Madagascar - Towards a school-based strategy for improving primary and secondary education

Report No. 1 3450-MAG Madagascar Towards a School-Based Strategy for Improving Primary and Secondary Education April 7, 1995 Population and Human Resources Division Central Africa and Indian Ocean Department ica Re,gion Document of the.World Bank Document of the World Bank  CURRENCY EQUIVALENCY Currency Unit: Malagasy Franc (FMG) Average Exchange Rate 1992: US$ 1.00 = 1864.0 FMG 1993: US$ 1.00 = 1913.8 FMG 1994: US$ 1.00 = 3871.1 FMG GLOSSARY OF ABBREVIATIONS AFA Auto-Formation Assistee (Assisted Self-Training) BEPC Brevet d'Etudes du Premier Cycle (Lower Secondary Cycle Diploma) CAP Certificat d'Aptitude Pedagogique (Teaching Capacity Diploma) CEPE Certificat d'Etudes Primaires Elementaires (Primary Level Diploma) CEG College d'Enseignement General (Lower Secondary School) CRESED Credit de Renforcement du Secteur de l'Education (Education Sector Reinforcemnent Credit) GER Gross Enrollment Ratio IDA International Development Association UPE Universal Primary Education MEN Ministere de I'Education Nationale (Ministry of National Education) MIP Ministere de l'instruction Publique (Ministry of Primary and Secondary Education) NER Net Enrollment Ratio PNAE Programme National d'Amelioration de I'Education (National Education Impro ement Program) UERP Unite d'Etudes et de Recherche Pedagogiques (Curriculum and Research Deve]opment Unit) UNDP United Nations Development Program ii  TABLE OF CONTENTS PREFACE ...................................................... v EXECUTIVE SUMMARY ..................... vi CHAPTER 1: INTRODUCTION ...... .....................1............. The Economic and Political Context ..... .................. . 2 The Education System's Response ..... ................... . 3 Scope of the Study ...... ...................... . 4 Organization of the Report ..... .................... . 5 CHAPTER 2: STRUCTURE, ACCESS AND INTERNAL EFFICIENCY IN PRIMARY AND SECONDARY EDUCATION ..... ................. . 7 Structure and Organization ..... .................... . 7 The Current Quantitative Performance of Primary and Secondary Education ... . 7 Access .............................................. 8 Internal Efficiency ........................................ 9 Madagascar Compared with the Rest of Sub-Saharan Africa .............. 9 Strategic Considerations .................................... 9 CHAPTER 3: THE QUALITY OF PRIMARY EDUCATION ........... ...1........ 11 Student Outcomes ..........................1.............. 11 School Process Elements .................................... 14 CHAPTER 4: THE QUALITY OF SECONDARY EDUCATION ................... 19 Student Outcomes ................ ................ 19 School Process Elements .................................... 19 CHAPTER 5: THE PUBLIC FINANCING OF EDUCATION ..................... 23 Public Funding of the Education Sector ........................... 23 Macro-Economic Perspective on Trends in Public Revenues ........... ... 24 Allocations to Primary and Secondary Education from the Education Budget ... 24 Expenditure Patterns within Primary and Secondary Education ....... .... 25 Other Sources of Financing for the Education Sector ........ ........ . 25 CHAPTER 6: ELEMENTS OF A STRATEGY TO IMPROVE THE QUALITY OF EDUCATION .......................................... 27 Macro-Economic Context ...................... ........... . 27 Constraints ............................ ............... 27 Implementation Issues ........................ ........... . 28 School-Related Elements of the Strategy .......................... 29 A Positive Climate for Effective Implementation: Clear Objectives and Local Control ... . 31 The Cost of Reform ....................................... 31 Financing Quality Improvements ............................... 33 Financing Quantity Increases ................................. 35 Issues to Resolve to Define a School-Based Strategy ................... 36 iii  TEXT TABLES TABLE 1 Enrollment Development Since 1980 3 TABLE 2 Efficiency Indicators for Primary and Secondary Schools Compared with Other African Countries 10 TABLE 3 Test Results of Pupils at End Year of Primary School, May 1991 14 TABLE 4 Simulated Additional Resource Requirements for Primary Education by 2000 33 CHARTS CHART 1 Hypothetical Network of Priority Factors that Determine Primary School Effectiveness (Before Research) 12 CHART 2 Conceptual Framework for a Strategy for the Reform of Primary Education (After Research) 13 ANNEXES .37 ANNEX 1 Educational System Structure (1993) ........................ 38 ANNEX 2 Organization of the Ministry of Education ..................... 39 ANNEX 3 Basic Education Statistics (Tables 3.1 - 3.7) .................... 40 ANNEX 4 A.Note on, Research Methodologies ......................... 46 B.Key Factors That Determine School Effectiveness .... .......... 48 ANNEX 5 Financial Statistics (Tables 5.1 - Tables 5.6) ................... 50 ANNEX 6 Goals and Objectives (Primary Level) ....................... 56 ANNEX 7 Financial Simulations (Tables 7.1 - 7.4) ...................... 57 MAP iv  PREFACE This report analyzes the results of research on primary and secondary education in Madagascar in order to identify the school-level factors which have the most influence on student learning and on whether students stay in school. Also included are crude estimates of the resources that will be needed and which might be available to invest in providing or establishing these factors by the year 2000. I)rafts of the report were discussed with the Govermment of Madagascar in July 1994 and January 1995. The study is based on a conceptual framework of school quality developed by a Technical Group in the Ministry of National Education and coordinated by the Ministry's Unit6 d'Etude de Recherche Pedagogiques. The Technical Group conducted qualitative field research in 36 schools using this framework and wrote case studies on each school. Their observations and findings add a richness of detail to this sector study that strengthen the report's findings. This report was written by Ward Heneveld (Human Resource Division, Africa Technical Department -- AFTHR), and Daniele Jaekel (AFTHR) and Gabriel Carceles (Consultant) assisted with the preparation and analysis of the statistical and financial information in the report. Report processing was finalized with assistance from Mesdames Donna McGreevy and Francoise Ryckebusch. Mr. Ulrich Thumm and Mr. David Berk are the Department Lead Economist and Managing Division Chief, respectively, for the operation. Mr. Makha Ndao (AF4PH) and Mr. Adrian Verspoor (ESP) are the peer reviewers. v   EXECUTIVE SUMMARY i. Primary and secondary education in Madagascar has suffered greatly during the economic decline of the eighties which continued through the political transition of the early nineties. The new government faces many difficult choices with significant political risks if education's contribution to human resource development is to improve. Conditions that need to be met in order to ensure successful sectoral reform include sustained appropriate macro-economic policies, investments in effective government that provides security, good management and popular participation. Within the context of the country's current and projected economic conditions and of current efforts in the reform of education, this sector study examines the quality of primary and secondary education and the internal efficiency of the system, and identifies the elements of a strategy that would focus on the schools to improve the quality of education. ii. The study defines educational quality as changes in the environnent in which children learn and learning gains that result from these changes. It draws on Madagascar's education statistics and research on primary and secondary education to analyze how well schools are doing. Central to the study are case studies of 12 primary, 12 junior secondary (CEG), and 12 senior secondary (lycee) schools that a Technical Group from the Ministry of Education (MEN) conducted in 1993/94. The results of this research at the primary level are compared with two other studies, one on determinants of academic achievement in French, Malagasy, and Mathematics and one on the reasons that children drop out of primary school. System Performance: Deteriorating Access. Poor Internal Efficiency and Learning Outcomes iii. Madagascar's education system comprises a five-year primary cycle and two secondary cycles, a first of four years and a second of three years. A small number of students attend technical schools at both secondary levels, but these schools are not considered in this study. The quantitative indicators on the performance of the system at the primary and secondary levels of education show that the system is deteriorating badly. Enrollments have declined at all levels between 1980 and the early nineties and most rapidly at secondary level. With many schools closing, the gross enrollment ratio dropped from 142% at the primary level and 30% at the secondary level in 1980 to 95% and 20% in 1992, respectively. The net enrollment ratio at primary level was about 64% in 1990. Recently as well more parents are choosing private schools over public ones for their children. By 1992, private schools enrolled 22% of all students in primary school, 40% in junior secondary, and 48% in senior secondary, in a period where quality of public education is criticized. These indicators suggest that the poor are abandoning the public schools altogether and that many of those who can afford to are choosing private schools. Both tendencies are probably in good part due to the poor quality of the education provided. iv. The low internal efficiency of the education system adds evidence about the poor quality of education provided. At the primary level, more than one third of all students repeat each year; fewer than 30% of those who start school reach the fifth (final) grade; and it takes close to three times the number of pupil-years of instruction that it should to produce a graduate. These indicators of internal efficiency are among the worst in Sub-Saharan Africa. At the secondary level, repetition and dropout rates are lower, approximating the average for the rest of Sub-Saharan Africa. Declining vi  vii enrollments, poor internal efficiency, and more and more students opting for private schools are a,l signs that a significant dimension of the crisis in Malagasy education is its poor quality. v. Available information on primary school learning outcomes suggests that the quality of primary education is weak. The fact that only about 30% of the entering students even reach grade 5 adds to the concem about quality of outcomes, as one may question what those who leave earlier have learned. Factors Determining The Quality of Primary and Secondary Education vi. The study analyzed the factors of the school environment that determine these learning outcomes. The factors studied include: