Education and training in Madagascar : toward a policy agenda for economic growth and poverty reduction : Education and training in Madagascar - toward a policy agenda for economic growth and poverty reduction

o~~oC 000 E3~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~L a4- 8~ 8 A~~~~~~~~M   A WORLD BANK COUNTRY STUDY Education and Training in Madagascar Toward a Policy Agendafor Economic Growth and Poverty Reduction The World Bank Washington, D.C.  ( 2002 The International Bank for Reconstruction and Development / The World Bank 1818 H Street, NW Washington, DC 20433 All rights reserved. 1234040302 World Bank Country Studies are among the many reports originally prepared for internal use as part of the continuing analysis by the Bank of the economic and related conditions of its developing member countries and of its dialogues with the governments. Some of the reports are published in this series with the least possible delay for the use of governments and the academic, business and financial, and development communities. The typescript of this paper therefore has not been prepared in accordance with the procedures appropriate to formal printed texts, and the World Bank accepts no responsibility for errors. Some sources cited in this paper may be informal documents that are not readily available. The findings, interpretations, and conclusions expressed here do not necessarily reflect the views of the Board of Executive Directors of the World Bank or the governments they represent. The World Bank cannot guarantee the accuracy of the data included in this work. The boundaries, col- ors, denominations, and other information shown on any map in this work do not imply on the part of the World Bank any judgment of the legal status of any territory or the endorsement or acceptance of such boundaries. Rights and Permissions The material in this work is copyrighted. No part of this work may be reproduced or transmitted in any form or by any means, electronic or mechanical, including photocopying, recording, or inclusion in any information storage and retrieval system, without the prior written permission of the World Bank. The World Bank encourages dissemination of its work and will normally grant permission promptly. For permission to photocopy or reprint, please send a request with complete information to the Copy- right Clearance Center, Inc., 222 Rosewood Drive, Danvers, MA 01923, USA, telephone 978-750-8400, fax 978-750-4470, www.copyright.com. All other queries on rights and licenses, including subsidiary rights, should be addressed to the Office of the Publisher, World Bank, 1818 H Street NW, Washington, DC 20433, USA, fax 202-522-2422, e-mail pubrights@worldbank.org. ISBN: 0-8213-5164-8 ISSN: 0253-2123 Cover Photo: Tsiroanomandidy, Madagascar, Girindre Beeharry. Library of Congress Cataloging-in-Publication Data: World Bank Education and Training in Madagascar: towards a policy agenda for economic growth and poverty reduction / World Bank. p. cm. - (A World Bank country study) Inlcudes bibliographical references (p.). ISBN 0-8213-5164-8 1. Education and state-Madagascar. 2. Economic development-Effect of education on. 3. Educational planning-Madagascar. 4. Education-Economic aspects-Madagascar. I. Title. II. Series. LC95.5.M42 W67 2002 379.691-dc2l 2002069160  Contents Foreword xi Abstract xiii Acknowledgments xv Abbreviations and Acronyms xvii Executive Summary xix I Demographic and Macroeconomic Context 1 The Demographic Context 1 Economic Growth and Current Conditions 3 Overall Government Finances 4 Overall Public Spending on Education 5 Prospects for Increased Spending on Education, and the Implications 7 Conclusion 8 Notes 9 2 Coverage and Structure of the Education System 10 Aggregate Enrollment Levels and Trends 10 Public and Private Sector Roles in Education 11 Levels and Trends in Gross Enrollment Ratios 12 Grade-to-Grade Student Flow Profiles in Primary and Secondary Education 14 Indices of Overall Student Flow Efficiency 18 Policy Perspectives on Grade Repetition and Dropout in Primary Education 21 Conclusion 24 Notes 24 Hii  Vi EDUCATION AND TRAINING IN MADAGASCAR 3.4 Government and Household Spending on Public and Private Primary Education, Madagascar, 1999 30 3.5 Government Employees in the Education Sector, Madagascar, 1998 31 3.6 Teacher Salaries and Distribution of Teachers by Salary Grade, Madagascar, 1998 33 3.7 Distribution of School-Level Nonteaching Staff by Salary Grade, Madagascar, 1998 36 3.8 Current Public Spending by Function, Ministry, and Level of Education, Madagascar, 1998 36 3.9 Distribution of Current Public Spending on Education by Business Category, Madagascar, 1998 38 3.10 Public Spending Per Student by Level of Education in the Public Sector, Madagascar, 1998 40 3.11 Current Public Spending Per Student, Madagascar and Developing Regions, 1990s 41 3.12 Decomposition of Public Spending Per Student in Public Schools, Madagascar, 1998 42 3.13 Teacher Salaries and Pupil-Teacher Ratios, Madagascar and Developing Regions, 1990s 43 4.1 Gross Enrollment Ratios by Province, Urban-Rural Locality, -and Gender, Madagascar, 1990-99 47 4.2 Distribution of School-Age Population and Enrollments by Locality and Gender, Madagascar, circa 1999 48 4.3 Disparities in Entry Rates to Grade 1, Madagascar, 1997 49 4.4 Disparities in Rates of Survival to Grade 5, Madagascar, 1997 50 4.5 Disparities in Grade 5 Enrollment Rates, Madagascar, 1997 50 4.6 Incidence of Public Spending on Education among Members of a Cohort, Madagascar, 1999 52 4.7 Incidence of Public Spending on Education, Madagascar and Developing Regions, 1990s 53 4.8 Share of Public Spending on Education Benefiting the Poorest and Richest Population Quintiles, Madagascar and Selected Countries, 1990s 54 4.9 Simulations of the Equity Impact of Alternative Structures of Unit Spending on Education and Student Flow, Madagascar 55 5.1 Number and Distribution of Primary and Secondary Schools, Madagascar, i998 58 5.2 Provincial Distribution of Primary and Secondary Schools, Madagascar, 1998 59 5.3 Characteristics of Public and Private Schools, Madagascar, 1998 60 5.4 Characteristics of Public and Private Primary Schools by Province, Madagascar, 1997-98 61 5.5 Public Subsidies to Private Schools, Madagascar, 1993-99 62 5.6 Subsidized Private Schools and Teachers, Madagascar, 1993-99 63 5.7 Student-Teacher Ratios in Primary and Secondary Schools by Sector and Locality, Madagascar, 1998 65 5.8 Provincial Regression Estimates of the Relation between Numbers of Teachers and Pupils, Primary Schools, Madagascar, 1998 66 5.9 Distribution of Public Primary Schools according to Adequacy of Teacher Complements, Madagascar, 1998 68 5.10 Regression Estimates of the Relation between Numbers of Teachers and Students in Public Secondary Schools, Madagascar, 1998 69 5.11 Distribution of Public Secondary Schools According to Staffing Patterns, Madagascar, 1998 70 5.12 Staff Redeployment and Recruitment, Madagascar, 1997-2000 71 5.13 Teacher Surpluses, Deficits, Redeployment, and Recruitment in Public Primary Schools, Madagascar, Late 1990s 71 5.14 Predicted Probability of Receiving at Least One of the Primary School Teachers Recruited during 1997-2000 in Relation to Adequacy of Staffing in 1997-98, Madagascar 72 5.15 Teacher Surpluses, Deficits, and Redeployment in Public Lower Secondary Schools, Madagascar, Late 1990s 73 5.16 Size Distribution of Primary Schools and Teaching Arrangements in the Public and Private Sectors, Madagascar, 1998 76 5.17 Size Distribution of Secondary Schools, Madagascar, 1997-98 78  CONTENTS VIi 5.18 Government Grants for Public Secondary Education, Madagascar, 1995 and 1999 79 5.19 Teaching Loads and Subject Specialization among Secondary School Teachers, Madagascar, 1998 80 5.20 Trends in Pass Rates on National End-of-Cycle Examinations, Madagascar, 1989-99 81 5.21 French and Mathematics Test Scores, Grades 2, 4, and 5, Madagascar, 1997-99 82 5.22 Correlates of Combined Scores in Mathematics and French, Grade 2, Madagascar, 1997-98 83 5.23 Private School Advantage over Public Schools in Test Scores, Madagascar, 1997-98 84 6.1 Enrollments in Vocational and Technical Education and Secondary General Education, Public and Private, Madagascar, 1990-97 94 6.2 Access to General Secondary Education and Vocational and Technical Education, by Household Quintile, Madagascar, 1997 95 6.3 Gender Differences by Field of Vocational and Technical Study, Public Sector, Madagascar, 1990-91 and 1998-99 95 6.4 Characteristics of Service Delivery in Vocational and Technical Education, Madagascar, 1990s 96 6.5 Number of Institutions and Students in Public Sector Vocational and Technical Education, Madagascar, 1990-91 and 1998-99 97 6.6 Course Offerings in Public Sector Vocational and Technical Institutions, Madagascar, 1998-99 98 6.7 Trends in Enrollments in Public Vocational and Technical Education, Madagascar, 1990s 99 6.8 Access to Public Vocational and Technical Education, by Province, Madagascar, 1991-92 and 1998-99 100 6.9 Student-Teacher Ratios in Vocational and Technical Education, Madagascar, 1990-91 and 1998-99 101 6.10 Regression Estimates of the Relation between Number of Teachers, Enrollments, and Diversity of Course Offerings, Vocational and Technical Education, Madagascar, 1999 102 6.11 Ratio of Students to Nonteaching Staff in Public Vocational and Technical Education, Madagascar, 1998-99 102 6.12 Current Unit Cost of Teaching Activities in CFPs and LTPs, Madagascar, 1998 102 6.13 Unit Costs of Vocational and Technical Education as a Multiple of Unit Costs of Secondary General Education, Madagascar and Selected Countries, 1980s and 1990s 103 6.14 Current Unit Cost by Field of Study, Madagascar, 1999 104 6.15 Regression Estimates of the Relation between Total Cost and Size of Enrollments, Vocational/Technical Education, Madagascar, 1999 104 6.16 Output of Graduatesfrom Public Vocational and Technical Education, Madagascar, 1984-99 105 6.17 Internal Efficiency of Vocational and Technical Education, Madagascar, Late 1990s 106 6.18 Entry Rate of LTP Graduates into