement of management and supervision Strengthening divisional 59 administrative centers 180% of target Centers constructed Offices in use offices to support constructed achieved decentralization  14 Components! Target Subcomponents Activities Achievement Outputs Outcomes Staff received training 80% of Trained directors Unknown provincial are more directors knowledgeable about 38% of zonal administrative directors matters Principals received 8,976 principals Trained principals Principals may be training trained (70% of reportedly know more effective, may be target) more about better able to monitor financial, instructional delivery administrative issues 5,042 English Teachers know Student perfornance teachers trained more English in English improved (new component) in 1996-97 exams 80% of target Strengthening school Clusters did not No results management through clusters effectively distribute resources and were abandoned School Boards were not Parent participafion development effective in many did increase boards were cases attached to each school Training of school principals 2-year overseas 32 Trainers were Trainers trained fellowships for trainers trained principals Preparation of training Training modules Modules given to Principals were trained modules in management principals supervision for principals Training of principals for 11,600 principals 65% of target Reportedly avery Type 2 and 3 rural were to be schools were positive impact on schools trained covered, i.e. 8,600 school administration principals Regional education tours Tours were carried Impact unknown for senior officials out; officials liked though reportedly them positive Education planning and policy International training 13 fellowships All fellows Unknown whether analysis awarded completed and they became more returned effective teachers Technical assistance Only short-term Marginal consultants were hired for studies Computer supply Computers provided Formed basis for Management Information System  15 Componentsl Target Subcomponents Activities Achievement Outputs Outcomes Development of MIS working in System partially All provinces able to management information all provinces working provide basic school system statistics Studies of management Study of teacher Carried out Proved very useful for systems and personnel attendance follow-on project procedures Teacher. handbook of Printed handbook Unknown whether it operational rules reduced the still high absenteeism rate; use mandatory Vocational education study Education-employment Contributed to the No results yet on linkages design of ADB recommendation secondary and effects, but increased vocational awareness of issues education projects Additional studies Physical facilities Developed a Recommendations No results yet school map used in follow-on project, closure of unneeded schools Management structures of Adjusted Three studies To be applied during MEHE GEP II Financing of education Revealed Findings to be used Financing formula to important in establishing be implemented in inequities in the financing formula 1998 system Teacher absentees and Study led to new Reduced absenteeism teacher education plan project from 35% to 20% in 1997 Maintenance handbook Handbook available Unknown Results 3.25 The project substantially achieved four of its five objectives. It did not calibrate education and training to labor market, but it significantly expanded capacity in primary education, improved management and regional cooperation, and helped reduce regional and local disparities. In terms of policy development and implementation, it achieved a great deal more than it set out to do. 3.26 Possibly as a result of principals' training, the schools seem to be under control. Schools were clean and neat despite poverty. Principals seem able to carry out the administrative work needed to run the schools, and teachers seem to be control of the classes (when they are present). Such characteristics are included in the hallmarks of effective schools. 3.27 An important factor in the successful implementation of this project was close and regular communication withprovincial officials. Due to devolution policies, MEHE had to plan  16 the project inputs but rely on provincial offices for imnplementation. After some initial frustration, the ministry established an educational development committee to coordinate all donor projects. Provincial directors and secretaries met every month at the MEBE offices in Colombo and for several hours discussed details and problems of implementation. To deal with procurement issues, the IDA, and other donor-financed projects, provincial accountants and engineers held monthly meetings before the directors' meeting; the problems that could not be resolved were brought before the committee. These contacts greatly increased the implementation capacity at the provincial level. It was very significant that these monthly meetings were held regularly and were maintained throughout the project. They continued after the GEP closed. Box. The Meetings of the Education Development Committee Many projects create coordinating boards and committees, but their meetings frequently become irregular and eventually cease altogether. What made it possible for Sri Lanka's educational committee meetings to be held every month for at least four years? The mission observed one such meeting. About 50 people-mainly provincial directors and secretaries of education-attended from all over the island. To accommodate Tatnil participants, the meeting was held in English. Some food was served, but there was no break; the work went