community and government contributions, the school director's management, what is expected of students, the capabilities and attitudes of teachers, school infrastructures and equipment, learning materials, and the teaching/learning process. The analysis of the research on primary education suggests that the most important factors influencing primary school effectiveness in Madagascar are: the management by the school's director; teaching materials, especially teachers' guides and textbooks; and community support, especially for facilities and equipment provision and maintenance. The analysis of the 24 secondary school case studies suggests that three main factors determine a secondary school's success: the management by the director of each school; the facilities of the school; and the competencies and attitudes of the teachers. But learning materials and parental support also appear important. Because these provide the country's only research results on secondary education, additional research on secondary education is needed before fully adequate policy conclusions about priorities can be drawn. vii. Although the school-level analyses suggest that there is not much of a financial contribu:ion by government to the life of the school beyond teachers' salaries, public expenditures on primary and secondary education constitute a significant part of the national budget, especially in the form of tea3chers' salaries and management of the education system. The report's analysis of current public financing of education looks at three policy-related issues: (a) the shares of GDP and of total government budgets that go to education, (b) the shares of education budgets that are allocated to primary and secondary education, and (c) expenditure patterns within these sub-sectors of education. Based on this analysis the report concludes that the education sector receives a meager share of GDP (2.3%) -- largely because the overall government budget is a relatively small percentage of GDP. The allocations to primary and secondary education should be increased somewhat at the expense of higher education. Expenditure patterns within education slight direct support to instruction in favor of administrative expenditures, especially at the primary level. Implementation Issues viii. If reform of the education system is to succeed, proposed changes must recognize, accept and work within the economic, environmental, cultural, and socio-political realities that characterize Madagascar. The report contends that these realities combine to create a crisis of public "will" about the possibility of improving primary and secondary education. The issues discussed are "soft" issues, ones that are not easily measured but that are very apparent in Madagascar. They include extensive poverty and an uneven income distribution, a loose political consensus in government, and a demoralized and indifferent civil service. If the quality of education is to be improved, leadership at all levels in the  viii education system needs to acknowledge the system's severe deterioration, but also to affirm the possibility of slow but steady improvement. It can then work on: building commitment for educational improvement; sustaining the effort to change civil servants' attitudes; supporting their efforts over time; and demanding and rewarding high self-expectations and discipline. Elements of a Strategv to Improve Quality ix. The study's findings on school quality and the implementation issues identified have helped clarify the elements of a strategy for improving the quality of primary and secondary education within the larger economic development framework. These elements include: focussing reform on improving learning outcomes; concentrating implementation on strengthening school leadership, providing more learning materials, and fostering community support; creating and sustaining a positive climate for the reform of education; shifting recurrent budget allocations towards instructional purposes while investing, through communities, in more and better physical facilities and equipment; supporting development of private education through, inter alia, regulations and supervision, performance contracts, incentives and/or subsidies; finding ways to get teachers to serve in rural areas; and increasing subsidies at lower income groups while increasing resources for the sector through more efficient and greater mobilization of household expenditure. x. Using high and low enrollment projections and World Bank projections of economic growth, high and low scenarios of possible funding needs and resource availability for these elements of a strategy for primary education were simulated. According to these simulations, between 9 billion FMG and 35 billion FMG at 1993 prices would be needed in the year 2000 to support the recurrent cost of additional primary school students and to increase present recurrent expenditures on primary level instruction inputs. This would be in addition both to the 103 billion FMG (including salaries) for primary and secondary education in the 1993 Government recurrent budget and to the private payments that were fully supporting at least 25 % of the students in that year. xi. The resources that might be available from Government to support these activities in 2000 have also been simulated. According to current projections of economic growth and under alternative assumptions about allocations of the budget to and within the education sector, the 1993 budget for primary and secondary education could increase by between 40 billion FMG and 154 billion FMG by the year 2000. This magnitude of increase is predicated on the share of the public budget going to education increasing slightly, on the allocation to primary and secondary education within the education budget increasing, and on the transfer of some of the funds now spent on administration to instruction. Even if primary education only received half of this increase, the additional funds would provide much of the additional government funding that is needed to support recurrent expenditures that will imnprove educational quality. xii. On the other hand, the simulations suggest that construction of new classrooms for additional students and rehabilitation of existing schools could cost between 245 billion FMG and 363 billicn FMG between 1993 and 2000. Government funding at this level for infrastructure will not be possible, even if the public in Madagascar continues to contribute at current or higher levels, and donor assistance will be required. This assistance should be used to leverage continued community participation and as an incentive to government to increase its own recurrent funding of learning materials and direct instructional support to improve educational quality.  ix xiii. Besides financing, much of which it seems reasonable to expect will be available, the government will also need to instill a sense of purpose and enthusiasm among educators and parents if the implementation of reform is to be successful. A clear set of objectives for primary education with a simple strategy that concentrates on the factors identified in this study should be communicated widely and backed up by action. A draft statement of the objectives for primary education is presented in this report. xiv. Given the current state of Madagascar's primary and secondary education system, changes of the magnitude and in the directions suggested here are imperative if the country is to build the human resource base for accelerated and sustainable economic growth and poverty reduction. The next step is to translate the results of this study into a coherent, integrated strategy of policies and activities that will improve primary education. The issues needing decisions by the Government of Madagascar in order to finalize this strategy include (a) clarifying assumptions about the conditions outside education that influence what can be done; (b) explicitly defining expected learning outcomes and the basic program design for improving primary education and (c) establishing financing policies that will continue the significant private contributions to education, increase the amount of public resources spent on instruction, and provide more equitable access to quality education for all children.  1 CHAPTER 1: INTRODUCTION 1. Over the last twenty years Madagascar's economy has declined dramatically. The country has gone through a period of piecemeal adjustment followed by political unrest and -- most recently -- a peaceful transition that has produced a democratically-elected new government. The new government faces many difficult choices with significant political risks if the economy is to recover and grow. Its most significant challenge beyond providing an enabling macro-economic environment for development in agriculture, industry, and tourism is to establish policies and implement changes that will develop the human resources necessary to sustain growth and alleviate poverty. In the short-to-medium term chere is an urgent need to stabilize and improve general education. This study examines the quality of primary and secondary education in the context of the development challenges facing Madagascar, and it identifies the kinds of investments in the sector that will improve the level of general education. 2. The study combines analyses of quantifiable issues -- economic decline, population growth, the size and internal efficiency of the education system, and education financing -- with more qualitative concerns -- school climates, the teaching/learning process, and the socio-political and cultural context -- in order to capture the multi-faceted reality that will shape the future of primary and secondary education in Madagascar. The focus is on schools. The study examines the factors that determine school effectiveness, and the conditions in schools that help children learn. 3. Even though Madagascar's education statistics fluctuate a lot, suggesting some unreliability in them, there are convincing indications that the quality of public primary education is currently very poor, and that the quality of secondary education is declining (Box 1 and Tables 3.1-3.7 in Annex 3). Repeater and dropout rates are high, exceedingly so at the primary level where fewer than one in three children who enter the five-year cycle complete it. The percentage of children passing terminal examinations in each cycle is below 50%. Furthermore, total enrollments in primary schools have fallen over the last 10 years, and those in secondary schools have fallen over at least the last five years, despite a population growth rate of about 3%. 