Burundi - Third Education Project

Document of The World Bank FOR OFFICIAL USE ONLY Report No. 9543 PROJECT COMPLETION REPORT BURUNDI THIRD EDUCATION PROJECT (CREDIT 1358-BU) APRIL 26, 1991 Population and Human Resources Operations South Central and Indian Ocean Department Africa Regional Office 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 SDR I = US$1.089655 - FBu 98.06895 MEASURES AND EQUIVALENTS (Metric system in use throughout report) ABBREVIATIONS AND ACRONYMS AfDB - African Development Bank BASE - Bureau Africain des Sciences de l'Education BEET - Technical Education Bureau BEPES - Secondary Education Bureau BER - Bureau d'Education Rurale (Office of Rural Education CODIFOR - French Agency EPU - Education Planning Unit ETP - Ecle des Travaux Publics (Public Works School) GTO - Offset Press IDA - Inrternational Development Association IIPE - International Institute for Education Planning INFRE - Institut National de Formation et de Recherche en Educ. JSS - Junior Secondary School (First Cycle Secondary) MLC - Multipurpose Learning Centers (Primary Schools equipped for practical works) MNE - Ministry of National Education MSA - Ministry of Social Affairs OAU - Organization for African Unity PCR - Project Completion Report PIU - Project Implementation Unit RPP - Regie des Productions Pedagogiques (Academic Printshop) SC - Social Centers (Young and Adult Education Center) TTI - Teacher Training Institute UNESCO - United Nations Education, Scientific and Cultural Organization UNDF - United Nations Development Program  FOR omcUL uSK ONLY THi WOItLO BANKt Washin5ton, D.C. 20433 U.S.A. OMc a Oiuctmmns,a Opwathm 1vo - tki April 26, 1991 MEMORA-IDUM TO THE SEECUTIVE DIRECTORS AND THE PRESIDENT SUEJECT: Project Completion Report - Burundi Third Education Project (Credit 1358-BU) Attached, for information, Is a copy of a report entitled "Project Completion Report - Burundi Third Education Project (Credit 1358-EU)J" prepared by the Africa Regional Office. No audit of this project has been made by the Operations Evaluation Department at this time. Attachment ' This document hs a rtrkicted distribution and may be jied by recipients only In the performance of their oflcial duties. Its contents may not otherwisc be disclosed without World Bank authoriation.  Page Table of Contents (Cont'd) 4. P:oject Costs and Financing 45 5. Project Results 46 6. Compliance with Credit Agreement Covenants4 7. Use of IDA Resources 49 ANNEX 1 Exchange Rate Fluctuations 51  - i - REPUBLIC OF BURUNDI THIRD VEDUCATION PROJECT (Credit No. 1358-BU) PROJECT COMPLETION REPORT Preface This is the project complotion report (PCR) fo: the Third Education Project in Burundi, for which Credit 1358-BU in the amount of SDR 14.5 million was approved on May 10, 1983. The Credit was closed on the scheduled date of March 31, 1988. The balance of SDR 73.56 was cancelled on September 13, 1988. The PCR was prepared jointly by the Population and Human Resources Division of the Africa South-Central and Indian Ocean Department, Africa Regional Office (Preface, Evaluation Summary and Parts I and III) and the Borrower (Part II). Preparation of this PCR began during a UNESCO miasion financed by AF3PH that visited Burundi from March 18 to April 5, 1989 and which produced a factual roport that served as a basis for Parts I and III. The appraisal report, Credit Agreement, supervision reports, correspondence between the Borrower and IDA, and IDA internal memoranda were also used.  ,- iii q REPUBLIC OF BURUNDI THIRD EDUCAkTION PROJECT (Credit No. 1358-BU) PROJECT COMPLETION REPORT Evaluation Summary Proiect oblectives 1. Designed as a continuation of the first two education projects (Credits 679-BU and 976-BU), the project was aimed at increasing the number of schools, primarily in rural areas, upgrading primary and secondary education, and training young people and adults without formal education. More specifically, project activities included the construction and equipping of 24 primary schools (MLCs), 4 junior secondary schools (JSSs), 2 primary teacher training institutes (TTIs) and 20 social centers (SCs) for adults and adolescents without formal education, production of textbooks, revision of curricula for the tt %ining of teachers and adults, and the improvement of product management. Two years before the end of the project implementation period, because of the excellent progress, the means for achieving the objectives were strengthened through education planning assistance, greater training of national specialists and the construction of additional primary and secondary schools (Part I, paras. 3 and 4). Product implementation 2. The project was implemented on schedule and achieved and even exceeded its quantitative objectives. It was flawlessly administered and managed by the Project Implementation Unit (PIU), which performed the customary administrative duties (personnel management, liaison with IDA -d technical government agencies, accounting and preparation of semi-annual activity reports), while also overseeing the construction on force account of most of the physical'elements of the project. Some construction was carried out by private companies under the supervision of the PIU, which also prepared the architectural programs (Part I, paras. 6, 7 and 8). 3. The project was amended on April 22, 1986 to take advantage of the advance of some 12 months over the timetable and the excellent financial situition due primarily to a favorable SDR/FBu exchange rate. The amendment made it possible to increase the number of places for boarders in the secondary schools, the number of which was also incresed. Proiect results 4. Quantitative results. All quantitative objectives regarding the number of student places were achieved and even exceeded, but some schools (TTIs and SCs) have not yet reached the expected enrollments (Part I, para. 15 and Part III, SA). The quality of construction for all schools was good, as  - iv - regards both those constructed on forc account and those handled by private firms, The construction costs for the two approaches were apparently very similar. Nevertheless, given the reticence of private entrepreneurs to build smoll schools on sites with difflcult access, work on force account will always be necessary for the primary schools and some rural JSSs (Part I, para. 10). 