DIRECTIONS IN DEVELOPMENT 161S5S Counting the Full Cost Parental and Community Financing ofEducation in East Asia MARK BRAY r~~~~~~~~~~~~~~~~~~~~ A collaborative report by the SWorld Bank and the Uniited Nationis Children's Fund DIRECTIONS IN DEVELOPMENT Counting the Full Cost Parental and Community Financing of Education in East Asia Mark Bray The World Bank Washington, D.C. © 1996 The International Bank for Reconstruction and Development/ THE WORLD BANK 1818 H Street, N.W. Washington, D.C. 20433 All rights reserved Manufactured in the United States of America First printing November 1996 The findings, interpretations, and conclusions expressed in this study are entirely those of the author and should not be attributed in any manner to the World Bank or the United Nations Children's Fund, to their affiliated organizations, or to mem- bers of their Board of Executive Directors or the countries they represent. Cover photograph by Curt Camemark, The World Bank. Mark Bray, a professor at the Comparative Research Centre of the University of Hong Kong, is author of several books and many articles focusing on aspects of the economics and financing of education. His books include (with Steve Packer) Eduication in Small States: Concepts, Challenges, and Strategies (Pergamon Press) and (editor, with Kevin Lillis) Commuinity Financing of Eduication: Issutes and Policy Implications in Less Developed Countries (Pergamon Press). Library of Congress Cataloging-in-Piublication Data Bray, Mark. 1952- Counting the full cost : parental and community financing of education in East Asia / Mark Bray. p. cm. - (Directions in development) "A collaborative report by the World Bank and the United Nations Children's Fund." Includes bibliographical references (p. 78). ISBN 0-8213-3827-7 1. Education-Asia--Costs-Statistics. 2. Parents-Asia-Finance, Personal-Statistics. 3. Community and school-Asia, Southeastern- Statistics. I. Title. II. Series: Directions in development (Washington, D.C.) LB2829.3.A78B73 1996 338.4'337'095021-dc21 96-39150 CIP Contents Foreword v Preface vii Acknowledgments ix Issues 1 Human Rights, Private Schooling, and Fees 1 Political Forces and Budget Constraints 3 Parental and Community Willingness to Pay 4 Policy Dilemmas 6 Profiles of Nine East Asian Countries 8 Private Education 10 Costs to Parents and Families in the Public System 17 Community Resources in the Public System 29 Proportions of Total Costs Met by Families and Communities 32 Burden on Household Budgets 38 Policy Implications 42 Privatization in Education Systems 42 Effect on Enrollment 45 Official Policies on Fees 47 Household Costs and Elasticity of Demand 51 Socioeconomic, Regional, and Urban-Rural Imbalances 54 Conclusion 56 Country Data 58 Notes 75 References 78 iii Foreword This book is the joint product of the United Nations Children's Fund (UNICEF) and the World Bank. It demonstrates the commitment of our two institutions to work together and in broad partnership with others to meet the challenge of Education for All, which was enunciat- ed at the 1990 World Conference on Education for All in Jomtien and reaffirmed at the 1996 Global Education for All Forum in Amman. In Jomtien 155 countries pledged to take the necessary steps to pro- vide primary education for all children by the end of the decade. Yet by 1996, 130 million children between the ages of 6 and 11 and 273 mil- lion children between the ages of 12 and 17 remained out of school. In Amman the Education for All commitments were reaffirmed, the resources required and their effective use were considered, and the international community was encouraged to rethink its resource pri- orities in order to meet the goal. This study identifies the level of assistance poor households may need to achieve education for all. It shows that households in nine East Asian countries already contribute a significant share of the total cost of public education. In some of these countries household contribu- tions to education are greater than government contributions; in all nine countries they represent a significant share of household income. The burden borne by the poorest households is much greater than that borne by the wealthiest households. Unfortunately, the policies and actions of many of the governments and international agencies involved in the support of education have shown little awareness of the magnitude of this burden. We hope that this book will serve to heighten awareness and inspire both further research and appropriate concrete action. David de Ferranti The World Bank Manzoor Ahmned UNICEF v Preface Many studies that purport to analyze the costs of education in fact pre- sent only the costs to governments. These studies either completely ignore the costs to parents and communities or provide only brief recognition of these costs but no figures or further discussion. Such neglect is unfortunate because in many societies household inputs are substantial. In many settings such inputs, even in the public education system, meet more than 20 percent of total costs (Cumming and others 1995; Togo 1994; UNICEF 1996; World Bank 1994a), and in two coun- tries covered by this study, they exceed 50 percent. The implications of these inputs are far-reaching. Such resources can increase both the quantity and quality of education and lead to fur- ther economic and social development through investment in human capital. Parental and community inputs may also strengthen the sense of partnership in the educational enterprise. Parental and community financing can also raise problems, how- ever. One difficulty arises from the fact that even when total provision is large, patterns are rarely balanced or uniform. As a result parental and community financing may exacerbate socioeconomic and region- al inequalities. Demands may also become a heavy burden at the indi- vidual and family levels, and situations in which this burden weighs down the poorest are particularly problematic. In some settings, more- over, awareness that such demands will be imposed discourages par- ents from enrolling their children in school, in the first place, and in families of children who do currently attend school, the costs may con- tribute to early dropout. The demands on households may thus obstruct the goal of universal basic education. As burdens grow in some countries, enrollment rates have been going down rather than up (Pepper 1995, 18.14; Watkins 1995, 79; Wu 1994, 15). This study focuses geographically on nine countries in East Asia- Cambodia, China, Indonesia, Lao People's Democratic Republic, Mongolia, Myanmar, the Philippines, Thailand, and Vietnam. These countries share some common features but are also diverse in several important ways. In some respects the contrasts are as instructive as the similarities because they help to highlight particular characteristics and the implications of specific contexts and policies. vii viii COUNTING THE FULL COST The discussion is chiefly concerned with the primary and secondary levels. The main focus is on formal education, but it also includes out- of-school supplementary tutoring, which in some societies is a major expenditure. Much of the focus is on financing in cash, but the study recognizes that demands may also be for materials and labor and that the full costs of education include the opportunity costs of forgone earnings. The work is based on information from four main sources. The first is the existing literature, not only on the nine countries on which the discussion focuses but also on other countries with comparable eco- nomic and social features. The second is data from questionnaires sent to the United Nations Children's Fund (UNICEF) officers responsible for education projects in each of the nine countries. In several cases these questionnaires were completed in conjunction with government officials. Third, the document is based on the author's experience and field work. He has worked in various capacities in each of the nine countries and visited Cambodia and Lao People's Democratic Republic for the specific purpose of this project.' Finally, the report draws extensively on information available in the World Bank on all the countries of the region. Acknowledgments The idea for this study originated with Sheldon Shaeffer, regional edu- cation adviser at UNICEF's Regional Office for Education in Asia and the Pacific in Bangkok. He deserves thanks not only for conceiving the project but also for helping it along at each stage. Particular thanks are also due to Marlaine Lockheed, principal education specialist in the Human Development Department of the World Bank. Much of the study was conducted while Mark Bray was a visiting research fellow at the World Bank in 1995. Marlaine Lockheed initiated the invitation for the fellowship and did everything to make the experience worth- while, enjoyable, and productive. Although the study was launched before the commencement of the fellowship, in a very real sense, it is also a product of that fellowship. Thanks go as well to the UNICEF staff and other individuals who facilitated the field work in Cambodia and Lao People's Democratic Republic, in particular Anne Dykstra and Ng Shui Meng. UNICEF staff in other country offices responded to questionnaires and pro- vided supplementary materials. A large number of individuals com- mented on drafts and exchanged ideas with the author; among them are Halsey Beemer, Nicholas Burnett, Shaikh Hossain, Emmanuel Jiminez, Ruth Kagia, Elizabeth King, Hena Mukherjee, Paud Murphy, Harry Patrinos, George Psacharopoulos, William Rees, Ompom Regel, Eluned Roberts-Schweitzer, Haneen Sayed, Antoine Schwartz, Jee- Peng Tan, Mun Tsang, Mathew Verghis, John Weidman, and Loraine West. American Writing Corporation edited and laid out the text. ix Issues An appropriate starting point is the concept of education as a human right, a notion generally accepted throughout the world. It was enshrined in the 1948 United Nations Declaration of Human Rights and has been reiterated in many international and national documents since that time. Among the most recent was the 1990 Jomtien Declaration of the World Conference on Education for All (WCEFA 1990b, 2). Human Rights, Private Schooling, and Fees If education is accepted as a human right, one must ask who should provide it to meet that right. On this point the philosophies of indi- viduals, governments, and agencies have varied. The East Asian coun- tries examined here include some that have always had capitalist regimes that permitted private education alongside government schooling. Some private institutions in these countries have been oper- ated by commercial entrepreneurs, but others have been run by reli- gious and other not-for-profit organizations. In contrast are countries with socialist regimes in which education has been seen as the exclu- sive right and responsibility of the state. In these countries, especially at the beginning of the socialist eras, private schools have been out- lawed. In recent years, however, the pattern in most socialist countries has changed. East Asia includes three socialist countries-China, Lao People's Democratic Republic, and Vietnam-that have moved toward market economies and have governments that now permit or actively encourage private education. These moves are part of a glob- al trend (Bray 1996c; Cummings and Riddell 1994; James 1991, 1994). Commercial forces are also being given stronger recognition and sup- port in countries that had long had capitalist economies. This is not only within the private sector but also within the government sector, where school authorities are increasingly expected to operate with market principles in mind (Lingard, Knight, and Porter 1993; Pring 1988). Allied to this shift in thinking about the role of market forces in edu- cation is an evolution of opinion on who should pay for education. The I 2 COUNTING THE FULL COST 1948 United Nations Declaration of Human Rights was unequivocal in the view that the state should bear most or all costs, and Article 26 declares: "Education shall be free, at least in the elementary and fun- damental stages." Similarly, Principle 7 of the 1959 Declaration of the Rights of the Child states: "The child shall be entitled to receive edu- cation, which shall be free and compulsory, at least in the elementary stages." Article 13 of the 1966 International Covenant on Economic, Social, and Cultural Rights states that: (a) Primary education shall be compulsory and available free to all. (b) Secondary education in its different forms. ..shall be made gen- erally available and accessible to all by every appropriate means, and in particular by the progressive introduction of free education. (c) Higher education shall be made equally accessible to all, on the basis of capacity, by every appropriate means, and in particular by the progressive introduction of free education. In a similar vein Article 28 of the 1989 Convention on the Rights of the Child indicated that the signatory states would: (a) Make primary education compulsory and available free to all; (b) Encourage the development of different forms of secondary education.. .make them available and accessible to every child, and take appropriate measures such as the introduction of free education and offering of financial assistance in case of need; (c) Make higher education accessible to all on the basis of capacity by every appropriate means. Since these declarations were made, however, the appropriateness of parts of their message has been increasingly questioned. Education is still generally accepted as a public good that can benefit societies as a whole as well as individuals (Levin 1987; Solomon and Fagnano 1995), but particularly in higher education a general worldwide swing of opinion now favors fees, supported as necessary by loans and other mechanisms to protect the poor (Albrecht and Ziderman 1995; Psacharopoulos and Woodhall 1985; Tilak 1994a). Reflecting this swing, the fact that point (c) of the 1989 Convention on the Rights of the Child did not completely echo point (c) of the 1966 International Covenant on Economic, Social, and Cultural Rights is significant. Support for cost recovery at lower levels has been more muted, but the issues have been actively discussed at that level too (Bray 1987; Jimenez 1994; Lewin with Berstecher 1989; Thobani 1984). ISSUES 3 Arguments in favor of fees at the tertiary level are partly based on the substantial private benefits that usually accrue to tertiary gradu- ates and on the fact that tertiary education has high unit costs that can- not easily be borne solely by governments. Arguments favoring fees are also based on concern for equity. Among tertiary enrollments stu- dents from rich families always form a much larger proportion than students from poor families, and it is widely considered unreasonable to subsidize the rich families when doing so will reduce the resources available for the poor. It is of course recognized that students from poor families also study in tertiary institutions, and that it is desirable for proportions of such students to be increased. However, because ter- tiary graduates usually receive greatly enhanced lifetime earnings, even the poor can finance their studies through loans that can later be repaid. The cost and equity arguments are to some extent valid in sec- ondary education and can also be used to justify cost recovery at that level. But unit costs to government are lower at the secondary level, and enrollment across socioeconomic groups is usually more balanced. These particular arguments are therefore less powerful at the sec- ondary level. At the primary level the arguments are even more difficult to sus- tain. Unit costs are generally lowest at this level. In countries with low enrollment rates school populations are likely to contain more children from middle-income and rich groups than from poor groups, but because all governments claim to wish to attract the poor to primary school, they are unwilling to discourage enrollments by imposing fees. Moreover, in systems in which enrollments approach universality, fee- free education is likely to give greater government resources to poor families than to rich ones, simply because poor families are likely to have larger numbers of children than are rich families. Political Forces and Budget Constraints The above-mentioned arguments would seem to favor charging fees for tertiary education to permit redistribution of resources to the poor. One way to serve the poor better, it has been argued, is to place greater proportions of government resources in the basic education sector (Colclough with Lewin 1993; Penrose 1993; Tan and Mingat 1992; World Bank 1995d). However, political forces may not easily permit governments to demand substantial cost recovery at the tertiary level, especially if university students are familiar with a long legacy of free provision, come from politically well-connected families, and are 4 COUNTING THE FULL COST prepared to be vocal in their protests. Moreover, even when govern- ments do recover some of the costs of tertiary education, it is not always easy to allocate the resources to primary and secondary educa- tion rather than to other sectors of public expenditure. And even in sit- uations in which most or all of the income generated from tertiary fees can be allocated to lower levels of education, the sheer numbers of pri- mary and secondary school students mean that the little money that is made available has to be spread thinly. Related to these matters is the acute budget austerity that has hit many governments during the 1980s and 1990s. The economies of many low-income countries have suffered severely from natural dis- asters, structural weaknesses, and external debt, and stagnation or decline in government revenues has been accompanied by continued rapid population growth (Graham-Brown 1991; UNESCO 1993). The last decade has brought particularly severe economic difficulties to the former and remaining socialist states, but pressures have also been heavy in many capitalist societies. Thus, although most governments would like to be able to provide fee-free primary and secondary edu- cation-and some even enshrine this in their constitutions-the prac- tical realities of making ends meet require at least some contribution from parents and communities. The fact that the 1990 World Declaration on Education for All did not include a statement that schooling should be free of charge reflects these realities. Instead the Final Report of the Conference (WCEFA 1990a, 31) included open discussion of fees, and Article 7 of the Declaration itself (WCEFA 1990b, 7) stressed the importance of partnerships: National, regional, and local educational authorities have a unique obligation to provide basic education for all, but they can- not be expected to supply every human, financial, or organiza- tional requirement for this task. New and revitalized partnerships at all levels will be necessary... [including] partnerships between government and nongovernmental organizations, the private sec- tor, local communities, religious groups, and families. This seems to be a subtle but important global shift of philosophy. Parental and Community Willingness to Pay Existing payments for schooling are the most obvious indications that many parents and communities are willing (and able) to pay for ISSUES 5 schooling. This willingness seems to be particularly evident in coun- tries where governments do not make satisfactory quantitative provi- sion. In Uganda, for example, households have sought to bridge the gaps left by government shortcomings by meeting 65-90 percent of the costs of primary schooling (World Bank 1993d, 35). In Kenya the value of self-help contributions to education was estimated in 1991 at 19.5 million Kenya shillings (Kenya 1994, 250). In addition to demonstrated demand, research in some countries has shown that some households are willing to spend even more for pri- mary and secondary education than they now do. In some cases this reflects dissatisfaction with access to schools. In Peru, for example, the views of parents whose children live close to secondary schools have been compared with the views of parents whose children have to trav- el two or more hours each day to attend school (Gertler and Glewwe 1989; Selden and Wasylenko 1995). The researchers found that house- holds in the second group were willing to pay fees that would more than cover the costs of opening new secondary schools in their villages. Important data have also been collected in Tanzania, in this case relat- ed to the quality of education rather than to access. At the primary level the government has retained control of the public sector, and few private schools are permitted to operate as alternatives to the public institutions. Public primary schooling is officially free of charge, but many parents are unhappy about its quality.2 Some parents are dissatisfied with the quality of teachers, and even more parents are dissatisfied with the sup- plies and physical infrastructure of schools. A 1993-94 survey found that 57.4 percent of respondents considered the ade-quacy of supplies to be poor or very poor, and 43.7 percent had a similar view about the ade- quacy of physical infrastructure (World Bank 1995f, 40). In the 1993-94 survey, questions on satisfaction or dissatisfaction were followed by questions on the amounts of money that respon- dents would be willing to pay for a primary school that delivered the types of services they thought were important. As might be expected, the answers showed wide variation. Although 17.9 percent indicated that they would not be willing to pay anything, 21 percent said they would be willing to pay more than 25,000 Tanzania shillings a year (World Bank 1995f). The average for the entire sample was 8,572 Tanzania shillings, several thousand shillings more than current aver- age household expenditures per primary student. Households in other countries also have demonstrated willingness to pay greater amounts for education. In parts of Nigeria, for example, communities have historically competed to gain advancement through 6 COUNTING THE FULL COST expanded secondary education and have gone to considerable lengths to secure government permission to open secondary schools (Okoye 1986). Likewise, in Bhutan community enthusiasm to construct lower- primary schools has been so strong that the government, which is responsible for supplying the teachers, has had to impose strict con- trols on the number of institutions that may be opened (Bhutan 1991; Bray 1995c). These examples, together with the evidence on household pay- ments for schooling, are of considerable importance. They suggest that in at least some settings previously untapped sources of income for schools can be exploited. However, the examples also raise many fur- ther questions, including the extent to which willingness to pay is dif- ferentiated by income group and other factors. Peruvian data provide some answers. Willingness to pay is substantially higher among the rich, and households are less willing to make payments for girls than for boys (Gertler and Glewwe 1991, 172). In this case, willingness to pay is still substantial even among poor groups and for the education of girls, but more research is needed in other contexts before such find- ings can be generalized. Policy Dilemmas One implication, therefore, is that even when households are willing to pay more for education, it does not necessarily follow that they should be encouraged or permitted to do so. Governments face many policy dilemmas, most of which revolve around issues of control and equity. Specifically addressing the clause of the WCEFA Declaration that stressed the need for partnerships, Shaeffer (1991, 7) highlights poten- tial benefits from collaboration with parents and communities. Partnerships not only increase the resources available for education, he points out; they also can improve the relevance, appropriateness, and sustainability of reform and innovation. Work conducted under the auspices of the International Institute for Educational Planning has shown some ways in which these forces operate (Shaeffer 1992, 1994). Accountability and control are also at issue. Local financing, it has been suggested, increases the legitimacy of communities to exercise con- trol over their schools (Lockheed and Levin 1993, 15; Williams 1993, 8). Even at its narrowest level, this may generate benefits (for example, in promoting the punctuality with which teachers arrive at school each day). Strengthened accountability to communities can lead to improve- ments in home-school liaison and in the local relevance of education. ISSUES 7 Research has shown, however, that such outcomes are far from certain (Opolot 1994, 112-13; World Bank 1993a, 8). Far from feeling empowered by partnership arrangements, many communities actu- ally feel exploited. The maxim that the one "who pays the piper calls the tune" does not seem widely applicable to household financing of education in low-income countries. An alternative interpretation of many government statements about the importance of partnership is simply that the governments want to keep basic control over the schools