Case- Studies- in Financing Quality gasic Ehrcation TEe World Bank Mducation and Social Pblicy-Department FILE COPY SbiaONdIEEMGbtUf UTERNWM NSCVERUn@CONPWAMM  ::;!‘…………乌…  Case Studies in Financing Quality Basic Acknowledgements These case studies, co-edited by Adriaan Verspoor and Mun C. Tsang, provide the background for the Bank-sponsored presentation on "Financing Quality Basic Education" for the second meeting of the EFA Forum at New Delhi in September, 1993. Mun C. Tsang is the author of the Introduction and of the case study on China. The case study on Chile was co-authored by Donald Winkler and Taryn Rounds. David Plank is the author of the Brazilian case study. Staff in the Education and Social Policy Department prepared the case study on India. The Sahelian case study was prepared by staff in the Population and Human Resources Operations Division of the Sahelian Department. Staff in the Population and Human Resources Division of the Middle East and North Africa Country Department 1B (MN2CO) prepared the case study on Yemen. Theresa Moran provided editing and coordinated the case studies' publication. Many thanks go to all who contributed to the development of these studies and much appreciation goes to Margot Verbeeck for her invaluable assistance and desktop publishing skills. These case studies are not a formal publication of the World Bank. The findings, interpretations, and conclusions expressed in these case studies are entirely those of the author(s) and should not be attributed in any manner to the World Bank, to its affiated organizations, or to members of its Board of Executive Directors or the countries they represent. The case studies on Chile and on China were previously presented at the International Symposium on the Economics of Education in Manchester, England, May 18-21, 1993. ia  Case Studies in Financing Quality Basic Table of Contents Page Acknowledgements ............................................. i Introducion ................................................. 1 I. Issues in the Financing of Quality Basic Education ................... 2 Resourc Reallocation for Quality Improvement .................. 2 Resource Mobilization for Basic Education ........ ........... 2 Targeting Resources to Benefit Females and Marginalized Populations .... 3 Decentralization ...................................... 5 II. Summary of Case Studies ................................... 6 The Sahelian Counties .................................. 7 Yemen ............................................ 7 C ins ............................................ 8 Chile ............................................. 8 India ............................................. 8 Brazil ............................................ 9 Financing Qualty Basic Educarion in the Sahel .......................... 12 1. Resource Mobilization for Basic Education ....................... . 14- More Efficient Resource Use............................. 15 Reallocation of Education Resources in Favor of Primary Education .... 23 Mobilizing Outside Resources ............................ 24- Conclusion ........................................ 25 II. Quality of Basic Education .................................. 25 Repetition and Drop-out in Primary Education .................. 26 Inservice Teacher Training .............................. 29 Textbooks......................................... 30 School Feeding Programs ............................... 30 Administrative Control and 1ocal Development ................. 30 V Case &xdies i FLnescdM QaIY Bastc Educian M. Basic Education for Girls .................................. 32 IV. Conclusion ........................................... 33 Republic of Yemen: Hunan Development, Societal Needs and Human Capital Response . 40 I. The Financing of Education: An Overview ...................... 42 II. Increasing Female Participation in Primary Education ................ 44 M. Quality Improvement in Basic Education ........................ 46 IV. Key Issues and Problems in the Financing of Basic Education ........... 47 V. Conclusion ........................................... 51 Financial Reform of Basic Education: The Chinese Experience ................. .53 I. Socio-Economic Context of Reform ........................... 53 U. Decentralization and Diversification ........................... .55 Decentralization ..................................... 55 Diversification ...................................... 56 Ill. An Assessment of the Financial Reform of Basic Education ............. 58 An Emerging Financing System for Basic Education .............. 59 Increased Resource Mobilization ........................... 59 Low Investment and Difficulties in Poor Areas .................. 62 Large Disparities ...................... . .*.* . ..... .64 Inefficiency and Management Weakness ...................... 65 IV. Lessons and Further Development ............................ 65 Increasing Resources to Basic Education ...... 66 Assisting Disadvantaged Areas and Populations ................. 68 Improving Efficiency .................................. 68 Education Reform in Cile: Decemraliation and Sdool Ooice ................ 72 1. Education Reform in Chile ................................. 72 II. Municipal and Private Sector Response ......................... .73 Revenues ......................................... 73 Changing Market Share ................................ 74 Demand for Private Education ............................ 74 Supply of Private Education ............................. 76 Model of Private School Market Share ....................... 77 Vi Tabe of CmmsIn M. The Equity and Efficiency Consequences of the Reform . 79 Equality in Spending .................................... 79 Equality in Outcomes .. ................................. 80 Effects on Costs....................................... 81 Cost-Effectiveness ................................... 82 IV. Conclusions .......................................... 85 Financing Quality Basic Educarion in India ............................. 88 1. Problems of Educational Development in India .................... 90 II. The New Direction in Indian Elementary Education: The National Policy of Education............................................ 93 M. The Uttar Pradesh Basic Education Project ....................... . 94 Community Participation . ............................... 97 Institutional Development ................................. 97 Improving Information Systems for Planning, Management and Professional Support .......................................... 98 Conclusion ........................................ 98 Financing Basic Education in Brazil ................................. 114 I. The Quality of Basic Education .............................. 115 II. Educational Finance in Brazil ................................ 117 M. Policy Options in Educational Finance .......................... 120 Administrative Decentralization ............................. 120 Diversification of Revenue Sources .......................... 121 Targeting Resources to Marginalized Populations ................ 122 Transparency and Accountability .......................... 123 Conclusion ........................................ 124 vii  子―-------------------------,--(- Caw SmAn in Plnemcdne Ousft Bwk Edumgm 1. ISSUIES IN THE MANCING OF QUALXrV BASIC EDUCATION Resource ReaUocafton for QuaUty Imprvvemext To achieve quality basic education for all, existing resources mist be used efficiently. Interventions in the education process have to be undertaken to allocate resources among the different inputs in such a way that they improve pupil learning with maximum cost-effectiveness. In its 1990 policy paper on primary education, the World Bank identified three categories of policy measures related to I i : provision of inputs that increase learning, such as tcabooks and. improved curricula; cost-reduction measures that, within reisonable ranges, have little negative effect on learning, such as multiple shifts and multigrade classes; and inputs dun indirectly improve learning by improving educationzoana ement such as incentives for teachers and information system. For example, the Republic of Yemen is currently wairin an effort to improve education quality by reallocating its resources. Mie Yemeni govemment has a specific budget item earniarked for quality improvements in basic education such as curriculmn and tettbwk I - - -1 -,- - and in4ervieg tearhin training. This budget item will increase over time so that by the yew 2000 ft will be approximately 700 million Rials (about one percent of total primary education mr; . Many of the resources used for quality improvement will come from savings realized Wough the 0 n a of the teachin profession. Local teachers are replacing more experienced outside teachers and are being trained to upgrade their pedagogical skills and topic knowledge. A study of rural northeastern Brazil by R. Harbison. and E. Hanushek argues that in many cam quality improvements in basic education can over time be . . The study shows that quality UMLLV .-.1 can be P ip P P! P d to lead to gain in efticiency (such as a reduction in the number of students that repeat grades or drop out) that more than offset the direct costs of such imptovements. In fwt quality i tovements can make schooling in= equally accessible by fiveing up places occupied by repeaters for new entrants. 7his further strengthens the case for investing in improved quality, even where universal access to basic education has not yet ban fully achmved. Central to cost-effective allocation of resources am strategies that will allow teachers' time to be used more efficiently through b=mmg the pupilhescher rado, operating double shifts in crowded sdkxds, and employing multigrade teaching in sparsely populated arm. Studi of sevend Sabelian countries show that careful implementum of thew strategies can minim their negative effects on pupil achievement. 7bese interventions; reduce teacher costs and permit the savings to be used for quality improvement. Resource Mobffiradon for Bask Mwadox In many countries, existing resources am insufficient to support quality basic education for all and it is necessary to mobilize additional resources. New resources may be obtained from both government and non-government sources. The case for allocation of additional resources to basic education will be strongest where the sector is underfunded because govannient's education spending has been, historically 2 low. Diversifying sources of funding can also be strategically important for obtaining additional government resources for basic education. It may include broadening the tax base, granting taxing powers and delegating financing responsibility to lower-level governments, and earmarking taxes for basic education. For example, during the period of 1950-1977 in China, the government devoted 4.3 to 8.2 percent of total government expenditures to education; this corresponds to less than 2 percent of GNP. The financial reform in the 1980s, characterized by the devolution of financing responsibility to provincial and local governments and the broadening of financing sources, has significantly increased total government spending on education, including basic education. During 1986-91, expenditure on education averaged 12.1 percent of total government spending, or 2.5 percent of GNP. Even without financial reform, resources for basic education can be mobilized without changes in the existing system of taxation by increasing the share of public expenditures allocated to basic education. Actual implementation of this strategy has been documented for the Republic of Yemen during the 1990- 93 period, for the Sahelian countries during the 1980-88 period, and for the state of Uttar Pradesh in India during the last 15 years. Contributions from parents and the community are an important source of additional non-government resources. Parents and the local community have been important contributors to school construction in many countries, including the Sahelian countries, the Republic of Yemen, India, China, and Thailand. These contributions are critical to making rapid progress towards the EFA goals and should be supported, as appropriate, by technical and financial assistance for the design, construction and maintenance of school buildings. In some countries, such as China, the local community even provides the livelihood of some or all of the teachers. But in other countries, such as Brazil, the attitude that public education is exclusively the government's responsibility has been a barrier to community involvement. This clearly makes progress towards the EFA goals more difficult. Community involvement in basic education has benefits that go far beyond the financial contributions. It results often in increased parental appreciation and support for the value of good education, pressure on teachers and students to produce high levels of learning and responsiveness of schools to parental and community involvement. Finally, external agencies can be an important new source for funding for basic education, so long as they do not supplant domestic resources. Bilateral and multilateral assistance has historically favored higher education. Much of the bilateral assistance has been given in kind through supply of highly qualified teachers and scholarships for studies abroad. The multilateral agencies, including the World Bank, have greatly increased their support for basic education in recent years. For example, World Bank lending increased from about USS500 million in 1990 to US$1,000 million in 1993. Further increases are, however, constrained by the limited availability of concessionary aid. For the increased assistance for basic education to be effective, its modalities will need to be reexamined, especially with regard to integration in national basic education programs, effective support for school level improvement, and support for innovation and experimentation. Tareing Resources to Benefit Femals and Marginalized Populatioa Marginalized populations are often low-income people of nondominant ethnic and racial groups, people living in rural and remote areas, aliens and refugees, and people with physical or mental handicaps. Because of cultural, economic, and other factors, females are more likely than males to be marginalized 3 Cae SaAdies i Aade Qmu oM Ed.a in most societies. Marginalized populations are the most likely groups to have no access to quality basic education. By themselves, strategies for resource reallocation and mobilization have not been able to ensure that females and marginalized populations have access to quality basic education. Few countries can afford to fully subsidize the education of the middle and upper classes aWg provide quality basic education to the marginalized groups. Unless resource allocation mechanisms are changed, allocating additional resources to basic education is likely to benefit non-marginalized populations disproportionately. To improve access for females and marginalized populations, targeting more resources specifically at these groups, the key challenge is for policymakers to mobilize the necessary political support for direct and significant interventions with such populations. A country should be concerned about providing quality basic education to marginalized populations for at least three reasons. First, from a moral or equity viewpoint, a society should take care of its most disadvantaged members. Second, from an efficiency viewpoint, investment in quality basic education for these populations can have a high rate of reurn. The investment costs can be substantially exceeded by the gains in productivity and the savings in social welfare costs. Third, given a policy of quality basic education for all, it makes strategic sense for the government to target these populations because they are the least capable of acquiring quality basic education using their own resources. In a country where females are marginalized, their education should be addressed by public policy. But even if females are not regarded as a marginalized group, additional public investment in their education is still highly desirable. There is substantial evidence from a wide range of countries that increased female education is linked to improved health, lower fertility, and other benefits, and that investment in female education has a high social rate of return. In a study of 23 targeted social programs in Latin America, a recent report sponsored by the World Bank concluded that targeted programs are more effective than universal provision in directing benefits at the poor and other marginalized groups. It also finds that the concern about the high administrative costs of targeted programs is overstated. In fact, a wide variety of programs that employed targeting mechanisms have cost very little. An encouraging development is that in the past few years, the countries selected for review in this document have given special attention to targeting basic education at females and marginalized populations. For example, the government of die Republic of Yemen has recently emphasized increased female participation in primary education by constructing additional schools and classrooms specifically for girls and by increasing the number of qualified female primary teachers. The Government of India also recognizes the importance of female education and has borrowed funds from the World Bank to assist the Government of Uttar Pradesh to make schooling and non-formal education more attractive to girls by using a combination of strategies. These strategies include providing latrines and improved facilities and offering early childhood care and education, better teaching, strengthened village education committees, gender sensitivity training for teachers, an increased proportion of female teachers in rural schools, removal of gender bias in curricula and materials, and the expansion of a female-empowerment program throughout the project districts. India and China are two Asian countries that have acted to improve basic education in poor areas. In 1992 India initiated a financing shift away from central assistance to states and toward targeted assistance 4 for elementary education in districts with low literacy and where the demand for elementary education exceeded the available supply. In some counties in China, towns and townships (government units immediately below the county level) are allowed to keep 80 percent of their education levies and surcharges. The remaining 20 percent is sent to the county government to support basic education in the entire county, including subsidies to poor areas within the county. For the first time since its borrowing from IDA a decade ago, the Chinese government is implementing an education project mainly for basic education in poor provinces. Some Latin-American countries also have explicitly addressed equity issues in their current primary education programs. Chile has a primary-education improvement program that runs from 1992 to 1998, with an estimated investment cost of USS243 million. The program supports 5,000 multi-year, school- specific, quality improvement projects. The selection of project schools will alleviate poverty by favoring "high-risk,* mostly poor, urban and rural schools. Explicit selection criteria are specified to minimize political interference in the program. Decentralizaioan Since the 1980s, policymakers in different parts of the world have increasingly recognized that the traditional methods of education finance and management are unable to deliver quality basic education to all children and that radical changes are needed. Central in these changes is the decentralization of basic education, which is the process of transferring decision-making power from higher levels of government to lower levels of government, community organizations, and schools. Available resources will be used more efficiently to the extent that local governments can make decisions that take individual community needs into account. Mobilizing local resources also reduces the financial burden on the central government. Finally, decentralization is a form of democratization. The potential benefits of decentralization can be substantial through the inputs that key education stakeholders can provide to the education process. Decentralization in basic education has taken two common forms in developing countries since the early 1980s. One form is devolution of decision-making, by which submational governments are primarily responsible for providing basic education and have the authority to raise and spend revenues. The decentralization that began in China in 1985 is an example. Based on the principle of "local responsibility and administration by levels,* provinces and autonomous regions are responsible for providing primary and secondary educition. Different levels of education are financed and managed by different levels of government. In rural areas today, primary education is mostly managed by the village government and financed by resources from the village; lower secondary education is mostly managed and funded by the town or township government. In cities, primary education is funded and managed by the district government and lower secondary education is managed and funded by the city government. In conjunction with an ongoing national economic reform, decentralization has resulted in additional resources for basic education and in increased authority for local governments. The fiscal decentralization, however, has been accompanied by widening educational inequality among different areas in China and the conditions of basic-education in poor areas are still very difficult. Another form of decentralization is deconcentration of decision-making, which involves the transfer of authority to lower levels within central or higher-level government agencies. This is often accomplished by expanding the power of regional directorates. Chile's 1981 decentralization reform is an example. In this reform, the central government remains the dominant source of public financing of primary and 5 cA S ås pb..me-~..my 8 5 secondary educadon (about 90 percent of total public education funds in 1990), bot municipal~ties are asslgned new revene sources and new expandimre responsibåiti. Municipal achools recive both enrollmet-base grants (known s vouchers) from the central government and funds from new sources of mnicipal revenue. The reform ansferred school property to municipalitis and moved teaclhwz from c~tral to municipal payrolls. Government gram ca also be provided to private schools which do not charge tdoi. Since governmt gram am tied to school aroloment, the financing scheme provida an meanrve for schools to provide quality educaon and an opportuniy for pupils to choose scaook they pr r. The aim is increased efficency and ufcdveness. A nixed form of decemraliztion has taken place in India. Eduation was devolved to stats governments by the nsti~o of 1950. But in 1992, the Nadonal Policy on Education was revisd to deonceame decision-making withn the stat mo that district governma have more audority to pla and manage elamntary eduin. Furthermore, village dnetion conmina. arm being reactvated to ineras enmmnity involvement in making decisions at the school level. In India, acdve village oduion cnmmiM are seen as the most cost-effective mtod to ensure that teachen work and schools fimction. To be effective, decmuraliation requires simlaneous effor to build and str~gthen aministrative capacity at lower 1evels and to incorporate mur for accountability and monitoring. Am.hown in Brazil, the devoludon of primary dadon to nmnipmiius alons dom not ensure that municipal mchools ara n efncietly ad tmathe distni~mtinn of remurcs is frac from polial interference. Alo, as demonstraed in th refor~s in'Cina and Chile, dm alizion can ruk in wideninginwqualty of resource given to schools in diffamret loc~litn because of disparitiø among ·fme·im. Dcemralizaion has to be linkad to the targeng of addina r.orces from higher levels of governmt at female and marginalized populadons. he "poor provinc project in China, the prinary-odumation improvement project in Chie, and the basic uduetion projeca in Northua Brazil are example of efforts by nional government to promote equity in basic udmuinm More generally, in a dmcunralized finmancing systm higher levels of goveriment should target their subsidia at femalu, narginaHzed populations, and poor a to acheve ffhiple polcy objcves. These objectivs should include moblizing adinal goverment remourcem, increating pupi learning, and promoting educational equy. In sumnary, notable progreis ha ben mede toward qualy biw edneation for all in recet yeari. In additin to increaued access, pupi learning ba becomne a focs of cural policy in many coumsira. Several countrie have alredy moved from ily discning stragia for finaneiny qualty basic Mducation to a y i~pmbdag ch srategim. Their spmrenc ha shown hat sven thgh rsource realocation and moblization atmeges are crdca~ly imporet, such stragi are not anffi~len to addres the finn~ial challenge bot mst be wle m~mn d by straegies for targeting ad deceralza~n. ID SUMMARY OF CASE STMDIES A brief summary of the case studie on the Sahulian co~ntrie, Yem, China, Chile, India, and Brazil is given in the folowing. The key elemntn of the financial intavenina in these countrie s lined in Table 1. 6- The Saliau Cout~u ThIy inciude Burkina Faso, Cape Verdi, Chad, Gambia, Mali, Mauritania, Niger, and Senegal; and are aong the countries in the world today with the hight ra-s of population growth. Basic ab==ainn in the Sahel falls behind most other African counris and low-income countries in other parts of the world. The financial challenge of quality basic education for all Is perhaps the most dnnting for thes countrin. In the 1980s, the Sahelian governments substantially increased the share of primary eduini in the total education budget and allocated a relatively high proportion of the total govern=et budget to education. But rapid population growth and unfavorable macroeconomic conditions put severe limitat~ona on q "ntittive and qualitative improvemen~ in basic cduatin in the near future. Additional reallocation of public resources in favor of primary ed~tin is likely to have negative impacts on other ea n subsetors and no.