Higher Education, Public Sector Only, Madagascar, 1997-98 107 6.19 Distribution of New Entrants into Higher Education by Type of Program, Public Sector Only, Madagascar, 1997-98 107 7.1 Entry Rates to Higher Education and Trends in Enrollments by Sector, Madagascar, 1987-99 112 7.2 Trends in Enrollments in the University Faculty Sector, Including the "Professionalized" Programs (Formations Professionnalisantes, FP), Madagascar, 1985-99 113 7.3 Characteristics of Course Offerings in the University Faculty Sector, Madagascar, 1998-99 114 7.4 Number of Students in Ecoles Superieures and University Instituts, Madagascar, 1985-99 115 7.5 Characteristics of Course Offerings in Ecoles Superieures and University Instituts, Madagascar, 1998-99 115 7.6 Enrollments in ISTs by Field, Madagascar, 1992-99 116 7.7 Characteristics of Course Offerings in ISTs, Madagascar, 1998-99 116 7.8 Enrollments in Distance Education by Field, Madagascar, 1992-99 116  Viii EDUCATION AND TRAINING IN MADAGASCAR 7.9 Characteristics of Private Higher Education, Madagascar, 1983-99 117 7.10 Number of Permanent Teaching Staff in Public Higher Education, Madagascar, 1990-99 118 7.11 Student-Faculty Ratio by Field, Madagascar, 1990-91 and 1998-99 119 7.12 Distribution of Permanent Teaching Staff by Grade and by University, Madagascar, 1995-99 120 7.13 Rate of Remuneration for Overtime Teaching Hours, Madagascar, 1992-2000 120 7.14 Number of Overtime Teaching Hours in Public Universities, Madagascar, 1992-98 121 7.15 Teaching Hours, Overtime, and Student-Faculty Ratio in Public Universities, Madagascar, 1997-98 121 7.16 Number of Administrative and Technical Staff in Public and Private Higher Education, Madagascar, 1998-99 122 7.17 Number of Administrative and Technical Staff in Universities, Madagascar, 1990-99 122 7.18 Number of Graduatesfrom Public Higher Education, Madagascar, 1990-91 and 1996-97 123 7.19 Number of Students and Repeaters in Public and Private Higher Education, Madagascar, 1989-90 and 1997-98 124 7.20 Schooling Career of a Cohort of Entrants to Public Higher Education, Madagascar, 1996-97 125 7.21 Unit Costs in Public Higher Education, Madagascar, Late 1990s 126 7.22 Unit Costs in Distance Education and Traditional Higher Education, Madagascar and Selected Developing Countries 127 7.23 Unit Costs in Private Institutions, Madagascar, 1994-95 127 7.24 Spending on Student Aid, and Number of Students in Higher Education Receiving Grants, Madagascar, 1990s 127 7.25 Student Grants in Public Higher Education in Relative Terms, Madagascar, 1997 128 7.26 Share of Grantholder Students by Field, Madagascar, 1990-91 and 1998-99 128 8.1 Trends in the Economically Active Population, Madagascar, 1993 and 1999 131 8.2 Distribution of Employment by Sector, Madagascar, 1993 and 1999 131 8.3 Distribution of Employment by Type, Madagascar, 1993 and 1999 133 8.4 Salaries of Wage Earners in the Formal and Informal Sectors, Madagascar, 1999 133 8.5 Rates of Return to Education, Madagascar, 1999 133 8.6 Output of Graduates and Their Absorption in the Labor Market, Madagascar, circa 1999 135 8.7 Percentage Distribution of Wage Earners by Educational Attainment Relative to Meanfor Workers in the Same Occupation, Antananarivo, 1996 137 8.8 Unemployment and Job Search Duration by Level of Education, Madagascar, 1993 and 1999 139 8.9 Educational Attainment of the Work Force, Madagascar, 1993 and 1999 141 8.10 Main Business Constraints Encountered by Malagasy Firms, 1998 142 8.11 In-Service Training by Formal Sector Firms, Madagascar and Selected Countries, 1990s 143 8.12 Predicted Probability of Participating in In-Service Training, by Educational Level and Age, Madagascar, 1999 143 8.13 Supply of In-Service Training, Madagascar, 1996 144 8.14 Hourly Cost of Training by Type, Madagascar, 1997 145 8.15 Hourly Cost of Trainingfor Skills Upgrading by Economic Sector of Training Offered, Madagascar, 1997 145 Figures 1.1 External Debt Service and Public Spending on Education, Madagascar, 1990-99 6 1.2 Relation between External Debt Service and Public Spending on Education, Madagascar, 1990-99 7  CONTENTS IX 1.3 Relation between Public Spending on Education and Expected Years of Schooling, Selected Countries, ctrca 1993 8 2.1 Share of Private Schools in Primary and Secondary Enrollments, Selected Countries, circa 1996 13 2.2 Student Flow in Primary Schooling, Madagascar and Selected Countries, circa 1993 20 2.3 Impact of Making Complete Cycle of Instruction Available and Reducing Repetition Rate, Primary Education, Madagascar, 1998 23 3.1 Staff Distribution by Function and Level of Education, Madagascar, 1998-99 32 3.2 Distribution of Current Public Spending on Education by Function and Level, Madagascar, 1998 39 4.1 Grade Attainment Profile of Cohort Age 15-19, Madagascar, 1997 51 5.1 Relation between Number of Publicly Paid Teachers and Number of Pupils, All Public Primary Schools, Madagascar, 1998 66 5.2 Relation between Numbers of Teachers and Pupils, Public Secondary Schools, Madagascar, 1998 68 5.3 Relation between Spending Per Pupil and School Size, Public Primary Schools, Madagascar, 1997-98 75 5.4 Relation between Spending Per Student and School Size, Public Lower Secondary Schools, Madagascar, 1997-98 77 5.5 Relation between Spending Per Student and School Size, Public Upper Secondary Schools, Madagascar, 1997-98 77 5.6 Fourth-Graders' Test Scores, Madagascar and Selected Developing Countries, 1999 81 5.7 Relation between Per-Pupil Spending on School Personnel and Test Scores, Public Schools, Madagascar, 1997-98 85 5.8 Relation between Examination Pass Rates and Spending Per Student on School Personnel, Madagascar, 1998 86 6.1 Vocational and Technical Students as Share of General Secondary Education Students, Madagascar and Selected Developing Countries, circa 1990 94 6.2 Relation between Number of Teachers and Number of Students, Madagascar, 1999 101 6.3 Relation between Total Current Cost and Number of Students Enrolled in Public Vocational and Technical Education, Madagascar, 1999 103 8.1 Relation between Per Capita GNP and Share of Labor Force outside Agriculture, Madagascar and Other Developing Countries, 1995 132 8.2 Relation between Per Capita GNP and Coverage in Higher Education, Selected Countries, circa 1995 140 Appendix Tables A1.1 Government Revenue and Expenditure, Madagascar, 1990-99 149 A1.2 Public Spending on Education, Madagascar, 1990-2000 150 A3.1 Public Spending by Level of Education, Madagascar, 1990-2000 151 A3.2 Personnel in Public Primary and Secondary Schools, Madagascar, 1990-2000 151 A3.3 Numbers of Government Employees by Education Sector Budget Vote, Madagascar, 1991-2000 152  X EDUCATION AND TRAINING IN MADAGASCAR A3.4 Distribution of Government Employees by Education Sector Budget Vote, Madagascar, 1991-2000 153 A3.5 GDP Per Capita and Per Worker, and Compensation of Government Employees in the Education Sector, Madagascar, 1992-98 153 A3.6 Salaries of Teachers and of Other Wage Earners with Similar Qualifications, Madagascar, 1997 154 A3.7 Distribution of Education Sector Staff by Function and Level of Education, Madagascar, 1998-99 154 A5.1 Distribution of Public and Private Sector Primary and Secondary Schools by Province, Madagascar, 1997-98 155 A5.2 Percentage of Primary Schools Offering Full- and Part-Day Instruction, Madagascar, 1997-98 155 A5.3 Regression Estimates of the Relation between Number of Teachers and Number of Pupils, with Provincial Dummy Variables, Primary and Secondary Schools, Madagascar, 1997-98 156 A5.4 Regression Estimates of the Relation between Teacher Allocation and Number of Lower Secondary Students, Public Schools, Madagascar, 1997-98 156 A5.5 Distribution of Public Primary Schools by Adequacy of Teaching Complement, Madagascar, 1997-98 157 A5.6 Distribution of Public Secondary Schools by Adequacy of Teaching Complement, Madagascar, 1997-98 158 A5.7 Correlates of the Allocation of Teachers Recruitedfor Public Primary Schools, Madagascar, 1997-2000 159 A5.8 Cost Functions by Type of Public Primary and Secondary School, Madagascar, 1997-98 160 A5.9 Fourth-Graders' Scores on Monitoring Learning Achievement (MLA) Tests, 1999 160 A5.10 Second- and Fifth-Graders' Year-End Test Scores in Mathematics and French in the PASEC Survey, 1995-98 161 A5.11 Sample Characteristics of Pupils and Schools in the 1997-98 PASEC Survey, Madagascar 161 A5.12 Probit Estimates of a Child's Probability of Attending Private School, Madagascar, 1997-98 163 A5.13 Correlates of Fifth-Graders' Combined Scores in Mathematics and French, Madagascar, 1997-98 164 A5.14 Correlates of Second- and Fifth-Graders' Combined Scores in Mathematics and French, Public Schools, Madagascar, 1997-98 165 A5.15 Logit Regression Estimates of the Relation between Pass Rates on National Examinations and Public Spending Per Student, Madagascar, 1997-98 166 A6.1 Organization of Vocational and Technical Education, Madagascar 167 A6.2 Number of Institutions and Number of Students by Field of Study, Public Sector Vocational and Technical Education, Madagascar, 1998-99 168 A7.1 Number of Students by Field in University Faculties, Ecoles Superieures, University Instituts, and ISTs, Madagascar, 1998-99 170 A8.1 Employment in the Formal Industrial Sector, Madagascar, 1994 and 1996 174 A8.2 Earning Functions, Madagascar, 1999 175 A8.3 Probability of Participation in Skills Upgrading, Madagascar, 1999 176 A8.4 Earnings Function, including Participation in Skills Upgrading, Formal Sector, Madagascar 177 Map Madagascar 181  Foreword W orld Bank support for educational devel- as they relate to poverty and the quality of life opment in Madagascar dates back to at enjoyed by the population. There is also agreement least 1989, when the first Bank-financed that because poverty reduction strategies cannot education project was approved. Two more educa- succeed without country ownership, it is appropri- tion projects (one of them still under implementa- ate for each country to take responsibility' for tion) followed, providing continuity in the Bank's preparing its own PRSP, relying on a process