on uninterrupted for about six hours. A detailed agenda was handed out, and each donor project was discussed separately. During the meeting observed by the mission, an official appointed by the prime minister to oversee education reforms explained the planned implementation of those refonns in detail and asked for feedback. Participants said they had to come every month to this meeting. If someone could not come, a replacement was sent. They found this the only workable means of coordinating the efforts of eight provinces with the center. Another important reason for attending was political. Some provincial governments represented opposition parties, who did not always communicate effectively with the central government. Participants saw themselves as the contact point between opposing political parties and viewed the meetings as a means of fnding out government policies in a neutral environment. Very important was the fact that participants had the means to come to the meetings. No honoraria were paid, but they traveled using provincial cars, many provided by the GEP. Most needed only 2 to 3 hours to get to Colombo and returned the same day. Participants from the northeastern part of the island had to travel for six hours and spend the night in Colombo. They received adequate allowances to buy gasoline and stay in hotels if they did not have relatives with whom they could stay. They were reimbursed easily and quickly. Are such meetings especially sustainable in Sri Lanka? Not necessarily. One meeting participant reported that the Ministry of Agriculture had once held coordination meetings, but they were not sustained in the long run. Along with the means to come, and with the political importance of the meetings, therefore, the leadership of MEHE may have been an important factor.  17 Monitoring and Evaluation 3.28 The project essentially had no monitoring activities. This is unfortunate, because its work could not be measured. It would have been particularly useful for future project activities, for example, to measure the effects of principals' training on teacher and student performance. Institutional Development 3.29 Institutional development is rated as modest. The successful training of master trainers and the effective implementation of principal training courses resulted in a wide dissemination of management principles. Also, the implementation experience that provinces gained and the effectiveness of the education development committee resulted in substantial institutional development at the provincial level. However, the extent of institutional development envisaged in the project was not achieved. Sustainability 3.30 The sustainability of the outcomes is rated as likely. Buildings, curricula, textbooks, and teacher training will probably have a sustained effect on the government's ability to provide education in Sri Lanka. IDA Performance 3.31 IDA's performance, while satisfactory overall, was uneven. Quality at entry was marginally satisfactory, because the objectives were many and disparate. Also, important policy issues and deficiencies (such as a lack of school mapping) were glossed over to concentrate on civil works. The quality and quantity of supervision were unsatisfactory for the first two years, when Sri Lanka operations changed country departments and task managers. Supervision missions were short and often did not include educators. 3.32 Supervision became highly satisfactory after the mid-term review. It is impossible to ignore in this project the qualifications of the task manager who took over at mid-term. This staff member had lengthy experience in classroom teaching, school management, provincial administration, and in-service teacher training, but he also had advanced academic credentials in education. This profile is unusual among Bank educators. The task manager's professional qualifications were enhanced by a willingness to take risks, to supervise consistently from headquarters, to provide sophisticated analyses to the government, to persist, and to help officials understand study results so that they could be adopted. 3.33 Government officials widely agree that IDA analyses and initiative were critical not only to project completion but also to the policy reforms the project initiated. However, some officials expressed concern that many of the reforms were generated outside the government system and warned of lack of ownership in such cases. Nevertheless, it is not apparent that task manager leadership compromised borrower ownership.  18 Borrower Performance 3.34 Overall, borrower performance was satisfactory. The fact that the government overcame the delays caused in the first half of the project indicates considerable implementation capacity, which is being tested in the two follow-on projects. 3.35 Examination of project files revealed some minor problems that were not resolved. The first project director, who did not perform satisfactorily, was allowed to serve during the entire project, thereby compromising the project. Annual audit reports in the project files have indicated some losses and use of funds for purposes not intended. For example, some funds for upgrading substandard Type 3 schools were allocated to national schools (1993 audit); some vehicles were found to be used about 50 percent of the time for nonproject purposes, and some stationary and chairs were lost. Govermment officials who were questioned did not recall such incidents.  