DIRECTIONS IN DEVELOPMENT.. l (c~ t C, S [.. Nr \i. IqiC Decentralization of Education Community Financing MARK BRAY ,'\,t ,q &jl 4 1: : F v X ; ~ ~ ~~~* '3,~~~~~~~~^t 1t , FE,~~~~~~~~~~~~~~~~~~~~~' DIRECTIONS IN DEVELOPMENT Decentralization of Education Community Financing Mark Bray The World Bank Washington, D.C. © 1996 The Intemational 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, to its affiliated organizations, or to members of its Board of Executive Directors or the countries they represent. Mark Bray, a professor at the Comparative Research Centre of the University of Hong Kong, is the author of several books and many articles focusing on the aspects of the economics and financing of education. His books indude (with Steve Packer) Education in Small States: Concepts, Challenges, and Strategies (Pergamon Press) and (editor, with Kevin Lillis) Commutnity Financing of Eduication: Issutes and Policy Implications in Less Developed Countries (Pergamon Press). He wrote this book as a visiting scholar in the World Bank's Human Development Department. Cover photo: Curt Camemark, The World Bank Library of Congress Cataloging-in-Publication Data Bray, Mark, 1952- Decentralization of education: community financing / Mark Bray. p. cm. - (Directions in development) Includes bibliographical references. ISBN 0-8213-3724-6 1. Education-Developing countries-Finance. 2. Federal aid to education-Developing countries. 3. Community and school- Developing countries. I. Title. II. Series: Directions in development (Washington, D.C.) LB2826.6.D44B73 1996 379. 1'2'091724-DC20 96-32177 CIP Contents Preface v Acknowledgments vii Communities and Their Characteristics 1 Scale, Motivation, and Nature of Community Financing 3 Mechanics of Community Financing 7 Initiating and Directing Operations 7 Strategies for Mobilizing Resources 8 Policy Issues Confronting Governiments 14 The Strength and Role of the State 14 Equity Issues 16 Planning and Resource Use 19 Government Incentives, Guidance, and Controls 21 Incentives 21 Guidance 26 Controls 29 Challenges and Risks 37 Political Dimensions and the Distribution of Authority 37 Disparities and Encouragement of Less Active Groups 39 Scale of Operation 40 Monitoring and Evaluation 42 Conclusion 44 References 47 iii Preface Although it is now generally accepted that governments should be the principal financiers of education, especially of basic education, in many countries communities are also significantly involved. Community financing has become especially important in countries where governments have been unable to meet the full demand for education (Assie-Lumumba 1993; Bray with Lillis 1988; Kemmerer 1992). In such instances, community resources may be a crucial sup- plement even in government schools. Where diversity in educational provision is encouraged, communities of different types have operat- ed schools alongside government systems to offer schooling with par- ticular religious, ethnic, political, or other orientations. In both low-income and more prosperous societies, community financing is caught up in the centralization-decentralization debate. To what extent and in what areas should communities be permitted to operate with full autonomy, under various kinds of guidance and sup- port, rather then being regulated and controlled? Although community financing may be desirable because it increas- es the resources available for education and reduces the burden on governments, community projects are not always well designed. Government planners sometimes feel that community initiatives lead to inappropriate use of scarce resources, and community financing may contribute to social inequalities. Finding an appropriate balance may be difficult, especially because of cultural, economic, managerial, and other differences within coun- tries. Different strategies may be required for different levels of educa- tion, and priorities may change over time. Community financing can be highly volatile. It can be affected by microforces such as the capac- ity and nature of leadership at the local level, as well as broader eco- nomic, social, and other forces. Moreover, although governments may want to encourage community involvement in some settings, in others they may need to discourage or redirect community activities to avoid petty rivalries and political instability. This book focuses on formal primary and secondary education. Although community financing may also be important at pre-primary and post-secondary levels and in nonformal education, this study does v vi DECENTRALIZATION OF EDUCATION: COMMUNITY FINANCING not consider those areas. It has been written to serve government policy- makers, planners, and project managers at national and subnational lev- els and their counterparts in nongovernmental organizations, external agencies, and similar bodies. It may also be useful to leaders of commu- nity projects who seek to understand better the perspectives of govern- ments and other partners. Although this study uses the word financing in its title, it is not con- cerned only with monetary inputs to education. Both communities and governments also contribute material, labor, expertise, and land. Because these inputs would have to be purchased if they were not pro- vided, they may be considered substitutes for cash. In addition, community financing may be important as a form and expression of participation (Colletta and Perkins 1995; Narayan 1995). People who pay directly for services are likely to take a stronger inter- est in them than are people who receive services free of charge. When community members contribute resources to schools, they are more likely to ensure that their children are enrolled in those schools and attend regularly. They may also scrutinize the performance of teachers, the curriculum, and other aspects of school operations more carefully, in turn improving the relevance and the effectiveness of educational processes (Shaeffer 1994). The scale of community financing may thus be an indicator of the extent of demand for specific education in spe- cific locations. The study begins by examining the volume, nature, and mechanics of community financing in a range of settings. Discussion then turns to policy issues facing governments and those who work with them. This leads to consideration of instruments for incentives, guidance, and control of community activities. Specific strategies from different countries are presented and evaluated. The next-to-last chapter under- lines the importance of monitoring and evaluation. The paper con- cludes by summarizing the lessons in the context of debate about cen- tralization and decentralization. Acknowledgments This paper is part of a larger effort by the Education Group of the World Bank's Human Development Department to understand best practice in the decentralization of education systems. The paper was prepared while Mark Bray was a visiting scholar in the department and was written under the direct supervision of Marlaine Lockheed, task manager, and the general guidance of Maris O'Rourke, senior education adviser. The paper benefited from discussion with other members of the team working on decentralization education- Rajayswar Bhowon, Robb Cooper, Cathy Gaynor, Eluned Schweitzer, and Carolyn Winter-and with department staff, including Nicholas Burnett, Tom Eisemon, Donald Holsinger, Peter Moock, Lauritz Hohm Nielsen, Maris O'Rourke, Harry Patrinos, Ralph Romain, Sverrir Sigurdsson, and Jee-Peng Tan. The paper was reviewed by the World Bank's Advisory Committee for the Education Decentralization Work Program, comprising Albert Aime, Sue Berryman, Hans Binswanger, Elizabeth King, Dzingai Mutumbuka, Juan Prawda, Alcyone Saliba, and Donald Wmkler. A draft was presented at the annual meetings of the Comparative and International Education Society on March 7, 1996. Margot Verbeeck provided excellent word processing, and American Writing Corporation edited and desktopped the text. vii Communities and Their Characteristics There are many different types of communities, but this study consid- ers only three: Geog-raphic communities. A geographic community is defined accord- ing to its members' place of residence, such as a village or district. In Tanzania and Nigeria, for example, locally formed development associations have become a major force in education. Rural com- munities are generally more cohesive than urban ones. In some sit- uations all the people in a region or even a country may see them- selves as a community, though this view is less relevant to the focus of this study. * Ethnic, racial, and religious communities. Schools may be run by peo- ple who identify themselves as Khmers or Yorubas, as Europeans, Chinese, or Hispanics; or as Muslims, Jews, or Christians. Within Muslim communities there may be separate groups of Ismailis, Ahmadiyya, and Sufis; and within Christian communities, separate groups of Roman Catholics, Baptists, Methodists, and Anglicans. Membership in communities based on ethnic, racial, or religious identification commonly cuts across membership based on geo- graphic location. This may lead to conflicts in identity and role but does not always do so. Communities based on shared family or educational concerns. Many schools have parents associations and similar bodies that are based on families' shared concern for the welfare of students. Some of these bodies may be defined narrowly to embrace only parents, while others have broader memberships. In Papua New Guinea, for example, they are deliberately called Parents and Citizens Associations (Bray 1988), and in the Philippines they are called Parent-Teacher Associations (PTAs) but commonly include alumni and other community members (Martin 1996). The degree of cohesion within communities also varies consider- ably. Policymakers cannot assume that a group of people living in a clearly identifiable village, for example, will have a strong sense of 1 2 DECENTRALIZATION OF EDUCATION: COMMUNITY FINANCING community. Goverunent and other personnel sometimes must help create a sense of community among people who would otherwise not work together. Government officers and project personnel must also be prepared for splits and disagreements within communities and for fluctuations in enthusiasm and effectiveness. Narayan (1995) suggests that the following characterize well-functioning groups: * The groups address felt needs and common interests. * The benefits to the groups of working together outweigh the costs. * The groups are embedded in existing social organization. * The groups have the capacity, leadership, knowledge, and skills to manage the tasks. * The groups own and enforce their own rules and regulations. Government personnel and project designers may need to treat dif- ferent types of community in different ways. This study will show some common ground among the three types of community, but it will also highlight differences. Scale, Motivation, and Nature of Community Financing Even in highly specific locations and at particular times, estimating the scale of community financing is difficult. Comprehensive statistics on this topic are not commonly collected, and calculating the monetary value of nonfinancial contributions is not easy. Nevertheless, such con- tributions are often considerable. Community financing is most dramatic where demand for school- ing is strong but government inputs are inadequate. This has been par- ticularly evident in parts of Africa. * In Uganda government provision of education all but collapsed during the 1970s and remained deficient during the 1980s and 1990s. The gap was partly bridged by parents and broader com- munities, who at the primary school level met between 65-90 per- cent of total costs (World Bank 1993e). Household inputs were also substantial at the secondary level and in 1994 were estimated at 70 percent even in government-aided schools (Cumming and others 1995). * Togo also experienced a major crisis in government funding, espe- cially after the mid-1980s. Communities and parents have had to provide one- to two-thirds of the resources needed to operate pub- lic-sector schools (Togo 1994a, 1994b). Also, in the mid-1990s about 400 conmmunity prirnary and secondary schools with about 27,000 pupils were operating outside the public system. - The situation in Chad is comparable. In 1991-92 communities employed about 40 percent of primary school teachers (World Bank 1993c), and communities also made major contributions to build- ings and infrastructure (Esquieu and Peano 1994). * In Malawi community-run primary schools unassisted by the gov- emment made up 20.5 percent of all primary schools in 1992-93 and enrolled 9.5 percent of all primary pupils (Malawi 1993). Community and parent contribution to public-sector schools were more difficult to quantify but have also long been substantial (Malawi 1992; Tilak 1989). 3 4 DECENTRALIZATION OF EDUCATION: COMMUNITY FINANCING Box 1. The Harambee Movement in Kenya Although the idea has a long history, President Jomo Kenyatta used the word harambee as a rallying cry at the time of Independence in 1963. It is commonly translated as "let us all pull together." Harambee activities finance a wide range of projects, including roads, wells, cattle dips, and medical dispensaries. Their role in education, however, far outweighs that in any other single sector. Harambee contri- butions to education in 1991 were estimated at 19.5 million Kenya schillings, compared with 11.7 mnillion Kenya schillings for all other sec- tors (Kenya 1994, 250). Harambee contributions are made to all levels of education, from kindergartens to tertiary institutions. In addition to cash donations, con- tributions are made of labor, materials, and land. Harambee has become part of the fabric of Kenyan society, and is a major force in development. In Kenya the self-help movement is known as harambee (box 1). At least in rural areas, communities are generally expected to take responsibility for building and maintaining primary schools and in 1987, harambee secondary schools made up nearly three-quarters of the total (Mwiria 1990). Almost 44 percent of all harambee sec- ondary schools were aided by the government, but others (almost 29 percent) were unaided. Since the early 1990s, all secondary schools have been eligible to receive government aid. The volume of community financing, however, remains substantial. Striking examples of community financing may also be found in Asia. In China the history of community-run minban schools dates from the 1940s (Robinson 1988). The numbers-of such schools fluctuated dur- ing the following decades, and official policy became to incorporate them into the public system. Community-employed teachers remained common in public schools, however. In 1990, for example, 41 percent of all full-time primary teachers and 10 percent of all full-time secondary teachers were employed by communities (Tsang forthcoming). A parallel system of community-funded schools has also long exist- ed in Nepal. In 1991, 18 percent of secondary schools were operated by communities with little or no support from the government (World Bank 1994c). As in other countries, communities are also expected to make substantial contributions to public schools. In neighboring Bhutan community contributions have had a short- er history because until the 1980s the government felt able to meet almost all needs. Demands for expansion of the educational system, SCALE, MOTIVATION, AND NATURE OF COMMUNITY FINANCING 