5. Qualitative results. The full qualitative effetes can be felt only over the long term. However, it is already clear that thanks to the manager training program, senior positions at the PIU, Education Planning Unit (EPO) and the Academic Print Shop (Regie des Productions Pedagogiques - RPP) have been entrusted to Burundi nationals. Regarding the other componentst (a) tho MLC program deviated from what was envisaged during project preparation, following the elimination of the practical activities of the primary education program. The result was underutilization of premises and equipment, which the Govi-nment and IDA are trying to remedy (Part I, para. 16); (b) TTI enrollments are low and in order to attract more students, the Government decided to turn them into higher-level teacher training schools (lycies Ptdazogiues) giving access to university *ducation, following two years of teaching at the primary level. One big risk is that these schools will provide training that is more theoretical than practical and the best students will drop out of primary education teaching in order to go.on to university (Part I1 para. 17); (c) although the quality of education at the JSSs seems acceptable, the training program is occasionally constrained by the latge number of students per class since some classrooms are usod as dormitories, as are areas that were planned for libraries. In fact, the project financed boarding facilities for only 50 percent of the students, whereas practieally 100 percent are boarders. A solution to this problem was discussed as part of the sectoral adjustment measures under the Fourth Education Project (Part I, para. 18). Proj ect sustainability 6. All components that help train mauagers for positions of administrative, technical and teaching responsibilities will have sustainable effects. This training will allow activities to continue, and help project management, preparation of new training programs, planning for the development of the education system and the production of teaching aids. For primary and secondary education, and to a lesser extent for socio-educative training, the new schools will unquestionably provide access to education for a growing number of children and adolescents in areas where such opportunities were previously limited. The only possible risk is financial in nature, with the Government having to carefully monitor its spending so that it can meet increased expenditures related to its quantitative and qualitative performance (Part I, para. 20). Conclusions and lessons 7. The financial and qualitative difficulties encountered during project implementation (Part I, paras. 17, 18, 19 and 20) have led to a new awareness of the close connection between the quantitative and qualitative  v dovelopment of the system and its financial impact. Financing problems, whether resulting from the Increasing costs of boarder students, political decisions decreeing universal education for children 7 years old, an lncrease in the primary/secondary transition rate or future reforms, can be resolved only if accompanied by measures to ensure savings through a carefully planned adjustment program. The Government is now aware of this, and IDA should ensure that its future participation in the development of Burundi's education system occur in this context and stress the structural and qualitatlve aspects as much as, if not more, the quantitative problems.  REPUBLIC Ol BURUNDI THIRD EDUCATION PROJECT (CREDIT 1358-BU) PROJECT COMPLETION REPORT PART I A. Project Identity and Key Indicators Project name: Third Education Project Credit: 1358-BU Country: Republic of Burundi Sector: Education and training Subsector: Primary education Secondtry education Socio-educative activities Amount of credits SDR 14.5 million Board approval: May 10, 1983 Original closing date: March 31, 1988 Actual closing date: March 31, 1988. B. Socioeconomic Background 1. When the Third Education Project was generated, the country's economy was beginning a decline in the wake of a relatively favorable period. At the end of 1981, the underlying weaknesses in Burundi's economy began to resurface, especially because of a plunge in coffee prices. In addition, decisionmakers were faced with difficulties related to the country's landlocked position, widespread poverty (in 1982, per capita GDP was US$218), scarce natural resources, a shortage of available farmland, few urban areas with the exception of tne capital city of Bujumbura, sparse settlement and jagged terrain, making communications difficult, and lastly, low literacy (in 1982, less than 10 percent of the adult population was literate and 29 percent of children 7 to 12 years old had been to school). The new Five-Year Development Plan (1983-1987), while setting ambitious objectives such as expansion ini rural economic activity, decentralization and intensification of social pzograms with a view to expanding access of the most disadvantaged to education and health, and increasing agricultural production to meet food needs, also called for limited growth in operating expenditures. 2. The expected economic austerity hit the Ministry of Education at a particularly delicate moment. Indeed, the 1983-1987 Plan called for it to increase enrollment rates to the extent possible, from 29 percent in 1981 to universal education of children 7 years old by 1992/93. Even though, as all national education experts agreed, that date was overly ambitious, the other education targets resulted from social pressure and had to be achieved. Among others, these included dealing with an *nnual increase of 2.8 percent in the school-age population, maintaining, and if possible, modestly increasing the  -2 - transition rate between primary and secondary educatlon, and upgrading the quality of education, its internal and external efficiency and adaptation to the country's socioeconomic realities. All this was to be done with a current expenditure budget that was cut froir 22 percent of appropriations in the State budget in 1978 to 15 percent in 1981, and which was expLcted Ander the new plan to be 18 percent at 'he end of the year. C. Prolect Obiectives and Description 3. Although designed with two totally distinct aspects, one of hich was to be the responsibility of the Ministry of National Education (MNE) and the other that of the Ministry of Social Affairs (MSA), the overall objective of the Third Education Project was a quantitative and especially qualitative improvement in education and training. Access to education and training in the most deprived regions was to be increased and, to the e: went possible, to affect segments of the population