4. It also appears that parents and their children, the consumers of education, are deserting the public education system. During the last five years enrollments in private primary and secondary schools have grown, while public enrollments declined significantly. In business terms, the public sector is losing market share as parents who can afford an alternative decide that government schools do not serve their children's needs. It may also be hypothesized that many parents who cannot afford private education are becoming even poorer and, at the same time, are pressured for additional contributions to their children's impoverished schools. These families may feel the same way as parents who are opting for private schools, that the investment of time and resources is not worth it. Therefore their children may stay home or drop out of school so that they can contribute to the meager family budget. If this assessment is accurate, improvements in the quality of education provided in public schools could regenerate demand, helping to reverse the declining enrollments and to improve participation.  2 The Economic and Political Context 5. Between 1971 and 1991 per capita income fell by 40% and the number of people living in poverty grew dramatically '. The economic decline has contributed to a dramatic decline in the quality and availability of educaticn in the country, thus jeopardizing the broad base for economic development that an educated population provides. Economic decline has also made regional antagonisms more pronounced as the deteriorating communication infrastructure made contact more difficult and less productive. The new government has responded to regional differences by recognizing the need for a less-centralized system of administration, and the means for establishing greater authority and responsibility outside the center are under discussion 2. 6. The Government also has the opportunity to follow up the accomplishment of peaceful political change with improved living conditions for all Malagasies, especially the poor. This can only be done if economic growth can be speeded up by 2000 to near 6% per annum, the estimated rate of growth required to improve people's lives appreciably given the rapidly growing population, and if the government has the will to set policies that emphasize human resource development for all of the population. Improving education levels is crucial to reduce poverty, but also to maintain Madagascar's imperative advantage in well-trained labor, and thus directly support the country's growth objective. Central to achieving this growth rate and human resource development objectives are sound economic strategies that will increase capital accumulation, savings and investment rates and, most significantly for this study, a greater investment in the quantity and quality of basic health, population, education, and training services. Later, as economic conditions improve, attention can turn from basic services to the more specialized needs that accelerated economic growth will create. 7. The new government faces significant challenges in implementing these strategies as macro- economic conditions continue to have a negative impact on the social sectors, including education. In a general climate of poor economic performance and deterioration of social indicators, the education system has also fared poorly (Table 1). In addition, demographic projections suggest that each year between now and the year 2000 at least 400,000 people will be added to the population 3. The impact of a weak economy and rapid population growth on primary and secondary education will cont,nue to pressure the education system. Sound policies and the successful implementation of new strategies are needed urgently to respond to these pressures. I Madagascar: Note de Sfrategie Econornique (May 1993). Washington, D.C.: The World Bank. 2 A far-reaching law on decentralization was approved in early 1994 and the implementation is currently being wor;:ed out. 3 Cf., ibid., pp. 5-7.  3 Table 1 MADAGASCAR: Enrollment Development Since 1980 (In Thousands) Junior Senior G.E.R. * G.E.R. G.E.R. Total Primary Secrv Seca Primarv Jr. Sec. Sr. Sec. 1980 2161 1724 414 23 142% 30% 2.9%; 1985 2085 1600 446 39 117% 31% 3.4% 1990 1947 1571 340 36 105% 22% --- 1991 1816 1497 319 --- 98% 21% --- 1992 1810 1490 320 --- 95% 20% % Annual Growth 1980-85 -1.5% +1.5% +11.1% % Annual Growth 1985-90 -0.4% -5.3% -1.3% % Annual Growth 1990-92 -2.7% -3.0% *G.E.R.: Gross Enrollment Ratio Source: Service des Statistiques, DPOE/MEN The Education System's Response 8. Primary education in Madagascar is in deep crisis, and secondary education is losing students, particularly in public schools. To respond to this situation, the newly-elected government has continued elements of the educational reform entitled Programme National d'Amelioration de I'Education (PNAE) that was initiated in the late eighties with the participation of the World Bank. Led by an Intemational Development Association (IDA) credit of $39 million (CRESED) and with participation by OPEC, UNDP and other donors, including the French government, changes that will allow for improvements in the quality of education are underway. First, the Ministry of Education (MEN) has reinforced essential central functions and has identified some of the needed refomis in the management of the primary and secondary education system, especially with respect to the assignment of more authority and responsibility for implementation to the regions, and has just begun experimenting with some of the proposed changes. Second, numerous studies to improve educational planning are nearing completion. Textbooks for the first year of primary school have been supplied to schools throughout the country, and books for higher  4 grades and other learning aids are being produced for distribution. Finally, training of personnel and planning for improved teaching through better teacher preparation and supervision are being carried out. It is too soon to see the results of these efforts in the national educational statistics presented in this report because their impact at the school level has not been felt yet. Scope of the Study 9. This study focuses on the quality of education defined as "an improvement in the environment in which the student work(s) (such that) this improved environment (produces) detectable gains in the knowledge, skills, and values acquired by students" 4. The study argues that neither the activities already underway nor the availability of additional resources will be sufficient to secure significant improvements in the quality of primary and junior and senior secondary education. A more comprehensive and integrated strategy is required, one in which the central government explicitly fosters an enabling environment throughout the system and, in each school, a climate which explicitly values children's learning and the teaching/learning processes that produce learning. 10. Other studies of primary education in Madagascar confirm the need for an approach focused on the school. A 1992 study of 181 primary schools and 2,874 students entitled Evaluation du Systyme Educatif Malgache recently concluded that "the disparity of resources among schools does not eWxplain, systematically, schools' difficulties. . . . The effectiveness (in learning outcomes) of a school . .. lies in the quality of school life." I Similarly, a school mapping pilot study of one prefecture ha. noted provisionally that among its twelve communes "the best examination results, in public schools, are found in the communes with the worst or mediocre staffing and material conditions for schooling." 6 Thus, something else is making a significant contribution to their effectiveness. 11. In order to complement and enrich the more quantitative research on the factors in Malagasy schools that determine educational quality, in 1993 a Technical Group in the MEN undertook a qualitative observational study in 36 schools (6 in each province, 12 at each of the primary, junior secondary and senior secondary cycles of the system). The study identifies school-level conditions, and their inter- relations, that contribute to improving academic results and reducing repetition rates. The concLusions on primary education from this field research have been compared with, and are largely corroborared by, the findings of the 1992 Evaluation study and by the preliminary findings of the planning studies being conducted on education in the current IDA-funded project. 4 Ross, Kenneth N. and Lars Mahick (1990). Planning the Quality of Education: The Collection and Use of D)ata for Infonned Decision-making. Paris: UNESCO/Pergamoni Press, p. 6. 5 Robin, Daniel and others (1992). Evaluation du Systme Educatif Malgache: Comnplments. Sevres, France: Centre International d'Etudes Pdagogiques, p. 46. 6 Ministere de l'Edt.cation Nationale, Etude Pilote de Carte Scolaire: Fivondronampokontany de Soavinandriana, Antananarivo. Provisional Version, September. 1993, p. 39.  5 Organization of the Report 12. The report presents an analysis of access and internal efficiency in both primary and secondary education (Chapter 2). The results of this research in primary education form the central part of this report (Chapter 3). The analysis of the quality of the school characteristics that are critical to improving educational quality secondary education is briefer and more tentative because there is only one study to draw on, namely the case studies done in 12 junior secondary schools and 12 lyckes (Chapter 4). Common themes can be seen in the results of these separate analyses, particularly regarding the importance of the school director and the community and parents; but differences also exist. Needs at the primary level are more focussed on learning materials than on teachers. At the secondary level, lack of effective teacher preparation seems to be more important, and individual parents' involvement appears to supersede the more encompassing community role at the primary level. Overall, the study concludes that the strategies for improving quality at the two levels need to be differentiated. 13. The report also examines the public financing available for education (Chapter 5). The final section of the report (Chapter 6) outlines the elements of a strategy for improving the quality of primary education in Madagascar and possible resource requirements. It takes into account the system's current performance and trends and considers the resources that can be expected for education if economic growth occurs and human resource development is given a higher priority. Before similar strategic conclusions can be reached on secondary education, additional research is required. Finally, the chapter suummarizes the policy and planning issues that the report suggests the Government of Madagascar needs to finalize decisions on in order to have a coherent strategy for the improvement of educational quality in primary schools.   