but want somebody else to help pay for the schools (Tembon 1994; Moorad 1992). Moreover, at the school level parent-teacher associations too easily become the vehicles through which teachers draw up long shopping lists of items that they expect parents and other community members to provide. As noted above, strong reliance on parental and community resourcing can also exacerbate various inequalities. The most obvious difficulty is that the individuals and groups who are best able to help themselves are the ones who are already well off. Within individual communities school boards may be sensitive to the burden of fees, levies, and other costs on the poorest families (Bray 1996b; Makau 1985). But such sensitivity cannot be guaranteed, and there is evidence from many settings of poor children being prevented from receiving schooling because of the direct and indirect costs (Brooks and Thant 1996; Galabawa 1994; Nam 1996). Moreover, in many countries a basic inequality between urban and rural areas is exacerbated by a ten- dency to build urban schools using government funds while rural schools are mostly built by communities (Hill 1991; Maravanyika 1995). And all countries in which community financing is a strong fea- ture also experience serious regional disparities (Assie-Lumumba 1993; Bray 1996a). Given these dilemmas, government policies must be devised care- fully, and they may need to vary to suit the circumstances of specific regions and groups within countries. In forming appropriate policies, the first need is to know precisely how much parents and communities are currently providing. The following sections will help answer this question. They show that in some countries the scale of nongovern- ment resourcing even in public systems of education is both large and growing. This should not be an unmitigated cause for satisfaction, however, and this study will highlight some of the policy implications associated with this pattern. Profiles of Nine East Asian Countries Basic statistics on the population, gross national product (GNP) per capita, real gross domestic product (GDP) adjusted for purchasing power parity (PPP), and gross enrollment rates in nine East Asian countries are discussed in this chapter (table 1). Table 1 also shows the human development index, which is based on life expectancies, edu- cational attainments, and real GDP per capita (UNDP 1995, 134). The diversity of the nine countries is immediately evident from the table. Population, for example, ranges from just 2.3 million in Mongolia to 1,187.4 million in China. GNP per capita ranges from only $220 in Lao People's Democratic Republic to $1,650 in Thailand. Although the data for real GDP adjusted for purchasing power parity show a slightly different scale and rank ordering, the range is still from PPP$751 in Myanmar to PPP$5,950 in Thailand. Among the 174 countries ranked on the human development index, Thailand was number 58, placing it in the group of high human development coun- tries (UNDP 1995, 155). Cambodia, by contrast, was number 153, among the least developed countries. All countries have high prima- ry school enrollment rates, but secondary school enrollment rates ranged from just 22 percent in Lao People's Democratic Republic to 73 percent in the Philippines. Further background, such as diversity in systems of government, is also discussed in this chapter (table 2). Both Cambodia and Thailand are monarchies, though in each case elected politicians have greater power than the monarchs. Myanmar, by contrast, is ruled by an unelected military government. China, Vietnam, and Lao People's Democratic Republic are still officially socialist countries, while the governments of Cambodia and Mongolia have formally renounced socialism in favor of capitalism. The Philippines is a democracy, but the military is the principal political power in Indonesia. In contrast to the socialist states, Indonesia, the Philippines, and Thailand have always had capitalist systems. Despite their diversity, these countries have some things in common. First, despite their varied histories, all nine countries have basically adopted a capitalist mode of operation. Even in the three countries that are officially socialist, market forces are permitted to operate and have 8 PROFILES OF NINE EAST ASIAN COUNTRIES 9 Table 1. Basic Statistics for Nine Countries, 1990s Gross enrollment rate GNP Real GDP Human (percent) Population per capita per capita development Primary Secondary Country (million) (US$) (PPP$) index school school Cambodia 9.7 238 1,250 0.337 114 29 China 1,187.4 370 1,950 0.594 125 48 Indonesia 191.2 710 2,950 0.637 117 45 Lao PDR 4.5 220 1,760 0.420 98 22 Mongolia 2.3 390 2,389 0.604 84 54 Myanmar 43.7 - 751 0.457 120 24 Philippines 65.2 740 2,550 0.677 112 73 Thailand 56.1 1,650 5,950 0.827 99 33 Vietnam 69.5 220 1,010 0.539 113 33 - Not available. Soturce: UNDP 1995; World Bank 1995e; various national sources. a substantial influence on educational financing. Second, all the coun- tries have significant subnational variations. The most obvious is the contrast between urban and rural areas in schooling and resourcing patterns. Also evident are regional imbalances arising from variations in the distribution of natural resources and from other factors. All the countries have significant national minority groups that have lan- guages, cultures, and socioeconomic forms of organization that differ from those of the majority groups. Finally, the governments in all nine countries argue that educational provision should be a partnership among the government, parents, and communities. One form that this partnership can take is financial support from parents and communi- ties. This philosophy is particularly necessary in countries where the governments are poor, but it is also found in the richer countries. Subgroups of countries also share common traits. To the subgroup of former socialist states may be added the states that officially remain socialist but that are in fact increasingly oriented toward a market economy. Cambodia, China, Lao People's Democratic Republic, Mongolia, and Vietnam have all faced major economic changes as part of wider transitions in the socialist world during the 1980s and 1990s.3 In all these countries education used to be exclusively provided by the state, but private schools are now permitted and in some cases are actively encouraged. Also, government schools now have to obtain supplementary support from nongovemment sources. Although the advent of capitalist forces has provided opportunities for some sectors of the population, it has created considerable hardship for others. 10 COUNTING THE FULL COST Table 2. Some Comparisons of the Nine Countries, 1995 Official Former Current exchange Dominant colonial political Country rate (to $1)a religion power system Cambodia 2,600 riels Buddhism France Constitutional monarchy China 8 yuan none Not colonized People's republic Indonesia 2,150 rupiah Islam Netherlands Republic Lao PDR 920 kip Buddhism France People's republic Mongolia 400 tugrics none Close ties to USSR Republic Myanmar 6 kyatsb Buddhism United Kingdom Military Philippines 25 pesos Christianity United States Republic Thailand 25 baht Buddhism Not colonized Constitutional monarchy Vietnam 11,200 dong Buddhism France Socialist republic a. As of rnid-1995. b. This was the official rate. On the free market, U.S. dollars could be exchanged for nearly twenty times the official rate. Source: Author. Colonial legacies and religion create other subgroups. Cambodia, Lao People's Democratic Republic, and Vietnam form a subgroup that shares the legacy of French colonialism, and all three countries share with Thailand and Myanmar a strong Buddhist heritage. This does not imply uniformity, but it does provide some comparable underpinnings in the social values attached to education. In China and Mongolia social transition has included a resurgence of religion among some sec- tors of the population. The legacy of strong official disapproval of reli- gion has almost entirely broken the links between religion and school- ing, however. Private Education One major determinant of resourcing patterns is the nature and scale of private education. This may take an institutional form of private schools, or it may take the noninstitutional form of private tutoring. Both categories are considered in this section. Private Schools Private schools may be of many different types. Some are commercial ventures that aim to make profits, while others are run by churches or other philanthropic bodies. Some private schools are high-quality insti- tutions that charge substantial fees and aim to serve the socioeconomic elite. Others are of acceptable quality but serve all socioeconomic class- es and have a religious or other philosophy that provides an alternative PROFILES OF NINE EAST ASIAN COUNTRIES 11 to the government system. Then there are low-quality enterprises that meet excess demand and give a second chance to pupils who have failed to gain places in public schools, particularly at the secondary level. A fourth group operates at low cost and serves families unable to meet the charges or traveling expenses needed for attendance at public schools, particularly at the primary level. Private education in the nine countries varies greatly (table 3). In the Philippines and Indonesia, where the private sector has always been highly visible, most private schools are operated by religious bodies. In the Philippines 90 percent of private elementary schools and 69 per- cent of private secondary schools are operated either by the Catholic church or by the Association of Christian Schools and Colleges (Arcelo 1991, 22). In Indonesia many of the private schools are madrasahs and pesantrens, which are run by Islamic organizations and give religion a strong place in the curriculum (Oepen and Karcher 1988; World Bank 1989). In most of the former or currently socialist states, by contrast, pri- vate education was outlawed for some time and has been permitted again only since the late 1980s or early 1990s. Yet although the private sector remains small in those states, it is of increasing significance. In China, for example, the first commercially oriented institute of private education was established in Sichuan Province, home of patriarch Deng Xiaoping, shortly after a tour of the province undertaken by Deng in 1992 (Mok and Chan 1995, 1). By 1994 nearly 22,000 private kindergartens and primary and secondary schools had been estab- lished (Pepper 1995, 18.21). There have also been important changes in Vietnam. Although pri- vate schooling has been permitted only since 1989, by 1995 nearly a third of kindergartens were private operations. The private sector was much smaller at the primary level, embracing just 0.5 percent of the total in 1994, but at the secondary level it embraced 3.3 percent of the schools. In many private primary and secondary schools the facilities, equipment, and curricula were provided by the state, but the funds for teachers' salaries and other operating expenses were raised through fees. Institutions were subject to the same quality checks as public schools. The quality of private primary schools is generally considered to be equal to or higher than that in public schools, but a large pro- portion of private secondary schools cater to low-performing students (Rorris and Evans 1994, 24). In several countries community-operated schools are permitted to operate paraliel to the government institutions to meet demand that the state system cannot meet. In Myanmar, for example, substantial numbers of affiliated schools have opened since they were first permitted in 1972 (Biak Cin and Scandlen 1988; Gibson 1992). The number of affiliated 12 COUNTING THE FULL COST Table 3. Privately Operated Schools in the Nine Countries (percent) Country Primary Secondary Comments Cambodia <1 <1 Private schooling was outlawed until the early 1990s. Now a few schools serve rich families or provide a curriculum for minority religious and ethnic groups. China <1 <1 Most community-run minban schools have been taken over by the state (though community- financed, minban teachers remain common). However, recent years have seen establishment of profit-making private schools. The the private school sector is small, but it is highly significant and growing fast. Indonesia 19 71 The government gives substantial subsidies to private schools. Many of them are madrasahs and pesantrens run by Islamic organizations while others are run by Christian bodies. The number of elite private schools is small but increasing. Lao PDR 2 1 Private schooling was outlawed in 1975 but has been permitted again since 1990. Some schools are commercial ventures but others are noncommercial and are run by such groups as the Vietnamese and Chinese Associations. Mongolia 0 0 Private education has been permitted since 1991, but the main initial growth was at tertiary rather than primary or secondary levels. Myanmar 3 31a Some communities, especially in rural areas, operate affiliated schools to meet demand that cannot be met by state schools. On some definitions, however, these would be called semi-public rather than private schools. Philippines 6 39 The Philippines has a long tradition of private schooling. Most private schools are run by churches. The majority are low-cost enterprises, though some private schools are high-cost ventures that aim to serve the economic elite. Thailand 4 12 Enrollments in private schools have declined over the past two decades as many low-quality institu- tions have closed. The remaining institutions are mostly in urban and suburban areas. These instructions chiefly serve middle- and high-income families. Many private schools receive subsidies from the government. Vietnam <1 3 The term private includes full-private, semi- public, and community schools. Note: Data are for the most recent year available, in most cases 1993-94. a. Includes middle schools (grades 5-8). Source: Author. PROFILES OF NINE EAST ASIAN COUNTRIES 13 schools fell in the period immediately following the 1988 accession to power by the State Law and Order Restoration Council because the new regime converted some affiliated schools into state schools.4 The govern- ment has not been able to shoulder the whole burden, however, and affil- iated schools remain a significant part of total provision. For several decades China had comparable institutions called min- ban, or people-run schools. Their numbers fluctuated during different periods of the post-1949 era, but, as in Myanmar, they were especially important in rural areas (Robinson 1988). Since the early 1990s cate- gories have been blurred because many profit-oriented private schools prefer to call themselves minban schools. In 1995 there were 4,000 so- called minban primary schools and 851 minban secondary schools (Wang 1995, 3). However, many of these were so-called aristocratic schools for the elite. The real minban schools, that is those operated by local communities to serve village needs, had been much reduced in numbers because the government had provided assistance and made efforts to incorporate them in the mainstream. However, community- employed minban teachers were still widely used to supplement the government (gongban) teachers. In 1994 there were 1.8 million minban teachers in primary schools, 32.3 percent of the total (Cheng 1995). Because most minban teachers were in rural areas, the figure probably represented more than 50 percent of the rural teaching force. Minban teachers are also common at the secondary level, where they comprise about 10 percent of the teaching force (Tsang forthcoming). Three observations can be made about elite private schools. First, these schools meet differentiated rather than excess demand: the pupils could get places in the public system, but the parents prefer that their children go to elite schools with better facilities and teachers. Second, these schools maintain and exacerbate social inequality because they are mainly attended by children of rich families who will use their superior education to secure better incomes. Third, the fact that rich families choose to send their children to private schools relieves the government of the burden of educating those children. Ideally this should release resources to assist the poor more effec- tively. In practice, this does not always happen, however, because political forces allocate the resources elsewhere. Private schools that are less elite but exist mainly to provide an alternative form of education, such as the madrasahs in Indonesia, raise different policy issues. In these cases one concern is the extent to which the government is prepared to tolerate diversity of provi- sion and perhaps to provide aid on the grounds that the pupils are 14 COUNTING THE FULL COST still citizens of the country and thus entitled to a share of the gov- ernment's resources. In practice substantial government resources are given to private schools of this type in Indonesia. The grants not only help the institutions provide education of an acceptable quality but also assist in the co-opting of groups that might otherwise threat- en government stability. A third group of schools meets excess demand, particularly at the secondary level, raising yet other policy issues. These schools are mainly populated with students who have been unable to gain places in the public system and who are academically weak. The poorest fam- ilies cannot afford to send their children to these schools, but few pupils in these schools are from rich families. Governments are some- times embarrassed by the low quality of these institutions and may provide subsidies. These subsidies are not usually given willingly, however, especially when the governments are financially hard pressed. A fourth group of schools serves people who cannot afford the charges even of the subsidized public sector. These may be exempli- fied by an old-style, private-tutor school (known as sishu) that has reemerged in China. Along with household tutors these small one- tutor, one-room classes were traditional China's principal means of providing elementary education. They were mostly phased out after the 1949 revolution but according to Pepper (1995, 18.19) have been revived in some communities to serve rural children who cannot afford the levies demanded by public schools. The current version is said to be able to provide basic literacy and numeracy at half the cost of regular schooling. Pepper cites a study from Hunan province that indicated an annual cost of 350 yuan or more per year in regular school charges, compared with 120-180 yuan for sishu instruction, which can be negotiated in kind for families with cash flow problems. Because few such schools are likely to register with or be reported by the authorities, it is difficult to estimate their number. However, their reemergence is of considerable significance. Pepper (1995, 18.20) adds that profits from thirty students are about 2,400 yuan per year, or twice the annual income of an elementary school minban teacher in that locality. Private Tutoring Private tutoring is widespread in some countries, especially in urban areas, and may consume substantial financial resources (box 1). PROFILES OF NINE EAST ASIAN COUNTRIES 15 Box 1. Private Tutoring in Myanmar A 1991 survey of 118 grade 9 and 10 students in seven townships of the Yangon Division indicated that 90.7 percent received private tutoring. Although a 1984 law had fixed the rate for tutoring at 10 kyats per sub- ject per month, the rates revealed by the survey were more like 100 kyats per subject. In some cases students paid as much as 300 kyats per subject for five subjects. These students were paying a total of 1,500 kyats a month, or 13,500 kyats a year. Private tutoring was also found to be common among students in grades 5-8. Among 131 students in nine Yangon Division townships, 65.6 percent were receiving tutoring. The average these students paid was less than the high school students. Among the sample in Dagon Township, which had the highest average, it was 2,802 kyats a year, but in the Hlegu Township sample the average was just 270 kyats a year. In addition to the tutoring, families had to bear the costs of travel, books, uniforms, and so on. Burdens increased proportionately for fam- ilies with many children. At the time of the survey, average annual household incomes in this part of Myanmar were estimated at just 36,000 kyats. Only the richest families would easily be able to meet all the direct and hidden costs of a full secondary education. Source: Gibson 1992, 5, 10; Myanmar 1992, 24. Some tutors specialize in this activity and take it as a full-time occu- pation. Most tutoring, however, is undertaken by teachers who are employed in the public sector but who work outside school hours to earn extra income. In some societies this type of work is an economic necessi- ty because teachers salaries are very low. Governments are not always happy with this situation, and recent official directives in Myanrar and Cambodia have endeavored to impose controls on private tutoring. The growing scale of tutoring has also created official unease in Indonesia. Restrictions have not proved easy to enforce, however. Among the most problematic cases are those in which teachers are employed in government schools but give private tutoring to the same pupils in the same subjects and sometimes even in the same class- rooms. This is common in Cambodia, for example, where the midday break between shifts in double-shift schools may be used for this pur- pose (box 2). Tutoring may also be provided in the evenings and early mornings, on weekends, and during school vacations. In Cambodia, as elsewhere, the scale of charges varies by subject and by grade. In 1995 one mother in Phnom Penh indicated that for her 16 COUNTING THE FULL COST Box 2. Private Tutoring in Cambodia The clang of a metal pipe against an old car wheel resounded across the school yard, signaling the end of morning classes, but none of the sev- enty students in Kung Vichet's fourth-grade class budged from their seats. Instead they put away the notes they had been taking on how magnets work and brought out thelr mathematics notebooks. Private schooling had begun. Using regular classrooms for voluntary private tutoring is built into the instructional program at the Stung Mean Chey Primary School, as it is in many schools in Phnom Penh. "The policy of the municipality is for students to pay for private instruction," explained Uth Makara, a school official. "Teachers asked for and got permission to do it in their regular classrooms." Students typically pay 200 to 500 riels ($0.08 to $0.20) an hour directly to the teacher. Private tutoring is not, as one might assume, an opportunity for indi- vidual students to get special help on material they did not understand in class. Instead it constitutes an extension of the regular curriculum offered by the same teacher in the same large group setting-this time with a user fee attached. Parents assume that children who do not take part in the shadow private system are likely candidates to stay in the same grade next year. Uth Makara explained that private instruction at Mean Chey is "not obligatory" and that "students who do not have the money are welcome to stay." But students seem to see it differently. One boy walking across the school yard said that he was going home because he did not have any riels to offer the teacher. "The teacher says that it is okay to stay any- way," he said. "But we don't because we are so ashamed." Source: Cambodia 1994a, 109. daughters in grades 9 and 10 she paid 500 riels an hour for four hours a day, six days a week, every week of the year except for the three days after the Khmer New Year. This adds up to 620,000 riels a year ($238) for each of those two children. In addition, she paid 300 riels an hour, three and a half hours a day for her children in grades 4 and 2, or 320,000 riels a child a year. Private tutoring for all four children, there- fore, cost 1,880,000 riels ($723). Her neighbor paid even more for a child in grade 8, which is an examination year (Bray 1995a, 13). These parents are clearly much richer than the average, and few parents insist on their children taking tutoring six days a week, fifty-two weeks a year. Yet fig- ures such as these expose the heavy burden many parents shoulder. PROFILES OF NINE EAST ASIAN COUNTRIES 17 When children are given private tutoring by their regular teachers, an element of blackmail is commonly involved because at least some teachers provide only the bare minimum during school hours and reserve the real teaching for their private classes. This seems a dismal and exploitative situation, but not all parents face such difficulties. Private tutoring is clearly a major activity in some countries, but nei- ther the number of children receiving tutoring nor the costs to the fam- ilies are usually recorded in official statistics. Costs to Parents and Families in the Public System Parents or other family members within the public education system must also make financial outlays. In the nine countries covered by this study, primary education in the public system is officially free of charge. Official fees may be demanded at secondary school, however. Many schools also inpose unofficial fees and levies of various sorts, and parents have to pay for uniforms, transport, and other items. Also important, as already indicated, are the opportunity costs. Fees In some parts of the world, governments that used to provide free edu- cation throughout the system now face such economic stringency that they have had to introduce fees for at least some levels. One such country is Vietnam, where official fees for post-primary schooling were introduced in 1989. Table 4 shows the fee structure for grades 6 to 12 in 1993. Urban fees were set at higher levels than rural ones, and the fees steadily increase through the grades. Table 4. Vietnam: Official Fees for Grades 6-12, 1993 (dong per student) Per month Per yeara Grade Urban Rural Urban Rural 6 3,000 2,000 27,000 18,000 7 4,000 3,000 36,000 27,000 8 5,000 4,000 45,000 36,000 9 6,000 5,000 54,000 45,000 10 7,000 5,000 63,000 45,000 11 8,000 6,000 72,000 54,000 12 9,000 7,000 81,000 63,000 a. Monthly fees are paid for a nine-month school year. Source: Government of Vietnam, reproduced in World Bank 1995g, 196. 