~euation sectors. Whle further effrts to obtain new resources from thaJocal cnmunmity and from foreign assistance are necessary, it is inportant that current resources are used efficiently. In the pat few years and in s~veral of the Sahulian countries, experimentation harbeen undertaken to improve educa~innl efficiency. Available evidence showa that, with proper planning and implenm 4inn strategics such as double shifting and multigrade teaching can reduce per-pupil.costa without negative impacta on pupil learning. Teachers cosa can bo further controlled by redploying teachers from aministration to classroom duties, and by ranaiing teacher-recruitment requirements to minimiz the hiring of teachers with %excesuivly high qualifications. The cost savings can be realocated to suppor improvementa in acc=s and learning. However, even when the efficiency and resource reallocaion measures are fully implemenued, most Sahlia countries will still have .an~o difficulty in providing quality bmic education to all in the n future. These countries may have to consider more radical changes in the financing of basic education. Ytmen Education in Yemen is also facing a difficult financial stuatinn Even. though ucation has experienced rapid quantaiive expansion in the past two decada, there arm stlllarge and unet educatin and training needs today. The epansin has alms ben achieved at the exponse of quality. At-the same tim, the financial base for education has deteriora~d. Tho country ha gon= through a period of slow growth and large budget deficits. External aid has decreased while the rra of emigrant workera has resulted in a large increase in demand for odu~tion and other services...In:.basic ~incain, there is a need to reasses development priorities and explore alternative mean for supporting suck prioriti:. New governi ent policie since unification in: 1990 have given more nn to pupil learning.al~the education of girs. Current quality-i~provemet ufforts include inservice training of teachers, curricuhun and textbook development, and upgrading of pyical ~aellitla. To ncrease fenale access to:p~imary educaion, additional classroots and schools specifically for girla are being built and tho number of qualified female primary teachers will bo increased. Four strategis are used to finance-these interventions. Firat, in the past two years, the government has inreas education's share intotal government expenditure by 2 percent. Second, t~achors' costa are being reduced by replacing expriate teachers with local teachers. Tlrd, much - of- th comsavings are reallocated to support-quality improvuments. There is an ite in the basic ednbation budget earmarked for qualiy im-provemel. Fourth, there is targeting of remources for girs education. Even though universal primary cduction is unliMy to be attained by the year 2000, continnous improvmen in access and quality are expeced. ?- caw sftkanes RRMItIanr Badse Shooa. China In contrast to Yemen and the Sahelian countries, China has implemented more drastic changes in the financing of basic education. As part of a comprehensive program to achieve the transition from a centrally planned economy to a Osocialist market economy," the Chinese government lanched a decentralization reform of public finance in 1982 and a reform of the structure, administration, and financing of education in 1985. Ihe financial reform of basic education was part of the larger education reform and an extension of the public-finance reform. The basic-education reform had two defining characteristics: the devolution of financing responsibility and authority to local governments and the diversification of revenue sources in support of basic education. It was aimed at mobilizing additional resources for basic education and increasing local decision-making power. By the early 1990s, a more decentralized and diversified system for financing basic education has emerged. And substatially more resources have been mobilized for basic education since the mid-1980s, in terms of both the percentage of GNP and the percentage of government expenditure. The financial gains since 1985, however, have not yet reversed the negative effects of the long-standing low investment situation before 1985. Basic education remains significantly underfunded and its conditions are especially difficult for marginalized populations and areas. Also, large disparities exist among schools and areas, and between government teachers and non-government teachers. These disparities are likely to widen over time if present practices continue. There is an urgent need to mobilize additional resources from higher levels of government to subsidize basic education for marginalized populations and areas. Chue Like China, the financial reform of basic education in Chile was not carried out in isolation; it was part of a larger decentralization reform which began in 1981 and which provided more authorities and responsibilities to municipal governments. Municipal governments could raise additional reveones and provide grants to schools. The aim was to increase local financing of basic education and to increase efficiency in education. But unlike China, the central govrnm t in Chile maintains primary responsibility for funding primary and secondary education. By providing a relatively large attendnce- based central grams to schools, all pupils are assured a relatively high minium level of school services; and municipal grants to schools are for augmentation purposes only. Municipal resources for nunicipal schools have increased rather rapidly throughout the 1980t; they accounted for 10.5 percent of total municipal school revenues in 1991. An important feature of the Chilean reform in basic education is that the amounts of central and municipal grants are tied to school enrollments and are provided to both government schools and subsidized privat schools. These practices are undertaken to promote choice and efficiency in basic education. he enrollment share for snbsidized private schools has grown since the implementation of the reform. IMdIa Compared to China and Chile, financial interventions in basic education in India are more recent. In 1992, the 1986 National Policy on Education was revised. In addition to reaffirming the goal of universal elementary education, the revised policy has initiated a financing shift from broadly based central assistance to the states to targeted assistance to states for supporting elementary education in low-literacy districts and in districts where demand for elementary education exceeds supply. WIthin a state, more planning and management responsibilities are being transferred from the state level to the district level. At the local level, village education committees are being reactivated to increase community participation a in school. Such participation consists of both community contributions to school construction and community involvement in decision-making at the school level. Finally, the revised policy contains special provisions to support the education of girls and minority pupils. The government of India has just approved a basic-education project in the state of Uttar Pradesh, supported by a US$165 million loan from the World Bank. The project incorporates a comprehensive strategy for promoting basic education for females. While the impacts of the above-mentioned interventions are yet to be evaluated, the interventions nevertheless reflect the government's policy shift towards targeting needy populations and decentralized decision-making. rawil The case study on Brazil points out potential pitfalls in decentralization. Historically, administrative responsibility for basic education in Brazil has been very decentralized. State and municipio governments administer rather autonomous primary-school systems with minimal federal oversight. The federal government still maintains a strong presence in basic education through its financial transfers to state and municipio governments for supporting non-salary expenditures. In 1988, a new constitution called for a modified fiscal federalism to strengthen the correspondence between administrative responsibility and revenue-raising authority and to reduce the dependence of state and municipio government on federal transfers. It also stipulated that the federal, state, and municipio governments spend respectively 18 percent, 25 percent, and 25 percent of their revenue on education To support universal access to primary education, all levels of governments are required to spend at least 50 percent of their education expenditure on basic education during the 1988-98 period. So far, there is little evidence that these stipulations and requirements are followed. And in the absence of monitoring and accountabiity measures, it is very difficult to ascertain the extent of compliance. In fact, available evidence indicates that large inequalities in basic education continue to exist across regions and among pupils of different backgrounds, and that many children are still denied access to schools of minimal quality. Such difficulties are related to weaknesses in the current school-finance system, including the lack of effective c oPcement of financial regulations, the complexity and lack of transparency of the system, and the openness of the system to political manipulation at various levels. The federal government recently has stepped up its effort to support basic education in poor regions. The World Bank and other international agencies have directed more of their loans to basic-education projects in Northeast Brazil. The targeting of additional government resources for basic education in poor regions is a positive step, but it must be carefully designed and implemented so that such resources will benefit the needy populations in the targeted regions, and not the best-off members in such regions. Also, targeting has to be accompanied by simultaneous actions to rectify the problems in the current decentralized system. These deficiencies have been recognized and the recently approved World Bank projects in Brazil will begin to address them. 9 Table 1. Sw~nry of Fnancial Reforms in Seral Dewioping Cowhies Cowry Key Inmw ions Policy Objec~iws Rulu and LMrsons Brazil 1988 consiufln initimed Resource mobili=tion, Acmal spending levels not eusily new fiscal feder~iam, red~ce reliance on ascmrtained, lack of complinnce stipulaed spending le . cemral transfer, wichconsitionareqirmmre, increase linkage large in.qualities persist, need beeween ardministrative nimetanspacy, aonnnbnhHIy, & financing power. &nn*~ring in flnncing synam. Chile Inuasing municipio More local decision- Municipio financing at 10.5% of financing. making. school revenne in 1991, cetral trfe maindomin source to ens relaivelyhigh in*um* leved of education servicu. Enrollmmmt-based gra for Promoting choice & Incrnasing enrollment share for both government & subvened efficiency. subv~nd privMe schoolk. pr~ school. China Devoluton of financing Resource mobil~ation M~rermmourceforbasicaetin, resposibity to local &äa part of public- dwnralizdsystemcauahihed, government, diversification finance refim widening inmqual,~nodmorm of revnu~ sources. targeda*g of governme~ r=urc at poor populations & arm. India Shift from broadly-based Targeting govermnet Results to be evaluMed. cenmral maitmn= to reourc at neded tagd aistne. ds . Moreclsim-muaking t Imcreasing efficiency. Need capacy building adistrict district & commny level. leve. Rasula to be evan~d. Specific policis supporting Promoting femule & New basic edeion project in girs & min~ntie. minority eanin Uttar Pradsh with focus on fenalfs. Sahelian Increauing baic cdu~atin's Ramource mobfiktn Amep share in~ed from 39% countrMies sar in education budget. in 1980 to 46% in 1988. Experimenting with Increaing intrnal MInlif negaiv. effecis on mudtigrad teaching and Cfflancy. learing with proper ifM - double shiftng. tdon. Can ralocM savings to support access & leamning. Yemen Increasing education's share. Resou mobliztion. Share inc from 17% to 19% during 1990-93 period. Replacing exprime teachers Cost rekin. Lng-term savings in recurret with local teach~s. cm about 8%. More classroomVachools for Promodng girls 600 classrooms to be built in giris, more female teachr. ed=in. 200 reral communities; 300 female taer. to be trained. Budget hem for quality. Quality improvemtn 1% of total primary education expnditure by year 2000. lo Colclough, C., and Lewin, K. (1993). Educating all the children. Oxford, UK: Clarendon Press. Grosh, M. (1992). From DIatitudes to oractice: Tarreting social rerams in Latin America. Volume 1. Washington, DC: Latin America and the Caribbean Technical Department, Regional Studies Program, Report No. 21. Harbison, R., and Hamishek, E. (1992). Educational Derformance of the Door: Lessons from rural Northeast Brazil. Oxford University Press. Inter-Agency Commission. (1990). World declaration on education for all and framework for action to meet basic learnine needs. New York. Lockheed, M., and Verspoor, A. (1991). IMDrVing Drimary education in develoniny countries. New York Oxford University Press. Schultz, T. P. (1993). Investments in the schooling and health of women and men: Quantities and returns. Paper presented at the International Symposium on Economic of Mucation, held in Manchester, United Kingdom, on May 19-21, 1993. Tsang, M. (1993). Cost analysis of educational incluion of m a poonit . Paris: IEP. Winkler, D. (1988). . on in education: An ec c ps t. Washington, DC: Public Sector Management Division/LAC, the World Bank. World Bank (1990). Pia educ . Washington, DC. World Bank (1991). StaffAgrisal Reot Chile Pia Education Imnvement Proie. Washington, DC. it Case Studies in Financing Quality Basic Financing Quality Basic Education in the Sahel The 1990 *World Conference on Education for All' defined ambitious targets, both regarding universal access to basic education, as well as quality requirements related to learning acquisitions. The conference called for an "expanded vision" that asurpasses present resource levels, institutional structures, curricula, and conventional delivery system while building on the best in current practices" (UNICEF, 1990) and Schooling for All, defined as 'the cirrmstances of having a school system in which all eligible children are enrolled in schools of at least minimally acceptable quality* (Colclough and Lewin, 1993). These general challenges are highly relevant to the Sahel. Sahelian countries, although rich in cultural and ethnic diversity, are economically poor. GNP per capita averages about US$425, ranging from about US$200 in Chad to US$750 in Cape Verde (World Bank, 1993b). According to UNDP's Hwuan Dewloprnea Index (UNDP, 1992), six out of the eight Sahelian countries are among the twelve lat advanced countries in the world. Although population density is comparatively low (15 inhabitants per km compared to 44 and 83 respectively of francophone and anglophone African countries), the Sahelian countries are characterized by climatic and economic conditions that create severe constraints on the countries' ability to rapidly increase the coverage of quality basic education. In order to put the education situation of the Sahelian countries into a comparative picture, four other groups of low income countries have been established as a basis for comparison: (a) francophone African countries other than the Sahel, (b) anglophone African countries, (c) low-income Asian countries (12 countries with per capita GNP below USS800), and (d) low-income Latin American countries (7 countries with per capita GNP below US$1,000). The figures are derived mostly from UNESCO and from national sector studies and refer in general to 1938 and 1989. Governments' revealed preferences for education can be studied in retrospect from actual spending on education. Traditional indicators in this respect are (a) the share of education in GDP which indicates the effort relative to the national product, and (b) the share of education within the Government's budget which indicates the relative weight given to the sector in comparison with other claims on public financing. The financial effort in favor of education is, on average, higher in the Sahel than in other world regions. In 1988, Sahelian governments allocated 19.4 percent of their budgets to education which is markedly higher than other low-income countries. The share of GDP per capita spent on education was 3.7 percent of GDP, compared to the average for other African countries of 4.4 percent, for low- income Asian countries 3.4 percent and Latin American countries 2.3 percent (see Annex 1, Table 1). However, great differences also exist among the Sahelian countries, ranging from Chad (1.5 percent) and Niger (2.5 percent) to Mauritania (4.9 percent). Given this comparatively high share and the many other M2 FhmendMg 0ay Basic EdAmsin a Suh t competing demands on their budgets, it is difficult to envisage, in the present economic context, that the Sahelian countries would be able to accelerate the development of primary education by spending a significantly greater share of their budgets on education. Structural equity' in the distribution of public education resources is much lower in the Sahel (Gini coefficient of 0.78 in 1988) than in the other countries in the comparison (see Annex 1, Table 2). A more direct way of presenting this result is to remark that, if the present pattern of resource allocation were to be maintained, more than 50 percent of public resources allocated to educating a generation of children would be appropriated by 5 percent of the age group. Inequities are most pronounced in Mali, Burkina Faso and Niger. Since 1980, there is a clear tendency towards less unequal distribution of public resources for education as a result of relative progress in terms of primary education coverage, but marked differences still exist between the Sahelian countries. In Mali, the distribution became even more unequal during the period 1980-88. Education coverage in Sahelian countries lags seriously behind other countries in Africa (particularly Mali, Niger and Burkina Faso) and even more seriously behind low-income countries in other parts of the world (see Annex 1, Table 3). The primary enrollment ratio in the Sahel averages only 35 percent for girls and 56 percent for boys, compared with the Sub-Saharan averages of 72 percent for girls and 82 percent for boys. Six of the countries in the region are among the only 17 countries in the world where primary school enrollment accounts for less than 60 percent of the school-age population. GrWh 1. Tads in Enrollment Ratios for Primary Edction 20 - 101 o *W Source: Band on data from UNESCO Swstical Yewbook, 1992 and mssu deat collected by APSPR. SUdIaics 0 IZuctual equky ak11 isle acOut data os: (a) dh distriuoion of the trinal level of education (proportions of a generation of chbldia who oeplts the dkffmu levels of an ofsiotal ssm); and (b) the public renources accumulated along the schooliag cae (er c yea of schooling pilies a privae apropmidin of public resources for an mount coepondingto the alk cost a differelevels of educastion). The distribudom of the legth of studies reius in a distriburion of public soores appoprad by diffMe aNabrs of a genraiun of chldren. This distribution (and concentration) of resernes ca be yahosed in the Gilni coefficient. 13 Cas ad ls Fbnmae Qamay Bac EdAss Progress has been made over the last decade. In the Sahel, rates increased at all levels of education (from 32 percent to 41 percent in primary education, from 7.4 percent to 9.7 percent in secondary, and from 0.9 percent to 1.5 percent in higher education). However, the Sahelian countries have developed differently. As Graph I shows, Chad, Senegal, and Burkina Faso increased their primary enrollment ratios steadily over the last years, while Mali is generally in a worse situation now than 15 years ago with a primary enrollment rate around 25 percent at the end of the period. By and large, demographic and macroeconomic constraints put severe limitations on improvents in enrollment rates in a short- and medium-term perspective. Assuming a scenario of no change in present modes of financing and organization, countries will find it difficult to match the rate of enrollment growth with that of population growth. Simulations based on country-specific data for five Sabelian countries suggest that in such a "no-change" scenario all of them would see enrollment rates decline. The negative consequences on primary enrollment rates vary in magnitude from one country to another and reflect, at least implicitly, current priorities. In Burkina Faso and Niger, the next decade may well compromise the progress accomplished during the 1980s. In conclusion, it is imperative for Sahelian countries to develop and implement an education development strategy that avoids stagnation at a very low level of basic education development. This case study examines some options for such a strategy. I. RESOURCE MOBILIZATION FOR BASIC EDUCATION Sahelian countries' intention to develop primary education is reflected in fairly ambitious medium-term policies and investment programs. Resources are being mobilized in three ways: (a) making more efficient (pedagogical/economical) use of existing resources; (b) increasing the share of governments' resources allocated to the education sector and to primary edimcation in particular; and (c) requiring increased contributions from parents and communities. The first type of action should be easier to implement than the second since the scope is limited to primary education itself. The second type requires a global strategy for the whole sector and reallocation of resources from secondary and, in particular, higher education, to primary education. This type of reallocation has proven to be politically very difficult in Sahelian countries. Any major educational reform has important short- and long-term effects on large segments of the population as well as on the nation's social and economic development process. Consequently, such reforms entail Government decisions that are politically highly sensitive, and their implementation needs to be considered and planned in the context of the political, economic and social interests affected. Reforms in this am normally-explicitly or implicitly-change the distribution of education costs, benefits and privileges among different population groups. As modern sector employment is becoming increasingly scarce and dependent upon education qualifications, benefits derived from public spending on education are becoming an even more important determinant of the distribution of influence and wealth in the society. Conflicts of interest between different regions, political parties or ethnic groups can also surface and be sharpened. Many educational reforms have a direct impact on the working conditions and living standards of teachers, who frequently constitute the largest single group within the civil service and have the strongest and most vocal union. Others affect the life of students who, especially at the post-secondary level, constitute a political momentum that may make or break governments in many developing countries. Consequently, given the highly political sphere encompassing educational policy decisions, strategies for 14 FW.d.R. Of..suy Badwc Eduamn bo ON SaIW consensus-building within a country are essential, in particular when policy reforms require changes in central and local behavior, or influence the distribution of education costs and benefits among different groups (Fredriksen, 1990). As this document shows, this is most often the case in Sahelian countries. Mobilizing more resources may also be difficult in the present fiscal and economic context. However, indications are that communities and parents would be willing to contribute more than at present to ensure basic education of their children. But this may require that they be more directly involved in controlling the functioning of the schools. Based on parents' demand for education for their children, decentralization may mobilize additional resources (including far more than economic resources) in a non- bureaucratic way, contribute to internal efficiency through transparency and direct involvement (e.g., by parent-teacher associations), and contribute to quality improvement by making the teachers more accountable vis-4-vis parents. Decentralization does not imply abandonment of government influence. The aim is rather to limit the detailed, central, bureaucratic control and to redefine the role of the district and regional administrative staff to become more resource persons for local school improvement. A side effect is that some evidence suggests that teachers' salaries may tend to drop as a consequence of increased local autonomy and accountability. When discussing reforms, adjustment and reallocation of resources within the education sector, it is essential to keep in mind that the objective is not to cut spending per se, but to seek an optimum balance between spending on the one hand and increasing coverage and quality in terms of student learning on the other. The objective is to move as far as possible in the direction of obtaining better quality universal primary education per dollar spent. More FTcient Reource Use Unit Cost Expressed in per capita terms, unit costs at all levels of education in the Sahel (see Annex 1, Table 4) exceed those of other African countries by 60-80 percent. During the 1980s, unit costs dropped in Sahelian countries but not as much as in other African countries. Great differences are found between countries. Unit costs are highest in Niger (especially in secondary education), Burkina Faso (in secondary and particularly higher education) and Mali (particularly in primary education). Unit cost of primary education in the Sahel is almost twice that of other African and low-income Asian countries. At secondary and higher education levels, unit costs are 44 percent higher than in other African countries and 3.5 times higher than in low-income Asian countries. Differences in the pupil:teacher ratio do not explain the difference in unit costs. Although pupil:teacher ratios range from 67:1 in Chad to 38:1 in Mauritania, on average they are higher in the Sahel than in other world regions. Teacher salaries, however, contribute largely to the differences in unit costs. At the primary level, teachers' average salaries amount to 10 times the per capita GDP compared to less than six and five in other francophone and anglophone SSA countries, respectively, and less than three times the per capita GDP in low-income Asian countries. A similar picture exists at the secondary level. At the secondary level, Sahelian countries spend more, on average, for scholarships and student aid (about 20 percent of secondary budget, see Annex 1, Table 5) or 2.5 times the share in SSA anglophone Africa and about 3.5 times that of low-income Asian countries. In francophone countries the gross enrollment ratio in secondary education is 23.9 percent and the share of the budget for student aid is 25 percent. Anglophone countries have a 22.8 percent enrollment ratio and the share of student aid is not higher than 7.5 percent. Secondary enrollment in the group of Asian countries used in this study is 34.4 is came ikod" i Puboine NUR amic Edwadon percent and the share for student aid in the secondary education budget is only 5 percent. Student aid is generous in the Sahel when the 10 percent of the age group who are enrolled in secondary education benefit from subsidies amounting to almost 20 percent of the budget. In higher education the picture is even more dramatic, at least when compared to the Asian countries. Sahelian countries spend a share of their budgets for student aid in higher education that is eight times that of the share in low-income Asian countries. The case of Burkina Faso is given in Box 1. Teachers Teachers' salaries constitute the principal budget component at the primary level. Therefore, more efficient use of teachers constitutes the core item of any policy aiming at reducing unit costs in primary education. Amroriate Level of Teachers' Oualifications It is to be expected that the more comprehensive the pre-service teacher training, the higher the quality of instruction and the capacity of teachers to adapt adequate pedagogical methods and adjust to curriculum changes. Many African countries have extended the length of pre-service training for primary school teachers over the last 15 years, in part because of an increased rate of graduate unployment. However, teachers' salaries are directly related to the level of the diplomas. Because research provides no strong link between student learning and the teachers' qualifications (Jarousse and Mingat 1991a, 1991b, 1991c, 1993), Sahelian countries have begun to recruit a larger share of iutisew a4oLnt. rather than btiturears to lower costs (as the 4:1 ratio between the two categories of teachers defined in Chad, Niger and Senegal). Given the rigidity of the civil service salary scale in Sahelian countries (linking wages to educational levels), using teachers who, prior to their training as teachers, have completed upper, rather than lower secondary education increases unit costs in primary education by 25-35 percent without any significant gain in student learning. A study in Mali (Fomba, 1992) concluded that primary school teachers performed significantly better (judged from student learning) when financed directly by the parents compared to the normal public mode of payment. Furthermore, teachers' salaries in these sprivaste schools were 40 percent lower than in Government schools. Similar differences in teacher salaries between public and community managed schools were noted in Chad (World Bank, 19934). This evidence could provide a basis for some suggestions: (a) rigidity in teachers' salaries in the public sector could be reduced by taking into account the conditions of the local labor market, thereby extending coverage of primary education; and (b) making teachers accountable to local users appears to boost motivation and efficiency. Some countries have also made teaching in primary schools an option through which to fulfill their national service obligation. In Burkina Faso, it was decided in the late 1980s to assign about 1,100 members of the National Civil Service to teach for one year. However, in all efforts to increase cost efficiency, a balance has to be found, taking into account teachers' accountability, professional qualifications and employment mode in order to ensure adequate quality teaching. 