7 CHAPTER 2: STRUCTURE, ACCESS AND INTERNAL EFFICIENCY IN PRIMARY AND SECONDARY EDUCATION 14. Overall, primary and secondary education in Madagascar has deteriorated in the last fifteen years. At the primary level, repetition and drop-out rates are among the highest in Sub-Saharan Africa. Secondary education's efficiency is not as bad, but there total enrollments have declined even more. At both levels, the proportion of students attending private schools is increasing. Structure and Organization 15. Primary school in Madagascar includes five grades (called Classe 1 1 - Classe 7), and those who complete this level and pass an examination receive the Certificat d'Etudes Primaires Elgmentaires (CEPE). General secondary schooling consists of two cycles, the College d 'Enseignement GMngral (CEG) lasting four years and granting the Brevet d 'Etudes du Premier Cycle (BEPC), and the lycge lasting three and granting a title of bachelier. Graduates of the lycee can go on to higher education. As already noted, private education is growing. While government has provided some limited assistance to private schools, quality control does not seem to be an important concern yet. A technical and professional education stream is organized in parallel with the general system starting after primary school, and other ministries and organizations provide technical and professional training. Pre-school education is under the charge of the Ministry of Population. Annex 1 presents a diagram of the education system. 16. Until the current government was formed in August 1993, the Ministere de lI 'nstruction Publique (MIP) served primary and secondary education, and a separate Ministry of Universities served higher education and scientific research. Since then, primary, secondary, and tertiary education and culture and sport have been integrated in the Ministre de I 'Education Nationale (MEN) with primary and secondary education comprising one of three Departments in the Ministry. In the regions, offices of the Ministry exist at the provincial (Directions Provinciales) and sub-provincial (Circonscriptions Scolaires) levels. An organigram of the ministry's structure showing particularly the units within the Secretariat for Primary and Secondary Education is in Annex 2. Currently the management of the system is highly centralized, but the new government is committed to decentralizing public administration. In addition, as part of the education reform process, experiments are being made with school mapping, decentralized management activities in one province, and increased consultation with personnel in the field. These efforts should lead to the general devolution of responsibilities and control of resources to the lower-level governments in implementation of the March 1994 decentralization 7. The Current Ouantitative Performance of Primarv and Secondary Education Access 17. In the seventies, Madagascar almost achieved universal primary education (UPE), but in the eighties total enrollments stagnated and then began to contract at all levels. As can be seen in Table 1, 7 See also Republic of Madagascar: Decentralization and Local Government Reform (Green Cover Report No. 13011- MAG), November 30, 1994.  8 there were fewer students in primary schools in 1992 than in 1980 or in 1990. But this is only part of the picture. Between 1988 and 1992 the percentage of new students enrolling in private schools increased from 17% to 25%, while the total number of new entrants declined only slightly (Table 3.1, Annex 3). There were however declines in both the number of teachers (by 8.8% between 1987/88 and 1992/93) and the number of schools (by 4.2% over the same period). School closings have been particularly striking and disturbing, with a cumulative total between 1000 and 1750 closings reported each year between 1985/86 and 1989/90 (Table 3.7, Annex 3). 18. With the population growing by about 3% annually, Gross Enrollment Ratio (GER) (including all under-age and, due to very high repetition rates, over-age pupils) had declined from 142% in 1980 to only 95% in 1992. In two regions where about half of the schools do not offer all five grades, the GER was only 64 % (Toliary) and 82% (Fiananrantsoa) (Tables 3.4 and 3.5). The national GER of 95 % compares with a modes, national primary Net Enrollment Ratio (NER) of about 64% in 1990. 19. At the secondary level, the total number of students has been declining. Of the small percentage of students who made it to the last year of primary school in 1991, about two-thirds continued to the first year of junior secondary school (CEG) in 1992. Of those who completed the CEG cycle in that year, 44% found a place in a senior secondary school. However, these relatively good transition rates are occurring as overall enrollments decline, particularly in public schools. Between 1987/88 and 1992/93, total junior secondary enrollments declined, even though private school enrollments increased each year. By 1992/93, private junior secondary schools accounted for 40% of total enrollments. Senior secondary school enrollments have also declined by 13% over the same period, with private school enrollments declining more slowly than in public schools to account for 48% of total enrollment. The total number of public school teachers for these two cycles increased before 1990 but has remained fairly stable since, while the number of teachers in private schools continues to increase each year. However, the number of secondary schools have remained about the same, senior secondary schools showing a slight increase. These trends suggest that, at the secondary level as at the primary level, the demand for education is declining, particularly for education in public schools. Internal Efficiency 20. The internal efficiency of an education system is its ability to minimize student repetition and drop-out so as to permit completion of a cycle as quickly as possible. The internal efficiency of Madagascar's primary education system is poor and below the Sub-Saharan Africa (SSA) average, while the efficiency profile of secondary education resembles many other African countries (Table 2). About one-third of all primary school students were repeating their grade in 1990. In the critical first year of the cycle, the rate is even higher, about 40%. The number of years of instruction per student required to produce a graduate is 15.4 years, three times the number of years in the cycle. These inefficiencies also reflect high drop-out rates with only an estimated 30% of those who enter the first grade completing the five-year primary cycle. There are also significant differences in internal efficiency among the country's six regions (Table 3.5) with Toliary and Fiananrantsoa, the regions with the most rural schools not serving all five grades, lagging well behind the other four regions. In these two regions less than 20% of the entrants to Grade 1 in 1987 were reaching Grade 5 in 1991. 21. Compared to primary education, secondary school promotion rates are consistently higher. At the junior secondary level (Table 3.6, Annex 3) promotion rates have been stable over time at between  9 65% and 80%, according to grades. Repetition is relatively low, 15% in the first three grades, then around 30% in the last grade. Drop-outs are also relatively low, around 15% except in the first year. At the senior secondary level, the picture is more complicated. Promotion rates have been consistently between 60% and 70%, but public school rates are lower than private school rates. Repetition rates in public schools during the same period were about 20%, with private schools doing better. Very few students drop out of the senior secondary cycle after the first year, especially from private schools. Madagascar Compared with the Rest of Sub-Saharan Africa 22. The performance of Madagascar's primary and secondary education system compares poorl) with other countries in the region. It is one of only six among the forty-four countries in the region to register a decline in total primary level enrollments during the 1980's, and one of seven to show a decline at the secondary level.. The rest of the countries in the region, almost all with population growth rates as high as Madagascar's, managed to continue to expand enrollments and, in some cases, enrollment ratios. 23. A comparison of the internal efficiency of Madagascar's system with the rest of the region (Table 2) accentuates the extremely poor performance of primary education. It is among the most inefficient two to four countries in terms of the percentage of pupils who repeat, the percentage of pupils who reach the highest grade, and the number of pupil-years of instruction it takes to produce a graduate 8. At the secondary level the internal efficiency of Madagascar's education system is comparable to that of the rest of the region in terms of the number of primary graduates continuing to secondary school and the percentage of students who repeat each year. Strategic Considerations 24. The decline in enrollments in both primary and secondary schools is probably a result of both supply and demand factors. The explanation for limitations in supply differ between rural and urban areas. In predominantly rural areas such as Toliary, the supply of places undoubtedly limits enrollments. Many primary schools do not serve all five grades, and some areas are probably not served at all by a school; there have been many school closings, although it is not known to what extent demand has dried up or and to what extent buildings have deteriorated and teachers abandoned their posts; and, finally, teachers resist postings and attendance at isolated rural schools. In more urban areas, the problem ts not so much one of providing places. Though class sizes are large, there is often a surplus of teachers (many of them having moved to urban areas) and the buildings suffice with minimal double-shifting. 25. The increasing demand for private schooling, especially in urban areas, suggests that the problem is probably much more one of providing effective education at a reasonable price. If the government could do this, it would stimulate demand for places in public schools, just as private schools are responding to those segments of the population that can afford to pay sustainable prices. However, Madagascar's economic decline since the early eighties and the resulting decline in direct government expenditures on education have obviously changed parents' decisions about education. The current weighing of factors in parents' decisions about the schooling of their children is not well understood, 8 Even though the number of years in Madagascar's primary cycle is at least one year less than in most other countries.  