that had not yet received formal education, through outreach activities at the social cen':ers. 4. More specifically, the project was to cover! (a) The construction, furnishing and equipping of: (i) 24 multipurpose learning centers (MLCs), each having six classrooms and one home economics workshop, and able to accommodate 480 students; (ii) 4 junior secondary schools (first-cycle secondary) (JSSs), each accommodating 320 students; (iii) 2 primary teacher traiaing institutes (TTIs), also able to accommodate 480 students. -- 160 teacher trainees and 320 general secondary students; -- (iv) 20 rural social centers (SCs), of which 2 were to be entirely new and 18 expanded and rehabilitated, each able to provide training for 120 adolescents or adults. (b) The provision of supplies and equipment necessary for the production and distribution of textbooks by the RPP. (c) Materials, equipment, services of specialists and fellowships needed to upgrade teacher training; development of training programs and methods for illiterate adults; implementation of the project construction program; and management of project activities as a whole. Later an amendment to the Credit Agreement dated April 22, 1986 authorized the financing of additional boarding places the JSSs and TTIs, construction of five additional MLCs, rehabilitation o& secondary schools, construction of premises for the RPP and Office of Rural Education (Bureau d'Education Rurale - BER), and additional technical assistance, (specialists and fellowships) for  - 3 - the RPP, social affairs department and two technical schools opened under the second project: the Public Works School (Ecole des Travaux Publics - ETP) and the Secondary Technical School (Ecole Technique Secondaire - ETS). D. Pro1ect Design and Organization 5. The initial request by the Government for financing a third education project was intended to continue the second project. Thus, from the beginning, the intention was to finance the construction and the functioning of new MLCs, and to upgrade the quality of education and training for young people and adults. As far as the management of the project was concerned, since the activities of the second and third projects were intended to be implemented simultaneously (between November 1983 and June 1986), and since the existing management had proved very efficient when implementing the first project and the first phase of the second project (see Completion Report on the Second Education Project, Credit 976-BU, dated November 1987), the intention was to use the same existing resources i.e. personnel, equipment, vehicles and physical infrastructures. 6. Three important conceptual points that emerged quickly were discussed during preparation and negotiations. The first lnvolved the number- -deemed high by IDA representatives--of boarders. While the Government planned to resolve this problem in part by phasing out boarding capacity in the cities, IDA technicians also wanted the government to consider eliminating the countrywide competitive examination for entrance into secondary education, replacing it with a competitive examination at the arrondissement level, which would mean more localized recruitment and thus less student travel. Despite the formulation of a detailed school map that was a condition of negotiations and was designed to facilitate such regional recr itment, the national competitive examination is still given. The s'"-ozid point discussed was related to the construction of practical workshops in the MLCs. Given the low use of the workshops built with financing under the Second Education Project Credit 976-BU, IDA was reluctant to authorize the consttuction of a2UIonal facilities that would meet the same type of problem. Ultimately, it was agreed to build a single room for home economics courses, with the exception of the two TTIs for which the cor,struction of three iooms for practical work was finally decided. The last point concerned the role of the PIU in construction and procurement. As one of the Bank's key objectives in Burundi was to foster the emergence of small and medium enterprises in the private sector, it seemed contradictory to award all construction works without bidding to the PIU, which during the implementation of the second project became the largest construction enterprise in the country, with some 1,6J0 contractual employees. Nevertheless, given the excellent results, both from the qualitative and financial standpoints, and the fact that no reliable firm would seek contracts for medium-sized buildings (MLCs, JSSs) on remote sites to which access was often difficult, it was finally agreed during negotiations that construction would largely be awarded to the PIU, but that small local firms would be used for all contracts of less than the equivalent of US$300,000. During negotiations, the Government and IDA also agreed that some 10 MLCs and/or SCs would be awarded to private firms in order to promote private enterprises and test their competitiveness.  - 4- E. Prolect ImDlementatimn 7. Given the good results of the first two education projects (Credits 679-BU and 976-BU), in which the objectives were achieved in full for r.ll physical and material elements, no major difficulties had been anticipated. Project implemtentation confirmed this, and the results show that among all projects in the sector, the Third Education Project in Burundi can be considered an example both for the speed with which the construction was carried out and the equipment delivered, and for the quality of the works and high level of discussions between the Government, IDA and other major donors such as Belgium, France and UNDP, on adapting the education system to local circumstances and the measures to be considered for its improvement. Nevertheless, problems related to costs and financing in the sector, control of growth in enrollment, and the quality of administrative and teaching personnel, as well as the contents of programs and curricula or training streams to be opened or abandoned depending on the real needs of the country, were neither broached nor discussed under the project. 