3 In Dawkima Fm k vas decmded is ts early 1900s es bsh a aw slory (ler deed as busarw e446) of primy school Uhmers based a I yer aer OM anng progpa es top of lower uecsindery. L, the duram of the aseer tMiming ptnrm has b meaded to two years. 16 FbMacmug awdUy Basc Edaan in es SAa Box 1. Scholrhips in Burkina In 1982, the total cost of fellowships in Burkina Faso to students at the secondary and higher educational levels amounted to 35 percent of the education budget, an amount higher than the total budgetary allocation for primary education. In 1988-89 new rules were adopted in order to limit the eligibility and thereby reduce the rate of increase. Fellowships were only to be awarded to first year students below the age of 23 with high performance at the baccalaurft, and aid given only to those who satisfied criteria related to academic merits and the economic status of parents. The number of new fellowships dropped accordingly, from 1,000 to less than 800. In the SY 1992-93, about 3,400 earlier fellowship holders received reduced amounts compared to what they received before. About 1,100 new fellowships were awarded (FCFA25,000 per month) in accordance with the new criteria. About 750 students with performance below average the previous year were awarded reduced fellowships (FCFA12,500 per month). The amount given to all fellowship holders for books and equipment is FCFA50,000 the first year and an anmal amount ranging between FCFA12,500 the first four years and FCFA50,000 for the 5th and later years. For the SY 1992-93, a special grant (FCFA100,000) was given to 523 first year students who were slightly older than the maximum age of 23 years, and prices on campus were reduced for fellowship holders. Consequently, 60 percent of the students at the university received support. There were 1,234 fellowship holders studying abroad (40 percent in other African countries, 35 percent in former USSR and 15 percent in France). Compared to 1982, the share of exponses for fellowships in post-primary education had dropped to 26 percent of the total budget for education and represented only 50 percent of the total budget for primary education. Within the budget for higher education, however, a large share (67 percent in 1993) is still allocated to fellowships. Soare: PpulatiM and Ha Remuma Operaions Divilso, Sabula Depaumm, Afri ReioM Redeployment of Teachers from Administration to the Classroom The centralized school systems of the Sahelian countries are expensive to operate and career patterns of teachers draw senior teachers away from teaching and into a rather voluminous administrative system that is neither transparent nor easy to manage.3 Efforts to redeploy teachers back to the classrooms have been quite successful in Mauritania and Senegal, and have permitted a reduction of the burden on the salary budget in a period of expansion. However, indications are that there is considerable scope for further gains in this area in many Sahelian countries. Increase Papil:Teacher Ratio It is commonly believed that pupil learning is better in small classes. However, given that unit cost is crucially dependent on class size, it is evident that the class size a given country can afford is closely related to economic conditions. But even for countries at similar levels of economic development, the Recen figum from Mali indicaa that 3,000 out of ab" 17,000 primary achool seec in the county are rniped to other fons. 17 caN &a&" i PWWRC*W 0"asy BeAdc EAMMn optimum class size is a matter of considerable controversy. Recent empirical studies in Benin, Burkina Faso and Togo (Jarousse and Mingat, I .-?) suggest that student learning deteriorates when class size increases; although it should be kept in mnd that a number of factors determine student learning. When, for example, a teacher lectures, class size is of little importance. Furthermore, large class size may be more detrimental to quality in certain grades. For example, when, the class size is high in grade 1 (as is often the case), the teacher will not be able to meet the individual needs of pupils for coaching, guidance and evaluation to the extent required, and learning may suffer more than in higher grades. Instead of a general increase in class size, a higher pupilteacher ratio can be achieved by better organizing rural schools better (e.g., by multigrade teaching), or by using double-shift teaching, mainly in urban schools. Use of these two techniques is discussed below. Double-shATeahin Although widely used in the rest of the world, arrangements for student grouping, especially as regards use of double-shift teaching, are objects of controversy in some Sahelian countries. It seems clear that this is an area where these countries can learn from the experience of other countries and where policies need to be more firmly based on research and current national economic realities. In urban settings worldwide where enrollment is high and lack of budgetary resources would otherwise result in very large classes and/or rejection of children, double-shift teaching has been introduced at different times. The basic idea is to utilize classrooms and teachers for a longer period of the day, for example by having one group of pupils attend in the morning and another in the afternoon. The time the pupils spend in school has to be reduced (for example from 30 to 20 hours a week, like in Senegal) and increased for the teacher (2 times 20 = 40 hours in the preceding example). Teachers are compensated for the extra work (in the Sahel, the bonus is generally 20-25 percent of average teacher salary). The rationale for double shifting is mainly related to the reduction of unit costs. In an experiment conducted in Niger, the per pupil recurree cost was only 60 percent compared to single-shift. The application of double-shi teaching in Dakar permitted-for the first time-the city of Dakar to accept all children who wanted to enroll in grade 1. As regards resource savings, introduction of double-shift teaching in Senegal permitted enrollment to increase by 15 percent without increasing the number of teachers. It is also interesting to note that the only country in the Sahel that has attained universal primary enrollment (Cape Verde) has even used triple-shift teaching. The impact of double-shift teaching on student achievements has been studied in many countries of the world. Several studies have also been conduCted-or are currently being undertaken-as part of IDA- financed projects in the Sahel. Studies conducted for two countries-Senegal and Niger-have given somewhat contradictory results. In Senegal, student achievement was almost at the level of pupils enrolled in single-shift classes, while in Niger, the achievement of double-shift pupils was below that of single-shift pupils. Observers familiar with the two cases strongly believe that the different outcomes are largely due to the way double-shift teahing was prepared and introduced in the two countries. It is not difficult to understand that double shifting may be rejected by parents who already have their childre enrolled in school because it is considered to be an inferior solution for their children, and by the teachers because 18 Fbncir Quulsy Basc Edodengm de ake the compensation they get may be considered to be too low to fully compensate for the extra work load. But it is also evident that the formula can be accepted and supported when properly introduced. In Senegal, double shifting was introduced with much enthusiasm among the teachers and substantial popular support (see Box 2). In Niger, the teachers' union (possibly also many inspectors) opposed double shifting. Faced with stiff opposition from the teachers' union, the Government decided to stop the experimental program. From experiences in the Sahel so far, it is evident that the way double shifting is introduced is essential. First, the quality of double-shift teaching is largely dependent upon training of teachers in the use of this method of teaching, that programs are revised to reflect the shorter duration, and that sufficient training materials are made available. Second, thorough preparation is required on the part of the Government as regards mobilization of public and professional support. The formula must be appreciated by parents and teachers as an important measure to reach crucial national objectives. It is especially evident that the views of parents of pupils who would be able to enter school thanks to double shifting must be taken into account. This is generally not the case. If properly prepared, double-shift teaching is likely to be a most effective way to increase internal efficiency without sacrificing educational quality. Even in cases where double shifting would cause student achievement to decline to some extent, this negative impact can be counterbalanced by such measures as reducing class-size and by providing more training materials. After all, priorities have to be made in the national context, and any negative effects of double-shift teaching would need to be weighed against the positive impact on increased access. Multiaade Teaching Multigrade teaching is a familiar method used in many countries in rural areas where, because of large catchment area due to low population density and/or low participation raze, the number of pupils in any single grade is low. Under such circumstances several options are possible: (a) each school may cover an even larger catchment area, resulting in some children having to walk a longer distance from their homes to school; (b) admission may be less frequent than every year; or (c) introduction of a multigrade system. A large catchment area has several inconveniences: the walking distance is not only strenuous and time-consuming, it also creates difficulties with food during the day and has been shown to have a strong negative impact on the admission of girls. Admitting pupils less frequently than every year has other disadvantages, e.g., it leads to a wider age-range in each grade. However, this is an approach that could be more frequently used, for example, admisin could be every second year. Nevertheless, of the three options, multigrade teaching is the most promising and widely used throughout the world. Multigrade teaching has to accommodate the needs of pupils at different grade levels in the same class. The teacher is therefore more or less forced to conduct teaching in a different way than talk-and-chalk, and, instead, organize the students for independent work with a larger share of the time devoted to individual/group tutoring than to talking in front of the whole class. Some argue, however, that teaching Howme, it should be aoed that Pn Niger,t sinalay the Teacher Uaion which oppoed the double shifting. Goveramt offacials ad eachers involved re la favor, the aMer group uialY becne of the e-r py, conasidering the dficulty in fiNdig sarasive mloyM e. In other coMriN (e.g., Mall), k aps ther saeemr an find alternative and aorn lucrative enpoymew by uoia ppils outide schooour sad thsioream les nered i working exs hours on taching another 19 cds~.a a ~ ~ Fbe onese £. d u in rura ara is difficult enough without the exta burden of having to teach several groups at the same dime. Bar 2. ~uc Adj~ unen inSenegal Depte Senegal's hrdiat= al rok a a camw for e ~a- and trainng in Fr-nch-perking West Afica, at indepmde the devepM~ of priry duc~aaa (27 pe~vet pss pinmry anoIn~ ra im 1960) was wel below the Sub-Sahaan Africa (SSA) average of 38 p t. Duing the fir two dds of ndseåda'. Sengal de major stides in develping priy aam despie week scon= growth (imreas the share of GDP devted to ducution from around 1 pe=e to around 5 phrent). Still, Smegal mnn d the 1980s with a pr~wy wro~t uatio of 46 peet, compr~d to the regioal average of 73 pemt. Stating in 1981, a now amha policy was developed and put imto ~ation in 1985: the key objctive beiog 30 purnt mnro~lnt for the age group 3-6 yes and 100 peret enroln for the ag group 7-16 yar by the yoa 2000. Although primry school anrolmn~ grw at an ~verage an mn ate of 6.8 perent from 1980 to 1985, and the amo~i ra inased frm 46.0 pr= to 55.5 p~04n, the developsnam was still not mfficient to rch the ~aget of aniversal pimy educmtian by the ymar 2000. Notwithtnding the Governme~t's ability to mms the ara~ of the *dumän budget dsvoted to primy @dmam~ frn abom t 40 permt in 1980 to 46 pr~ in 1985, c d anada t of expmin~e in higher eamann wa neah ~ry in ord r to a "aaoda e furthr Doain e ig foer of pI~ry Genem Quality slipped n primry *å~cati- the uc rate at the prinary school nvig exs: -me daelined from 59 percen in 1966 to 42 p=r=n in 1983 due to poor meaing efficincy, overy manaic canmet, and inadequae maname ad co~tol at the regional lvaL la addition to finacag invments to ipove quality =n access, to hep the Gover~nmt muilis era remoues for puinry educatiam, the Pinmmy Edcation Develpa~t Project (spported by the Wodd Bank and the African Developt B ) that buame aiive in 1987 was, im adinäm to frnlauiny invemes to i ove quality ma~ndamed ~, designed to hep the Gove t mobilin e ts emos for prin ry acmann by uch aus ra: (i) reducing at ct by introducing doubleshift teaching inrban areas ad m~igrade teachimg i ~ rul arn, and (5) cmmiing expendm im post-Primy edmcan. As regurds the cost-ving , the muanie n rkm the Semugal experi~ce fer 1987 arm: 1. Double-s~ift teaching wa msaef It was und im aut 1000 clasmeoms during the SY 89-90 and an ev cnhm~am cueinad that pupil achievet was mo low than im regular-hift clamses. 2. Mulhigrude aching was also --o~AuDy iepka~ m 221 clasom. The quality a bects have not ben eval 3. It proved to be far more difficat to mnain or bring dow the vel of p«ding in po-primary e& ~.mam Stu entbeidi~s which a~omumed for 36.5 percet of the high r educatioM budget in 1985, had inmraed harply in spike of Goverments efforts. The Snegal experence «hibits the difficsim of budget adtm i highb~ eo~catia= im a framem" Sain country where mamam systcdu cea gcnuamy ~ m l, faiudy cosy, and ammially elitit, with primary eacaun~ catag to meinly urn - and posmury meotin. banefitting the urban elit. S r Fredrn:um, 1992 Fbmi QuSay Basic Edeismbes he Sabol In order to assess the benefits and drawbacks of multigrade teaching, studies from Senegal, Burkina Faso and Togo show that multigrade (most often double-grade) teaching is pedagogically beneficial for student learning.- Allowing for the possibility that existing knowledge as regards effective use of muligrade teaching methods has not been fully used in the caes that have been studied (see for example Thomas and Shaw, 1992), the results are encouraging. A learning gain, accompanied by 50 percent lower unit cass, is obviously important and so encouraging that it might even be worth considering a more flexible grouping of pupils in schools where multigrade teaching is not a necessity because of small class size. Training of teachers in use of this technique and provision of additional materials considered essential is required, both for the sake of maintaining quality and to convince parents that use of muligrade teaching does not impair quality. Sdkool Buildings and Furniat" School buildings in the region have been mostly made from local materials, in particular in banco.' The schools financed by the Government, often through donor support, are made of concrete. An evaluation of projects supported by the World Bank with regard to construction components (World Bank, 1993c) indicates that substantial efficiency-gains can be made by improving planning and management. For example, under an IDA-financed education project in Niger in 1991, the average cost per classroom made of concrete was about 50 percent lower (in nominal terms) than similar classrooms constructed in the mid 1980s. T1e key factors were the use of standard designs and especially private, contracting procedures. Existing studies on pupil learning in the region clearly indicate that there is no relationship between building materials and student achievement. In fact, classrooms constructed in local materials offer some advantages for multigrade teaching. The poles often found in traditional classrooms serve as natural separations between groups of pupils and are not disturbing at all, as they are for single-grade teaching. Schools made from concrete can be justified on other grounds than those related to student learning (e.g., lower maintenanc casts). However, given the budgetary constraints and the obvious necessity to increase coverage of primary education without sacrificing quality, the transfer of the responsibility of school building construction to local communies is an option that should be seriously considered (both by countries and donors). For this approach to work well, local communities would need both technical support and supply of some key building elements. Not only could self-help schemes reduce budgetary costs to a minimum, they could also ensure that schools are constructed where there is a real effective demand and avoid the classical situation in which local comm ities wait for the Government to build the school as it did in the neighboring village. No evaluations of the relationship between school furniture and learning are available in the Sahel, but it is obviously a precondition for learning that at least some minimum requirements are met. When children are requested to bring their own tables and chairs, and even this equipment is stolen, irritation turns into despair in families who sacrifice other necessities for the education of their children. Cost- S wa Medif WlY Verads 1-2 ad 56 w aMdIed. T1e MeuI seW gtb, aU other coandiions being equal, pupils is double- grade clases with 40 chldra gin sessa which wege peor ewn schieved in single-rads clserooms of 20 pupils. Tbs dithereas in grades 1-2 wia, on average, six poima, and in grades 56 fou points o a scal, with a 15-poikt standard deavieson (Jargags ad Miag, 1992). 1 Ba w isn faess argil and eiher ed as bricks or apWHd disu@ly as building measfw. 21 Ca "hadi FbM9cV OMOmAy BoW Edms effectiveness is vital in this are as well; properly conducted competitive bidding and adequate quality control can keep expenses down. Estimated Results from Imprving Cosr Efficiency The factual information available on the impact of school factors on student learning needs to be more fully used in the design of more cost-effective approaches to help solve the very serious crises in coverage and quality of basic education faced by Sahelian countries. To illustrate the scope for cost savings, simulations have been performed to estimate the implications for year 2000 of introducing the following measures: - a pupil:teacher ratio of 55:1, - only recruitment of bntaews adioints, - teacher salaries slip by 1.5 percent per annum in real terms. Table . Inpa of Measurer in Primary Education on die Gross Primary Enrollment rare n the Year Burkina Mali Mauritania Niger Senegal Actual Enrollment Ratios, 198 34- 23 51 30 59 Project Enrollment Ratios, Year 2000: 1. With no policy change -9 4 -6 -12 -5 2. With Single Measure: Pupil:Teacher Ratio - 55 - +9 +4 +5 +6 Recruitment of Insitume Adjoainr - + 1 - - +3 Teacher Salary Drop (1.5 percent p.a.) +5 +6 +11 +5 +14 3. Total of 3 Measures: With Current Share of GDP for Education -3 +14 +10 -1 +14 With 3.5 percent of GDP for Education +13 +19 +10 +25 +14 Nr: Numbers for eac meas aow nwmem paesagpoimt ch sn mlm ms uing oimS de dinureamselisd. The simulations indicate that, except for Burkina Faso and Niger, the countries would be able to increase the primary enrollment ratio by applying the three measures considered. For the other two countries, a real improvement in the coverage of primary education is dependent upon the mobilization of more resources. However, to reach the objective of universal primary education would require mobilizing additional resources for all these five countries. ' Msiag da pnee aimulaes a be mais fr Chd. 22 F6ieiN Queasy Betc Edwim in e S&Wul Reallocation of Eds ion Resources in Favor of Prfaary Educaton The capacity of labor markets in the Sahel to absorb the growing number of school-leavers is more and more limited, and open unemployment among graduates from secondary and higher education has become a major social concern and represents a destabilizing political factor in some countries. Some regulation of student flows can probably not be avoided. Students continue to seek admission to higher education in spite of bleak prospects for work in order to benefit from the strong public financial support provided to a majority of them. Efforts to control admission to upper secondary and, especially, higher education would imply limiting the provision of scholarships and/or adopting some cost-recovery formula. On the basis of simulations made for the Sahelian countries as a whole in 1989, estimated gains of eight to 10 percentage points in primary education coverage by the year 2000 would be the result of a freeze of scholarships in higher education at their 1989 nominal level, provided that all the savings were transferred to primary education (Jarousse and Mingat, 1992). The enrollment ratio in secondary education is very low in the Sahel-on average only about 10 percent as compared to the Sub-Saharan average of 21 percent. Thus, there is little justification for severely limiting enrollment growth at this level in order to transfer resources to primary education. This being said, in the future, a larger share of the enrollment growth at this level should be in privately financed schools, or financed through increased levels of cost-recovery in public schools. Based on the above, two conclusions can be drawn: (i) an excessively high share of post-primary resources is devoted to student subsidies; and (ii) it would be in the national interest to reallocate a large share of these resources for: (a) pedagogical and research purposes within higher education; and (b) to primary education. However, provisions of Government scholarships is a highly sensitive issue in most Sahelian countries, and experience shows how difficult it is, in practice, to have sufficient space to maneuver. With the exception of a couple of countries where the scholarship budget is extremely high and needs to be reduced in absolute terms, reduction can probably only be effected in relative terms and made gradually over time. This means that the share of higher education in the total education budget would decline over time because the higher education budget would grow less rapidly than the budget for primary education. It should also be kept in mind that the high expenditures on student subsidies in higher education have impeded budget allocations for development of the sector itself, and that increased funds for quality improvement are sorely needed. A strong private involvement is normally actively encouraged in the post-primary educational levels, partly to secure relevance of curriculum and partnership in practical training. However, with the exception of Burkina Faso, no Sahelian country has a strong private sector in secondary education. Moreover, even in Burkina Faso, private funding of secondary schools and higher education institutions has dropped over the years, and, thereby, more financial demands are put on the Government. Consequently, given the juxtaposition of strong social demand for post-primary education and severe budgetary constraints, promoting stronger private involvement in the financing of post-primary education is important, even for primary education, as it would offer more scope for the Government to focus its financing on basic education. 