18 COUNTING THE FULL COST Official provision is made for fees in Vietnam to be waived or halved for certain groups. Handicapped students, boarders in minority areas, orphans, children of killed or seriously wounded soldiers, and children in mountainous or remote areas can obtain exemptions. Children of slightly wounded or seriously ill soldiers, children of government employees disabled by work injuries, ethnic minorities, and children of families who are poor or face other diffi- cult circumstances as certified by their local authorities pay half fees. Figures for students paying half fees were not readily available, but in 1993 only 14 percent of lower secondary students and 10 percent of upper secondary students were fully exempted (World Bank 1995g, 197). Some students had to pay double fees, which is among the penalties for having to repeat a class at the secondary level (Rorris and Evans 1994, 24). Fees of various types are also charged in China. Officially, no tuition fees are supposed to be demanded through the end of junior sec- ondary school.5 Other fees are permitted, however, and some are mon- itored by the authorities. Table 5 shows average miscellaneous fees col- lected in the whole country during 1993. Students in rural schools paid slightly less than students in urban schools, but the difference was smaller than in Vietnam. Students in senior high schools paid 2.5 times as much as students in junior high schools, who in turn paid twice as much as students in primary schools. Table 5 does not show the great variation both between and within provinces, however. West (1995b) presents more detailed data from three locations in Guangdong Province (reproduced in table A.2). Among these locations, Zhongshan was the most prosperous while L.iannan was very poor. In Liannan the miscellaneous fees were below the national average, but in the other two locations they were consid- erably higher. Moreover, in all three locations the miscellaneous fees Table 5. China: Average Miscellaneous Fees and Unit Costs, 1993 (yuan per student) Tuition and Total unit Fees as a share Type of school otherfees cost of unit cost (percent) General senior high school 118 917 12.9 General junior high school 47 477 9.9 Rural general junior high school 44 413 10.7 Primary school 23 247 9.3 Rural primary school 22 225 9.8 Rural vocational school 182 993 18.3 Souirce: State Education Corunission, cited by West 1995a, 55. PROFILES OF NINE EAST ASIAN COUNTRIES 19 Box 3. Items Listed on a Receipt from a Junior Secondary School in China (Yuan) Meaning of each item: 1. Tuition fee 2. Miscellaneous fee 3. Education funds 4. Office expenses 5. Desk and chair fee 6. Teaching aids, sports equipment, and library books 7. Water and electricity 8. Management fee for night study 9. Maintenance of fans 10. Minban teachers 11. Fee for students who are not locals 12. Chaoyuan fee (for students whose parents bought places) 13. Textbooks for this semester (half fee paid last semester in advance) 14. Teaching materials 15. Exercise books 16. Boarding fee 17. Payment to the cook 18. Class association 19. Health checkup 20. Public security 21. Movie fee 22. Taking care of students' bicycle 23. Textbooks for next semester (half fee paid in advance) 24. Funds for teacher retirement 25. Tutoring fee were dwarfed by the sum of all the other fees. West adds (16) that some of the fees collected by the schools were not used by the schools but instead passed on to other units. In this respect the schools were acting as agents for other organizations. The involvement of the schools in collecting some of these fees, such as those for insurance and physical checkups, has aroused controversy. To illustrate demands in China further, box 3 shows the detailed cat- egories from a receipt for fees charged by a junior secondary school in a disadvantaged rural area for one semester during 1992-93. Although no tuition fee was demanded (item 1), as many as twenty-four other charges were made. The largest item for this particular student was the 20 COUNTING THE FULL COST chaoyuan fee, which was for pupils whose parents had bought places in the school. The school was not the best in the district, but because it was better than some others and also conveniently located, at least some parents were willing to pay for instruction there. Parents of stu- dents who scored highly on the primary school leaving examination did not need to buy places, but the schools routinely kept additional places for students whose parents were prepared to pay this fee. Students who were not locals had to pay an additional fee. Other minor fees covered water, electricity, maintenance of fans, office expenses, textbooks, exercise books, the class association, movies, pub- lic security, and a health checkup. And although the categories seem to cover almost every conceivable expense, the receipt still shows the general charge for miscellaneous expenses (item 2). In Indonesia primary school fees in the public sector were officially abolished in 1977. Secondary school fees were retained until 1994, however, and even at the primary level many schools continue to charge fees for initial enrollment. A survey by Moegiadi and Jiyono (1991, 31) identified primary schools in Jakarta that charged 100,000 Indonesian rupiah (Rp) (US$55) a child admitted to grade 1 in 1989-90. Admission fees were lower elsewhere, but they existed even in rural areas. For example, one remote primary school in Central Kalimantan included in the same survey charged Rp 500. A more comprehensive national survey conducted by the Indonesian government in 1989 found that fees in public junior secondary schools ranged from Rp 1,000 to Rp 125,000 and averaged Rp 11,700.6 Private schools in Indonesia also charge fees. In most cases private school fees are higher, but this is not a uniform pattern. The 1989-90 admission fee in one high-cost private primary school in Jakarta investi- gated by Moegiadi and Jiyono (1991) was Rp 325,000, on top of monthly fees of Rp 10,000-15,000, depending on the grade. But another private school that the researchers investigated had no admission fee, and even the annual tuition fee (Rp 42,000) was comparatively modest. This school served working-class parents, some of whom found it cheaper to send their children to the private school than to pay for transport to a more distant public school. Because income from fees could not cover all expenses, the school received a subsidy from its sponsoring organization. Private schools in other countries also charge fees. The average tuition fee for private primary schools in Thailand was 1,420 baht in 1987 (table A.15), though this figure conceals wide variation among institutions. Wide variation is also found in the Philippines (Arcelo 1991; Schwartz 1995). PROFILES OF NINE EAST ASIAN COUNTRIES 21 Levies Imposed by Parent-Teacher Associations and School Boards In addition to explicitly named fees, families in many countries have to pay levies to parent-teacher associations (PTAs) or similar bodies. In Myanmar an official regulation limits PTA levies to 10 kyats per child a year, with a maximum of 20 kyats for parents with more than one school-aged child. Inflation has seriously eroded the value of such contributions, however, and many schools demand higher levies. Some schools directly contravene the regulation, but others avoid confrontation by describing demands as contributions to school funds rather than PTA levies. Among the three schools surveyed by Evans and Rorris (1994), two charged 20 kyats per student, with a maximum of 40 kyats for families with more than one child in the school. The third charged 15 kyats per student, or 40 kyats for three or more children. In Indonesia most schools have boards known as Badan Pembantu Penyelenggaraan Pendidikan, or BP3. Officially, these bodies are not supposed to include teachers unless they are mem- bers because they are also parents. In practice, however, the activi- ties of the boards are commonly guided by school staff. The 1974 decree that established the boards explicitly instructed them to seek contributions from broadly defined communities rather than from parents (Moegiadi and Jiyono 1991, 2). However, most BP3 boards impose levies on parents. In Jakarta a separate arrangement was made in 1982 for each school to form a body incorporating both par- ents and teachers, known formally as the Persatuan Orang tua Murid dan Guru, or POMG. Schools in Indonesia frequently impose BP3 or POMG charges on top of admission and class fees. For example, in one Surakarta prima- ry school documented by Moegiadi and Jiyono (1991, 15), in addition to the monthly levy of Rp 1,500 a student, parents were encouraged to contribute Rp 500 a student each quarter to cover general contingen- cies. A 1992 national survey indicated that BP3 demands in that year averaged Rp 1,473 in primary schools, Rp 3,400 in junior secondary schools, and Rp 4,874 in senior secondary schools (table A.5 shows the distribution of BP3 payments by families in different income groups). Gaps are wide between the top and bottom deciles, but they are fairly narrow between the ninth and the second deciles. At the primary level, for example, families in the top decile paid BP3 fees that were 4.7 times those paid in the bottom decile. The fees paid by families in the ninth 22 COUNTING THE FULL COST Table 6. Vietnam: Average PTA Contributions in Public Schools, 1993 (dong per student) Poorest Richest All Level of school I II III IV V Vietnam Urban Rural Primary 7,600 8,300 9,400 11,800 19,200 10,900 22,200 8,700 Lower secondary 8,500 10,600 12,300 15,400 24,600 16,300 28,800 11,600 Upper secondary 27,900 9,000 14,300 16,000 29,000 22,400 26,600 18,300 Source: World Bank 1995g, 188. decile were only 1.9 times those of the second decile. Such data assist analysis of the burden on the poorest groups, discussed below. Detailed data are also available from Vietnam, though they have been classified by quintile rather than by decile (table 6). Table 6 also shows patterns by urban or rural location. As in Indonesia, richer fam- ilies and urban families generally pay more. Even poor and rural fam- ilies have to make substantial payments, however, and the poor devote considerably larger proportions of their incomes to these payments than the rich. Demands increase significantly at higher levels of the system. Rural dwellers and poor households probably also contribute labor and materials in lieu of cash. This practice is common in most parts of East Asia and in other parts of the world (Assie-Lumumba 1993; Bray 1996b; Hill 1991).7 Other Direct Costs Other direct costs include uniforms, food, books, stationery, and trans- port. Although children need clothing regardless of whether they attend school, uniforms sometimes cost more than ordinary clothes.8 Moreover, some schools require three or four different uniforms, including ones for brdinary use and for sports in different seasons, plus perhaps scouts or other special-activity clothes. In these cases it is legitimate to include at least some uniform costs as a school expense. The same applies to food. Research in Cambodia indicated that in at least some cases the food provided at school was cheaper than that available in the market, but in other settings it is more expensive (Bray 1995a). In some parts of Indonesia pupils are required to purchase uni- forms and food through the schools so that the schools or teachers can earn extra money. Average household costs of education in Indonesia are detailed in this section (table 7). Information on the components of these costs is contained in tables A.4-7. Table A.8 desegregates the data by public PROFILES OF NINE EAST ASIAN COUNTRIES 23 and private sectors. Nearly a third of primary-level expenditures were devoted to pocket money. A large proportion of this item is for food and snacks, which are not necessarily luxury items-even the poorest households allocated a nontrivial sum to this category (table A.7). The second largest item in primary education was uniforms, which con- sumed one-fifth of the total. Study materials were the next largest item, followed by tuition fees and textbooks. At the junior secondary level, pocket money was again the largest item, though it formed a smaller proportion. Tuition fees were the second largest item, followed by uni- forms and transport. By the senior secondary level, pocket money and tuition fees were almost equal in proportion, while transport had grown to 17.2 percent. The expenditure on transport at this level pre- sumably reflects the fact that senior secondary schools were thinly scattered. Absolute amounts of expenditures are worth noting as well. Junior secondary uniforms cost, on average, nearly twice as much as prima- ry ones, and senior secondary uniforms cost about 1.3 times as much as junior secondary ones (table A.6). This probably reflects not only the fact that the pupils are larger but also that secondary schools are more likely to demand sports uniforms and other clothing that may not be required in primary schools. Expenditures on textbooks and learning materials at the junior secondary level are more than twice those at the primary level, and increase by another third at the senior secondary Table 7. Indonesia: Annual Household Education Expenditures per Student by Level and Type, 1992 Primary Junior secondary Senior secondary Type of expenditure Rupiah % Rupiah % Rupiah % Registration and entry fees 1,200 2.3 8,510 5.8 17,080 6.7 Tuition fees 5,500 10.5 26,410 17.9 52,680 20.8 BP3 fees 1,470 2.8 3,380 2.3 4,600 1.8 Examination fees 20 0.0 130 0.1 650 0.3 Other school fees 1,790 3.4 6,180 4.2 10,870 4.3 Uniforms 10,980 21.0 19,920 13.5 25,370 10.0 Textbooks 3,020 5.8 8,520 5.8 12,450 4.9 Study materials 5,570 10.6 10,090 6.8 12,490 4.9 Pocket money 17,170 32.8 36,890 24.9 55,670 21.9 Transport 1,950 3.7 17,970 12.2 43,530 17.2 Miscellaneous 50 0.1 270 0.2 870 0.3 Additional courses 240 0.5 1,410 0.9 3,790 1.5 Others 3,400 6.5 8,180 5.5 13,550 5.3 Total 52,370 100.0 147,870 100.0 253,600 100.0 Source: Indonesia 1992. 24 COUNTING THE FULL COST level. Most dramatic are transport expenses, which at the junior sec- ondary level are about 9.2 times the average at the primary level, and at the senior secondary level are more than twice the average at the junior secondary level. Further insights can be gained by desegregating the data to show differences between public and private schools (table A.8). Total household expenditures are of course higher in private than in public schools, though the gap is less than might be expected. Household expenditures on private primary schools are 79.9 percent more than on public primary schools, but the figures for junior and senior secondary schools are only 21.6 and 20 percent. The greatest gaps at each level are in the costs of tuition, pocket money, and additional courses. At the primary level, households also spent considerably more on transport in private than in public schools. A further breakdown of the public and private sectors reveals marked differences between religious and nonreligious schools (table A.9). Among private primary schools, average expenditures for nonreligious schools were Rp 130,437, compared with only Rp 56,988 in the religious schools. This implies that the nonreligious private primary schools chiefly serve the economic elite. At the primary level, unit expenditures in the religious private primary schools (Rp 56,988) were only a small margin greater than those in both the public primary schools (Rp 43,241) and the nonreligious public-primary schools (Rp 48,518). At the junior and senior secondary levels, unit expenditures in religious private schools were actually lower than in nonreligious public schools. The data indicate that at the junior secondary levels, average expenditures for households in the poorest decile were actually lower for private than for public schools (table A.4). This is presumably because the children from those households who attended private schools were mainly in religious institutions. At the senior secondary level, unit expenditures were lower for private than for public schools in the first three deciles. Data from Vietnam echo aspects of the Indonesian pattern (tables A.10-14). Pocket money is not a category in these tables because it was not asked for in the questionnaire, though food was a specific item (table A.13). Books are the most expensive category, though sig- nificant sums are also spent on uniforms, food, and lodging. Transport costs form a small proportion, but, as in Indonesia, in absolute terms increase dramatically at post-primary levels. In most categories urban expenditures on transport are higher than rural ones. The chief exception to this is transport to public upper secondary schools. Rural families also spend more on food and lodging for PROFILES OF NINE EAST ASIAN COUNTRIES 25 upper secondary students, presumably because a higher proportion of rural students must be boarders. As might be expected, per pupil expenditures in most categories rise with family incomes. The extent of variation, even in uniforms, is dramatic (table A.10) and shows a much wider spread than in Indonesia (table A.6). It would be instruc- tive to learn more about this cross-national difference. Rich families perhaps not only buy better-quality uniforms but also buy multiple sets. They also purchase sportswear and other items the poor cannot afford. Substantial variation by income group is also seen in the expenditures on food and lodging. Turnmg to Thailand, data on primary and lower secondary schools in 1987 show that, once again, books and materials are a big item, together with uniforms and, especially at the lower secondary level, transport (table 8). As one would expect, the breakdown of primary school figures shows (table A.15), as in Vietnam, higher transport costs for urban than for rural children. The families of urban children also spent more on writing supplies and uniforms, though for Thailand these figures include contributions in kind to teachers and schools that are slightly larger from rural families. The table does not include infor- mation on upper secondary schools, but a broad figure has been pro- vided by Myers and Sussangkam (1992, 39). They indicate that direct household costs for students in upper secondary schools were 6,000-10,000 baht a year per pupil, and that for rural households the costs would have been toward the upper end because of the need for transport or boarding. Table 8. Thailand: Household Direct Costs of Schooling, 1987 (baht per student) Primary Lower secondary Public Private Public Private Cost schools schools schools schools Tuition fees 0 1,417 621 1,813 Books and materials 280 625 428 520 Uniformsa 302 654 520 583 Transport 32 326 852 1,656 Other costsb 153 62 149 138 Total 767 3,084 2,570 4,710 a. For prirnary schools this includes school bags, shoes, and sportswear. For secondary schools no breakdown was provided. b. For primary schools this includes the value of household contributions in kind. The source for lower secondary schools did not indicate what was included, but it does not seem probable that it includes contributions in kind. Source: Tsang 1994b, 220; Myers and Sussangkam 1992, 38. 26 COUNTING THE FULL COST Table 9. Myanmar: Average Household Expenditures by Level of Education, 1990 and 1994 (kyats per year) Primary Middle High Type of expenditure (1994) (1990) (1990) Uniforms, school bags, and clothing 378 1,178 1,304 Books and stationery 53 335 315 Transport and pocket money 211 941 1,100 Other 61 67 123 Total 703 2,521 2,842 Note: For more detailed breakdown, see tables A.23-25. Source: Evans and Roms 1994; Gibson 1992. Data for Myammar are available only from targeted field work. Nevertheless, these data are useful (table 9). Because the primary school data are from 1994 and the middle and high school data are from 1990, comparison between the two sets should bear in mind the pressures of inflation. Comparing the middle and high schools, how- ever, the difference between the two columns of figures is perhaps less than might be expected. Expenditures are shown in greater detail in tables A.23-25, and uniforms are once again shown to be the largest. The second-largest item is labeled transport and pocket money In the primary school sample there were in fact no transport costs (table A.23), so at this level the figure is all for pocket money. Evans and Rorris (1994, 47) stated that this item was not originally included in the questionnaire but that parents clearly saw it as an educational expense and nomi- nated it as an item. Among the fifty-five households in Village A with at least one child in school, fifty independently identified pocket money as a cost item. This seems to create a parallel with patterns in Indonesia. Again, it seems likely that a large proportion of the money was spent on food. Table 9 also includes a category labeled "other." In one village in the primary school sample this covered a fee for examinations. Fees for examinations are also normal in Cambodia, where additional charges may be made for providing answers during the examination process and for adjusting grades when the marking has been completed. Figures on expenditures in Cambodia highlight the burden of uni- forms and books (tables A.20-A.21). These figures also show the importance of private tutoring, which was omitted from the figures of the other countries (see box 3).9 PROFILES OF NINE EAST ASIAN COUNTRIES 27 In Mongolia the system has remained mostly free of direct charges. However, economic crisis has required the government to insist on some cost recovery. In particular, the government now requires board- ing schools to charge for the full price of meat. To observers from soci- eties in which meat is a luxury, this might seem very reasonable. In Mongolia, however, meat is not considered a luxury. Instead it is a major component of the diet, partly because vegetables are scarce for much of the year because temperatures are below freezing a great deal of the time. Moreover, because much of the population is nomadic or thinly scattered, a high proportion of pupils are boarders (18 percent in 1985 and 9.6 percent in 1992 [Mongolia 19931 ).IO Many schools have in fact charged more than is required to pay just for the meat, using charges for food to cover other costs (Subbarao and Ezemenari 1995, 13). While the economic restructuring has benefited some households, others have been hit hard (Abbai 1995; Harper 1994). Opportunity Costs In addition to direct costs, families must bear various opportunity costs. The time that adults devote to meetings and to labor on schools is normally treated as a direct cost. The time that children devote to school is normally treated as an opportunity cost. Opportunity costs are a function not only of the hours that children spend in or traveling to and from school, but also of the labor market and the nature of home production. The value of labor generally becomes greater as children get older; but even young children can help care for their siblings who are younger still, thereby releasing adults to earn incomes outside the home. Mason (1994, 65) points out that in rural Indonesia, opportunity costs jump quite dramatically at about the time that a child completes primary or begins junior secondary school. Whereas the opportunity cost for seven-year-old children is low and associated predominantly with the contribution they can make at home, by age twelve or thirteen young people can begin to earn substantial wages in agriculture. Mason cites research that indicates that in rural Java, the main focus of his work, children age 12 to 14 work nearly four-tenths as many hours as adults age 20 to 29. Children age 15 to 19 work more than nine- tenths as many hours. Mason illustrates the importance of such a contribution to the income of an agricultural household. The household is assumed to have five members: two parents who work primarily as farm laborers, 28 COUNTING THE FULL COST a thirteen-year-old child who has graduated from primary school, an eight-year-old