of par- engagement in policy development and implemen- ticipatory consultation with civil society to develop tation in the sector. In addition to project financing, a shared vision on the way forward. World Bank assistance also takes the form of ana- These two features in the new dynamic of devel- lytical activities and advice. This report is an exam- opment aid motivate the need to deepen our under- ple of such assistance. As with project financing, the standing of the current status of educational devel- ultimate objectives are to help the country expand opment and the obstacles that stand in the way of its educational opportunities, particularly at the progress, both to provide a factual basis for policy lower levels of the system, and to improve quality dialogue and to identify specific outcomes as goals throughout the system. Unlike in project financing, around which to build mutual accountabilities for however, the focus is on diagnosing major con- the use of national and donor resources. This report straints on educational development so as to help is an attempt by the World Bank to respond to that identify appropriate policy responses to those con- need. Although it does not take up the full range of straints, rather than on setting forth detailed speci- issues in the sector, its focus on cost and financing fications for project inputs and their design, imple- addresses some of the key concerns of policymakers mentation plans, and so on. and others engaged in the HIPC/PRSP process. Such analytical work and advice are becoming I am particularly pleased to note that the prepara- increasingly relevant in the context of debt relief tion of the report has involved substantial and close under the Heavily Indebted Poor Countries (HIPC) collaboration between the World Bank team and its Initiative and the related preparation of Poverty Malagasy counterparts. At the outset, a national Reduction Strategy Papers (PRSPs). These develop- team made up of representatives from the three edu- ments reflect a growing recognition among donors, cation ministries and from other government agen- including the World Bank, that although external cies was constituted. Interaction between this team financing through projects is important for sector and the Bank team was maintained through all development, the aggregate flow of resources, both stages of the work, from defining its scope to data external and domestic, has to be linked more closely collection and analysis, report writing and dissemi- to the achievement of tangible results, particularly nation, and policy dialogue. As the long list of names xi  Xii EDUCATION AND TRAINING IN MADAGASCAR in the Acknowledgments section attests, others also tion of this summary of the main findings in an participated in various ways-in particular, repre- easily accessible format will make a small but sentatives from the donor community, nongovern- nonetheless tangible contribution to Madagascar's m6ntal organizations, and members of civil society. efforts to shape an effective response to the chal- The extensive collaboration was critical in preparing lenge of poverty reduction. the report and is essential for strengthening country ownership of the analytical results and of their impli- Hafez Ghanem cations for policy development. Country Director for Comoros, Madagascar, Mauri- Wide dissemination and discussion of the diag- tius and Seychelles nostic results are first steps in the process of policy Africa Region development. It is my sincere hope that the publica- The World Bank  Abstract T he prospects for educational development education the challenges include: (a) raising the sur- T are excellent in Madagascar today in light of vival rates to the end of the primary cycle and the increasingly favorable policy environ- reducing grade repetition; (b) rationalizing teacher ment for the sector. During the first half of the allocation across schools and enhancing time use of 1990s, public spending on education relative to the teachers; and (c) improving student learning gross domestic product (GDP) declined more than through adequate provision of learning materials, 40 percent, coinciding with a five-fold rise in the but especially by strengthening management of the country's interest payment on extemal debt. As the' pedagogical process. At the secondary level, the debt service burden began to ease after 1995, public policy issues that warrant attention include: (a) spending on education began to recover, reaching expanding enrollments at a moderate pace in the the same level of spending in 2000 as at the start of lower secondary cycle, while focusing more on the 1990s, around 3.0 percent of GDP. As the gov- quality improvements than on expansion in the emments' efforts to reduce poverty is put into upper secondary cycle; and (b) taking advantage of place, the sector's claim on public spending can be scale economies and increasing the practice of expected to grow, a prospect made possible in part multi-subject teaching to manage the high cost of by earmarking a significant share of the resources service delivery. In the field of vocational and tech- freed up by debt relief for investments in education. nical education and training, the main issue is to A key challenge for policy makers and managers in rationalize the supply of training services so as to the sector is to transform public spending in the reduce their costs and align them more closely to sector into educational outcomes that would make labor market demand. Finally, in higher education, a significant contribution to poverty reduction. The the government faces two immediate issues: (a) analysis in this report identifies challenges at all rationalizing the system's structure in order to levels in the formal system. An important goal in improve quality and responsiveness to the demand the medium term is to universalize access to basic for skills in the labor market; and (b) improving education of reasonable quality while closely link- personnel management, particularly that relating to ing expansion of other levels and types of education compensation for over-time teaching and the use of and training to labor market demand. In primary administrative staff. xiii   Acknowledgments his study is the product of close collabora- * Ministere de l'Enseignement Technique et Forma- tion between the World Bank and the gov- tion Professionnelle (METFP): Daniel Robijaona, enment of Madagascar. Jee-Peng Tan led the Emmanuel Rakotovao, and Ernest Randriama- Bank team. Gerard Lassibille (Institut de Recherche nampisoa, with overall management support sur l'Economie de l'Education, France) was the from Hyacinthe Ralantonirina main team member. Lucia Navarro (Universidad de Malaga, Spain), Julia Lane (American University, * Ministere de l'Enseignement Superieur (M[NE- United States), and Dandan Chen (World Bank SUP): Ymg Vah Zafilahy, Chantal Rahalison, and Young Professionals Program) served as ad hoc Adolphe Randriantsoa, with overall management members providing specific inputs. Alain Mingat support from Lucien Rabenantoandro Rakotoma- was overall adviser, and Zafiris Tzannatos, Keith nantsoa and data assistance by Elisabeth Dama, Hinchcliffe, and Jaap Bregman were peer reviewers. Lala Rambeloson, and Daniel Razafimandimby The study benefited from overall management sup- port from Arvil Van Adams, Hafez Ghanem, and * Ministere charge du Budget et du Developpement Michael Sarris. des Provinces Autonomes (MBDPA), Fleurette The Malagasy technical working group consisted Nirinaharindrasana; Programme d'Appui Insti- of senior policy and technical staff from the three tutionel a la Gestion des Depenses Publiques/ education ministries, as well as representatives Comit6 Technique pour la Reforme du Service from other parts of government, as follows: Public/Secretariat Technique de l'Ajustement (PAIGEP/CTRSP/STA), Heri-Zo Ralambomahay; * Ministere de l'Enseignement Secondaire et de and Institut National de la Statistique (INSTAT)/ l'Education de Base (MINESEB): Albert Rafali- Madagascar DIAL/INSTAT/Orstom (MADIO), manana, Dominique Fofa, and Joseph Sylvain, Harivelo Rajemison and Eric Ramilison. with overall management support from Andre Juckely Boto and assistance with data com- Members of the technical working group con- pilation by Rabiary Andriaherisaina, Roger tributed background papers (see Madagascar Andrianaivoson, Georges Gniarck, Jean-Claude 2000a, 2000b, 2000c; Nirinaharindrasana 2000; Landry, Mariamo, Roland Rabeson, Honore Rajemison 2000; Ralambomahay 2000; Ramilison Rajerison, Jose Rakotonaivojaoniriana, Jean 2000) and provided critical help in assembling Pierre Rakotovao, Jacqueline Ralisiarisoa, and the databases used for the analyses presented in Jules Razafimandimby the report. xv  Xvi EDUCATION AND TRAINING IN MADAGASCAR The work has also benefited from inputs of various P. G. Le Boul and G. Thuaudet (Cooperation kinds from, among others, N. Andrianaivo, C. Andri- Francaise); B. Coquelin (United Nations Population analimanana, F. Chenakamkul, R. Kalidindi, C. Fund, UNFPA); V. Hubert (European Union); P. Lam- Kostopoulos, E. Murray, A. Onivololona, S. Pater- biliotte (Aide et Action); Y Polard (Association nostro, T. Rabenandrasana, E. Rabeson, J. Rabe- Francaise des Volontaires du Progres); J. Randrianan- tokotany, V. Rahelimanantsoa, M. Ranaivoarivelo, A. drasana (United Nations Educational, Cultural, and Ranarisoa, D. Randriamanampisoa, D. Ramiandrisoa, Scientific Organization, UNESCO); and L. Rama- R. Raoilison, J. Razafindravonona, S. Sosale, A. Velom- narivo (United Nations Development Programme, panahy, D. Viens, and J. Zalduendo. Nancy Levine UNDP). and Paule Bolo provided expert editorial advice, and The report was financed by the World Bank, with Angel Mattimore managed the production process. contributions from the governments of Madagascar, Helpful comments and information were received France, and Norway. from the following representatives of the donor com- A summary of this paper, in French and English, munity and nongovernmental organizations: 0. has been published in the World Bank's Africa Region Akpaka (United Nations Children's Fund, UNICEF); Working Paper Series.  