19 4. Issues and Recommendations 4.1 An important factor in improving the quality of education in poorer schools is the presence of teachers in schools and their performance. The GEP started out with a study on teacher absenteeism and proceeded to a separate project to deal with these problems. Nevertheless, these problems still existed at the time of the audit. Teacher Absenteeism 4.2 Although teachers do not abandon classes at whim, as they used to, the liberal leave policy enables them to be absent for 1.5 month of the school year. The number of leave days that teachers have in Sri Lanka is excessive by international standards. Certainly changes in labor conditions are difficult, and acquired rights, such as extensive leave, often cannot be taken away without much struggle and unrest. But reductions in leave could be traded for other benefits and for future increases in salaries.4 4.3 Compounding the problem is the fact there is no system of substitute teachers in Sri Lanka. Sometimes excess teachers take over, sometimes students from nearby teacher colleges are asked to do so. In rural areas, principals do not have recourse even to these solutions. Use of volunteer teachers is prohibited for security reasons. Unsatisfactory arrangements may also be made when a teacher is pregnant and gets three months of maternity leave. There is clearly a need for substitute teachers or other suitable arrangements. Instructional Deficiencies in Teacher Training Institutes 4.4 Given the emphasis on teacher development, the mission visited one pre-service teacher training college, which had been built with Japanese funds. Despite the excellent buildings, the college did not appear to impart teaching skills to students effectively. The problems were: - Negative role modeling. Despite the emphasis on teaching child-centered and participatory methods, the classes themselves consisted of lectures. The teacher stood in front of the class, often at a considerable distance from the students, and lectured. Because there were no textbooks, teachers spent much of the class time dictating notes to students. * Lack of textbooks and materials. None of the classes observed (physics, music, mathematics) had textbooks. Students took notes and then studied from textbooks in the library. Instructional materials were also insufficient. The physical education teachers did not have sports equipment, and the music teachers had very few instruments. College administrators expressed concern about the limited equipment. 4. Frustration with political influence seems widespread, though in some areas it was reported as a larger problem than in others. Some provincial staff members expressed a dislike of principals who had been appointed for political reasons rather than on merit. There was widespread concern that angering politically active educators could result in difficulties and unfavorable transfers.  20 4.5 Textbooks for most subjects do exist, and students can buy them at reasonable prices, but they are not asked to buy a single textbook for a course. As in many other countries, it is thought that university-level students are mature enough to study from a multitude of texts and learn enough about a subject the professor has assigned. However, research does not support this assumption, and the use of multiple sources is a very inefficient use of time. Many students do not spend the needed hours in libraries, and they study mainly from notes. The result is that they cover very little material and in insufficient depth. 4.6 To improve the quality of pre-service education, professors must teach trainees as they expect them to teach later in classes. Research shows that role modeling changes teaching behavior more than textbooks discussing the benefits of the modeled behavior. By memorizing information about participatory processes in a lecture environment, teachers receive conflicting messages. It is not surprising, therefore, that they do not necessarily practice what they learn. The mission observed much wastage of classroom time that resulted because the teacher worked with one student, and the rest of the class had no meaningful work. Reportedly, teachers had been lectured about the wastage of time, but college experiences teach them differently. 4.7 Classroom instruction will become much more efficient if a single textbook is assigned for the course, and students unequivocally know what to study from day to day. Most textbooks are available in Sinhalese. Methodology textbooks may have to be written in Sinhalese, but many existing in-service instructional modules could serve for this purpose with some modifications. Extra Burdens of the Poor Students' Schools 4.8 The open classrooms of Sri Lankan rural schools are an unusual feature by international standards. The lack of self-contained classrooms is a major problem that interferes with the instructional process and diminishes the quality of learning. Clearly research on building designs is needed to balance the needs for air circulation versus attention to the lesson. Noise levels can be measured in various classroom types and related to teacher satisfaction levels, quality of air, and student achievement. Results will probably indicate that in the majority of schools, sound- blocking walls must go up. These may be built from relatively inexpensive materials, using various soundproofing techniques. 4.9 Overall, building condition can become a considerable source of work for principals and teaching staff. Because schools have a limited budget to carry out repairs, they depend considerably on parents' good will for maintenance and must make contact with them. When buildings have defects, principals have to write letters requesting repairs and deal with teacher and student complaints. Since building and maintenance problems have been identified in remote and poor schools, the burden of extra work seems to weight disproportionately on the staff of poor schools. There is a concern that such principals and teachers may wish to transfer to better schools and escape the problems. 