child in primary school, and a five-year-old child. In 1989 the father could have earned about Rp 1,400 a day, including meals, for a variety of activities, while the mother could have earned Rp 930 a day. If both parents worked twenty days a month but none of the children worked, the family would have earned Rp 46,000 a month, or Rp 6,320 per capita. This would have placed the household in the lowest income quintile. If the oldest child were male and earned as much as his mother (that is, nearly two-thirds as much as his father), his contribution would raise household and per capita income by 40 percent, to Rp 65,200 and Rp 13,040 a month, respectively. This would place the household in the second income qumitile. If the oldest child were a girl who earned two- thirds as much as her mother, her financial contribution to the house- hold would increase per capita income by 25.4 percent, to Rp 11,853 a month."1 Conducting research on primary education in Thailand, Tsang and Kidchanapanish (1992, 182) addressed the question of opportunity cost from a slightly different angle. They defined opportunity cost as the number of hours a day parents would have liked children to help them if the children were not in school. They then set the monetary value of one hour of a grade 6 student's time at 25 percent of the adult minimum wage. At the time of their research the rural and urban min- imum wages were about 6.5 and 7.5 baht an hour. The total indirect private cost per year was the product of 25 percent of the minimum wage, the opportunity cost (in hours per school day), and the number of school days per year. Tsang and Kidchanapanish (1992) found that patterns of indirect costs across school types were the reverse of patterns of direct costs. For direct costs urban families and parents of pupils in elite private schools made the greatest inputs of fees, uniforms, books, and so on, and pupils in rural government schools made the least such inputs. However, urban and elite parents were less concerned about their chil- dren being occupied in school rather than in other activities. Although the value assigned to an hour's labor was higher in the towns than in the rural areas, the total opportunity cost to rural families was thus much greater than to urban families (table A.15). Empirical estimates of the opportunity cost of labor have also been made in Myanmar. In the three villages that Evans and Rorris (1994) studied in detail, the average value of paid labor for children age 11 to 15 was 14-20 kyats a day, or 5,000-7,300 kyats a year. In the context of PROFfLES OF NINE EAST ASIAN COUNTRIES 29 average annual household incomes between 19,100 and 22,066 kyats, this was clearly of major importance to some families. The researchers found a strong correlation between paid child labor and both nonen- rollment and dropout from school. Returning to broader labor market conditions, in Mongolia oppor- tunity costs increased with the advent of the capitalist economy. Livestock herds were privatized, and families are now left on their own to look after the animals. This has contributed to substantial dropping out of schools, especially by boys (Mongolia 1993, 4-14; Harper 1994, 27). Although the economies of China, Lao People's Democratic Republic, and Vietnam differ from that in Mongolia, similar forces have operated there. The advent of the market economy has substan- tially changed the context in which school systems operate. With the sudden availability of wage-earning opportunities in the new eco- nomic system, many families face a difficult decision on whether chil- dren should stay in school in the hope of greater future returns or enter the labor force now to maximize present earnings and seize the best openings while they are available. Not all families have the luxury of this dilemma, however. Some families are forced to require their chil- dren to enter the labor market simply to survive. It is worth empha- sizing that this type of decision often rests with the family or wider household rather than simply with the individual. Community Resources in the Public System In addition to the inputs of parents and the direct family members of schoolchildren are those of broader communities. These inputs have been alluded to in connection with the BP3 boards in Indonesia and deserve further comment. A substantial literature exists on community financing of education (Assie-Lumumba 1993; Bray with Lillis 1988; Kemmerer 1992). It shows that, partly for sociological reasons, communities tend to be more cohesive in rural than in urban areas and that governments exploit this cohesion by demanding greater degrees of community financing in rural schools. Dominant patterns in much of East Asia match those in many other parts of the world. Urban-rural bias is insti- tutionalized in China, for example, where urban primary schools are almost fully funded by government but rural primary schools rely on village communities (Tsang forthcoming, 6). These communities not only provide construction materials and labor but may also provide 30 COUNTING THE FULL COST food for the teachers, which may be their only remuneration (World Bank 1991, 45; West 1995a, 12). Community financing may also be linked to religious behavior. In the Buddhist countries strong links are found with temples. For rea- sons that deserve to be explored more fully, links appear to be stronger in Cambodia than in Lao People's Democratic Republic and Thailand, but in all three countries they are substantial. Strong ties to Buddhist temples also exist in Myanmar. Biak Cin and Scandlen (1988, 171) highlight the importance of cetana, which is translated as loving kindness, and point out that attitudes that in the past sup- ported monastic schools have to a substantial extent been trans- ferred to secular schools. Donations by individuals and families to schools are commonly seen as an expression of virtue within the reli- gious context. Among the minority religious groups in Thailand, Tsang and Kidchanapanish (1992, 188) found that Christians made above-average household inputs to education, while Muslims made below-average inputs. In other countries, however, Muslim commu- nities may make substantial inputs. This is particularly evident in Indonesia. Community resourcing is also seen in Lao People's Democratic Republic (box 4). In this village each household was levied both in cash and in labor. The leaders in the village argued that schooling was a good investment for the whole community and said that the arrangement was therefore generally accepted. The leaders did, how- ever, recognize that some households were too poor to pay the cash levy. Those households provided extra labor instead (CARE International 1996). In some settings community leaders are extremely well organized and are able to impose fines or other sanctions on households that are reluctant to participate (Hill 1991; Igwe 1988; Moegiadi and Jiyono 1991). This is not the case in all settings, however. At the microlevel, variations among villages arise from variations in the enthusiasm, vision, and organizational ability of village leaders. More broadly, variations arise from divergence in living standards and in cultures and forms of social organization. Community activities in Lao People's Democratic Republic, for example, are said to be stronger in the more prosperous parts of the country (such as Luang Prabang, Savannakhet, and Vientiane) than in the poorer and more sparsely populated north- ern and mountainous areas (Bray 1995b). Also, different racial and eth- nic groups have different cultural traditions about the extent and nature of cooperation. The schools run by the Vietnamese and Chinese PROFILES OF NINE EAST ASIAN COUNTRIES 31 Box 4. Community Resources for a Rural Primary School in Lao People's Democratic Republic The village of Namouang is in a rural part of Vientiane Province, about three hours' drive from Vientiane Municipality. The village has made major inputs to its primary school. The most visible contributions are two buildings, One was constructed in fiscal 1993 with materials that cost 1.2 million kip; the other was constructed in fiscal 1994 at a cost of 1.3 million kip. Adding of the value of labor would at least double these sums. The community has also erected a perimeter fence and has con- structed benches and desks. To collect the money, each of the fifty-three households was levied an equal sum. Although the households had varying numbers of children and some had none at all, nobody complained about this system. Education was seen, at least by the leadership, as a sound investment for the community as a whole. Four of the fifty-three households were rec- ognized as unable to contribute in cash, so they contributed extra labor instead. The chairman of the Community Association oversaw the work. To carry out the construction, the village was divided into four groups that provided labor on separate days. Each household was required to send a strong worker, and careful records were kept. Anybody arriving late for work had to have a good explanation. Each building took about a week to construct. Source: Bray 1995b. Associations in Lao People's Democratic Republic are strong institu- tions with considerable community inputs, but the activities of other minority groups are much less vigorous. These pattems have parallels in other countries. In addition to materials and labor, communities commonly donate land for schools. In many cases such land has a high opportunity cost, especially when it is flat and located close to a population center. Were the govemment to purchase such land, it would require a substantial budgetary provision. Community resourcing is not strong in all countries, however. Mongolia is the most sparsely populated country in the world, with a population density outside Ulaanbaatar of just 1.05 people a square kilometer. Also, 28.5 percent of the total population, or 38.8 percent of the non-Ulaanbaatar population, is nomadic, and seven 32 COUNTING THE FULL COST decades of antireligious activity by the communist government have resulted in a largely atheist society. Under such circumstances com- munity cohesion of the type needed to support schools cannot be expected. Proportions of Total Costs Met by Families and Communities The total costs of education can be calculated by adding the inputs of governments, families, communities, and other agencies where rele- vant. Ideally, this calculation should include the value of labor, mate- rials, and land as well as cash and should include expenditures on pri- vate supplementary tutoring. It is then instructive to identify the pro- portion of the total represented by nongovernment inputs. The discus- sion that follows presents such figures where possible, though data on some costs-particularly nonmonetary and noninstitutional ones-are scarce. A complete picture would include the opportunity costs, not only for the resources of families and communities but also for those of governments. However, these costs are not discussed in this section because they differ from direct inputs. Figure 1. Household and Government Public Primary Education Resourcing in Nine Countries Percent U Household 100 Government 80 60 , .u 40 20 0 Cambodia China Indonesia Lao Mongoha Myanmar Philippines Thailand Vietnam PDR Note: See text for explanation of what items are included in household expenditures in each country. Soturce: Evans and Rorris 1994; Hossain 1996; King 1994; Schwartz 1995; Thomas 1996; Tsang and Kidchanapanish 1992; West 1995a, 1995b; World Bank 1995g. PROFILES OF NINE EAST ASIAN COUNTRIES 33 Figure 2. Aggregate Flow of Funds in Elemenatary Education, 1986 (millions of pesos) Aggregate flow of funds Operating costs = 8,221 Private share = 6% Direct social costs = 9,445 Private share = 18% Householcls Fees School- Other private Local Central 316 q related sources governments government expenditure 145 209 7,551 316 \ '1276 / \126 209 7,551 Private schools Public schools Operating Direct Operating Direct costs social costs costs social costs 335 611 7,886 8,834 Souirce: World Bank 1988, 26. Figure 1 shows the proportions of total expenditure in public pri- mary schools borne by households and governments in the nine coun- tries covered by this study. In all countries the household proportions are significant, but in Cambodia, the Philippines, and Vietnam, they are dramatic. Households in Vietnam meet half the total costs, and households in Cambodia meet three-quarters. Only in Indonesia are household costs estimated at less than one-tenth. Within each country are substantial variations, and various cautions need to be emphasized on comparability of the figures across countries, particularly in rela- tion to what was included or excluded in calculation of household costs. Nevertheless, the portrait is striking. It should be stressed that figure 1 refers only to the public sector and to the primary level. The shares of total costs met by households would be much greater in the private sectors. Available data indicate that they are also considerably higher at the secondary level. In Vietnam, for example, households were estimated in fiscal 1993 to be meeting 52.2 percent of the costs of public primary education but 67.7 percent of the costs of public lower secondary education and 34 COUNTING THE FULL COST 72.2 percent of the costs of public upper secondary education (World Bank 1995g, 192). Figure 2 gives some historical perspective with statistics that are not recorded in figure 1. Showing the aggregate flow of funds in Philippine elementary education in 1986, it indicates that in the public sector direct social costs (that is, excluding opportunity costs) amount- ed to 8,834 million pesos, of which 948 million pesos came from house- holds and 126 million pesos came from other private sources such as churches and companies. Among private schools, direct social costs amounted to 611 million pesos, of which 316 million came from fees, 276 million came from other household expenditure, and 19 million came from other private sources. Overall, households and other pri- vate sources met 6 percent of total operating costs and 18 percent of total direct social costs. Figure 3. Aggregate Flow of Funds in Secondary Education, 1986 (millions of pesos) Aggregate flow of funds Operating costs = 3,216 Private share = 37% Direct social costs = 4,355 Private share = 54% Households Local governments School- Fees _ - National government NCR Non-NCR related 1,200 1,343 521 178 expenditure 1,139 560166 18721 5217 57^ 40 0 \< ~~~~/ -895 57 14 Private high schools Public hi gh schools Operating Direct Schools National National Local costs social costs attached to high schools high high 994 1573 state universities schools schools and colleges NCR Non-NCR 198 895 524 605 Operating costs = 2,222 Direct social costs = 2,782 Note: NCR is National Capital Region. Source: World Bank 1988, 29. PROFILES OF NINE EAST ASIAN COUNTRIES 35 Table 10. Philippines: Cost-Sharing in Public Education, 1986 and 1994 (percent) 1994 index 1986 1994 (1986=100) Elementary Govermnent 87.9 69.1 139 Households 10.6 30.4 507 Total 100.0 100.0 177 Secondary Government 66.5 57.3 132 Households 33.5 42.5 195 Total 100.0 100.0 154 Source: Schwartz 1995, 50. The pattern at the secondary level was more complex (figure 3). At that time there were at least eight types of secondary schools in the Philippines: two types of nationally funded schools (those adminis- tered directly by the government's Department of Education, Culture, and Sports and those attached to state universities and colleges); four types of locally funded schools (provincial, city, and municipal sec- ondary schools financed by local governments and local schools financed by communities with varying amounts of government sub- sidy); and two types of private schools (sectarian schools affiliated to religious groups and nonsectarian schools operated either as stock or nonstock corporations). The figure shows that the operating costs in the sector amounted to 3,216 million pesos, of which 37 percent was pro- vided by households in the form of fees. However, the total direct social costs, which included household expenditures on books, uniforms, transport, and so on plus private donations and income from school activities, amounted to 4,355 million pesos. Of this, 54 percent was pro- vided by the private sector. In other words, nongovernment expendi- tures at the secondary level exceeded government expenditures. Moreover, this was an underestimate of nongovernnent expenditures because it omitted the costs of supplementary out-of-school tutoring and the value of nongovernment inputs of materials, labor, and land. Since 1986 several important changes have occurred in the Philippines. In 1988 the government abolished tuition fees in all pub- lic secondary schools and agreed to pay the salaries of all teachers in community-run local secondary schools. The burden on households has not, however, been reduced in the way that was promised at the time of the reforms. Comparing proportions of cost sharing in 1986 and 1994, the secondary level total costs increased 54 percent (table 10). 36 COUNTING THE FULL COST Table 11. Indonesia: Private Components of Primary School Costs, 1989 (percent) Public Private Rural Urban All schools schools schools schools schools Jakarta 29.0 91.9 26.6 88.2 39.0 Other Java 9.3 48.7 8.6 30.2 10.0 Outside Java 3.5 12.6 3.5 8.5 4.2 Indonesia 7.8 30.0 6.8 13.4 8.8 Source: Calculated from King 1994,18. The bulk of this increase (an index of 195, compared with 132) was cov- ered by households rather than by the government.12 In public ele- mentary schools the shift toward private subsidy was even more dra- matic. Whereas the government index increased from 100 to 139, the household index increased from 100 to 507. The chief reason unit costs increased so much was that teachers salaries improved in both private and public schools. An official cap on the ceiling of private teachers salaries was removed. Public teach- ers' compensation packages increased 5 percent a year in real terms between 1985 and 1995. Moreover, the raise in basic salaries and other compensations granted in 1994 and 1995 was part of a longer-term package with scheduled annual adjustments up to 1998. To pay for this, the government has had to cut other expenditures in the public schools. The result has been a de facto transfer of the burden to par- ents, who, ironically, have ended up paying more for education than they were before of the 1988 reforms that had promised free education. In Indonesia the picture is less dramatic but still quite marked. At the primary level in 1989, private inputs to public primary schools met 7.8 percent of school budgets and in Jakarta the proportion was as high as 29 percent (table 11). In the private sector nongovernment inputs were naturally much higher (note the scale of government inputs to private schools, especially outside Java). The figures suggest that in general pri- vate inputs were higher in urban than in rural schools, though the cal- culation omits allowance for the value of donated labor, materials, and land, which were probably more forthcoming in rural areas.13 Vietnam provides a rather different setting because it is officially a socialist society in which, at least until recently, strong emphasis has been placed on government provision of education. Given this context, the large proportions of costs being met by households are all the more noteworthy. In fiscal 1993 households met 52.3 percent of costs at the primary level and even greater proportions at post-primary levels PROFILES OF NINE EAST ASIAN COUNTRIES 37 (table 12). Once again the figures exclude allowances for household and community inputs of labor, materials, and land. In Cambodia no detailed national surveys have been conducted of the types presented from Indonesia, the Philippines, and Vietnam, but a picture can still be constucted from existing data. Tables A.19 and A.20 present some figures on household expenditure per pupil by level of education. The tables are not easily compared with each other because this period was characterized by high inflation and also by some fluctuation in prices. To convert the figures into real costs, therefore, it would be necessary to know the months in the year during which expenditures occurred. Still the figures do provide some indication and can be compared with estimates of government expenditures per student in selected provinces (table A.20). The fig- ures suggest that, including private coaching, families may have been spending at the primary level between three and nine times the amount the government was spending per pupil, and that the gap between family and government expenditures was even wider at the secondary level. The statistics for other countries do not include the costs of private tutoring (see figure 1). These costs were deleted from the figure for Cambodia, but the average for other household costs for primary schools (table A.20), compared with the average of gov- ernment unit expenditures (table A.22), still shows households meet- ing 75 percent of total costs. Estimates in Thailand (Tsang 1994b, 220, 226) placed household con- tributions at 14.1 percent of the total costs of public primary education Table 12. Vietnam: Public and Household Resources for Public Schools, Fiscal 1993 (billions of dong) Lower Upper Item Primary secondary secondary Total contribution 1,590 963 407 Government 759 311 113 Households 831 652 294 Fees 36 95 57 PTA 109 53 15 Uniforms 122 110 40 Books 237 160 54 Other 327 234 128 Household contribution as a percentage of total 52.3 67.7 72.2 Source: World Bank 1995g, 192. 