10 especially at the secondary level where standards of instruction seem to have been better maintained but where the enrollment decline is the greatest. In order to improve its understanding of the demand for education, the goverrnment is planning a study within CRESED. Whatever a study on demand for education may conclude, the analysis here makes it clear that: (a) on the supply side, efforts to improve public primary and secondary education must pay at least as much attention to improving quality as to increasing the number of places; (b) private education plays an important role in education supply and government might consider support in the future in the form of enabling regulations and supervision, performance contracts, incentives andior subsidies; and (c) ways need to be found to get teachers to serve in rural areas. Table 2 MADAGASCAR: Efficiency Indicators For Primary And Secondary Schools Compared With Other African Countries INDICATOR AVERAGE Madagascar Sub-Saharan Africa Madagascar's Ranking' PRIMARY: Percent who repeat (1990) 36% 21% 31 out of ?2 Percent who arrive in final 37.7% 67.1% 27/29 year (1989) Number of years invested per 13.8 years Not available 26/29 student completing first cycle (1989) 2 SECONDARY Percent of primary school grads who continue to first 43% 40% 9/23 year of general secondary Percent who repeat (1990) 20% 20% 14/28 ' Madagascar is compared to countries with available data. 2 Prescribed duration of first cycle in most countries is six years; in Madagascar, five years. SOURCE: Bourdon, lean, T. Hartnett, F. Orivel, C. Shaw. A Statistical Profile of Education in Sub-Saharan Africa in the 1980s. Paris: Donors for African Education, in edit.  11 CHAPTER 3: THE QUALITY OF PRIMARY EDUCATION 26. The study investigated the input factors that affect the system, the process characteristics that make for good schools, and the learning results achieved by children. Chart I presents the factors included in the study and shows their hypothesized relationships before the research. The scope and content of the curriculum have not been assessed in this study, but would not affect the differential performance of individual schools because they are common to all. Case studies covering the factors have been prepared on twelve schools each at the primary, junior secondary and senior secondary levels (hereinafter, "the Quality study"). Additional sources of information on primary education include the study of student achievement in Mathematics, Malagasy and French sponsored by French Cooperation ("the Evaluation study") and a study of the causes of student repetition and drop-out undertaken within the framework of CRESED ("the Wastage study"). The methodologies for these studies are summarized in Annex 4.A. This chapter presents the findings on primary education, the next chapter those on secondary education. 27. The analysis of primary school quality in Madagascar identified three key factors affecting school quality: school leadership, community participation, and teachers' guides and textbooks. The first two of these factors are critical to the life of the school, corroborating conclusions that material resources do not alone explain success. The analysis also found that teacher competencies and a school's facilities and equipment contribute to student success. However, these appear to be less critical than the three priority factors. Also, the importance of facilities and equipment is probably mainly a reflection of the community's very significant role in the life of the school (and the government's absence). Chart 2 summarizes the factors analyzed and depicts their estimated relation to each other based on the analysis of research results. "Government contribution" is included in the diagram, connected by thinner lines, to show where public resources, beyond teachers' salaries, would be best directed to strengthen the school factors that this study has found to be the most important. Student Outcomes 28. The best source of information on learning outcomes in Malagasy primary schools is the 1991 Evaluation of 2,874 students in fifth grade (Classe 7), the last year of primary school. In May 1991, these students from 181 schools were tested in Malagasy, French, and Mathematics using tests based on the Madagascar curriculum. Table 3 shows the average scores and the percentage of students scoring 60% or better in each subject. The average scores range from 56% in Malagasy to 48% in Mathematics. In Malagasy, but not in French, the students performed 11 points better, on average, in reading than in writing. If the exclusion of isolated schools from the sample is taken into account, probably less than half of the small number of the pupil cohort who reach the last year of primary school currently fulfill the curriculum's objectives in these three major subjects. 29. Results on the CEPE, the primary school leaving examination, suggest a similar conclusion, though the results are not standardized. Nationally, until the mid-eighties about half of the students taking the CEPE examination passed it. Starting in 1984 pass rates fluctuated between 30% and 40%, except in 1987 when the rate was again near 50%. The rate dropped to 22.7% in 1989, the last year for which examination results have been released. Among the 12 primary schools in the Quality study, three of  12 Chart 1: MADAGASCAR: Hypothetical Network of Priority Factors that Determine Primary School Effectiveness (Before research) The School Condition Positive Attitude iof School of Teachers / ..Optimal Conditions A K / Government of Work epetition Contribution Academic Teaching Director's > Expectations Results Materials Management of High / Performance Contribution of Classroom Parents and the Learning Activities Community Competent Teaching Staff CONTEXT  13 Chart 2: MADAGASCAR: Conceptual Framework for a Strategy for the Reform of Primary Education (After Research) eac er ttitudes epeti ton Leadership - ~~Learnin CONTEXT Thick Arrows show relations among factors from the research. Thin Arrows are the suggested foci for the government's strategy.  14 the schools had pass rates of over 70% and five of them had rates of 20% or less, suggesting wide disparities among schools. From this study, the disparities appear to be partially explainable by irregularities in test administration that local responsibility for the examination may encourage, suggesting that better standardization of the primary-school-leaving examination is desirable for monitoring system performance. Table 3 MADAGASCAR: Test Results of Pupils At End Year of Primary School, May 1991 (N=2,874) Average Score % Scoring 60% or Better Malagasy (Language) 56% 40% French 54% 30% Mathematics 48% 23% Source: Robin, Daniel, et al. "Evaluation des Acquis des Eleves en fin de T5. " Paris: Ministere de l'Education Nationale et de la Culture. September, 1992, pp. 32-36. 30. Both of these sets of information on primary school learning outcomes suggest that the quality of primary education is weak. The fact that only about 30% of the entering students even reach grade 5 adds to the concern about quality of outcomes, as one may question what those who leave earlier have learned. Only about 15% of the age-group is leaving school with learning achievement at primary level or better, which is totally inadequate. These observations suggest a deterioration of the quality of education and increasing illiteracy in Madagascar. School Process Elements 31. Leadership and Pupil Expectations. The field observations in the Quality study found that the school directors in better schools pay attention to order and discipline and make teacher assignments carefully. On the other hand, in almost all of the schools the other indicators of leadership are weak: there is no evidence of school goals being articulated and shared; there is scant evidence that students are being rewarded for success other than through annual award ceremonies; and the bulk of the observations made of valuing students' efforts and giving them meaningful responsibilities were found in individual teachers' classes. It appears that the expectations that are set by the school leadership, even in good schools, reach the students through individual teachers. Where the head's attention to the orderliness of school life and administration encourage teachers to hold high expectations for students and to reward them, students' results should be better. 32. From the Evaluation study there is even stronger evidence that the expected relationship between school leadership and academic success is valid for Madagascar's primary schools. That study found that the schools with better test results had directors with a higher personal investment in the school in terms of their availability in the school, their attitudes about their responsibilities, their control of the school, and their ability to motivate people. In the national seminar, held in March 1994 to discuss the primary school results of the Quality study, national educational leaders corroborated the importance of the school  15 director. However, the study concluded that none of the school directors in the small sample paid direct attention to supervising teaching or stimulating improvements in teaching practice. Instead, they are loaded with the administrative responsibilities which, when heads handle them well, allow teachers to be more effective with students. These studies suggest that classroom practice, teacher collaboration, and delegation of responsibilities to students are left almost entirely to the teachers. 33. The Capabilities and Attitudes of Teachers. The teaching force has, for the most part, an adequate minimum formal preparation and experience. With the exception of two schools out of twelve in the Quality study, all of the teachers had initial teacher training, and in those two schools, 80% of the teachers had had training. However, only four schools had at least one teacher with the Certificat d'Aptitude Pedagogique (CAP), the highest professional qualification for primary level teaching. Similarly, all but a few teachers in these schools had been on the job for at least two years. In the Quality study, training and years of service were not found linked to outcomes. However, the achievement and drop-out studies found a connection between formal teacher competencies and student outcomes, whether achievement or dropout. The Wastage study found that schools with better-prepared teachers have less dropouts, and the Evaluation study found that length of service and length of time in current post contributed to higher student competencies. Interestingly, both studies also found that female teachers obtain better results than male teachers. Using its teacher interview data, the Evaluation study explains how male teachers usually have numerous obligations, many of them economic, outside their teaching. 34. While the studies agree on the presence and importance of basic professional capabilities, teachers' attitudes appear to detract from their performance. The Wastage study found that teacher absenteeism is higher and the number of hours of teaching per day is less in schools with higher dropout rates. Also, absenteeism without a reason poses a problem in many of the twelve schools in the Quality study. These conditions suggest that some teachers are not dedicated to their work. Beyond this, in these schools, very few examples of a personal commitment to students -- counseling individuals or motivating the community in providing writing materials or a school blouse for needy students -- were found. 