8. As for the two previous projects, the construction was carried out on force account by the PIU, with the exception of two MLCs and three SCs. which were built by private firms. The PIU was also responsible for the preparation of bidding documents, review of bids, contract signature and distribution of furnishings and equipment to the various schools built under the project. All these tasks were carried out in a highly professional manner; record-keeping regarding accounts, inventory control, personnel and daily activities done with the assistance of the most modern means, including computers, largely facilitated the supervision of the successive IDA missions. The PItl was also responsible for implementing an education project financed by the African Development Bank (AfDB) and the end of the second IDA project. Lastly, the PIU's efficiency meant that the project could be implemented ahead of schedule and below the cost estimated at the time of appraisal, largely owing to favorable changes in the SDR/FBu exchange rate. These changes were included in the Credit Agreement dated April 22, 1986, which content is presented in details at the end of paragraph 4 above. F. Prolect Results 9. Results relating to construction and equipment. All project objectives were achieved. Overall, the quality of work was good, and can even be considered excellent for most MLCs and SCs. Supplies accounted for 55 percent of the total cost of construction. Of that total, 54.8 percent was purchased following international competitive bidding and 1.5 percent after local bidding. The balance of 43.7 percent of the total amount was obtained through negotiated contracts, primarily involving purchases of cement (5 percent), roofing materials (27 percent), fuel, transportation and tools (44 percent); and electricity and plumbing (25 percent). 10. As agreed (para. 6), two MLCs near Bujumbura (Magara and Mageyo) and three SCs with relatively easy access (Kivyuka, Kinama and Giheta) were built by private firms. According to cost records and PIU calculations, the construction cost on force account was about 38 percent to 50 percent lower  - 5 - than that of private firms. A UNESCO mis3ion financed by IDA for the preparation of the Project Completion Report proposed a mesor revision of the PIU's cost estimates to reflect services by expatriates, depreciation of vehicles and equipment, taxes and cash reserves at the disposal of the PIU. The mission thus suggested that to obtain comparable figures the cost of construction on force account had to be multiplied by 1.68. In point of fact, any comparison of the two approaches is very difficult. If all work had been done by private firms, it would still have been necessary to use expatriates to supervise preparation of the bidding documents, bid review, contract preparation and frequent worksite inspections. Similarly, many private entrepreneurs are given tax exemptions for work done on 1iehalf of the State. Moreover, some bidding for small structures (MLCs) or sites with difficult access did not attract any bids from the private sector. Lastly, as the PIU added two dwellings with three bedrooms to all MLCs and a three-bedroom house to all SCs, which were not built by private firms, a comparison was difficult. The preparation mission for the Fourth Education Project, after having scrupulously examined the cost and calculation bases that had been used in their appraisal, recommended that all small and medium rural structures (MLCs and JSSs) continue to be built on force account, an approach whose advantages had been demonstrated in terms of cost, quality and timetable, whereas the larger and more complex constructions, including the rehabilitation of the major urban secondary schools, would be contracted to the private sector, under the close and constant monitoring of the PIU. 11. Furnishings were, in ,eneral, of average quality. However, some problems were experienced during their procurement: three firms did not comply with the contracts signed and wanted to deliver low-quality furniture causing delays in delivery time. To avoid other problems of this type, the Government dezided that under the third project, local artisans would be contracted to manufacture simple but sturdy furnishings from prototypes designed by the PIU. As far as the MLC equipment purchased under the Credit Is concerned, it was covered by two rounds of bidding, the first international for 97 percent of the total and the balance being purchased locally. The equipment, well adapted to needs at the time of appraisal, was not all delivered to the MLCs following the elimination of the practical work component of the primary education program. 12. Results relating to technical assistance and fellowships. The projei.. received significant amounts of technical assistance. General PIU management received the services of an accountant, a procurement specialist, three archi1ects, an engineer and a head of the vehicle fleet. Belgium financed the services of an architect. This situation was inherited from the second project, still incomplete at the time implementation of the third project started, and the assistance was also used for the implementation of the ifDB project. The various IDA supervision missions had all insisted on the urgent need to train local counterparts and technicians, and the situation in this regard was significantly changed with the departure of the accountant, the head of the vehicle fleet, and an architect, who were re'laced by local technicians. Two Burundese architects were also recruited. 13. In order to ensure that the project's effects would be sustainable, technicians assisted and trained local counterparts in initial  -6- and in-service teacher training and in the preparation of TTI curricula (two UNESCO experts for a total of 3.8 staff years), preparation of an adult literacy program for the Ministry of Social Affairs (MSA) (a UNESCO expert for 1.5 years), training in printing techniques (a UNESCO expert for 5.5 years), and the continued services of a public works engineer begun with financing under the second project and continued for two years under Education III. The services of these specialists were provided in accordance with projections, were deemed to be of good quality and greatly boosted the qualitative results. 14. Originally, the project fellowship program provided for the equivalent of three years of training abroad for four RPP officials, at the rate of nine months each. In actuality, nearly seven years were used, 2.85 years for the RPP, 1 year for the Kamenge FTS and 2 years for the BER. The project also financed internships for two teachers from the ETP (Public Works School) (five months each); nine specialists from the BER who traveled to Africa and Europe to study education reforms, preparation of curricula and teacher training; three officials from the Ministry of Social Affairs who went to Benin, Ethiopia and Mali for advanced training in adult education and literacy action. Lastly, with the assistance of the International Institute for Education Planning (IIPE), the education planning office organized four seminars on regional planning, collection of statistics, school mapping and use of computer aids. The training, which kept to the schedule and the conditions set in advance, largely helped improve the technical skills of the beneficiaries. 