38 COUNTING THE FULL COST in 1987.14 In Myanmar a 1994 Ministry of Education study of primary education in Sagaing Division estimated household costs at 30 percent of the total (Evans and Rorris 1994, 32). In China, estimates based on field work in Guangdong province place household costs at 16-25 per- cent of total costs of basic education (West 1995b, 18).15 All these observations may be summarized in a few important points. First, in most countries households make significant contributions to schooling. The research basis of the figures suffers from various short- comings, in particular from the fact that most studies exclude nonmone- tary contributions of land, labor, and materials. Also, the studies in dif- ferent countries used different items in their calculations. It is possible, however, to make comparisons across countries. Most striking is the dif- ference between the figure for Indonesia, where the proportion was about 75 percent. Thailand was also on the low side, while Vietnam was on the high side. The fact that proportions met by households were high- est in the countries that had recently shifted form socialist economies toward capitalist ones to a large extent reflected the crisis in government funding in those countries. The picture presented by these figures may of course be transitory, but much depends both on the scale of government revenues and on the nature of government policies. Burden on Household Budgets It is obvious that education is a major item of household expenditure in most parts of the region. This finding deserves further investigation, particularly to identify the burden on families in different income groups and also to note variations within particular countries. Table 13 summarizes data on the proportion of household expen- ditures consumed by education in the nine countries. The fact that the reported studies used varying bases for assessment means that they cannot be placed on a single numerical base. Among the methodolog- ical variations are whether incomes refer only to cash or whether they also refer to incomes in kind; whether disparities are calculated by decile, quintile, or some other basis; and whether all household expenditures or only nonfood expenditures are the focus. Variations are also found in the basis of educational costs. Some studies cover only cash expenditures, but the study in Thailand was notable for having added both contributions in kind and opportunity costs.16 Despite this methodological diversity, however, some patterns may be noted. First is that the poor in all countries are hit hard, whether they are in countries that overall are rather prosperous such as Thailand, or PROFILES OF NINE EAST ASIAN COUNTRIES 39 in countries that have a low average GNP per capita such as Vietnam. The number of absolute poor varies by country and by localities with- in those countries, but in most settings the cost of education is a heavy burden for the poor. Poor households are also likely to have children, a further challenge to their budgets. Another consistent finding is that secondary education consumes more than primary. This means that even if poor families can afford primary education, they are unlikely to be able to afford secondary education. Yet although rich families spend much more on education in absolute terms, both because their children stay longer in school and because their children are more fully resourced during each year of schooling, rich families spend much less in proportional terms. The size of the gap between rich and poor varies (see table 13). The figures for Mongolia indicate the smallest gap, though the rich pur- chased much more with given percentages of their incomes and in par- ticular had considerably greater enrollment rates in higher education (World Bank 1996). The gaps between rich and poor were much greater in Indonesia and Vietnam. In Vietnam the direct costs at pri- mary education consumed 22 percent of nonfood expenditure for the poor, compared with just 12 percent for the richest quintile, and gaps widened at lower and upper secondary levels. The findings on the proportions paid by urban as opposed to rural households are mixed. The Cambodian figures, although they should in general be treated with caution,17 show the proportion of household incomes consumed by education to be considerably higher in urban than in rural areas (Cambodia 1994c, 3-6). This is no doubt partly due to the fact that urban families are more likely to have students in post- secondary education. The study by West (1995b), which focused on primary and lower secondary education in China's Guangdong Province, showed that although households in the more urbanized middle-income and rich areas spent more per child in school (and were more likely to have children who stayed to the end of lower sec- ondary school), those expenditures formed a smaller proportion of total household expenditures. West's indication of regional disparities within Guangdong Province are echoed in interprovincial figures computed by the nation- al government and quoted by Tsang (1994a, 300-01). The combined average of extrabudgetary expenditures (includes schools' self-gener- ated revenue and social contributions, fees, surcharges, and levies but not transport, food, uniforms, or other expenses met by households) at primary and secondary levels in 1989 formed 5.3 percent of average 40 COUNTING THE FULL COST Table 13. Household Education Expenditures Country Expenditures as a proportion of household incomes Cambodia A 1993-94 survey reported that households in Phnom Penh devoted 3.3% of their incomes to education. The figure was 1.7% in other towns, and 1.9% in rural areas. A smaller survey of three villages in 1992 recorded proportions ranging from 4.9 to 11.8% (table A.19). China A 1994 study in Guangdong Province found that in the poorest part education consumed 29% of household incomes, while in middle- income and rich parts it consumed 19% and 14% (West 1995b, 24). A 1988 study found that the direct private costs of primary education in rural Shaanxi and Guizhou Provinces were 41 and 39 yuan. For average farm families those costs were not too heavy because total annual family incomes approached 2,000 yuan and cash incomes about 1,250 yuan. However, the poorest 8%, with total incomes below 1,200 yuan and cash incomes below 500 yuan, would have found the burden heavy (World Bank 1992, 85). In Beijing households spent 2.9% of their incomes on education in 1989 (Tsang 1994a, 294). Indonesia In 1989 direct primary school costs consumed 38% of per capita incomes in the poorest quintile and 17% in the richest quintile. Junior secondary figures were 47% and 11%, and those for senior secondary were 84% and 17% (World Bank 1993b, 56). Lao PDR Precise numerical estimates have not yet been calculated for Lao People's Democratic Republic. However, informal evidence together with some more formal studies suggests that the burden on house holds is comparable to that in Cambodia and Vietnam (CARE International 1996; World Bank 1995a). Mongolia A 1995 survey estimated average per capita household expenditure on education at 6.1% of per capita nonfood expenditure. The figure for the poorest quintile was 7.5%, and for the richest quintile was 5.3% (table A.27). Myanmar The cost of high schooling, not including private tutoring, was estimated in 1990 at 15.6% of average family income (Gibson 1992, 4). Inclusion of tutoring would have raised the proportion to 26.8%. Figures for middle schooling were 13.1% and 16.8%.However, case studies estimated primary education consumes only 2.6% to 4.2% of average village household incomes (Evans and Rorris 1994, 36). Philippines According to a 1994 estimate, households in the lowest seven deciles spent 2.7% of their incomes on elementary and 4.5% on secondary education (Schwartz 1995, 52). More detailed work is needed to desegregate expenditures by decile. Thailand For households with incomes below 650 baht per month the direct private costs in 1987 of sending a child to public primary school consumed 16.2% of total income. Households with incomes above 10,000 baht spent only 0.6% of total income per child in primary school. The greatest marginal gulf was between the poorest and those earning 651 baht to 1,600 baht per month, for those earning 1,600 baht per month spent only 3.8 baht per child on the direct costs of primary PROFILES OF NINE EAST ASIAN COUNTRIES 41 Country Expenditures as a proportion of household incomes school (table A.16). For the poorest groups, contributions in cash and in kind were equivalent to a further 4.3% of household income; and the opportunity costs were equivalent to 26.2% of household incomes. Adding these up, the burden on the poorest was shown as a huge 46.7%, compared to 10% in the next income group and just 1.9% in the top income group. Vietnam For the families in the poorest quintile direct costs per primary student in 1993 were equivalent to 22% of nonfood consumption. This was nearly twice the figure for the richest quintile (12%). At the lower secondary level, the cost-nonfood ratio was 45%, more than twice the ratio for the rich, and upper secondary schooling cost as much as the entire nonfood budget of the poorest quintile, which was three times as much as the relative cost facing the richest quintile. According to these figures, a family with two children in primary and one in junior secondary would have spent 89% of its nonfood budget on school expenses (World Bank 1995g, 94). per capita income in the poorest province (Guizhou) and 9.4 percent in the richest province (Liaoning). Based on this and other data, Tsang (1994a, 305) concluded that not only were households being required to contribute ever-greater amounts to schools but also that regional disparities were widening. Little information is available on the differential burden on minority groups. In general, minorities in Cambodia, China, Lao People's Democratic Republic, Thailand, and Vietnam have low enrollment rates, though they also tend to have low incomes and so may still be allocating high proportions of incomes to education (Nam 1996). The chief exceptions are the minorities of Chinese ethnicity outside China itself. In Vietnam, for example, Fong (1994,6) indicates that among the majority Kinh population, expenditures in 1993 averaged 140 dong, compared with 439 dong for the Chinese minority but only 45 dong for other minority groups. Policy Implications It might be argued that the input of resources by those who can afford it is thoroughly desirable. From a financial perspective, it increases the total resources available for education and releases some of the pres- sure on seriously strained government budgets. Cost sharing may also contribute to a feeling of partnership in the educational enterprise, and this partnership may in turn generate more resources. In China, for example, Tsang (forthcoming, 8) has stated that decentralization reforms initiated in 1985 encouraged a sense of community ownership of schools, which increased willingness to make contributions. This study has, however, highlighted various problematic features of recent trends. The greatest problems concern equity. The heavy bur- den on some sections of the population contributes to high dropout rates. Parental and community financing may also lead to major regional imbalances. Privatization in Education Systems In some of the countries covered by this study, the private sector has always been prominent, and in others it has greatly increased in promi- nence. Much of the privatization is hidden in the sense that it has occurred in the public sector rather than in the private sector per se. One message from this study is that the extent of privatization should be rec- ognized more widely. In addition to private schools, it includes private tutoring. It also includes increased private financing of public schools. At the beginning of this study a distinction was made among four types of private school. First are elite private schools, which generally provide good-quality education at a high price; second are schools run by religious or other not-for-profit organizations, which provide an alternative to the public system and may be qualitatively superior, equal, or inferior to public schools; third are low-quality institutions, which operate on low budgets but cater to excess demand and give a second chance to those who have failed to gain places in public schools; and fourth are low-cost institutions, which cater to students who could go to public schools but who are discouraged from doing so by the various charges made in those institutions. 42 POLICY IMPLICATIONS 43 Each form of private school has strengths and weaknesses. On the positive side, many people value diversity within the system, feel that elite schools can act as a model for other institutions, and applaud the fact that private second-chance institutions provide opportunities that would otherwise not be available. Even the low-cost institutions that serve children whose families cannot afford the charges in the public system have some positive role, because in the absence of such insti- tutions the children would probably have to leave school altogether. The fact that parents and communities are prepared to pay for private schooling also relieves the government from educating some children and permits redistribution of resources to other areas. On the negative side, however, elite private schools may maintain and exacerbate social inequalities, alternative religious schools may threaten national cohesion, low-quality secondary schools sometimes raise individual aspirations to unrealistic levels and waste resources that could be bet- ter invested in other activities, and the reemergence of low-quality pri- mary schools serving people who are excluded from the public system is felt by some observers to be evidence of retrograde trends in the education system. Ultimately, policies on private schools are based mainly on value judgments. Economics certainly plays a role, but even in poor coun- tries the policymakers are able to prohibit private schooling if they are prepared to deny second chances to those who are unable to gain pub- lic places or to demand that children whose families cannot pay pub- lic school charges should withdraw from schooling altogether. Few governments are willing to take such strong stances, however. In Indonesia the authorities have sought to discourage lower-quality private schools but to provide subsidies and find other ways to upgrade their quality. The authorities in Myanmar have permitted communities to operate affiliated schools that help to extend oppor- tunities especially in rural areas and have held out the possibility that the government will relieve the burden on these communities by tak- ing over the schools. The Myanmar government has not welcomed operation of more elitist private institutions but has left the door open to an informal shadow system at the secondary level by permitting private candidates to register through public schools for public exam- inations (Myanmar 1992, 23). In China, Lao People's Democratic Republic, and Vietnam the authorities are much more tolerant of elite and alternative-model pri- vate schools. In Lao People's Democratic Republic the Asian Development Bank has given technical assistance on mechanisms to 44 COUNTING THE FULL COST support the private sector (Murugasu 1993), and a prime minister's decree setting out guidelines and incentives was issued in 1995 (Weidman 1995). The Chinese government is ambivalent about the growth and implications of private aristocratic schools (Mok and Chan 1995; Wang 1995), but growth in the sector has been so vigorous that it would now be difficult to stem the tide. Likewise, although for a time the Vietnamese authorities were somewhat coy about the newly emerged private sector, preferring instead to call it semipublic, the true nature of the increasingly vigorous private institutions is now being widely recognized and accepted. Private tutoring also has both positive and negative sides. One major recommendation arising from this study is that the activity be researched more thoroughly to determine more accurately its nature, extent, and implications. Meanwhile, it is worth remembering that pri- vate tutoring is a way through which pupils can receive more individ- ual attention, and it can have a clear influence on their academic achievement. It is also a mechanism through which poorly paid teach- ers can increase their earnings. On the negative side, in most settings private tutoring seems to tip the curriculum further toward academic book-learning to pass examinations, and it appears to favor rich and urban populations. It may also impose a heavy burden on household budgets, and in some settings an element of blackmail is involved, par- ticularly when teachers give private tutoring to the same children whom the government employs them to teach during school hours. Other forms of privatization in the public sector have been particu- larly evident in Cambodia, Lao People's Democratic Republic, Mongolia, and Vietnam, where education used to be almost fully fund- ed by the government but where a wide range of contributions is now demanded from households. China and Myanmar were socialist states that diverged from this model and encouraged or at least tolerated what in China were called minban schools and in Myammar were called affiliated schools. But even in those two countries the past decade has brought a great increase in the proportion of funds solicit- ed directly from households rather than through the taxation system. Such an increase has also been evident in some of the countries that have always had capitalist regimes. The governments of Indonesia, the Philippines, and Thailand are all making major thrusts to expand enrollments in junior secondary education, and to that end have extended the official abolition of fees from primary to junior secondary education. Policies and practice are not, however, always the same thing. In the Philippines the financial burden on households has POLICY IMPLICATIONS 45 mcreased rather than been reduced since 1988, and in Indonesia offi- cial government-determined fees have been replaced by fees, levies, and contributions determined at the school level. These observations underline the need for policymakers first to assess carefully the advantages and disadvantages of privatized edu- cation, and second, if they decide that they are willing to tolerate or encourage greater privatization, to examine its many different forms and manifestations. Although evidence suggests that private schools may be more efficient than public ones (Bashir 1994; Lockheed and Jimenez 1994), considerably more research on the matter is needed before findings can be considered firm (Bray 1996c; Kingdon 1994; Riddell 1993). Advocates of privatization often point out that an increase in enrollments in private schools can save the state money by reducing enrollments in public schools, but it is far from certain that the money that might be saved would actually be used to benefit dis- advantaged groups. Of course, it should also not be assumed that pub- lic education is necessarily less elitist or discriminatory than private education. Effect on Enrollment One can note positive and negative effects on enrollment from parental and community financing. On the positive side, the extra resources have extended the quantity of education. This is particularly clear in the case of Myanmar's affiliated schools, but is also true of other set- tings where local contributions have helped establish or maintain schools that could not be sustained by the government alone. While further research is needed, it might also be argued that direct family contributions to the cost of education encourage those families to pay greater attention to pupils' attendance and to the nature of what their children learn in school. This in turn could help ensure an appropriate return on the investment of the resources, including the children's time. By contrast, the combination of direct and indirect costs imposes a heavy burden on the poorest households. In these settings there is a likelihood either that children will not be enrolled or that they will drop out at an early stage. The World Bank has noted (1991, 62) that in China "poor parents are finding it increasingly difficult to send their children to school and many parents have come to doubt the value of schooling." A subsequent report observed that although the Chinese government had increased funding for education, most of 46 COUNTING THE FULL COST that funding had benefited the well-off urban population and middle- and upper-income rural inhabitants (World Bank 1992, 81). Very little of the money had reached the absolute poor. Moreover, because schooling in rural areas was inferior in quality, potential returns to investment in education were relatively low for those groups. Evans and Rorris (1994, 31), have made comparable comments about Myanmar. They pointed out that "[t]he difficult economic situa- tion in Myanmar households means that even the most negligible financial costs associated with primary education may be of critical significance." They added (35) that although education was highly val- ued in the settings they investigated, this was more as a means of improving status and of being better able to function in society than because schooling, at least at the primary level, was seen as an invest- ment. In such circumstances it is no surprise that the poorest house- holds have simply found that they could not afford the direct and indi- rect costs of school attendance. Even if they had been willing to take loans to finance the schooling, the virtual absence of credit markets would have precluded this possibility. Developments have been particularly dramatic in Vietnam and Mongolia, where economic transition has brought a break with the past. Although Vietnam used to boast universal primary education, during the late 1980s and early 1990s enrollment rates fell markedly. By 1993 only 78 percent of children age 6 to 10, the official age group, were reported to be enrolled in primary schools (World Bank 1995g, 84). In Mongolia dropout and nonenrollment in primary and sec- ondary schools were negligible before 1991, but by 1994 an estimated 30 percent of the grades 1-10 age group were not in school (World Bank 1995b, 6). The differences between the rural and urban areas were even more alarming: 37 percent in rural areas, compared with 16 percent in towns.18 These figures of course also reflect other developments, including the fact that young people in Mongolia were now needed to look after privatized herds and that in both Vietnam and Mongolia the new opportunities arising from the market economy made many young people feel that their time would be better spent outside school rather than within it (Abbai 1995; Brooks and Thant 1996). It seems clear, however, that the direct costs of schooling were certainly a major fac- tor depressing enrollment rates (World Bank 1996). In Indonesia and Thailand the situation is rather different. The economies of these countries have not been subjected to the pains of transition from socialism to capitalism, and in general the societies are POLICY IMPLICATIONS 47 much more prosperous. The main focus of current government initia- tives in those two countries is on raising secondary enrollments. The authorities certainly recognize that the poorest of the poor are exclud- ed from primary schooling and that these families need attention and help, but they feel that they are a fairly small minority. Raising lower secondary education rates to 95 percent or more is an even greater challenge than at the primary level, partly because lower secondary education has higher direct costs both for governments and for house- holds, but also because for that age group families must bear much heavier opportunity costs. Official Policies on Fees The figures on the amounts that households devote to education, whether voluntarily or not, should be considered in conjunction with the severe resource constraints faced by many governments and should prompt reconsideration of official policies on fees. The outset of this study presented several declarations from forms convened by United Nations organizations that asserted the desirability of fee-free education. This philosophy has become widely accepted, and in three of the nine countries covered by this study (Cambodia, Mongolia, and the Philippines), free public provision of education is enshrined in the national constitutions. Education is not free, however, in the sense that no one has to pay for it (box 5). The politicians who promise fee-free education usually state that all the resources needed by schools will be provided by govern- ments. In most low-income countries, however, governments are simply unable to meet all the needs of public schools. The result is either that schools are starved of resources or that parents and other community members decide or are requested to contribute funds to bridge the gap. One irony of free-education campaigns is that, at least in the short run, they often deprive schools of essential resources. Because the govern- ment promises to meet all needs, parents quite reasonably withhold their contributions. Only when it becomes obvious that the government will not fulfill its promise do parental and community inputs come forth again, and even then the volume of such inputs is generally less than it would have been in the absence of official policies of fee-free provision. This point suggests that governments might be better advised not to declare that education in public schools will be free of charge, particu- larly if those governments are unable to provide all the resources that schools need. This is not necessarily to suggest that governments 48 COUNTING THE FULL COST Box 5. High Cost of Free Schooling Public education in Cambodia is ostensibly free, but an elaborate system of user fees and other charges can strain the budgets of poor families and, in some cases, put it out of reach entirely. Take the case of Chann Mom, who lives with her husband, Chann Sophal, and five children in a one-room thatched hut adjacent to a rice paddy in the village of Russey. Chann Mom and Chann Sophal awake long before dawn each mom- ing to stoke the fire in an open charcoal pit and to make small cakes known as akow for her to take to the Russian Market in nearby Phnom Penh. She brings home about 3,000 riels ($1.20). He eams a comparable amount driving a rented motorbike taxi, and the family is also raising two pigs that will yield about 19,000 riels. The Chann family is fortunate because a charitable organization is picking up the 1,200 riels entrance fee for the four children in Russey Primary School, and her sister is doing likewise for the oldest son, Chann Phan, at the Beng Tra Bek Secondary School. Nevertheless, there are other costs for schooling. Phan needs 500 to 1,000 riels each day to pay for lunch and voluntary private tutoring, and the younger children need 100 to 200 riels for their own private lessons. "When I have the money for tutoring, I give it to them," she explained. "Other times I do not have the money and they feel ashamed, so they do not go to school those days." Mom ran into a problem two years ago when, because of heavy med- ical expenses, they had to sell their house in Phnom Penh and move to Russey. She said that the director of the local school wanted 20,000 riels to enroll each of the three children in school, and the director of the school they were leaving in Phnom Penh demanded 30,000 riels for each child before he would sign the papers releasing them. Because she could not pay either official, the children dropped out of school for a year. Constant fundraising is another problem for poor families. The Russey School is located on the grounds of a Buddhist pagoda, so there are fundraising festivals virtually every month. "If we know they are going to ask for money," said Mom, "the children stay home." Source: Cambodia 1994a, 86. should impose fees. Rather, school boards should have the formal right to demand fees and other contributions, perhaps subject to a maxi- mum ceiling. This is the de facto situation in many countries, includ- ing those in which free education is a constitutional right. The clauses on free education should be deleted from constitutions, first because free education is not necessarily a good policy and second because in many contexts it is clearly unworkable. POLICY IMPLICATIONS 49 The recommendation that fees be permitted does, however, raise several questions, and requires attention to accompanying frame- works. The first question concerns the safety net for the poorest to ensure that