Abbreviations and Acronyms AGENATE Agence Nationale d'Evaluation ARIF Association Regionale Inter-Professionnelle pour le Developpement de la Formation Professionnelle BEPC Brevet d'enseignement du premier cycle CEPE Certificat d'etudes primaires elementaires CERES Centre de Ressources des Personnels des Etablissements d'Enseignement Technique et Professionnel CFP Centre deformation professionnelle CISCO Circonscription Scolaire (school district) CNFTP Conseil National de la Formation Technique et Professionnelle CNTEMAD Centre National de Tele-Enseignement de Madagascar CONFEMEN Conference des Ministres de l'Education des Pays ayant le Francais en Partage DEA Dipl8me d'etudes approfondies DETFP Direction de l'Enseignement Technique et de la Formation Professionnelle DEUG Diplome d'enseignement universitaire g6neral DGGETFP Delegation Generale du Gouvemement a l'Enseignement Technique et a la Formation Professionnelle DIRESEB Direction de l'Enseignement Secondaire et de l'Education de Base ENI Ecole Nationale d'Informatique ENS Ecole normale superieure ENSET Ecole Normale Superieure pour l'Enseignement Technique ESP Ecole superieure polytechnique ESSA Ecole Superieure des Sciences Agronomiques FIFP Fonds d'Intervention a la Formation Professionnelle FRAM Fikambanan'ny Ray-Amandrenin'ny Mpianatra (school-based parent-teacher association) GIREFTPs Groupements Inter-Regionaux d'Etablissements de Formation Technique et Professionnelle HIPC Initiative Heavily Indebted Poor Countries Initiative IHSM Institut Halieutique des Sciences Marines INSTAT Institut National de la Statistique IOSTM Instifut d'Odontologie-Stomatologie Tropicale de Madagascar xvii  XViii EDUCATION AND TRAINING IN MADAGASCAR IST Institut superieur de technologie LTP Lycee technique professionnel MBDPA Minist&e charge du Budget et du Developpement des Provinces Autonomes METFP Ministere de l'Enseignement Technique et Formation Professionnelle MINESEB Ministere de l'Enseignement Secondaire et de l'Education de Base MINESUP Ministere de l'Enseignement Sup6rieur ONCE Observatoire National des Comp6tences pour l'Emploi ONETFOPP Office National de l'Enseignement Technique et de la Formation Professionnelle Priv6e PASEC Programme d'Analyse des Systemes Educatifs de la CONFEMEN PAT Personnel administrif et technique PDES Plan Directeur de l'Enseignement Superieur PDI Plan de Developpement Institutionnel PNAE Programme National pour l'Amelioration de l'Education PRAGAP Programme de Renforcement et d'Amelioration de la Gestion Administrative et Pedagogique TEFISO Teknica Fiofanana Sosialista VSNs Volontaires du service national Exchange rate (une 21, 2001): 1 Malagasy franc (FMG) = US$0.00015. 1 U.S. dollar = 6.657 FMG.  Executive Summary adagascar is a poor country; nearly three- the central issue is to advance that agenda through M quarters of its population has subsisted good stewardship of the available resources. This below the poverty line for at least the past report seeks to inform the discussion by document- two decades. Yet it is endowed with abundant and ing the current status of educational development unique natural resources that can be mobilized to in Madagascar and the key constraints on the per- raise the standard of living of its people. Several formance of the sector. recent developments have made the country's prospects for reducing poverty and improving the Scope of the Report and Its Audience well-being of its population more promising. Most important, the government has renewed its The report deals with the main education sectors- commitment to poverty reduction by articulating a primary, secondary, vocational/technical, and publicly discussed strategy containing explicit goals higher education-but its focus is necessarily lim- for macroeconomic management and sectoral devel- ited by the complexity of the subject. In particular, opment in key areas. Signs of economic recovery are nonformal education has been set aside for elabo- emerging. Gross domestic product (GDP) grew an ration by others, even though it is an important average 4.1 a year during 1997-99 (as against only element in any poverty reduction strategy. Simi- 1.9 percent a year during 1995-96) and is projected larly, neither early childhood education nor such to accelerate to 8 percent or more in the coming traditional educational issues as curriculum devel- years. The country's burden of external debt is being opment, teacher education, and pedagogical meth- lightened under the World Bank-International Mon- ods are addressed in this report. Rather, the etary Fund Heavily Indebted Poor Countries (HIPC) emphasis is on key aspects of education cost and Initiative, with projected savings in debt service finance and the link between spending choices and payment on the order of US$50 million annually for educational outcomes. at least the next 10 years. The report is primarily intended for education and Government spending on education has been other policymakers in Madagascar. Given the subject rising over the past five years, and the increase can matter and the context, it should also be of interest to be expected to continue, in part because a large the multilateral and bilateral donor community, to share of the savings from the HIPC Initiative is ear- nongovernmental organizations that are active in marked for the education sector. Expectations that education, and to teachers, students, parents, and education will make a significant contribution to other members of the general public who are directly the government's poverty reduction agenda are or indirectly concerned about the broader policy con- high. For policymakers and managers in the sector, text of their involvement in education. xix  XX EDUCATION AND TRAINING IN MADAGASCAR A key objective of the report is to contribute tion, must be calibrated, in qualitative as well as toward developing a widely shared vision for edu- quantitative terms, to the labor market's capacity to cational development in Madagascar based on a absorb educated labor. common understanding of the challenges involved Acceptance of this broad understanding of the and the role that each participant can play in realiz- direction for educational development provides a ing that vision. Open discussion of the report is an benchmark for assessing the various layers of important part of the process for creating a common management decisions needed to optimize the per- understanding. formance of the education system as a whole. Initial Such a process is, of course, only the first step conditions are important in this regard, for they toward policy development. Given its diagnostic delineate the system's current strengths and weak- focus and its selective scope, this report offers no nesses, as well as the tradeoffs and policy chal- policy prescriptions. Rather, it provides background lenges that must be managed to produce the best that can assist in the design of policies. Ideally, results within the inevitable budget constraints. policy development should be carried out with the Cross-country experience also imparts a useful per- participation of all the actors involved. spective by illustrating the scope of potentially viable choices and by bringing into focus the dis- The Economic Development Context tance that Madagascar must travel to match other countries' achievements in education. In an increas- Madagascar is a predominantly agricultural coun- ingly globalized world in which the basic skills and try; about 75 percent of the labor force is engaged in the technical know-how and competencies of the farming and other rural pursuits. In all, 85 percent whole population make the difference between of workers are in the informal sector. Accelerated competitive and lagging economies, the education economic growth is a necessary, but obviously not sector cannot afford to fall behind. sufficient, condition for overcoming the country's deep-seated poverty. Given the current structure of Strengths of Madagascar's Education System the economy, achievement of such growth in the coming decades will depend on two parallel devel- Madagascar starts out with some important assets opments: increased productivity in agriculture and in education. The country was one of the earliest to the informal economy (which will continue to recognize the importance of mass basic education, employ the bulk of the labor force), and expansion and over the years the government has extended of the modern, and typically more productive, the network of public primary schools to almost all sector. Investments in education can help facilitate of the country's some 11,000 neighborhoods, or both developments, but the kinds of interventions fokotanys. The social demand for education appears needed are likely to differ according to the sector of to be generally strong, as indicated by the high the education system. entry rates to first grade (averaging 81 percent for Worldwide historical experience suggests that no the country and 75 percent or more in all provinces country has managed to develop its economy with- except Toliara); the large private sector presence at out putting in place a system of mass basic educa- all levels of education; and the substantial spending tion capable of producing a literate and numerate by households throughout the system, which in population that can contribute to overall economic 1997 made up an estimated 35 percent of the coun- modernization. That experience also suggests that try's national investment in education. above the basic level of education, the production of Unlike the situation in many low-income coun- educated labor needs to be linked closely to the tries, girls in Madagascar are as likely to enroll as pace of job creation in the modern sector. Thus, boys; the gender gap widens (and then, not dra- while at the primary level the goal is clearly to matically) only in certain fields of technical educa- ensure that all children benefit from at least five