35. The Quality study also found problems in how primary school teachers use their knowledge and skills. First, the majority of the teachers in the study's case schools have great difficulty using French, both for instruction and for communication. Second, there is little evidence of pedagogical collaboration among teachers, the examples found being limited to teachers who teach the same grades in the same school. Third, while the teachers have had teacher training, the skills leamed are often applied ritualistically. For example, lesson-planning is done regularly but without being specific to the classes being prepared for. The lack of supervision and staff development by school heads and weaknesses in the teacher training curriculum may contribute to this lack of meaningful application of what was learned in training, but teacher attitudes probably also contribute to weakening the effective application of leamed teaching skills. 36. Even though some teachers teach conscientiously under difficult circumstances, the cluster of teacher attitudes reflected by teachers' attendance and their impersonal behavior towards students needs greater consideration in planning school improvement. The participants in the national seminar in March, 1994, that reviewed the results of the Quality study probably realize this as well because they singled out "professionalization" of teachers, along with school leadership, as a critical problem to resolve in order to improve student learning.  16 37. School Infrastructure and Equipment. Two of the studies analyzed here suggest that in Madagascar the physical facilities of the school contribute to student learning. Attributes that have been found important include the number of classrooms, water and toilet facilities, an adequate-sized compound for recreation and productive activities and equipment (blackboards, desks, other classroom furniture, and tools for productive activities). The physical characteristics of the twelve schools in the Quality study suggest shortages of facilities: only four of the twelve primary schools had enough rooms for all class sections; four had only half as many classrooms as class sections; and in nine schools sanitary installations were inadequate. The Evaluation study found most schools had toilets, but only a few had electricity and running water, with the schools that performed well on the achievement tests having significantly better services than those that performed poorly. On the other hand, the Pilot School Mapping exercise in one commune found similar levels of infrastructure availability, but concluded that schools with "mediocre material conditions for schooling" performed well. These mixed findings on the relation of infrastructure to students' results suggests that the relationships that have been found may be more a reflection of community support for both the school and its pupils than a direct cause of success (see below on community support). 38. School equipment has also been shown to be associated with student outcomes. The finclings of the studies used in this analysis suggest that perhaps two-thirds or more of all schools have sufficient blackboards, pupil seats, and other types of equipment. Also, two studies found that the schools that have more and better blackboards, greater numbers of seats for the number of students, and other classroom furniture tend also to have better CEPE examination results (the Quality study) or higher test scores (the Evaluation study). 39. Learning Materials. Textbooks, slates, notebooks, and pens and pencils constitute the main pupil learning materials in Madagascar's primary schools. The Evaluation study and the Quality study both found that schools do not differ much in the availability of paper and pens for students, both being available for almost all students. However, only limited numbers of textbooks are available. For example, the Quality study found that all schools had some French texts (A Toi de Parler) for all five grades which were provided by French Cooperation and sets of newly-arrived grade 1 Malagasy texts (Garabola), written with German technical and financial assistance and published and distributed under CRESED. Not one mathematics textbook was found in any of the twelve schools studied, and nine of the schools had insufficient numbers of teachers guides for the texts that they did have. fBoth the Evaluation and the Wastage study found that schools with better test scores or lower dropout rates had more books for their pupils than schools that performed less well. Even with limited textbook availability, it is noteworthy that the presence of textbooks seemed to differentiate schools with good and poor results. 40. The Teaching/Learning Process. The only study to look directly at the teaching/learning: process has been the Quality study. In this study, between two and six classes were observed in each school to get a sense of how efficiently class time is used, what the diversity of teaching methods are, how students are evaluated, and h3w frequently homework is given and checked. Recognizing its informal method of classroom observation, the study concluded that there is some diversity of teaching methods, though it is limited by the paucity of materials and the teachers' lack of imagination. The study also found that in the higher grades teachers tend to use a more formal, expository, teaching approach than in lower grades. Conclusions were not reached about homework and about how efficiently class time was used, though a few examples of good practice in both these areas were found. Generally, the study indicates  17 that teachers know the basic teaching skills, but they are not being challenged to use them effectively or creatively or to grow in the practice of their profession. 41. Supporting Input Elements: Government and Community Contributions. While neither of the quantitative studies reviewed looked directly at communities' and governments' contributions to schools, the multi-variate analysis of the factors contributing to dropouts indicated that about 20% of the variance in dropouts among schools could be explained by family context, including the family's aid to the student. On the other hand, the Quality study looked at parents' and communities' contribution to the school in terms of money, labor, materials and participation in school life and at government's contribution in terms of money, building construction and repairs, and teacher development services. Only two of the twelve schools received any financial assistance from government beyond teachers' salaries, in both cases a small subsidy from local government; four had help with building and equipment renovations; and seven had received token amounts of materials. Government investments at the primary level have been almost nil. 42. While all twelve schools reported teachers participating in in-service activities provided by government, these were either short courses on teaching French sponsored by French Cooperation or long-standing regular meetings, called F4, led by local supervisors. Most of the teachers reported that the teacher's meetings are formalistic and/or social and not at all useful for their teaching. Only two of the twelve primary schools in the study reported having a supervisory visit or inspection in the last couple of years, and in both cases there were special reasons why a visit had occurred. 43. Not one of the twelve schools studied was found to have poor community support. In fact, ten of the communities were classified as giving good material support to the school, including locally-set fees that ranged from 500 FMG to 3000 FMG per year. The uniformity of community support across all twelve schools, when the academic results of the schools varies widely, leads to the counterintuitive hypothesis that community support is not important to learning outcomes. However, two key factors that do differentiate schools -- leadership and infrastructure -- derive their significance from interplay with parents and the community. First, as has been found, the variety in facilities and equipment among the schools derives from the communities' participation. If good school results are related to the infrastructure, it is because commnunity support exists to create and maintain the infrastructure. Similarly, a school director's influence on outcomes can be expected to be stronger through interaction with the community supporting it. As was noted, the Evaluation study found that school heads in good schools tended to be available in the school and to be able to motivate people. Both of these attributes add depth and quality to the relationship between a school and the community. For both these key factors, there is justification to conclude that they are important for school effectiveness in concert with support from the conmnunity. 44. The Quality study concluded that "parents assure the life of the school," and "there's not much of a contribution by government" (other than salaries). Since the schools in this study were all near to or in the larger regional cities, this picture of very little government participation in public schools is probably even more true for rural areas. As the study found, the absence of government support, other than salaries, leaves the supply of school infrastructure, equipment, and pupil supplies to the parents and the community. Where parents are keen on the education of their children, more material support for education will be found. The finding that good facilities and equipment contribute to better academic results is probably, in fact, a proxy for the broader influence of parental and community support.   