5 however, have led to calls for increased community contributions. In 1986 the government encouraged communities to construct satellite classrooms attached to but some distance from existing primary schools, and in 1990 these entities were formally separated and renamed community schools. By 1995, 102 community primary schools operated alongside 143 fully government primary schools and just 19 government junior high schools (Bhutan 1995). Community contributions have also been increasingly evident in the mainstream public sector schools (Bray 1996b). Community financing appears in general less vigorous in Latin America than in Africa and Asia, although prominent examples may also be found in that part of the world. For example, El Salvador has a substantial group of community-managed schools. Much of the fund- ing for these schools comes from the government and from external aid, but a 1994 report noted that about 30 percent of the community educa- tion associations responsible for the schools embarked on their own fundraising initiatives (World Bank 1994b). A parallel report on Bolivia has noted that communities played a major role in that country, partic- ularly in school construction and maintenance (World Bank 1993a). In most of these cases community financing was primarily provid- ed to meet excess demand. The communities would have preferred the government to make adequate provision, and the communities' prin- cipal motivation was to fill in where government inputs fell short. As already noted, however, community financing may also arise from demand for alternative forms of education. Minority Chinese communities in Malaysia have formed vigorous schools that teach the basic curriculum required by the national gov- ernment but also provide a firm grounding in Chinese language, his- tory, and culture (Tan 1988). The history of these schools dates back to the nineteenth century, but they have continued to flourish in Malaysia's multiracial society. In 1995 independent Chinese schools enrolled more than a quarter of the pupils in private secondary schools and represented 3.3 percent of total (public plus private) secondary school enrollment. In Indonesia many schools with a religious bias are run by Islamic communities and come under the aegis of the Ministry of Religious Affairs rather than the Ministry of Education and Culture. In the mid- 1980s enrollments in these schools formed about 14 percent of the total at the primary level (World Bank 1989). Corresponding figures for the lower and upper secondary levels were 12 percent and 9 percent. In Papua New Guinea a national education system was created in 1970 from the existing government schools and the majority of church 6 DECENTRALIZATION OF EDUCATION: COMMUNITY FINANCING schools. The Seventh Day Adventist Church, however, stayed outside the national education system. Although the Adventist school system is small, in the mid-1980s serving only about 1.2 percent of the primary- age population and 0.4 percent of the secondary-age population (Bray 1988), it is a clear example of an alternative education system operated through nongovernmental funds. A similar model exists in the Solomon Islands (Gannicott and McGavin 1987). Just as motivations for community financing vary, so do the nature of those investments. Particularly in urban areas, some communities rely entirely on cash contributions, but others encourage inputs of materials. These goods may include wood, stones, and other items for buildings and food for students and teachers. Some societies donate cattle or other animals. Labor may be contributed not only for con- struction and maintenance but also for planting and harvesting crops either to be used directly in school meals or to earn money for school supplies. Community members serve on school committees and often provide free services in accounting, architectural design, and so on. Finally, land is frequently donated for classrooms, teachers' houses, and sports grounds. Communities aim to make up for government shortfalls, and offi- cial government policies often anticipate shortages in buildings, furni- ture, and land. In some countries the government is able to provide all teachers and discourages or prohibits communities from employing teachers because it is anxious to retain control of standards and access to education. In others government resources are inadequate even for providing teachers, and communities have to employ their own. Such countries include Chad, China, Mali, Myanmar, and Nepal. Mechanics of Community Financing The mechanics of community financing depend on how the school is operated and the purpose of the financial need. Initiating and Directing Operations Some schools are associated with individuals of vision and personal energies who not only have founded but also manage those institu- tions. There are examples in many countries. In most cases, however, schools spring from movements such as the harambee movement in Kenya. This description also applies to church schools and to most institutions sponsored by political bodies and cooperatives. Even schools that owe their origins to individuals require committees or broader structures for long-term stability. In some countries schools operate under the direct control of gov- ernments or other sponsoring bodies and have no school-level man- agement boards. In government schools in Sri Lanka, for example, headteachers receive instructions directly from government officers. Some schools do not even have parent-teacher associations (PTAs). Where school boards do exist, their mandates, composition, func- tions, and the way they operate may be of considerable importance, not only for financing but also for other aspects of school life. Some governments have recently introduced school boards where they did not exist before. This has occurred in Bhutan, for example, and has been actively discussed in Zambia (Bhutan 1992; Zambia 1993). The fact that some schools have PTAs or similar bodies that operate separately from school boards deserves further comment. In Papua New Guinea each primary school is required by law to have a board of management, the powers and responsibilities of which are determined by law (Bray 1988; Preston 1991). As indicated above, most primary schools also have Parent and Citizen Associations (PCAs). These bodies have less formal status but usually coordinate with the boards of man- agement. In some schools these associations are seen as bodies for car- rying out decisions made by the boards, although the relationship is not always entirely smooth. Moorad (1992) has made a similar observation in Botswana about the boards of governors of cornmunity junior 7 8 DECENTRALIZATION OF EDUCATION: COMMUNITY FINANCING secondary schools and their PTAs. Coordination is improved when at least the chairpersons of the associations also belong to the school boards. The greatest potential for confusion and conflict arises when there is no overlap in membership between the boards and the execu- tives of the PTAs and PCAs, when the two organizations have separate bank accounts, and when their roles have not been prepared in writing. In contrast, schools in Vietnam have only PTAs and no formal school boards (Rorris and Evans 1994). The PTAs link schools with families in a variety of ways and include fundraising among their major functions. Not all PTAs have formal constitutions or legal status as in Papua New Guinea, but they may perform some or all of the functions of school boards in other systems. Tembon's (1994) study in Cameroon shows some parallels (box 2). Strategies for Mobilizing Resources The ways school boards, PTAs, and similar bodies operate depends not only on their formal powers but also on the cultures within which they work. Strategies for mobilizing resources also depend on the nature of the task at hand. In particular, mechanisms to raise funds for buildings and other capital works are often different from those for recurrent needs. Capital Works For capital works, community leaders commonly organize focused activities such as launching ceremonies and rallies. Hill (1991) has pre- sented a portrait from Kenya. Referring specifically to the research he conducted among Kamba communities, he indicated that: "Harambee Days" were govermnent-licensed fundraising rallies at which the fundraising collections were presented in the pres- ence of guests and local notables, with additional contributions by guests. Ministers, MPs [members of parliament] and politi- cians regularly attended the functions. Entertainment was pro- vided by myethya [self-help] song groups, by church and school choirs, and sometimes by traditional dance groups. Prominent guests made speeches. The highlights were the presentation of locally collected funds and the competitive fundraising where all guests contributed, including administrative officials. Guests made their contributions publicly by presenting them to the mas- ter of ceremonies, or chairman of fundraising, who was usually MECHANICS OF COMMUNITY FINANCING 9 the chief guest, and they often made a short speech at the micro- phone. The most prominent guest-for example, a govermanent Minister-was expected to out-give the others. "Modem" fundraising methods were also sometimes used to enhance guests' contributions-for example, auctioning donated goods. Guests were given food and drinks afterwards. Hill reports that social sanctions were employed to ensure full atten- dance at such ceremonies and also to solicit contributions from indi- viduals from the village who had migrated elsewhere for employment or other reasons. To avoid disputes, the organizers usually stressed Box 2. PTA Fundraising in Cameroon In Cameroon govemment resources for education have been increasing- ly stretched by general fiscal crisis. To help make up the difference, many schools have established parent teacher associations (PTAs). The main functions of the associations are fundraising and provision of facilities. A 1990 government circular, seeking to legalize what was already happen- ing, sanctioned association fundraising. Tembon (1994) studied sixteen secondary schools in one division of West Cameroon. Her sample covered public, private-secular, and pri- vate-mission schools. Two of the private-secular schools were able to derive sufficient income from fees and did not have PTAs. The other schools all had associations, but levies were particularly high in govern- ment schools, which were not permitted to charge for tuition. Tembon found general satisfaction with the association among parents. Of 267 respondents, 80.1 percent reported that they had attended association meetings; 97.4 percent, that they had contributed the levy; 83.9 percent, that they had received the report on expenditure; and 82.6 percent, that they were satisfied with this report. Some respondents felt that the report had presented inadequate detail, but the fact that so many were satisfied with the report seems remarkable. Average PTA levies in govemment secondary schools were 13,000 CFA francs per pupil, or about 5 percent of estimated per capita GNP. Parents also had to pay for books, uniforms, travel, and other expenses. Tembon comments (277) that "the enormous contributions mobilized by PTAs in government secondary schools that are tuition free signifies par- ents' willingness to pay for their children's education." However, she also observes (294) that excessive dependence on PTAs contributes to instability because PTAs depend so strongly on the nature of local lead- ership and the inspiration and energy of particular individuals. 10 DECENTRALIZATION OF EDUCATION: COMMUNITY FINANCING that the schools would benefit everyone in the community, either directly or indirectly. Igwe's (1988) account of launching ceremonies among Igbo com- munities in Nigeria is similar. He highlights the strong incentives for individuals to enhance their prestige by donating large amounts and perhaps having buildings named after them. Igwe also notes the role of age groups in the fundraising process. In striking parallel to the Kenyan setting, Igwe indicates (111-12) that: Whenever it is decided that there is to be a launching ceremony for fundraising, the village improvement union issues a circular letter making it compulsory for all to attend. Special invitations are also issued to in-laws, relations, friends, and well-wishers outside the village. Normally, the circular states that heavy fines will be imposed on any absent branch unions, age groups, and individuals. Such fines range from N50 to N200 [Nigerian naira] or more, and sanctions are enforced with a roll call. In the village, a town crier goes round several times to announce the event and to invite people to decorate the venue before the date of the ceremony. All the village cultural dance, masquerade, and other entertainment groups are invited to attend, and the women prepare food for the occasion. Public address systems are mounted, and each branch union and age group has a colourful stand. Sons and daughters who work in town charter buses to get to the village, and many arrive a few days earlier. The ceremonies are usually held at weekends to enable all public servants to attend and to prevent them from using their jobs as an excuse for absence. At the ceremony, people compete to see who will donate the highest amount. There is a similar competition among the age groups. Everybody in the community endeavours to donate some amount, no matter how big or small, and each donation is announced on over the public address system. By the general spirit of the people, it is better to be dead than to be alive and not be able to contribute to the development of one's village or clan. Of course these practices are not found uniformly, even within Kenya and Nigeria. Other communities are less cohesive, and their leaders are not able to impose sanctions in the ways mentioned here. In these cases, the volume of fundraising may be more limited, and alternative methods may be required. MECHANICS OF COMMUNITY FINANCING 11 Other ways to raise money for capital works include fetes, carni- vals, cultural shows, harvest festivals, raffles, and sponsored walks. These are common throughout the world (Gillespie and Collins 1988). In these endeavors, school personnel tend to take more initia- tive than community leaders, but the largest and most successful events attract community members from a wide range of age groups and occupations. Another way to raise resources for capital works is through a levy on each household or individual. Ruda (1993, 45) describes the ways decisions are typically made in Botswana. Some individuals and households contribute cattle, but others contribute money according to minimum and maximum rates determined at village Kgotla meet- ings. Meetings are organized by the school board of governors and presided over by the village chief. The chief usually begins by intro- ducing the board members and indicating what the deliberations will be about. "Then he will call upon the chairman of the board who will introduce the topic and allow the assembly to discuss it and ask questions where clarity is lacking. When the discussions have been exhausted, the chairman will then request the commun- ity to decide how much has to be contributed." A common range, Ruda reports, is 200-300 pula for men, and half that amount for women. Strikingly similar systems exist in parts of Asia. In Lao People's Democratic Republic, for instance, chairmen of village community