tion and in higher education. Student learning in years of primary schooling, the development of primary education compares favorably with out- other levels in the system, especially higher educa- comes in other low-income countries, with Mala-  EXECUTIVE SUMMARY XXi gasy children scoring above the sample averages in 1 are high, only about a third of the entrants suc- language, mathematics, and life skills. cessfully complete the full five-year primary cycle. Policy development has already led to some Furthermore, repetition rates are uniformly high major achievements. In particular, in the decade just throughout the primary cycle, averaging about 31 past, Madagascar successfully managed the diffi- percent in 1998. These patterns have remained cult and often politically tense tradeoffs in resource unchanged for most of the past decade, and they allocation between higher education and the lower make Madagascar's system one of the least efficient levels-tradeoffs that many low-income countries in terms of student flow. continue to confront today. Higher education now absorbs 13 percent of total public spending on edu- Deteriorating skills levels. The latest data from cation, instead of the 30 percent or so that prevailed household surveys show a decline in the share of in the early 1990s. Furthermore, there is growing the labor force of those with no schooling and a rise recognition of the need for greater local participa- in the share of those with secondary schooling. The tion in school management, and institutional mech- improvement, however, is largely illusory, since the anisms have been put in place to formalize decen- same data also indicate that younger workers have tralization in education. far fewer years of schooling than their elders (4.2 Perhaps the greatest asset of the system is its years, on average, among those age 20-24, for potential for continued growth, given the highly example, compared with 5.2 years among those age favorable prospects for increased public investment 35-39). The clear deterioration in the educational in education. In such a context, the political econ- attainment of the labor force raises questions about omy of policy reform will be easier to manage, in the system's ability to meet the skills needs of a that some of the transitional costs of reforming the modernizing rural economy. system's operational arrangements can be absorbed through strategic allocation of the incremental Mismatch between jobs and education. At postpri- spending. An important factor is that overall levels mary levels, the student flow indicators are better, of teacher salaries in primary and secondary educa- although there is still room for improvement. The tion do not appear out of line with market condi- most important issue here concerns the output of tions and therefore do not present a major problem, graduates and their absorption into the labor beyond the need for appropriate incentives to moti- market. In the aggregate, the volume of postpri- vate improvements in service delivery. mary graduates seems consistent with the demand for labor. At the top end of the educational ladder, Challenges for the Next Decade however, there is evidence of substantial overpro- duction: jobs for highly educated labor are currently Policymakers in Madagascar face an exciting situa- expanding at perhaps 2,000 to 2,500 positions a tion in which the choices they'make today can truly year, but the system is producing about 3,700 higher help propel the system toward an efficient and equi- education graduates annually. One consequence of table path of expansion. This section reviews the this overproduction, which has been sustained over main problems and challenges to be taken into many years, is that graduates find themselves in account in developing education policies. jobs for which they are overqualified, and the adverse effects percolate down to school leavers Managing output from the education system with less education. Had enrollments in higher edu- cation not been brought under control during the Incomplete primary education. A major problem of 1990s, the problem would have been much worse. the education system is that many children leave The foregoing picture suggests that a major effort school without a complete primary education, is needed to boost the performance of the primary which is generally regarded as the minimum school system. At the same time, as cohort survival investment necessary for achieving permanent lit- rates improve, selection mechanisms need to be put eracy and numeracy. Although entry rates to grade in place to ensure that enrollments at postbasic  xxii EDUCATION AND TRAINING IN MADAGASCAR levels expand at a rate commensurate with the mary schooling. A combination of demand- and capacity of the labor market to absorb graduates supply-side interventions will be needed to amelio- into productive employment. The recent creation of rate the situation. the distance learning system in higher education, and the decision to admit private suppliers to the Policy options. As in most other settings, poverty sector, are important mechanisms for managing the and parental income play an important role on the excess demand for education at this level. demand side, but these factors are not directly within reach of the policy instruments currently Improving equity in education available to policymakers. Interventions on the supply side can have a direct impact and can also Income and education spending. Madagascar has work indirectly on demand by improving schools one of the most inequitable patterns of public spend- and making school attendance more attractive. ing on education among low-income countries. In Supply-side measures might include, for example, 1997 the bottom 20 percent of the population adjusting the school calendar to reduce the oppor- received only 7 percent of government spending on tunity cost of schooling and providing school feed- education, while the top 20 percent received 36 per- ing as a form of income transfer to poor families. cent, implying that the rich received more than five According to our estimates, ensuring that children times as much as the poor. Across members of each have access to a complete cycle of instruction and birth cohort, the best educated receive an estimated reducing repetition rates to about half their current 63 percent of the spending on the entire cohort. levels would, together, boost survival rates from 33 percent at present to more than 58 percent. Equity across education levels. Making the pattern of spending more equitable will require a restruc- Managing public spending on education turing of public spending per student in favor of primary education, as well as substantial progress Primary education a priority. Bringing about in ensuring that children persist to the end of the improvements in outcomes requires good steward- primary cycle. Public spending per student in ship of public spending on education. At the aggre- higher education is currently 1.57 times per capita gate level, current spending on education is modest, gross national product (GNP), compared with at 3.1 percent of GDP, compared with 4 percent in about 0.83 for low- and middle-income countries in most other low-income countries. A priority in Latin America, Asia, and the Middle East. Simula- overall public finance is to seek ways to increase the tions suggest that reducing spending on higher public resources available for education. Beyond education by about half while, at the same time, the aggregate increase, it is also important to ensure raising the cohort survival rate in primary educa- that resources are channeled to primary schooling. tion from 33 percent to about 50 percent would During the past decade, as total spending -on edu- bring Madagascar's system closer to the more equi- cation was squeezed under the weight of growing table patterns found in countries of those other external debt, the government, to its credit, man- developing regions. aged to redirect the shrinking resources for educa- tion toward basic education. The main beneficiary, Provincial differences. Across provinces and income however, turned out to be secondary rather than groups, there are substantial gaps both in the rate of primary education. In the coming years it will be entry to grade 1 and in survival patterns to the end important to ensure that increases in spending do in of the cycle. More needs to be done, particularly in fact reach all the way down to the primary level. Toliara but also in Fianarantsoa and Mahajanga, to raise entry rates. Survival rates are poor in all Shifting spending patterns toward teaching. A provinces, but the need to improve is especially closer examination of the pattern of aggregate pressing in Fianarantsoa and Toliara, where barely a spending suggests that more could be done to fifth of first-graders currently complete their pri- improve the distribution of public spendmg across  EXECUTIVE SUMMARY XXiii functions. Ideally, a substantial share of spending the pattern of inconsistency persists, although to a would go to support the core activities of teaching lesser extent. At all levels, the allocation favors and learning-the main business of education, after Antananarivo Province (and especially the city of all. Especially at postprimary levels of schooling, Antananarivo) and urban areas in general. too large a share appears to be allocated to adminis- trative and support services at the individual school Redeployment. The government did take steps to level. In lower and upper secondary schools, for improve the situation through two rounds of rede- example, only about half the government's total ployment that involved nearly 5,000 primary school spending goes for the core business of classroom teachers and 1,000 secondary school teachers. It is teaching, while school administration and support unclear, however, to what extent the inconsistency services at the school level absorb about 20 percent in staffing across schools was in fact rectified. For of spending. In vocational/technical