19 CHAPTER 4: THE QUALITY OF SECONDARY EDUCATION 45. Even though the same conceptual framework and definitions of quality as for the primary schools were used to study 24 secondary schools, it is difficult to draw firm conclusions about secondary education from the analysis because there are no other research results on the determinants of the quality of secondary education to reinforce this small study. Still, the study begins to define a focus for future efforts at reform. As with the primary schools, the director's management and the school's physical facilities are important factors in school success. It also appears that the teachers' attitudes and competencies are more important in secondary schools (including, for the lycees, their expectations of pupils) than are learning materials. However, the data from the case studies also suggest that learning materials may be important and that parents play a critical role in supporting the school, despite the fact that all schools receive some government financial assistance. Because of the limitations in the number and the representativeness of the cases, additional research is required to clarify the tentative findings of this study before the current strategy for the improvement of secondary education is revised. Therefore, this study does not follow the treatment of primary education presented in Chapter 3, in simulating the cost implications of changes in policy. The sections that follow briefly discuss each of the seven main factors in the conceptual framework, combining results for the CEGs and the lycees, on the way to drawing these tentative conclusions. Student Outcomes 46. The only national data available on student outcomes at the secondary level are examination results through 1989. According to the published results, between 1984 and 1989 35% or fewer of the students taking the junior secondary leaving examination passed. For the five years for which data exist on the Iycees, 1985-1989, the national average pass rate was always less than 20%. There is no evidence that the pass rates have changed since 1989, and the Quality study gives some idea of the more recent variation in rates among schools. Among the twelve CEG the BEPC pass rates varied from 49% to 10% in 1992, and for the twelve lycees, results on the leaving examination (the "baccalaureat") varied from 40% to one school in which no one passed. Assuming that the examination tests what is covered by the curriculum, these pass rates suggest that the educational environment in a large number of secondary schools is not very effective. School Process Elements 47. Leadership and Pupil Expectations. In both the CEG and the lycees in the Quality study, the school directors delegate to coordinators most of the responsibility for making teaching assignmenis, for coordinating teaching activities, and for classroom supervision. Pedagogic supervision was found to be done by only 3 CEG directors and by no Iycee directors. The study found that this delegation of responsibilities, usually through assigned coordinators, puts directors at a distance from the life of the school. Similarly, most of the responsibility and rewards given to students are done through teachers, particularly in class, and directors articulate school goals and communicate them to students in only a few schools. Instead, Iyc&e directors seem to spend most of their time on administrative matters. 48. The Capabilities and Attitudes of Teachers. The academic and professional preparation and experience of the teachers in the secondary schools that were studied is higher than in many other African  20 countries. In all but four CEGs, more than 80% of the teachers had received some formal 1:eacher training, and in the lycees all teachers had an undergraduate degree (licence) and/or formal teacher training. In all the CEGs and half of the Iycees, all of the teachers had more than two years of experience. However, at both levels the number of teachers with the currently prescribed l-vel of professional preparation was very limited, and smaller rural schools with fewer teachers tend to lack expertise in all subjects that are taught, partly because teachers only train in one subject. 49. In the 24 secondary schools studied the utilization of teachers' knowledge and skills is affected by their attitudes towards teaching and the students. While there are examples of dedication to students, the signs of a poor commitment to teaching outweigh them. In the CEGs, seven of the twelve schools reported teacher attendance problems, and at both levels there is little collaboration among teachers even though Ministry regulations officially establish a Pedagogic Committee in each school. Furthermore, the study found that the majority of teachers, especially the young ones, lack confidence in their ability to teach, partially because of their perceived lack of knowledge and skills but most importantly because of their inability to teach effectively in French. The research team estimated that half of the 44 teachers contacted in the study do not speak French. In the lycees, the problem of French does not seem to be as significant as in the CEGs, but there are still signs that teacher confidence is low there too, perhaps because of the high proportion of the teachers who do not have any teacher training. 50. School Infrastructure and Equipment. As at the primary level in the Quality study, the schools' physical conditions seem to be related to students' results. This is especially true for the lyce&s where the two schools with the best examination results are also the only two schools judged to have "good" facilities. However, the general quality of the facilities is better than in the primary schools. A,most all secondary school settings and grounds were judged adequate, and all but three schools have running water and enough classrooms. Based on the analysis of the twelve CEG case studies, "enough" is defined as four rooms for each five sections, the level at which it was found that directors could schedule all classes during established school hours. The three lycees that have half the number of rooms required to house all classes run on two shifts. Also, sixteen of the twenty-four schools have libraries. However, only two lycees have laboratories, and laboratories at the CEGs are "non-existent or inadequate." The frequency and intensity of use of these special facilities was not noted. 51. Similarly, classroom desks and blackboards were evaluated to be adequate in all but one of the twenty-four schools. On the other hand, there is a general lack of teachers' desks in the classrooms, perhaps because the teachers move from classroom to classroom to teach, and only two or three schools at each level had tools for productive activities. In general, facilities and equipment for the secondary schools are basically adequate, though improvements could be made in some schools. Future efforts should seek to make these improvements where they are needed and to increase the use of lahoratories and libraries. 52. Learnin2 Materials: Contrary to what would be expected, the Quality study could niot find a relation between the availability of learning materials and examination results or dropout rates in junior or senior secondary schools because conditions across schools did not vary. There was an almost complete absence of textbooks and teachers guides in any of the schools. No student textbcoks were observed in ten of the CEGs and in six of the lyc6es, and only two 1ycees (no CEGs) were judged to have even a modest level of textbooks. Even more striking, not one teacher's guide was found in any of the  21 lycees, nor in eight of the CEGs. In the face of this situation, the research team did find a few teachers who had purchased books for their own use or who had prepared or borrowed materials for use with the students. On a more positive note, as in the primary schools, students seem to be adequately equipped with notebooks, pens and pencils. 53. The hypothesis is obvious that the nearly complete lack of printed learning materials for both students and teachers is contributing significantly to the poor learning outcomes in Madagascar's secondary schools. However, even though the Quality study has documented this deplorable lack in all the schools studied, it does not provide a compelling case for this conclusion. There is a need for more sophisticated research on learning outcomes and their determinants, as in the primary-level Evaluation study, in order to determine the importance of learning materials in comparison to other inputs for secondary schools in Madagascar. 54. The Teaching/Learning Process. Two general findings about teaching methods at both levels of secondary education stand out from the Quality study. First, even with the paucity of materials, the foreign language teachers attempt to vary their teaching methods. This is not surprising given that donors have been supporting in-service training of language teachers. On the other hand, teachers in the sciences tend to center all instruction on themselves, either talking, writing on the blackboard, correcting student mistakes, or dictating material. Among lyc&e teachers, their discourse tends to be even more theoretical and didactic than in the CEGs. Consequently, students at both levels tend to write a lot and to speak verv little. Here, as with the implications about learning materials, additional research could help clarify how different teaching methods have an impact on learning outcomes. 55. Supportinz Input Elements: Government and Community Contributions: At the secondary level, as at the primary level, community resources provide most of the financial and material support for the schools in the Quality study, but with important differences related to financial assistance and local responsibilities. First, contrary to primary education, every secondary school in the study receives some financial or material assistance from the central government, albeit in such limited amounts that it does not seem to have an impact on the quality of the education provided. Neither has the government provided very much support in terms of in-service training or of inspection, especially for the lycees. While all the CEGs had staff who had had a recent opportunity to participate in some kind of in-service course, and three schools had been visited by an inspector, no Iycee had been visited by an inspector in the last couple of years and only one school reported teachers participating in in-service training, though some counseling and mentoring of teachers was reported. 56. With such limited support from government, local resources have to fill in for them. The Quality study found that at the secondary level it is much more the parents, and occasionally former students, than the community who provide support to the school. Not only do the parents pay fees, but examples were also found where they constructed latrines and classrooms, purchased school sports equipment, desks and prizes for students, and paid water and electricity bills. Also, more than at the primary level, the parents' support often includes fairly active participation in school life. 57. While the financial allocations made to secondary schools by the government suggest a greater commitment to supporting this level of education over primary education -- and comparative per-student expenditures for instruction corroborate this commitment -- the government's role in ensuring quality schooling beyond teachers' salaries is minimal. This lack of significant involvement in preparing the  22 nation's future educated manpower probably accounts for many of the weaknesses in teachers' attitudes and classroom methods and in school directors' failures to provide much leadership. A successful strategy for improving the quality of secondary education will have to ensure that the government's role is strengthened to better address these central issues, directly in the public schools and through oversight of private schools.  