associations usually oversee construction of rural schools. Levies may be imposed on individual adults, or they may be by household. Often households are levied equal sums, regardless of the incomes of those households or the numbers of their school-age children, though per- haps with some allowance for substitution of labor in the case of very poor households (Bray 1996a). Recurrent Needs Community fundraising for recurrent needs is in many ways more problematic than for capital works. Continually donating salaries, food, and other consumables is less appealing to many potential donors than constructing buildings or other items that are more easily visible. However, recurrent demands cannot wait for the moods of the populace or the inspiration of organizers. When communities need to meet teachers' salaries and make other cash payments, they commonly raise much of the necessary revenue through school fees. The schools are thus able to secure regular, 12 DECENTRALIZATION OF EDUCATION: COMMUNITY FINANCING Box 3. Racially Based Groups and Community Levies in Singapore Among Singapore's 2.7 million people, 78 percent are Chinese, 14 percent Malays, 7 percent Indians, and 1 percent others, including Eurasians. Several racially based community associations have been established that receive government sanction and support and place high priority on edu- cational work. The Council on Education for Muslim Children was founded in 1981. Almost all its members are Malays, and the organization is better known by its abbreviated Malay name, Mendaki. The main rationale for founding the association was to help the Malay community catch up with the Chinese and Indians in educational performance. This, it was argued, would pro- mote Malay participation in the economy and avoid the racial disharmony caused by social imbalances. Mendaki's educational activities include weekend tutoring for primary and secondary students, provision of schol- arships and loans, and promotion of Islamic values through booklets, speeches, and courses for parents (Tan 1995). Mendaki receives strong government support. The Ministry of Education provides grants, leases premises at nominal rent for the association head- quarters, and permits Mendaki to use public schools as tuition centers. The government also helps train tutors and has established a mechanism for securing contributions through the Central Provident Fund (CPF). Levies are made automatically on all Malays unless they opt out. In 1995 the levies ranged from S$5 (Singapore dollars) per month for people eaning more than predictable income and ensure that services are at least partly paid for by their principal beneficiaries. In some societies, fees in commu- nity schools are substantial and raise questions about equity and effi- ciency that have also been examined in the broader literature on user charges in the social sectors (Cornia 1987; Jimenez 1987; Lewin with Berstecher 1989). Community leaders may also devise levies comparable to those for capital works. Such levies then become a form of regular taxation. In Singapore the government helps racially based associations raise funds through monthly Central Provident Fund deductions (box 3), and in parts of India purchases made at village shops are levied (Dandekar 1996). When collecting cash is a problem, teachers may be willing to accept at least partial payment in food, accommodation, or labor on their agricultural plots. Schools sponsored by religious organizations may also meet recur- rent needs from collections made during church, temple, or mosque ceremonies and from other regular sources of income such as rent on MECHANICS OF COMMUNITY FINANCING 13 S$3,000 to S$2 for those earning less than S$1,000. In 1994 these levies provided 39.3 percent of Mendaki income. Donations and governrnent grants provided 31.4 percent, and income from projects earned 29.3 percent. Following the Mendaki lead, other ethnic groups established associa- tions in the 1990s. The Singapore Indian Development Association was founded in 1991, and the Chinese Development Assistance Council and the Eurasian Association in 1994. These bodies also raise funds through Central Provident Fund deductions and irnpose a higher rate than Mendaki. For example, in 1995 the Eurasian Association levies ranged from S$10 per month for those earning more than S$4,000 to montly on incomes less than S$1,000. Fewer than 5 percent of Eurasians opted out of this scheme and the opt-out rate was only marginally higher among Chinese and Indians. The operation of similar educational activities by Chinese, Eurasians, and Indians makes it more difficult for the Malays to bridge develop- mental gaps. Yet all these associations contribute in ways the government could not do alone. As Prime Minister Lee Kuan Yew noted in 1982: "A government-run scheme cannot achieve a quarter of the results of this voluntary, spontaneous effort by Malays/Muslims to help themselves." Recognizing the limits of government actions, he added: "You can better succeed because you will be more effective with Malay/Muslim parents than government officers. You can reach them through their hearts, not just their minds... .This emotional/psychological support can make a vast difference between a student who tries, fails, and tries again, and another who fails and gives up." (Tan 1995, 344) properties. When religious organizations have more than one school, they may be able to subsidize the deficits of some institutions from the surpluses of others. Also, teachers and other workers in religious schools usually work at less than the market rate, therefore providing a hidden subsidy. Finally, many religious schools in low-income coun- tries are able to attract funds from religious organizations in more prosperous countries. Cash-crop cooperatives operate another type of community school. In Tanzania, for example, the Kilimanjaro Native Cooperative Union supports more than sixty self-help secondary schools, and the Bukoba Cooperative Union runs another five secondary schools (Galabawa 1994). Funds are generated through a levy on each kilogram of coffee sold through the cooperative. Arrangements for voluntary levies on the licenses issued to merchants who are members of communal asso- ciations resemble this model. Abreu (1982, 123) indicates that some schools operated by Hindu communities in Kenya have been partly financed in this way. Policy Issues Confronting Governments The extent to which community financing is tolerated or encouraged partly depends on the strength of the state and on official perceptions of its role. To see this fact more clearly, it is useful to present several contrasting models and examples. The Strength and Role of the State The most dramatic examples of community financing are where the demand for schooling is substantial but government inputs are inade- quate. Chad, Togo, and Uganda are "weak states" where government inputs are lacking because economic crisis has combined with general disorder to deprive the government of resources (Togo 1994a; World Bank 1993c, 1993e). Macau, a Portuguese territory in East Asia, was a weak state at least up to the late 1980s, the result of colonial neglect rather than economic poverty. The government could have generated greater taxation revenues had it wished and could have devoted them to the education sector but was not sufficiently organized to do so (Bray 1995). In all these societies, some schools have long been operated by reli- gious, racial, or other groups that value control over their institutions and that would be willing to continue investing to keep control even if government resources were available. However, most other schools in these societies get community resources by default. The communities would prefer the governments to take more active roles and make donations themselves only because they realize that in their absence, the schools would be severely starved of resources and might not exist at all. Indonesia, by contrast, may be characterized as a fairly strong state in which the military dominates the government. As noted above, in Indonesia substantial numbers of schools are operated by various Islamic religious groups. Prohibiting these schools would be more dif- ficult than in Singapore, and the authorities tolerate their existence. In the Indonesian case, however, government grants are used to strength- 14 POLICY ISSUES CONFRONTING GOVERNMENTS 15 en the power of the state and to encourage the schools and their own- ers to take a positive stance toward the ruling regime. The states of the former Soviet Union present a third model. During the socialist era government funding of education in these states was considered both a responsibility and a right, so community financing was almost nonexistent. With the economic crisis that accompanied the collapse of socialism, the governments of these countries have no longer been able to provide resources for their schools at satisfactory levels. This has forced the authorities to look for other ways to finance schools, including privatization and imposition of fees in public schools. Racial and religious communities could be another source of funds for education, but few governments have actively embraced that option because they fear further political fragmentation. Azerbaijan, for example, is a weak state in which the government is trying to main- tain legitimacy despite major religious and ethnic tensions. Encouraging Armenian Christians or Azeri Muslims to operate their own schools could strengthen religious and ethnic divisions and exac- erbate already serious political tensions. Singapore exemplifies a fourth model. A strong state in which the government has resources, authority, and a robust self-image, Singapore could prohibit community financing of schools if it want- ed to. In this case, the government has decided to encourage com- munity financing because it recognizes that such activities help gen- erate interest and commitment to the educational process (see box 3). Earlier in Singapore's history, the state was fragile, and the govern- ment felt threatened by the notion of racially-based community bod- ies organizing their own schools (Gopinathan 1995; Wilson 1978). The fact that during the 1980s and 1990s the government actively encouraged such community groups reflected the government's increased self-confidence. The Singaporean example also illustrates the dilemmas all states may face when govemment funding becomes dominant. In India govemment revenues have been sufficient for the govern- ment to take over full payment of salaries in aided schools, but one result of the increased govemment inputs has been decreased com- munity involvement and control. As Siddhanti (1996, 14) says of Kamataka, under the old system of just 50 percent subsidies, the school sponsors "of course had to struggle very hard to collect money every month for payment to the staff. But that had brought both of them closer, and the management had moral control over its staff." This closeness between teachers and community sponsors, Siddhanti reports, has been lost in the new system. 16 DECENTRALIZATION OF EDUCATION: COMMUNITY FINANCING Rath (1996) makes a related point about Goa. Although community- sponsored schools continue to exist, they gain much less support than before. "In the past, community and other resources played an impor- tant role. Now the managements find it difficult to approach the peo- ple to collect donations because the donors hesitate to donate for pri- mary education due to the co-existence of government schools in the same area." Similar decline in community involvement has been observed in Botswana, where governnent grants to the so-called community junior secondary schools have increased to the point at which the institutions have effectively become government rather than community schools (Moorad 1992; Ruda 1993). Swartland and Taylor (1988, 151) note: What was originally seen as government support for community initiative is fast becoming a government programme with commu- nity support. Communities identify less closely with buildings pro- vided for them by the government and the World Bank than they do with ones they have worked long and hard to build themselves. One of the strengths of a system of local control in education is the diversity and flexibility which is possible, and the responsiveness to localy perceived needs. lncreased government support too easi- ly turns into confonnity and control. It also heightens expectations of what government can and should do and reinforces a dependent mentality rather than one of self-reliance. Having got subsidies, teachers, and buildings from the government, the schools are now asking for film projectors, security fences, "official-free" stamps, school vehicles, and graders to level their sports fields. Swartland and Taylor add that there is "no easy escape from the dilem- ma that the more government is able to do, the more government is expected to do, and the less the community expects to do for itself." These examples help illustrate the diversity of circumstances in which government policies must be made. The tensions facing weak states are perhaps more obvious than those facing strong states. But even in strong states the role of government may be problematic because it may be perceived as undermining community initiative and involvement in the education system. Equity Issues In all states, both weak and strong, community financing may raise major issues of equity. The most obvious issue arises from the fact that POLICY ISSUES CONFRONTING GOVERNMENTS 17' some communities are in a better position to help themselves than are others. In Kenya disadvantaged districts have remained disadvantaged because they had low incomes in the first place and therefore found it difficult to help themselves (Mwiria 1990). Similar problems have been encountered in Botswana (Moorad 1989), Cameroon (Tamukong 1995), Nepal (World Bank 1994c), and many other countries. In some societies regional variations reflect cultural variations. In Nigeria and Kenya, for example, regional imbalances have less to do with the distribution of natural resources than with the fact that peo- ples of certain ethnic groups have stronger traditions of community organization than do others. The same is true of Tanzania. Samoff (1990) points out that the Chagga people, because of their strong tradi- tions of self help, have a disproportionate number of educated people in the population. In 1982 the Chagga formed 3.7 percent of the nation- al population but provided 20.5 percent of Form 5 students. Ethnic rivalries, which are given expression in and are fueled by community self-help projects, may present major political challenges to governments. In some countries racial inequalities are the issue. In Zimbabwe great emphasis was placed on community financing during the 1980s as a way to generate resources and expand educational provision. However, the policy operated differently. As Maravanyika (1995, 12) recounts: Schools in former white areas established management agree- ments with government. These enabled management committees to levy parents so that the schools could buy additional school equipment and other teaching resources or recruit additional staff to reduce the goverunent stipulated teacher/pupil ratio, which some white parents considered too high for effective teaching, or introduced specialist subjects not covered by gov- ernment such as music and computing. In short, management