education and example, in Antananarivo many schools have more higher education the share of teaching services teachers than would be consistent with systemwide drops to 34 and 29 percent, respectively, while the norms based on enrollments, and only 104 teachers share of administration rises to 22 and 28 percent. in that province were deployed in the two exercises. Furthermore, within higher education the current system of remuneration for ovrertime teaching Recruitment. Since 1997-98, there have been two results in nearly 80 percent of teaching being paid rounds of teacher recruitment. The placement of the on overtime. There also appears to be scope for new recruits has tended to favor understaffed rationalizing the current arrangements for student schools: the probability that such schools received a finance. Overall, although the benchmarks for iden- new teacher is estimated to be 13.5 percent (rising to tifying an ideal allocation of spending are lacking, 35.4 percent among those short by 3 teachers or current spending appears to be insufficiently more), compared with 8.9 percent for a school that focused on the system's core business and strategic was adequately staffed. In future rounds of teacher goals. Closer examination, either to justify the pat- recruitment, it will be important to strengthen the tern or to reform it, would be beneficial. focus on understaffed schools, particularly those in rural areas. Given the current pattern of staffing, Improving teacher allocation across schools any new teacher allocation to schools in the city of Antananarivo would need careful and convincing Beyond the systemwide problems in resource allo- justification. There are plans, too, to enhance the cation, there is also room for improving the alloca- incentives for rural posting. The design of these fion of spending among schools. Because teachers measures warrants thorough assessment to ensure represent the bulk of the public spending that that they are effective both in attracting teachers to schools receive, their deployment across schools rural schools and in retaining them, by offering suf- provides a good picture of the current patterns of ficient inducements to compensate for the disad- allocation of spending in the system. The results vantages of working in remote areas. suggest that there is substantial inconsistency in the allocation: primary schools with about 100 pupils, Taking advantage of scale economies and for example, may have as many as 7 teachers, but managing unit costs in small schools some of these schools receive no allocation at all and must rely on teachers who are financed entirely Madagascar's sparse population and the govern- through parental contributions to the school's ment's policy of ensuring that at least at the pri- parent-teacher association. On average, Antsiranana mary level, schools are physically close to the pop- and Toamasina appear to receive fewer primary ulation they serve combine to produce a significant school teachers than other provinces. The differ- number of small schools. Indeed, more than 30 per- ences across provinces, however, pale in compari- cent of the public primary schools enroll no more son with the inconsistency in allocation among than 75 pupils each, far below the minimum eco- schools within a province. In secondary education nomic size of about 250 pupils. In lower and upper  XXiV EDUCATION AND TRAINING IN MADAGASCAR secondary education, economies of scale are even of children in the public schools with less than 3 larger, but small schools continue to be a prominent hours of instruction a day. In public schools feature in the system. In vocational/technical edu- enrolling fewer than 50 pupils, for example, only cation and in higher education the small size of the one-third of the children receive full-day instruc- systems and fragmentation across multiple loca- tion, compared with more than two-thirds among tions combine to make it difficult to realize private schools of similar size. economies of scale in service delivery. In the centres deformation professionnelle (CFPs), for example, more Teacher specialization. At the secondary level, cost than half the unit cost of. service delivery is management in small schools would require a accounted for by fixed costs; the high share reflects closer look into teachers' workloads and subject the effect of the small enrollments in many schools. specialization. At both the lower and upper second- At these levels the problems are compounded by ary levels, many more public school teachers spe- the duplication of course offerings across institu- cialize in a single subject area than do their counter- tions. Thus, the entire system is both costly and parts in the private sector. As a result, their teaching inadequately diversified. loads are significantly lighter. All else being the same, the difference implies that unit costs in the Potential for school consolidation. The results public system are 28 percent higher than in private have two implications for cost management. The schools in lower secondary education and 45 per- first is that the spatial distribution of institutions at cent higher at the upper secondary level. Clearly, all levels has to be examined carefully to ensure that any policy for cost management would need to feasible options for school consolidation are not review this situation and evaluate the implications ignored. School consolidation can entail combining for teacher training and development and perhaps nearby schools that offer the same level of instruc- also for teacher incentives. tion or contiguous cycles. (Many of Madagascar's private schools, too, are small, but they invariably Enhancing student learning take advantage of economies of scope by combining under one roof different levels of schooling-for As noted earlier, Malagasy pupils perform well rel- example, primary and secondary, or lower and ative to children in other low-income countries. Yet upper secondary-whereas most public schools in absolute terms their scores are low, with children offer only a single cycle of instruction.) Any plan for correctly answering only half of the test items m consolidation will inevitably increase the cost of mathematics and French. Examination pass rates schooling for some students, who may have to remain low, as well. Comparison between public travel farther or even board away from home. All and private schools suggests that while social selec- such plans would therefore need to be combined tion does account in part for the superior perform- with arrangements for student finance to ensure ance of the private sector, a significant portion of the that deserving students from poor families are not difference can be traced to differences in the effec- adversely affected. Doing so will call for a more tiveness of school management. Looking within the strategic use of the resources currently allocated to public system itself, the results suggest that while student subsidies at all levels in the system. there is a positive relation between school inputs and student learning, there is great diversity across Multigrade instruction. The second set of implica- schools; schools with the same level of spending per tions pertains to cost management measures in student may achieve widely different outcomes. small schools. Multigrade teaching will be impor- What this implies is a lack of sufficient management tant at the primary level, and apparently a large oversight of the pedagogical process in schools and number of schools in Madagascar already practice classrooms. A challenge in this area is therefore to it. Its use should, in theory, not affect children's equip teachers with skills for more effective teach- access to full-day instruction, but the way it is ing while also putting in place a system of account- implemented in Madagascar leaves large numbers ability to motivate the application of those skills.  EXECUTIVE SUMMARY XXV Responding to the Challenges for the system to move forward, and the process of dialogue is needed to clarify and agree on roles. This report attempts to offer a broad diagnosis of the The central ministries will obviously have key main issues in the education sector in Madagascar. parts in leading and shaping the dialogue. In the As indicated at the outset, its coverage is selective, areas in which they exercise direct control, as in and the results must clearly be viewed within the staffing and resource allocation across schools and country's broader context of human capital develop- institutions, there is much scope for action to ment. This context includes policies pertaining to improve the management of the system. The nmn- two areas specifically excluded here, nonformal istries also recognize the importance of partnerships education and early childhood development. and local management of the educational process. Even more important, as a diagnostic product, the The concept of contrats-programmes has been report is only a first step toward policy development. launched at the primary and secondary education That process must begin with a broad discussion levels and may be extended elsewhere in the system. among interested parties concerning the assessment As with all innovations, including those that may be of the system's strengths, the challenges it faces, and motivated by the findings of this report, a system of the options for moving forward. The discussion is monitoring and evaluation is needed to provide needed not just because it has become fashionable to feedback to enable adjustment as plans are imple- talk about participatory approaches to poverty mented and their impact materializes. The process reduction and policy development but, rather, of learning by doing is a critical ingredient in policy because the business of education naturally involves reform and needs to be put in place even as discus- a myriad of actors, each with a specific contribution sions get under way on the future direction for edu- to make. Everyone must pull in the same direction cational development in Madagascar.   