23 caS.es an wancilg Quiy Basie Edmedon Mob&king Ouside Resourcs Self-help and Cos-recowy Most Sahelian countries practice some form of self-help and cost-recovery schemes. In the successful cases of building classrooms by community participation, the ownership and concern among villagers is strengthened and the cooperation between teachers and parents is improved. However, such schemes often face quite a number of practical problems. The necessary planning capacity at various administrative levels does not always exist, and communication channels that enable the various parties to share information and coordinate actions are not always functioning. Self-help schemes are often more costly and the quality is often inferior due to lack of supervision, poor construction and inferior materials. Cost-recovery schemes that have been developed and put in place in the field of education are most often related to textbooks. Revolving funds could relieve some of the stress on the public budgets, when properly managed, accounted and audited. However, increased economic burdens placed upon parents might contradict other efforts to achieve universal primary education. A recent study on community participation in the health sector in Burkina (Ciardi, 1993) points to the problematic and troublesome nature of the mechanisms and structures that have been devised. In general, community participation is neither conceptually clear nor easily defined in practical terms. Therefore, schemes to raise resources through school fees, cost recovery on textbooks and provision of free labor (even materials and equipment) for classroom construction must be carefuly designed to avoid hitting the weakest groups hardest. However, in the Sahel there is scope for increased contribution by parents even to the financing of teacher salaries, provided the teachers are recruited by, and accountable to, the local community. Foreign Aid The volume of foreign aid by level of education showed a marked imbalance in favor of higher levels of education in the early 1980. In Sub-Saharan countries (198446), only 7.8 percent of aid to the educational sector concerned primary schooling (compared to 59 percent for higher education), and the situation for Sahelian countries in this respect was even worse since the average share of aid to primary education was only 2.8 percent of the total amount allocated to education (Jarousse and Mingat, 1992). The situation has probably improved. As far as the World Bank is concerned, being the leading development agency in the education and training sector, the percentage of funds allocated to primary education Bank-wide (as share of the total involvement in education) has increased from 10 percent in FY88 to 34 percent in FY91. A substantial part of foreign aid (mostly bilateral) is given "in kind,* and donor countries often assume they have comparative advantages in supplying highly qualified teachers and scholarships for studies abroad. Given the extremely low primary enrollment ratios in the Sael, a substantial change is needed in the foreign aid profile if aid is to contribute to accelerated development of primary education. This would imply a shift from the direct supply of teachers for post-primary education to contributing to the financing of recurrent costs (possibly also teachers' salaries) in primary schools. Obviously, a shift towards more recurrent cost financing by donors would be encouraged in cases where the countries themselves are clearly engaged in a major adjustment of their educational systems, materialized in the adoption of measures showing a clear budgetary priority in favor of the primary level. 24 Faemecig QuaMy Basic Edeaus is Mr SaWed Coadusion Current educational trends in the Sabel are very disturbing. Universal primary education is still a very elusive target and to achieve rapid improvements from current low levels in coverage and quality represents a great challenge to governments. It is obvious that major efforts are required, first and foremost from the countries themselves in improving efficiency in resource use and allocation to reach standards already reached by other low-income countries, and in involving other domestic partners in the national endeavor to offer primary education to all. International donors must reconsider their aid profile and contribute far more forcefully to the development of primary education. Experience from other developing countries indicates, beyond reasonable doubt, that primary education enrollment above 60-70 percent and retention of the majority of those enrolled for the whole first cycle of education are prerequisites for modernization and industrial development, and, subsequently, for achieving improvements in key social indicators such as fertility, mortality, nutrition and health. The expansion in coverage of primary education will have to go hand in hand with improving quality which is already considered to be low: "...the near collapse of educational services is so obvious that parents and pupils have no other rational choice than to vote with their feet" (UNESCO, 1990). The challenges are great, and governments are trying to develop and put into action educational policies and programs that aim beyond the short-term perspective. The interest and support for primary education is growing in Sahelian countries, and experiences with coherent, well-planned community-based school development programs (however few) seem encouraging. IL QUALITY OF BASIC EDUCATION Quality in education can be defined in various ways. In general terms, quality is the extent to which education contributes to the development of the society in promoting economic growth, health, social and cultural integration, preservation of the environment, etc. The definition must, however, be adjusted to match the historical, social and political conditions of a nation or region. With regard to primary education, development of basic skills in reading, writing and computing are essential elements in any definition of quality. In order to retain basic literacy and numeracy skills acquired in the school, at least four years of primary education is normally considered a minimum, and completion of the primary level strongly encouraged. Because of the pupils' general low proficiency in the language of instruction, to remain in school beyond grade 4 is particularly relevant for the Sahel (para 3.21). However, basic definitions of quality must be translated into operational terms. As a minimum, learning results must be related to inputs in order to define quality as a relationship between achievements and resources spent. However, until the early/mid 1980s, because of the lack of data on learning outcomes, the level and configuration of inputs was generally used as a good proxy for quality, anticipating a close relationship between input and student learning. It is only recently that the focus has shifted from judging quality on the basis of inputs such as teacher qualifications, class size, pedagogical materials, grouping of students, etc., to the outcomes of the educational process, i.e., student learning, schooling careers, employment opportunities and inequities. It is not easy to define quality standards in education. Again, as indicated above, these are often defined in terms of inputs, for example, as a set of minimum requirements as to class size, classroom furniture 25 CaW Sua dn FMscing "Ilay Basic Edumam and equipment (like lightning, temperature etc.), staffing norms, textbooks, repetitions, etc. Such regulations would be useful when properly developed and enforced, and there is obviously a need for such basic management tools to be developed and put in place in Sahelian countries. However, such quality standards should not be taken as indicators of outputs. The emphasis traditionally put on them is probably due to the fact that: (a) the management, planning and financing of education mainly concerns these items; (b) the actors are primarily concerned with their actual conditions of teaching; and (c) that the outcomes are less visible and more difficult to ascertain. Factual data show that there are neither obvious nor linear relationships between input factors and student learning. Empirical studies on quality of learning outcomes are not at all frequent, but all indicators are that education in African countries is subjected to erosion of quality (World Bank, 1988). In the Sahel, there is at least one recent study from Burkina Faso that seems to confirm the general picture (see Box 3). Repetition and Drop-out in PrWwy Euwaion An important dimension of a system's operational efficiency concerns the amounts of resources spent on pupils repeating grades and/or dropping out prior to the completion of the cycle. Research in the area (e.g., Harbison and Hanushek, 1992, Cuadra and Fredrikson, 1992) shows that investments in quality improvement measures could reduce costs per graduate and the savings generated by the resulting improvements in internal efficiency may in some cirrumntanes largely compensate for the costs of these investments. Flow indicators (see Annex 1, Table 6) show the high repetition rates in the Sahel. Reducing repetition must, therefore, be a priority. Drop-out depends on both school and home factors, and it is probably a more challenging task to keep the children in the school than to enroll them in the first place. Chad (where repetition and drop-out combined represent a waste of two out of every three student years) and Mali depart substantially from other Sahelian countries in that they have much higher repetition and lower retention rates. However, the fact that Niger and Burkina Faso are much more successful in this respect suggests that improvements would also be realistic in the low-performing countries. Senegal (see Box 4) has an intermediate position. One of the main issues in the design of primary education curriculum concerns the tensions between the preparation of pupils who will continue to secondary education and the others who will leave school after completing primary education. Priority given to universal primary education will result in a growing proportion of pupils in the second category. Siace primary education traditionally has been directed towards the needs of the first category, in the Sahel the pupils leaving school after grade 6 are often regarded as dropouts. A major change in attitude is required for the completion of primary education to be considered as an objective in itself. Primary school graduates are not dropouts; they are the right products of a system that produces human capital at various levels of qualification for the economic and social development of the country, and maximizes the overall educational opportunities for the population. The objective-to make the pupils proficient in reading, writing and computing-clearly has a general value, and strengthening of schools in this respect remains a central objective (see Box 5). 26 Pmhse*ay Oally Badc Educutms i A SahWe Box 3. Quality of Primary Education in Burkina While the primary education enrollment ratio in Burkina Faso increased from 5 percent in 1960 to 30 percent in 1990, in absolute terms more children are excluded from the system now than in 1960. The pass rate for the primary school leaving certificate (CEP) has been consistently below 50 percent. Although this fact may reflect the role of CEP in the selection for secondary school, it means that about half the pupils actually fail the final examination and leave primary school without any certificate. A recent study by a joint Burkinab6-World Bank team concludes: (a) The great majority of pupils have not mastered the essentials of what they are expected to know after six years of schooling. (i) In French, pupils perform especially poorly in oral and written communication, particularly when expressing their own ideas. By grade 6, only 25 percent could read fluently, and less than 20 percent could write simple texts. Test results as well as classroom observations suggest that French is not taught effectively, particularly in the lower grades. (ii) The French curriculum contains eight subject areas, but teachers mainly emphasize theoretical, formal aspects of the language (i.e., grammar and vocabulary) leading to poor communication and shortcomings in basic reading and writing skills. (iii) French takes up a substantial part of the curriculum. Pupils study mathematics, for example, less than 4 hours per week in the first 3 grades. (iv) In mathematics, few students have learned how to do basic addition, subtraction and multiplication computations by grade 6. (v) The final grades have only about two and a half hours per week of science instruction, health and agriculture included. While attempts are made to teach students to apply science to familiar situations and problems, most of the time is devoted to studying vocabulary rather than to initiating scientific inquiry. (b) A suitable environment for learning is largely absent. Pupils and teachers work in ill- equipped and poorly lit buildings. Instructional materials and exercise books are scarcely found. Average class size in grades 3 and 6 are 75 and 50 students, respectively. Grade I classes are often larger, well over 100 in some urban schools. Teaching is conducted in French from grade 1 on. (c) About two-thirds of the primary school teachers have completed secondary school and about half of them hold the baccalaurdat (or are about to complete it). However, only about 50 percent have had any professional training at all, and, in addition, between 300 and 800 untrained teachers (special missions, i.e., recruits of Service national populaire) are recruited annually. (d) Great emphasis is placed on preparing the pupils for the CEP. Content analysis of CEP test items indicates a lack of instructional validity when compared to the curriculum. SoWc: World Bank, 1993 27 Cam Snua i FkmsMt Qa"N1y Bose EdaiMs Box 4. Internal Efficiency in Senegal The effectiveness of education in Senegal is low at all levels as evidenced by poor internal efficiency and low examination pass rates. In primary education, repetition rates average 16 percent, ranging between 10 percent in grade 1 to 34 percent in grade 6. The average number of student years necessary to graduate from the 6-year primary cycle is 8. In 1990-91, only 56% of students passed the school leaving exam. In lower- and upper-secondary education (colliges and lycies), about 15 percent and 16 percent of students are repeaters, respectively-ranging between 12 percent in the first grade to 20 percent in the fourth grade. The average number of student years necessary to graduate is 6 in the 4-year lower cycle and 5.5 in the 3-year upper cycle. In 1990-91, only half the students passed the baccalaurdat exam. In higher education, internal efficiency is particularly low. More than 40 percent of students at the University of Dakar are repeaters. Promotion rates between the first and second year average only 25 percent. The number of student years necessary to graduate from a 4-year program averages 18 in economics, 21 in the humanities and 27 in the sciences. Sace: World Bank, 1993. Efforts to improve means and methods of teaching, examinations, etc., often prove more beneficial than major curriculum reforms. Cearly, more time could be devoted to practical subjects (e.g., issues concerning health, nutrition, water sanitation, protection of the environment, etc.). The relevance of primary education can also be improved, however, through making everyday pedagogical methods more inductive and more oriented towards solving of problems with practical relevance. But it would be questionable to use a substantial proportion of time to teach applied subjects (like agriculture) because (a) schools may not have a comparative advantage to do so, and (b) it may hamper the basic cognitive and functional acquisitions that remain the first economic and social objective of primary schooling. Promoting these basic acquisitions by relating the contents of training materials and instruction to agriculture and village life holds higher potentials. The pivot in this domain is the examination system. For all countries in the region, and in particular Chad and Mali, improvements in survival and repetition rates in primary education are a major concern of educational policy. The relevant target is not the gross enrollment ratio, but the proportion of the age group reaching a reasonable level of proficiency in reading, writing, computing and understanding of the environment. An important step towards achieving this objective is the enhancement of student learning. It might seem that much of the repetition and drop-out in primary education could be eliminated by one administrative stroke of the pen, through which repetition would be abolished in favor of automatic promotion and drop-out would be restricted by introducing compulsory education. However, the discouraging results of the past two decades' efforts to reduce the extent of these two phenomena clearly indicate the complexity of the educational, economic, social and cultural factors that cause pupils to leave school prematurely. Two broad strategy areas must be considered: (a) stimulate demand for education, i.e., encouraging parents to ensure that children attend regularly and complete the cycle; and (b) improve supply conditions, i.e. enhance schools' ability to reduce failure at exams. There is little systematic evidence on how demand for education is affected by poor quality. While studies indicate that school quality seems to have little influence on parents' decisions to enroll their 2s Fbnanc*g Quality Basic Educedon in SOe Saet children, it plays a significant role in decisions affecting withdrawal from school. Enrollments decisions are mainly related to direct costs, according to the Burkinab6 study quoted above in Box 3, perceived employment opportunities and opportunity costs. Decisions on withdrawals are related to meager learning outcomes and poor learning and teaching conditions. Box 5. Importance of Primary Education Numerous studies have established (mainly in non-African countries) that primary education with a classical content (reading, writing, computing) has a very positive impact on the productivity of farmers. Similarly, ex-post analyses of the success rate of development projects in both agricultural and other projects was positively related to literacy level. It is further observed that the effectiveness of specific training offered in conjunction with the project increased with the educational level of the trainees. A literacy rate of 50% was found to be a threshold value below which progress becomes extremely difficult. Finally, entrepreneurs in the informal sector in Niger were significantly more successful when they had completed primary school. As a corollary it was found that primary enrollment enabled young people to profit appreciably more from apprenticeship periods in the informal sector. The examples demonstrate the beneficial effects of primary education in the exercise of a number of activities for which the school offers no specific preparation. This has been labelled the "hidden curriculum" of primary education; in fact, cognitive skills appear to constitute general assets that can be effectively mobilized in a wide range of concrete contexts. Source: Jamison and Lau, 1982; Miagat and Tan, 1988; Boardan. Jmamusn and Mingst, 1919 The examination systems in Sahelian countries are mostly directed towards selection and teaching is, to a large extent, coaching for tests and examinations. Efforts to utilize tests and examinations for monitoring and feedback purposes have been launched in several countries, but the success of such efforts remains to be seen. In-service Teacher Training Although not conclusive, research indicates that in-service training is generally more efficient than pre- service training beyond a certain duration. Pre-service training tends to be theoretical and not so much related to actual classroom work. In-service training holds a higher potential, although focus and content is of primary importance. When efforts are made to adjust the level of theoretical qualifications to the actual teaching functions in primary school classrooms, programs for in-service training become even more important. To increase the likelihood that teachers put into practice what they acquire through in-service courses, various complimentary measures are worth considering: (a) increase the responsibility of headmasters in pedagogical management and supervision; and (b) strengthen the role of inspectors and involve them in diagnostic evaluations at the regional/national level. In-service training in Niger is currently run via teacher study groups in nearby schools. Thus far, this method has proven to be more efficient and less costly than the traditionally large seminar organized at the central level. Regarding in-service training as one of the elements within a comprehensive effort to support decentralized school management, present plans in Chad may serve as example of contemporary approaches. 29 CMWSA #X huckv QM.hY Bask EdMUMAM Textbooks The proportion of pupils who have textbooks is generally low in Sahelian schools, although there are great variations between countries and, most important, between rural and urban schools. Empirical studies in Benin, Burkina Faso, Niger and Togo (Jarousse and Mingat, 1992) confirm how important textbooks are for academic performance: readers are essential. In Togo, where only 37 percent of students have a reader, a gain in academic performance as high as 9 points (standard deviation - 15) was estimated to be the consequence of all pupils having a reader. In particular when the teacher has little training, the impact of textbooks is great. In mathematics, the impact is more limited, both in intensity and scope. Given the low cost of textbooks, the cost effectiveness of providing reading textbooks is particularly high, and it should be a first priority in the educational policy for primary education to provide all pupils with readers. In the IDA-supported projects in the Sael, the countries have managed to improve the situation substantially,' partly by creating a national capacity for writing and editing manuscripts for textbooks and partly by bringing unit costs down through Iuernational competitive bidding (for example, the price for a reader in Burkina Faso was brought down from about USS8 to less than USS3, and reprinting to less than USS2). Sdool Feeding Progrm In the regional context, empirical data on the impact of school canten on student learning is only available from Burkina Faso (Jarousse and Mingst, 1992) where it was found that their existence is associated with a significant gain of about 4 points (standard deviation - 15). One might think that the existence of school feeding programs increases attendance and thereby increases time-on-task, often considered the most influential factor of student learning. One might also consider school feeding programs for reasons other than the effect on student learning. After all, a school is the place where one finds more children than any other arena and where they can be reached when easy access is sought. At least supplementary micronuritional Items can be distributed as is the case in an increasing number of IDA-financed education projects. The challenge is to establish school feeding programs that are financially sustainable. Adinisatidve Co trol ald LocWa Drelopaea The major challenges relate to the revigoration and redirection of the admniristration of education. It follows from the previous analysis that major efforts have to concentrate on a general upgrading of the central, regional and local admninisration; but even more to develop a better balance between government policies, management and control on one hand and delegation of authority to lower levels on the other. The tremendous challenges require concerted efforts which also rely heavily on local participation and commitments. The countries are, however, in an early stage of this process. To achieve this is certainly not an easy task. One example might illustrate the point. 7he question of medium of instruction is a matter of debate in most of the countries. Teaching is conducted In a language not often spoken in the family or the local community (French in most countries, Arabic in Mauritania A partiady elling exAmple of the oWtag of ambook is hs h tha the 790,00 ebooks provided uader Eduamia IV (Cr. 1950-CD) to primary school papils la Chad were the firistoms provided by ti" Goverment in about 10 years. 