4.10 The problems of open classes, building defects and low-quality furniture seem to accumulate on the schools of poorer children. It may cost more in contractor and supervisor fees to build schools in remote areas, but this issue can be taken into account. The problems of poor schools are compounded with existing teacher recruitment policies. To improve equity, at the very least, teachers must be deployed on the basis of clear and objective criteria, such as the "ready reckoner" formula that was used in the 1980s.  21 5. Lessons * Organized, long-term communication between ministries and regional offices clearly improves implementation outcome. The challenge is to provide the circumstances to make this possible. Incentives and easy means for commuting are important in allowing committees to meet repeatedly over time. * Task manager diligence, knowledge, and implementation orientation can significantly influence project outcomes. The Bank should use educators who have detailed knowledge of the field and who are willing to innovate and to take risks. * Insufficient attention may be paid to the instructional consequences of building designs. Projects should study noise and heat levels in schoolrooms of different types relative to instructional outcomes. * Remote schools are more likely to have cumulative problems, including low-quality construction. More support in time and effort should be made to supervise those sites. D Instructional materials, particularly for science, may stay unused until expertise and a real motive for using them exists. In addition to a clear rationale for choosing each item, training must be provided in their use, and teacher supervision must include use of instructional aids.   23 Annex Basic Data Sheet GENERAL EDUCATION PROJECT (CREDIT 2072-CE) Key Project Data (amounts in US$ million) Appraisal Actual or Actual as % of estimate current estimate appraisal estimate Total project costs 75 60.02 80% Loan amount 49 55.80 114% Cofinancing 0 Cancellation 0 0 0 Date physical components completed 12/31/95 12/31/96 one year Cumulative Estimated and Actual Disbursements (amounts in US$ thousands) FY90 FY91 FY92 FY93 FY94 FY95 FY96 FY97 Appraisal estimate 2,500 8,100 16,700 26,200 36,600 46,100 49,000 49,000 Actual 0 1,033 3,656 11,403 15,867 26,497 46,317 55,802 Actual as % of appraisal 0% 13% 22% 44% 43% 57% 95% 114% Date of final disbursement: June 1997 Project Dates Steps in project cycle Original Actual Identification March 13, 1988 Preparation August 1988 August 22, 1988 Appraisal April 1989 May 7, 1989 Negotiations June/July 1989 October 4, 1989 Board presentation August 1989 December 12, 1989 Signing - February 14, 1990 Effectiveness November 1989 April 19, 1990 Project completion June 30, 1995 December 31, 1996 Loan closing December 31, 95 December 31, 1996 Staff Inputs (staff weeks) Steps ofproject cycle Planned Revised Actual Weeks US$000 Weeks US$000 Weeks US$000 Through appraisal 39.0 72.5 39.0 72.5 39.0 72.5 Appraisal-Board 19.3 39.7 19.3 39.7 19.3 39.7 Board-effectiveness 3.7 9.1 3.7 9.1 3.7 9.1 Supervision 58.6 159.7 56.5 159.8 64.8 180.6 Completion 5.0 7.4 5.5 17.2 4.7 14.8 Total 125.6 288.4 124.0 298.3 131.5 316.7  Annex 24 Mission Data Performance rating Stage of Date No. of Staffdays Specializ. Implement Develop. project cycle (monthlyear) persons infield representeda status objectives Types ofproblems Identification March 88 1 29 EC Preparation Aug. 88 3 23 EC ,GE ,C Preappraisall May 89 5 24 EC, GE,C,C,C Appraisal Supervision I Aug. 90 2 7 EC, GE. I I Supervision II April 91 1 6 EC 2 1 Supervision III Jan. 92 2 11 EC, ED 2 1 Disburs. lag; absence of progress & audit reports; & using societies for construction Supervision IV June 92 2 17 ED, EC 2 1 Disburs. lag and absence of progress and audit reports Supervision V June 93 3 14 IS, ED, EC. 2 2 Disb. procedures markedly behind; no submission of progress & audit reports; lack of close superv. by a sen. manager of MOEHE; & pr.steer. commit. inactive. Supervision VI October 93 1 6 EC 2 2 Disb. & accounting procedures remain unsat.; pr. not supervised by sen. ministry official; & MDU has not become the agency for admin develop. Supervision VII Feb. 94 4 13 SEP, A, POO, 2 2 Mid-term review; reviewed SA project objectives Supervision VIII June 94 2 7 SEP, SA 2 2 Project staffing; slow disbursement; slow procurement Supervision IX Sept. 94 1 13 SEP. S S Serious disburs. gap; slow proces. of procurement & disburs. applic.; lack of adequate staff at PCU; & inadeq. monitor. & lack of importance attached to implemen. schedule Supervision X Jan. 95 2 7 SEP, EC S S Supervision XI April 95 2 10 SEP, EC S S Supervision XII Oct. 95 2 5 SEP, EC S S Supervision XIII March 96 2 7 SEP, EC S S Supply of counterpart funds Supervision. XI July 96 2 2 SEP, EC S S Slow disbursement Supervision. XV Nov. 96 2 4 SEP, EC S S a. A=Architect; C=Consultant; ED-Educator; EC=Economist; GE=General educator; IS=Implementation Specialist; POO=Project Operations Officer; SA=Senior architect; SEP=Senior Education Planner.  25 Annex Other Project Data Borrower/Executing Agency: FOLLOW-ON OPERATIONS Operation Credit no. Amount Effectiveness (US$ million) Date Teacher Training Teacher Development 2881-CE 64.1 August 1996 Project General Education II Project 3014-CE 40 January 1998