15. Quantitative results. Quantitatively, the project results exceeded, sometimes substantially, the objectives of the Credit Agreement. Instead of 24, 29 MLCs were built and through double-shifting (one group of students in the morning and another in the afternoon, with a single teacher for both), 13,920 students could be accommodated instead of the 11,570 envisaged. The new TTIs offar two cycles, i.e., four years of the first cycle of secondary educaw.ion (JSS) (grades 7 to 10) and two years of primary teacher training (grades 11 and 12). Despite an increase in the number of boarders that could be accommodated at the two TTIs, the number of trainee teachers was lower than expected, whereas that for first-cycle students was higher (about 850 instead of 640). However, in December 1988 the Cibitoke and Mukenke TTIs had only 115 trainee teachers instead of the 320 r3timated at appraisal. The four additional JSSs built under the project can accommodate 1,450 students, i.e., 10 percent more than the number of places expected at appraisal. Following an amendment to the Credit Agreement, 50 percent of those students can be boarded. Under the fourth project, the increase in the number of boarder students in rural areas should be offset by a phasing out of boarders in Bujumbura and Gitega. The quantitative results for the social centers were, however, more disappointing, as only an average of 52 people attended each center instead of the 120 projected. This is a qualitative problem that will be discussed below. Lastly, the production of textbooks and teacher guides met and exceeded the targets. Between January 1984 and October 1987, the RPP printed 61 different book titles represent'ng some 1,350,000 books. 16. Qualitative results. The most readily observable qualitative results involve the training of local officials at the PIU, planning office, BER and RPP. Nearly 80 percent of those trained, either on the job by  - 7 - technical assistants or via fellowships abroad or local seminars, hold high- level administrative, technical or teaching positions. However, for the other components, as is always the case of an education project, the qualitative impact is measurable only over the long run, when the students have completed the primary or secondary cycle using improved educational materials or studying untar better-trained teachers. Thus, for the MLCs it is difficult to predict the long-term qualitative impact of the double-shift system and the elimination of the practical subjects program. For the time being, the premises and equipment for practical subjects are underutilized, and some workshops should be converted immediately into classrooms in order to reduce congestion and thereby mitigate the negative effects of two shifts, wherever possible and compatible with budget projections. Over the longer term, the new workshops could be converted into classrooms for postprimary courses. This could help reduce the number of years of study in the JSSs and pare their enrollments, which are very high. It was also proposed that the rooms be allocated for social programs such as kindergartens, health care centers and additional training for artisans. No decision has been made yet, but the Government is evidently leaning tow_.rds the establishment of several pilot MLCs that could benefit from appropriate equipment and materials, and whose qualitative effects could be assessed later. 17. The first teachers trained ir the project TTIs are not yet exercising their professions, but their curriculum, prepared by two UNESCO specialists financed by the project &nd Belgian and French civil servants, is satisfactory. However, major changes are likely to occur in the near future. In the recent past, three streams -- general high school, technical high school and TTI -- were offered to the students leaving the JSS. For social and financial reasons -- low salaries, difficult job constraints, obligaticn to accept a job in remote rural areas -- students were not enthusiast to become teachers and TTI were functioning at half capacity. In responding to this problem, the Ministry of Primary and Secondary Education decided to convert some TTIs into higher-level teacher training schools (lycees pedagogiques), thereby introducing a second secondary cycle. The graduates of these high schools must teach for two years at the primary level and then, if they so desire, can go on to de university-level studies. Although this approach seems to attract more students to the TTIs, it may result in more theoretical training programs an" siphnn the majority of the primary education teachers away from primary schools, part of them being admitted at university and others,Vfrustrated, resigning and leaving the profession. A solution to this problem on how to attract more students to TTIs is an important point in the continuing dialogue between Government, IDA and other major bilateral donors (Belgium, France). Obviously the solution is not to hamper students to go to university, but, on the contrary, to create favorable conditions for the primary school teachers -- competitive salaiies, housing, school equipment and career development plans. 18. The quality of JSS education could be acceptable if the boarder problem does not undermine the conditions under which it is dispensed. In effect, in rural areas, whereas existing facilities can house only 50% of the JSSs'students, all of them are, in fact, boarders. As a result, the libraries and some classrooms are used as dormitories, and these are overcrowded. The ongoing dialogue between the Government and IDA is stressing the need to  -8 - reduce the costs of secondary education by restricting boarding only to those students living in remote rural areas. The others would have to be selected on the basis of a regional competitive examination for a JSS near their home. In the meantime, to make up for the cost due to the increase in the number of boarding places in rural areas, urban boarding will have to be phased out. The Government is already taking some important measures in that direction. 