committees were concerned with maintaining former colonial privileges and standards. Maravanyika points out that the large amounts charged by the man- agement committees were generally out of the reach of the ordinary black parents and that the method of financing perpetuated racial inequalities. Most societies also suffer from rural-urban disparities in educa- tional provision. Government policies tend to compound rather than reduce inequalities. Because it is usually more difficult to foster 18 DECENTRALIZATION OF EDUCATION: COMMUNITY FINANCING community initiatives in urban than in rural areas, governments com- monly provide extensive resources for the urban schools while expecting the rural ones to help themselves (Korboe 1995; Maravanyika 1995; World Bank 1995). This has been recognized in Malawi, where authorities have tried to devise balancing mecha- nisms. For example, in 1983 the Lilongwe Urban District Council imposed a levy of 5 kwacha per family in lieu of self-help contribu- tions (Bray 1987). At the same time schools operated by some types of community are more likely to be located in towns than in rural areas. The independent Chinese schools in Malaysia are in this category, chiefly because the Chinese communities are themselves concentrated in urban areas. One effect of this form of community activity, therefore, is to extend the imbalance in facilities between rural and urban areas-a fact that becomes of added significance when substantial numbers of non- Chinese families also choose to enroll in the schools (Hiebert 1995). Further inequities may be socioeconomic. When community financ- ing expands access to education, it presumably makes schooling avail- able to some groups who would otherwise have been denied it. However, schools may be vulnerable to exploitation by elites. As Bhatnagar and Williams (1992, 4) observe, "Sometimes resources for development can be captured by local elites and used primarily for their own benefit rather than [that of] the intended beneficiaries. This happens because local elites usually have advantageous ties to nation- al elites, because they have access to and information about resource allocation procedures, and because they can use threats and force against the disadvantaged." Specifically in education, Muya and others (1995) report that in Kenya many parents feel exploited by school committees, which are considered demanding and unsympathetic to the burdens they impose. Similar comments have been made about Uganda by Opolot (1994). Particularly problematic are situations in which the poorest groups contribute resources to schools that are mainly occupied by the children of the middle classes and the rich (Galabawa 1985). These issues may also be related to taxation. From the official stand- point community financing may be an alternative to government tax- ation, and it may be much more effective in generating resources. Evidence from Kenya indicates that harambee activities for projects to which communities are committed release physical and financial resources for education that otherwise would be far beyond the reach of the government's taxation system (Mbithi and Rasmusson 1977,16; Wilson 1992, 4). Community taxes, however, are not necessarily more POLICY ISSUES CONFRONTING GOVERNMENTS 1'3 progressive than government ones. Although some community lead- ers take differing incomes into account when setting levies on individ- uals and households, this is not always the case. Planning and Resource Use Questions about the ways in which resources are used in self-help projects are also relevant. Although advocates of reliance on the market assert that decisionmakers are best left to identify for themselves what is in their best interest, not all planners agree. Examples can easily be found of: * Schools sited in inappropriate places because communities lacked broader perspectives and planning skills. * Schools that are undesirably small because two or more competing communities have insisted on their own separate schools. * Projects that were embarked on but never completed because of loss of interest, change of priorities, underestimation of resource requirements, or other factors. Concerning uncompleted projects, official data from Kenya showed that during one year, 1971, the number of abandoned self-help projects exceeded the number completed (table 1). Such problems suggest that self-help schemes sometimes lead to misdirection of resources that would have been better invested elsewhere. Along these lines, Shaeffer (1994, 28) comments on the danger of participatory processes leading to the domination at the local level of narrow community self-interest that may be based on short-sighted perspectives and aim at short-term benefits. He notes, "Such self- interest may ignore longer-term political or economic implications for the community or the larger society and contradict more general goals of national integration, the dissemination of scientific truth, and the modernisation of society." The community leaders in charge of self- help schools sometimes insist on building schools that may not have Table 1. Flow of Self-Help Projects in Kenya, 1971-73 Continuedfrom Continued into previous year Newly next year Year (cumulative) started Abandoned Completed (ongoing) 1971 5,074 6,579 3,325 1,841 6,505 1972 6,010 2,841 1,057 1,800 5,994 1973 7,190 3,064 740 1,704 7,810 Source: Government of Kenya, reproduced in Mbithi and Rasmusson (1977, 40). 20 DECENTRALIZATION OF EDUCATION: COMMUNITY FINANCING long-term futures or that operate curricula strongly at variance with those favored by national planners. Issues of quality relate to these problems. Some self-help schools, especially those operated to provide an alter- native to government institutions, are excellent. Most schools operated by the Seventh Day Adventist Church in Papua New Guinea and the Solomon Islands are in this category, and so are the majority of Malaysia's independent Chinese schools. However, the schools that pro- vide a second chance to students unable to gain places in government schools are usually very poor in quality. They attract less capable pupils, are usually unable to employ well-qualified teachers, and generally operate with the barest facilities (Esquieu and PNano, 1994; Wellings 1983; World Bank 1994c). Again the question arises whether these schools are not mere instruments for inappropriate investment of resources better invested elsewhere. Government Incentives, Guidance, and Controls Given that community financing may be very desirable but can also be problematic, how could governments stimulate, guide, and control it? Much depends on the particular community, but some general points can usefully be set out. Incentives Governments can establish a set of incentives designed to stimulate community financing. These include assurance of long-term opera- tions, promise of ultimate government takeover, provision of person- nel and materials, matching grants, and taxation incentives. Assurance of Long-Term Operation For communities that want to provide religious, cultural, or political education, perhaps the most important requirement is confidence that the government will at least tolerate and preferably encourage their efforts on a long-term basis. Few individuals and agencies will invest substantial resources in schools if they fear that those schools will not be permitted to continue operating. The record has been mixed. During the heyday of socialism most socialist regimes considered education not only the responsibility but also the right of the state. They therefore prohibited private education, including that operated by not-for-profit voluntary agencies. Some capi- talist countries have prohibited it as well. For example, in 1972 the gov- ernment of Pakistan nationalized all schools, including those that church- es and other noncommercial organizations operated not-for-profit. In 1974 the government of Zaire made a similar move. In due course, both these policies were reversed, but the confidence of nongovernmental agencies had been severely shaken. Thus governments should assure communities that their education- al enterprises would indeed be permitted to operate long-term. Governments should think through the extent to which they are 21 22 DECENTRALIZATION OF EDUCATION: COMMUNITY FINANCING willing to allow communities to control admissions, curricula, and other facets of the educational process and should make detailed pub- lic statements on these matters. Some governments may feel that com- munity control creates unacceptable threats to social harmony and national unity, but other governments would be more willing to take risks and to tolerate diversity. Promise of Government Takeover By contrast, communities that aim to expand provision more or less on the model of government schools are concerned not that the govern- ment keep out of their affairs but that the government take over fund- ing schools in a few years' time. Thus a major incentive for communi- ties to establish secondary schools in Kenya and Myanmar has been the promise of future government takeover (Lillis and Ayot 1988; Biak Cin and Scandlen 1988). That promise allows communities to feel that although initial burdens are heavy, their responsibilities will be reduced in the medium or longer run. The system in Myanmar illustrates one type of arrangement. Community-sponsored schools are attached to government institu- tions and are known as affiliated schools. When the government begins supporting affiliated schools, it still expects some community inputs but pays the teachers and in other ways treats the institutions like government schools. In 1989-90, for example, the government took over responsibility for one-quarter of the 1,259 affiliated prirnary schools then existing, together with one-quarter of the 1,322 affiliated middle schools, and one-third of the 342 affiliated high schools (Gibson 1992, 2). This move attracted considerable popular support because the government was seen as relieving the burden that had been shouldered by some dedicated communities. It is not always easy to decide which schools should be taken over and which should not. In many cases decisions are based on geo- graphic criteria (such as whether particular areas currently have gov- ernment schools) and on political criteria (such as whether the author- ities wish to demonstrate support for neglected groups). India's Maharashtra state has established a formula for deciding whether unaided schools should be eligible for the government to take over (box 4). While setting minimum criteria on class size, this formula bal- ances other factors in deciding whether the governmnent can take on the main costs of running an institution. The formula also provides ini- tial targets for communities in academic performance, construction of buildings, and purchase of science materials. GOVERNMENT INCENTIVES, GUIDANCE, AND CONTROLS 23 Personnel and Materials Some governments are able and willing to provide teachers for all community schools. Because salaries are usually the greatest cost in any education system, this is a significant form of support. Government agreement to supply teachers provides a major incentive for communities to donate land and construct facilities. In other cases governments agree to subsidize the salaries of community-employed teachers (see, for example, Cheng 1990, 59). Governments may also provide textbooks and other learning materials, and local govermnent personnel sometimes assist community schools when they urgently need mechanics, electricians, or other skilled personnel. In some societies, however, governments require communities to make their own initial commitments before they will allocate any gov- emnment resources. For example, in the Baluchistan Province of Pakistan, a community-based project for girls' schooling requires com- munities to operate schools independently for at least three months before the government will provide school supplies or pay teachers (O'Grady 1995). Government requirements on aid for construction are even more stringent, for communities may receive offers of government- Box 4. Eligibility for Government Grants: An Indian Formula In India's Maharashtra State the government is willing to take over the main costs of operating community secondary schools, but only after they have demonstrated initial capacity and reached certain targets. First, no school will even be considered for government funding unless it has been in operation for at least three years. Second, schools must achieve at least 70 points out of 100 on the following criteria: * Strength of enrollment, 20 points (subject to a minimum enrollment of thirty students per class). * Academic performance of class 10 pupils, 30 points. * Buildings, 10 points for three classrooms, one office, one staff room, one laboratory, and one library. * Furniture, 10 points if adequate. * Science materials, 1 point for every 1,000 rupees invested. * Trained teachers, 10 points if at least 50 percent of teachers have been trained. * Proximity to another school, 10 points if there is no other secondary school within a radius of 10 kilometers. Source: Dandekar 1996, 11. 