23 CHAPTER 5: THE PUBLIC FINANCING OF EDUCATION 58. Before examining the policy implications of the findings on the quality of education, it is necessary to examine the extent to which the financing of education has contributed to the problems of access, internal efficiency and quality that have been discussed, as well as to identify where possible sources of additional funds may be found to support improvements. Although school-level analyses suggest that there is not much of a contribution by government to the life of the school, public expenditures on primary and secondary education are significant. The government pays teachers' salaries and supports a large number of administrators of the education system. Parents and communities contribute what money, labor and materials there are for facilities and non-salary recurrent expenditures at the school level. This section examines the amount of and trends in public financing of education in order to identify whether and how government can supplement private contributions to operating costs to improve the quality of education. The analysis looks at (a) the share of GNP and of total government budgets that goes to education, (b) the share of education budgets that are allocated to primary and secondary education, and (c) expenditure patterns within these sub-sectors of education. 59. Madagascar had a modest public budget in relation to GNP because its fiscal effort is unduly low for a country with its economic level and structure. However, it allocates a reasonable share of this budget to education. Within education, higher education takes a higher percentage than in many other countries, and over two-thirds of non-salary funding in higher education is spent for non-instructional purposes. The result has been a very small public resource base for direct support to instruction, especially at the primary level. 60. There has been insufficient information to quantify total spending on education. Information on actual public expenditures on education is only available for the period 1988-91, and private spending is not recorded, even though it is significant. Future research should look more closely at how households finance education, both with contributions to government institutions and to the private schools that absorb a growing proportion of the students. Public Funding of the Education Sector 61. Allocations to education as a percentage of GDP (for which data were available) fluctuated between 2.3 % and 2.6 % during the period 1988-1991. This is low, the average being 4.3 % of GNP for all of Sub-Saharan Africa in 1990 and 3.0% for the low-income countries. The modest size of Madagascar's total budget in relation to GDP seems to be the major contributor to the low education/GDP ratio, suggesting a need for attention to ways to increase government's total revenues. 62. As is presented in Table 5.1, Annex 5, the total national budget in Madagascar has grown more slowly than the GNP in 1991, but the reverse was true in 1989 and 1990. In this context, allocations from the national budget for education have fared well. Although the share of the national budget going to education declined to 16% of the total in 1989 and 1990, it increased to 25% in 1991, even as the total national budget decreased significantly (Table 5.2, Annex 5). These allocations to education compare favorably with the averages for country groupings in the rest of Sub-Saharan Africa which range from 13% to 20%.  24 Macro-Economic Perspective on Trends in Public Revenues 63. The government is currently putting in place a program of macroeconomic reform, aimed at achieving a high rate of economic growth. Reform of government finances is the keystone of macroeconomic equilibrium; rapid growth cannot take place without it. In order to achieve high GDP growth, the overall investment rate will have to increase without either producing undue inflationary pressures or depriving the private sector of the resources it needs for development. The increase in the investment ratio as a percentage of GDP required to meet the country's growth objectives should result from expanded private investment. Government financial policy will aim at increasing public savings, controlling government expenditure (with redirection towards high priority sectors, such as human resources), and raising fiscal revenue and user fees. 64. Revenue expansion is a key objective of the government's proposed reform program. Tax collections represented only 8% of GDP in 1993. The government intends to increase fiscal receipts by two percentage points of GDP during each year of its reform program. To this end, a review of the fiscal system will be undertaken in 1995; this review will focus on improvement of tax administration, simplifying the tax structure, modernizing the tax on companies, and establishing a simpler, more efficient system of indirect taxation. 65. In order to reduce Madagascar's budget deficit, government expenditures will be carefully controlled. At the same time, the public sector investment program should be redirected to focus more attention on key areas like basic education and health, and the government intends to increase spending on basic education, to maintain the current level in real spending per capita. Moreover, the quality of expenditure has to be increased, so that expenditures actually benefit the target population by being spent on instruction. Other categories of expenditure will have to decrease. Allocations to Primary and Secondary Education from the Education Budget 66. Primary and secondary education's share of public expenditures for education in Madagascar has stayed around 70% of the total education budget, including expenditures for secondary technical education, between 1988 and 1991 (Table 5.4). Budget figures for 1992 and 1993 suggest that this share has been rising to about 75%, so that the budget for 1993 foresaw per-student expenditures, salaries included, of 29,704 FMG ($15.63) at the primary level and 103,490 FMG ($54.47) at the secondary level. However, actual non-salary expenditures during the year have in the past usually favored the higher levels in the system, reducing the actual share going to the primary level. This may have occurred in 1992 and 1993, but data are not available to check. If the actual expenditures on primary and secondary education have reached 75 % of the total education budget, they will be roughly at the average for all of Sub-Saharan Africa, but below the 83% in 1990 that characterized other countries in the region with conditions similar to Madagascar's. 67. Because data on the shares of the salary budget for primary and secondary education have not been available for this study, the only analysis of primary education's share of expenditures that has been possible is a comparison of the percentage distribution and per pupil non-salary operating allocations among primary, secondary and technical education. Table 5.5, Annex 5, summarizes this data for 1988- 1993 according to the three main categories of non-salary expense: Administration, Aid and Transfers, and instructional expenditures (Enseignement). Only the last of these is budgeted by level of education.  25 As can be seen in Table 5.5, Annex 5, primary education's share of instructional expenditures has fluctuated greatly, from 7% of the total in 1988 to 35% in 1990 and back to 17% in 1992. The budget for 1993 shows a major increase in the funding for primary level instruction (55% of the total), suggesting the new government's recognition of the current crisis and of the potential social benefits of primary education. If that amount has actually been spent on direct support to primary education in 1993, it could mark a significant turnaround in the financing of those expenditures benefitting students most directly, even though the absolute per-student amount in 1993 is still small (3900 FMG, or US$2.05) compared to a recommended target of about US$ 5 per student 9. Expenditure Patterns within Primary and Secondary Education 68. Besides giving an indication of how little money is spent directly on primary education, T'able 5.5, Annex 5 also summarizes the overall imbalance between non-instructional expenses and instructional expenses. Over a six-year period, 1988-1993, the proportion of non-salary expenditures for primary and secondary education going to administration averaged about 60%. If the additional 10% or so for "Aid and Transfers' (mainly scholarships) is taken into account, only about 30% of the non-salary education budget has been available for instructional purposes, probably including supervisors' expenses. It is not surprising that the Quality study found that central government contributions to secondary education at the school level are small and that there are no contributions to primary schools. Future policies need to enlarge the amount for instructional purposes, however the actual expenditures on instruction are defined, while maintaining the most necessary elements of administrative expenditures. 69. To emphasize just how little can be spent on instruction, Table 5.6 summarizes the consequences of these decisions for per-student expenditures on instruction in 1991, the last year for which actual expenditures are available and the worst year reported on for primary education. The average per-student non-salary allocation for primary, secondary, and technical education in that year was 5,365 FMG ($2.82); but the non-salary expenditure for the fewer than 10,000 technical school students was 83,359 FMG (US$ 45.45) per student, 1100 times more than the 78 FMG (US$ .04) per student spent on the 1.3 million primary school students. General secondary school expenditures were 6,951 FMG ($3.66) per student, eighty-nine times more than that spent on a primary school student, but still well below US$5 per student. These comparisons were much better in 1992 and 1993 where significant increases in non- salary budgets have occurred as the new government has taken hold. Other Sources of Financing for the Education Sector 70. In addition to the public budget, other sources of financing for the sector comprise an important share of both capital and recurrent expenditure and may become increasingly important given the limited ability of the public budget to accommodate both quality and quantity expansion. These sources include households, the private sector and foreign aid. 71. Households, especially parents, already contribute a substantial portion of the recurrent and capital costs of public schools. According to recent estimates, households spend about 1 % of their income on 9 In A World Bank Policy Paper - ImDrovinM Primary Education (1990), the Bank recommended raising the level of per- student expenditures on teaching-leaming materials to US$5 per annum in low-income countries.  26 education, for both capital and recurrent costs. This is equivalent to about a third of public expenditure on education. However, the absolute levels of expenditures seem to be roughly constant in all income groups, implying higher percentages of household incomes spent by the poor. In addition, averages conceal the fact that while some children may have all their materials purchased by parents, others have none. In fact, the beneficiary assessment of poverty '