they are not barred from school by an inability to pay fees. The importance of local-level decisionmaking on this matter becomes more apparent. It is at the local level that school boards can know per- sonally the individuals and families who face hardship. Experience in a wide range of contexts shows that school boards can and do exempt poor families from payment of cash, either by proposing contributions of labor and materials instead or in some cases by not even demand- ing those contributions (Bray 1996b; Evans and Rorris 1994). Many school boards also give discounts to families that have more than one child at school. However, experience in various settings has also shown that local school boards are not necessarily sensitive to the needs of the poorest. Opolot (1994) has pointed out that secondary school boards in Uganda are often dominated by local elites who are not particularly concerned about exclusion of the poor. Similar observations have been made about Tanzania by Galabawa (1994) and Kenya by Makau (1985). While provision does exist in Vietnam for exemption from the nation- ally determined secondary school fees, fee remission is not heavily skewed toward the poorest groups. In 1993 only 17.3 percent of lower secondary students from the poorest quintile paid no fees, compared with 14.1 percent in the richest quintile (World Bank 1995g, 197). These facts underscore the need for some consciousness raising, guidance, and monitoring of school-level decisions. Workshops for school man- agers could be a good investment not only for this goal but also for other aspects of administration. One useful model might be that introduced in Nicaragua, where a group of secondary schools entered a special scheme in 1993 (Arnove 1995, 39). Schools were expected to charge monthly tuition fees of 10 c6rdobas ($1.70). The very impoverished, high academic achievers, and children of teachers, demobilized soldiers, and former revolu- tionary combatants were exempt from payment. For these students the government provided a subsidy of 20 c6rdobas a student in lieu of the fee. Such an arrangement perhaps creates a danger of abuse and generates demand costs for documentation and verification, but it gives schools a clear incentive to enroll these particular types of students. Where whole districts are so impoverished that very few families can pay fees, governments must make regional adjustments (Besley 50 COUNTING THE FULL COST and Kanbur 1990). In China, for example, some localities are so poor that they can have little hope of unaided development. In such cases the authorities should aim to transfer resources from richer areas.19 Providing reasonable-quality, fee-free education, primarily paid for out of revenues in other parts of the country is one way to effect such resource transfers. Even poor districts have not-so-poor people who would be capable of contributing to schools, but policymakers may decide that providing fee-free education to such people is acceptable in the effort to achieve the overall goal and reach the most disadvan- taged (Van de Walle and Nead 1995). In Indonesia the World Bank (1993b, 64) has estimated that if subsidies were provided throughout the thirty poorest districts grouped according to their average per capi- ta income, less than 1 percent of the beneficiary families would belong to the richest 20 percent of the national population. This might be con- sidered a modest degree of leakage, especially because the system would cover more than three-quarters of Indonesia's youths in the poorest quintile. Targeting by language group is an alternative approach. Analysts have recommended this strategy in Peru, where 70 percent of Aymara speakers and 32 percent of Quechua speakers are extremely poor and where targeting all Aymara and Quechua speakers would capture 41 percent of the extremely poor (World Bank 1993c, 52). Almost all the East Asian countries covered here have ethnic minorities who are eco- nomically disadvantaged and who might be targeted in the same way. Differential treatment for both geographic regions and language groups can raise political controversies, but such arrangements deserve consideration. Costs of fee collection and accounting also deserve consideration. Before the fee-free policy for junior secondary schools launched in Indonesia in 1994, parents were required to pay fees that were sent from the schools to the national government. Once they reached the government, they were supposed to be converted into grants to cover school supplies. One study, however, showed that the value of the grants that came back to the schools ranged from 35-75 percent of the value of the fees and averaged only 51 percent (King 1994, 9). The bal- ance had been absorbed by administrative expenses and other factors. Fee retention at the school level would cut out this element. Of course, fee retention by schools raises the danger of loss and theft at that level. School boards should be assisted with mechanisms to handle money and to encourage both transparency and accountability. Most school boards are already handling cash and other contributions, POLICY IMPLICATIONS 51 however, so retaining fee income at the school level is not a dramatic change from existing practices. And although central monitoring and support for school boards would demand resources, investment of those resources would reap major dividends in improved administra- tion at the school level. Official private sector policies in two countries are worth nothing. In the Philippines and Thailand the government used to place a cap on private school fees (Arcelo 1991; Jessadachatr and others 1992). The main objective of this cap was to limit the burden on parents and to prevent excessive profiteering by the institutions. The strategy did not work satisfactorily in either case, however, and in both cases the pol- icy was either abandoned or loosened in the early 1990s.20 In Thailand the cap was widely circumvented by demands for tea money or other contributions from parents, and in the Philippines the cap on fees restrained teachers' salaries and prevented low-quality private schools from upgrading themselves. These experiences suggest that the pri- vate sector is best left to regulate its own fees. Household Costs and Elasticity of Demand Researchers in a range of countries have shown that the poorest groups are more sensitive to changes in the price of schooling than are middle- income and rich groups (Evans and Rorris 1994; Gertler and Glewwe 1989; Ilon and Moock 1991; Mason 1994). One study in Indonesia sug- gested that although abolishing fees would be unlikely to have a significant effect on enrollments at the primary level, for groups in the lowest income quintile abolishing fees at the junior secondary level would raise the transition rate from primary school by 12.2 percentage points, bringing it to 58.3 percent (King 1995, 27). At the senior sec- ondary level, the study suggested, abolition of fees would raise the transition rate by 7.6 percentage points, bringing it to 33.2 percent. Such studies underscore the desirability of ensuring that the poorest households do not have to pay fees, especially if those fees are in cash and do not allow for inputs of labor or materials instead. Several coun- tries in the region covered by this study, especially those that have moved from socialist to capitalist economic systems, have experienced sharply increased poverty during the decade. In Mongolia, for ex-ample, the proportion of households below the poverty line was estimated to have increased from 16 percent in March 1992 to 26 percent two years later (World Bank 1995b, 4).21 In the countries with long-standing capi- talist systems-Indonesia, the Philippines, and Thailand-important 52 COUNTING THE FULL COST reductions in poverty have been achieved during recent decades, but even there the pace of improvement has been less satisfactory than had been hoped (World Bank 1995c, 3).22 Moreover, a recent estimate using comparable measures of dietary and nonfood consumption found that rural poverty in the Philippines was lower than in Thailand but more than half again higher than in China, and a third higher than in Indonesia. Urban poverty, in contrast, was higher than in Thailand but lower than in China and Indonesia (World Bank 1995c, 6). As both the proportions and the direction of change in the numbers of the poorest may vary in different countries, all the countries covered by this study feel the challenges of absolute poverty. The timing of fees also determines households' ability to make pay- ments. Studies in various parts of the world have pointed out that demands for fees may be made at inconvenient times of the year. In the Gambia, for example, fees are required at the beginning of the school year even though the months September to November are when farm- ers are poorest because they have sold their cash crops, consumed most of what food they have stored, and are waiting to harvest in December (Gambia 1995, 45). Similar observations have been made in Zambia, and they are likely also to apply to parts of the East Asian region covered here (World Bank 1994b, 182). Governments that do have centrally imposed fees could alleviate the burden on households by being flexible in the timing of their demands. In most settings, however, official fees are only a fairly small portion of total household expenditures on education, and even supplemen- tary school-level fees do not raise that proportion to enormous levels. This implies that removing other costs could have an even greater effect. Table 14 presents information from Indonesia and Vietnam. The Indonesian figures show that fees of all kinds comprised 16.8 percent of total household expenditures on primary education in 1992 and rose to 26.3 percent in upper secondary education. The figures for the low- est income decile were 16.9 percent at the primary level and 37 percent at the upper secondary level.23 The Vietnamese figures show that fees of all kinds consumed an average of 4.3 percent of expenditures at the primary level, rising to 19.3 percent in upper secondary education. Respective figures for the lowest income quintile were 2.3 and 23.7 percent. Thus even if governments were to succeed in prohibiting fees of all kinds, the costs borne by households would not have been reduced to anything like zero. And were opportunity costs to be added to the equation, the proportionate reduction in the burden on house- holds would have been even smaller. POLICY IMPLICATIONS 53 Table 14. Indonesia and Vietnam: Proportions of Total Household Education Expenditures Devoted to Fees of All Kinds (percent) Lower Upper Primary secondary secondary Indonesia (public and private systems), 1992 All households 16.8 24.4 26.3 Lowest income decile 16.9 34.2 37.0 Vietnam (public system), 1993 All households 4.3 14.6 19.3 Lowest income quintile 2.3 26.4 23.7 Source: Indonesia 1992; World Bank 1995g, 187,191. Given this situation, governments that really want to help the poor- est households may need to take more drastic steps. The challenge is to find a measure or combination of measures that on balance best meet the need (Van de Walle 1995). One approach, particularly for rural communities, would be to construct schools nearer poor com- munities. This would cut both the direct expenses of travel and the opportunity costs of the children's time while traveling. It could also reduce the children's fatigue helping them to concentrate on their lessons. King's (1995) simulations in Indonesia suggested that for the lowest income quintile, constructing new schools so that no child would have to travel more than one kilometer would increase the tran- sition rate from primary to junior secondary school by 6.5 percentage points. This was only half the predicted effect of abolishing fees but was nevertheless substantial. For enrollment in primary schools the effect would be almost negligible, but at the senior secondary level the predicted effect would be to raise transition rates by 3.8 percentage points, which was again about half the impact that was predicted from abolishing fees. Operating numerous small schools would of course raise questions about class size and unit costs. At the senior secondary level, it would also raise questions about the number of specialist options that could be offered in each school, but it might be a strategy worth considering in some settings. Another strategy would be to provide subsidies in cash or in kind (for example, through free food, books, materials, and uniforms) to off- set the direct and opportunity costs. A scholarship project for female secondary students in Bangladesh, which has been considered suc- cessful since its launch in 1977, is one example of such a measure (Herz and others 1991, 44). In this system, cash payments are made only if 54 COUNTING THE FULL COST students enroll and stay in schools for the duration of the year. Payments are made through banks, and recipients have thus become familiar with the banking system. Recipients can be identified through income tests or, where the culture permits, through affidavits of pover- ty.24 Such targeting can cause political opposition when individual households or children are treated differently from others, which caused modification of the Bangladesh project so that subsidies were given to all girls in target areas rather than just to the poorest (World Bank 1993b, 64). The fact that the poor require special consideration is not an argu- ment for making education free for everybody. Even less is it an argu- ment for prohibiting households from making contributions to the costs of schooling. As noted at the outset of this study, evidence from a range of countries suggests that at least some households are willing to make extra payments to reduce the distances that children travel to school or to improve quality. Indeed, given that the demand for schooling among rich households is much less elastic than among poor families, there may be a case for increasing the fees and other contri- butions demanded from rich families (Mingat and Tan 1986; World Bank 1993b, 69). Some school boards already impose greater levies on more prosperous families and solicit donations from those willing to pay for them. Such practices could be encouraged to generate further resources for education. Socioeconomic, Regional, and Urban-Rural Imbalances One side effect of these proposed measures is that the gap between prosperous communities and poor communities is likely to widen. Even if the authorities successfully operate fee exemption and subsidy schemes for the poorest, the fact that richer households are permitted or encouraged to inject resources to improve the quality of access to education is likely to widen disparities. In the past a dominant philosophy in socialist societies was that equal poverty was preferable to a situation in which some were pros- perous while others were poor. This view is no longer held so strong- ly, and even the governments in China, Lao People's Democratic Republic, and Vietnam are much more tolerant of inequalities than they used to be. The new philosophy seems to be that even if some communities are poor, that is not a reason everyone should be poor. Nevertheless, unbridled exacerbation of regional inequalities is not to be recommended because it sows seeds of future national instability. POLICY IMPLICATIONS 55 Governments should use other measures to assist disadvantaged areas and to reduce imbalances where they can. Such measures may need to rely to a large extent on increased taxation revenues from more pros- perous socioeconomic groups and regions. The alternative to a school system so heavily financed by user charges is a system primarily financed through taxation revenues. The difficulty for many of the countries covered by this study is that the mechanisms for extracting taxation revenues and for redistributing resources are weak. This is par- ticularly the case in the societies making a transition from socialism (Wong and Zahid 1996). But improvement m such mechanisms would be one long-term mechanism to alleviate the hardships imposed by the current system. The same is true for urban-rural imbalances. Some governments exploit the fact that people in rural areas are likely to have stronger community bonds, demanding greater community contributions from rural than from urban areas. This does not seem reasonable. If people in towns are less able or willing to contribute labor and materials to schools, a case can be made for demanding cash contributions instead. Conclusion This study has highlighted the need for much more detailed attention to the cost of schooling incurred by parents and communities. In some societies these costs are greater than even the costs to governments. Quite apart from overt forms of privatization, the growth of household resourcing of public education has been a hidden form of privatization of enormous significance. While this study has primarily focused on nine countries of East Asia, it has clear parallels with experiences in some other parts of the world (Assie-Lumumba 1993; Esquieu and Peano 1994; Mason and Khander 1996; Tilak 1995; UNICEF 1996). Patterns are far from uni- form but one striking finding from this study is that costs to house- holds have increased in long-standing capitalist countries such as the Philippines, as well as in former socialist countries such as Cambodia. The scale of the increase varies widely, but it is significant that in these countries there is an increase at all. Some household and communities have welcomed the opportunities to improve access or upgrade the quality of institutions. Others have not welcomed the demands on them but have nevertheless been willing to respond. Yet others found the burdens intolerable and have been forced to withdraw their chil- dren from school. Much more research is urgently required. This study has presented some empirical findings, but they are in need of implementation. Both broad surveys and detailed case studies are required, focusing not only on the direct costs of schooling but also on the opportunity costs. Such studies should examine all levels and types of education and should identify variations in particular societies at different points of time. Such research should look at the different units of decisonmaking, including communities, households, individuals, local governments, school com- mittees, and central governments. The 1990 Jomtien Declaration on Education for All identified the need for new and revitalized partnerships in basic education. Such part- nerships will certainly be an important element for the coming decades for a wide range of reasons. Questions arise, however, on the circum- stances in which governments are prepared to extend partnerships to real decisonmaking or whether the concept is restricted mainly to 56 CONCLUSION 57 mechanisms for sharing the burden of resourcing. The nature of exist- ing partnerships of course varies widely, even within the region cov- ered by this study. Relationships between the Indonesian government and Islamic private schools are very different from those between the Chinese authorities and the operators of village minban schools. Shaeffer (1991, 1994) has conducted valuable work on such variations, and the present study has endeavored to build on these insights. Even in the absence of further research, some points are already clear. Above all, governments that seek to achieve universal primary education and expanded enrollments in secondary education must consider the costs and benefits at the household level. Their resulting policies must focus not only on supply but also on demand for educa- tion. Included in demand will be complex considerations of the quality as well as the price of education. When assessing the cost side of the equation, policy analysts must count the full cost, not only to govern- ments but also to parents and communitites, and not only the monetary costs but also the costs of donated labor, materials, and land. Country Data Table A.1. China: Direct Private Costs of Education in Shaanxi and Guizhou Provinces, 1988 (yuan per student) Primary Lower secondary Upper secondary Urban Rural Urban Rural Urban Rural Shaanxi Province Tuition 0 0 0 0 12 10 Other school fees 7 5 18 14 50 40 Textbooks 21 16 43 30 30 30 Writing supplies 8 5 15 10 20 15 Miscellaneous 15 15 20 20 25 25 Total' 51 41 96 74 137 120 Guizhou Province Tuition 2 1.5 2.5 2 3.5 2.5 Other school fees 7 7 13 12 52 44 Textbooks 24 10 40 35 40 40 Writing supplies 8 5 15 10 20 15 Miscellaneous 15 15 20 20 25 25 Totala 56 39 91 79 141 127 a. Does not include boarding costs in some secondary schools, which averaged about 150 yuan a year. Totals are rounded up. Source: Tsang 1994a, 303. 58 COUNTRY DATA 59 Table A.2. China: Average Fees in Selected Parts of Guangdong Province, 1992-93 (yuan per student) Primary schools Junior high schools Fee Liannan Yunfo Zhongshan Liannan Yunfo Zhongshan Miscellaneous 14.34 43.99 139.25 16.00 140.22 286.68 Community 5.10 9.37 50.22 3.33 16.17 81.40 education Class association 1.87 2.39 6.02 1.83 13.63 13.18 Capital construction 26.49 21.12 n.c. 20.00 2.35 n.c. Physical checkup 2.00 2.12 2.36 12.00 2.13 4.06 Books 18.81 25.75 34.68 30.00 69.58 108.16 Personal insurance 4.29 2.93 8.33 4.33 3.15 16.71 Water and electricity 7.63 5.74 3.24 5.66 n.c. n.c. Fuel 20.49 n.c. n.c. 15.50 n.c. n.c. Teaching instruments 4.74 2.61 12.73 7.67 2.84 n.c. School desks 4.72 n.c. 58.06 5.00 34.84 71.81 Other 6.33 n.c. n.c. 5.00 n.c. n.c. n.c. Not charged. Note: These figures reflect the average only for those schools that charged the specific fee. Source: West 1995b, 32, 33. 60 COUNTING THE FULL COST Table A.3. Indonesia: Average Household Education Expenditures by Location, 1989 (rupiah per student) junior Senior Primary secondary secondary Urban Public schools 7,300 17,200 26,200 Private schools 13,700 20,900 33,400 Rural Public schools 3,600 11,200 19,200 Private schools 3,900 11,700 23,100 Average Public schools 4,500 13,700 23,000 Private schools 8,000 15,700 29,100 Source: King 1994, 11. Table A.4. Indonesia: Average Household Education Expenditures by Income Decile, 1992 (rupiah per student) Income Junior Senior decile Primary secondary secondary 1 24,216 70,648 104,234 2 30,496 80,047 125,422 3 35,611 92,384 143,746 4 39,822 102,421 145,970 5 45,849 116,298 169,252 6 52,562 128,146 186,199 7 58,075 136,358 197,855 8 69,939 161,987 223,848 9 89,056 188,076 265,890 10 165,973 287,212 395,114 Average 52,369 148,812 251,721 Source: Indonesia 1992. COUNTRY DATA 61 Table A.5. Indonesia: Average Household Expenditures on Education Fees by Income Decile, 1992 (rupiah per student) Allfeesa BP3fees Income Junior Senior Junior Senior decile Primary secondary secondary Primary secondary secondary 1 4,093 24,175 38,517 818 3,195 1,546 2 4,736 25,181 41,585 1,183 3,257 4,216 3 5,436 26,129 46,595 1,144 3,015 4,235 4 5,681 27,132 46,934 1,225 2,783 3,903 5 6,781 29,863 50,960 1,309 2,973 3,614 6 7,876 31,699 52,926 1,454 3,180 4,386 7 9,015 33,727 54,958 1,626 3,372 5,692 8 11,612 37,300 59,410 1,731 2,925 4,628 9 16,128 43,440 70,084 2,202 3,555 4,713 10 35,466 63,713 94,244 3,819 5,242 5,893 Average 8,782 36,314 66,296 1,473 3,400 4,874 a. BP3, registration, entry, and tuition fees. Source: Indonesia 1992. Table A.6. Indonesia: Average Household Expenditures on Uniforms, Textbooks, and Learning Materials by Income Decile, 1992 (rupiah per student) Uniforms Textbooks and learning materials Income Junior Senior Junior Senior decile Primary secondary secondary Primary secondary secondary 1 6,887 14,119 19,256 4,234 9,489 10,319 2 8,361 14,388 18,630 5,431 10,877 13,795 3 9,119 16,038 18,527 6,160 12,947 15,585 4 9,968 16,258 19,929 7,212 14,149 15,852 5 10,839 18,338 21,016 8,023 15,337 17,903 6 11,784 19,270 23,529 9,280 17,977 20,854 7 12,226 19,546 23,251 9,951 18,071 22,154 8 13,530 20,921 25,024 11,448 21,004 24,811 9 15,237 22,907 27,157 13,971 22,894 26,864 10 20,705 28,627 32,022 21,316 31,490 36,834 Average 10,977 19,977 25,757 8,595 18,833 25,651 Source: Indonesia 1992. 62 COUNTING THE FULL COST Table A.7. Indonesia: Average Household Expenditures on Transport and Pocket Money by Income Decile, 1992 (rupiah per student) Transport Pocket money Income Junior Senior Junior Senior decile Primary secondary secondary Primary secondary secondary 1 67 3,019 10,577 6,877 11,216 10,675 2 217 4,535 17,174 9,101 14,968 22,040 3 165 8,096 19,758 11,785 18,833 24,203 4 348 10,359 21,952 13,197 22,728 24,139 5 632 11,823 26,846 15,555 27,502 31,961 6 792 12,993 31,143 18,148 30,953 35,188 7 1,045 14,994 29,719 20,636 34,258 42,724 8 2,168 21,498 40,276 24,584 42,023 47,094 9 4,607 26,841 49,649 30,644 49,866 61,721 10 21,061 42,463 65,136 48,249 78,381 103,828 Average 1,952 18,047 42,390 17,166 37,126 57,757 Source: Indonesia 1992. Table A.8. Indonesia: Average Annual Household Expenses by Item, Level of School, and Public or Private Ownership, 1992 (rupiah per student) Primary Junior secondary Senior secondary Public Private Public Private Public Private Registration or entry fee 919 3,756 7,323 11,046 13,595 21,864 Tuition 4,112 18,145 20,845 39,540 39,854 71,493 BP3 contributions 1,420 1,916 3,579 2,995 4,864 4,042 Examination fees 18 18 113 162 617 691 Other school fees 1,698 2,634 6,298 5,891 10,426 11,189 Uniform expenses 10,810 12,203 20,148 19,188 26,080 24,164 Textbooks 2,807 4,832 8,291 8,997 12,359 12,391 Study materials 5,452 6,484 10,225 9,643 12,457 12,298 Pocket money 16,463 22,954 34,907 40,729 52,505 59,166 Transport 1,222 8,507 17,652 18,288 43,643 42,389 Miscellaneous 50 86 249 263 671 819 Additional courses 154 1,030 1,080 2,229 3,373 4,423 Others 3,263 4,505 7,695 9,385 12,713 14,877 Total 48,388 87,070 138,405 168,356 233,157 279,806 Source: Indonesia 1992. COUNTRY DATA 63 Table A.9. Indonesia: Average Annual Household Expenses by Affiliation and Ownership, 1992 (rupiah per student) Public Private Religious 43,231 56,988 Nonreligious 48,518 130,437 Junior secondary Religious 123,280 133,545 Nonreligious 139,634 188,150 Senior secondary Religious 177,921 209,653 Nonreligious 236,660 293,887 Source: Indonesia 1992. 