30 PAWsiV OaNY DasAc Educde In A bSA and in Madrasa schools). Research suggests that pupil achievements would improve if the children were taught in their mother tongue, at leat during the first years at school. A policy like that would, however, complicate a number of factors in the provision of schooling (textbooks and teaching materials, allocation of teachers, etc.), in particular in countries with a number of local languages. The policy might also contradict other measures taken to promote national unity in nations with short democratic traditions. Countries like Burundi and Guinea have reverted to French as the medium of instruction after large-scale experiments on the use of local languages. One reason for failure in these and other cases was that the introduction of mother tongue in the school was not well-prepared. After a long period of experimentation with use of national languages as the means of instruction in many Sahelian countries, it is now important to evaluate the experiences gained and to draw operational conclusions as regards the role of these languages within the education system. Education systems are rather conservative organizations and changing their practices and processes is generally difficult, always time- and energy-consuming, and often costly. The key to success is to ensure that the changes advocated can actually be implemented at the classroom level and that procedures and routines at all administrative levels are put in place in such a way that they do not contradict the desired activities and processes in the classrooms. The technical complexities and the political sensitivity of the education sector must be appreciated in order to understand the challenges encountered in the implementation process. Our present knowledge with respect to the most cost-effe way to implement euational reforms, is often incomplete. While Sahelian countries have a lot to learn from countries that have been more successful in developing their educational systems, transferring experiences and solutions from one country to another has to be conducted with extreme caution as so much is dependent upon local circumstances. TIe experience gained over the last decades demonstrates that most developing countries have weak capacities to implement the changes in educational and administrative processes required to upgrade their education systems. Significant capacity strengthening is normally necessary to (a) develop the knowledge base needed to prepare viable reforms on which a national consensus can be reached, and (b) plan and manage the implementation of these reforms. Through its role as guardian of the education of a nation's youth, and manager of a large share of public budgets and employees, the Ministry of Education is probably the national organization which most directly affects the lives of the largest number of people. It is a paradox, therefore, that research, planning and manageial capabilities of this key ministry often lag behind those of other ministries (Fredrikson, 1990). It is equally important to observe, however, that improved functioning of central ministries and agencies is .certainly not sufficient. Decentralization is not a prominent feature of education in the Sabel. The distance from policy and decision makers at the central level to the individual schools is normally far too great and, consequently, local commitment in the schools to central decisions is generally weak-at best, school directors and teachers get some information, or rather some new rules to follow. Improving implementation of government policies is not only a matter of refurbishing the organization (for example by providing transport facilities for inspectors) to make it operate as it should. It is more a matter of shortening the distance between decision makers and teaching staff. 31 Casme is Pkmhe Qu&y Bac EdaWdM Norm .1Iy, accountability is low among teachers, not because of low motivation (although that might also be th: case, i.e. when several months salary arrears occur), but mainly because teachers have a split loyalty to a remote and distant administrative center on one hand and to the parents and the local community on the other. The main objective of decentralization is to make school directors and teachers more responsible and accountable to parents and thereby more sensitive to their concerns, and to facilitate cooperation with the local community. In the Sahel, the experience gained in Mali and, especially in Chad is significant, although it is too early to know exactly how far the decentralization of decision making should go. A comprehensive policy combining elements of decentralization, diversification of financing and local school improvement seems to have high potential as a coherent strategy. Chad has had scattered examples which have now been brought together (see Box 6). From the initial pilot phase, important experience will be gained that hopefully can be disseminated to other countries in the region. A keyword in present policy making is diversificaion. Sahelian countries already have diversity in educational provisions, in primary education mainly in the form of madrasas, Christian and non- denominational schools as well as community-schools. The term madrasa ("school* in Arabic) is most often used to denominate a school with secular subjects in the curriculum, although the proportion of religious/secular teaching varies. III. BASIC EDUCATION FOR GIRIS Gender differences in enrollment are pronounced at all levels of education in the Sahel, both absolutely and relative to other world regions; particularly rural areas. Primary school enrollment has historically been among the lowest in the world, and the gap in enrollment between boys and girls is among the highest. This has resulted in literacy rates, among Sahelian women, which average only 19 percent, compared with the Sub-Saharan average of 34 percent. Low enrollment is explained by low admission to the first grade of the cycle and by high dropout. In the Sahel, the main factor causing the comparatively low enrollment of girls is the very low admission rate. Low admission enrollment is a greater problem in rural than urban areas, and the rural girls who start school seem to be a selected group that tends to remain in school. The relative "handicap" of girls in primary education in Sahelian countries is caused by a complex set of factors related to poverty, geographic location and government policies concerning school locations. Costs of schooling, an unsupportive school environment and socio-cultural factors that limit female mobility and opportunity (the stams of girls in patrilinear kinship systems, the practice of forced and early marriages, betrothals and bridewealth arrangements which require submissiveness of girls to their parents' decision, the utilization of girls for domestic chores, parental fears that education may give girls a will of their own and cause them to rebel against or disobey/embarrass their families) are all contributing factors. However, little information is currently available on the relative importance of these factors across communities and countries. Most of the countries are currently conducting studies that could improve the knowledge basis for policies and programs. Apart from the fact that enrollment targets cannot be met without significant increases in girls' participation, educating girls is closely linked to development efforts in general: 32 Fbucukeg augay Baic Educestisw SA W. b (a) Investing in educating females is probably the single most effective investment to improve standards of living in developing countries, particularly among the poorest countries. (b) Women's education is strongly linked to lower fertility and child mortality, better family health and nutrition, longer life expectancy and higher levels of education in the next generation (see Box 7). Bor 6. Commuiry-based Primary Education in Otad The level of community-based primary education in Chad is unique in the Sahel. First, 400 local communities have created their own primary schools (dcoles spontandes) which are managed with little support from the Government. Forty-six percent are incomplete and offer less than three grades. In total, they cover approximately 19 percent of primary enrollment. Building on this proven, local decision making, the Government wants to improve school quality by experimenting with three types of school-based programs covering 36 percent of total primary school enrollment in 1990-91: 90 pedagogic improvement projects developed by and involving the community: Projects designed at the school level by teachers, parents, and NGOs will include activities such as: (a) improving reading, writing, mathematics, and social science skills; (b) implementing new pedagogical models and teaching approaches; (c) increasing community participation in school activities; (d) developing supplementary instructional materials for independent study; and (e) implementing pedagogical and supervision workshops (atllers itegrds) for teachers in a set of neighboring schools. Comprehensive diagnostic testing of students in 100 schools to: (a) monitor performance changes over time among participating schools; (b) identify problem areas in the curriculum; (c) identify schools among the sample which need special assistance; (d) evaluate the effectiveness of double- shift and multigrade teaching and other investments; and (e) evaluate the effectiveness of the local pedagogic improvements. Support to 100 APE (Association des parents des lvs) to organize themselves, to administer their funds, and to mobilize the communities in support of primary education. School inspectors, directors and staff of NGOs will advise APE boards on: (a) the selection process of teachers; (b) information on low-cost building design; and (c) accessing support for community schools. Srce: Wodd Bank, 1993d To date, with the exception of Cape Verde which has close to universal primary education, no major improvements can be reported from the Sahel. The above-mentioned studies on factors hampering girls' education conducted under Bank-fmanced projects, aim at developing strategies for improving the situation. IV. CONCLUSION Sahelian countries find themselves at a crossroads. In spite of efforts made, and some successes in the development of primary education, with the notable exception of Cape Verde, the attainment of universal 33 Car .adl bs FbMWeV ONuAVy Badc Emsdms primary education is an elusive goal, and pursuing the present policies and priorities could even imply that the meager achievements made could be compromised. Consequently, it is time to launch a vigorous and determined action program to avoid stagnation at a very low level of basic education development. Box 7. Factors Contributing to Gurls'Low Enrollment 7he under-representation of girls in education is mostly a demand side phenomenon since schools are a priori equally open to both sexes. 7he weakness of the demand from girls is in general linked both to the cost side (girls may take care of younger siblings and they often play a significant part in household production from an early age) and to the benefit side (weakness of labor market opportunities for women). Parents may also have a different view concerning the temporality of investing in boys and girls. Girls often have to leave the family early to marry. The view that educating girls is less profitable is linked to private family perceptions which are not in line with concerns on welfare in the society. Recent studies show that there are obvious social benefits related specifically to investing in educating girls. For example, different empirical studies show that an extra year of primary schooling for girls would later result in a 5 -10 percent reduction in fertility, and that it would reduce child mortality by as much as 10 percent. These two socially desirable objectives are accomplished at a lower cost through education than via direct and specific interventions in health programs. Given empirical results of this type, it is justified to more vigorously promote girls' education than through policies specifically designed for this purpose. The design of such policies and programs requires a broad understanding of the determinants of girls' access and retention in schools. Sawre: Smunr. 1991, SubeaM and RanWy, 1992 Given the disturbingly low enrollment rates, quality issues have not been equally prominent in discussions on educational policies. It is important to regnize that progress in both coverage and quality is critical in the Sahel. Thus, while focusing on improving the efficiency of resource use and allocation in order to bring down unit costs and thereby create room for maneuvering for further developments, adequate attention has to be given to quality improvement measures. There is considerable evidence to show that investment in such measures will give handsome returns in the form of cost savings per graduate. In the expansion of coverage, two groups of children become the main targets: (a) rural youth, and (b) girls. Biases in enrollment rates are so strongly in favor of the cities and boys that real progress cannot be made without a substantial shift in favor of the former two groups. Encouraging developments have been recorded in some areas, and it seems that multigrade and double- shift teaching can contribute to bringing down unit costs, together with other targeted strategies that have been discussed. Greater involvement of communities and parents in the financing and management of schools also offers promise. To date, Sahelian countries have neither pursued decentralization nor diversification to any great extent. In summary, the political challenge of developing basic education in the Sahel is enormous. It is, in fact, a matter of changing the basic conceptions of education away from an elitist system benefitting the few towards a more egalitarian one for the benefit of development of the nation. 34 Flownsng Quky Basc Edadn bi do SahSl REFERENCES Bellew, Rosemary and Elizabeth King. Promoting Girls' and Women's Education. Lessons from the Past. World Bank Staff Working Paper, no. 715. Washington, D.C., 1991. Bourdon, J., Jean-Pierre Jarousse and Alain Mingat. Formation et revenus dans le secteur informel: Veemple de l'arrisanar et du commerce & Mamey. Dijon: IREDU, 1989. Ciardi, Paola. Qualitative Survey of Utilization of Health Services, Participation and Health Needs in Burkina Faso's Rural Communities. Washington, D.C.: The World Bank, 1993. Colclough, Christopher and Keith M. Lewin. Educating All the Ciildren. Oxford: Clarendon Press, 1993. Cuadra, Ernesto and Birger Fredriksen. Scope ofEfficiency Gains Resultingfrom Reduction n Repetion and Dropout. A Simulation Erercise. PHREE Background Paper Series. Washington, D.C.: The World Bank, 1992. Fomba, C. Efet du tye d'1cole primaire sur la performance des lves: otude comparative des "Ecoles de Basew et des dcoles publiques du district de Bamako. Dijon: IREDU, 1992. Fredriksen, Birger. Increasing Foreign Aid for Primary Education: The Challenge for Donors. PHREE Background Paper Series no. PHREE/90/30. Washington, D.C.: The World Bank, 1990. -. Social and Political Constraints on Education Reform. The Case of Senegal. In Fuller and Habte (Eds.): Adjusting Educational Policies. World Bank Discussion Paper, Africa Technical Department Series, 132. Washington, D.C., 1992. Harbison, Ralph W. and Eric A. Hanushek. Fducational Performance of the Poor. Lessons from Rural Northeast Brazil. Oxford: The World Bank/Oxford University Press, 1992. Jamison, D. and L. Lau. Education and Farm Efficiency. Baltimore: The Johns Hopkins University Press, 1982. Jarousse, Jean-Pierre and Alain Mingat. Euluaiaon de Venseignement primaire au B&Min. UNESCO/BAO-PSA, 1991a. Les Facteurs agissant sur les acquisitions des Ms 4 1'dcole primaire au Burkina Faso. Dijor. IREDU, 1991b. Evaluation pddagogique et Economique de la double vacation au Niger. Dijon: IREDU, 1991c. OpuionrforAccelerated Development ofPrimary Education in the Sahel. Washington, D.C.: The World Bank, 1992. (First draft) 35 cam a*" An FkMMMg O.May Bde ".M.. -. 1993. LEcole primaire en Afrique. Foncionnement, qualitd, prodults: le cas duTogo. In press (I'Rarmattan) Mingpa, Alain and Jee-Peng Tan. "The Economic Return in Investmens in Project Related Training: Some Empirical Evidence.* International Review of Education, 34, 1988. Ministbre de la Coopdration et du D6veloppement. Ls Crises iconomiques et fmsancitres en Afrique Sub-Saharienne et les rdorientarious souhliables des actions de la coopiration franpaise. Paris: Ministbre de la Coopdration et du D6veloppement, 1992. Subbarao, Kalanidhi and Laura Raney. Social Gains from Female Education. World Bank Staff Working Paper, no. 1045. Washington, D.C., 1992. Summers, L. Invrsting in All the People. World Bank Informal Sector Study. Washington, D.C., 1991. Thomas, Christopher and Christopher Shaw. Issues in the Developmem of Multigrade Sdools. World Bank Technical Paper, no. 172, Washington, D.C., 1992. UNESCO. Pr6mary Education and Econwic Recession in the Developing World since 198). Document prepared by D. Berstecher and R. Carr-Hill for the World Conference on Education for All, Paris, 1990. UNDP. Hwnan Development Report. Oxford: Oxford University Press, 1992. UNICEF. World Conference on Education for All. Meeting Basic Learning Needs. Final Report. New York: UNICEF, 1990. World Bank. Burkina Faso: An Assesment of the Quality of Prmary Education. World Bank Formal Sector Study, Population and Human Resources Operations Division, Sahelian Department, Africa Region. Washington, D.C., 1993a. -. Education in Sub-Saharan Africa. Policies for AdJuswneM, Revitalization, and Expansion. A World Bank Policy Study. Washington, D.C., 1988. -. Social Indicaors ofDevelopment, 1993. Baltimore: The Johns Hopkins University Press, 1993b. -. Social Ifrastrucrure Construcrion in the Sahel: Opom for Improving Current Pracrces. World Bank Regional Sector Study, Population and Human Resources and Infrastructure Operations Divisions, Sahelian Department, Africa Region. Washington, D.C., 1993c. -. "Staff Appraisal Report. Republic of Chad. Basic Education Project (Education V). "Population and Human Resources Operations Division, Sahelian Department, Africa Region, Washington, D.C., 1993d. -. "Staff Appraisal Report. Republic of Senegal. Second Human Resources Development Project (Education V)." Population and Human Resources Operations Division, Sahelian Department, Africa Region, Washington, D.C., 1993e. 36 Fumne6sg Qgality Bade Ed"cadm hi As Shl Table 1. Budgetary Indicators ANNEX 1 Page 1 of 3 Educational Budget in percent Education Budget GDP Gov. Budget Primary Secondary 1980 1988 1980 1988 1980 1988 1980 1988 Burkina Faso 3.0 2.9 19.8 14.9 38 39 23 26 Chad - 1.5 - 12.0 - 53 - 30 Mali 3.7 3.3 30.8 17.3 42 57 22 27 Mauritania 5.0 4.9 21.0 22.7 30 33 36 40 Niger 3.1 2.5 22.9 18.0 37 46 46 34 Senegal 4.5 4.8 23.5 24.0 44 47 27 25 Sahel 3.6 3.3 22.0 18.2 391 46 31' 30 - Sahel without Chad 3.9 3.7 23.6 19.4 38 44 31 30 Non Sahel Africa - French Speaking 4.7 4.0 20.5 17.6 42 43 30 31 - English Speaking 4.6 4.7 14.3 15.0 45 47 32 34 Asia < USSS0 per capita 2.8 3.4 10.3 12.4 42 48 27 29 Latin America< US$1,000 3.2 2.3 19.7 17.4 50 54 23 29 No: * E0imatio incuding Chad on the basis of the average pefr of SahWl. Some: Mos of the figurm preented i ths = the foilowing tables ae derived eroe Jrousse ead Miagat (1992). Table 2. Disribution of Educational Resourcesv. Gini Coefficie 1980 1988 Burkina Faso 0.94 0.85 Chad - .69 Mali 0.84 0.86 Mauritania 0.85 0.76 Niger 0.90 0.85 Senegal 0.77 0.68 Sahel 0.84 0.78 Non Sahel Africa: - French Speaking 0.62 0.58 - English Speaking 0.60 0.59 AsiaEEEaARY EDUCATION: THE NATIONAL POLICY ON EDUCATION (DISTRICT PRIMARY EDUCATION PROGRAMME) Cognizant of the problems of inequitable access, and the high internal inefficiency of education in India, in 1992, GOI, in consultation with the states, updated the 1986 National Polic, on Eduation. The new strategy is codified in the District Primary Education Programme (D.P.E.P). The Programme builds upon implementation experience gained from basic education projects implemented in Bibar, Rajethan, and Andhra Pradesh. The D.P.E.P is a multi-pronged approach that confronts the problems of. (i) resource mobilization; (ii) inequity; (iii) poor educational quality. Mobilization and utilization of resources will take place within a more efficient education management and delivery system. Based on national experience of implemening UEE, the National Policy drew the following conclusions: (a) UEE was context-specific, and the exact nature of problems in elementary education varied from state to state. In states where near-uaiversal enrollment has been achieved, quality improvements were necessary; in other states both access and equity were a problem; (b) the contextuality of educaton problem entails local-level planning with disaggregated targets and decentralized planning, and management. It recognizes the importance-of bonom-up planning and involving local-level hatitutions, i.e., teachers, parents and non- governmental organi tins who will be "owners* of these educational services; (c) resources are an critical but not sufficient condition for attmining UEE; (d) the new Policy emphasizes a holistic plan which perceives the complexity of educational problems, particularly those related to gender, and disadvantaged groups, school ineffectiveness, and poor learning achievement (District Primary Education Programme, 1993). 93 CaN SaM, e U..M*1 Queay Basic Edmadm In terms of financial resources. the Central Government will boost allocation to elementary education. Thirty-seven percent of planned outlays to education under the Seventh Plan (1985-90) will be spent on elementary education. Under the Eight Plan (1992-97) GOI will commit 46.8 percent of planned outlays in education on elementary education. In total, development outlay on elementary and adult education will be tripled under the Eight Plan. (India-Education for All, 1993) (Annex 10). An important shift will be made in the allocation of resources to the states. Resources will be a[Wl to educationally backward states with guidelines for targeting the most disadvantaged districts within these states. The DPEP will strengthen the existing decentralized structure. Districts will take more responsibility for the management and delivery of education. Districts will be selected on three sets of criteria: (i) female literacy rates which are below the national average; (ii) demand for education as generated by the total literacy campaign, and (iii) high enrollment ratios and low dropout but poor learning achievement. After districts are selected, they will be assisted in developing education plans (India-Education for All, 1993). In keeping with the objectives of the DPEP, the forandation of the district plans will be through a process of capacity-building rather than by entrusting the job on consultants, individuals, or institutions. In recognition of the poor institutional capacity of local-level institutions, there is emphasis on capacity- building through networking with state, and central level institutions that will provide training (GOI - The District Primary Education Programme). I. THE UITAR PRADESH BASIC EDUCATION PROJECT The case study of Uttar Pradesh will be used to illustrate how the District Primary Education Program is being operationalized in one of the least developed states of India with World Bank assistance. Uttar Pradesh is one of India's least developed states. Twenty-one percent of its population is Scheduled Caste. Per capita income is 28 percent below the average for the country. Forty-one percent of its population lives below the poverty line. Educational Devrlopmear Uttar Pradesh is one of the most educationally backward states in India. Total literacy rate is 41.7 percent, in comparison to an All India Average of 52.1 percent. Female literacy is only 26 percent which is below the India Average of 39 percent. Grass enrollment ration is 98.5 percent but drop-out rates are high (Annex 12). At the root of the problem is a State Primary Education System that enrolls 98 percent of boys and 62 percent of girls, but loses approximataly forty percent of enrolled boys and sixty percent of enrolled girls before they complete five years of education and acquire literacy skills. High drop-out rates can be attributed to poor educational quality which is charactrized by: (i) poor teacher performance caused by high absenteeism, inadequate training; (if) a difficult curriculum; and (ii) weak school management. Financing of Educaion The financial situation of Uttar Pradesh is extremely constrained. Growth of the State's own tax revenues has been slow, ranking U.P. number 13 lowest among the 15 largest states. Budget deficits have led to increased loans and indebtedness. Central government grants to U.P have been steadily expanding. Within three years (89-90 to 91-92) the contribution from grants rose five percentage points. In the education sector, the percentage share of plan expenditure for education has increased from 4.0 94 Fbmnncba Quaiy Basic Educedmin hads in 1985-90 to 4.2 percent in 1992-3. Planned outlays for the Eight Plan (1992-7) are estimated at 6.5 percent. In addition to state resources, there are several centrally sponsored schemes, financed by the central government. During the Seventh Plan education expenditures on the schemes constituted 25 percent of State Plan spending on education. Most of these resources were focused on primary schooling, non-. formal education and literacy programs. Administraive Structure The State has a decentralized administrative structure with administrative units at the district, block, and village level. Basic Education is the responsibility of the State Department of Education through the Directorates of Basic Education (elementary and non-formal education). The Basic Shiksha Parishad (BSP) is in charge of routine management of elementary education. The BSP receives annual grants from the UP Department of Education for this purpose. A branch of the NCERT, the State Center for Education Research and Training provides training (SCERT) to education personnel at the state level. District Institutions for Training and Education (DIETs) exist at the district level, and are supposed to co-ordinate with SCERT. The Zilla Parishad at the district level, the Khetra Vikas Samiti at the block level, the Nyaya Parishad at the Sub-Block level, and the Village Education Committees at the village level are the main administrative units at the administrative units at the block, and village level (Annex 11). Acute centralization of management and decision-making functions, and the lack of participation at lower levels of government have been identified as institutional obstacles to an efficient basic education system. While on paper decentralized units exist at the district level (DIETs) little effort has been made to activate these institutions and involve them in the management, and delivery of basic education. Institutional arrangements at the block and village level are extremely weak. For example, there are no institutional arrangements at the block level; many blocks contain as many as 100 primary schools. In addition, schools have very little decision-making powers. This has been identified to have negative impact on the motivation of head teachers (school principals who also teach) to provide leadership, and innovate at the school level (Uttar Pradesh Education for All, 1992). The lack of a well-developed supervision, monitoring, testing and evaluation capacity has meant that there are no conduits to identify and correct problems in the current system (UP Basic Education Project, 1993). The Project The UP Basic Education Project has two objectives: (i) to develop the institutional capacity of the State to manage the long-term, State-wide basic education program; (ii) target financial resources to improve basic education in an initial 10 of the State's 63 districts. The Project has a strong gender-focus and will target women through the basic education improvement program. The Government of India (GOI) has mobilized external assistance (World Bank), and is channeling these resources to the Government of Uttar Pradesh (GOUP). The GOI will provide additional