19. The 20 social centers (SCs) have two objectives: to raise the cultural level of young people through literacy action and to give them a trade by teaching them farming and stock raising, health and cooperative management, and giving instruction in civics and patriotism. The Government hopes this will keep them from moving out of the rural areas. The training, which lasts three years and is held three mornings each week, is provided by teachers and outreach workers. The training is very theoretical, probably too much so to attract a large number of students, and consequently the facilities and teachers are generally underutilized. G. Project Sustainability 20. The most lasting positive effect of the project is the development of local capacity for carrying out major administrative, technical and teachir- tasks. Whether involving administrators and/or architects at the PIU, sp.vialists in curriculum development at the BER, planners, education statisticians or technicians at the RPP, the training achieved under the project will mean that in the future technical assistance will be needed for only very specific requirements. For primary and secondary education, and to a lesser extent socio-educative training, the new schools will unquestionably provide access to education and training for growing numbers of children and adolescents in areas where such opportunities were previously limited. Also, because of the qualitative improvements that will certainly ensue from the production and dissemination of teaching aids in greater number and of better quality, anA more teachers, who are better trained and recruited on the basis of appropriate planning and school maps, the effects of the project should be sustainable and, if the Government can allocate the necessary financing, should continue to grow. The only foreseeable risk for this part of the project is associated with the budgetary aspects. Meeting operating expenses, including expenditures for boarders, requires that the Government adopt urgent measures regarding the control of unit costs and the distribution of the education budget. Likewise, decisions concerning boarding and the recruitment of future teachers, who will have a major impact on operating costs, must be taken without delay. The preparation and implementation monitoring of these measures will be under the responsibility of the Education Planning Unit which prepared, under the second education project, a school map and a first education development plan. However, the EPU was not sufficiently equipped to face the cost and financing problems of the education system. The Fourth Education Project is therefore stressing sectoral adjustment measures, in response to a situation that could quickly become critical. The EPU will be strengthened to handle efficiently this situation.  -9- H. IDA's Performance 21. IDA should be credited with several points. First, the project was identified, prepared and negotiated before the completion of the second education project, showing that IDA staff involved was aware that the needs were real and that the status of activities under the second project could serve as a reference for the third. Additionally, the project benefited from the fact that a large proportion of its supervision missions included an architect and an educator. This made it possible to continue a dialogue that, although it did not lead to a resolution of all problems, at least enabled them to be raised and clearly identified, so that they can be taken into account in the formulation of sectoral adjustment measures, which are the cornerstone of the Fourth Education Project. Lastly, IDA demonstrated great flexibility in amending the project and financing elements not envisaged at appraisal, but which were subsequently shown to be extremely beneficial (study trips, organization of seminars, reorganization and expansion of the RPP and the BER, technical assistance to the ETS). 22. However, three important aspects were not sufficiently taken into account by IDA. First, IDA should have taken a firmer position regarding the MLC workshops and, on the basis of what occurred during the second education project it has been proved that practical works in primary schools were non- efficient, should have refused to finance facilities and equipment for practical works. Instead IDA should have agreed on construction of more rural primary schools. Second, IDA should have taken a less ambiguous position on the boarding problem. Linking this problem to the education financing global problem, IDA staff should have made it clear that boarding facilities and functi-oning expenses would be neceseary for most of rural students and that, to finance them, phasing out boarding expenses in rural areas were almost mandatory. Third, regarding the management of the project (PIU), IDA should have proposed a plan to decrease the number of expatriate specialists and to train local counterparts. All those three sections are now under implementation as part cf the fourth education project. In the future, IDA will have to ensure that its contribution to education development in Burundi stresses the structural and qualitative problems as much as, if not more than, the quantitative aspects. I. Borro*er's Performance 23. This report has heavily stressed the quality of project management. In addition, the Government was at all times able to use project results optimally, particularly by maintaining the officials in the positions for which they had been trained. Another positive aspect to the borrower's credit was its willingness to discuss the issues and its constant availability, either during exchanges of correspondence or missions. The quality of the dialogue led to greater awareness of the close linkages between quantitative and qualitative improvement in the system and its financial impact. During the course of project implementation and thereafter, the Government adopted, and is currently implementing, several drastic measures which demonstrate its willingness to link quantitative and qualitative improvements in the system, to their financial impact. This is a major step  - 10 - forward performed by the Burundese officials and technicians as action' now under consideration and, for some of them, under implementation, concern student boarding, preparation of teacher training and teacher employment policies adopted to the socio-political