64 COUNTING THE FULL COST Table A.10. Vietnam: Average Household Expenditures on Uniforms, 1993 (dong per student) Poorest Richest All I II III IV V Vietnam Urban Rural Public Primary 3,400 5,400 9,200 15,800 32,300 12,200 27,600 9,200 Lower secondary 5,700 10,600 20,100 32,900 63,600 34,000 63,200 23,200 Upper secondary 0 33,900 20,300 57,200 77,700 59,000 83,900 34,700 Semipublic Primary 1,100 8,400 400 5,300 12,800 4,700 13,300 2,100 Lower secondary 30,000 25,000 30,000 62,500 97,100 64,300 81,300 1,000 Upper secondary n.a. 0 0 50,000 51,700 40,400 53,800 22,500 n.a. Not applicable. Source: World Bank 1995d, 188. Table A.11. Vietnam: Average Household Expenditures on Books, 1993 (dong per student) Poorest Richest All I I IIl IV V Vietnam Urban Rural Public Primary 17,900 19,800 22,100 25,900 35,100 23,600 34,600 21,500 Lower secondary 32,900 39,400 44,600 51,700 59,900 49,300 59,000 45,700 Upper secondary 83,900 73,100 71,200 77,700 82,000 78,900 78,300 79,600 Semipublic Primnary 9,300 10,500 10,600 26,800 21,000 15,100 24,400 12,000 Lower secondary 95,000 35,000 34,000 73,800 52,400 61,200 75,300 16,000 Upper secondary n.a. 0 0 75,000 66,700 53,600 71,300 30,000 n.a. Not applicable. Source: World Bank 1995d, 189. COUNTRY DATA 65 Table A.12. Vietnam: Average Household Expenditures on Transport, 1993 (dong per student) Poorest Richest All I II III IV V Vietnam Urban Rural Public Primary 0 100 100 300 700 200 600 100 Lower secondary 100 100 900 1,800 6,700 2,700 4,200 2,100 Upper secondary 0 1,100 700 5,500 4,400 4,800 2,300 7,100 Semipublic Primary 0 0 0 0 15,000 1,800 7,900 0 Lower secondary 0 0 12,500 0 0 1,200 0 5,000 Upper secondary n.a. 0 0 0 1,100 700 1,300 0 n.a. Not applicable. Source: World Bank 1995d, 189. Table A.13. Vietnam: Average Household Expenditures on Food and Lodging, 1993 (dong per student) Poorest Richest All I II III IV V Vietnam Urban Rural Public Primary 3,000 7,500 15,600 27,800 76,300 23,500 54,300 17,500 Lower secondary 2,400 7,400 17,900 27,500 91,000 39,600 75,800 26,400 Upper secondary 0 600 10,900 112,000 96,200 81,100 77,200 84,800 Semipublic Primary 3,100 4,700 24,200 49,400 132,400 34,500 30,900 35,500 Lower secondary 0 0 25,000 22,500 205,700 79,500 90,000 46,000 Upper secondary n.a. 0 0 0 271,100 174,300 92,500 283,300 n.a. Not applicable. Source: World Bank 1995d, 190. Table A.14. Vietnam: Average Household Expenditures on All Schooling Items, 1993 (dong per student) Poorest Richest All I II III IV V Vietnam Urban Rural Public Primary 35,100 45,000 63,400 91,900 214,900 82,900 199,600 61,100 Lower secondary 73,800 94,700 127,300 180,000 359,600 201,100 357,900 144,900 Upper secondary 160,700 253,200 222,900 420,400 537,200 431,300 541,700 327,500 Semipublic Primary 28,600 45,600 66,000 140,800 224,900 91,400 133,100 78,000 Lower secondary 189,500 113,500 395,000 365,500 564,200 414,300 392,800 469,400 Upper secondary n.a. 0 60,000 224,800 731,600 511,000 493,600 534,200 n.a. Not applicable. Source: World Bank 1995d, 191. 66 COUNTING THE FULL COST Table A.15. Thailand: Household Resources for Primary Education by Type of School, 1987 (baht) Government All ONPEC municipal Bangkok government Private schoolsa schools schools schools schools Direct private costs (baht per student per year) Nontuition costs Uniform 24.6 37.8 39.3 26.6 46.9 School bag 50.8 108.0 134.7 61.9 158.7 Textbooks 107.0 195.7 236.7 123.5 340.7 Writing supplied 150.0 215.0 182.3 156.6 284.6 Transport 14.9 66.7 164.3 32.2 325.9 School fees 63.2 204.1 146.0 78.2 33.3 Shoes and sportswear 189.1 367.3 330.4 213.0 448.0 Subtotal 599.6 1,194.6 1,233.3 692.0 1,638.1 Tuition 0.0 0.0 0.0 0.0 1,417.1 Total (X) 599.6 1,194.6 1,233.3 692.0 3,055.2 Household contribution (baht per household per year) To school in cash 92.8 40.2 125.9 92.5 31.4 To school in kind 54.0 39.3 48.6 52.7 14.6 To teachers in cash 2.4 2.1 3.7 2.5 2.0 To teachers in kind 5.5 5.6 2.8 5.3 13.8 Total (Y) 154.7 87.2 181.0 153.0 61.8 Indirect private cost (baht per year per student) (Z) 536.3 461.0 285.9 508.1 450.7 Total private resources (X+Y+Z in baht per year) 1,290.6 1,742.8 1,700.2 1,353.1 3,567.7 a. Office of the National Primary Education Commission. ONPEC schools are mostly in rural areas. Source: Tsang 1994b, 220. COUNTRY DATA 67 Table A.16. Thailand: Household Resources for Primary Education by Family Background and Type of School, 1987 (baht) Total direct Total household Indirect private cost private cost (baht per contribution (baht per student per year) (baht per year) student per year) Government Private Government Private Government Private schools schools schools schools schools schools Household income (baht per month) Less than 650 419 58 614 651-1,600 483 81 660 1,601-3,000 654 2,048 83 44 495 434 3,001-6,000 810 2,798 164 22 480 434 6,001-10,000 900 3,394 76 100 301 337 More than 10,000 1,087 3,025 2,056 109 457 507 Father's education No education 364 43 * 531 * Some primary 609 2,258 86 32 452 268 Primary graduate 554 2,691 146 58 532 573 Lower secondary 821 2,844 187 46 400 260 Upper secondary 1,165 3,184 270 142 361 419 Diploma level 789 2,886 208 20 550 638 Bachelor and higher 1,227 2,863 365 35 643 601 Father's occupation Agriculturalist 497 1,523 169 69 552 1,110 Trader 706 3,031 114 86 378 339 Clerk 907 3,056 214 308 574 300 Manual worker 839 2,366 56 12 415 195 Government employee 1,125 3,670 88 26 341 349 Craftsman 779 2,700 90 15 383 563 Professional 1,043 2,462 320 15 767 649 Executive 1,131 3,292 467 89 34 475 Family assets (baht) Less than 25,000 534 2,415 154 26 537 570 25,001-50,000 803 2,851 104 8 548 174 50,001-75,000 791 3,006 82 8 600 494 75,001-100,000 943 2,958 341 79 478 697 More than 100,000 1,052 3,378 218 130 354 557 Entire sample 692 3,055 153 62 508 451 * Subsample too small (five or fewer cases). Source: Tsang 1994b, 222. 68 COUNTING THE FULL COST Table A.17. Thailand: Household Resources for Primary Education as a Share of Household Income, 1987 (baht) Total direct private Total household cost as a share contribution as a share of household income of household income Government Private Government Private schools schools schools schools (1) (2) (3) (4) Household income (baht per month) Less than 650 16.20 * 4.28 651-1,600 3.84 0.70 * 1,601-3,000 2.42 6.69 0.31 0.13 3,001-6,000 1.57 4.84 0.32 0.03 6,001-10,000 0.98 3.24 0.08 0.10 Above 10,000 0.58 1.53 1.11 0.04 Father's education No education 7.14 0.80 0.00 Some primary 5.62 4.59 0.59 0.06 Primary graduate 6.28 4.69 1.61 0.07 Lower secondary 3.71 4.30 1.61 0.10 Upper secondary 2.54 2.77 0.55 0.06 Diploma level 1.18 2.02 0.32 0.01 Bachelor and higher 1.15 2.51 0.22 0.02 Father's occupation Agriculturalist 7.36 6.00 1.90 0.07 Trader 2.14 3.39 0.27 0.13 Clerk 1.40 4.05 0.30 0.09 Manual worker 3.14 4.15 0.20 0.02 Government employee 1.61 4.84 0.12 0.06 Craftsman 2.51 3.67 0.69 0.02 Professional 1.41 2.96 0.26 0.01 Executive 0.85 2.10 0.12 0.04 Family assets (baht) Less than 25,000 6.61 5.36 1.81 0.03 25,001-50,000 3.83 3.66 0.39 0.01 50,001-75,000 2.47 3.31 0.19 0.004 75,001-100,000 4.41 3.88 0.97 0.08 More than 100,000 2.53 3.08 0.38 0.11 Entire sample 5.79 3.88 1.46 0.06 * Subsample too small (five or fewer cases). Source: Tsang 1994b, 223-24. COUNTRY DATA 69 Total private Indirect private resources as a share cost as a share of of household income household income Government Private Government Private schools schools schools schools (7) (8) (5) (6) [1+3+51 [2+4+61 Household income (baht per month) Less than 650 26.20 46.48 651-1,600 5.43 9.97 1,601-3,000 1.90 1.42 4.63 8.24 3,001-6,000 0.97 0.73 2.86 5.60 6,001-10,000 0.33 0.35 1.39 3.69 Above 10,000 0.25 0.26 1.94 1.83 Father's education No education 12.05 19.99 * Some primary 8.27 0.74 14.48 5.39 Primary graduate 7.86 2.23 15.75 6.99 Lower secondary 3.09 0.48 8.41 4.88 Upper secondary 0.91 0.41 4.00 3.24 Diploma level 0.74 0.73 2.24 2.76 Bachelor and higher 0.72 0.50 2.09 3.03 Father's occupation Agriculturalist 10.13 8.82 19.39 14.89 Trader 1.09 0.31 3.50 3.83 Clerk 1.02 0.31 2.72 4.45 Manual worker 1.92 0.38 5.26 4.55 Government employee 0.84 0.65 2.57 5.55 Craftsman 1.44 0.49 4.64 4.18 Professional 1.30 0.73 2.97 3.70 Executive 0.01 0.25 0.98 2.39 Family assets (baht) Less than 25,000 9.13 2.62 17.55 8.01 25,001-50,000 2.72 0.31 6.94 3.98 50,001-75,000 1.90 0.28 4.56 3.59 75,001-100,000 1.28 1.00 6.66 4.96 More than 100,000 1.49 0.52 4.40 3.71 Entire sample 7.22 1.14 14.47 5.08 70 COUNTING THE FULL COST Table A.18. Philippines: Household Education Expenditures by Public or Private Institutions, 1992-93 (pesos) Elementary Secondary Type of Public Private Public Private expenditure Amount number Amount number Amount number Amount number Tuition and fees 159 2,319 3,207 232 389 904 2,766 437 PTA 24 2,496 56 111 29 870 51 216 Other school fees 87 1,802 432 100 138 696 545 152 Educational materials 80 1,466 846 209 168 628 677 362 School supplies 319 2,690 691 235 427 947 611 413 Other expenses 182 1,694 939 97 246 639 738 144 Clothing 752 2,657 1,267 237 1,036 964 1,363 413 Transport 670 1,635 1,948 166 1,050 820 1,908 365 Board and lodging 451 1,222 5,663 30 1,158 478 5,903 77 Source: borromeo 1995. Table A.19. Cambodia: Household Education Expenditures in Selected Communities, 1992 (riels) Average yearly Share of education expenditure on Total annual in total Name of village children's education expenditure expenditure (%) Chamkar 19,464 164,433 11.8 Srai 14,419 216,880 6.6 Cheay Krong 20,805 428,344 4.9 Note: Chamkar is near Phnom Penh; Srai is near the border of Kandal and Takeo Provinces; and Cheay Krong is a suburb of Phnom Penh. Source: Ledgerwood 1992, 62; quoted in Tilak 1994b, 171. COUNTRY DATA 71 Table A.20. Cambodia: Family Education Costs per Student per Year, Main Urban Centers of Selected Provinces, 1993 (thousands of riels) Textbooks and Contributions Private materials Uniforms to school Transport coaching Total Primary schools Battambang 20 50 .. .. 100 170 Banteay Meanchey 50 50 20 120 PhnomnPenh 50 100 1 40 191 Kampot 30 100 35 165 Kratie 7 36 3 30 .. 76 Siem Reap 30 30 2 60 100 222 Average 31 61 7 45 69 157 Secondary schools Phnom Penh 80 195 7 250 120 652 Svay Rieng 140 135 2 25 .. 302 Siem Reap 40 50 3 300 393 Average 87 127 4 138 210 449 .. Negligible. Source: Carmbodia 1994b. Table A.21. Cambodia: Household Education Expenditures per Student, 1994 (thousands of riels) Fees and Books and Private Level of school other charges stationery tuition Uniforms Transport Others Total Pre-primary (2) 5.8 1.5 0.0 2.5 0.0 37.5 47.3 Primary (82) 50.4 9.9 18.6 31.0 6.1 7.4 123.4 Lower secondary (23) 102.5 25.3 46.0 58.9 6.7 8.3 247.7 Upper secondary (11) 213.9 28.8 82.1 63.3 35.6 0.1 423.8 Higher (8) 112.5 214.3 82.9 230.8 48.8 0.0 689.3 A11 (126) 77.4 27.2 32.9 51.1 11.4 6.9 207.0 Note: Figures in parentheses Tefer to the nunber of students in each category. Source: Tilak 1994b, 173. 72 COUNTING THE FULL COST Table A.22. Cambodia: Government Expenditures per Student, Selected Provinces, 1992-93 (thousands of riels) Lower Upper Name of province Primary secondary secondary Kratie 22.9 77.0 97.9 Ratanakiri 27.5 0.0 0.0 Phnom Penh 20.8 44.6 41.7 Kompong Thom 21.3 127.0 46.5 Kompong Chhnang 22.9 0.0 217.3 Kampot 21.1 91.0 62.4 Source: Tilak 1994b, 140. Table A.23. Myanmar: Average Household Expenditures for Primary Students, 1993-94 (kyats) Type of expenditure Village A Village B Village C Textbooks 7.5 8.2 29.3 Exercise books 9.2 17.0 56.3 Pen 4.0 4.9 22.4 School fees 11.6 15.0 8.8 Uniform 245.0 186.0 285.2 School bag 33.3 41.2 71.3 Umbrella or rainsheet 30.2 66.9 100.0 Slippers 31.2 28.4 14.8 Pocket money 206.8 119.0 308.3 Contribution to extra teacher 119.9 0.0 0.0 Others 3.1 6.1 19.5 Total 701.8 492.7 915.9 Source: Evans and Rorris 1994, 46, 55, 63. COUNTRY DATA 73 Table A.24. Myanmar: Annual Private Costs for Middle School Students, Mon State and Yangon Division, 1990 (kyats) Books and Bus fare and Uniform stationery pocket money Others Total Mon State Tanbyuzayat 1,210 346 785 81 2,422 Mudon 1,277 217 391 150 2,035 Kyaikmaraw 947 237 533 150 1,867 Mawlamyine 1,093 343 553 150 2,139 Yangon Division Hlegu 935 243 400 6 1,584 Thingangyun 1,162 399 491 16 2,068 Insein 1,010 311 393 7 1,721 S. Okkalapa 1,134 533 887 23 2,577 Sanchaung 1,292 314 1,322 29 2,957 Kamayut 1,618 505 1,557 25 3,705 Dagon 1,306 328 1,762 125 3,521 Botahtaung 1,246 269 1,867 52 3,434 Pazundaung 1,081 312 1,287 56 2,736 Source: Gibson 1992, 9. Table A.25. Myanmar: Annual Private Costs for High School Students, Mon State and Yangon Division, 1990 (kyats) Books and Bus fare and Uniform stationery pocket money Others Total Mon State Tanbyuzayat 1,379 311 932 141 2,763 Mudon 1,430 339 1,043 111 2,923 Kyaikmaraw 1,227 341 1,190 155 2,913 Mawlamyine 1,288 343 762 80 2,473 Yangon Division Hlegu 1,035 260 829 147 2,271 Thingangyun 1,000 250 800 .. 2,050 Insein 1,368 244 1,062 31 2,705 S. Okkalapa 1,351 262 1,062 67 2,742 Sanchaung 1,823 344 1,233 31 3,431 Dagon 1,278 495 1,810 200 3,783 Botahtaung 1,116 270 1,327 170 2,883 Pazundaung 1,349 317 1,154 347 3,167 .. Neghgible. Source: Gibson 1992, 8. 74 COUNTING THE FULL COST Table A.26. Mongolia: Per Capita Household Education Expenditures, 1995 (tugrics per month) Tuition Books Rooms Quintile andfees and supplies Clothing Transport and meals Other Total I 1.3 20.7 70.6 2.6 4.4 3.9 103.5 II 1.9 29.6 90.2 0.8 12.6 7.2 142.3 III 11.7 38.3 127.0 14.0 30.7 12.2 233.9 Iv 43.0 37.9 145.8 13.0 44.8 16.9 301.4 V 212.3 65.1 176.8 29.2 51.9 52.4 587.7 Mongolia 54.0 38.3 122.1 11.9 28.9 18.5 273.8 Source: World Bank 1996. Table A.27. Mongolia: Per Capita Household Education Expenditures as a Share of Per Capita Nonfood Expenditure, 1995 (percent) Tuition Books Rooms Quintile and fees and supplies Clothing Transport and meals Other Total I 0.1 1.5 5.1 0.2 0.3 0.3 7.5 II 0.1 1.2 3.6 0.0 0.5 0.3 5.7 III 0.3 1.0 3.4 0.4 0.8 0.3 6.2 IV 0.8 0.7 2.7 0.2 0.8 0.3 5.6 V 1.9 0.6 1.6 0.3 0.5 0.5 5.3 Mongolia 0.6 1.0 3.3 0.2 0.6 0.3 6.1 Source: World Bank 1996. Notes 1. The author's previous work was chiefly on projects for the Asian Development Bank, the United Nations Educational, Scientific, and Cultural Organization (UNESCO), and the United Nations International Children's Education Fund (UNICEF). 2. Each primary child does have to pay a universal primary education fee of 200 Tanzanian shillings, but this is considered little more than a token. In most parts of the country the fees is equivalent to the price of a bottled soft drink. 3. Myanmar was also a socialist state between 1962 and 1988, though it followed its own brand of socialism and was less affected by the collapse of the Soviet Union and other wider transitions in the early 1990s. The ideology was officially known as the "Burmese Way to Socialism." 4. In 1989-90 there were 905 affiliated primary schools, compared with 1,259 the year before; 968 middle schools, compared with 1,322 the year before; and 220 high schools, compared with 342 the year before. Affiliated primary schools in 1989-90 com- prised 2.6 percent of the total number of primary schools. Respective figures for middle and high schools were 34.3 and 21.9 percent. 5. One study in Guizhou Province (table A.1), however, noted that in 1988 at least some primary and junior secondary pupils were charged specifically named tuition fees. Whether such charges have continued could be answered only through field work, but in any case so many other fees may be demanded that the actual nomenclature may not seem too important to the people concerned. Among the fees charged that are unique to China is a penalty charge on parents who have more than one child when they are not supposed to. 6. The survey, known as SUSENAS (Survei Sosial Ekonomi Nasional), is conducted every two or three years by the Indonesian government. The 1989 survey covered 148,000 individuals in 32,700 households; the 1992 survey covered 356,600 individuals in 66,700 households. Mason (1994, 67) found some lower secondary schools with even higher registration fees in 1989. The highest fee that he identified was 143,000 rupiah. 7. Neither the Indonesian nor the Vietnamese data included estimation of contribu- tions of labor and materials. Had they done so, the figure on total value of inputs would have been higher, and the distributions across income groups and across the urban-rural division would have looked more balanced. 8. One study of Zambia (World Bank 1994b, 182) noted that its regulations are very constraining. For instance, regulations require children to wear black leather shoes, not just shoes or black shoes. The uniforms are made by a parastatal agency and are strictly enforced by some (especially urban) schools to the extent that children without socks or the appropriate shoes are sent home. Although this seems an extreme case, similar pat- terns prevail in some of the East Asian countries covered by the present study. 9. However, some figures on costs of private tuition in Myanmar were reported sep- arately and are summarized in box 3. The figures were only for Yangon Division, where the average paid by those middle school students who took private tuition (65.6 percent of the total sample) was 1,088 kyats. In the high school sample 90.7 percent of students in the sample took private tuition and paid an average of 2,556 kyats. Thus inclusion of 75 76 COUNTING THE FULL COST private tuition would have substantially increased the unit costs to households shown in table 9. Allowing for the fact that some students took no private tuition, which would therefore have decreased average payments, household costs at middle school would have increased from 2,521 to 3,235 kyats (an increase of 28.3 percent), and those at high school would have increased from 2,842 to 5,158 kyats (an increase of 81.5 percent). 10. In 1985,18 percent of primary and secondary pupils were boarders. By 1992, part- ly because of the newly imposed direct costs of boarding, this proportion had fallen to 9.6 percent (Mongolia 1993). 11. Both these estimates understate the full opportunity cost because they do not cap- ture the value of home production. In general, failing to include the contribution made at home will lead to greater understatement of girls' than boys' opportunity costs because girls tend to spend more time in home-related activities. Mason (1994, 66, 90) also notes that in general, wage labor markets in the poor, dryland, and upland regions of rural Java operate only in the morning, which is also when schools operate. In the more prosperous irrigated rice areas, markets for agricultural wage labor operate in both the mornings and the afternoons. During Mason's field work he identified one area in East Java in which the government junior secondary school convened in the afternoon. This was also an area in which the market for agricultural wage labor existed primarily in the mornings. Because the school convened in the afternoons, children from poor households could work before coming to school, thereby significantly reducing their opportunity costs. Mason's example highlights the importance of taking regional factors into account when designing and operating school systems. 12. Table 10 in fact underestimates household inputs. It excludes the cost of uniforms, meals, and boarding; nonmonetary inputs of labor, material, and land; and private sup- plementary tuition outside school hours. The calculation for government inputs does include an estimate of the government's management and operational support (curricu- lum development, examinations, inspections, and so on). 13. The figures also omit expenditures on supplementary tutoring, which would have been higher in urban areas. However, supplementary tutoring was less important at the primary than at the secondary level. 14. These figures exclude the opportunity costs of students' time but include the costs of land. If land were excluded from the calculation, the figure for the whole coun- try would be 15.8 percent, while those for Bangkok and rural areas would be 26.8 and 14 percent. 15. To provide some sort of a cross check, it is useful to note that the World Bank (1991, 69) estimated recurrent expenditures met in 1988 by nonbudgeted funds in Shaanxi and Guizhou Provinces. This category included education surcharges and levies, social contributions, overseas contributions, fees, and funds generated by schools. In Shaanxi nonbudgeted funds met 16.9 percent of primary and 25.3 percent of secondary recurrent costs. Respective figures for Guizhou were 7.6 and 13.3 percent. These figures illustrate regional variations within China. They do not include parents' expenditures outside the institutional budgets. It seems likely that proportions have risen since 1988. 16. Methodologically it might be queried if these should be added up in this way. This is partly because columns 1 and 2 in table A.17 refer to direct costs per student, whereas columns 3 and 4 are costs per household (which may have more than one stu- dent). A similar comment applies to table A.15. A further query concerns the indirect (opportunity) cost in columns 5 and 6 of table A.17. Tsang's article, from which the table was taken, does not indicate whether the percentage shown is the indirect cost as a pro- portion of the actual incomes of those households, or as a proportion of the actual NOTES 77 incomes of those households plus the opportunity costs (that is, what the incomes would have been had the children not gone to school). Only the proportion of the actu- al incomes of those households plus the opportunity costs would be fully meaningful; but the percentages in columns 1-4 should be proportions only of the actual household costs. Nevertheless, despite these methodological questions, the table is very instructive. In particular, it clearly shows that the poor were paying considerably greater propor- tions of their incomes for primary education than were the rich. 17. The figures show only expenditures during a period of one month between October and December 1993. They have not been adjusted for seasonal patterns and seem in general to be an underestimate. Moreover, because of security problems the researchers went only to fifteen of the nineteen provinces; and because of budgetary problems they only went to the relatively accessible parts of those provinces (Cambodia 1994c, 3-6). The published report of this survey did not show a breakdown of education expenditures by income groups. 18. The figures on government expenditure are also startling. During 1990-92 real public expenditure on education fell 56 percent. In 1990 the allocation for basic educa- tion services was equivalent to 11.3 percent of GDP, but by 1993 it had dropped to 3.1 percent (World Bank 1995b, 6). 19. The problem in China, however, is that the goals of fiscal decentralization are having the opposite effect of increasing disparities (Chung 195; Wong and West 1996). 20. In the Phihppines the policy was abandoned, whereas in Thailand it was only loosened. Even nonsubsidized private schools remained subject to a de facto fee ceiling, which in 1994-95 was 3,900 baht per pupil per year for lower secondary, 4,300 baht for upper secondary. 21. In March 1994, 6 percent of the total population was classified as being absolute- ly poor, with monthly per capita incomes less than 40 percent of the poverty line (World Bank 1995b, 4). Figures on the absolute poor had not previously been compiled, but informal evidence indicated a sharp increase in this group. 22. Annual reductions in poverty incidence are reported to have been 2 percent in Indonesia for 1970-90, 0.6 percent in the Philippines for 1971-91, and 1.4 percent in Thailand for 1962-68 (World Bank 1995c, 3). The decline in poverty in the Philippines is described as "relatively modest," and has been accompanied by much greater inequali- ties in the distribution of incomes than in the other two countries. 23. These figures, in contrast to those from Vietnam, include both public and private schools. However, the private schools mainly attended by the lowest income groups in Indonesia were religious institutions. Total household expenditures on these private reli- gious institutions were not markedly different from those on public schools (particular- ly nonreligious ones). Indeed at the secondary level per pupil expenditures for the pri- vate religious schools were less than those for public nonreligious schools (table A.9). The private schools probably charged higher explicitly named fees, but they appeared to have lower charges on other items. 