support through the NCERT and NIEPA. The GOUP will direct funds to the UP Education For All Project Board, an autonomous institution set-up under the Project. This Board will direct funds to districts. Districts will have chief responsibility for administration, and management of elementary education and will direct 95 Car &"a i P%mr Om Badir EdmROe funds to Village Education Committees for school improvement programs, and management of female and minority scholarship programs. Selection of Disricts and Pqoject Planning Project districts have been selected to represent the range of initial conditions in the State, as well as the extent of female illiteracy. Eight districts have an average female literacy rae of 17 percent, well below the average female literacy rate for the state as a whole (26 percent), and for India (39 percent) (UP Basic Education Project, 1993). Project planning was preceded by large scale sample based baseline beneficiary assessments comprising National Council of Education Research and Testing achievement tests for language and mathematics for Grades 2 and 5, field work, simple literacy test for out of school children, surveys of teachers, analyses of school records, and focus group discussions with community members. The key findings of the surveys were used for project planning. The initial Project proposal was prepared by the Directorate of Basic Education in close collaboration with district-level staff. The Government of India provided technical assistance and on-going appraisal for the process. State-wide discussions of basic education were held at the district, block and community level in September 1992. The consultations confirmed the broad strategic approach being taken by GOUP (UP basic Education Project). Improving Access Three hundred thousand primary school places will be established. One hundred and flMy thousand upper primary places will be established. An additional 150,000 children will be enrolled in non-formal primary education program. Special attention will be paid to improving access in undeserved are., and provision of schooling for undeserved groups, i.e., females, and schedules castes and tribes (UP Basic Education project, 1993). Targering Females The Project will target females through overcoming supply and demand constraints to female enrolment and retention. FIy percent of the new upper primary schools will be exclusively for girls. Improving facilities (latrines), early childhood care, provision for female teachers, and betterment in school quality will help overcome supply-side constraints. The demand-side impediments will be overcome by including Village Education Committees, and Mabils Samakhya Programs (Women's Empowerment) (UP Basic Education Project, 1993). Targering Sdeduled Castes and Shedduled TIbes Preference will be given to opening new schools in areas occupied by scheduled castes and scheduled tribes. The VECs will expand membership to include SCIST females, and a special scholarship program for schedules caste girls will be implemented on a pilot basis (UP Basic Education project, 1993). Improving Educational Quality Poor educational quality has high costs in terms of reduced internal efficiency, and poor learning acquisition. The Project has adopted a holistic approach to educational quality improvement. It will nt 96 Fbncing QuN1y BedAc Educdadms i hiai only provide improved educational inputs, but strengthen the institutional, and management structure within which these inputs will be utilized. Community Paticipa*ion Before this Project was planned, Village Education Committees had been involved in school construction. Encouraged by the positive participation of Village Education Committees (VECs) in school construction, the Project will strengthen their role and participation. To ensure that females and minority groups are represented in the Committees, the Project will expand membership to SC/STs and females. The VECs will also be granted financial Esnibility and will disburse funds for school construction, and manage the distribution of scholarships. Efforts will be directed at improving the authority of VECs with respect to teachers, maintaining bank accounts, and distributing funds for educational materials. The VECs will meet monthly to discuss school performance. District and block education officers will participae in these meetings. The Project will finance training for VEC members. In addition, it will support a program of annual cash grants in the amount of Rs25,000 (US$950) to those VECs that show evidence of achievements in completing village surveys, increasing enrollments and retention in school (especially for girls and minority students). The awards would be used by the VECs for school improvement activities. School Management The GOUP will upgrade teachers to head teachers and appoint new head teachers. Training of head teachers will concentrate on the skills needed to provide professional supervision and support for teachers, and for working effectively with Village Education Committees, and other community organizations. School clusters will be strengthened to provide professional support to teachers, and head teachers, and to create a forum for exchange of information between school personnel. Improving Educational Inp&ts The project will support the development of the primary school curriculum to the Minimum Level of Learning (MLL) standard. Teacher handbooks, improved measures for testing and assessment, and new textbooks would be developed and tested. Efforts will be directed at eliminating gender bias in textbooks. During the later stages of the project, DIETs will work with teachers to develop supplementary reading materials which reflect local conditions (UP-Basic Education for All, 1993). Intitutional Developmmat Strentheninr the Canacift of Existini Institutions and Develoning New Institution One of the primary objectives of the project is to enhance the capacity of institutions at the state, district, block, and community level to manage, and deliver educational services. To achieve this objective the Project will establish new institutions, strengthen, and activate pre-existing one, and provide linkages between them to ensure efficient management. At the state level, the Uttar Pradesh EFA Project Board will be set up for the implementation of the Project. It will have two bodies: The Executive Committee, and The General Council. The Chief Minister will head the General Council, and will be responsible for periodically reviewing 97 com. &M~ 69 Fwlcf QWuNIY I.*c UNZIAN implementation, and providing guidelines. The Executive Committee will be in-charge of project strategy formulation, monitoring and supervision. A new State Institute of Education Management and Training (SIEMT will become fully operational during the life of the project, and provide 1000 person weeks of training, and complete at least five major policy studies. The SIMET will co-ordinate with national-level institutions, i.e., the National Institute of Educational Planning and Administration (NIEPA), and the National Council for Educational Research and Training (NCERT) (UP Basic Education Project, 1993). At the District level, a District Education Project Committee will be set up under the District Magistrate. The D.E.P.C. will include people involved In the Project at the block level, teacher representatives, and non-governmental organizations. The D.E.P.C. will have primary responsibility for preparing district education plan, administrative and financial control of the project, improving access and educational quality, and implementation and monitoring of the Project. The District Institutes of Education and Training (DIETs) will be strengthened, and associated Block Resource Centers (BRC) will be established in each blocks. The BRCs will in turn support Cluster Resource Teachers, each of whom will provide continuous in-service support to teachers in 8-10 schools. The DIETs will serve as the principal professional support agency for the work of the schools, and co- ordinate with Block Resource Centers. They will organize and provide regular in-service training for teacher trainers, head teachers, VEC members, and non-formal education instructors and supervisors. DIETS will eventually serve as centers for curriculum and materials development and evaluation studies. At the block level, a Block Education Project Advisory Committee, and a task group will be set-up to provide co-ordination at the Block level (See Annex 12 for further details on functions of units at State, District, and Block level). Zmproving Infonnation System for Plni, Moaemeat and Professinal Support Information is a key ingredient for the successful co-ordination between decentralized administrative units. Through the project the GOUP will improve the flow of information in four ways, through: (i) school mapping; (ii) management information system development; (i) state-wide assessments of learning achievement; and (iv) research and evaluation studies. The gathering and processing of information will involve a variety of institutions at the local, district, and state levels. The Village Education Committees (VECs) will be involved in undertaking household surveys. Results from this survey will be used to develop village education plan. The National Institute for Education Planning will assist in the process of micro-planning through the provision of guidelines. The MIS system will have linkages to the district level where monitoring cells would be set-up. A systematic program of assessing learning achievement will be used to monitor learning acquisition. Finally, empirical and special policy stndies will capture the on-going innovations in education reform, and disseminate at the State, and National level (UP Basic Education Project, 1993). The baseline assessment studies, which were carried during the project planning stage, will be completed in all project districts by the end of the first project year, and replicated at three year intervals in order to measure project impact. Condusion As was stated in the Introduction, India is a large and diverse country with complex economic, socio- cultural, and political problems. In this context, the New National Policy on Education, and the District 98 FbMinig Qualy Basdc Edwdan in Indla Primary Education Programme, as illustrated through the UP Basic Education Project, provides a more adequate framework for achieving a Quality Basic Education for All in India. From the institutional perspective, decentralization is the most viable administrative structure for India. For example, the population of Uttar Pradesh exceeds the populations of 168 member states of the United Nations (Human Development Report, 1993). This was recognized at Independence when India was declared a Federal Republic. But, what was perhaps underestimated was the complexity of implementing such a structure. A successful implementation strategy for decentralization should: (i) analyze the role of each decentralized unit, and optimally utilize each unit; (ii) provide coherent linkages between these units; (iii) develop accountability mechanisms; (iv) redefine the role of the central government; (v) build the capacity of decentralized units to fulfill their duties. Prior to the D.P.E.P, there were lacunas in the decentralized administrative structure. For example, district, block, and village-level organizations were not being utilized. This resulted in under-utilization of financial, and human resources at the local level, and inefficient management and administrative systems. The DPEP addresses these pitfalls, and provides a more comprehensive framework for decentralized administration. For example, it grants financial and management responsibility to local levels of government. Functions at the state-level are restricted to channeling resources to lower levels of government, state-wide monitoring and evaluation, setting state-level standards, and training. The role of the Center also undergoes a transformation as it takes on responsibility for mobilization and targeting of resources, setting national standards, and providing assistance for capacity-building. The State and Center will collaborate on capacity building, research and evaluation. Implementation of the D.P.E.P should pay special attention to capacity-building at the community and district level. A process of constant monitoring during implementation should be applied to address any problems that might surface. A sustainable fmnial strategy for elementary education should: (i) mobilize additional domestic resources for elementary education; (ii) improve the internal efficiency of the elementary education system; (iii) ensure equitable distribution; (iv) diversify the sources of funding for elementary education (Verspoor, Lockheed and Associates, 1991). Prior to 1992, these issues were not being addressed in India. The D.P.E.P. provides a more sustainable financial strategy for achieving Education for All . By adopting a strategy that is a blend of mobilization of additional resources, and decentralization, the DPEP incorporates all of the above into the Plan. Mobilization of additional resources can be achieved through increasing budgetary allocations to elementary education, and mobilizing non-governmental sources of finance. The present Strategy does both by increasing the amount going to elementary education (47 percent in the Eight Plan), and mobilizing financial and non-financial resources at the district and community levels. Communities will provide financial and human resources, and be involved in school construction. At the District level, there will be significant utilization of human resources. Both the Center and the State will target resources to elementary education, economically backward states, districts within states, and disadvantaged groups. Allocation and utilization of resources will take place within a more efficient delivery system. However, two areas need attention. First, efforts should be made to mobilize additional domestic resources for elementary education. Presently, the D.P.E.P is heavily funded by external donors. Second, more efforts should be directed at exploring non-governmental sources of finance for elementary education, particularly in districts that are not as poor as some others. An incremental approach should be adopted to tackling these problems. 99 Cas samx& i Ibsmmcg Qunkoy Badc Educaam Educators are increasing recognizing that teaching and learning is a complex process that requires a flexible environment to respond to the unexpectec and meets local needs. Therefore, improving educational analit requires providing educational nputs (books. materials, trained teaches), and improving the participation of the school and community in the utilization of these inputs (Verspoor, Lockheed and Associates, 1991). A flexible learning environment is specially required for a country like India which is socio-culturally diverse. Prior to the DPEP, schools and communities had been not been involved in quality improvement. The DPEP provides a more holistic framework for improving educational quality by involving communities and schools, and providing educational inputs. The problem of injqitable access to schooling is caused by the absence of quality schools, and lack of demand. This is particularly true for females, and minority groups and children from very poor households. The DPEP provides a comprehensive approach to targeting females, and minority groups by: (i) providing scholarships, (economic incentive); (ii) expanding the membership of village education committees to include women from minority groups, involving Mahila Samakhya Program (institutional development and social change); and (iii) increasing the number of female teachers, building schools in areas with high concentration of scheduled castes (pedagogical). 100 Fioane Quslay Badc Educaws India ANNEX1. India A. Percent Literate Above Age 7 (1991) Total 52.11 Male 63.85 Female 39.42 B. Primary (Age 6-11) PeSADn 1. Total cohort population (1990) 98,111,300 100 Male 50,390,900 51.3 Females 47,720,400 Scheduled Castes (SC) 15,452,260 15.7 Male 7,936,600 51.3 Females 7,516,000 Scheduled Tribes (ST) 7,613,400 7.7 Male 3,910,300 51.3 Females 3,703,100 2. Primary enrollments 99,118,320 100 Male 58,094,716 59.4 Females 41,023,604 Scheduled Castes (SC) 15,794,427 15.9 Male 9,736,924 61.6 Females 6,057,503 Scheduled Tribes (ST) 7,868,187 7.9 Male 4,957,611 63.0 Females 2,910,576 3. Gross Enrollment Ratios (GER) Total 101.03 Male 115.29 Female 85.97 Scheduled Caste 102.21 Male 126.78 Female 80.60 Schedule Tribe 103.35 Male 126.78 Female 78.60 4. Number of Schools 558,392 (primary and upper primary combined) 5. Number of Teachers n.a. 6. Enrollments in Primary NFE n.a. 101 cm AN h hnoor ONAY Bder Educmdon C. Upper Primary (Age 11-14) 1. Total Cohort Population (1990) 55,372,400 100 Males 29,405,800 51.3 Females 26,966,600 Scheduled Castes (SC) 8,721,100 15.7 Males 4,473,900 51.3 Females 4,257,200 Scheduled Tribes (ST) 4,296,900 7.7 Males 2,204,300 51.3 Females 2,092,600 2. Upper Primary Enrollments 33,282,999 100 Males 20,844,291 62.8 Females 12,438,708 Scheduled Castes (SC) 4,160,526 12.4 Males 2,747,100 66.0 Females 1,412,416 Scheduled Tribes (ST) 1,706,906 5.1 Males 1,131,388 66.3 Females 575,518 3. Gross Enrollment Ratios (GER) Total 60.11 Male 73.38 Female 46.13 Scheduled Caste 47.70 Male 61.40 Female 33.17 Scheduled Tribe 39.72 Male 51.33 Female 27.50 4. Number of Schools n.a. 5. Number of Teachers na. D. Other Statistics Male: 51% Drop-out, Grades I-VI (1986/87) Female: 67% Soane: 1986 All India Survey 102 Fwhmig QualSy Basdc EdHcOan in Ims ANNEX 2. Distribution of Perrons Newr Enrolled as Students. By Fracrile Groups, 6-11 Age Group (percent) Fractile Group Rural Male Rural Female Urban Male Urban Female 0-20 32.88 33.51 50.64 52.48 20-40 28.12 27.22 31.27 30.00 40-60 20.30 21.08 13.78 12.50 60-80 13.87 12.95 3.51 8.91 80-100 4.83 5.24 0.80 1.11 Sorce: NSS (198687) Odidren Never Enrolled, 1988 (6-11 age group) Population (million) Newr Enrolled (million) Percent Never Enrolled Rural Male 46.5 16.4 35.3 Rural Female 41.4 22.7 54.8 Urban Male 12.4 2.0 15.8 Urban Female 10.8 2.5 22.8 All Children 111.1 43.5 39.2 Source: NSS (198687) 103 cae aina S. Fb.co.e Q..Uy a* Ej E - ANNEX 3. Cildren Dropping Our From Schools: Ihe Primary State as Percentage of All Drop Outs by Sex, CaselrIbe and Rural-Urban Residence Rural Areas Urban Ars Stares Caste/ Trbe Boyr Giris Total Boy Giris Total All India SC 61.2 68.8 63.3 46.2 49.1 47.2 ST 67.2 67.3 67.2 44.0 55.8 47.9 Odheu 49.4 56.9 52.0 36.9 42.8 39.2 All 52.6 58.9 54.7 38.0 43.8 40.4 And~ra Prade SC 73.8 82.8 77.0 34.3 49.6 41.7 ST 68.7 100.0 73.7 68.0 54.8 62.2 Other 64.1 71.6 66.8 41.9 58.1 48.7 All 65.7 73.6 68.5 41.7 57.1 48.2 Asmm SC 54.9 76.4 62.8 48.9 39.3 44.3 ST 63.4 56.6 60.7 9.3 100.0 16.9 Ohrs 40.7 50.5 44.4 28.7 31.9 30.1 All 44.1 53.3 47.4 29.4 32.9 30.9 Bihar SC 64.3 100.0 67.1 59.4 43.7 57.3 ST 59.4 49.6 58.5 28.8 100.0 32.9 Othes 52.8 78.5 59.1 52.1 62.6 55.4 All 54.8 78.4 59.9 51.9 62.1 54.8 Jammu & Kashmir SC 23.7 0.0 20.8 62.8 50.7 66.3 ST 34.8 53.0 39.7 0.0 0.0 0.0 Othrs 36.7 48.9 39.1 24.5 27.5 25.7 All 36.3 49.3 39.0 29.6 28.5 30.0 Madhya Pradesh SC 81.7 92.5 83.6 62.2 50.8 58.9 ST 83.3 64.8 80.4 42.0 80.0 56.2 Ohrs 69.8 82.0 72.8 48.1 55.4 50.8 All 74.5 80.8 75.8 50.6 55.9 52.6 Orissa SC 59.8 70.9 62.6 50.3 65.2 52.9 ST 70.2 81.8 72.7 38.8 30.7 35.7 Oheru 62.3 75.9 67.1 36.9 52.1 42.8 All 63.4 76.0 67.4 39.0 50.9 43.3 Rajasthan SC 62.3 100.0 67.0 73.9 100.0 80.2 ST 49.7 95.4 51.8 86.7 0.0 86.7 Otheru 53.1 75.7 58.0 41.6 60.0 47.7 All 53.8 78.3 58.3 47.2 63.6 52.4 Uttar Pradesh SC 56.8 4.9 59.6 47.5 61.7 50.3 ST 85.6 80.6 85.3 58.2 0.0 44.9 Others 54.2 75.2 57.9 39.8 54.9 44.3 All 55.6 76.5 58.9 41.3 55.5 45.3 West Bengal SC 63.3 71.1 65.3 49.2 47.1 48.5 ST 52.9 36.2 48.0 41.0 41.2 41.0 Otbern 45.5 54.0 48.7 38.0 39.4 38.7 All 50.7 56.4 52.7 39.5 40.4 39.9 Sa.~e: Na~ionai SchooI Suai., (July 196 - Jun 19"),. rport No. 365. Vol. I r II 104 Fbmmng QOdky DeicEdm&ndM ' Imd ANNEX 4. Expenditure on Education as Percentage of GMP in Indian 1950-51, 1989-90 % of % of % of % of Year GNP Year GNP Year GNP Year GNP 1950151 1.2 1960/61 2.5 1970171 3.1 1980/81 2.9 1951/52 1.3 1961/62 2.7 1971/72 3.3 1981/82 3.3 1952/53 1.5 1962/63 2.8 1972/73 3.2 1982/83 3.2 1953/54 1.5 1963/64 2.7 1973/74 2.7 1983/84 3.2 1954/55 1.8 1964/65 2.5 1974/75 2.9 1984/85 3.7 1955/56 2.0 1965/66 2.8 1975/76 3.2 1985/86 3.9 1956/57 1.8 1966/67 2.8 1976/77 3.2 1986/87 3.8 1957/58 2.1 1967/68 2.7 1977/78 3.2 1987/88 4.4 1958/59 2.1 1968/69 3.0 1978179 3.0 1988/89 3.5 1959/60 2.3 1969/70 3.0 1979/80 3.1 1989/90r 3.4 Note: Expcditurs of the Deprnment of Educion only. Percent 5 4 ----------------------------------------------------- 3 ------------------------------------ 2 ---- ------------------------------------------- 1------ --------------------------------------- 0 1950-51 1960-61 1970-71 1980-81 1989-90 105 CaW &AN in Fbncbq Qauay Bac Educam ANNEX 5. Secroral Outlays in Fwe-Year Plans in India (percen) 1s Ples 2nd Plan 3rd Ples Amd. Pies 4*r Pkm 5th Ples drk Plan 7th Pes Mth Ples Agricultual & Allied 14.30 11.30 12.70 16.70 14.70 12.30 13.70 14.30 12.70 Irrigation & Flood Conrol 22.00 9.30 7.30 7.10 8.60 9.30 10.00 7.50 9.50 Power/Eaerty 7.70 9.50 14.60 18.30 18.60 18.80 28.30 26.60 30.50 Industry & Miuerals 4.90 24.10 22.90 24.70 19.70 24.30 15.30 12.50 10.80 Tramaort & Comunicaeons 26.40 27.00 24.60 13.40 19.50 17.40 16.10 17.15 19.60 Educama 7.90 5.30 6.90 4.60 4.90 3.30 2.70 3.70 4.50 Health 5.00 4.90 2.90 3.20 3.90 3.20 3.10 3.00 3.20 Total 100.00 100.00 100.00 100.00 100.00 100.00 100.00 100.00 100.00 (196) (467) (53) (663) (1578) (394) (10965) (22292) (43410) Noe: Figures in 0 a s. in 1oo minios. ANNEX 6. Expenditure on Elmenrary Education, by Objects. 1982-83 Item Primwy Middle Mem. Recurring Ex2enditure Salary of Teaching Staff 93.8 90.7 92.5 Salary of Non-Teaching Staff 2.7 3.8 3.1 Maintenance of Buildings 0.6 0.8 0.7 Maintenance of Equipment and Furniture 0.2 0.2 0.2 Apparatus, Chemicals, etc. 0.1 0.1 0.1 Libraries 0.0 0.1 0.1 Scholarships and Other Aids 0.5 1.6 1.0 Games and Sports 0.1 0.2 0.1 Hostels 0.1 0.2 0.2 Other Items 1.8 2.3 2.0 Total Recurring 100.0 100.0 100.0 Non-Recurrinr Exoenditure Libraries 0.8 2.6 1.5 Buildings 60.1 44.0 53.5 Equipment 5.2 7.5 6.1 Furniture 5.7 7.6 6.5 Other Items 28.2 38.3 32.3 Total Non-Recurring 100.0 100.0 100.0 Distribution of the Grand Total (261) (181) (442) Recurring Expenditure 97.7 97.8 97.8 Non-Recurring Expenditure 2.3 2.2 2.2 Grand Total 100.0 100.0 100.0 (11580) (8211) (19791) 106 FROciit Qamlisy Basic Educeis as hadle ANNEX 7. Total Education Expenditure in Central and State Sectors (Rs. 10 Million) CEN7RAL STATE TOTAL Period Rs. Percent RS. Percent Rs. Percent 1st FYP 39.8 26.0 113.2 74.0 153 100 2nd FYP 149.0 17.5 701.0 82.5 850 100 3rd FYP 322.1 19.6 1322.9 80.4 1645 100 4th FYP 464.4 8.2 5178.6 91.8 5643 100 5th FYP 891.7 9.6 8402.3 90.4 9294 100 6th FYP 2063.0 7.8 24316.0 92.2 26379 100 7th FYP 5710.5 10.7 47462.2 89.3 53173 100 8th FYP 74430.0 38.0 121566.8 62.0 195997 100 Nar: 8th FYP figre ar (or plan expeodare alo Somce: MHRD, 1992 ANNEX 8. Education Finance Stare Level ED. AS % OF GSDp SOC. SVCS ED. AS % OF GSDP 1990191 1990/91 1990/91 Andhra Pradesh' 308.9 49.4 3.7 Bibarb 251.0 61.8 5.0 Gujarat 285.6 53.9 3.3 Haryana 127.6 45.9 2.6 Karnataka 230.7 51.1 3.6 Kerala 130.0 57.7 6.0 Madhya Pradeshb 250.9 46.5 3.7 Maharashtra 605.9 46.5 3.7 Orissa 109.8 51.2 4.9 Punjab 191.9 56.5 3.0 Rajasthn 186.0 51.9 4.6 Tamil Nadu 291.7 49.3 3.8 Uttar Pradesh 538.8 60.3 4.0 West Bengal' 344.8 57.0 4.3 Total 275.2 53.4 3.9 NOe: Oros Stae Domestic Product (Ra. bilia) Educatioally Backward Stares 107 Case .mdes i" PWO&fg Quaay Basic EdMit ANNEX 9. Central Grants to Stare Gowrrnments (Rs. billion at current prices) 1991/92 (R.E.) 1992/93 (R.E.) Plan Non-Plan Total Plan Non-Plan Total A. Block Grants 67.57 21.39 88.95 72.73 20.39 93.12 B. Social Services 24.34 0.94 24.28 26.52 0.97 27.48 Education, Youth Services, Sports, Art and 4.50 0.87 5.38 3.90 0.86 4.76 Culture Education 4.35 0.73 5.08 3.74 0.73 4.47 Elementary Education 2.47 0.00 2.47 2.29 0.00 2.29 Operation Blackboard 1.69 0.00 1.69 0.98 0.00 0.98 Teacher Training Program 0.39 0.00 0.39 0.62 0.00 0.62 Non-Formal Education 0.40 0.00 0.40 0.68 0.00 0.68 Other 0.00 0.00 0.00 0.00 0.00 0.00 Secondary Education 0.95 0.00 0.95 1.08 0.00 1.08 Vocationalisation of Education 0.64 0.00 0.64 0.75 0.00 0.75 Other 0.31 0.00 0.31 0.33 0.00 0.33 University and Higher Education 0.00 0.60 0.60 0.00 0.60 0.60 Improvement in Salary Scale of Teachers 0.00 0.60 0.60 0.00 0.60 0.60 Adult Education 0.31 0.03 0.34 0.32 0.03 0.35 Special Central Assistance-Border Areas 0.54 0.00 0.54 0.00 0.00 0.00 Other 0.23 0.24 0.47 0.21 0.24 0.45 Youth Services and Sports, Art and Culture 0.15 0.14 0.29 0.16 0.14 0.30 103 State Dept. of Educ. Policy, Planning O Directorate of Directorate Basic SCERT Basic Policy, Ed. Policy, (Autonomous) Planning (B.E) Planning (AL) Assist and Basic Shiksha DIETS (District Parishad Amin. Level) Training Elem. Edue BSP BSP Financial Administration Management District Accounts Officers Officer Basic Shiksha Adhikarl Cam Smands in Flbssedx QualltY Basc Educeson ANNEX 11. Uttar Pradesh A. Percent Literate Above Age 7 (1991) Total 41.71 Male 55.35 Female 26.02 B. Primary (Age 6-11) ereAD 1. Total cohort population (1990) 17127-20 i1 Males 9,016,600 (52.6) Females 8,110,600 Scheduled Castes (SC) 3-624-10 LD Males 1,907,900 (52.6) Females 1,716,200 Scheduled Tribes (ST) 35-90 L=l2 Males 18,900 (52.6) Females 17,000 2. PriMar Enrollments 13-90.000 a Males 8,889,785 (63.4) Females 5,050,215 Scheduled Castes (SC) 2.429-29 L21 Males 1,753,066 72.1) Females 676,244 Scheduled Tribes (ST) _34-209101 Males 21,594 (63.0) Females 12,625 3. Gross Enrollment Ratios (GER) Inta IL32 Male 98.59 Female 62.26 Scheduled Caste A=.0 Male 91.80 Female 39.40 Scheduled Tribe 2 Male 114.20 Female 74.30 4. Number of Schools (1991) 76,734 (GOUP data) 5. Number of Teachers 264,713 (18% female) 6. Enrollments in Primary NFE 1,500,000 (GOUP estimate) C. Upper Primary (Age 11-14r 1. Total cohort population (1990) 9,448.100 1 Male 4,993,900 (52.9) Female 4,454,200 Scheduled Castes(SC) 1.999-20 1.2 Male 1,056,700 (52.8) Female 9,42,500 110 Ffsocbig OraIy Basic Edcedm is hadle Schedulgd Tribes(ST) 19.90 az.2 Male 10. Female 9,400 2. Unocr Primary Enrollments 4470.010 Male 3,240,428 (72.5) Female 1,229,583 Scheduled Castes(SC) 4200 QAI Male 337,523 (80.4) Female 82,523 Scheduled Tribes(ST) fL.2m2 Male 5,353 (76.6) Female 1,634 Gross Enrollment Ratios(GER) Male 64.88 Female 21.15 Scheduld Castes ZLQ Male 31.94 Female 08.75 Scheduled Tribe .L1 Male 50.90 Female 17.38 4. Number of Schools 14,687 5. Number of Teachers 95,562 (24% female) Some: Slecte Educational Statistics. 