situation, countrywide selection of JSS students, and preparation of a new policy for young and adult training in order to use the SC more efficiently. J. Prolect Relationship 24. Dialogue between the Government and IDA was constant, frank and direct, and has continued as part of preparation for the fourth project. Relationships with other organ .,zations, whether bilateral or multilateral, were excellent, with one exception. Co-financing and/or parallel financing was discussed and established with the AfDB, UNICEF, Belgium and UNDP/UNESCO. Contacts with France, which play an important role in the training of primary school teachers and in reforms, were possible only among technicians, enabling the additional activities to be carried out. Lastly, during the project, in close conjunction with the UNDP, the Government organized a round table on education issues, which enabled all countries and organizations represented to plan for future activities. The conclusions of the round table were taken into account in the preparation and identification of the Fourth Education Project. K. Consulting Services 25. The private firms to which the construction contracts were awarded performed satisfactorily, and because of the quality of the works and the fact that they kept to schedule, it was decided to use the private sector for the rehabilitation of large secondary schools under the fourth project. However, as the furnishing contractors did not perform satisfactorily, precautions must be taken in that regard for any future undertakings. The technical assistants recruited through the PIU for project management, construction, the RPP and the consultants for the BER, the Education Planning Unit (EPU) and the social affairs department ail performed satisfactorily. L. Proiect Documentation and Data 26. The project generated on the customary documentation: Credit Agreement, Appraisal Report and President's Report. These documents had been preceded by a November 1981 memorandum on the education sector, which gave a general analysis of the sector and made some global recommendations for its development. Project documentation also included a comprehensive working paper on project architectural aspects, including costs per square meter for each school. Unfortunately, no working paper was prepared for the educational aspects of the project, which would have been extremely useful, primarily for the MLC and TTI programs. This omission has been overcome for the fourth project.  - 11 - 27. The information needed for the preparation of the PCR was available either from the PIU or the files of the various offices of the Ministry and IDA. The supervision reports were clear and complete and included numerous annexes and technical documents. In addition, the semi- annual PIU reports, which were submitted regularly and were well documented, provided excellent reference points throughout the implementation of the project.  - 12 - PART II REPUBLIC OF BURUNDI MINISTRY OF PRIMARY AND SECONDARY EDUCATION PROJECT IMPLEMENTATION UNIT COMPLETION REPORT THIRD EDUCATION PROJECT Credit No. 1358-BU 1. Introduction 2. Objectives 3. Description 4. Amendments 5. Project (a) Project management implementation: (b) Building design and construction (c) Equipment (d) Furnishings (e) Procurement (f) Technical assistance (g) Logistical support to the Ministry of National Education and the Ministry of Social Affairs (h) Internships, study trips, seminars 6. Financial tables: - Breakdown of project amounts at appraisal - Breakdown of project amounts: April 1986 .. reallocation - Exchange rate fluctuations - Actual disbursements - Project financing by expenditure category - Comparison of actual project costs - Utilization of surplus - Expenditures under the Third Education Project - Comparison of areas, cost and enrollments 7 Results 8. Conclusion  - 13 - 1. Introduction The Project Implementation unit (PIU) was established by Decree No. 100/258 of November 5, 1977 to increase student capacity at the schools and upgrade the quality of teaching. Although the Unit experienced several problems at the start of the First Education Project, its implementation of the two subsequent projects was extremely successful. This evaluation report seeks to determine to what extent the Third Education Project has helpcd the Government to raise enrollments in primary and secondary educatlion, and to continue the educational reforms begun in 1973. The report will also examine the manner in which the project helped improve education and training opportunities for young illiterates. 2. Obiectives The objective of the Credit Agreement for the Third Education Project, for an amount of SDR 17,712,000 signed June 9, 1983 with the International Development Association, was to: - increase access to primary education through the establishment of new multipurpose learning centers; - develop and upgrade teacher training so that the reforms could be implemented more efficiently; - develop first-cycle secondary education with a view to better distribution of schools throughout the national territory and phase out boarding; - develop nonformal education and training in the rural areas with a view to reducing illiteracy and provide educational follow-up; improve the production and distribution of textbooks; upgrade personnel; provide technical and logistical assistance to the support services of the Ministry of National Education. 3. Description The project called for: (a) The construction, furnishing and equipping of: - 24 multipurpose learning centers;  - 14 - - two primary teacher training institutes, including two junior secondary schools; - four separate junior secondary schools; - 20 social centers. (b) Purchase of paper, veh4cles and equipment for the Academic Print Shop (Regie des Productions P;dagogiques - RPP). (c) Technical assistance to the PIU, BER, RPP, Ministry of Social Affairs and the Planning for carrying out the countrywide school mapping exercise. 4. Amendments The first amendment was made on October 18, 1985 following a letter from the Minister of Education requesting the establishment of additional places for boarders in the junior secondary schools and TTIs under the Third Project. The second amendment involved the allocation of the balance at closing from the 1986 budget exercise, which coincided with the completion of the initial program. The surplus of FBu 314,098,718 made possible by the PIU's efficiency was used for the construction of 5 additional MLCs, the renovation and expansion of some 10 secondary schools, expansion of the RPP and the BER, financing of technical assistance for the RPP and the PIU, support to the Ministry of Social Affairs for the printing of literacy brochures, financing of a consultant mission, payment of fellowships and study trips for the Kamenge ETS, Gitega ETP, the BER and the Ministry of Social Affairs. 