Demographic and Macroeconomic Context M adagascar's abundant and unique natural Madagascar's demographic and macroeconomic PA resources hold great potential for improv- context. It sets the stage for subsequent chapters, ing the livelihood of her population. (See which take up the second question in greater detail. Box 1.1 for basic country information.) Since the mid-1980s various economic reforms have been The Demographic Context launched to reverse the downward spiral of the economy over the past 40 years. Although the Madagascar's population in 1999 was estimated at reforms have yet to produce a pattern of consistent 14.6 million, up from 5.6 million at the country's and strong growth, signs of an upturn are begin- independence in 1960. As in most low-income coun- ning to emerge. Given the anemic economy, the tries, the rate of population increase has been and state's investmnent in education has been modest at continues to be high, averaging 2.4 percent a year in best, and even that meager amount suffered cut- the two decades following independence and accel- backs during periods of fiscal austerity, as exempli- erating to just under 3.0 percent a year in the 1990s fied by the pattern of spending in the 1990s. Hope- (Table 1.1). This rate of growth is projected to con- ful signs for the decade to come are the Malagasy tinue into the next decade and a half, in contrast to government's explicit commitment to poverty the projected growth rate of 2.2 percent a year for reduction (Madagascar 2000d) and the Heavily Sub-Saharan Africa as a whole and 1.3 percent a year Indebted Poor Countries (HIPC) Initiative recently in other low-income countries (World Bank 2000d). embraced by the international community. A funda- In most high-growth populations, young people mental principle underlying the initiative is that make up a sizable share of the population, and countries benefiting from it would conimit to using Madagascar is no exception. For example, children the resources freed up by debt reduction to expand of prirnary school age (6- to 10-year-olds) account and improve investments in the social services, for 13 percent of the total population. including education. The underlying assumption is An estimated 75 percent of the population lives that such investments are essential for achieving in rural areas, where agriculture is the main source sustainable and sustained poverty reduction.1 of livelihood and employment. About 40 percent of What, given these developments, is the scope for the urban population lives in Antananarivo, the increasing resources for education in Madagascar? capital city. Most people there are engaged in petty How can the resources-those currently allocated trade in the informal sector.2 and any extra that might become available-be As Table 1.1 shows, poverty is rampant in Mada- used to maximum effect? This chapter focuses on gascar, and the situation has deteriorated over the the first question by examining relevant features of past few decades. Whereas an estimated two-fifths I  2 EDUCATION AND TRAINING IN MADAGASCAR Box 1.1 Madagascar at a Glance Geography. Madagascar is an island nation located in the Indian Ocean off the southeast coast of the African continent With a surface area of 592,000 square kilometers, it is the fourth-largest island in the world, after Greenland, New Guinea, and Borneo. The country has diverse soils and climates and many species of plants and animals found nowhere else in the world Parts of Madagascar are vulnerable to drought and cyclones, and the environment is under increasing stress from massive deforestation as trees are cut for fuelwood People. Most of the Malagasy people trace their ongin to ancestors from Afnca and Indonesia The two main ethnic groups, the Merina and the Betsileo, representing 26 and 12 percent, respectively, of the population, are mostly of direct Indonesian ongin and live in the central high- lands The other major ethnic groups live mainly in the coastal areas They include the Betsimisaraka in the east (about 15 percent of the pop- ulation), the Tsimehety in the north (7 percent), and the Antandroy in the south (5 percent) Economy. With a per capita gross domestic product (GDP) of US$235 in 1999, Madagascar is one of the poorest countries in the world Agri- culture is the dominant economic activity, employing three-quarters of the work force and yielding about one-third of the country's GDP The creation of export-processing zones in 1990-part of a policy of attracting foreign investment by offering favorable tax breaks to qualified com- panies-is helping to expand the country's modest manufacturing sector in such areas as textiles, clothing, essential oils and essences, wood- crafts, sisal, and fisheries Tourism and mining are potential high-growth sectors Political system. After decades of French rule, Madagascar became fully independent in June 1960. The country is now organized as a republic with a pluralist, democratic constitution that was approved by referendum in 1992 Political power is wielded at three levels the pres- idency, the national assembly, and local and regional authorities Both the president and the members of the national assembly are elected by universal suffrage Constitutional changes approved in March 1998 are expected to strengthen the presidency and expand the role of local authorities. The political landscape is complex, with shifting alliances among the many parties. Relations with the World Bank. Madagascar became a member of the World Bank in 1963 and received its first loan in 1966. Since then, the Bank has provided policy advice, technical assistance, and project and program financing to support the country's economic development Total cumulative lending to date amounts to more than US$1 6 billion. Of this, completed projects account for US$650 million, US$550 million is for projects still being implemented; and US$400 million is for economic adjustment operations Source World Bank (1999), Economist Inteiligence Unit (2000) Table 1.1 Population and Incidence of Poverty, Madagascar, 1962-97 Annual percentage changeb 1962 1975a 1980 1993a 1997 1960-80 1990-99 POPULATION Total size (millions) 5.7 7 6 8 7 12 2 13.7 2.4 2 9 Of primary school age (6-10) - - - 1 6 1 8 - 2.9 Share in rural areas (percent) - 84 - 77 77 - SHARE OF POPULATION LIVING IN POVERTYC All Madagascar 43 - 47 70 75 Urban areas - - 65 68 Rural areas - - 72 78 - Not avaliable Not applicable a Population census years The dates shown in the other columns coincide with the dates for which data on the incidence of poverty are available b Growth rates are computed based on the population sizes reported in IMF (2000) for the end years of the indicated penods, growth of the school-age populabon is for the penod 1993-99 c Defined as those unable to afford a food basket yielding the equivalent of 2,100 calories per day and to satisfy minimum nonfood needs Source For overall population size, IMF (2000), for size of the pnmary-school-age population, Minist6re de l'Enseignement Secondaire et de l'Education de Base (MINESEB) data files, for share of populabon living in poverty, Madagascar (2000d), for share of population in rural areas, Madagascar (1997c, 2000d)  DEMOGRAPHIC AND MACROECONOMIC CONTEXT 3 of the population lived below the poverty line in Table 1.2 1962, the proportion had risen to three-quarters by Economic Performance, Madagascar and 1997. With such a high overall rate of poverty, the Selected Countries, 1970-99 share of poor people is high in both rural and Index of per capita GDP Annual GDP growth urban areas. (Madagascar 1.0) rate, 1970-99 Despite the country's poverty, most of Madagas- Country 1970 1999 (percent per year) car's health indicators show up favorably against Madagascar 1 0 1 0 those of other Sub-Saharan African countries. Aver- -adagascar_1_0_1_0_0.5 age life expectancy at birth was about 58 years in (Per capita GDP, 1998, as against an average 52 years for Sub-Saha- 1997 US$) (473) (227) ran Africa, and the under-five mortality rate was Benin 0 9 1 7 3 3 about 162 deaths per 1,000 live births in 1997, com- pared with an average for the region of 147 deaths Burkina Faso 0 5 1 1 36 (World Bank 2000c). The World Bank estimates, Kenya 0.6 1.4 45 however, that half of all children under age 5 are malnourished (World Bank 2000c). Lesotho 0.6 2.2 55 The prevalence of HIV/AIDS is still relatively low, Zimbabwe 1 3 2 9 4 6 with an estimated infection rate of about 0.13 percent in the population age 15 and above (World Bank India 06 15 4.4 2000c). But there is no cause for complacency: expe- Indonesia 0.5 4.4 7 3 rience in nearby countries such as Tanzania, Uganda, and Zambia shows how easily and rapidly the situa- Pakistan 1.0 2.0 5.3 tion can get out of control in the absence of early pre- Source World Bank (1999), updated with data for 1999 from the World Bank Statistical ventive action. The government recognizes the risks Informabon Management and Analysis (SIMA) database and has taken some important precautions against an outbreak, including the establishment in 1988 of a Niger, and the Democratic Republic of Congo (World national program to control sexually transmitted Bank 2000a). Not surprisingly, Madagascar now lags diseases and HIV/AIDS; the creation of a sero-sur- far behind countries that had the same or a smaller per veillance system to gather data and monitor the situ- capita GDP in the 1970s (Table 1.2). ation; and the use of high-level publicity to increase Given Madagascar's many endowments and nat- general awareness of the disease and encourage ural advantages and the fact that demographic pres- changes in personal behavior to minimize its spread. sures are no greater than those experienced by other low-income countries, the poor performance of the Economic Growth and Current Conditions economy must be laid squarely at the door of failed economic policies (World Bank 1999; lIMF 2000). The dismal performance of the economy is the main From the 1970s until the mid-1980s, the country pur- reason for Madagascar's widespread poverty. sued a development strategy based on economic Between 1970 and 1999 gross domestic product (GDP) nationalism, self-sufficiency, and pervasive state grew at only 0.5 percent a year (Table 1.2), while the involvement in and control of economic production. population increased at about 3 percent a year. As a Excessive government regulation not only distorted result, per capita GDP (in 1997 prices) fell from incentives and production but also discouraged the US$473 in 1970 to US$227 in 1999, a decline of more domestic and foreign investment needed to boost than 50 percent. The poor performance of the Mala- economic growth. The mid-1980s marked the begin- gasy economy comes into particularly sharp relief in ning of new policies directed at fiscal and adminis- cross-country comparison: among the 33 African trative reform, economic and financial liberalization, countries for which relevant data exist, Madagascar and disengagement of the state from direct economic ranks near the bottom in growth of per capita GDP production. Progress has been halting and was inter- between 1970 and 1999, ahead of only Sierra Leone, rupted by political instability during 1991-93.  