24 DECENTRALIZATION OF EDUCATION: COMMUNITY FINANCING supplied buildings only when they have been operating their schools for at least three years. As in India (see box 4) this arrangement is con- sidered necessary to ensure that communities uphold their role within the partnership and that projects are a response to real and sufficiently sustained demand. One Ghanaian project is also worth mentioning. In this secondary school project the authorities were aware that communities had the capacity to construct latrines but often failed to do so to the extent that the lack of sanitary facilities threatened health. Recognizing at the same time that most communities wanted vehicles but would under normal circumstances be unable to purchase them unaided, the gov- ernment decided to give vehicles to communities that built satisfactory latrines (World Bank 1991). Matching Grants Rather than donating buildings or other facilities outright, govern- ments can require matching funds. This is attractive because ratios can be varied according to the wealth of particular communities and the resources available to governments. In Botswana, for example, the government originally expected communities to meet 50 percent of the capital costs of community junior secondary schools but later reduced this to 25 percent and then to 10 percent (Ruda 1993, 13). Discretion can be exercised over whether or not to recognize labor, materials, and land as community contributions in the matching grant formula. Mauritius uses a formula for making grants (table 2). The scheme consists of a per capita grant based on school population, a fixed grant for minor projects, and a matching grant equivalent to the amount Table 2. Annual Grants Payable to PTAs in Mauritius (rupees) Per capita Upper limit of Type of school grant Basic grant matching grants Government and aided primary One stream 1 1,500 7,000 Two stream 1 2,000 10,000 Three or more stream 1 2,500 12,000 State secondary Form V 1 5,000 20,000 Form VI 5,000 30,000 Source: Mauritius 1991,112. GOVERNMENT INCENTIVES, GUIDANCE, AND CONTROLS 2I raised by PTAs, with an upper limit. A special grant is payable tc schools in deprived areas that cannot raise the basic sum of 7,000 rupees. Additional matching grants (not shown in table 2) are payable to private secondary schools for purchase of computers, photocopiers, and audio-visual equipment. Also, an ad hoc grant of up to 100,000 rupees is payable to state secondary schools for specific projects under- taken by PTAs. To strengthen the system, the Mauritian government has encouraged schools to coordinate and exchange information through PTA federations and through newsletter production. In China the policy on matching grants has been described as fishing: the government fund is the bait to attract the fish of larger contributions from the communities. Cheng (1990) indicates that there are mixed pub- lic reactions to the strategy. Some community leaders welcome the sys- tem as a way to attract government resources to their schools, but others feel discouraged because they are unable to respond to the government offers. Moreover, some local governments that are accustomed to issuing top-down directives have made local contributions mandatory. In these situations, Cheng reports, the policy has caused resentment. This dis- parity in reactions highlights the need for governments to monitor reac- tions and to be flexible in carrying out policies. Ghanaian authorities have also used cash matching grants. Under the Community Initiative Project launched in 1988, grants were made available not only for schools but also for health centers, public latrines, and markets (World Bank 1993d, 14). During the first five years more than a thousand grants were made. An evaluation high- lighted four lessons that may also have wider applicability: - During periods of high inflation it is essential to make prompt pay- ments to communities and to adjust grant levels frequently. * To maintain morale, disbursements for sizeable projects should be made in stages rather than in large lump sums on completion. • Most communities should be expected to contribute primarily in labor and materials rather than in cash. * Approval and payment processes must be decentralized because central authorities cannot be aware of all local conditions. Taxation Incentives Governments can use the taxation system as a way to encourage community financing. Schools that do not aim to make profits may be exempted from taxation, and in some industrial countries it is standard practice for individuals and companies not to be taxed on 26 DECENTRALIZATION OF EDUCATION: COMMUNITY FINANCING Box 5. School Sponsorship by Philippine Commerical Enterprises Under an innovative partnership scheme in Manila, several corporations have become involved in an adopt-a-school program. Three elementary and two secondary schools have benefitted and have received inputs from an oil refinery, a match manufacturer, a detergent company, and a large multinational hamburger outlet. Assistance has included cash grants, construction, computers, scholarships, and maintenance support (Martin 1996). donations made to registered charities. Developing countries may not have such sophisticated taxation and accounting systems, but in most cases there could be scope for at least some incentives of this sort. The case of the tea producer, Brooke Bond Kenya Limited, illus- trates the potential role of commerical companies. Providing educa- tional facilities for the children of employees has long been a prior- ity for this company, which by the mid-1980s had established seventeen schools (Lillis and Ayot 1988, 121). In 1984 Brooke Bond embarked on a program worth 3.5 million Kenya shillings to con- struct thirty-four classrooms in the area where its tea operations were concentrated. Activities of the Singapore Indian Development Association, which in 1992 launched a raffle to raise 2.5 million Singapore dol- lars, were arranged differently. Banks, supermarkets, and other com- panies sponsored prizes in cash and in kind and were able to claim exemption from the government for the donations in cash (Straits Times, 9 August 1992). Revenue from these activities went into the association's general endowment funds. Yet another strategy tried in the Philippines is to encourage companies to become partners with individual schools (box 5). Governments can encourage this type of activity by publicizing the donations by the companies. They can add extra incentives by grant- ing taxation rebates on approved donations. Guidance Government can provide information and guidance on pedagogical aspects of schooling, accounting and budgeting procedures, and orga- nization and management techniques. GOVERNMENT INCENTIVES, GUIDANCE, AND CONTROLS 27 Technical and Professional Information While community leaders may be able to identify needs for education and strategies for mobilizing resources, it is rare for school committees to possess the full range of technical and professional skills required for building and operating schools. Governments can provide techni- cal guidance on such matters as architectural design and planning of facilities. Central or local government officers may also need to inform conmmunity leaders about legal requirements that affect their schools. Information and advice may be provided in documents or through district education officers, inspectors, surveyors, and other specialist personnel. School committees may also need guidance on curriculum, peda- gogy, and academic standards. This includes identification of learning goals for pupils in each grade, strategies to meet the needs of children with learning difficulties, mechanisms for recruiting appropriate teach- ers, and procedures for supervising and monitoring teaching and learn- ing. Following their research in Uganda, Cumming and others (1995, 5-13) pointed out that, "While boards may have the knowledge to choose between hiring or buying transport, when it comes to establish- ing how to improve poor [scholastic] performance, professional exper- tise is needed. This is where the Ministry of Education and Science should come in as the provider of professional services to boards." They added that "those who pay should be given the assurance that what they are buying is what they think they are buying." This is per- haps particularly important for parents with low levels of education who would be unable to make judgements by themselves. In this de facto decentralized system in which many initiatives are taken at the local community level, one role for government is providing such technical support where needed. Accounting and Budgeting Many community projects have inadequate records of income and expenditure, which create several problems. Adequate accounts are needed: * To prevent fraud. Many schools have had trouble because of accusa- tions that individuals have taken school money. Sometimes these accu- sations are justified; sometimes they are not. But in the absence of ade- quate accounts it is impossible to prove the case one way or the other. * To allow school authorities to forecast future expenditures. 28 DECENTRALIZATION OF EDUCATION: COMMUNITY FINANCING * To record how much has been contributed, and by whom, and therefore to ensure that contributions are appropriately recognized and that defaulters can be identified. • To help ensure a continuing flow of grants from governments and donors, who are more likely to maintain support if presented with clear and regular accounts. * To obtain loans, which is much easier when schools can show lenders that they are well organized, and know how and when they can pay back money. Accounts do not need to be complicated. They should show how much money was received from each source and how much was spent for each purpose. The best-organized communities also keep detailed records of contributions of labor and materials. In some countries the government is able to arrange for official audi- tors to handle school accounts. In Botswana, for example, the govern- ment audits the accounts of community junior secondary schools annu- ally. Similar arrangements have been instituted in parts of Indonesia. Such arrangements are not, however, possible in the majority of cases. Few governments have sufficient resources for such tasks, especially because many primary schools are located in remote and inaccessible areas. In these cases, communities may be guided to organize their own accounting and auditing systems. Accounts in church schools, for exam- ple, may be audited by the church education secretaries or parish priests. Or schools can make an arrangement with neighboring schools similar to that used when schools invigilate each others' examinations. These do not need to be complex operations. The chief goal is to establish a system in which at least one outsider looks at school accounts. Basic Organization and Management Community leaders may need help with simple aspects of organiza- tion and management. These may include strategies for running meet- ings, building consensus, meeting deadlines, and dealing with conflict. As Narayan noted (1995, 16), "Groups have failed because too much was expected of them too soon without any supportive training in management or specific skills. Getting local groups to become self- managing organizations can extend over several years and does not happen without investment in capacity building." Some governments have sought to strengthen operations at the community level by orga- nizing workshops for community leaders (box 6). Such workshops need not be costly affairs, and governments often find that communi- ties are willing to contribute to the cost of running them. GOVERNMENT INCENTIVES, GUIDANCE, AND CONTROLS 29 Controls Governments often find some regulation of community schools necessary. Registration of Schools and Managers Most governments insist that all schools be officially registered, including schools that are independently managed and financed. They do this first to assess the geographical coverage of education and sec- ond to permnit enforcement of regulations. In the majority of systems, schools must pass through several stages before they can be fully registered. In Imo State of Nigeria, for instance, schools should apply twelve months before they are due to open. The forms should be submitted to the zonal inspectors of edu- cation, who send them to the State Ministry of Education with com- m [ents. On receipt of the application the ministry checks whether the application conforms with government policies, whether the pro- posed school is in a suitable place, and whether the government will be able to pay the required grants. The ministry may also demand Box 6. Workshops for Community Leaders In some countries governments have sought to strengthen community- level operations and promote partnerships in the delivery of education by organizing workshops. In Papua New Guinea, for example, work- shops have been held for primary school boards of management. Provincial education officers usually organize these workshops, which have focused both on the respective roles of participants in the education system and on the ways that board members can fulfill their roles more effectively. Topics discussed have included ways to organize effective meetings and formats for systematic accounting systems. In some dis- tricts, the workshops have made a major contribution to the operation of the education systems. Workshops for community personnel have also been organized in Ghana. In the Primary School Development Project, the govemrnment has been willing to provide roofs and pillars for classrooms for communities that build the walls. The project has also introduced a formal role for communities in selecting head teachers. The workshops have explained the project to community leaders and have helped government officials understand both the constraints and the variations that should be expected in different communities (World Bank 1993d). 30 DECENTRALIZATION OF EDUCATION: COMMUNITY FINANCING evidence that the community has sufficient land and financing. When satisfied, the ministry secures the community's agreement to the regulations and grants permission to commence construction. The school is registered and assigned teachers only when buildings have been deemed to be satisfactory. This is a tight system, and com- munities must meet many conditions before registration. Other sys- tems may be less rigorous, but in most places, refusal to register a school is an important sanction. Some governments also require communities to register the man- agers of their schools. They do this so that the authorities: * Have proper records of who the managers are. * Can ensure that managers are suitable individuals or bodies. * Have evidence on the extent to which communities are well organized. * Have ways to require managers to declare their understanding of their responsibilities within the system. Teacher Recruitment, Qualifications, and Salaries As noted above, some governments insist that all teachers in both gov- ernment and community schools be employed by the public service. This does of course require those governments to have sufficient bud- gets, but it allows them to control teacher recruitment, qualifications, and salaries. In the absence of such an arrangement, communities might employ less-qualified staff. Staff qualifications are perhaps not a major problem in systems that have a reputation for quality, such as those run by the Seventh Day Adventists in the Solomon Islands or by the Chinese communities in Malaysia. However, qualifications may become a major issue in schools run by village self-help groups. Such groups are likely to seek inexpensive staff, which usually means those with low qualifications. Communities are also likely to give preference to employing their own members. Where the communities have access to well-qualified per- sonnel, this need not be a problem. The practice may be incestuous, however, and can seriously limit the pool from which teachers are drawn. For rural communities there may be no way to recruit staff from beyond the immediate neighborhoods even if the leaders are keen to do so. In these cases governments may decide to accept the fact that the skills of community-recruited teachers are likely to leave much to be desired and seek to upgrade their skills by offering (or insisting on) in-service training in government-run programs. GOVERNMENT INCENTIVES, GUIDANCE, AND CONTROLS 31 Where governments do not themselves feel able to take responsibil- ity for employing and paying all teachers, they may nevertheless regulate their minimum qualifications. For example, during the 1970s and 1980s all harambee secondary school teachers in Kenya were sup- posed to have at least school certificate qualifications, although this regulation was not always closely adhered to. Governments also could set minimum salary levels, as in some provinces in China (Cheng 1990), to indirectly influence the levels of qualifications. Enrollment of Marginal Groups One danger of leaving most decisions to community leaders is that school systems may discriminate against marginal groups. As Narayan pointed out (1995, 5): "Experience shows that community- based development does not automatically include marginalized groups, the poor, women, or ethnic minorities unless their partici- pation is specifically highlighted as a goal, both at the agency and community levels." In Anambra State in Nigeria government officials have used their powers over allocation of pupils to ensure that community schools provided sufficient places for girls. Okoye (1986) explains that during the 1970s and 1980s the government usually deferred deciding on the gender of pupils to be admitted to community-built schools until after the final inspection had indicated that the buildings were ready and the available places in the area had been compared with the numbers of successful candidates from primary schools. She adds (271) that, "In a few cases, community leaders rejected the allocation and the advice of ministry officials and went to higher authorities to change a girls' school to a boys' school, only to regret this when they had to explain to the local people why they had sacrificed so much to build an empty school while neighbouring schools were full." So seriously did gov- ernment officials pursue this policy that the female participation rate increased from 37 percent of total secondary school enrollment in 1975-76 to 57 percent in 1982-83, giving Anambra State the highest female participation rate in the country. Governments may also wish to insist that community schools be open to pupils who are not members of those communities. One dan- ger of permitting a religious organization to establish a school in a par- ticular village might be that such a school would make unviable the operation of a separate school for children who were not members of that religion. Similar comments might apply to schools sponsored by organizations based on particular racial or ethnic groups. 