5. Proiect implementation GenerAl Credit Agreement No. 1358-BU for the Third IDA assisted Project was signed on June 9, 1983. Credit effectiveness was to take place 90 days later, i.e., September 9, 1983. Effectiveness procedures and conditions, in particular the release of the first tranche of FBu 31,500,000 representing the financial participation of the Government in the project, were delayed and the Credit Agreement did not become effective until November 16, 1983. The project, in its original form, was completed ahead of schedule on December 31, 1986.  - 15 - (a) Proiect manaaement The Project Implementation Unit established in 1976 was responsible, as under the first two projects, for coordination between the various ministries and agencies, liaison with IDA and the execution on force account of most project construction works. The PIU now employs 14 higher- level staff, of which 8 are Burundi, and 65 skilled workers. Temporary personnel consist of 1,600 workers of various categories. During the project, Burundi nationals took on increasing responsibilities. The heads of finance, personnel, and vehicle fleet are Burundi, as are all deputy heads of the four main services: technical, procurement. accounting and human resources. As part of its bilateral assistance program, Belgium continues to finance the services of an architect. (b) Building design and construction Design Four junior secondary schools and two TTIs - ,General design The original plan was to build the junior secondary schools (JSSs) and TTIs to house a bare minimam of boarders (80 boarders/320 students)--160 boarders/320 students or 160 boarders/480 students, as appropriate, with the area of the buildings in question to be studied as a result: kitchen, cafeteria, bathrooms, dorms (which could later be converted into classrooms). However, almost all the sites selected for the new schools were very far from the urban centers, which made boarding necessary. As soon as the schools opened, capacity was oversaturated, which led to inappropriate use of the premises and facilities: - classrooms or science labs used as dorms; - dorms for 64 students occupied by more than 100 (two students per bed or beds crammed together), with the ensuing problems of hygiene and security; - undersized septic and sanitary tanks; - insufficient water reserves or distribution networks; - undersized cafeteria and kitchen (plus pantry); - etc. These "arrangements" became unbearable and pressures grew with each new academic year. Expansions were considered, but appropriate solutionia were not always possible (impossibility of expanding the kitchen and pantries, water system remaining inadequate, etc.).  - 16 - The design of the various school buildings was simple and occasionally even overly so for the schools studied and built In the first phase (Cibitoke TTI, Bukirasazi JSS, Bubanza JSS), and was based on a single module and standardized construction materials, regardless of the future use of the buildings (window and door frames, etc.). The result was a certain degree of monotony and, beyond the signs posted above the doors, it is virtually impossible for an uninformed visitor to know what the building is used for. While still respecting considerations of simplicity, another approach was adopted for the phase-two schools (Mukenke TTI, Nyankanda JSS and Murore JSS). In accordance with the base construction options adopted, certain elements were not deemed necessary (e.g., ceilings and mosquito netting). Actual v'e of the buildings showed the error of such judgements (impossibility of holding classes when it is raining and students bothered by insects during clas3 or in the dorms), and additional works had to be done while the premises were occupied. For eAample, modifications had to be made to the electrical installatiors. The master plans could have been covered by more comprehensive research, at least regarding the first phase (Cibitoke TTI and Bukirasazi JSS). The same is true for walkways between the buildings, which were virtually nonexistent. The base options overlooked the "lodging" aspect to the benefit of the school proper, the main objective of the investment. Guidelines thus limited expenditures in that respect to the strict minimum: - minimal surface area; - very low-cost masonry (sometimes mud brick); - rudimentary door and window frames; - no ceilings; no mosquito netting; outdoor pit latrines; no furniture for the bedrooms. These arrangemebts were harshly criticized by the teaching staff and resulted in additional and time-consuming works.  17 - 24 MLCs and 20 Social Centers General Desian The base options for the Jesign of the buildings were as followss - Average price/m2 of + FBu 16,000 - Window openings with grillwork, no glass Note: Because broken window glass was not replaced in the other schools. - No ceiling - Two 3-bedroom dwellings for the MLCs and one dwelling for the SCs, of rudimentary construction initially (improved worksite shack). These buildings were subsequently built with more care and of durable materials. Considerations These low-cost construction options resulted in a sizeable balance, at least for the 24 MLCs. However, the elimination of glass for the window frames of the social center classrooms was highly criticized by the users, and me-;ures had to be taken to remedy the situation. The houses built for the MLC teachers and SC instructors were for the most part in practice requisitioned by the commune goveriunents for their personnel (administrator, policeman, agronomist, etc.), with the result that the original intent was not fulfilled. It appears that the teachers refuse to pay a modest rent for maintenance. CONSTRUCTION Four JSSs and TTIs - These six schools were built on force account by the PIU ahead of schedule (with the exception of the Mukenke TTI). The budgetary savings were used for the construction of additional buildings, which could not have been done if the works had been awarded to private firms. - The trial manufacture of industrial roofing tiles was carried out for the Mukenke TTI with a view to covering all school buildings (