24. In Indonesia poor people can request their village headman to issue an affidavit that exempts them from paying health charges. This system is called Surat Lurah. The World Bank (1993b, 84) notes that the system is not used as widely as it should be but nevertheless suggests (64) that the framework could be adapted to assist poor people with school expenses. References Abbai, Belai. 1995. "An Assessment: Financing of Social Services in Mongolia." World Bank, East Asia and Pacific Regional Office, Country Department II, Washington, D.C. Albrecht, Douglas, and Adrian Ziderman. 1995. Financing Universities in Developing Countries. London: Falmer Press. Arcelo, Adriano. 1991. Financing Education for Social and Economic Development in the Philippines. Manila: Congressional Commission on Education. Arnove, Robert F 1995. "Education as Contested Terrain in Nicaragua." Comparative Education Review 39 (1): 28-53. Assi&Lumumba, N'Dri Therese. 1993. Cofits,financement de l'education de base et partici- pation des familles et communautes rurales dans les pays du Tiers Monde. Paris: Institut International de Planification de l'Education. Bashir, Sajitha. 1994. "Public versus Private in Primary Education: Comparisons of School Effectiveness and Costs in Tamil Nadu." Ph.D. thesis, London School of Economics, University of London. Besley, Timothy, and Ravi Kanbur. 1990. "The Principles of Targeting." Policy Research Working Paper 385. World Bank, Research Advisory Staff, Washington, D.C. Bhutan. Government of. 1991. Seventh Five Year Plan (1992-1997). Thimphu: Government of Bhutan Planning Commission. Biak Cin, Robert, and Guy Scandlen. 1988. "Cultural Values and Community Financing of Education in Burma." In Mark Bray with Kevin Lillis, eds., Community Financing of Education: Issues and Policy Implications in Less Developed Countries. Oxford: Pergamon Press. Borromeo, Roberto T. 1995. Cost-Effectiveness of Education in the Philippines: Comparative Analysis of Public and Private School Programs. Manila: Fund for Assistance to Private Education. Bray, Mark. 1987. "Is Free Education in the Third World Either Desirable or Possible?" Journal of Education Policy 3 (2): 119-29. - 1995a. "Community and Parental Contributions to Schooling in Cambodia." United Nations Children Fund (UNICEF), East Asia and Pacific Regional Office, Bangkok. - 1995b. "Community and Parental Contributions to Schooling in the Lao People's Democratic Republic." UNICEF, East Asia and Pacific Regional Office, Bangkok. - . 1995c. The Costs and Financing of Primary Schooling in Bhutan. Thimphu: UNICEF, Division of Education, and Government of Bhutan. - 1996a. Decentralization of Education: Community Financing. Washington, D.C.: World Bank. - 1996b. "Equity Issues in Local Resourcing of Education: Community Financing of Primary Schools in Bhutan." International Review of Education 42 (4). - 1996c. Privatization of Secondary Education: Issues and Policy Implications. Paris: United Nations Educational, Scientific, and Cultural Organization (UNESCO), Division of Secondary Education. 78 REFERENCES 79 Bray, Mark, with Kevin Lillis, eds. 1988. Community Financing of Education: Issues and Policy Implications in Less Developed Countries. Oxford: Pergamon Press. Brooks, Douglas H., and Myo Thant. 1996. "Health and Education Issues in Transition Economies." Paper presented at the Pacific Rim Allied Economic Organizations Conference, January 11-12, Hong Kong. Cambodia, Government of. 1994a. Education Sector Review. Vol. 2A. Manila: Asian Development Bank and Brisbane: Queensland Education Consortium. - 1994b. Rebuilding Quality Education and Training in Cambodia. Phnom Penh: Ministry of Education, Youth, and Sport. 1994c. Report on the Socio-Economic Survey-1993/1994 Cambodia (First Round). Phnom Penh: Ministry of Planning, National Institute of Statistics. CARE International. 1996. "Assessment of Community Participation in School Construction: Lao People's Democratic Republic." Education Development Project CR 2488-LA. World Bank, Washington, D.C. Cheng, Kai-ming. 1995. Minban Teachers in China. Hong Kong: University of Hong Kong, Comparative Education Research Center. Chung, Jae Ho. 1995. "Central-Provincial Relations." In Chi Kin Lo, Suzanne Pepper, and Kai Tsui Yuen, eds., China Review 1995. Hong Kong: Chinese University Press. Colclough, Christopher, with Keith Lewin. 1993. Educating All the Children: Strategies for Primary Schooling in the South. Oxford: Clarendon Press. Cumming, C. E., A. Brock, P. K. Kayiso, and C. Opio-Owalu. 1995. Secondary Education Costs and Finance Study. Cambridge, U.K.: Cambridge Education Consultants and Kampala: Serefaco Consultants. Cummings, William K., and Abby Riddell. 1994. "Alternative Policies for the Finance, Control, and Delivery of Basic Education." International Journal of Educational Research 21 (8): 751-76. Esquieu, Paul, and Serge Peano. 1994. L'enseignement prive et spontane dans le systeme edu- catif Tchadien. Rapport de Recherche 103. Paris: Institut International de Planification de l'Education. Evans, Kiri, and Adam Rorris. 1994. Cost-Effectiveness of Primary Education in Myanmar. Yangon: UNICEF. Fong, Monica S. 1994. "Poverty and Ethnic Minorities in Vietnam." ESP Discussion Paper. World Bank, Education and Social Policy Department, Washington, D.C. Galabawa, J. C. J. 1994. "Characteristics, Financing, Unit Costs, and Selection Issues of Non-Government Secondary School Provision in Tanzania." Paper for the workshop on Investing in Human Capital, April 15-20, Arusha. Gambia, Government of. 1995. 1993-94 Community Education Survey Report. Banjul: Ministry of Finance and Economic Affairs, Central Statistics Department. Gertler, Paul, and Paul Glewwe. 1989. The Willingness to Payfor Education in Developing Countries: Evidence from Rural Peru. LSMS Working Paper 54. Washington, D.C.: World Bank. -_____ 1991. "The Willingness to Pay for Education for Daughters in Contrast to Sons: Evidence from Rural Peru." World Bank Economic Review 6 (1): 171-88. Gibson, U Saw. 1992. "Non-Government Expenditure on Education." Working Paper 4.3. Myanmar Education Research Bureau, Education Sector Study Project, Yangon. Graham-Brown, Sarah. 1991. Education in the Developing World: Conflict and Crisis. London: Longman. Harper, Caroline. 1994. An Assessment of Vulnerable Groups in Mongolia: Strategies for Social Policy Planning. Discussion Paper 229. Washington, D.C.: World Bank. 80 COUNTING THE FULL COST Herz, Barbara, K. Subbarao, Masooma Habib, and Laura Raney. 1991. Letting Girls Learn: Promising Approaches in Primary and Secondary Education. Discussion Paper 133. Washington, D.C.: World Bank. Hill, Martin J. D. 1991. The Harambee Movement in Kenya: Self-Help, Development, and Education among the Kamba of Kitui District. LSE Monographs on Social Anthropology 64. London: Athlone Press, Hossain, Shaikh. 1996. Unpublished data on costs and financing of education in Mongolia. World Bank, Washington, D.C. Igwe, S. 0. 1988. "Community Financing of Schools in Eastern Nigeria." In Mark Bray with Kevin Lillis, eds., Community Financing of Education: Issues and Policy Implications in Less Developed Countries. Oxford: Pergamon Press. Ilon, Lynn, and Peter Moock, 1991. "School Attributes, Household Characteristics, and Demand for Schooling: A Case Study of Rural Peru." International Review of Education 37 (4): 429-51. Indonesia, Government of. 1989, 1992. SUSENAS (Survei Sosial Ekonomi Nasional) Survey. James, Estelle. 1991. "Public Policies toward Private Education: An International Comparison." International Journal of Educational Research 15 (4) 359-76. - 1994. "The Public-Private Division of Responsibility for Education." Internaticnal Journal of Educational Research 21 (8): 777-83. Jessadachatr, P., K. Jiratatprasot, S. Kohengkul, and C. Tantithamthavorn. 1992. "Educational Management: Public and Private Sector Roles in the Provision of Education." In Educational Options for the Future of Thailand: The 1991 Year-End Conference. Bangkok: Thailand Development Research Institute. Jimenez, Emmanuel. 1994. "Human and Physical Infrastructure: Public Investment and Pricing Policies in Developing Countries." Policy Research Working Paper 1281. World Bank, Pohcy Research Department, Washington, D.C. Kemmerer, Frances. 1992. "Community Support of Schools." In David W. Chapman and Herbert J. Walberg, eds., Advances in Educational Productivity. Vol. 2. Greenwich, Conn.: JAI Press. Kenya, Republic of. 1994. National Development Plan for the Period 1994 to 1996. Nairobi: Government Printer. King, Elizabeth. 1994. "Who Pays for Education in Indonesia? The Role of Government and Families." Paper prepared for the workshop on New Strategies for Financing Education and Health in Developing Countries, February 28-March 2, University of Sussex, Institute of Development Studies. 1 1995. "Does the Price of Schooling Matter? Fees, Opportunity Costs, and Enrollment in Indonesia." World Bank, Washington, D.C. Kingdon, Geeta Gandhi. 1994. "An Economic Evaluation of School Management-Types in Urban India: A Case Study of Uttar Pradesh." Ph.D. thesis, University of Oxford. Ledgerwood, Judy. 1992. Analysis of the Situation of Women in Cambodia: Research on Women in Khmer Society. Phnom Penh: UNICEF. Levin, Henry M. 1987. "Education as a Public and Private Good." Journal of Policy Analysis and Management 6 (4): 628-41. Lewin, Keith, with Dieter Berstecher. 1989. "The Costs of Recovery: Are User Fees the Answer?" IDS Bulletin 20 (1): 59-71. Lingard, B., J. Knight, and P. Porter, eds. 1993. Schooling Reform in Hard Times. London: Falmer Press. Lockheed, Marlaine E., and Emmanuel Jimenez. 1994. "Public and Private Schools in Developing Countries." HRO Working Paper 43. World Bank, Washington, D.C. REFERENCES 81 Lockheed, Marlaine E., and Henry M. Levin. 1993. "Creating Effective Schools.' In Henry M. Levin and Marlaine E. Lockheed, eds., Effective Schools in Developing Countries. London: Falmer Press. Makau, B. M. 1985. "Equity and Efficiency in Financing Secondary Education in Kenya: Key Issues in State-Community Partnerships." Working Paper 429. University of Nairobi, Institute of Development Studies. Maravanyika, 0. E. 1995. "Community Financing Strategies and Resources within the Context of Educational Democratization: The Case of Zimbabwe." Paper presented at the conference on Partnerships in Education and Development: Tensions between Economics and Culture, May 24-26, University of London, Institute of Education. Mason, Andrew D. 1994. 'Schooling Decisions, Basic Education, and the Poor in Rural Java." Ph.D. diss., Stanford University. Mason, Andrew D,, and Shahidur R. Khandker. 1996. "Household Schooling Decisions in Tanzania." World Bank, Poverty and Social Policy Department, Washington, D.C. Mingat, Alain, and Jee-Peng Tan. 1986. "Expanding Education through User Charges: What Can be Achieved in Malawi and Other LDCs?" Economics of Education Review 5 (3): 273-86. Moegiadi, and Jiyono. 1991. Research Report on Community Participation in Education at the Primary Level in Indonesia. Jakarta and Paris: Office of Educational and Cultural Research and Development (Balitbang Dikbud), Ministry of Education and Culture, in cooperation with the International Institute for Educational Planning (IIEP). Mok, Ka-ho, and David Chan. 1995. "An Exploration into the Implications for the Development of Private Education in the Pearl River Delta." Paper presented at the International Symposium on Education and Sociopolitical Transitions in Asia, May 29-31, University of Hong Kong, Comparative Education Research Center. Mongolia, Government of. 1993. Mongolia: Education and Human Resource Sector Review. Ulaanbaatar: Ministry of Science and Education and Manila: Asian Development Bank. Moorad, E R. 1992. "Community Management of Schools: Some Useful Insights from Botswana and Kenya." In G. W. Kouwenhoven, S. G. Weeks, and C. Mannathoko, eds., Proceedings of the Comparative Education Awareness Seminar. Gaborone: Botswana Educational Research Association. Murugasu, V. 1993. Encouraging Private Sector Education Study (TA No. 1776-LAO): Final Report. Manila: Asian Development Bank. Myanmar, Government of. 1992. Education Sector Study, Phase 1: Final Report. Yangon: Ministry of Education, United Nations Development Programme (UNDP), and UNESCO. Myers, Charles N., and Chalongphob Sussangkam. 1992. "Education and Economic Development: Issues and Options for Policy and Reform." In Educational Optionsfor the Future of Thailand: The 1991 Year-End Conference. Bangkok: Thailand Development Research Institute. Nam, Jane. 1996. "A Comparative Study of Pai Yao and Han Chinese Junior Secondary School Dropouts in Liannan Yao Autonomous County, Guangdong Province, the People's Republic of China." Ph.D. thesis, University of Hong Kong. Oepen, Manfred, and Wolfgang Karcher. 1988. The Impact of Pesantren in Education and Community Development in Indonesia. Berlin: Friedrich-Naumann Stiftung. Okoye, Mary. 1986. "Community Secondary Schools: A Case-Study of a Nigerian Innovation in Self-Help." International Journal of Educational Development 6 (4): 263-74. Opolot, Jethro A. 1994. "Study on Costs and Cost-Effective Approaches to Primary Education in Uganda." UNICEF, Kampala. 82 COUNTING THE FULL COST Penrose, Perran. 1993. Affording the Unaffordable: Planning and Financing Education Systems in Sub-Saharan Africa. Occasional Papers on Education 7. London: Overseas Development Administration. Pepper, Suzanne. 1995. "Regaining the Initiative for Education Reform and Development." In Chi-Kin Lo, Suzanne Pepper, and Kai Yuen Tsui, eds., China Review 1995. Hong Kong: Chinese University Press. Pring, Richard. 1988. "Privatization." Educational Management and Administration 16 (2): 85-96. Psacharopoulos, George, and Maureen Woodhall. 1985. Education for Development: An Analysis of Investment Options. New York: Oxford University Press. Riddell, Abby R. 1993. "The Evidence on Public-Private Educational Trade-Off in Developing Countries." International Journal of Educational Development 13 (4): 373-86. Robinson, Jean. 1988. "State Control and Local Financing of Schools in China." In Mark Bray with Kevin Lillis, eds., Community Financing of Education: Issues and Policy Implications in Less Developed Countries. Oxford: Pergamon Press. Rorris, Adam, and Kiri Evans. 1994. Towards Universalized Primary Education in Vietnam: A Study of the Cost and Cost-Effectiveness of the Primary Education System. Hanoi: UNICEE Schwartz, Antoine. 1995. The Philippines: Cost and Financing Issues in Education. Washington, D.C.: World Bank. Selden, Thomas M., and Michael J. Wasylenko. 1995. "Measuring the Distributional Effects of Public Education in Peru." In Dominique Van de Walle and Kimberly Nead, eds., Public Spending and the Poor. Baltimore: Johns Hopkins University Press. Shaeffer, Sheldon. 1991. School and Community Collaborationfor Educational Change: Report of an IIEP Seminar. Paris: International Institute for Educational Planning. , ed. 1992. Collaborating for Educational Change: The Role of Teachers, Parents, and the Community in School Improvement. Paris: International Institute for Educational Planning. . 1994. Participation for Educational Change: A Synthesis of Experiences. Paris: International Institute for Educational Planning. Solnon, L. C., and C. L. Fagnano. 1995. "Benefits of Education." In Martin Carnoy, ed., International Encyclopedia of Economics of Education. Second Edition. Oxford: Pergamon Press. Subbarao, K., and Kene Ezemenari. 1995. "Transition, Poverty and Social Assistance in Mongolia." World Bank, Education and Social Policy Department, Washington, D.C. Tan, Jee-Peng, and Alain Mingat. 1992. Education in Asia: A Comparative Study of Cost and Financing. Washington, D.C.: World Bank. Tembon, Mercy Miyang. 1994. "The Financing of Secondary Education in Mezam Division, North West Province Cameroon: An Uneasy Partnership between Family and State?" Ph.D. diss., University of London, Institute of Education. Thobani, Mateen. 1984. "Charging User Fees for Social Services: Education in Malawi." Comparative Education Review 28 (3): 402-23. Thomas, Christopher. 1996. Unpublished data on costs and financing of education in Lao People's Democratic Republic. World Bank, East Asia and Pacific Regional Office, Country Department I, Washington, D.C. Tilak, Jandhyala B. G. 1994a. "Cost Recovery Approaches in Education." Paper prepared for the Workshop on New Strategies for Financing Education and Health in Developing Countries, February 28-March 2, University of Sussex, Institute of Development Studies. . 1994b. Financing Education in Cambodia. Brisbane: Queensland Education Consortium for the Asian Development Bank. REFERENCES 83 - 1995. HIov Free is 'Free' Primary Education in India? Occasional Paper 21. New Delhi: National Institute of Educational Planning and Administration. Togo, Government of. 1994. Analyse de la situation des enseignements du primaire pris en charge par les commnunautes locales. Lome: Ministere de l'Education Nationale et de la Recherche Scientifique. Tsang, Mun C. 1994a. "Costs of Education in China: Issues of Resource Mobilization, Equality, Equity, and Efficiency." Education Economics 2 (3): 287-312. - 1994b. "Government and Private Schools." In W. Steven Barnett, ed., Cost Analysis for Education Planning: Methods and Examples. Adzvances in Educational Productivity Vol. 4. Greenwich, Conn.: JAI Press. - Forthcoming. "Financial Reform of Basic Education in China." Econontics of Education Review. Tsang, Mun C., and Somsri Kidchanapanish. 1992. "Private Resources and the Quality of Primary Education in Thailand." International Journal of Educational Research 17 (2): 179-98. UNDP (United Nations Development Programme). 1995. Human Development Report 1995. New York: Oxford University Press. UNESCO (United Nations Educational, Scientific, and Cultural Organization). 1993. World Education Report 1993. Paris: UNESCO. UNICEF (United Nations Children's Fund). 1996. Public and Private Costs of Primary Education: Evidencefrom Selected Countries in Asia and Africa. New York: UNICEF. Van de Walle, Dominique. 1995. "Incidence and Targeting: An Overview of Implications for Research and Policy." In Dominique Van de Walle and Kimberly Nead, eds., Public Spending and the Poor: Theory and Evidence. Baltimore: Johns Hopkins University Press. Van de Walle, Dominique, and Kimberly Nead, eds. 1995. Public Spending and the Poor: Theory and Evidence. Baltimore: Johns Hopkins University Press. Wang, Ying Jie. 1995. "Minban Schools in China: Trends and Issues." Paper presented at the International Symposium on Education and Sociopolitical Transitions in Asia, May 29-31, University of Hong Kong, Comparative Education Research Center. Watkins, Kevin. 1995. The Oxfam Poverty Report. Oxford: Oxfam. WCEFA (World Conference on Education for All). 1990a. World Conference on Education for All: Final Report. New York: WCEFA Inter-Agency Commission. - 1990b. World Declaration on Education for All. New York: WCEFA Inter-Agency Commission. Weidman, John C. 1995. "Action Plan for Lao PDR Private Sector Education Development Project." TA 2097-LAO. Asian Development Bank, Manila. West, Loraine A. 1995a. "Provision of Public Services in Rural PRC." In Christine P. W. Wong, ed., Financing Local Government in the People's Republic of China. TA 2118-PRC. Manila: Asian Development Bank. - 1995b. "Regional Economic Variation and Basic Education in Rural China." Robert S. McNamara Fellowship Report. World Bank, Washington, D.C. Williams, James H. 1993. "Improving School-Community Relations in the Periphery." In William K. Cummings, ed., Reaching Peripheral Groups: Community, Language and Teachers in the Context of Development. Studies in Comparative Education 31. Buffalo: State University of New York Press. Wong, Christine P. W., and Loraine A. West. 1996. "Fiscal Decentralization and Growing Regional Disparities in Rural China: Some Evidence in the Provision of Social Services." Paper presented at the Pacific Rim Allied Economic Organizations Conference, January 11-12, Hong Kong. 84 COUNTING THE FULL COST Wong, Christine P. W., and Shahid N. Zahid. 1996. "Financing Government during the Transition: Problems and Issues in Public Finance and Intergovernmental Relations." Paper presented at the Pacific Rim Allied Economic Organizations Conference, January 11-12, Hong Kong. World Bank. 1988. "The Philippines: Education Sector Study." Vol. 2. East Asia and Pacific Regional Office, Country Department I, Washington, D.C. - 1989. "Indonesia: Basic Education Study." East Asia and Pacific Regional Office, Country Department III, Washington, D.C. - 1991. "China: Provincial Education Planning and Finance-Sector Study" East Asia and Pacific Regional Office, Country Department 11, Washington, D.C. - 1992. China: Strategies for Reducing Pozverty in the 1990s. A World Bank Country Study. Washington, D.C. - 1993a. "Bolivia Education Sector: Sector Development and International Assistance." Latin America and the Caribbean, Country Department II, Washington, D.C. - 1993b. "Indonesia: Public Expenditures, Prices, and the Poor." East Asia and Pacific Regional Office, Country Department III, Washington, D.C. - 1993c. "Peru: Poverty Assessment and Social Policies and Programs for the Poor." Latin America and the Caribbean, Country Department II, Washington, D.C. - 1993d. Uganda: Social Sectors. A World Bank Country Study. Washington, D.C. - 1994a. "Nepal: Critical Issues in Secondary Education and Options for Reform." South Asia, Country Department II, Washington, D.C. - 1994b. "Zambia Poverty Assessment." Vol. 1. Southern Africa Department, Washington, D.C. - 1995a. "Lao PDR: Social Development Assessment and Strategy." East Asia and Pacific Regional Office, Country Department I, Washington, D.C. - 1995b. "Mongolia: Poverty Alleviation for Vulnerable Groups Project Staff Appraisal Report." East Asia and Pacific Regional Office, Country Department II, Washington, D.C. - 1995c. "Philippines: A Strategy to Fight Poverty." East Asia and Pacific Regional Office, Country Department I, Washington, D.C. - 1995d. Priorities and Strategies for Education: A World Bank Review. Development in Practice series. Washington, D.C. 1995e. Social Indicators of Development. Baltimore: Johns Hopkins University Press. - . 1995f. "Tanzania: Social Sector Review." Eastern Africa Department, Washington, D.C. - . 1995g. "Vietnam: Poverty Assessment and Strategy." East Asia and Pacific Regional Office, Country Department I, Washington, D.C. 1996. "Mongolia: Poverty in a Transition Economy." East Asia and Pacific Regional Office, Country Department II, Washington, D.C. Wu, Kin Bing. 1994. Mongolia: Financing Education during Economic Transition. Discussion Paper 226. World Bank, Washington, D.C. Directions in Development Begun in 1994, this series contains short essays, written for a general audience, often to summarize published or forthcoming books or to highlight current development issues. Africa's Management in the 1990s and Beyond: Reconciling Indigenous and Transplanted Institutions Building Human Capital for Better Lives Class Action: Improving School Performance in the Developing World through Better Health and Nutrition Decentralization of Education: Community Financing Decentralization of Education: Politics and Consensus Decentralization of Education: Teacher Management Deep Crises and Reform: What Have We Learned? Early Child Development: Investing in the Future Financing Health Care in Sub-Saharan Africa through User Fees and Insurance Global Capital Supply and Demand: Is There Enough to Go Around? Implementing Projects for the Poor: What Has Been Learned? Improving Early Childhood Development: An Integrated Program for the Philippines India's Family Welfare Program: Moving to a Reproductive and Child Health Approach (with a separate supplement) Investing in People: The World Bank in Action Managing Commodity Booms - and Busts Meeting the Infrastructure Challenge in Latin America and the Caribbean Monitoring the Learning Outcomes of Education Systems MIGA: The First Five Years and Future Challenges (continued on the following page) Directions in Development (continued) Nurturing Development: Aid and Cooperation in Today's Changing World Nutrition in Zimbabwe: An Update Poverty Reduction in South Asia: Promoting Participation of the Poor Private and Public Initiatives: Working Togetherfor Health and Education Private Sector Participation in Water Supply and Sanitation in Latin America Reversing the Spiral: The Population, Agriculture, and Environment Nexus in Sub-Saharan Africa (with a separate supplement) A Strategy for Managing Water in the Middle East and North Africa Taxing Bads by Taxing Goods: Pollution Control with Presumptive Charges Toward Sustainable Management of Water Resources Trade Performance and Policy in the New Independent States The Transition from War to Peace in Sub-Saharan Africa Unshackling the Private Sector: A Latin American Story The Uruguay Round: Widening and Deepening the World Trading System I I I I ; i I II 1 i (I t i i i i ] i I  11i q g - ,:; 11U78 Eggs,