32 DECENTRALIZATION OF EDUCATION: COMMUNITY FINANCING Curriculum Some projects seek to stimulate community activity by making major changes in curricula. In Mali, for example, USAID and Save the Children USA supported a community-based primary education pro- ject (De Stefano 1996). The project aims to find an alternative to the mainstream system. The curriculum, which was developed in collabo- ration with the government's National Pedagogical Institute, uses Bambara as the medium for instruction in the first three years, and makes French an option only from the third year. The project is yet to be fully evaluated, but it is an example of deliberate curricular modifi- cation to encourage community interest. In other cases community demand may already be strong, and govemment concem may be to insist that community schools embrace elements of national curricula that might otherwise be omitted. This stipulation might be most needed in the case of alter- native-curriculum schools sponsored by religious, ethnic, racial, or Box 7. Government Controls on Community Projects in Nigeria Nigeria's Anambra State has strong traditions of community self-help. During the mid-1970s the state government harnessed these traditions in a major program for expanding secondary education. The number of sec- ondary schools increased from 95 in 1975-76 to 403 in 1982-83. Nearly 90 percent of the new schools were built by communities. When launching the program, the government widely publicized the opportunities for communities to establish schools. It agreed to provide teachers for approved institutions if communities provided land, buildings, and furniture. However, because the government was mnindful of the dan- gers of uncontrolled expansion, it set stringent conditions. To ensure that communities understood the magnitude of tasks, the gov- ernment required all proposals to include building programs showing requirements for each year of growth of the schools. Because the government could not compel communities to complete schools once they had been start- ed, however, the authorities retained power to refuse allocation of Form 1 students at the beginning of any school year. Thus, where a school had insuf- ficient facilities for the next session, the normal Form 1 intake was canceled or reduced, depending on the gravity of the situation. Warnings were given both to the schools and to the communities well ahead of time, with detailed listings of specific inputs needed to avoid the penalty. Community leaders were often grateful to receive notification in writing because the warnings could then be read out at monthly town union meetings. GOVERNMENT INCENTIVES, GUIDANCE, AND CONTROLS 33 other groups. Governments might insist national languages and civic education be included. In a third situation comnmunity demand for mainstream schools may be high. As illustrated by the Nigerian case (box 7), govern- ments can control access to public examinations as a means of ensur- ing that communities both build schools with adequate facilities and recruit teachers with sufficient qualifications. In this case the control is on support for the curriculum rather than on content of the curriculum. Building Designs Some authorities are content to allow communities to decide on their own buildings and to use either local or imported designs. Arguments favoring local designs include the following: * It is important to encourage and respect local cultures, of which building designs are a prominent part. A local style may be The government also retained controls over both the geographical spread of schools and the intakes of pupils. Communities in disadvantaged areas were treated leniently, while stringent conditions were imposed on communities that already had several schools. Also, aware of the danger that many communities if left to themselves would establish only boys schools, the government insisted that some institutions would either be coeducational or only for girls. Another tool used by government concerned approval to take the school certificate examninations. Inspections were conducted in the fourth year of each school's life, to see whether the institutions had achieved the required stan- dard to present candidates for examinations the following year. hispections focused on facilities, curiiculum, and administration. When schools were unable to reach the standards in specific subjects (such as providing a fully equipped laboratory for each of the sciences), these subjects were excluded from the general approval given to the school. Conimunities generally saw the inspection as a judgement on their own efforts and were ready to remedy shortcomnings to ensure that pupils were permitted to sit the examninations. Patterns in Anambra reflected the dynamism of communities and the strong demand for education in this part of Nigeria. Applying such con- trols elsewhere might not be easy. However, the model indicates what is possible when community resources can be mobilized and demand for education is high. Source: Okoye 1986. 34 DECENTRALIZATION OF EDUCATION: COMMUNITY FINANCING especially desirable for school buildings because of the status and role of education. * In many remote areas schools can be built only with locally available materials. It may be impossible to carry roofing sheets, metal windows, and cement to such areas. * Villagers may find it easier to maintain buildings when they are familiar with them and have built them from local rather than imported materials. o If a school already has some buildings that do not meet the govern- ment's standards, demolishing these buildings can do more harm than good. It can destroy the very spirit that the government is try- ing to encourage. * Many government buildings are themselves low standard, and it could be both hypocritical and unfair to require self-help com- munities to put up buildings of a higher standard than govern- ment ones. There are also counter arguments in favor of government insistence on standard designs and minimum quality in construction: * Smart buildings can be a source of pride and can raise the prestige of education. - Roofs and buildings needs to be adequately constructed to with- stand heavy rainfall. * Furniture, books, and equipment are expensive and should be pro- tected from rain, termites, and thieves. * Villagers may not know where to locate chalkboards to avoid glare (an especially difficult task in round buildings). * Village houses-even for chiefs and headmen-often exclude so much daylight that prolonged reading and writing are difficult and bad for the eyes. * Some buildings may be cheap to construct but have such high main- tenance costs and short life spans that it is wiser to build more expensive but sturdier ones. Because each of these sets of arguments has force, it is no surprise that patterns vary in different countries. Some authorities consider it appropriate to insist on standard designs, but others are more flexible. A compromise policy followed in Lao People's Democratic Republic is for communities to construct the buildings but for governments or external agencies to provide zinc roofs. A zinc roof is much more effec- tive than a grass or palm-frond one for protecting the rest of the build- ing, but zinc roofs are also costly and therefore difficult for communi- ties to provide by themselves. GOVERNMENT INCENTIVES, GUIDANCE, AND CONTROLS 35 Minimum School Size Some governments seek to ensure that schools are viable and have reasonable unit costs by insisting on minimum enrollments and school sizes before they will be officially registered. While the pre- cise requirements would have to be determined locally in each case, the minimum appropriate size is usually greater for secondary than for primary schools. In Uganda, Cumming and others (1995) recom- mended that no new secondary school should be approved unless: * Enrollment of 160 students could be guaranteed as day pupils from a population no more than 8 kilometers from the school site. * At least 120 students were enrolled in arts and science, or 100 stu- dents in arts only, for higher school certificate classes. * Classrooms provided at least 15 square feet per student, and labo- ratories, at least 20 square feet. Controls on minimum school size also discourage community splits in which individual villages have two or more schools. Such regulations, however, might mean that some children are denied access to any education simply because they inhabit sparsely popu- lated areas or live in communities that are unenthusiastic about education. Maximum Fees Governments concerned about the potential for exacerbating social inequalities may set regulations on the maximum levels of charges that can be imposed on parents and other members of the community. In Ghana, for example, elimination of unauthorized PTA levies was made a condition for government assistance in the Primary School Development Project, a point emphasized in workshops for communi- ty leaders (World Bank 1993d). Official regulations have also placed ceilings on charges in Cameroon, Myanmar, Tanzania (Evans and Rorris 1994; Galabawa 1994; Tembon 1994). One problem with such regulations, however, is that they may be, or may become, unrealistic. The proclamations may serve a political purpose, assuaging the doubts of those local and international observers who do not scrutinize the situation too closely, but unless the ceilings are appropriate in the first place and are frequently raised in line with inflation, the practicalities of making ends meet are likely to require school managers to ignore regulations. Evans 36 DECENTRALIZATION OF EDUCATION: COMMUNITY FINANCING and Rorris (1994), investigating patterns in Myanmar in 1993-94, found that regulations on maximum PTA levies set more than a decade earlier were still in force. Although the regulations set a ceil- ing on those levies of 10 kyats per child, among the three primary schools that the researchers examined in detail, one charged 15 kyats per child, and the other two charged 20. Had the schools been pressed, they would simply have renamed the charges as contribu- tions to school funds rather than levies. Galabawa (1994) reports similar patterns in Tanzania, where regulations are evaded by calling the extra charges "commitments." Challenges and Risks Some challenges and risks have already been identified, but several additional ones need to be recognized and elaborated upon. The list of effective instruments for incentives, guidance, and control may look logical, but their use may encounter difficulties. Political Dimensions and the Distribution of Authority Finding an appropriate distribution of authority and accountability within the government hierarchy is a challenge. The majority of detailed decisions on incentives, guidance, and control can be made only at the local level, where the specific circumstances of individ- ual communities can be identified and responded to. However, national governments may be reluctant to delegate strong decision- making power to local officers. This may be partly because they doubt if all local officers have sufficient capacity to make appropri- ate decisions within the overall vision of the national government. There is also danger of acrimony when some communities find that they have been treated less generously than others by local admin- istrative officers. At all levels in the system tension may also arise between civil ser- vants and politicians. Civil servants are often suspicious of the motives of politicians, and politicians often resent the constraints imposed upon them by civil servants. Tembon's (1994, 255) views are primarily on the side of the planners when she points out that in Cameroon many new schools experience resource shortfalls even before they begin: because the decision to create a government secondary school in any given area is often a political one. Members of parlia- ment and/or other influential members of government per- suade the Minister of National Education who signs a decree creating a school irrespective of whether or not buildings exist. As soon as a school is created, community leaders, in collabo- ration with officials from the Ministry of National Education, will choose an appropriate site for the school. Once the site is 37 38 DECENTRALIZATION OF EDUCATION: COMMUNTrY FINANCING known, members of the community then start to raise funds for the construction of school buildings. The school might begin at a temporary site initially while classrooms are constructed as the funds become available. In addition to causing major problems for the institutions themselves, Tembon adds, this process results in some areas having many schools while others which have fewer politically influential people are deprived. Rival political groups can simply destroy each others' efforts. Dandekar (1996) gives an example from India in which one political group's work to initiate a secondary school in a particular village caused another political group to launch a competing school in the same village. The result was that both schools collapsed, and the two groups were unwilling to compromise on a single, joint institution. As a consequence, the students had to walk 6 kilometers each day to attend a school in a neighboring village. Such scenarios understandably make government planners cau- tious. They might, however, overreact and fail to find the best bal- ance of forces. Barkan and Homquist (1989) have ably analyzed this in the Kenyan context. Beginning with the peasants' perspectives, they note (377-8) that, "As private landowners, the majority of peasants support the patron-client capitalist structure of the Kenyan political economy. But they also perceive abuses in the form of cor- ruption, misuses of office, ethnic favoritism, and a great political distance between themselves and the dominant elites that control the Kenyan state." As in other countries, one response from govern- ment officials has been to limit both the incidence of uncontrolled initiative from below and the involvement of members of parlia- ment and other elected personnel seeking to expand their political bases. By exerting greater control over the planning and initiation of new projects, the state