4 EDUCATION AND TRAINING IN MADAGASCAR Efforts in these directions have been renewed the decade. Total revenues, induding such grants, since the mid-1990s, and signs of a turnaround are stood at 16.4 percent of GDP at the start of the 1990s but just beginning to emerge. For the first time in many had fallen to 15.1 percent by 1999. The 1999 figure rep- years, real GDP growth has outstripped population resents a recovery since mid-decade, when the share growth, averaging 4.1 percent a year between 1997 was barely over 11 percent. Current revenues (exclud- and 1999, compared with only 1.9 percent a year in ing grants) averaged less than 10 percent of GDP over 1995-96. As a result, per capita income grew at the decade, with the lowest percentages again at mid- about 1 percent a year during the period. The decade. Since 1997, the trend has showni a strong upswing is good news indeed, but it represents upturn, reflecting increases in income tax revenues gen- only the beginning of a trend that obviously would erated by renewed economic growth and improved tax have to be sustained to produce tangible reductions collection. To enhance the government's capacity to in poverty. The World Bank estimates that bringing finance the country's many unmet development needs, the poverty rate down to 32 percent would be pos- the World Bank has suggested that by the late 2010s it sible only if GDP grew by about 7.5 percent a year would be desirable to raise current revenues (i.e., over the next two decades (World Bank 1999).3 income from taxes and other domestic sources) to around 15 percent of GDP (World Bank 1999). Overall Government Finances Turning to the expenditure side, Table 1.4 shows that total expenditures consistently exceeded total Better management of government finances is an revenues, often by exceptionally large margins, for important part of efforts to renew economic growth. most of the decade 1990-99. In 1994, for example, Tables 1.3 and 1.4 document changes in various indica- total expenditures were 1.74 times the government's tors in this area over the past 10 years. The government total receipts. Current spending-that is, total expen- depends to a significant extent on grants from external diture minus capital expenditure and debt service donors for its revenues, with this source representing payment-represented between 53 and 65 percent of one-quarter or more of total revenues for most years of total spending throughout the 1990s and averaged Table 1.3 Overall Government Revenue, Madagascar, 1990-99 Total revenue Current revenue Billions Grants as of current As share of Index As share of Index share of total Year Malagasy francs GDP (percent) (1990 = 100) GDP (percent) (1990 = 100) revenue (percent) 1990 752 8 16.4 100 11.8 100 26.9 1991 533.2 10 9 66 8 5 72 19 5 1992 752.6 13.5 82 9.8 83 26.0 1993 864.5 13.4 82 9.7 82 26 1 1994 1,035.9 11.3 69 8 3 70 26 5 1995 1,541.8 11.4 70 8 5 72 25.4 1996 2,090.7 12.9 79 8.7 73 32.7 1997 2,703.4 15 0 92 9 7 82 35 4 1998 2,864.3 14 1 86 10 6 90 24.4 1999 3,522.0 151 92 11.4 96 24.5 Source Computed using data on government finance from the World Bank Afnca Live Database and GDP data from the World Bank SIMA database, see also Appendix table Al 1  DEMOGRAPHIC AND MACROECONOMIC CONTEXT 5 Table 1.4 Overall Government Expenditure, Madagascar, 1990-99 Interest on all Interest on Total expenditure Current expenditurea public debt extemal debt As share of Billions As share of GDP As ratio As share As share total debt of current of total of total As share of total As share interest As share Malagasy Index govemment expenditure of GDP expenditure of GDP payment of GDP Year francs Percent (1990 = 100) revenue (percent) (percent) (percent) (percent) (percent) (percent) 1990 784 1 17.0 100 1.04 53.4 9.1 8.9 1.5 68.3 1 0 1991 805 6 16.4 96 1.51 59.6 9 8 12.3 2.0 56 0 1.1 1992 1,120.9 20.0 118 1.49 58.8 11.8 17.1 3.4 40 3 1.4 1993 1,328.4 20 6 121 1 54 58.9 12.1 19.4 4.0 78 7 3.2 1994 1,803.6 19.8 116 1.74 64 8 12.8 27 3 5.4 94.0 5 1 1995 2,373.8 17 6 103 1.54 64.2 11.3 29.0 5.1 90 8 4.6 1996 2,883 4 178 104 1 38 59.1 105 26.3 4.7 82.8 3.9 1997 3,137.1 17.4 102 1.16 62.5 10.9 17.5 3.0 91.7 2.8 1998 3,819.2 18.8 110 1.33 56.2 10.5 14.5 2.7 84.2 2.3 1999 3,624.7 15.5 91' 1.03 57.9 9.0 13 6 2.1 65 9 1.4 a Total expenditure net of capital expenditure and debt service payment Source Computed using data on govemment finance from the Wordd Bank Afnrca Live Database and GDP data from the Wodd Bank SIMA database, see also Appendix table Al 1 around 10.8 percent of GDP over the decade. Current in the 2000 budget (see Table 1.5). Spending on edu- expenditures and current revenues match each other cation as a share of total public spending followed a more closely than do total expenditures and total similar U-shaped pattern, with highs of more than 17 revenues. The table highlights the role of debt inter- percent at the beginning and end of the decade and est as a source of fiscal distress during the 1990s. In a trough of 10-11 percent in the middle years. The the most difficult years, at mid-decade, total debt share averaged around 14 percent over the decade interest payment accounted for nearly 30 percent of and did not rise above 20 percent in any year. total government expenditure and for more than 5 As expected, most of the public spending on edu- percent of GDP. Most of the interest payment was to cation is accounted for by current spending, in external creditors. The situation has improved con- which teacher salaries are the main item. The share siderably since 1997; by 1999 total interest payments of current spending rose to a high of 96 percent in had declined to 2.1 percent of GDP. 1995, just when overall spending on education was squeezed, and fell to a low of 71 percent in 2000, Overall Public Spending on Education when spending was rising again. The trend reflects the unsurprising fact that it has been easier to Total public spending on education averaged 2.5 expand and contract capital spending than current percent of GDP over the 1990s, but there was sub- spending in response to overall fiscal conditions. stantial variation around that average. The share fell Thus, while total spending on education as a per- from a high of around 3.0 percent or more at the centage of GDP bounced back in 2000 to the levels beginning of the decade to a low of 1.8 percent in at the beginning of the 1990s, current spending as a mid-decade and then recovered to over 3.0 percent percentage of GDP was still only 86 percent as high  6 EDUCATION AND TRAINING N MADAGASCAR Table 1.5 Overall Public Spending on Education, Madagascar, 1990-2000 Current and capital spending on educations Current spending on education As share of total As share of total As share of total public current spending net of As share of GOP public spending spending on education As share of GDP debt interest payment Index Index Index Index Index Year Percent (1990 = 100) Percent (1990 = 100) Percent (1990 = 100) Percent (1990 = 100) Percent (1990 = 100) 1990 3.0 100 17.6 100 85.9 100 2 6 100 28.3 100 1991 32 105 193 110 84.9 99 2.7 104 275 97 1992 2.8 93 13.9 79 85.0 99 2.4 92 20.2 71 1993 2 6 87 12.7 72 90.2 105 2.4 91 19.4 69 1994 2.1 71 10.8 62 91.5 107 2 0 76 15.3 54 1995 1.8 58 10.0 57 96.2 112 1 7 65 14.9 53 1996 2.0 66 11.1 63 85.9 100 1.7 66 161 57 1997 2.3 78 13 5 77 81.4 95 1.9 74 17.6 62 1998 2 6 85 13.7 78 81.8 95 2 1 81 19.9 70 1999 2 8 92 17 8 101 77.0 90 2.1 83 23 7 84 2000 3.1 103 - - 71.1 83 2.2 86 - - -Not available a Indudes current and capital spending by the Minist6re de rEnseignement Secondaire et de l'Education de Base (MINESEB), the Miistere de l'Enseignement Technique et Formation Professionnelle (METFP), and the Ministere de l'Enseignement Sup6neur (MINESUP), as well as the salanes of former nabonal service personnel (volontaires du servce naftonal, or VSNs) paid unbl 1998 by local govemments, the salanes of physical educabon teachers paid by the MinistAre de la Jeunesse et du Sport (MJS), and central govemment transfers to communes for schools Source Computed using data on public spending on education from Madagascar (various years, a) for 1990-92, Madagascar (vanous years, b) for 1993-96, Madagascar(vanous years, c) for 1997-2000, and GDP data from the Wodd Bank SIMA database, for absolute amounts, see also Appendix table At 2 in 2000 as in 1990. Because the functioning of the Figure 1.1 education system depends on adequate levels of External Debt Service and Public Spending current spending, these trends suggest that the on Education, Madagascar, 1990-99 system's recovery from the fiscal austerity of the 30 1990s is probably still incomplete. The pattem of public spending on education in F 25 Madagascar during the 1990s exemplifies the _ - / _ adverse impact that a heavy burden of external debt c can have on the education sector. Figure 1.1 illus- ; IS - ' trates how as interest on the country's external debt ,"'