199O-91. Ministry of Human Rasource Development, Deartmeat of Education, New Delhi, 1992. Estimated Figures. 1 STATE Sa Deptmn LEVEL oEciN R NP FINANCIAL AND EDUCATION TRAINING 4 ADMINISTRATION UP EFA OUALITY MANAGEMENT Board SCERT SIMET MANAEM p* Wianct era con P xecgtim am --- --___LN UHRT Chief~LIN MWS COOtRDINATION IET for Education DItrlcd LevetTWaiVn adW .Coordinaion SEIRT nd BRCsft IMLME*TON*FNCI* LOock Level Training and . Block Resourc Co-ordination wfth DIETS * Centers *SCHO- DISTRICT ADMINISTRATION AND *HOtS LEVEL FNANC MANAGEMENT D.E.rC. OF THE PROJECT District Magistrate FINANCIAL PROJECT IMPLEMENTATION MANAGEMENT Finance I . mProgram AND CO-ORODINTION SbCmitSub-Committee LINE OF AUTHORITY CP I *.*..*.**.*. LINE OF COORDINATION PROJECT ........" " ADVISORY IMPLEMENTATION : FUNCTIONS B3LOCK BPCBlock-Level LEVEL BEAC . ask Group a VILLAGE EDUCATION COMMITTEES Implementation of School Quality improvement, Scholarship Program SCHOOLS Implementation of School Qualty Improvement Source: UP Basic Education Project, 1993. FbMnss Qea&y Bas*c Edcgudms hi 1ads REFERENCES NIEPA: SeminU on Im IMMtion Strategjes of Naional EolijX on ducation, New Delhi, India, 1986 UNESCO: National Studies: India: UNESCO Principal Regional Office for the Asia and the Pacific, Bangkok, 1991 Tan and Mingat: Education in Asia: A Comarative Study of Cost and FinancinL World Bank Regional and Sector Studies, Washington DC, 1993. Government of India: Tle District Prima Education Prorramme: The Government of India, New Delhi, 1993 Staff Appraisal Report: Uttar Pradesh Basic Education Proiect, The World Bank, Washington D.C. 1993. NIEPA: Develonment of Education in India. 1988-1990 Renort to the 42nd Session of the International Conference on Education, New Delhi, Government of India, 1992 Tiak, Jandhyala B.G.: Financing Education in India, NIEPA, New Delhi, India. 1992 Government of Uttar Pradesh: Uttar Pradesh: Education for All. Lucknow, 1992. Rondinelli Dennis A. John R. Nellis and G. Shabbir Cheema: Dcnaization in Developing Countries: A Review of Recent Experience: World Bank Staff Working Paper No 581, Washington DC; World Bank, 1984. Winkler Donald: Decentralization in Education: An Economic Persnective, The World Bank, washington D.C, November 1988. Verspoor, Lockheed and Associates: IThCm Educato in Devloing CnUntries Th Wodd Bank, Washington DC, 1991. World Bank: Gender and Povert, in India. IBRD, Washington DC, 1991. UNDP: Human Develoment Reprt UNDP, New York, 1993 Lockheed and Levin: Effeie Schools in Develonina Couries, Falmer Press, 1993. 113 Case Studies in Financing Quality Basic Education FINANCING BASIC EDUCATION IN BRAZIE Per capita incomes in Brazil are among the highest in Latin America, but the rate of literacy among adults and the rate of enrollment among primary school-aged children are among the lowest in the region. Around the world, countries at similar income levels are almost invariably far ahead of Brazil on the main indices of educational development (see Table 1). The relative backwardness of the Brazilian educational system has been recognized and analyzed for more than six decades, but the problems of the system have proven remarkably durable. Policy initiatives to address problems of access and quality have recurred with depressing frequency, but with few lasting effects. The problems of the system have changed little, and are little closer to being resolved now than they were in 1930, but recent policy initiatives at federal, state, and municipio levels mark a renewed effort to tackle persistent organizational and management issues. In the basic education system the problems include obstacles to access and retention for large numbers of children and low standards of instructional quality in schools throughout Brazil. The primary school enrollment rate has stagnated at approximately 80 percent since the mid-1970s; official enrollment estimates suggest that almost five million children in the ages of compulsory attendance (7 to 14) were not in school in 1989 (MEC, 1990, 34). More than half of all students in 1985 were enrolled in schools that offered four or fewer grades, despite the fact that eight grades of instruction had been compulsory since 1971 (Fletcher and Castro, 1986, 27). With respect to quality, per-pupil expenditures in primary schools average less than two hundred dollars per year, and many schools lack basic equipment including desks, chairs, and toilets, as well as textbooks and other instructional materials (Fletcher and Castro, 1986, 20; Xavier and Marques, 1987; Castro and Fletcher, 1986, 21, 23). In the rural Northeast fewer than half of primary school teachers have themselves completed primary school (Harbison and Hannahek, 1992, 32). Instead of contributing to the solution of these problems and the improvement of basic education, the Brazilian school finance system tends to exacerbates the country's relative educational backwardness (World Bank, 1986; Melchior, 1987; Plank, 1990). Control over discretionary (i.e., non-salary) financial resources remains highly centralized, while responsibility for the administration of schools is almost entirely decentralized to state and local governments. State and municipio governments maintain parallel and sometimes competing networks of primary schools, and many depend on federal transfers for a large share of their revenues. Policies governing federal transfers are ill-defined, and political criteria can often supersede technical considerations in the distribution of revenues (Plank, 1990; Plank, Xavier, and Sobrinho, 1991). Mechanisms to keep track of state and local education expenditures are absent or often ineffective (Costa, 1991; Plank and Verhine, 1993). 114 Aunmde Basic Edcaik e rdJ Reliable data on Brazilian education finance are hard to come by, but the available evidence prompts three observations. First, recent estimates suggest that total public expenditure on education is equivalent to approximately 4.5 percent of GDP. This level of expenditure is comparable to the level in other developing countries, which suggests that the problems facing the educational system are not entirely attributable to scarcity of resources. Second, the share of educational expenditure allocated to basic education varies greatly across levels of the government. At the federal level approximately 70 percent of educational spending goes to the support of higher education, with less than 20 percent allocated to basic education, while states and municipios devote much larger shares of their educational resources to the support of basic education. The shares allocated to basic education by state and municipio governments have declined in recent years, especially as state university systems have expanded. Third, aggregate educational expenditures and expenditures on specific functions fluctuate significantly and sometimes widely from year to year, which renders the planning and investment prerequisite to improvement in the system virtually impossible (World Bank, 1986; Plank, 1990; Sobrinho, 1992; Gomes 1993). The system thus incorporates many of the disadvantages of both centralized and decentralized systems, and few of the advantages of either. To address these problems, the Ministry of Education is currently pursuing a number of strategies aimed at addressing issues of access and quality. Federal transfers to states and municipios have been subjected to closer control through the application of explicit, need-based criteria, and the transfer of resources has been made conditional on the adoption of measures to increase school effectiveness and enhance teacher professionalism. The government has continued to increase the number of Cmntres de Atengie Integral & Crlans (CAICs), which provide an integrated program of educational, nutritional, and health services to poor children. Full implementation of these programs would go some way toward resolving some of the most pressing problems of the basic education system. In this paper basic education finance in Brazil is discussed, with a focus on the financing of educational quality. In the following section some indicators of educational quality that display the relative backwardness of the Brazilian educational system are presented, while in the second section the main features of the financing of basic education are described. The third section discusses options for policy reform in educational finance, including administrative decentralization, diversification of funding sources for basic education along with increased transparency and accountability in the distribution and expenditure of funds. In conclusion, the prospects for educational finance reform and improved quality in the Brazilian education system are briefly discussed. L THE QUALITY OF BASIC EDUCATION The quality of basic education is poor in Brazil. More than half of all primary schools (enrolling approximately 10 percent of primary school students) comprise a single room; many of these one-room schools have no formal premises at all, and are convened in teachers' houses (Fletcher and Castro, 1986, 27). The survival of such schools is extremely precarious. A survey of schools in the rural Northeast found that one-third of all sampled primary schools disappeared between 1981 and 1983; an additional 17 percent disappeared between 1983 and 1985 (arbison and Hanushek, 1992, 39). Many primary schools offer only two or three grades, instead of the nominally compulsory eight, and several grades may be taught by a single teacher. In urban areas primary schools are almost invariably taught in shifts, with most schools providing half a day of instruction to two shifts of students. Some schools in SIo Paulo serve as many as six shifts a day. 115 Coe &MeA in Pbmng Quaiy Badc EdmM Even in well-established schools conditions are far from ideal. In a survey conducted in 1985 fewer than 30 percent of primary schools were found to be in good condition, while an equivalent number were found to be in need of major renovations. Onequarter of all primary schools lacked bathrooms, and half lacked electric light. Nearly 30 percent of all schools had too few seats for the students enrolled, and 25 percent had no desk for the teacher. Most schools in all regions lacked maps, posters, and other instructional materials (Castro and Fletcher, 1986, 20-23). Teachers in primary schools are often poorly-qualified and poorly paid. Nearly half of all teachers in rural primary schools lack the requisite formal qualifications for their positions; approximately three and a half million students attend schools taught by unqualified teachers (Barreto, 1991). In some schools the performance of teachers on standardized tests is hardly better than that of their students (Harbison and Hanushek, 1992, 113). Teachers' salaries are low, and under constant threat from chronically high rates of inflation. In 1987 teachers in the first four grades of Brazilian primary schools were on average paid less than US$150 per month. In the rural Northeast, however, teachers' salaries averaged only US$35 per month, with some paid as little as US$10 (Barreto, 1991; Plank, 1990).' Low salaries contribute to frequent strikes by teachers, which further reduce the already limited quantity of instruction received by primary school students (Costa, 1990). In some regions teachers are hired and fired at the discretion of local politicians, and many or most lose their jobs after each election (Barretto, 1983; Leroy, 1986). The consequences of these problems are seen in data on student attainment and achievement. Rates of grade repetition, overage enrollment, and drop-out are very high: more than 40 percent of all students in primary schools are enrolled in the first two grades, with 25 percent in the first grade alone (IBGE, 1989). Of students entering the first grade in 1981, only 32 percent were promoted to the fourth grade "on time" in 1991. Only 2.5 percent of those who begin primary school complete the course in eight years, and only 40 percent ever finish at all. Approximately 16 percent of children between the ages of fifteen and nineteen are enrolled in secondary schools (MEC, 1990, 42, 46). Standardized achievement tests are rarely administered in Brazil, but one study in a sample of rural schools found that students' scores fell far short of minimum performance standards (Harbison and anusahek, 1992, 84). Problems of instructional quality and access to schooling are closely related. Most significantly, very high rates of grade repetition in the first grade of primary school mean that repeaters occupy school places that might otherwise be available for newly-entaring students. Fletcher and Castro (1986, 15) have argued that reducing the repetition rate could open up enough places in existing primary schools to accommodate all children who are now excluded. A similar argument to this effect has recently been made by Harbison and Hanushek (1992), who conclude that various improvements in instructional quality (e.g., renovation of classrooms, provision of furniture and materials) might actually pay for themselves by increasing the internal efficiency of rural schools. The quality of inputs provided in Brazilian schools varies widely and systematically across regions, income classes, and ethnic groups, and so do educational outcomes. Schools in the relatively prosperous states of the South and Southeast spend more than twice as much per pupil as schools in the Northeast, and pay their teachers more than twice as much. As a result, the employment of underqualified teachers is far more common in the Northeast than elsewhere (see Table 2). Inequalities are observed not only across regions, however, but also between urban and rural areas within regions. Schools in urban areas Teaching jobs provide a variety of bomentboyod salary. including partiction in the social mcutity sysm and soas inflamse over Lhe diribution of chool rmource including food. 116 Fbamcag Base EdcM hs JiU throughout the country are better equipped and better staffed than those in rural areas, and are more likely to be supported by state rather than municipio governments. Urban/rural inequalities are in many respects more powerful than those across regions; for example, "lay" teachers in the urban Northeast are paid more than their rural counterparts in the Southeast (Barreto, 1990, Table 32). Inequalities of race and class are also pervasive in the Brazilian education system. Whites are approximately three times more likely than Afro-Brazilians to complete eight grades of primary schooling, and are less likely to repeat grades or drop out at all levels (Silva and Hasenbalg, 1992). Poor children of all races are less likely than richer children to enroll or to stay in school. In addition, they are more likely to be enrolled in public rather than private schools, and in schools administered by municipio rather than state governments. In contrast, there is little overt discrimination against girls in the Brazilian education system. Females outnumber males at all levels of the system except the earliest grades, and there is no reason to suppose that the quality of instruction varies systematically by gender in primary schools (see Table 3). Discrimination against girls and women in the legal system and in the labor market are widespread and very powerful (Lovell, 1993), but it is not immediately obvious whether or how gender inequalities are perpetuated through the basic education system. The several dimensions of inequality in the Brazilian education system tend to be mutually reinforcing. For example, Afro-Brazilians are more likely than whites to live in the Northeast, to live in rural areas, and to be poor. Similarly, the population of the Northeast is disproportionately poor, rural, and black when compared with the populations of other regions. The complex interplay of race, class, and region poses important challenges for the design of public policies aimed at reducing inequalities. U. EDUCATIONAL FINANCE IN BRAZL The basic elements of the Brazilian school finance system comprise a set of Constitutional rules to govern the expenditure of general revenues, a number of revenue sources "earmarked" for educational purposes, and a variety of institutions created to accomplish specific objectives. In political rhetoric and policy statements priority in the allocation of public resources has generally been assigned to basic education, but the distribution of resources has not always reflected formal policy priorities. Since 1934 Brazilian Constitutions have included provisions fixing the percentage of their revenues that federal, state, and municipio governments must spend on education. Under the 1988 Constitution, for example, state and municipio governments are required to spend 25 percent of their revenues from taxes and transfers on education, while the federal government must spend 18 percent (Senado Federal, 1988, Article 212). In addition, the so-called Transitional Provisions of the 1988 Constitution require all levels of the Brazilian government to direct at least half of their educational expenditures to basic education in the decade following the adoption of the Constitution (i.e., 1988-1998), in order to eliminate illiteracy and ensure universal access to primary education (Senado Federal, 1988, "Ato das Disposi;,es Constitucionais Transitdrias," Article 60). These Constitutional rules have not always been fully effective in ensuring minimum levels of educational expenditure. The question whether the federal government is in compliance with the requirement that a designated percentage of federal revenues be spent on education has aroused considerable controversy 117 cae aNdfes 6. PNmig auNry Befdc E..m over the years, including a Congressional investigation in 1988 (Melchior, 1987, 102-124). It is widely assumed that state and municipio governments spend at least as much on education as they are required to do, but in the absence of effective mechanisms for reviewing public expenditure or sanctions for non- compliance it is hard to know whether this is true (Gomes, 1992a, Costa, 1991; Plank and Verhine, 1993). There is not much evidence that the Constitutional requirement that all public authorities devote 50 percent of their educational expenditure to basic education has brought about changes in existing patterns of expenditure at any level of government. The main earmarked revenue source for basic education is the saltrlo-educago, a federal wage tax of 2.5 percent that is imposed on the total wage bill of most employers (World Bank, 1986; Melchior, 1987; Plank, 1990). Specified shares of revenue from federal lotteries and oil and gas royalties have also been earmarked for education; tax exemptions and fiscal incentives are used to generate additional support for educational activities including private education and literacy training. Revenue from the saidrio-educagio must be spent on primary education, with two thirds returned to the state where the tax was collected and one third retained by the federal government for distribution among states and municipios according to criteria that are in principle equalizing. Firms may withhold their tax payments in order to maintain their own schools or to reimburse the educational expenditures of employees or their dependents. They may in addition direct their payments to private schools in order to provide tuition scholarships for the children of their employees (FNDE 1990). The administration of the saiddo-educaoo has long been controversial, for two main reasons. The first is that a disproportionate share of the revenues collected under the saidrlo-educagio have gone to support private rather than public schools. Since 1985 firms have directed more than half of their aldArio- educagio payments to private schools, to pay the tuition of their employees and dependents (FNDE, 1990; Plank, Xavier and Sobrinho, 1991).' The second is that saidro-educagio revenues are readily susceptible to political manipulation: revenues from the federal share of the saldrio-educagio are often transferred to state and municipio governments according to political rather than technical criteria. In 1989, for example, two-thirds of the revenues collected under the safido-educa;io were distributed off- budget, for "special projects" selected at the discretion of those in charge of the Ministry of Education. The consequences are displayed in Table 4, which shows how transfers to particular states tended to rise when a politician from that state was appointed Minister of Education (see Table 4). Other earmarked revenue sources are subject to similar problems (Melchior, 1987; Plank, 1990). A third feature of the Brazilian school finance system is the proliferation of autonomous or semi- autonomous institutions to fulfill specific educational purposes. To address the problem of adult illiteracy, for example, the military government organized the Movimento Brasiluro de Alfabethaio (MOBRAL), a literacy campaign which was expected to reduce the number of illiterates by half between 1969 and 1972 (Passarinho, 1970). Special fiscal incentives were established, which gave MOBRAL a resource base independent from the regular education budget. The literacy campaign failed to achieve its goal; instead, MOBRAL quickly evolved into a parallel education system offering a variety of 2 Progms mder which revenues my be drected to privefe schools hve periodically resulted in fraud. inthe ady 1930s firms we permmned to diseact iir saldhno-dc0opaymnts to privers schools to pay tuition oe only for eaployee and their childre but for other adultsad childre (abiesida co....a*)as wUll. The resuk was the diversion of large amounts of revenue from abs goveamet as the suport of leaginay Madas and even umagiary schools (Vellose 1937; Lea, 1990). The clietelw of the program was once again e 'mid mo amloyees and their childre in 1933. I18 Fbwaicag Beadc Educales di Brad programs including vocational training, health education, and pre-schools (Cunha da Costa, 1986). MOBRAL was brought under the control of the Minister of Education in 1981, and closed down in 1985. Subsequent governments have established similarly autonomous and independently-funded institutions (Funda;io Educar, Programa Nacional de Alfabetiza;o e Cidadania) to take its place, however. Another important and semi-autonomous institution in the basic education system is the Funda;ho de Assist&ncia ao Estudante (FAE). The FAE is responsible for the distribution of textbooks and instructional materials to primary schools throughout Brazil, and for the administration of school feeding programs. Like MOBRAL and its successors, the FAE's resources are obtained independently from those of the Ministry of Education; in 1987 three-quarters of the budget was funded by the Fundo de In7estimento Social, which is under the direct control of the President's office (Melchior, 1987; FAE, 1987). Also like MOBRAL, the FAE maintains an administrative structure which runs parallel to that of the Ministry of Education, and it has expanded into a variety of areas (health education, teacher training, distribution of scholarships) that are at best peripheral to its original mission. A third set of semi-autonomous institutions in the basic education system is the network of educational and social service centers for poor children (CAICs) inaugurated under President Color (as CIACs) and expanded under President Franco. The premises on which this initiative is based are first, that the multiple disadvantages (educational, nutritional, medical, social) suffered by poor children must be addressed simultaneously; and second, that the educational deficit that holds poor children back cannot be overcome in the three or four hours of the typical Brazilian school day. The CAICs therefore provide three meals and a variety of social and medical services in addition to a full day of instruction. Like MOBRAL and the FAE, the CAICs represent yet another effort to address the problems of basic education by establishing a new set of institutions rather than improving the performance of the existing education system. The establishment of autonomous institutions to pursue specific educational objectives has led to a variety of problems. The first is the diversion of large quantities of resources into the maintenance of parallel administrative structures; these resources might better be devoted to the achievement of substantive objectives. Alongside this is the burden that administrative duplication places on educational officials at state and municipio levels, who are obliged to seek funds for related educational programs from a number of different agencies. A third problem is the absence of systematic oversight over autonomous institutions, which has resulted in frequent scandals in the FAE. Finally, as with the saldro-educago, the establishment of multiple channels for the distribution of resources to state and municipio governments lends itself to exploitation by politicians seeking to build political support at these levels. There are a variety of reasons why the Brazilian school finance system does not ensure access to schools of even minimal quality for so many children, of which three are especially important. First, despite the incorporation in successive Brazilian constitutions of requirements that specific shares of public revenue be allocated to educational uses, the education finance system as a whole lacks effective rules to govern the distribution of resources. Moreover, such rules as are defined-including Constitutional rules-are frequently evaded or ignored, and there are few sanctions for failure to comply. Access to educational resources thus remains contingent on a host of factors that typically includes region, wealth, and political affiliation, but often excludes need or effort (Plank, 1990; Plank and Verhine, 1993). A second reason for the failure of the school finance system to provide enough school places of acceptable quality is that the system is inordinately complex. The problems posed by the lack of "transparent" decision rules for the distribution of funds and the proliferation of funding sources are 119 CNAlse Ina bnckg Qnsbyt Badc Edwam. compounded by the fact that most funding sources require state and local governments to apply for funds in support of specific projects. As a result, access to information and political influence are of tremendous value in the competition for resources. Impoverished rural mimcipios far from Braflia face obstacles in learning about the availability of funds from particular sources, and experience difficulties in complying with complex administrative and accounting procedures. Such municipios may consequently not gain access to financial support to which they are ostensibly entitled. In Bahia in 1988, for example, nearly 25 percent of the municipios in the state received no federal transfers, though all were entitled to them (see Table 5). To avert this possibility, many municipios employ specialized intermediaries to facilitate their access to federal transfers (Goies, 1992b). The third reason for the failure of the school finance system to guarantee the right to basic education is that the system is open to political manipulation at several levels. Ministers of Education steer resources to their home states, and members of the National Congress distribute scholarships and "special projects" to their electoral redoubts (see Tables 4 and 5). At the local level funds are used to reward clients and prospective supporters with administrative and teaching jobs, and with contracts for the provision of construction and other services (Barresto, 1933; Leroy, 1986; Leal, 1990; Mello e Souza, 1989). These interventions in the administration and financing of the education system undermine the quality of schooling provided, while infisting its cost. The wide array of funding sources and the absence of rules to govern resource distribution in the Brazilian education finance system combine to diffuse responsibility, maximize administrative discretion, and minimiz public accountability. The results can be am in the poor quality of many Brazilian schools, and in the relatively low levels of educational attainment among Brazilian students. Il. POuCY OFTIONS IN EDUCATIONAL FINANCE The Brazilian school finance system illustrates how educational finance is embedded in local and national politics. The way the system works is determined to a substantial extent by the uses to which it is put by political authorities. The changes that administrative and policy reforms by themselves can be expected to bring about in the system are therefore limited; solutions to the many problems of the Brazilian educational system will necessarily depend on broader changes in the political system, including a sustained commitment to Education for All and measures to give a more effective voice to those who are now disenfranchised in local and national politics. There are nevertheless a variety of policy reforms that have been adopted or proposed in Brazil that represent significant first steps toward increased access and improved quality in the basic education system. Four of thes-administative decentralization, diversification of revenue sources, targeting of resources to marginalized populations, and increased transparency and accountability-sre discussed in the following sections. Adminitadive Decenatliaion As a practical matter, administrative responsibility for basic education is very highly decentralized in Brazil. State and municipio government administer separate and independent "systems" of primary schools, with minimal coordination among systems and little effective oversight from central authorities. At the same time, the federal government retains control over the largest share of discretionary financial resources for basic education, and most states and unmicipios rely on federal transfers to fund much of their non-salary educational expenditure (World Bank, 1986; Melchior, 1987; Plank, 1990). 