tends to give a green light only to a select group of viable projects; but in the process it dampens initiative out- side the plan. Initiative in this type of situation lies primarily with the state. Local involvement diminishes, and the balance shifts from self-help projects with possible state participation to state projects with pos- sible local participation. The result, Barkan and Holmquist (1989) point out, is a risk of overregulation and choking off of peasant ini- tiative at a time that the state is faced by financial stringency and is keen to shift a greater proportion of the costs of local services back to the communities. CHALLENGES AND RISKS 39 Disparities and Encouragement of Less Active Groups Allied to questions about the locus of control is the possibility that regional and other disparities will be widened by local decisionmak- ing. As already noted, inequalities are inherent in the whole concept of community financing. One practical question confronting deci- sionmakers concerns the effects of different incentives and controls not only on the communities to which they are targeted but also on other communities that are interested observers. It is useful to take the hypothetical case of one part of a country in which the people do not commonly act together as a community and have not expressed much interest in schooling. In this case, should the government itself undertake most, or even all, construction? That would at least get the schools built, would in turn provide the opportunity for families to learn about the value and nature of education, and would reduce regional inequalities in access. It might, however, convey a signal that communities can gain "free" resources from the government simply by refusing to organize themselves. It could also create resentment from better-organized neighboring communities. These communities might feel penalized for their hard work because the government resources flow to the less developed area in which the people are not only poorly organized but also less interested in receiving the government's resources. In this type of situation the government faces a risk whatever it does. Some authorities react by avoiding the issue. Either they adopt a fully market-oriented strategy in which they respond to requests from active communities and leave passive ones to suffer the consequences of their passivity, or they prohibit all elements of comnmunity financing and instead operate a centralized system dominated by government distribution of resources. Both of these responses would miss oppor- tunities, however, and would lead to further problems. More positively, governments can employ locally based community- development officers who are responsible for working with less active groups and help to identify both needs and strategies to meet those needs. Such personnel require skills in consciousness raising, promo- tion of cohesion, and management of conflict. On the practical level, they would advise on the location of schools, seeking to ensure that schools serve population large enough but not so large that individual microcommunities find it difficult to identify with the institutions. They would also advise on management practices, fundraising procedures, and other matters required for effective and sustained operation of institutions. 40 DECENTRALIZATION OF EDUCATION: COMMUNITY FINANCING Scale of Operation Successes in small projects cannot always be multiplied in large ones. Governments that wish to expand enrollments considerably, for exam- ple, to achieve universal primary or junior secondary education, may decide that it can only be done with community participation, but the authorities would experience problems simply because of the size of the task. This has been the experience in Ghana's Community Secondary Schools Construction Project, which was designed to build 140 sec- ondary schools in underserved areas (World Bank 1991). As a national project, it had to produce a national formula balancing government and community inputs despite the fact that the project designers were aware of great variations in capacity around the country. As formula for this project, communities would contribute one- third of the construction costs for schools, chiefly through clearing the ground, preparing the foundations, and supplying some raw materi- als. The government would cover the remaining two-thirds. However, two years into the project, the construction phase was already behind schedule. This was partly because communities rarely operate on rigid time schedules in the ways that ministries and external agencies do and partly because the project designers had overestimated the extent to which communities could be mobilized. As a result, many contrac- tors who had expected buildings to be one-third complete before com- mencing their own work found that they had to start at lower levels; because payments from the ministry covered only fixed amounts of work, buildings were being left uncompleted (World Bank 1993b). The project was therefore modified to allow for community inputs worth just 10 percent of construction costs. In one way this was an admirable example of flexibility, but it could also be seen as a reward for those communities that did not meet their side of agreements and a penalty for those communities that did. In at least two instances district inspection teams made false certifi- cations that one-third of the work had been completed. The ministry therefore made payments, only to find that the work had not been done and that earlier advance payments had not been used as agreed. Such false reporting is a regrettable fact of life in many countries, but it is a particular challenge to large projects in which the location and the center are separated by several layers of bureaucracy and by long distances. To avoid such problems in the future, project personnel rec- ommended cross-certification of reports by technical auditors. This consumed further resources on the government side, however, and further bureaucratized the project. CHALLENGES AND RISKS 41 Gopal and Marc (1994) have identified other challenges arising from large-scale operation. The fact that large projects generally require competitive bidding favors large corporations and discriminates against smaller communities, most of which are not even legal entities. Also, the pace at which the bureaucracies of large projects operate can- not easily be coordinated with the paces at which communities work. Either the bureaucracies move too slowly, in which case projects fail to catch the tides of community enthusiasm and activity, or the bureau- cracies move too fast, in which case decisions are forced on communi- ties before they have completed the necessary consultation processes. Recognizing these difficulties, the World Bank has produced a manual to help project designers be more sensitive to the issues and in partic- ular to improve procurement and disbursement mechanisms for pro- jects involving community participation (World Bank 1994a). Preparation of this manual is a major step forward, but most project designers agree that many challenges remain. Monitoring and Evaluation Monitoring and evaluation are important for almost all kinds of poli- cies. They are especially important in community financing. So much depends on the cultures of particular communities and on the dynam- ics of individual partnerships that predicting outcomes in specific cir- cumstances may be far from easy. Without careful monitoring and evaluation, it is impossible to achieve the necessary fine-tuning and adjusting of policies. In most countries the quality of data on community financing and on the impact of government policies leaves much to be desired. Govermnents commonly lack detailed information even on how their own funds are spent in partnership schemes, let alone on the expendi- tures of others. These facts stress the need for much more extensive evaluation, monitoring, and research. Especially needed are detailed case studies to show the dynamics of operation in communities of different types and at different times. Such studies would expose the shortcomings of partnership schemes as well as their successes. Although research is Box 8. Steps for Designing Large-Scale Community-Based Projects 1. Clarify, simplify, and set priorities for objectives; link them to outputs. 2. Identify the major social actors, capacity, and interests at community and agency levels. 3. Assess demand. 4. Craft a self-selection process for subprojects, groups, or communities. 5. Structure subsidies that do not violate demand. 6. Restructure fund release to support demand. 7. Plan for learning and for several models. 8. Invest in outreach and social organization. 9. Institute participatory monitoring, evaluation, and feedback. 10. Redefine procurement rules to support community-level procure- ment where possible. Source, Narayan 1995. 42 MONITORING AND EVALUATION 43 sometimes costly, these studies could generate considerable returns by identifying the ways in which existing resources could be used more wisely and in which additional resources could be raised for the edu- cation sector. Referring particularly to projects operated by the World Bank and similar agencies, Narayan (1995, 39) makes an important further point. The temptation, she points out, is to freeze project designs once imple- mentation starts. Even though much is written in staff appraisal reports about learning, monitoring, and evaluation, action is not always taken during project implementation. Learning is more effective and efficient when feedback is listened to and when changes are put in place. The process of listening to practitioners may be empowering and may encourage innovation and responsibility. For this reason, Narayan included outreach and participatory monitoring as important steps in the design of community-based projects (box 8). Narayan's concern was with large projects, but the point is equally applicable to small ones. Conclusion Debates about the appropriate locus of control within administrative hierarchies for education are vigorous and seemingly defy resolution (see, for example, Lauglo 1995; McGinn 1992; Rideout and Ural 1993). The forces that may be identified as justifying centralized, mixed, or decentralized systems are complex and variable. No models can be recommended with universal applicability, and even in specific places demands are likely to change over time. From the official perspective, much of the attractiveness of commu- nity financing lies in the extent to which it is able to relieve govern- ments of the burden for educational financing by providing an alter- native source for channelling resources to the education sector. Where governments are prosperous, community financing seems to attract relatively little official attention, but where governments are hard pressed, it is given more recognition. The financial crisis that has beset many governments during the 1980s and 1990s has greatly increased the interest with which governments look at the phenomenon of com- munity financing. This, however, is a somewhat negative reason for interest. More positive reasons are linked to the value of partnership and joint responsibility in enhancing commitment to the education enterprise. Yet enthusiasm for community financing is not always accompanied by enthusiasm to relinquish controls in the education system. Thus PTAs are widely seen as mere fundraising bodies for schools, and it is not necessarily the case that the one who pays the piper is pernitted to call the tune. This comment in a World Bank report on Ghana (1993d, 15) seems to have wide applicability, "PTAs are mainly used to raise development funds for the school, or to help organize commun- ity labor. PTA members are actively discouraged from becoming involved in any management issues, and teachers and district officials are horrified at the idea of having non-professionals supervise the work of teachers." The Ghanaian government has taken a major step toward changing this situation by involving community leaders in the selection of head teachers. However, the long-term effectiveness of this step remains to be evaluated. Meanwhile, few education systems elsewhere in Africa, 44 CONCLUSION 45 or indeed more widely, accord communities strong roles in decision- making within public systems. Interestingly, in some societies the willingness to decentralize weakens as the competence and expertise of administrators at the cen- ter increases. Robinson (1988, 194) has made this point about China, where criticism of community-managed schools grew as the personnel in the Ministry of Education and provincial offices gained experience and skill: It is not only the increasing competence of experts that is crucial but also the growing gap in knowledge between experts and the general population. Especially in terms of education, peasants who at best have only basic literacy skills cannot compete with highly trained officials. Whereas decisions about hiring teachers, choosing texts, and writing examinations were once made at the local level, these functions are now performed at provincial or central levels. This is because experts are concentrated at these levels and expertise is valued in its own right. This analysis presents another reason for the variation in patterns in different countries. Some societies may have decentralization by default because the central administrators lack the competence and professionalism to operate a tightly controlled system; others, in slightly more sophisticated societies, have more centralized systems because the officers, as described by Robinson, perceive a substantial gap between their competences and those of the periphery; and yet more sophisticated societies may have high degrees of decentraliza- tion because the periphery has both the necessary skills and sufficient political strength to demand the power to take local decisions. Meanwhile, in some societies aspects of government ambivalence about community activities are well founded. This study has highlighted the dangers of community financing both exacerbating regional, rural- urban, and social inequities and contributing to what planners may see as an inappropriate use of resources. These issues justify the use of vari- ous mechanisms for guidance and control. In other contexts, planners may consider it desirable to encourage community financing. The chal- lenge is to find tools that can encourage desirable forms of community financing in particular locations and among particular communities without creating other political, social, and economic problems. One further lesson from analysis of experiences in different settings is that circumstances vary widely within as well as between countries. This has been illustrated by Okoye's (1986) study of Nigeria's 46 DECENTRALIZATION OF EDUCATION: COMMUNITY FINANCING Anambra State. She pointed out that different policies were needed for different parts of the state. In the rest of the federation, the divergence of conditions was wider still. Different policies may also be needed for secondary and primary schools. Secondary schools generally require greater capital invest- ments in laboratories, libraries, and equipment and also have higher per capita running costs because teachers are usually paid higher salaries. Secondary schools may also demand wider catchment areas, which, particularly in the case of rural institutions, may require sever- al communities to operate together for a single school. Such factors may require stronger government financial inputs at the secondary level and more careful attention to incentives, guidance, and controls. Also apparent in this study has been the need to treat different types of communities in different ways. For political as well as practical rea- sons, the nature of government relationships with communities based on ethnicity, race, and religion is likely to be rather different from that concerning geographically defined communities. Malaysia's elite independent Chinese schools, for example, are very different from Myanmar's poorly endowed affiliated schools. Such diversity does of course create challenges, not only for gov- ernments but also for such agencies as the World Bank. Forming of effective and productive partnerships with communities may be espe- cially difficult in large projects. However, large does not necessarily mean inflexible, and with further effort, improved understanding can be reached about the optimal balances between government and com- munity roles in the financing of education in a range of settings and at different times. References Abreu, Elsa. 1982. The Role of Self-Help in the Development of Education in Kenya 1900-1973. Nairobi: Kenya Literature Bureau. Assie-Lumumba, N'Dri Therese. 1993. Couits,financement de l1education de base et partici- pation des familles et communautes rurales dans les pays du Tiers Monde. Paris: Institut International de Planification de l'Education. Barkan, Joel D., and Frank Holmquist. 1989. "Peasant-State Relations and the Social Base of Self-Help in Kenya." World Politics 41(4):359-80. Bhatnagar, Bhuvan, and Aubrey C. Williams. 1992. "Introduction." In Bhuvan Bhatnagar and Aubrey C. Williams, eds., Participatory Development and the World Bank: Potential Directions for Change. Discussion Paper 183. Washington, D.C.: World Bank. Bhutan, Govemment of. 1992. Ninth Quarterly Policy Guidelines and Instructions. Thimphu: Department of Education. 1995. General Statistics, May 1995. Thimphu: Education Division, Ministry of Health and Education. 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. Bray, Mark. 1987. New Resources for Education: Community Management and Financing of Schools in Less Developed Countries. London: Commonwealth Secretariat. - 1988. "Community Management and Financing of Schools in Papua New Guinea." In Mark Bray with Kevin Lillis, eds., Community Financing of Education: Issues and Policy Implications in Less Developed Countries. Oxford: Pergamon Press. - . 1995. "Macau." In Paul Morris and Anthony Sweeting, eds., Education and Development in East Asia. New York: Garland. - . 1996a. Counting the Full Cost: Parental and Community Financing of Education in East Asia. Washington, D.C. and Bangkok: World Bank-UNICEF. - 1996b. "Equity Issues in Local Resourcing of Education: Community Financing of Primary Schools in Bhutan." International Review of Education 42(4). Bray, Mark, with Kevin Lillis, eds. 1988. Community Financing of Education: Issues and Policy Implications in Less Developed Countries. Oxford: Pergamon Press. Cheng Kai-ming. 1990. "Financing Education in Mainland China: What Are the Real Problems?" Issues and Studies 26(3):54-74. Colletta, Nat J., and Gillian Perkins. 1995. "Participation in Education.' World Bank, Environment Department, Washington, D.C. Cornia, G. A. 1987. "Social Policy-making: Restructuring, Targeting, Efficiency." In G. A. Comia, R. Jolly, and F. Stewart, eds., Adjustment with a Human Face: Protecting the Vulnerable and Promoting Growth. 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. 47 48 DECENTRALIZATION OF EDUCATION: COMMUNITY FINANCING Dandekar, V. S. 1996. "Community Resource Mobilization." Paper presented at the con- ference on Reforming School Education, National Institute of Educational Planning and Administration, February 1-2, New Delhi. De Stefano, Joseph. 1996. "Community-Based Primary Education: Lessons Learned from the Basic Education Expansion Project (BEEP) in Mali." Technical Paper 15. United States Agency for International Development, Washington, D.C. Evans, Kiri, and Adam Rorris. 1994. Cost-Effectiveness of Primary Education in Myanmar. Yangon: UNICEE 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. Galabawa, J. C. J. 1985. "Community Financing of Schools in Tanzania." Paper present- ed at the Commonwealth Secretariat Workshop on Community Financing of Schools, Gaborone, June 7-10, Botswana. - 1994. "Characteristics, Financing, Unit Costs and Selection Issues of Non- Government Secondary School Provision in Tanzania.' Paper prepared for the Workshop on Investing in Human Capital, April 15-20, Arusha. Gannicott, K., and P. McGavin. 1987. "Resource and Management Requirements for Recurrent Financing of Primary and Secondary Schools in Solomon Islands." Ministry of Education and Training, Honiara. Gibson, U Saw. 1992. "Non-Government Expenditure on Education." Working Paper 4.3, Education Sector Study Project. Myanmar Education Research Bureau, Yangon. Gillespie, Roselyn R., and Colin B. Collins. 1988. The Fete and the Flea Market: The Many Faces of Productive Labour in Schools. Report S.134. Paris: UNESCO, Division of Educational Policy and Planning. Gopal, Gita, and Alexandre Marc. 1994. World Bank-Financed Projects with Community Participation: Procurement and Disbursement Issues. Discussion Paper 265. Washington, D.C.: World Bank. Gopinathan, S. 1995. "Singapore." In Paul Morris and Anthony Sweeting, eds., Education and Development in East Asia. New York: Garland. Hiebert, Murray. 1995. "Golden Tongue: Non-Chinese Flock to Mandarin-Language Schools." Far Eastern Economic Review 158(29). Hill, Martin I. 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. lgwe, 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. Jimenez, Emmanuel. 1987. Pricing Policy in the Social Sectors: Cost Recovery for Education and Health in Developing Countries. Baltimore: Johns Hopkins University Press. 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 Planfor the Period 1994 to 1996. Nairobi: Government Printer. Korboe, David. 1995. "Access and Utilisation of Basic Social Services by the Poor in Ghana." University of Science and Technology, Kumasi. Lauglo, Jon. 1995. "Forms of Decentralisation and their Implications for Education." Comparative Education 31(1):5-29. Lewin, Keith, with Dieter Berstecher. 1989. "The Costs of Recovery: Are User Fees the Answer?" IDS Bulletin 20(1):59-71. REFERENCES 49 Lillis, Kevin, and Henry Ayot. 1988. "Community Financing of Education in Kenya." In Mark Bray with Kevin Lillis, eds., Community Financing of Education: Issues and Policy Implications in Less Developed Countries. Oxford: Pergamon Press. Malawi, Government of. 1992. Report on Primary Education in Malawi. Lilongwe: Ministry of Education. -1 1993. Basic Education Statistics 1993. Lilongwe: Ministry of Education. 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. Martin, Anita H. 1996. "A Case Study of School-Parent-Community Partnerships via the Parent Teacher Associations (PTA)." UNICEF, East Asia and Pacific Regional Office, Bangkok. Mauritius, Government of. 1991. Master Plan for the Year 2000. Port Louis: Ministry of Education, Arts, and Culture. Mbithi, Philip M., and Rasmus Rasmusson. 1977. Self-Reliance in Kenya: The Case of Harambee. Uppsala: Scandinavian Institute of Development Studies. McGinn, Noel F. 1992. "Reforming Educational Governance: Centralization/ Decentralization." In Robert F. Amove, Philip G. Altbach, and Gail P. Kelly, eds., Emergent Issues in Education: Comparative Perspectives. Albany: State University of New York Press. Moorad, E R. 1989. "An Evaluation of Community Junior Secondary Schools in Botswana with Particular Emphasis on Community Participation and Decentralization." Ph.D. diss, University of London, Institute of Education. - . 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. Muya, D., I. Wan'gombe, J. Muthuri, M. Mwaura, P. Wanjala, S. Cottingham, and V. Sathyabalan. 1995. Sixteen Years of ACTIONAID's Support to Primary Education in Kenya: A Review. London: ACIIONAID. Mwiria, Kilemi. 1990. "Kenya's Harambee Secondary School Movement: The Contradictions of Public Policy." Comparative Education Review 34(3):350-68. Narayan, Deepa. 1995. "Designing Community-Based Development." Paper 7. World Bank, Environment Department, Washington, D.C. 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. O'Grady, Barbara. 1995. Teaching Communities to Educate Girls in Balochistan. Washington, D.C.: Academy for Educational Development. Opolot, Jethro A. 1994. "Study on Costs and Cost-Effective Approaches to Primary Education in Uganda." UNICEF, Kampala. Preston, Rosemary. 1991. "Organisational Perspectives on the Local Management of Schools: Papua New Guinean Perspectives." Intenational Journal of Educational Development 11(4):275-88. Rath, Seema P. 1996. "Financing of Primary Education in Goa." Paper presented at the conference on Reforming School Education, National Institute of Educational Planning and Administration, New Delhi. Rideout, William, and Ipek Ural. 1993. Centralised and Decentralised Models of Education: Comparative Studies. Policy Working Paper 1, Halfway House. Johannesburg, South Africa: Development Bank of Southern Africa. 50 DECENTRALIZATION OF EDUCATION: COMMUNITY FINANCING 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 Universalised Primary Education in Vietnam: A Study of the Cost and Cost Effectiveness of the Primary Education System. Hanoi: UNICEE Ruda, Gaonosi Buster. 1993. "Implications of Community Financing of Community Secondary Schools in Botswana." M.Ed. diss., University of Botswana. Samoff, Joel. 1990. "Modernizing a Socialist Vision: Education in Tanzania." In Martin Carnoy and Joel Samoff, eds., Education and Social Transition in the Third World. Princeton: Princeton University Press. Shaeffer, Sheldon. 1994. Participation for Educational Change: A Synthesis of Experiences. Paris: International Institute for Educational Planning. Siddhanti, K. A. 1996. "Financing and Resource Mobilization for School Education: A Study of Microlevel Experiences." Paper presented at the conference on Reforming School Education. National Institute of Educational Planning and Administration, February 1-2, New Delhi. Straits Times. August 9, 1992. "Sinda Plans Donation Draw to Raise $2.5 m." Swartland, J. R., and D. C. Taylor. 1988. "Community Financing of Schools in Botswana." In Mark Bray with Kevin Lillis, eds., Community Financing of Education: Issues and Policy Implications in Less Developed Countries. Oxford: Pergamon Press. Tan, Jason. 1995. "Joint Government-Malay Community Efforts to Improve Malay Educational Achievement in Singapore." Comparative Education 31(3):339-53. Tan, Liok Ee. 1988. "Chinese Independent Schools in West Malaysia: Varying Responses to Changing Demands." In Jennifer Cushman and Gungwu Wang, eds., Changing Identities of the Southeast Asian Chinese since World War II. Hong Kong: Hong Kong University Press. Tamukong, Joseph Azia 1995. "An Analysis of Local Support in Financing Secondary School Education in the North West Province of Cameroon." Ph.D. diss., University of Jos, Nigeria. 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. Tilak, Jandhyala B. G. 1989. "Financing and Cost Recovery in Social Sectors in Malawi." World Bank, Southern Africa Department, Population and Human Resources Operations Divison, Washington, D.C. Togo, Government of. 1994a. Esquise d'une nouvelle approche de la contribution des parents d'eleves aufinancement de l'education. Lome: Ministere de l'Education Nationale et de la R&herche Scientifique. .1994b. Analyse de la situation des enseignements du primaire pris en charge par les com- munautes locales. Lome: Ministere de l'Education Nationale et de la Recherche Scientifique. Tsang, Mun C. Forthcoming. "Financial Reform of Basic Education in China." Economics of Education Review. Wellings, Paul Anthony. 1983. "Unaided Education in Kenya: Blessing or Blight?" Research in Education 29(11-28). Wilson, H. E. 1978. Social Engineering in Singapore: Educational Policies and Social Change 1819-1972. Singapore: Singapore University Press. Wilson, L. S. 1992. "The Harambee Movement and Efficient Public Good Provision in Kenya." Journal of Public Economics 48(1):1-19. REFERENCES 51 World Bank. 1989. "Indonesia: Basic Education Study." East Asia and Pacific Regional Office, Country Department III, Washington, D.C. .1991. "Staff Appraisal Report: Republic of Ghana Community Secondary Schools Construction Project." West Central Africa Department, Washington, D.C. .1993a. "Bolivia Education Sector: Sector Development and International Assistance." Latin America and the Caribbean, Country Department II[, Washington, D.C. - 1993b. "Ghana-Community Secondary Schools Construction Project, Supervision Report." West Central Africa Department, Washington, D.C. - 1993c. "Republic of Chad: Basic Education Project-Staff Appraisal Report." Central and Indian Ocean Department, Washington, D.C. . 1993d. "Staff Appraisal Report: Republic of Ghana Primary School Development Project." West Central Africa Department, Washington, D.C. 1993e. Uganda: Social Sectors. World Bank Country Study. Washington, D.C. 1994a. "Bank-Financed Projects with Community Participation: A Manual for Designing Procurement and Disbursement Mechanisms." World Bank, Washington, D.C. - 1994b. "El Salvador Community Education Strategy: Decentralized School Management." Latin America and the Caribbean, Central America Country Department, Washington, D.C. . 1994c. "Nepal: Critical Issues in Secondary Education and Options for Reform." South Asia, Country Department II, Washington, D.C. . 1995. "Staff Appraisal Report: Republic of Togo Education Rehabilitation Project." West Central Africa Department, Washington, D.C. Zambia, Ministry of Education. 1993. "Decentralization of Education Services and Increase in Local Self-Governance and Community Involvement through Introduction of School and District Management Boards: Recommendations of a Ministry of Education Committee." Ministry of Education, Lusaka. 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 Counting the Full Cost: Parental and Community Financing of Education in East Asia Decentralization of Education: Community Financing Decentralization of Education: Politics and Consensus Decentralization of Education: Teacher Management 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 Monitoring the Learning Outcomes of Education Systems Nurturing Development: Aid and Cooperation in Today's Changing World Nutrition in Zimbabwe: An Update Private and Public Initiatives: Working Together for Health and Education The Transition from War to Peace in Sub-Saharan Africa THE WORLD BANK A partner in strengthening economies and expanding markets to improve the quality of life for people everywhere, especially the poorest HEADQUARTERS 1818 H Street, N.W. Washington, D.C. 20433, U.S.A. TELEPHONE: (202) 477-1234 FACSIMILE: (202) 477-6391 TELEX: MCI 64145 WORLDBANK MCI 248423 WORLDBANK CABLE ADDRESS: INTBAFRAD WASHINGTONDC INTERNET: http://www.worldbank.org// EUROPEAN OFFICE 66, avenue d'1ena 75116 Paris, France TELEPHONE: (1) 40.69.30.00 FACSIMILE: (1) 40.69.30.66 TELEX: 640651 TOKYO OFFICE Kokusai Building 1-1, Marunouchi 3-chome Chiyoda-ku, Tokyo 100, Japan TELEPHIONE: (3) 3214-5001 FACSIMILE: (3) 3214-3657 172 TOVEL DEX: GN: 26838 MGZNGOPIN-21-32-9 780821 337240 COVER DESIGN: THE MAGAZINE GROUP ISBN 0-8213-3724-6