120 Fnibnng Basc Educadm in Bri The new fiscal federalism introduced by the Constitution of 1988 marks an important step toward a closer correspondence between the distribution of administrative responsibility and the assignment of control over tax bases, as does the continuing effort by the FAE to shift responsibility for school feeding programs to the local level (FAE, 1987; Gomes, 1988; Shah, 1991). Reducing the dependence of state and municipio governments on federal transfers could result in a net gain in the quantity of resources available for basic education, if it limited opportunities for political intermediation and increased the efficiency of revenue collection and expenditure. Assigning control over significant tax bases to state and municipio authorities might also serve to ensure the availability of resources to all local governments, and so increase their accountability to their constituents. Beyond this, allowing greater discretion to local authorities in the determination of educational expenditures would make it possible to adapt schedules, curricula, and even menus in ways that might increase the propensity of local children to enroll and stay in school or improve the quality of instruction provided in local classrooms. Policies that shift additional responsibility to administrators and councils at the school level have been adopted in several states, and the pace of innovation appears to be accelerating (Plank, Xavier, and Sobrinbo, 1991). Decentralization is no panacea for the problems of the Brazilian education system, however. Indeed, decentralization will make many educational problems worse if the central government does not simultaneously expand and strengthen its role in monitoring local taxation and expenditures, equalizing the distribution of resources across and within regions, and protecting the educational interests of otherwise disadvantaged children. With respect to the first point, for example, decentralizing control over tax bases will bring about no improvements in local school systems if local tax bases are too small to support minimal quality standards, or if state and municipio governments fail to tax their constituents. The latter has historically been a problem in the rural Northeast, where municipio governments have been reluctant to impose taxes on property, preferring to rely on federal transfers as a revenue source (Shah, 1991). With respect to the second point, fiscal decentralization can increase inequalities in the availability of resources across states and municipios, both because of differences in wealth ad income and because of variation in tax effort (Plank, Sobrinho, and Xavier, 1991). Finally, in the absence of central government action increasing the resources available to state and municipio governments may do little to improve the educational opportunities available to poor children, whose access to schools of acceptable quality depends on local decisions about the allocation of resources. A policy reform even more urgent than administrative decentralization is the unification and rationalization of what are now parallel and competing basic education systems maintained by state and municipio governments (Mello e Souza, 1989). Apart from the waste of scarce resources entailed by the duplication of services, the existence of two systems encourages the evasion of responsibility by both state and municipio governments, neither of which can at present be held accountable for problems of access and quality in the basic educational opportunities available to Brazilian children. The unification of competing school systems might be accompanied by the integration of autonomous institutions providing ancillary educational services (e.g., the FAE) into the local education system, to further reduce the number of agencies with administrative responsibility for basic education. Dierniaion of Reeae Seawv Efforts to diversify the sources of funds available for basic education have been rare in Brazil, apart from the *earmarking" of new tax bases (e.g., salrlo-educaoo, lotteries) for educational uses. The 121 Cam S.diw in Famdn Owsuethy Bac Ed.c.m participation of parents, communities, and business organizations in public school finance remains minimal, as virtually all educational services (including textbooks, materials, and lunches) are either provided directly at public expense or are heavily subsidized through the tax system. A growing number of states and municipios have encouraged self-help maintenance and repair projects (mutir5es) in local schools, but the practice is not yet common (Castor, 1987). The presumption that public education is the exclusive responsibility of *the State" remains widespread. The relatively large share of children enrolled in private schools marks an important exception to the overwhelming predominance of the public sector in educational finance (see Table 6). Private schooling perforce shifts a significant financial burden from public authorities to households and other private agencies, but the magnitude of this shift is greatly reduced in Brazil through public subsidies and tax exemptions. In the 1980s, for example, growth in the share of primary school students enrolled in private schools came about mainly in response to the provision of public subsidies to poor schools and parents, and to the regulation of the fees that elite schools were permitted to charge their clients (Plank, Xavier, and Sobrinho, 1991). There is little evidence to suggest that growth in private school enrollments has brought about a net increase in the supply of primary school places available to poor children, though it may well have produced some financial savings for the public sector. Targeting Resourc to Marginahred Populatfar A third strategy for improving the quality and coverage of the basic education system is to target a greater than proportional share of resources to constituencies who would otherwise be marginalized by the system. In Brazil marginalized groups include Afro-Brazilians, rural people, residents of the Northeast, and the urban poor. As discussed earlier, the categories of disadvantage are to some extent overlapping and mutually reinforcing: Afro-Brazilians are on average poorer than whites, and more likely to live in rural areas or in the Northeast; residents of the Northeast are disproportionately black, poor, and rural as compared with other Brazilians. In Brazil efforts to target educational resources have focused primarily on regional rather than racial or income inequalities. Policies pursued by the federal government over several decades have transferred revenues from other regions to the Northeast, in education as in other sectors. More recently, the World Bank and other international agencies have directed a disproportionate share of their loans to projects in the Northeast, including EDURURAL and the so-called *Nordest" (Harbison and Hanushek, 1992). Underlying these efforts is the assumption that accelerating the development of the Northeast will in turn increase the prosperity of marginalized groups within the region. In fact, however, several decades of experience with a strategy focused on the reduction of regional inequalities have produced meager benefits for disadvantaged populations in Brazil. The income distribution has grown more unequal since 1960, and differences between Afro-Brazilians and other groups in income and status remain wide (Hoffman, 1989; Silva and Hasenbalg, 1992). The Northeast continues to lag far behind other regions on virtually all indices of social and economic development (IBGE, 1989). More direct efforts to reduce the effects of income inequality in the educational system, including the public supply of textbooks and school lunches and the construction of CAICs, have had positive but still limited effects on the educational opportunities provided to poor children. In conam, increasd private provision at post-secondary levels ba greatly imcreasd soom for young people who would otherwise be excluded from further education. 122 FLWcg Base Edcades dn Brd The success of a strategy that targets resources to the disadvantaged depends on political commitment to benefiting otherwise marginalized groups. Recent developments in Brazil offer basis for hope that such commitment is stronger now than it has been in the past. Policy debates have begun to turn away from a preoccupation with the control of resources to a focus on the quality of instruction provided in schools. State governments in states including Ceari and Minas Gerais have adopted reforms that promise dramatic changes in the ways in which resources are distributed and schools are administered. Municipio governments in all parts of Brazil have begun to engage in administrative reform and policy experimentation on an unprecedented scale. Ile Ministry of Education has devoted large quantities of resources to the construction of CAICs, and has taken steps to ensure that resources are distributed so as to reach the schools and classrooms where they are most needed. Together these developments mark a promising shift in the political system in favor of policies that could in time provide quality basic education for all Brazilian children. Traspaecy and Accountabity A final strategy for increasing access and improving quality in basic education is the introduction of mechanisms to foster transparency and accountability in the educational finance system. As long as the local availability of educational resources depends significantly upon political connections and administrative sophistication, the final assignment of responsibility for the provision of basic educational services remains problematic. Similarly, in the absence of reliable information on transfers, receipts, and expenditures it is difficult to establish accountability for the performance of the basic education system, or to hold public officials responsible for the institutions under their authority. Only when the availability of funds for basic education is assured and public access to information on educational expenditures is guaranteed will the right to basic education for all acquire meaning in Brazil. The simplest and most powerful step toward establishing accountability in the educational system would entail the institution and enforcement of binding decision rules for the distribution of federal transfers, in order to ensure that all states and municipios have sufficient resources to provide schools of acceptable quality for all local children (Plank, Sobrinho and Xavier, 1991). Some steps in this direction have already been taken; the distribution of saldrio-educaoo revenues has in principle been determined by a "funding formula" since at least 1981 (Gomes, 1992b; SEB/MEC, 1988; MEC, 1993). The potential effectiveness of the funding formula has often been vitiated in practice, but the current government has demonstrated a renewed commitment to allocating resources according to technical criteria including need. A second reform closely related to the first would require the timely publication of data on receipts and expenditures in state and manicipio education systems, and on federal transfers, in order to facilitate public monitoring of the flow of educational resources and increase accountability in the educational system (Plank and Verhine, 1993). For the present, however, the respective responsibilities of federal, state, and local authorities in the educational system remain undefined, and the extent to which these responsibilities have been fulfilled remains difficult to evaluate. Recent policy shifts at all levels indicate an increased commitment to improving the quality of basic education and ensuring access for all children. The institution of mechanisms for the monitoring and evaluation of public expenditures would help to ensure that this commitment is realized. 123 Can &udies in PInescing aurly Basc EducdM Coadusion In a recent evaluation of a project that sought to raise the quality of primary schools in the rural Northeast, Harbison and Hanushek (1992) have argued that there is no necessary tradeoff between spending to improve quality and spending to expand access in the basic education system. Investments in the quality of schooling may be self-financing, and may even free resources to finance new school places, by reducing rates of grade repetition and drop-out and opening up places for additional children in primary schools. The value of these benefits may be quite large in some parts of Brazil, but capturing them will require political and institutional changes for which a constituency has only begun to emerge. A shift in the structure of interests and incentives in the broader political system is therefore prerequisite to lasting improvements in the education system. Increasing the accountability of local officials by decentralizing control over tax bases and encouraging transparency in the distribution and use of educational resources can contribute to this shift, but cannot take the place of political change. Expanded participation in educational governance at both national and local levels and the enhancement of citizens' rights against those of "the States" finally represent the surest guarantees of expanded access and improved quality in basic education. There is encouraging evidence of a shift in this direction within the Brazilian educational system. The Constitution of 1988 made the notional "right" to education actionable (un direlto sujetivo) for the first time, which opens the way for citizens to demand that their "rights' be honored by public authorities (Senado Federal, 1988, Art. 208). Schools in a growing number of states are at least partly governed by school councils comprising parents, teachers, and students; in some states school principals are elected by these same constituencies, instead of being appointed by local politicians (Plank, Xavier, and Sobrinho, 1991). The present challenge is to ensure that political commitment to these and related reforms is sustained in order to bring about lasting improvements in the educational system and guarantee basic educational opportunities for all Brazilian children. 124 FaOVng Badc Edaens Bradw Table 1. Development Statistics, Brazil and Comparable Counries, 1990 Per Capita Life Adult Net Primary GNP Expectancy Literacy Enrollnent (USS) (years) (percent) (percent) Brazil 2,680 66 81 84 Argentina 2,370 71 95 96 £ Mexico 2,490 70 87 100 Chile 1,940 72 93 89 Peru 1,160 63 85 95 Bolivia 630 60 77 83 Latin America 2,180 68 84 87 Poland 1,690 71 - 97 Korea 5,400 71 - 100 Note: * 1987 Some: Would Bak, World Deeooment Rct 1992, pp. 218-19,274-75. Table 2. Comparative Education Indicarors, Northeast and Brazil, 1988 Northeast Brazil Persons 5 + (in percent) Literate 54.6 74.2 Persons 10+ (in percent) No Instruction 36.6 19.7 1 - 4 Years 37.7 41.3 8 Years or more 14.1 23.0 12 Years or more 2.4 5.2 Enrollments Municipio schools 44.7 29.4 State schools 40.4 56.9 Per pupil expenditure (in USS) Municipio schools 28.6 52.2 State schools 67.3 146.3 Teachers' salaries (in USS) Urban 124.0 207.3 Rural 35.0 76.0 "Lay" teachers (in percent) Urban 4.6 2.7 Rural 66.0 48.4 Note: * Includes persons with less tham one year of in&raion Sore: Brazil. AnaMoEaraNsco, 1989. pp. 19199; MEC (1990), p. 39; Xavier ad Marques, 1987, Table 2; Barreto, 1990, Table 32; Barreto, 1991, Tale 1. 125 Car SIaNAs A flm:Mang Osiv Basic Edamels Table 3. Gender Distriburion of Student, 1987 (percent) Male Female Pre-School 50.5 49.5 Primary (1-8) 49.1 50.9 Secondary 41.5 58.5 Post-Secondary 47.4 52.6 Source: Brazil, Anaudo Esadsico. 1989. p. 200. Table 4. Shares of the Quota Federal Received by Selected States, 1985-1988 (percent) 1985 1986 1987 19W Population Pernambuco (Marco Maciel, 198546) 5.7 6.6 4.1 3.5 5.2 Santa Catarina (Jorge Bornhausen, 1986-87) 2.1 3.8 5.0 3.0 3.1 Piauf (Hugo Napoleao, 1987-89) 5.4 6.1 5.3 5.9 1.8 Bahia (Carlos Sant'Anna, 1989) 6.1 9.7 6.5 7.9 7.9 Maranhlo (President Jose Sarney) 5.7 7.9 6.5 9.2 3.4 Noe: Excluding Distrito FederWal. Source: FNDE. Saldrio-Educacio: Sdries Hist6ricas, pp. 4144. 126 Fbncing Basic Edu:aton in Brail Table 5. DLstibution of Revenues from Salarlo-Eucacao in Bahia, 1988 Number of municipios 367 Number receiving "standard" SE funds 230 Number receiving "special projectsm 160 Number receiving both 104 Number receiving nothing 89 Total value of transfers (in millions of cruzados) 2,302 Value of "standard" transfers 773 Value of "special projects" 1,474 Mean per capita value of transfers (in cruzados) 198 Maximum per capita value of transfers 3,521 Minimum per capita value of transfers 0 Number affiliated with governing parties (PDS/PFL) 304 Number affiliated with other parties (PMDB/FTB/other) 63 Percent PDS/PFL receiving "standard" funds .64 Percent PMDB/FTB receiving "standard" funds .52 Percent PDS/PFL receiving "special projects" .48 Percent PMDB/PTB receiving "special projects" .24 Percent PDS/PFL receiving both .31 Percent PMDB/PTB receiving both .16 Percent PDS/PFL receiving nothing .22 Percent PMDB/FTB receiving nothing .30 Source: Ddecis do MEC, Salvador. Upulisd data. Table 6. Enrolments in Prhe Schools, Brazil, 196L-1987 (percent) PrimWy Secondary Tetiary 1960 11.5 65.0 44.3 1970 9.0 40.1 50.5 1980 12.9 46.5 63.3 1985 12.1 33.3 59.2 1987 13.2 35.1 60.2 1989 14.5' 34.6 61.5 Nma: *Emass Source: Brazil, Anos Eastis , various yea; MEC 1990; and Levy 1986. Table 5.1. 127 Case a.~ 9M aeca Q.ary BPkad Educ~den REFERENCES Barreto, Angela Maria Rabelo Ferreira (1990). A Geografia do Professor Leigo: Situaçao Atual e Perspecrivs. Brasília: IPEA. Barreto, Angela Maria Rabelo Ferreira (1991). Professores do Primeiro Grau: Quem São, Onde Esao, e Quanto Ganha~. Brasília: IPEA. Barretto, Elba Siqueira de Sã (1983). "Novas Políticas Educacionais para Velhas Escolas Rurais: Um Estudo de Caso no Sertão do Piauí." Cadernos de Pesquisa 46: Agosto. Castor, Belmiro Valverde Jobim (1987). "Fortalecimento das Instâncias Locais de Planejamento, Financiamento, e Administração Educacional: AçIes da Secretaria da Educação do Paraná." Paper presented to the Seminario Sobre Planejamento, Financiamento, e Descentralização da Educação no Brasil, Petropolis RI, October. Castro, Cláudio de Moura and Philip Fletcher (1986). A Escola que os Brasileiros Freqfentaram em 1985. Brasilia: IPEA. Costa, Messias (1990). O Rendimento Escolar no Brasi e a Experincia de Outros Pases. São Paulo: Ediçes Loyola. Costa, Thais Helena (1991). "O Financiamento da Escola Básica: O Papel do Tribunal de Contas." In O Desafio da Escola Básica: Qualidade e Eqidade, ed. Ligia Cademartori. Brasília: IPEA. Cunha da Costa, Roberto Mario (1986). MOBRAL Nascimento, Vida, PaLxao, Morre. Rio de Janeiro: Shogun Editora e Arte. Fletcher, Philip and Cláudio de Moura Castro (1986). "Os Alunos e as Escolas no Brasil de Hoje." Paper presented to a conference on Social Change in Brazil, University of Wisconsin - Madison, May. Fundação de Assistência ao Estudante (1987). Relatrio 1987. Braslia: FAE. Fundo Nacional de Desenvolvimento da Educação (1990). Sadrio-Eduan: Séries Histórcas. Brasília: FNDE. Gomes, Candido Alberto (1988). "Os Quatro Anos da Emenda Calmon: Qual o seu Impacto?" Revissa Brasileira de Estudos Pedagogicos 162: AbrillJunha. Gomes, Candido Alberto (1992a). "Financiamento e Eqidade." In Qualdade, Eficiéncia e Eq9idade na Educaçao Bdsica, ed. Candido Alberto Gomes and José Amaral Sobrinho. Brasilia: IPEA. Gomes, Candido Alberto (1992b). "Transferéncias Intergovernamentai no Ensino Fundamental: Critérios e Processos Decisdrios." Unpublished manuscript. Gomes, Candido Alberto (1993). "Despesa Pública no Brasil: Educação é Prioridade?" Unpublished manuscript. 128 Fusce Basc Edcedes bs BAard Harbison, Ralph W. and Eric A. Hanushek (1992). Educational Performance of the Poor. Lessons from Rural Norheast Brazil. New York: Oxford University Press. Hoffman, Helga (1989). "Poverty and Prosperity in Brazil: What Is Changing?" in Social Oiange in Brazil. 1945-1985: The Incomplete Transition, ed. Edmar L. Bacha and Herbert S. Klein. Albuquerque: University of New Mexico Press. Instituto Brasileiro de Geogrifica e Estatistica (1989). Anudrio Estartstico do Brasil, 1989. Rio de Janeiro: IBGE. Leal, Maria Cristina (1990). "Os Des(mandos) do Clientelismo de Estado sobre os Recursos Pdblicos da Educa;so de 1' e 2* Graus." Doctoral dissertation, Universidade Federal do Rio de Janeiro. Leroy, No&ia Maria Inez Pereira (1986). 0 Gatopardismo na Educardo: Reformarpara ndo Mudar (0 Caso de Minas Gerais). Rio de Janeiro: Dois Pontos Editora. Lovell, Peggy A. (1993). "Race, Gonder, and Development in Brazil." Paper presented to the conference on Racial Politics in Contemporary Brazil, Austin, April. Melchior, Josd Carlos de Aradjo (1987). 0 Financiamento da Educan no Brasil. Slo Paulo: Editora Pedag6gica e Universitiria. Mello e Souza, Alberto (1989). "Consideraoes sobre a Distribui;io dos Recursos Educacionais." Em Aberto 42: Abril/Junho. MinistWrio da Educaio (1990). A Educado no Brasil na Dicada de 80. Brasilia: MEC. Minist6rio da Educalo e do Desporto (1993). Stzemdicapara Financamento de Projetos na Ara de Educacd Bdsica. Brasilia: MEC. Passarinho, Jarbas Gonalves (1970). "A Problemica da Educao no Brasil." Separata da revista A Defesa Nacional (Mar/oAbril). Plank, David N. (1990). "Issues in Brazilian School Finance." Journal of Education Fance 15: Spring. Plank, David N., Antonio Carlos da Ressurrei;Ao Xavier and Jos6 Amaral Sobrinho (1991). "Why Brazil Lags Behind in Educational Development." Paper presented to the conference on "Education, Growth, and Inequality in Brazil," Rio de Janeiro, March. Plank, David N., Josd Amaral Sobrinho and Antonio Carlos da Ressurreio Xavier (1991). "Uma Estrat6gia Para a Qualidade do Ensino Fundamental: Os Padres Mfnimos de Oportunidades Educacionais." Estuds em AvoiaWpo Educacional 1: Julbo/Dezembro. Plank, David N and Robert E. Verhine (1993). "Determinants of Educational Expenditure across Municipios in Bahia, Brazil." Paper presented to the annal meeting of the Comparative and International Education Society, Kingston, Jamaica, March. 129 Caae aSues, in ~~nda Qua~y Raa E&aw Secretaria de Ensino Básico/Ministério da Educaço (1988). "Critérios de Distribuiçio de Recursos." Paper presented to the Seminírio sobre Modelos de Alocação de Recursos para o Ensino Pdblico, CENDEC/PEA, Braslia DF, April. Senado Federal (1988). Constituiçao da República Federativa do Brasil. Shah, Anwar (1991). 7he New Fiscal Federalism in Brazil. Washington: IBRD. Silva, Nelson do Valle and Carlos A. Hasenbalg (1992). Rela~4es Raciais no Brasil Conze rmporneo. Rio de Janeiro: Rio Fundo Editora/IUPERJ. Sobrinho, José Amaral (1992). 'O MEC e o Ensino Fundamental: O que os gastos revelam,' in Esado e Educaçao, ed. Jacques Velloso, Guiomar Namo de Mello, Lilian Wachovicz, e outros. Campinas: Papirus/CEDES; S *o Paulo: ANDE/ANPED. Velloso, Jacques. 1987. "Política Educacional e Recursos para o Ensino: O Salãrio-Educaçao e a Universidade Federal." Cadernos de Perquisa 61: Maio. World Bank (1986). Brail: Finance of Prmy Educarion. Washington: IBRD. Xavier, Antonio Carlos da Ressurreição and Antonio Emílio Sendim Marques (1987). Quanto ~ata wn Aluno nas Escolas que os Brasileiros Frequentan. Brasília: PEA. 130